(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Proceedings of the annual convention"

V/ . C7 . K^i, OL/ / rrt. ^ y ^ r 




m 




American Railway Engineering Association 

May 1997 

Volume 98, Bulletin 760 



BOARD OF DIRECTION 

1997-1998 

President 

Mr. p. R. Ogden, Norfolk Southern, Vice President-Engineering, 99 Spring Street, 
Room 801 , Atlanta, GA 30303 

Vice Presidents 

Mr. D. C. Kelly, Illinois Central Railroad, Vice President-Maintenance, 17641 S. 
Ashland, Homewood, IL 60430 

Mr. Rick Richardson, Jr., Canadian National Railways, Chief Engineer, 935 
de la Gauchetiere Street, West Montreal, PQ-H3B 2M9 

Past Presidents 

Mr. B. G. Willbrant, Amtrak, Assistant Chief Engineer, 21 Berkshire Drive, Wayne, 
PA 19087 

Mr. J. R. Beran, Union Pacific Railroad, Chief Engineer-Maintenance of Way 
Structures, 1416 Dodge Street, Room 1000, Omaha, NE 68179 

Treasurer 

Mr. W. B. Dwinnell, III, SEPTA, Railroad Division, Chief Line Maintenance Officer, 
1234 Market Street, Philadelphia, PA 19107 

Directors 

Mr. J. R. Clark, Jr., CSX, Assistant Chief Engineer, 500 Water Street, Jacksonville, 
PL 32202 

Mr. E. p. Reilly, Union Pacific Railroad, Chief Engineer-MAV Central, 1 860 Lincoln 
Street, 14th Floor, Denver, CO 80295 

Mr. L. Anderson, Illinois Central, Superintendent Engineering, P.O. Box 2600, 
Jackson, MS 39103 

Inc. Lorenzo Reyes R., FNM, Director Ferrocarril Sureste, Estacion Terminal, 
Montesinos S/N Colonia Altos, Veracruz, Mexico 91700 

Mr. W. C. Thompson, Union Pacific Railroad, Director Engineering Research, 1416 
Dodge Street, MC 3300, Omaha, NE 68179 

Ms. C. D. Wylder, MARTA, Executive VP Operations and Development, 2424 
Piedmont Road, N.E., Atlanta, GA 30324-3330 

Mr. R. L. Keller, Montana Rail Link, Chief Engineer, PO. Box 8779, 210 
International Way, Missoula, MT 59807 

Mr. John Cunningham, Amtrak, Assistant Chief Engineer-Track, 3rd Floor, South 
Tower 30th Street Station, Philadelphia, PA 19104 

Mr. Michael Roney, CP Rail System, General Manager-Engineering Services and 
Systems, Suite 500, Gulf Canada Square, 401 -9th Avenue, S.E., Calgary, Alberta T2P 
4Z4 Canada 

Mr. Walter Heide, Conrail, Assistant Chief Engineer-MAV, 2001 Market Street, Room 
10-B, Philadelphia, PA 19101-1410 

Mr. Gary Woods, Norfolk Southern, AVP-Maintenance of Way Structures, 99 Spring 
Street, Box 142, Atlanta, GA 30303 

Mr. Michael Armstrong, BNSF, AVP-Maintenance Planning, 2600 Lou Menk Drive, 
Ft. Worth, TX 76131-2830 

Executive Director 

David E. Staplin 

50 F Street, N.W., Washington, DC 20001 

(202) 639-2190 



American Railway 
Engineering Association 

BULLETIN 
No. 760 

MAY 1997 

Proceedings Volume 98 (1997) 

D. E. Staplin, Editor 

CONTENTS 

Cover Story: Reconstruction of Saco River Rail/Highway Bridge 

Crossing: A Partnering Project in Design, Construction 

and Finance 1 

Proposed 1997 AREA l\/lanual Revisions 

(Chapters 5, 8, 16, 18 and the AAR Scale Handbook) 9 

Presentations at the 1997 Annual AREA Technical Conference — March 1997 

1. Norfolk Southern Trackside Lubrication Studies — 1997 Interim Report. . . 67 

2. The Philosophy and Development of AREA Seismic Design Criteria .... 77 

3. Fracture Toughness Testing of AREA Grade B Hand Tool Steel 83 

4. The Potential of the CFS MOW Estimating Relationships as a Basis 

for Cost Allocation 91 

5. Longitudinal Forces in an Open-Deck Steel Deck Plate-Girder Bridge . . 101 

6. Improving Infrastructure Reliability Through Engineering Facility 
Management System Standardization 1 07 

7. Managing Rail Resources 139 

8. On the Benefits of Rail Maintenance Grinding 149 

9. Service Level Live Load Stress Ranges on Hangers and Floor Beams 

of Steel Railway Bridges 169 

Technical Papers: 

1. Development of a Recycled Plastic/Composite Crosstie 181 

Memoirs: 

Eldon E. Farris 189 

Edward Q. Johnson 190 

Index of Advertisers 201 

Index of Consulting Engineers 201 

Front Cover: Reconstructing the Saco River Rail/Highway Bridge in Saco, Maine. 

Published by the American Railway Engineering Association, March, May, October and December at 

50 F St.. N.W.. Washington, D.C. 20001 

Second class postage at Washington. D.C. and at additional mailing offices 

Subscription $81.00 per annum 

Copyright© 1997 

AMERICAN RAILWAY ENGINEERING ASSOCIATION (ISSN 0003-0694) 

All rights reserved 

POSTMASTER: Send address changes to American Railway Engineering Association Bulletin. 

50 F Street. N.W.. Washington, D.C. 20001 

No part of this publication may be reproduced, stored in an information or data retrieval system, or transmitted, in any form, or by 

any means — electronic, mechanical, photocopying, recording, or otherwise — without the prior written permission of the publisher. 

Published by the 

American Railway Engineering Association 

50 F St., N.W. 

Washington, D.C. 20001 




SUPPLIERS OF 



equipment and Quality Repair Ports 



RECONSTRUCTION OF SACO RIVER 

RAIL/HIGHWAY BRIDGE CROSSING: 

A PARTNERING PROJECT IN DESIGN, 

CONSTRUCTION AND FINANCE 

By: Vinay Mudholkar* 
(This is a synopsis of the presentation made to the 
AREA Technical Conference on March 18, 1997) 

Guilford Rail System's east-west main-line traverses thru the heart of the City of Saco, Maine 
and is carried over Saco River falls on the bridge called Cataract Bridge. The bridge structure also 
carries busy Main street (Rt. 9) at grade across the railroad. The railroad bridge and highway bridge 
were interconnected structurally and were supported on common piers. This unique complexity 
posed a difficult challenge to the engineers of the railroad, DOT and City and required a strong part- 
nership between designers, constructors, and suppliers for the success of the project. 

The highway/railroad bridge crossing is composed of 5 interconnecting super structures, which 
were replaced during the construction. The 4 bridge spans carry 2 main-line tracks, but at present 
time only one track is in operation. Due to the existing conditions, track speed was restricted to 10 
MPH. Upon completion of the project, the speed was raised to 30 MPH. Cross vehicular highway 
traffic today is 22,250 (AADT) travelling at 25 MPH limit. Highway traffic growth was projected at 
31,150 (AADT) by the year 2012. During the summertime tourist traffic is significant in the area and 
is a major revenue source to local merchants.The rail corridor has also been a subject of study for 
passenger trains and intermodal traffic and, therefore, the provision of double track was an essential 
capacity element. 

A joint partnering effort was undertaken right from the start of the design process. This part- 
nering effort addressed concerns of highway traffic, railroad operations, requirements of utility com- 
panies, needs of local merchants and citizens affected by the project. Without such cooperation dur- 
ing the design and construction this project could not have been accomplished within the limited 
resources and time frame. The Project also was safely completed using safe construction practices 
and good project management. 

The bridge has an interesting long history and has played an important role in local as well as 
interstate transportation. Major historic events were as follows: 

1837: Town of Saco built 2 span roadway bridge carrying Main street over Saco River 

1872: Boston and Maine Railroad builds railroad bridge thru, replacing roadway spans; the 
structure is now modified to carry two modes of transportation. 

1884: Boston and Maine rebuilt the structure for higher capacity. 

1887: Biddeford and Saco Street railway was built along the center of Main street crossing; 
intersecting the freight lines. Structure is now carrying three modes of transportation. 

1919: All bridge spans were replaced; disputes arose amongst the owners; Public Utilities Com- 
mission of State of Maine stepped in and established a cost allocation for the construction and future 
maintenance. 



*Deputy Director; MBTA — Former Chief Engineer; GTI — Boston, MA. 



Bulletin 760 — American Railway Engineering Association 



1939; Biddeford and Saco Sheet Railway discontinued service and removed the street railway 
tracks. Main street became part of Maine State Highway system and is now designated as Route 9. 

1995: Design, specifications, plans were prepared and cost allocation was agreed upon for con- 
struction and future maintenance. 

1996: Thru the partnering efforts, the highway and railroad superstructure was replaced within 
24 days and a budget of 2.5 million dollars. Work was performed 24 hours per day, 7 days per week. 

Public meetings and hearings were held in the City of Saco for public information and com- 
ments. City staff was extremely helpful in accommodating the project needs and clo.sely worked with 
the engineers in the entire design and construction process. It was realized that the project construc- 
tion could be speeded up by working round the clock, 7 days a week for 4 to 5 weeks. The very crit- 
ical element was the closure of the highway for that construction time period. Detour arrangements 
were worked out on the local roads and toll free access was provided to motorists to travel between 
two local exits of the Maine Turnpike. A bus shuttle was also arranged to transport local citizens who 
used to walk across the bridge using the sidewalks. 

Amongst the suppliers. Premier Company of Portland. Oregon played a significant role in spe- 
cially manufacturing the railroad bridge crossing concrete deck, which supports the railroad's two 
main-line tracks, on the steel deck plate girders. For over 28 years Premier modular concrete railroad 
crossings have been successfully used in the rail industry and it provided expertise in modifying basic 
modules with appropriate structural requirements. The conventional wood crossing deck supported 
on wood ties was a constant troublesome area for repairs and was a rough riding surface. Roadway 
salt had leaked thru the timber deck causing major corrosion of the bridge steel. The ride over the 
timber deck at the crossing was certainly not comfortable to motorists, and timber fasteners used to 
cause problems. Frequent crossing deck repairs required street closure periodically, causing incon- 
venience to motorists and local businesses. The selection of Premier concrete crossing provided a 
smooth and durable ride quality surface eliminating repairs. 

Speed and ease of total construction of the project was achieved thru prefabrication and simul- 
taneous work activities by a good work plan (Figures 1 & 2). Premier bridge deck crossing modules 
were specially designed and cast for this project. Premier modules were pre-cast reinforced concrete 
modules which used 7,500 psi concrete at 28 day strength. The average strength tested was 8,000 psi 
for concrete. Reinforcement used was 70,000 psi. Rails are encased in a continuous 85 durometer 
non-conductive TPE rubber boot, to dampen vibration and prevent abrasion of concrete against rail- 
steel. This also results in electrical isolation and less rail wheel noise. Rails can easily be replaced by 
removing center panels. Rails are locked into exactly formed recesses supported on a solid reinforced 
concrete base slab unit, installed on the bridge girders. The heaviest pre-cast concrete section 
weighed under 29,000 lbs. The standard module was 8.00 ft. long and 13 ft. wide and designed to 
carry highway loading of HS25 and railroad loading of E80. Units were 1 8 inches thick and rein- 
forced to carry the loadings. In all, there were 6 different module shapes used to accommodate the 
required bridge layout. Steel channels were used as shear connectors to provide connections to bridge 
plate girders. The units were cast at Premier's pre-cast facility in Ottsville, Pennsylvania. The units 
were quick to install and connect over the deck, and within 2 hours 90 feet of bridge deck was assem- 
bled for the support of the main-line rails. A standard boom truck was used to pull the rails and the 
center section was bolted down to secure the rail in the position (Figure 3). A % inch Fabrica pad was 
placed between the Premier modules and steel girders as the bearing pad to soften the deck impact 
on steel. Upon final placement, the modules were grouted at the connectors (Figure 4). The entire 
process was efficient and the main-line traffic was quickly restored over the bridge crossing without 
major delay to rail traffic (Figure 5). 



Paper by V. Mudholkar 




Figure 1 



P 



T ^ 








Figure 2 



Bulletin 760 — American Railway Engineering Association 




Figure 3 




Figure 4 



Paper by V. Mudholkar 




Figure 5 



Railroad girders were fabricated by National Eastern of Plainville. Ct.; using ASTM A588 
structural steel, with yield strength of 50,000 psi. Highway pre-cast prestressed concrete girders were 
fabricated by StresCon of New Brunswick, Canada using 6,000 psi concrete. Prestressing strands 
were 1/2 uncoated 7 wire low relaxation ASTM A416 Grade 270 steel, with yield strength of 270,000 
psi. The highway deck slab was cast in place using 3,000 psi concrete and 60,000 psi reinforcing 
steel. The new highway superstructure was supported on common piers of the railroad structure; but 
is not interconnected: thus separating ownership and maintenance responsibility. 

Two railroad spans which were not under the crossing were strengthened by repairs performed 
by railroad crews and additional cover plates were placed, using high strength bolts. Structural 
replacement work was performed by CPM Contractors of Portland, Maine; with good quality and 
efficiency. Track work and signal work was performed by railroad forces very effectively. Roadway 
approaches were built by the City. Utilities were also relocated during the construction; from old 
structures to new structures. 

These were: NYNEX Telephone conduit, water-line of Saco Water Company, sanitary sewer of 
City of Saco. fiber optics cables of U.S. Sprint and AT &T , Railroad signal conduit, overhead cables 
of Continental Cablevision, Central Maine Power and Saco fire department. All of the interested par- 
ties worked together in partnership and at reasonable costs. Excellent cooperation resulted in a good 
safety record, and no undue inconvenience was caused. Local merchants whose business were 
affected due to the Main Street closure were particularly to be thanked for their accommodation. 
Without their cooperation, this project could not have been accomplished within such a short con- 
struction period. 



Koppers inausxnes... solving nroDiems wr America s Maiiroaas 



Why Koppers Industries? 

Thirteen Strategically Located Wood 

Treating Plants • 920 Dedicated 

Employees • Over Eighty Years of 

Service to Railroads • A National Tie 

Procurement Network • Wood and 

Concrete Crossties • Wood and Concrete 

Switch Ties • Used Tie Disposal • Car 

Cleaning • Tie Pre-Plating • Track Panels 

• Concrete Turnouts • Wood and 

Concrete Grade Crossings 

Enough Said? 




Tie Pre-Plating 



Track Panels 



Tie Disposal 



Turnouts 



Railroads around the world know 
Koppers Industries, Inc. for high 
quality crossties. But there is more... 
much more. Today, KM is helping 
America's railroads with a wide array 
of innovative products and services... 
all designed to solve problems and 
improve efficiency. The Railroad 
Products and Services Division offers 
track panels, concrete turnouts. 



wood and concrete ties, switch ties 
and wood and concrete grade cross- 
ings. Our cost cutting services 
include tie pre-plating, tie disposal, 
warehousing, crosstie rehabilitation, 
car cleaning and special packaging. 
Call us today. Let's talk about our 
solutions to your problems. 
Koppers Industries. ..in partnership 
witli America's Railroads. rpsdi 



KOPPERS 

INDUSTRIES 



Railroad Products 
and Services Division 

Koppers Building, 

436 Seventh Avenue 

Pittsburgh, PA 15219 

Phone 1 -888-567-8437 

Fax 412-227-2328 



The Single-Component Solution 

OMNI's Standard Concrete-Rubber (SCR) System, 

concrete panels with built-in rubber flangeway fillers: 

Streamlines handling 

and installation 

• 

Eliminates the hassles 

associated with 

top-down rubber 

• 

Saves time and money 

1-800-203-8034 




OMBU^ 

GRADE CROSSING SYSTEMS ^_^ 




amON HEIPS YOU li 

(OHM JUL OF m ^ 
Kin CHEATER^ 
EFHOENCV m 





USA: Tel: 412/962-3571 • Fax: 412/962-4310 
CANADA: Tel: 905/873-9440 • Fax: 905/873-9449 
EUROPE: Tel: +44(0)1932-247511 • Fax: *44(0)1932-220937 
CATTRON" Info, via World TOde Web: www.cattron.com 

LJObS 




m®ra 



FREE BROCHURE 



Meet the family of track experts 

Tie Tampers in six sizes, Tie Handiers, Tie inserters, Grinding Moduies and 
Trains, Uitrasonic Raii Flaw Detection and a full range of Rail Measurement 
Sen/ices. Quality Equipment manufactured to meet or exceed the demand- 
ing standards of ISO 9001. When the bottom line counts, count on us. 




"Total Quality 
Keeps Us On Track"® 






-^ 


Pandrol 
Jackson 


















Pandrol Jackson, Incorporated 








200 South Jackson Road 
Ludington Michigan 49431 
Telephone 616-843-3431 














FAX 616-8434830 





















Proposed 1997 AREA Manual 
and Portfolio Revisions 

The following proposed Revisions of the AREA Manual for Railway Engineering and Portfolio 
ofTrackwork Plans have been recommended to the Association by the Technical Committee respon- 
sible for each after a letter ballot is approved by: (1) a two-thirds majority of the eligible members 
\ oting. and (2) by at least fifty percent of the total eligible voting members on the committee. They 
are being published here for comment by the general AREA membership and any other interested 
parties. Comments should be sent to AREA headquarters by June 15, 1997. These comments will be 
considered by the AREA Board of Direction in deciding whether to give final approval for inclusion 
of the proposed changes in the Manual and Portfolio Revisions, which if approved, go into effect 
August 1, 1997. 



Proposed 1997 Manual Revisions to 
Chapter 5 — Track 

Part 6 — Specifications and Plans for Track Tools 

Page 5-6-4. Substitute the following revised te.xt for Part 6, Specifications and Plans for Track Tools. 

Section 6.1 Specifications for Track Tools (1997) 
GENERAL 

1.0 Workmanship 

1 . 1 The steel used in the manufacture of all tools shall be free from pipe, porous centers, gross 
non-metallic inclusions or any other defects. 

1.2 The chemical composition of percussion tools will be as stated in 8.3.1. 

1.3 Unless specifically stated otherwise in the section on non-percussion tools, the chemical 
composition of non-percussion tools made from carbon steel will be as follows: 





Carbon 


Manganese 


Phosphorous 


Sulfur 


Grade 


Min Max 


Min Max 


Min Max 


Min Max 


Carbon 


.55 .70 


.60 .90 


.05 


.05 



1.4 All tools shall be made in a workmanlike manner and shall be free from cracks, seams, laps 
and other injurious discontinuities. Tools shall be free from burrs and sharp edges not specifically 
shown on the plans. 

1.5 Eyes of tools with handle holes must be on center and in true alignment. 

2.0 Finish 

2.1 Percussion Tools 

The body of the tool will be unpainted. The entire tool will be coated with a transparent lacquer 
type rust preventative. 

2.2 Non-Percussion tools 

The body of the tool will be coated with paint, oil or varnish to prevent corrosion. Each pol- 
ished cutting edge will be oiled or coated with a transparent lacquer type rust preventative. 



10 Bulletin 760 — American Railway Engineering Association 



3.0 Marking 

3. 1 Each tool shall be legibly marked by stamping the following. 

3.1.1 The manufacturer's name and/or trademark. 

3. 1 .2 A code indicating the production lot. 

3. 1 .3 For tools manufactured for use in the United States, any information required by the U.S. 
Department of Labor, Occupational Safety And Health Administration (OSHA). For tools manufac- 
tured for use in other countries, the requirements of that country will apply. This pertains primarily 
to lifting devices used by cranes but may also be required for other tools. The manufacturer will also 
furnish certified testing and/or other information with each item shipped as needed to comply with 
OSHA Standards or the requirements of other countries. 

3.1.4 If requested by the purchaser, a specific marking indicating the railroad for which the tool 
was made. 

3.2 The marking shall be located in a position which will not interfere with the quality or per- 
formance of the tool, and will not be removed by subsequent redressing. 

4.0 Inspection 

4. 1 The inspector representing the purchaser shall have free entry, at all times while the work 
on the contract of the purchases is being performed, to all parts of the manufacturer's works which 
concern the manufacture of the materials ordered. The manufacturer shall afford the inspector free of 
charge, all reasonable facilities and necessary assistance to satisfy the inspector that the material is 
being furnished in accordance with these specifications. Tests and inspections shall be made prior to 
shipment at the place of manufacture unless otherwise specified. 

4.2 The purchaser may make tests to govern the acceptance or rejection in the purchaser's lab- 
oratory or elsewhere. Such tests shall be made at the expense of the purchaser. 

4.3 Rejection — Material represented by samples which fail to conform to the requirements of 
these specifications will be rejected. 

4.4 Material which, subsequent to test and inspection at the manufacturer's plant or elsewhere, 
and the acceptance shows injurious defects will be rejected and the manufacturer shall be notified. 

5.0 Shipment or Delivery 

Tools shall be properly packed for shipment to avoid damage. All bundles and boxes shall be 
plainly marked with the name of the purchaser, purchaser's order number, the name of the supplier, 
and the point of shipment. 

6.0 Warranty 

The manufacturer shall warrant that all tools are free from defects in material, workmanship and 
heat treatment, that the tools meet all requirements of this specification, and that any defective tools 
will be replaced free of cost to the purchaser. Certified test report may be requested by the purchaser. 

PERCUSSION TOOLS 

7.0 Scope 

7.1 This section of specifications covers the contouring and metallurgical requirements for the 
manufacturing, ordering, inspection and acceptance of the following percussion tools. 

7.1.1 Metal to Metal Contact Striking Tools 

Spike Maul Plan No. 3-83 

Double-Face Sledge Hammers Plan No. 13-83 



Proposed Manual Changes 



7.1.2 Metal To Metal Contact Struck Tools 



Track Chisels 


Plan No. 


17-83 


Round Track Punch 


Plan No. 


19-83 


Track Spike Lifter 


Plan No. 


32-83 


Nut Cutter 


Plan No. 


35-83 


3 lb. Hot Cutter 


Plan No. 


36-83 


5 lb. Hot Cutter 


Plan No. 


37-83 


Drift Pin-Short 


Plan No. 


38-83 


Drift Pin-Long 


Plan No. 


39-83 


Spiking Too! 


Plan No. 


41-94 



8.0 Manufacture 

8. 1 Process — The shock resisting steel shall be made from carbon deoxidized special quality 
fine grain size alloy bar produced in accordance with ASTM A576, Standard Specification For Steel 
Bars, Carbon, Hot- Wrought, Special Quality. 

8.2 Heat Treatment 

8.2.1 Each tool classified in 7.1.1 and 7.1.2 shall be hardened by liquid quenching and subse- 
quent tempering in such a manner that the hardness range will be maintained to a sufficient depth to 
absorb the normal working stresses. This heat treatment shall be such that a fracture test of the tool 
will exhibit a silky, fine grained appearance according to Shephard Standard No. 6 or finer. 

8.2.2 All tools made with alloy steel to be redressed without subsequent heat treatment shall be 
initially heat treated so that the hardness specified in 8.3.2 is maintained to depth from the end not 
less than the average cross sectional thickness. 

8.3 Chemical and Hardness Requirements 

8.3. 1 Ail striking and struck tools (7.1.1 and 7. 1 .2) shall be made of shock resisting alloy steel 
chemical composition with standard AISl residuals. 





Carbon 


Manganese 


Phos. 


Sul- 
fur 


Silicon 


Vanadium 


Molyb- 
denum 


Grade 


Min Max 


Min Max 


Max 


Max 


Min Max 


Min Max 


Min Max 


Alloy 


.51 .60 


.75 1.00 


.025 


.025 


1.80 2.20 


.45 


.35 .50 



8.3.2 Hardness — All hardness tests shall be performed according to the latest revision of 
ASTM Spec. E-18. Frequency of testing should be performed to the requirements in the latest revi- 
sion of MIL-STD-105D, Military Standard Sampling Procedure Tables for Inspection Attributes." 

8.3.2.1 All struck surfaces shall be 44/48 Rockwell "C" Hardness. 

8.3.2.2 All striking surfaces shall be 51/55 Rockwell "C" Hardness. 

8.3.2.3 All cutting surfaces shall be 56/60 Rockwell "C" Hardness. 

8.3.2.4 All punch ends shall be 52/56 Rockwell "C" Hardness. 

8.4 Hardenability 

8.4.1 Alloy Steel — Composition of the steel shall be such that in the standard Jominy test the 
hardness is greater than 50 Rockwell C at 8/16 inch from the quenched end of the specimen. 

8.4.2 Frequency of Testing — The steel manufacturer shall have conducted a Jominy test from 
the first, middle and last ingot of each heat of steel purchased. 



Bulletin 760 — American Railway Engineering Association 



8.5 Microscopic Inclusion Evaluation 

8.5.1 Alloy steel shall meet the following requirements for inclusions. 

8.5.2 Test Specimens — Specimens shall be taken from approximately 4 inch (100 mm) forged, 
square section taken from the top and bottom of the first, middle and last ingot. The specimen shall 
be Vs by -A inch (9.5 by 19 mm) and shall be taken from an area midway between the center and out- 
side of the test section. Procedures outlined in the latest revision of ASTM Method E 45 shall be fol- 
lowed. 

8.5.3 Examination and Limits — Specimens shall be examined in accordance with the latest 
revision of ASTM Method E 45, Method D, using the modified JK Chart Fig. 12 of Plate III. The 
worst field in any specimen shall not exceed the following limits: 





A 


B 


C 


D 


Thin 


3.0 


3.0 


2.5 


2.0 


Thick 


2.0 


2.5 


1.5 


1.5 



8.6 Nondestructive Test Requirements 

8.6. 1 To insure that all tools are free from defects listed in 1 .0, each tool shall be inspected after 
finished grinding by the supplier according to one of the following procedures: 

8.6.1.1 Magnetic Particle Inspection in accordance with the latest revision of ASTM Method 
A-275. 

8.6.1.2 Liquid Penetrant Inspection in accordance with the latest revision of ASTM 
Recommended Practice E-165. 

9.0 Design 

9.1 All tools shall conform substantially when applicable to the dimensions set forth. 
Dimensions for head contours as shown in Plans A-83, B-83 or C-83, D-83. 

9.1.1 Head Contour 

9.1.1.1 Heads of tools with a round cross section shall be ground to the comer contours pre- 
scribed in Plans A-83, B-83 or C-83. 

9.1.1.2 Heads of tools with a hexagonal or octagonal cross section should also be ground to the 
comer contours prescribed in Plans A-83, B-83 or C-83. In addition, the arcs not tangent to the hexag- 
onal or octagonal comers shall be "blended" into a smooth contour similar to that shown in Plan D-83. 

9.1.1.3 Punch ends shall have comer radii according to 9. I.I.I, but with no crown radius. 

9.1.1.4 All ground surfaces shall be free of decarburization. 

Non-Percussion Tools 
(Materials, Inspection And Physical Tests) 

10.0 Clay Pick— Plan No. 1 

Chemical composition for carbon steel as specified in 1.3, or alloy steel as specified in AISI 
4140. No special tests required. 

11.0 Tamping Pick— Plan No. 2 

Chemical composition for carbon steel as specified in 1.3, or alloy steel as specified in AISI 
4140. No special tests required. 

12.0 Spike Maul— Plan No. 3 

See percussion tools. 



Proposed Manual Changes 13 



13.0 Track Wrenches— Plan No. 4 

Chemical composition for carbon steel as specified in 1.3. One wrench to be tested from each 
lot of 10 dozen or less by applying for 1 minute a load of 400 lb. at a point distant from the jaw end 
equal to 95 percent of the total length of the wrench without any spreading of the jaw or any perma- 
nent set in the handle. If requested by the purchaser. Section 8.6. Nondestructive Test Requirements 
will be adhered to. 

14.0 Lining Bars— Plan No. 5 

Chemical composition for carbon steel as specified in 1 .3. One bar to be tested from each lot of 
10 dozen or less by applying a load of 350 lb. 9 in. from the end of the handle, with the point suit- 
ably secured 6 in. from the end, without leaving a permanent set in excess of 'A in. 

15.0 Rail Tongs— Plan No. 6 

Chemical composition for carbon steel as specified in 1.3. No special tests required. 

16.0 Tie Tongs — Plan No. 7 

Chemical composition for carbon steel as specified in 1.3. No special tests required. 

17.0 Timber Tongs— Plan No. 8 

Chemical composition for carbon steel as specified in 1 .3. Three pair of tongs to be tested from 
each lot of 10 dozen or less by suspending a load of 300 lb. or 400 lb. Work wi.se in the tongs with 
the handles in a horizontal position and supported 2 in. from the end. Deflection with 300 lb. weight 
shall not exceed 1 in. with no permanent set, and with 400 lb. weight deflection shall not exceed 
l-'/j in. with a permanent set not to exceed 'A in. 

18.0 Spike Puller— Plan No. 9 

Chemical composition for carbon steel as specified in 1 .3. One puller from each lot of 10 dozen 
or less to be tested in actual use by pulling a spike with a standard claw bar. 

19.0 Rail Fork— Plan No. 10 

Chemical composition for carbon steel as specified in 1.3. No special tests required. 

20.0 Claw Bar— Plan No. 11 

Chemical composition for carbon steel as specified in 1.3. In the manufacture of claw bars, 
Section 8.6, Nondestructive Test Requirements will be adhered to. One bar from each lot of 10 dozen 
or less to be tested by placing the claws of the bar Vi in. under the head of a standard spike, rigidly 
placed and so located as to hold the bar in a horizontal position while a shock load equivalent to that 
of a 200 lb. weight falling a distance of 1 ft. is applied to the handle at a point 5 in. from its end, with- 
out the toes showing any cracks or the handle taking any permanent set. 

21.0 Track Adz— Plan No. 12 

Chemical composition for carbon steel as specified in 1.3. Test one adz in each lot of 10 dozen 
or less by subjecting cutting edge to 5 normal blows on metal of the same composition as a railroad 
spike without breakage or serious nicking. 

22.0 Carpenter's Adz— Plan No. 12A 

Chemical composition for carbon steel as specified in 1.3. No special tests required. 

23.0 Double Face Sledge— Plan No. 13 

See percussion tools. 



14 Bulletin 760 — American Railway Engineering Association 



24.0 Tamping Bar — Plan Numbers 14-15 

Chemical composition for carbon steel as specified in 1.3. No special tests required. 

25.0 Tie Plug Driver— Plan No. 16 

Material as shown on plan. No special tests required. 

26.0 Track Chisels— Plan No. 17 

See percussion tools. 

27,0 Round Track Punch— Plan No. 19 

See percussion tools. 

28.0 Track Gage— Plan No. 20 

Material as shown on plans. No special tests required. 

29.0 Track Gage with Wood Rod— Plan No. 20-A 

Material as shown on plans. No special tests required. 

30.0 Track Shovel— Plan No. 21 

30.1 Scope and Design 

This specification covers the welded or riveted type and the solid shank type with either wood, 
malleable iron, combination wood metal, or user approved composition handle tops. Dimension shall 
conform to plans, which are made part of this specification. A variation of Vi in. more or less from 
the dimensions shown on the plan for the length of the strap or shank and handle will be allowed. A 
variation of % in. more or less from the dimensions shown on the plan for the width or length of the 
blade will be allowed, but the total variation in the overall length of shovels shall not exceed V2 in. 
more or less of the dimensions shown on the plan. 

30.2 Materials 

Blades shall be of carbon or alloy steel, with a Rockwell (Re) hardness for carbon steel of 45 
to 50. 

Carbon steel blades shall have a thickness of not less than No. 13 gage and alloy blades shall 
be not less than No. 14 gage U.S. Standard, the gage to be measured at the point where the hardness 
is taken. For welded or riveted types, the straps shall be welded or riveted to the blade. 

30.3 Handles and Tops 

This specification covers either wood, malleable iron, combination wood metal, or user 
approved composition handle tops. Wood handles shall be made of ash and shall conform to Grade 
AA and be in accordance with the general Specifications for Handles for Track Tools. 

30.4 Tests 

One shovel from each lot of 10 dozen or less shall be selected and metal straps (curved to fit 
the contour of the handle) shall be clamped to the upper and lower parts of the handle, after which 
the shovel shall be placed in a prying position, supported at the end of the blade by clamps and shall 
be capable of sustaining a load of 200 lb. suspended from the end for a period of 2 minutes without 
showing any permanent set, fracture or distortion. 

Alloy steel shovels which have been given heat treatment to insure uniformity in hardness shall 
be subject to shock test to insure against brittleness. The test shall be made by forcibly striking the 
blade of the shovel with a hand hammer at several places when placed on an anvil. 



Proposed Manual Changes 15 



31.0 Ballast Fork— Plan No. 22 

31.1 Scope and Design 

The dimensions shall conform to the plans, which are made part of this specification. The total 
variation in the overall length of the forks shall not exceed Vi in. more or less of the dimensions 
shown on plan. 

31.2 Material 

Forks shall be made of high grade carbon steel. Tines of forks shall show Rockwell (Re) har- 
ness of 35-45. Straps shall be 0.04 U.S. Standard gage steel. 

31.3 Handles 

This specification covers either wood, malleable iron, combination wood metal, or user ap- 
proved composition handle tops. Wood handles shall be made of ash and shall conform to Grade AA 
and be in accordance with the general Specifications for Handles for Track Tools. 

32.0 Track Tool Handles— Plan No. 25 

See Specification For Ash And Hickory Handles For Track Tools for material requirements. No 
special tests required. 

33.0 Scoop— Plan No. 26 

33.1 Scope and Design 

This specification covers the welded or riveted type and the solid shank type with either wood, 
malleable iron, combination wood metal, or user approved composition handle tops. Dimension shall 
conform to plans, which are made part of this specification. A variation of Vi in. more or less from 
the dimensions shown on the plan for the length of the .strap or shank and handle will be allowed. A 
variation of Vi in. more or less from the dimensions shown on the plan for the width or length of the 
blade will be allowed, but the total variation in the overall length of scoops shall not exceed '/2 in. 
more or less of the dimensions shown on the plan. 

33.2 Materials 

Blades shall be of carbon or alloy steel, with a Rockwell (Re) hardness for carbon steel of 45 
to 50. 

Carbon steel blades shall have a thickness of not less than No. 13 gage and alloy blades shall 
be not less than No. 14 gage U.S. Standard, the gage to be measured at the point where the hardness 
is taken. For welded or riveted types, the straps shall be welded or riveted to the blade. 

33.3 Handles and Tops 

This specification covers either wood, malleable iron, combination wood metal, or user 
approved composition handle tops. Wood handles shall be made of ash and shall conform to Grade 
AA and be in accordance with the general Specifications for Handles for Track Tools. 

33.4 Tests 

One scoop from each lot of 10 dozen or less shall be selected and metal straps (curved to fit the 
contour of the handle) shall be clamped to the upper and lower parts of the handle, after which the 
shovel shall be placed in a prying position, supported at the end of the blade by clamps and shall be 
capable of sustaining a load of 200 lb. suspended from the end for a period of 2 minutes without 
showing any permanent set, fracture or distortion. 

Alloy steel scoops which have been given heat treatment to insure uniformity in hardness shall 
be subject to shock test to insure against brittleness. The test shall be made by forcibly striking the 
blade of the scoop with a hand hammer at several places when placed on an anvil. 



16 Bulletin 760 — American Railway Engineering Association 



34.0 Aluminum Track Level And Gage — Plan No. 27 

Material as shown on plans. No special tests required. 

35.0 Scythe— Plan No. 28 

No special tests required. 

36.0 Snath— Plan No. 29 

Material as shown on plans. No special tests required. 

37.0 Spot Board— Plan No. 30 

Material as shown on plans. No special tests required. 

38.0 Rail Tongs for use with crane — Plan No. 31 

Material as shown on plans. In the manufacture of the rail tongs. Section 8.6, Nondestructive 
Test Requirements will be adhered to. 

39.0 Track Spike Lifter— Plan No. 32 

See percussion tools. 

40.0 Drive Spike Extractor Socket Wrench — Plan No. 33 

No special tests required. 

41.0 Rail Thermometer— Plan No. 34 

Material as shown on plans. No special tests required. 

42.0 Nut Cutter— Plan No. 35 

See percussion tools. 

43.0 Hot Cutter (3 Pound)— Plan No. 36 

See percussion tools. 

44.0 Hot Cutter (5 Pound)— Plan No. 37 

See percussion tools. 

45.0 Drift Pin (Short)— Plan No. 38 

See percussion tools. 

46.0 Drift Pin (Long)— Plan No. 39 

See percussion tools. 

47.0 Spiking Tool— Plan No. 41 

See percussion tools. 

48.0 Switch Clip Wrenches— Plan No. 43 

Chemical composition for carbon steel as specified in 1.3. If requested by the purchaser, 
Section 8.6, Nondestructive Test Requirements will be adhered to. 



Proposed Manual Changes 



17 





PLANS FOR TRACK TOOLS (1997) 




Plan 




Grade of 






Number 


Description 


Steel 




Hardness 


1-62 


Clay Pick 


Carbon or 


Alloy 


425-500 BHN 


2-62 


Tamping Pick 


Carbon or 


Alloy 


425-500 BHN 


3-83 


Spike Maul 


Alloy 




51-55 Re 


4-62 


Track Wrenches 


Carbon 




375-450 BHN 


5-62 


Lining Bars 


Carbon 




300-375 BHN 


6-62 


Rail Tongs 


Carbon 






7-93 


Tie Tongs 


Carbon or 


Alloy 




8-93 


Timber Tongs 


Carbon or 


Alloy 




9-94 


Spike Puller 


Carbon 




375-450 BHN 


10-97 


Rail Fork 


Carbon 




275-350 BHN 


11-97 


Claw Bar 


Carbon 




300-375 BHN 


12-62 


*Track Adz 


Carbon or 


Alloy 


375-450 BHN 


12-A-62 


*Carpenters Adz 


Carbon or 


Alloy 




13-83 


Double Faced Sledge 


Alloy 




5 1 -55 Re 


14-62 


Chisel End Tamping Bar 


Carbon 




425-500 BHN 


15-62 


Spear End Tamping Bar 


Carbon 




425-500 BHN 


16-62 


Tie Plug Driver 


Carbon 






17-83 


Track Chisel 


Alloy 




44-48 Re (Head) 


19-83 


Round Track Punch 


Alloy 




44-48 Re (Head) 


20-62 


Track Gage — Pipe Center 


See Plan 






20-A-62 


Track Gage — Wood Center 


See Plan 






21-62 


Track Shovels 


Carbon or 


Alloy 


45-50 Re 


22-62 


Ballast Forks 


Carbon 




35-45 Re 


25-83 


Track Tool Handles 








26-62 


Scoop 


Carbon or 


Alloy 


45-50 Re 


27-80 


Aluminum Combination 
Track Level And Gage 
(Insulated) 








28-62 


Scythe 


See Plan 




54-58 Re 


29-62 


Snath 


See Plan 






30-62 


Spot Board 


See Plan 






31-97 


Rail Tongs For Use With 
Crane (Type 1 And 2) 


See Plan 






32-83 


Track Spike Lifter 


Alloy 




44-48 Re (Head) 
44-48 Re (Claw) 


33-83 


Drive Spike Extractor 
Socket Wrench 


Carbon 




300-350 BHN 


34-71 


Rail Thermometer 








35-83 


Nut Cutter 


Alloy 




44-48 Re (Head) 
56-60 Re (Point) 


36-83 


(3 lb.) Hot Cutter 


Alloy 




44-48 Re (Head) 
56-60 Re (Point) 


37-83 


(5 lb.) Hot Cutter 


Alloy 




44-48 Re (Head) 
56-60 Re (Point) 


38-83 


Drift Pin (Short) 


Alloy 




44-48 Re (Overall) 


39-83 


Drift Pin (Long) 


Alloy 




44-48 Re (Overall) 


41-94 


Spiking Tool 


Alloy 




44-48 Re (Head) 
52-56 Re (Point) 


43-97 


Switch Clip Wrench 


Carbon 




375-450 BHN 


*When specif 


led, the small eyed track tools will be furnished with AREA handles. The handles are 



to be properly fitted and wedged. 



18 



Bulletin 760 — American Railway Engineering Association 



Page 5-6-27. Insert revised Plan 10, Rail Fork. 




PLAN 10 - 97 - AREA RAIL FORK 



Proposed Manual Changes 



19 



Pase 5-6-28. Insert revised Plan 11, Claw Bar. 




^ r 



r 



r 




u 




[-3'/,-H 






PLAN 11-97 AREA CLAW BAR 



20 



Bulletin 760 — American Railway Engineering Association 



Page 5-6-39. Insert revised Plan 31, AREA Rail Tongs for Use with Cranes (Type 1 and Type 2). 



. -I SS_i Q- 



y? X Q. 



-" u. (E -"IT » L.. 

—I «< a)< ■< — .fo: 




.'Ki zi 



PLAN 31 -97 AREA RAIL TONGS FOR USE WITH CRANES (TYPE 1) 



Proposed Manual Changes 



2 o o 



:dT^ 



st:^^ 



inr 



-.% i 




PLAN 31-97 AREA RAIL TONGS FOR USE WITH CRANES (TYPE 2) 



Bulletin 760 — American Railway Engineering Association 



Page 5-6-46. Add new Plan 43, AREA Switch Clip Wrench. 



3iywixoaddy ,.yi sz 




o o 

l\l UT 





?- 








_) 


rr 


m 




1. 


n 


rr 






r- 


m 












<I 


1. 1 


, 




iL 






<I 
















y 








_J 


■a- 


(M 






0-) 


(M 




*— ' 


(0 


(Tl 




^ 










. , 


, , 




^ 






ri 


^ 


?^ 




" 


- 


t— LlI 






_l M 


t 




O >-• 


■— • 




OQ cn 




— 










t 


"^ 












^^ 







i 




-H y i-^ <x 



..yi E 



n 



T~ 



..Vi c 



PLAN 43 - 97 - AREA SWITCH CLIP WRENCH 



Specialists In keeping 
Bridges In service 

•TIMBER •STEEL •CONCRETE 

Over 40 years of railroad experience. 
Inspect. . . Repair. . . Treat. . . Strengthen 




RAILROAD DIVISION 
P O Box 8276 • Madison, Wisconsin 53708 
608/221-2292 • 800/356-5952 



WE'RE THE BRIDGE PRESERVERS 



URVE 

CONQU EROI 




Rocia prestressed concrete ties conquer the curves. They 
provide a stiffer track modulus, improved lateral stability 
and gauge control, increased rail life, greater locomotive 
fuel economy, and significantly lower maintenance costs. 

Two North American manufacturing plants to serve you. 
Unsurpassed quality and durability. Go with the leader in 
North American prestressed concrete ties. Call: 



uaii: ^^ 



^^^t"-:. 






h-iJ^^^i^^' 



Concrete Tie, inc. 

701 West 48th Avenue • Denver, Colorado 80216 
(303) 296-3500 • Fax (303) 297-2255 

268 East Scotland Drive • Bear, Delaware 19701 
(302) 836-5304 • Fax (302) 836-5458 



23 




Engineered To Handle Higher Speeds, 
Tighter Turns And Heavier Loads, 
Today's Wood Crossties Offer You Superior Cost 
And Performaince Advantages. 





Maybe not. Maybe even tougiier wlien you consider ttie 
tremendous variables that track engineers contend with. 
Materials with differing physical properties and maintenance 
needs. Track structures that must stand up to an enormous 
range of loadings and speeds, for long periods of time, in all 
types of weather, over all kinds of terrain. Enter yet another 
factor — high-speed passenger trciffic — and the complexity 
increases. 




Fortunately, there is a proven performer you can count on 
to handle your toughest demands. The treated wood crosstie. 

Pre-engineered by nature and enhanced by man, the wood 
crosstie is a marvel of natural science and applied technology. 

Concrete Cuidence That UJood Is Vour Best Choice. 

For 150 years, the wood tie has taken everything 
that man and nature have dished out. Steep 
grades. Brutal environments. High speeds. 
Heavier axle loads. Tighter curves. 

Through it all, resilient wood has been and 
continues to be the material of choice for durability, 
economy and strength. No other material matches 
wood's value on freight and passenger lines. 

And innovations in hardware, installation 
methods and wood preservation have taken a 
good thing and made it even better By improving gage 
retention. Minimizing maintenance. And increasing tie life. 




Building track structures is decidedly 
complex. But choosing the best crosstie 
isn't exactly "rocket science". 
Just say wood. 



m 

Railway Tie 
Association 



Wood crossties. Something to build on. 

115 Commerce Drive • Suite C 
Fayetteville,GA 30214 

Phone (770) 460-5553 
Fax (770) 460-5573 



Proposed Manual Changes 25 



Proposed 1997 Manual Revisions to 
Cliapter 8 — Concrete Structures and Foundations 

Part 21 — Inspection of Concrete and Masonry Structures 

Page 8-21-1. Insert revised Part 21, Inspection of Concrete and Masonry Structures, including the 
Effects of Fire in a new Commentary. 

Part 21 — Inspection of Concrete and Masonry Structures 

21.1 General 

21.1.1 All concrete and masonry structures and components should be given thorough, detailed 
inspections at scheduled intervals. For timber and steel components, refer to Chapter 7 and Chapter 
15, respectively. The depth and detail of the inspection should be based on the condition and age of 
the structure, and traffic type and tonnage in order to determine that the physical condition of each 
structure is suitable for the imposed loading. A record of physical conditions should be kept. 

21.1.2 A special inspection may be required when the structure is subjected to abnormal con- 
ditions which may affect the capacity of the structure such as: floods, storms, fires, earthquakes, col- 
lisions, overloads and evidence of recent movement. Refer to commentary for information related to 
inspection of fire damaged concrete. 

21.1.3 The inspector should review prior inspection reports before making the inspection. 
Previously noted defects should be examined in the field and any changes in conditions recorded. Field 
book, sketch pad, inspection form, camera, monitoring gages, etc. should be used to record the inspec- 
tion data. Appropriate personal safety equipment should be employed throughout the inspection. 

21.2 Reporting of Defects 

2 1 .2. 1 When the inspector finds defects that appear to be of such a nature as to make the pas- 
sage of traffic unsafe, the condition should immediately be reported. After steps have been taken to 
protect traffic, the train dispatcher and appropriate officers should be notified, consistent with estab- 
lished policies, recommending a speed limit and briefly describing the conditions which prompted 
the action. The inspector should follow this immediately with a report so that a detailed investigation 
and recommendation for repair can be made. 

21.2.2 Upon completion of the inspection, a written record covering the inspection should be 
forwarded to the engineer or other officer in charge of maintenance. Upon receipt of the report, a 
review should be made to determine the need for remedial action. 

21.3 Inspection 

21.3.1 The inspection of concrete and masonry structures should be carried out in a methodi- 
cal manner. Of primary importance in all structures is evidence of distress, misalignment, deflection, 
settlement, cracks, and general deterioration. Evidence of deterioration of concrete such as width and 
length of structural cracks, size and location of spalling and .scaling, and location and extent of water- 
saturation of concrete should be recorded. 

21.3.2 The inspector should report indications of failure in any portion of the structure and any 
conditions which could contribute to a future failure. 

2 1 .3.3 If practical, the inspector should observe the structure during passage of a train, so that 
the effects of vibration, sidesway and deflection may be noted. 

21.3.4 Reference points should be established for monitoring misalignment, deflection, settle- 
ment, and cracks. The amount of tilt, separation between components, width and length of cracks, efflo- 
rescence and rust-staining and other measurements necessary for future checking should be recorded. 



26 Bulletin 760 — American Railway Engineering Association 



21.3.5 The inspection should include the structure and all related features. The following addi- 
tional items should be covered in detail. 

21.3.5.1 Track 

The inspector should note the alignment, profile and surface of the track on the structure, its 
approaches and bridge ends. Any irregularities in line or surface should be noted along with their 
magnitude, location and any other information that may indicate the cause of the irregularities. Depth 
of ballast and condition of ballast, ties and hardware should be noted. Line swings may be an indi- 
cation of pier movement. Sags in the track over the structure may indicate settlement. 

21.3.5.2 Site and Crossing 

a) Where a structure cros.ses over a waterway, the inspector should note the condition and 
alignment of the waterway. The condition of the slopes and any slope protection (such as riprap) 
should be noted along with any indication of debris accumulation. The inspector should note any 
indication of damage from marine collision, ice or debris. 

b) Where scour is possible, the channel bottom at piers and abutments should be checked by 
sounding, probing or other means. 

c) Where a structure carries tracks over a roadway or another track, the inspector should note 
any indication of collision damage from high or wide loads. Roadway clearances should be measured 
and signage verified for accuracy. 

21.3.5.3 Foundations, Piers and Abutments 

a) The inspector should note any settlement and/or rotation of foundations, piers, abutments or 
their component parts. Reference points should be established for monitoring of structural movement 
if appropriate. 

b) The type of foundation and type and condition of material used in the various structural 
components should be noted. Location and extent of exposed or corroded reinforcing bars should be 
reported. The condition of the structure at the bridge seats, bearings and near the waterline should 
also be investigated. 

c) Crack width, orientation and location should be noted. Widths and lengths of structural 
cracks should be marked and dated to monitor crack progression. On masonry structures note 
cracked, shifted, or missing stones, and condition of mortar. 

d) Location, size and description of unsound areas, spalling, scaling or other deterioration 
should be noted. 

e) Condition of retained fill, drainage and slope protection at abutments should be inspected. 
Water-saturated masonry or concrete and extent of efflorescence and rust-staining should be noted. 
Check weepholes and drains for clogging. 

21.3.5.4 Pile and Pile Bents 

a) Inspection of piling and pile bents should be in general conformance with Article 21.3.5.3. 
For timber and steel components, refer to Chapter 7 and Chapter 15, respectively. 

b) Alignment and condition of piling should be recorded. Impact damage from debris, vessels 
or vehicles should also be noted. 

c) Condition of piles should be investigated for .soundness. Loss of section and cracking should 
be noted. These may be especially severe in a marine environment, particularly in the tidal zone. 

d) Connections between cap and piling should be inspected. 

e) Bracing members and their connections should be inspected. 



Proposed Manual Changes 27 



21.3.5.5 Underwater Inspections 

The need and frequency for underwater inspections should be evaluated for every structure hav- 
ing submerged components. These inspections should identify the channel bottom conditions and 
presence of any scour, extent of foundation exposure and any undermining, and all deterioration and 
damage below water. 

a) Divers should be experienced in the inspection of bridges. 

b) Inspection data should be recorded by written description, sketches, reports, photography 
and/or video. 

c) During high water events when scour conditions may be expected, channel activity should 
be monitored, which may include the use of sonar readings, until inspections can be made. 

21.3.5.6 Retaining Walls 

a) The inspector should note any settlement and/or rotation of retaining walls. Changes in wall 
alignment or cracks in earth embankment which parallel the wall should be noted. 

b) Concrete inspection should be in general conformance with 21.3.5.3. 

c) Condition of retained fill and drainage at walls should be inspected. The extent of water-sat- 
urated concrete and exposed or corroded reinforcing bars should be recorded. 

21.3.5.7 Slabs and Beams 

a) Inspector should note if prestressed or conventionally reinforced concrete is used in the struc- 
ture. Method of construction, cast-in-place or precast, simple or continuous, should also be indicated. 

b) Any cracks that open and close under traffic, diagonal cracks near supports, or wide or 
numerous cracks in any location should be reported immediately to the proper authority. Acute cor- 
ners of skewed bridges should be examined for cracking. 

c) Structural members should be inspected for excessive deflection or misalignment. 

d) Curbs, ballast retainers, walkways and handrails should be inspected, noting the condition as 
to soundness and security of fastening devices. Soundness, uniformity and condition of bearings and 
bearing areas should also be noted. Areas exposed to drainage should be checked for spalling and 
cracking. 

21.3.5.8 Box Girders 

a) Type of box construction (precast, segmental, pre-tensioned, post-tensioned, simple or con- 
tinuous spans) should be recorded. 

b) General inspection guidelines should be as outlined in 21.3.5.7. Top flange, bottom flange 
and web walls should be inspected when accessible. Chamfers of boxes should be inspected for 
cracking which may extend along the sides or bottom of the girders. 

c) Shear transfer devices between adjacent box girders should be inspected, where accessible. 
Condition of grout, hardware, tie rods, and other materials used in tying together adjacent box gird- 
ers should be noted. Evidence of differential box deflections or misalignments should be recorded. 

d) Condition of void drain holes and evidence of leakage between adjacent boxes should be noted. 

21.3.5.9 Arches 

a) Type of arch construction, such as segmental, open spandrel, closed spandrel, single or mul- 
tiple span should be noted. Shape of arch span (circular, elliptical or parabolic) should be recorded, 
if known. Type and general condition of material (brick, stone, mortar or concrete) should also be 
recorded. 



Bulletin 760 — American Railway Engineering Association 



b) Arch foundations should be investigated for settlement, shifting, scour and undermining. 

c) Arch ribs and bearing areas of arches at springings should be inspected for loss of cross sec- 
tion due to spalling or cracking. 

d) Open spandrel columns should be inspected with particular attention to areas near the inter- 
face with the arch rib and cap. 

e) Arch ribs connected with .struts should be inspected for diagonal cracking due to torsional 
shear. 

f) Floor systems of open spandrel arches and closed spandrel arches with no till material 
should be inspected as outlined in 21 .3.3.7. 

g) Inspect areas exposed to drainage and seepage for deteriorated and contaminated areas. For 
closed spandrel arches, note if weepholes are working properly. 

21.3.5.10 Structural Protection 

Structural protection devices including crash walls, cellular dolphins, pile clusters, sheer fences, 
floating sheer booms, anchored pontoons, fender systems, navigation lights and warning mechanisms 
should be inspected as part of the scheduled inspection of their related foundation or substructure ele- 
ment. The inspection should identify all deterioration, damage, displacement, misalignment, insta- 
bility, and any other detrimental conditions which would inhibit these devices from protecting the 
structure or cause them to create an obstruction. All submerged portions of structural protection 
devices should be inspected underwater based on the recommendations set forth in Article 21.3.5.5. 
The inspection of structural protection devices should also note any aspects which may present a haz- 
ard to navigation, and identify the necessary measures to correct the situation. 

21.3.5.11 Culverts 

a) Inspector should note any settlement, variations in cross-sectional shape and misalignment 
along the horizontal axis of a culvert. All joints between end treatments and within the culvert itself 
should be examined for differential movement, and all transverse or longitudinal cracking within a 
culvert should be noted. Look for holes appearing in the track structure as an indication of open cul- 
vert joints. 

b) A culvert should be inspected for any scour or undermining at either end. Any embankment 
damage around the culvert openings and debris or vegetation within the culvert should be noted. All 
submerged portions of a culvert should be inspected underwater based on the recommendations set 
forth in Article 21.3.5.5. 

c) Inspection of a concrete or masonry culvert in general should be in conformance with Article 
21.3.5.3 

21.3.5.12 l\innels 

a) Important features of a tunnel might be obscured by a shield or lining, therefore the inspec- 
tor should review plans, if available, prior to the inspection. Note the structural configuration, provi- 
sions for drainage, ventilation and lighting. Note if secondary passageways that would provide addi- 
tional access for inspection are present. 

b) Concrete inspection should be in general conformance with 2 1 .3.5.3. In exposed masonry con- 
struction, make special note of bulges in walls and displacement, shifting or loss of masonry or mortar. 

c) Walls should be inspected for indications of water leakage or ice buildup. The condition and 
effectiveness of drainage systems should be noted. 

d) Note whether ancillary systems for lighting, ventilation, and fire prevention are in working 
order, if discernible. 



Proposed Manual Changes 29 



e) The accumulation of trash or foreign debris or the blockage of safety niches should be noted. 

Any new construction above or adjacent to the tunnel should be noted. 

g) Horizontal and vertical clearances should be verified. Items causing changes in clearance 
should be noted. 

h) The inspector should note the alignment, profile and surface of the track and clearance of 
the tunnel. 

21.3.6 Bibliography 

Bridge Inspection Seminar Manual, American Railway Bridge and Building Association, 
Atlanta, Georgia, February 22-24, 1993. 

Bridge Inspector's Training ManuaI/90, FHWA-PD-9I-0I5, U.S. Department of Transporta- 
tion, Federal Highway Administration, May, 1991 

Underwater Inspection of Bridges, FHWA-DP-80- 1 , Federal Highway Administration, Novem- 
ber, 1989. 



Part 21 -Commentary 
Inspection of Concrete and Masonry Structures 

C21.1.2 Guidelines for Evaluating Fire Damaged Concrete Railway Bridges 

C21. 1.2.1 General 

Concrete structures exposed to fire may experience a permanent loss of strength, formation of 
structural cracks, surface spalling, and reinforcing damage. However, concrete structures exposed to 
fire generally perform well and usually are repairable. The heat conductivity of concrete is low and 
thus heat from a fire is usually confined to shallow depths. The extent of structural damage is related 
to the intensity and duration of the fire, and the mass and details of the concrete structure. 

The exposure of concrete to a temperature of 300°C (572°F) is significant for two reasons: 

• Below this temperature the effects of heat on concrete are likely to be insignificant. 

• Above this temperature concrete coloration changes may indicate permanent damage. 

Water directed on hot concrete may cause spalling, crack developinent and the embrittlement 
of steel. Fire fighting efforts should be directed to extinguishing the combustible material and not 
cooling the structure. 

Traffic should not cross the structure if significant deflection or distortion is noted or if there 
are reasons to doubt that adequate strength remains. 

C21. 1.2.2 Inspection 

Prior to the inspection of a damaged concrete structure, it should be determined whether the site 
is safe for entry. 

Damage may include the deflection of concrete beams and slabs, di.stortion of columns, crack- 
ing, spalling and un.sightly appearance. 

Inspection observations should include looking for and measuring any unusual component 
deflection, recording the location and extent of structural cracks, spalls and exposed reinforcing. Fire 
exposed surfaces should be mapped to indicate those areas having structural and cosmetic damage. 
If fire exposed surfaces exhibit colorations of pink, white or buff, those surfaces should be mapped 
and color noted. Surfaces may need to be cleaned of soot to make these observations. 



30 Bulletin 760 — American Railway Engineering Association 



Information concerning the combustible material, duration, intensity indicators and method for 
extinguishing should be obtained from eyewitnesses or other reliable sources for assistance in eval- 
uating the damage. Although any concrete coloration from the fire may provide sufficient informa- 
tion concerning the intensity of the fire, if coloration is not evident, to a lesser degree other materi- 
als associated with the fire site may have melted and may provide some indication of the fire 
intensity; such as: lead 327°C (62 IT), plastics 300-450°C (572-842°F), glass 400-500°C 
(752-932°F), aluminum 660°C (I2I8°F) and copper 1083°C (IQSTF). Other information concerning 
the original concrete strength, age, reinforcing details and types of aggregates may be obtained from 
structural plans, specifications and construction records. 

C21. 1.2.3 Evaluation 

Generally, all concrete that has coloration changes (pink, white, bufO is considered damaged. 
The pink coloration 300°C (572°F) experienced by heating concrete is the formation of ferrous salts 
and is more pronounced in concrete with siliceous aggregates. At approximately 600°C (IIIOT), 
concrete may have a whitish coloration from the hydration of lime. At 900°C (1650°F) the coloration 
may be grey-buff. 

Indications of possible structural damage may be evident by visual examination, but the extent 
of damage will require tests and analysis. Evaluation tools for testing include: surface hammer sound- 
ing, impact hammers, coring and/or drilling and pulse-echo non-destructive testing. Sounding the 
concrete surface with hammers may be sufficient to determine if there is any internal concrete delam- 
ination. Calibrated impact hammers can give direct measurements of the concrete compressive 
strength and may be used on sound and unsound concrete for quick strength comparisons. Coring will 
assist in determining the depth of damage and corings destructively tested will ascertain accurate 
compressive strength. A petrographic analysis of cored samples will give a detailed analysis of the 
concrete condition but the analysis is time consuming. Pulse-echo testing can give a rapid and accu- 
rate determination of internal concrete conditions relative to micro-cracking and bond loss. 
Additional testing may be needed for prestressed and post-tensioned concrete. 

Concrete strength decreases as temperature is increased and further decreases on cooling as a 
result of micro-cracking. Approximately 75% residual strength remains in most concrete after expo- 
sure to fire. This loss may be offset by excess residual strength of mature concrete. Internal induced 
stresses from differential heating may result in the formation of cracks. Young concrete may experi- 
ence more damage than mature concrete due to larger amounts of internal moisture that may convert 
to steam and increase internal tensile stresses. 

Damage may result from aggregate spalls due to physical or chemical changes. Explosive 
spalling may occur from the release of tensile stresses by the formation of steam within aggregates. 
Slough-off or the detachment of layers of concrete may occur where reinforcement is restrained. 
Igneous aggregates (granite, basalt) generally perform well when exposed to fire, carbonate aggre- 
gates (limestone) perform well to about 700°C ( 1290°F), and siliceous aggregates (quartz) do not per- 
form well due to expansion and cracking. 

The absence of deflection or distortion in any element may indicate that the steel was not damaged. 
Reinforcing steel usually recovers in strength unless exposed to temperatures over 600°C (lllOT). 
Anchorages of post-tensioned members may require special evaluation. The tension in pretensioned steel 
or post-tensioned ducts exposed by spalling should generally be assumed to be zero. Prestressed mem- 
bers may suffer substantial relaxation losses, additional to those allowed by normal design. Low relax- 
ation strands may have improved fire performance. At 300°C (572°F) the residual bond strength is 
approximately 85% and at 500°C (932°F) the bond strength is approximately 50% of initial bond. Bond 
strength losses of epoxy coated reinforcing steel subjected to fire may require special evaluation. 

Resins used in construction bonding of concrete elements and in repairs may not perform well 
in the presence of elevated temperatures. 



Proposed Manual Changes 3 1 



Hydrochloric acid fumes occurring in fires involving PVC and other plastic ducts may react 
with hardened cement paste to form calcium chloride which may constitute a hazard to the rein- 
forcement. A silver/chromate test can confirm the presence of calcium chloride ions. 

C21.1.2.4 Repairs 

Repair procedures, as applicable, are outlined in Part 14. 

Pulse-echo or other non-destructive testing may be used to confirm that all damaged concrete 
is removed and can be u.sed to confirm proper bonding of new concrete to old concrete and bonding 
to reinforcement. 

C21.1.2.5 Bibliography 

Evaluation and Repair of Fire Damage to Concrete, AC! SP-92, American Concrete Institute, 
Detroit, Michigan. 1986 

Fire Safety of Concrete Structures, ACI SP-80, American Concrete Institute, Detroit, Michigan, 
1983 

Guide for Determining the Fire Endurance of Concrete Elements, ACI216R-89, American 
Concrete Institute, Detroit, Michigan, 1989 

Reinforced Concrete Fire Resistance, CRSl Engineering Practice Committee, Concrete 
Reinforcing Steel Institute, Chicago, Illinois, 1980. 

C21.3.1 Inspection — Commentary 

There are many common defects that occur on concrete bridges. The following definitions are 
provided as a guideline for consistency in reporting of defects. 

Abrasion — Abrasion damage is the result of external forces acting on the surface of the con- 
crete member. Erosive action of silt-laden water running over a concrete surface and ice flow in rivers 
and streams can cause considerable abrasion damage to concrete. 

Cold joint displacement or deterioration — Unbonded concrete resulting from intended sepa- 
rate concrete placement or by lack of consolidation. 

Cracking — A crack is a linear fracture that may extend partially or completely through the con- 
crete member. When recording cracks, the inspector should describe the type, width, depth, length, 
direction, location and appearance of the crack as appropriate for the inspection. 

Delamination — Delamination occurs when layers of concrete .separate at or near the level of 
the top or outermost layer of reinforcing steel. The major cause of delamination is expansion of cor- 
roding reinforcing steel. Delaminated areas can generally be identified by a hollow sound when 
tapped with a hammer. 

Efflorescence — Efflorescence is a white deposit on concrete cau.sed by crystallization of solu- 
ble salts (calcium chloride) brought to the surface by moisture in the concrete. 

Honeycombs — Honeycombs are hollow spaces or voids that may be present within the con- 
crete. Honeycombs are caused by improper consolidation dunng construction, resulting in the segre- 
gation of the coarse aggregates from the fine aggregates and cement paste. 

Pop-Outs — Pop-outs are conical fragments that break out of the surface of the concrete leav- 
ing small holes. Generally, a shattered aggregate particle will be found at the bottom of the hole, with 
a part of the fragment still adhering to the small end of the pop-out cone. 

Scaling — Scaling is the gradual and continuing loss of surface mortar and aggregate over an 
area. When reporting scaling, the inspector should note the location of the defect, the size of the area, 
and the depth of penetration of the defect. 



32 Bulletin 760 — American Railway Engineering Association 



Spalling — A spall is a roughly circular or oval depression in the concrete. Spalls result from the 
separation and removal of a portion of the surface concrete, revealing a fracture roughly parallel to 
the surface. Spalls can be caused by corroding reinforcement and friction from thermal movement. 
Reinforcing steel is often exposed. When reporting spalls, the inspector should note the location of 
the defect, the size of the area, and the depth of the defect. 

C21.3,5.2 Site and Crossing — Commentary 

The inspector should note any changes in the alignment of a waterway both upstream and 
downstream and the resulting effect that they may have on the structure. A major change in the align- 
ment of a waterway may place it outside the spans intended for the crossing. 

Sedimentation deposits may fill scour holes after high water events. Underwater investigations 
may be required as per €21. 3. 5. 5. 

Structures located downstream of spillways or locks may be subject to increased scour potential. 

C21.3.5.3 Foundations, Piers and Abutments — Commentary 

Concrete and masonry structures are placed on foundations of earth, piling, cribbing, rock or 
other similar material. Cracks may be evidence of settlement which has occurred during consolida- 
tion of the foundation. Settlement may occur without cracking. Noticeable changes in track surfaces 
and alignment, plumbness or elevation may indicate foundation settlement. Changes in backwall 
alignment or cracks in the earth embankment parallel to the backwall may indicate movement. 
Constant wetting may indicate swelling, premature loss of mortar, deterioration of facing or exces- 
sive water pressure behind backwalls. Exposure of timber mats or untreated timber piling may lead 
to rapid deterioration of the timber. 

C21.3.5.5 Underwater Inspections — Commentary 

In evaluating the need for an underwater inspection, consideration should be given to type and 
depth of foundation, depth of water, normal and peak flow rates, nature of channel bottom and sus- 
ceptibility to and history of scour, type of aquatic environment, typical extent of drift and ice accu- 
mulation, and amount and type of watercraft traffic. The inspections should be performed with suf- 
ficient frequency to provide early detection of any detrimental conditions, and between inspections, 
the measuring of water depths should be considered to monitor channel bottom activity. In the event 
of a high water and/or flow occurrence, an excessive accumulation of ice or drift, a watercraft colli- 
sion, a significant change in channel bottom configuration, or any submerged component movement, 
consideration should be given to performing an emergency inspection as soon as conditions will 
safely permit. 

C2 1.3.5.6 Retaining Walls — Commentary 

In addition to structural deficiencies, retaining wall failures may result from: 

(1) Softening of the supporting material by moisture. 

(2) Overloading of the embankment behind the wall. 

(3) Scour or erosion beneath the foundation. 

(4) Expansive backfills. 

(5) Hydrostatic pressure behind wall. 

Cracks in the earth embankment which parallel the wall may be signs of wall movement. 

C21.3.5.7 Slabs and Beams — Commentary 

b) Transverse cracks in the bottom of simple span slabs and beams can indicate overload, par- 
ticularly if cracks open and close during passage of a train. Hairline cracks on the tops of simple span 



Proposed Manual Changes 33 



prestressed beams are generally due to shrinkage of the concrete. Hairline cracks in the top or bot- 
tom of simple span reinforced concrete slabs and beams are generally not significant. Diagonal 
cracks running up the sides of the slab or beam from near the supports may indicate excessive shear 
stress in the member or the beginning of shear failure. 

Transverse cracks in the top of continuous beams over support locations or in the bottom of 
continuous beams within the span can indicate overload. 

c) Sagging or excess deflection may indicate a loss of prestress. Loss of prestress may be 
caused by strand slippage, which may be visible at the ends of beams. 

d) End spalling can lead to a loss of bond in the prestressing tendons. Note any deterioration 
that has exposed or damaged prestressing tendons. 

C21. 3.5.8 Box Girders — Commentary 

b) Horizontal or vertical cracks in the top of girder ends are probably due to stresses created at 
the transfer of prestressing forces. Flexural cracks in the lower portion of the girders, particularly at 
mid-span, may indicate a problem resulting from overload or loss of prestress. 

c) Individual girder deflection under live load may indicate shear keys between boxes have 
been broken and that boxes are acting independently of each other. 

C21.3.5.9 Arches — Commentary 

a) A true arch has an elliptical shape and functions in a state of pure compression. Many arches 
are not elliptical and resist loads by a combination of axial compression and bending moment. 

c) Changes in alignment, sags in the arch crown, bulges in the sidewalls, transverse and longi- 
tudinal cracks and expansion joint failures may be signs of settlement, overload or impending arch 
failure. 

d) The area between the arches and the deck is called the spandrel. Open spandrel concrete 
arches receive traffic loads through spandrel bents which support a slab or tee beam floor system. 
Horizontal cracks in spandrel columns within several feet of the arch indicate excessive bending in 
the column, which may be caused by overloads and differential arch rib deflection. 

g) The spandrel area in closed spandrel arches is typically occupied by fill retained by vertical 
walls. Surface water should drain properly and not penetrate the fill material. 

C21.3.5.11 Culverts — Commentary 

a) Horizontal alignment of a culvert can be inspected by sighting along one of the culvert walls. 
Sag in the culvert axis may be identified by a location of sediment build-up on the culvert floor. 
Spalls or cracking in the vicinity of a joint may be a sign of movement at the joint. Both longitudi- 
nal and transverse cracking may be an indication of differential settlement. Longitudinal cracks can 
also be caused by a structural overloading of the culvert. 

b) Insufficient hydraulic capacity, either by design or due to obstructions, may cause upstream 
ponding and lateral flow movements which can erode the embankments and supporting material 
around the culvert end treatments. Culverts often convey short-term, high volume flows, and conse- 
quently, all culverts should be carefully inspected for scour and undermining. Tipping, cracking or 
separation of the headwalls, wingwalls or apron may indicate the presence of undermining. For arch 
and frame type culverts with earthen floors, undermining beneath the wall foundations along their 
full length should also be investigated. 




Recognized throughout the industry. . . 
For all the wood products and service you depend on. 



Companies like yours — whether 
Class I, regional, short line or transit 
railroads or contractors — have been 
ordering wood products from 
Burke-Parsons-Bowlby for more than 
30 years. That's a good sign BPB 
delivers the quality products and 
flexible service you can depend on. 
BPB offers the broadest wood 
product lines in the industry, in- 
cluding: crossties; switch ties; tie 
plugs; highway grade crossings; 



bridge timbers; paneled bridge 
decks; and more. 

With plants located in Virginia, 
Pennsylvania, West Virginia and 
Kentucky, BPB has the capacity to 
fill your daily and emergency product 
needs. Our own trucking subsidiary. 
Timber Trucking, allows us to deliver 
your order right on schedule. Today 
we're providing all the wood prod- 
ucts you depend on and developing 
the ones you'll need for tomorrow. 



Quality products, service, inno- 
vation, our commitment to you and 
the future of railroading. 

Get BPB's wood products and 
prompt delivery on track for your 
operation. Call 1-800-BPB-TIES 
(1-800-272-8437). 



Bt 



h 



The Burke-Farsons-Bowlby 
Corporation 

P.O. Box 231, Ripley, WV 25271 
(304) 372-2211 



34 



Proposed Manual Changes 35 



Part 24— Drilled Shaft Foundations 

Page 8-24-2. Insert revised Part 24, Drilled Shaft Foundations, as follows. 

24.1 General 

24.1.2 Scope 

This specification covers the description and general aspects of design, installation, inspection 
and testing of drilled shafts, also frequently referred to as drilled caissons, drilled piers, or bored piles. 

This specification is intended to serve as guidelines in developing specific designs and con- 
struction specifications on a project basis. 

For the purpo.se of this specification, the minimum diameter of these units shall be 30 inches 
(0.75 meter). Drilled shafts with smaller diameters have been constructed, but are not included in this 
specification. 

This specification relates primarily to single vertical drilled shafts. 

Factors to be used in modifying the capacities of single drilled shafts for determination of the 
capacity of a group of drilled shafts which support a common rigid cap are included elsewhere in this 
specification. 

The use of battered drilled shafts to accommodate lateral loads by the horizontal component of 
the shaft axial resistance is not recommended and is not addressed by this .specification. Lateral loads 
applied to drilled shafts are to be resisted by the effect of soil/rock interaction between the shaft and 
ground. 

24.1.2 Purpose and Necessity 

Drilled shafts are used to transmit loads through soils of poor bearing capacity into rock or soil 
formations having adequate bearing capacity. Generally, single drilled shafts have load capacities 
much larger than piling due to their larger size and capability of belling to increase the bearing area 
without enlarging either the footing or the drilled shaft. 

The selection of foundation treatment for a given site should be determined by subsurface con- 
ditions, and by economic considerations as there is often a choice of several types of foundations for 
a structure. 

24.1.3 Definitions (see Figure I) 

Drilled Shaft — A machine and/or hand excavated shaft, concrete filled, with or without .steel 
reinforcing, for the purpose of transferring structural loads to bearing .strata below the structure. 

Protective Casing — Protective steel unit, usually cylindrical in shape lowered into the excava- 
tion to protect workmen and inspectors from collapse or cave-in of the side wall. 

Bell or Underream — An enlargement at the bottom of the drilled shaft made by hand excava- 
tion or mechanical underreaming with drilling equipment for the purpose of spreading the load over 
a larger area. 

Socket — A shaft of equal or smaller diameter extended into the bearing material. 

Toe — Vertical section at bottom of bell. 

Permanent Casing — A steel cylinder that is installed for the purpose of excluding soil and 
water from the excavations. It is used as a form to contain concrete placed for the drilled shaft and 
remains in place. 



Rd'crcncc. Vol. »5. iy84. p.2y. 

Latest page con.sisl: I to X. incl. ( iyX4). 



36 



Bulletin 760 — American Railway Engineering Association 



NOMINAL SHAFT DIAMETER (B 



I J 1 r] 



CASING 



60* MIN 



-REINFORCING 
STEEL 




BELL DIAMETER 




SOCKET DIAMETER 



OPTIONAL 
SOCKET 



1. FIGURE IS FOR ILLUSTRATIVE 
DEFINITION OF DRILLED SHAFT 
NOMENCLATURE 

2 THE NEED FOR OR EXTENT OF CASING 
IS DEPENDENT ON LOCAL SOIL AND 
GROUND WATER CONDITIONS. 

3. THE NEED FOR AND EXTENT OF REIN- 
FORCING STEEL SHALL BE AS SPECIFIED 
BY THE DESIGN ENGINEER. 



Figure 1 
Drilled Shaft 



Temporary Casing— A steel cylinder that is installed for the purpose of excluding soil and 
water from the excavations. It may also be used as a form for the shaft concrete but is withdrawn as 
the shaft concrete is placed. 

24.1.4 Design Loads 

Loading for drilled shafts shall be the design loads from the supported structure without applica- 
tion of load factors used for Load Factor design procedure. Design loads shall include the following: 



Proposed Manual Changes 37 



Primary Forces: 

Dead Load 

Live Load 

Centrifugal Force 

Earth Pressure 

Buoyancy 

Negative Soil Friction 

Secondary Forces (Occasional): 
Wind and Other Lateral Forces 
Ice and Stream Flow 
Longitudinal Forces 
Seismic Forces 

When drilled shaft foundations are designed for both primary and secondary forces, the allowable 
load on the drilled shafts may be increased by 23 percent, provided that the size or number of drilled 
shafts is not less than that required for primary forces alone. In soils where downward movements of 
surrounding soil relative to the drilled shaft are expected to occur, axial loads shall include negative soil 
friction forces, acting downward on the drilled shaft. Under special conditions swelling soils can pro- 
duce upward forces, with fluctuation of the water table, which should also be considered in design. 

24.2 Information Required 

24.2.1 Field Survey 

Sufficient information shall be furnished in the form of profile and cross sections to determine 
general design and structural requirements. The location of overhead and underground utilities, exist- 
ing foundations, roads, tracks, or other structures shall be indicated. Records pertaining to high and 
low water levels and depth of scour shall be provided for stream crossings. 

24.2.2 Subsurface Investigation 

Foundation material shall be investigated as specified under Chapter 8, Part 22, Geotechincal 
Subsurface Investigation, in order to determine soil or rock properties, ground water elevations, and 
any other pertinent conditions. 

Where a large portion of the required shaft capacity is to be generated from tip resistance of the 
shaft or rock .socket, the geotechnical investigation shall be of sufficient scope to permit the deter- 
mination that the .strata in which the tip is founded is of sufficient depth and .strength to carry the loads 
to which it is subjected. 

Reference is also made to Chapter 8, Part 4.3.1, for additional information. 

24.3 Design 

24.3.1 General 

The design is divided into three basic parts: ( 1 ) transfer of load from the drilled shaft to the rock 
and/or soil bearing strata; (2) the drilled shaft itself; and (3) the connection between the supported 
structure and the drilled shaft. 

24.3.2 The transfer of Load from the Drilled Shaft to the Rock or Soil Bearing strata. 

24.3.2.1 Drilled shafts transfer load to the bearing strata as follows: 

a. An end bearing-type drilled shaft transfers essentially all of its load through weaker soils to 
a layer of soil or rock with adequate bearing capacity. 

b. A friction-type shaft whereby the drilled shaft load is transferred to the surrounding material 
primarily through friction between the shaft wall and the adjacent material. 



38 Bulletin 760 — American Railway Engineering Association 



c. A combination end bearing and friction-type drilled shaft is a shaft in which some of the load 
is transferred into the stratum by soil friction and the remainder by end bearing. 

24.3.2.2 Lateral Loads and Moinent: When the drilled shaft is subjected to lateral load and 
moments, as well as axial load, the distribution of soil pressures and the variation of moments and 
shear in the shaft must be determined. 

24.3.2.3 Belled Shafts: Where the bearing strata has insufficient strength to support the load on the 
ba.se of the shaft, the shaft bottom may be enlarged by belling or underreaming to reduce the pressure by 
distributing the load over a greater area. Belled shafts shall be used only where the soil/rock in which the 
bell is placed will not collapse due to the underreaming. Bells are normally unreinforced. The base diam- 
eter of the bell shall not exceed three times the shaft diameter and the sides shall not be less than 60 
degrees from the horizontal. The toe height of bottom edge shall not be less than 6 inches (150 mm). 

24.3.2.4 The uhimate axial capacity of a drilled shaft (Q^,,) shall be based on the summation of 
the ultimate shaft tip capacity and ultimate side resistance capacity minus the weight of the shaft. The 
allowable shaft capacity shall be the ultimate capacity divided by a factor of safety. 

The ultimate shaft tip capacity (Q,) shall be Q^ = qr • A^, where qr is the ultimate unit soil tip 
resistance determined by geotechnical analysis and A^ is the area of the shaft tip. 

The ultimate side resistance (Q,) of the shaft shall be equal to the circumference of the shaft 
multiplied by the embedment length in a soil layer of uniform unit side resistance (qs) multiplied by 
qs. The value(s) of qs shall be determined by geotechnical analysis. Where a shaft passes through 
stratified soil having different values of qs for the various soil type layers, the value of Qs shall be the 
shaft circumference multiplied by the summation of various q^ values multiplied by the depth of the 
respective layer. In general, the top five feet of an embedded shaft and a bottom length equal to the 
diameter of the shaft tip or perimeter of the bell shall be considered as non-contributing to the side 
resistance of the shaft. Where the drilled shaft is located in scour susceptible areas, the probable depth 
of scour shall also be deducted when calculating the ultimate shaft side resistance. 

Where rock sockets having a diameter equal to or less than the nominal diameter of the shaft are 
used, the ultimate tip capacity of the shaft shall be equal to the area of the socket tip multiplied by the 
uniaxial ultimate unit rock capacity. The ultimate socket side resistance shall be the product of the socket 
circumference, socket embedment and ultimate unit side shear resistance along the socket/rock interface. 

Unless an analysis is used which accounts for the load/deflection relationship of the various soil 
or rock strata encountered, the ultimate capacity of a drilled shaft which utilizes a rock socket shall 
be based on the sum of the ultimate tip and side resistance capacities of the rock socket only, neglect- 
ing nominal capacities of the shaft in the soil overburden. 

24.3.2.5 Uplift Capacity: The ultimate uplift capacity of a drilled shaft shall be equal to or less 
than the weight of the shaft plus 0.7 times the ultimate side resistance of the shaft. If belled, the uplift 
capacity of the shaft may be increased by taking into consideration the reinforcement details of the 
shaft and bell together with the strength characteristics of the adjacent material. 

24.3.2.6 Factors of Safety: For drilled shafts in soil or socketed in rock, a minimum design fac- 
tor of safety of 2.5 shall be used against bearing capacity failure. A factor of safety of 2.5 shall be 
used when designing for conditions which produce uplift. 

24.3.2.7 Shafts Under Water: Wherever practicable, the drilled shaft shall be designed to per- 
mit the placing of the concrete in the dry, and for visual inspection of the hole, the bearing strata, and 
the rock socket. 

When it is impractical to dewater the excavation for rock-socketed shafts, the concrete may be 
placed under water by means of a tremie or pumped concrete and appropriate allowances made in the 
concrete mix design. The water level shall have reached a static condition before concrete placement 
begins. 



Proposed Manual Changes 39 



When concrete cannot be placed in the dry and a thorough visual inspection cannot be made by 
television or by divers, the Design Engineer shall reduce the allowable bearing and side resistance 
stress appropriately. 

Any free water in belled shafts shall be removed by pumping or bailing, and the bottom rein- 
spected before placing concrete in the dry. 

24.3.3 The Drilled Shaft 

24.3.3.1 The drilled shaft is generally designed as a short column for axial loads due to the lat- 
eral support provided by the soil/rock. In muck or water, slenderness effects of the column must be 
taken into consideration. 

When the drilled shaft is subjected to moment and lateral forces at the connection to the sup- 
ported structure, the shaft must be designed for bending and shear in addition to axial force. Moment 
and shear along the length of the shaft must be calculated, and adequate reinforcement provided. 

24.3.3.2 The shaft shall satisfy the design requirements of part 2 of this Chapter. 

24.3.4 Connection Between Supported Structure and Drilled Shaft 

The connection between the drilled shaft and the supported structure (parts above the top of 
shaft) shall be capable of transferring the design loads, including direct load, shear and moment. This 
can be accomplished by the following means: 

a. When the supported structure at the top of shaft is of concrete, the reinforcing steel cage shall 
be extended into the cap so that the load is transferred into the reinforcing steel of the drilled shaft 
by bond and into the concrete by compression. 

b. When the cap section is a steel element, appropriate design shall be developed to transmit all 
loads, conforming to the requirements of Chapter 15, Part 1 or 2. 

24.3.5 Group Action of Drilled Shafts 

Evaluation of group shaft capacity assumes the effects of negative soil friction (if any) are neg- 
ligible. 

24.3.5.1 Cohesive Soil: Evaluation of group capacity of shafts in cohesive soil shall consider 
the presence and contact of a cap with the ground surface and the spacing between adjacent shafts. 

If the cap is not in firm contact with the ground, or if the soil at the surface is loose or soft, the 
individual capacity of each shaft having a diameter B should be reduced to a reduction factor times 
Qt for an isolated shaft. This factor equals 0.67 for a center-to-center (CTC) spacing of 3B and 1.0 
for a CTC spacing of 68. For intermediate spacings, the reduction factor may be determined by lin- 
ear interpolation. The group capacity may then be computed as the lesser of: 

1 ) the sum of the modified individual capacities of each shaft in the group, or 

2) the capacity of an equivalent pier defined in the perimeter area of the group. 

For a shaft group with a cap in firm contact with the ground Q„„ may be computed as the lesser 
of: 

1) the sum of the individual capacities of each shaft in the group or 

2) the capacity of an equivalent pier as described above. 

For the equivalent pier, the shear strength of soil shall not be reduced by any factor to deter- 
mine the Q, component of Q^,, the total base area of the equivalent pier shall be used to determine the 
Q, component of Q„|, and the additional capacity of the cap shall be ignored. 

24.3.5.2 Cohesionless Soil: Evaluation of group capacity of shafts in cohesionless .soil .shall 
consider the spacing between adjacent shafts. Regardless of cap contract with the ground, the indi- 



40 Bulletin 760 — American Railway Engineering Association 



vidual capacity of each shaft should be reduced to a reduction factor times Qj for an isolated shaft. 
This factor equals 0.67 or a center-to-center (CTC) spacing of 3B and 1.0 for a CTC spacing of 88. 
For intermediate spacings, the reduction factor may be determined by linear interpolation. The group 
capacity may be computed as the lesser of: 

1) the sum of the modified individual capacities of each shaft in the group, or 

2) the capacity of an equivalent pier circumscribing the group, including resistance over the 
entire perimeter and base areas. 

24.3.5.3 Group in Strong Soil Overlying Weaker Soil: If a group of shafts which is embedded 
in a strong soil deposit overlies a weaker deposit (cohesionless and cohesive soil), consideration shall 
be given to the potential for a punching failure of the tip into the weaker soil strata. 

If the underlying soil unit is a weaker cohesive soil strata, careful consideration shall be given 
to the potential for large settlements in the weaker layer. 

24.4 Material 

24.4.1 Concrete 

Unless otherwise stipulated in this specification, concrete shall be produced and place in accor- 
dance with Part 1 of this chapter. Concrete shall have a minimum compressive strength of 3,000 psi 
(21 megapascals) in 28 days. Approved additives, such as set retarders, may be used to improve 
workability. Slump at time of placement shall be not be less than 4 inches (100 mm), and not more 
than 6 inches ( 1 50 mm). If temporary casing is to be used, the slump should be not less than 5 inches 
(125 mm), and a set retarder may be necessary. 

24.4.2 Reinforcing Steel 

Unless otherwise stipulated in this specification, any required reinforcing steel shall conform to 
the requirements of Part 1 of this chapter. 

24.4.3 Steel Casing Material 

If the steel casing is relied upon as a structural element, the steel casing material shall conform 
to the requirements of ASTM A252. 

24.5 Construction 

24.5.1 Contractor Qualifications 

Drilled shafts shall be installed by the Owner with experienced personnel, or by a Contractor 
or Subcontractor who specializes in such work. Availability of all required special equipment, tools, 
and experienced personnel are important items to be considered when determining Owner installa- 
tion or selecting an installation contractor. 

24.5.2 Shaft Excavation 

When excavating a drilled shaft, earth walls shall be adequately and securely protected against 
cave-in, subsidence and/or displacement of surrounding earth, and for the exclusion of ground water 
by means of temporary or permanent steel casings. 

Whenever personnel are required to enter the shaft, a protective casing shall be used and there 
shall be adequate provisions for fresh air, light and protection from falling objects and toxic gases. 
Operation of harmful gas-producing equipment in the shaft must be prohibited. 

Rock grapples or special tools for removal of boulders or other obstructions must be readily 
available for use. Blasting will be permitted only upon obtaining written approval from the Engineer. 

Inspection of the shaft base, and any socket, by a qualified inspector is highly recommended 
and should be omitted only with the approval of the Engineer. 



Proposed Manual Changes 4 1 



No shaft excavation shall be made within 15 feet (4.5 meters) of an uncased shaft filled with 
concrete that is less than one day old. The construction procedure used shall be approved by the 
Engineer in charge before commencing work. 

24.5.3 Casing 

Where called for, permanent steel casing shall be installed to the plan elevation or to the eleva- 
tion designated by the Engineer in the field. When the top of the drilled shaft is below the surface of 
the ground, installation of additional large diameter casing may be required to extend above the work- 
ing level to minimize possibility of foreign materials or water entering the top of the shaft. 

Casings shall be of adequate size and thickness to safely retain the adjacent earth materials and 
water form entering the shaft excavation, without exceeding allowable steel stresses, distortion, or 
collapse of the casing. 

A protective casing is also to be provided, where required, to serve as protection for personnel 
entering the shaft excavations not provided with casings as specified above. Casing size and thickness 
shall meet the requirements stated above. The outside diameter of the protective casing shall be as large 
as possible, yet small enough to be lowered and removed without damage to the sides of the shaft. 

If conditions are such that casing withdrawal will cause dislocation of the reinforcing steel or 
permit sloughing soils to enter the shaft, a double casing should be used. By this method, the shaft is 
drilled oversize and a temporary casing installed. A light gage permanent inner casing the same size 
as the required shaft diameter is then installed. This inner casing shall be of sufficient strength to 
serve as a form for the concrete shaft but need not be designed for soil pressure. Concrete is then 
placed within the permanent inner casing. After the concrete has set, the annular space between the 
permanent casing and surrounding soil is filled with grout, lean concrete, sand or by other approved 
method and the temporary outer casing is withdrawn. 

24.5.4 Bells or Underreams 

Before belling, the Engineer shall determine that the formation encountered at the plan eleva- 
tion is adequate. When shafts are required to be belled, the bells shall be formed either by hand or 
use of special belling equipment to the angle and slope called for on the drawings. The bottoms of 
bells shall be thoroughly cleaned of all loose materials and inspected before the concrete is placed. 

24.5.5 Sockets 

When sockets are required, they shall be formed by machine or by hand to the proper size and 
depth called for in the plans. Sides and bottom of sockets must be thoroughly cleaned of all loose 
material since the bond of the concrete to the socket sides is used in design. 

24.5.6 Tolerances 

The center of the top of each shaft shall not vary form its design location by more than '/:4 of 
the shaft diameter, or 3 inches (75 mm), whichever is less, and the shaft shall not be out of plumb by 
more than 1.5 percent of the length not exceeding 12.5 percent of shaft diameter, whichever is less. 

24.5.7 Dewatering 

Suitable dewatering procedures shall be as agreed upon between the Engineer and Contractor 
as determined at such time as conditions warrant. Unless otherwi.se agreed, the shaft at the time of 
placement of .steel reinforcing cage, if any, and concrete shall be essentially free of standing water in 
excess of two inches (50 mm) deep. 

24.5.8 Inspection 

Immediately prior to placement of any required reinforcement or concrete, each shaft shall be 
thoroughly inspected as directed by the Engineer to ascertain that the shaft has been properly pre- 
pared, that the bearing material is compatible with design requirements, and whether additional 



42 Bulletin 760 — -American Railway Engineering Association 



investigation of the bottom is required. If conditions vary from the assumed conditions determined 
by the borings, additional investigation shall be conducted as directed by the Engineer 

24.5.9 Placing Steel 

When reinforcing steel is specified, it shall be prefabricated and placed as a unit immediately 
prior to concrete operations. In order to minimize displacement of reinforcing steel cage when cas- 
ing is pulled, the cage may be reinforced below the zone of significant bending moment by welding 
horizontal bands to the cage at about five-foot (one and one-half meters) intervals. 

24.5.10 Placing Concrete 

Dry Hole — Prevent segregation of concrete through use of tube, sectionalized pipe or other 
means to direct the free fall of concrete so that it does not strike the sides of reinforcement in the shaft. 

Under Water — Utilize a tremie or pumped concrete in accordance with Chapter 8, Part 1, 
Article 1.14.10 and Part 24, Article 24.3.2.7. 

Rodding or mechanical vibrating is necessary only for the top five feet (one and one-half 
meters) of shaft. Any special requirements for concrete placement shall be approved by the Engineer. 

24.5.11 Casing Removal 

In situations where temporary casing is to be removed, the head of concrete inside the casing 
must be adequate to preclude infiltration of water and sluffage of the shaft face and sides. 

Elapsed time from beginning of concrete placement in cased portion of shaft, until extraction 
of casing is begun, shall not exceed one hour. 

Extreme care shall be taken when a casing is removed to prevent subsidence of the surround- 
ing ground if this condition is critical due to the presence of surrounding structures or utilities. 

Elevation of top of the steel cage should be carefully checked before and after casing extrac- 
tion. The top of the concrete shall not raise during extraction of the casing. 

The exterior temporary casing, if a double -cased shaft, shall not be removed until three (3) days 
after the shaft is poured. 

24.5.12 Continuity of Work 

Drilled shaft construction work shall be planned so that all required operations proceed in a 
continuous manner until the shaft is complete. A precise time schedule agreement between the 
Contractor and the Engineer should be established. Provision shall be made for protecting the shaft 
and adjacent construction in case of unforeseen interruptions. Such provisions shall be approved by 
the Engineer before drilling begins. 

24.5.13 Records 

An accurate record shall be kept of each drilled shaft as installed. The record shall show the top 
and bottom elevations, shaft and bell diameters, depths of test holes if required, date the shaft is exca- 
vated, inspection report of the bearing stratum, depth of water in excavation at time of placing steel 
and concrete, quantity of concrete placed compared with theoretical quantity, and any other pertinent 
data. Records shall be made and signed by both the project superintendent and inspector and distrib- 
uted to proper authorities daily. 

24.6 Testing 

Materials used in construction of drilled shafts should be sampled and tested as specified else- 
where in Chapter 8. At least two (2) concrete test cylinders shall be taken for each shaft. 

Further testing of the shafts may be required by the Engineer in order to determine the quality 
of the concrete by coring of the bearing capacity of the shaft, by test loading. 




Connect with 

Parker for 
all your 
hose and 

fittings needs. 



For your air brake applications, fuel 
and oil lines or hydraulic system 
connections, Parker has the hose 
and fittings you need. 

Including skive-type hose and fittings 
or no-skive hose with permanent or 
reuseable fittings. 

Or our patented Parkrimp system 
that allows you to make 
permanent, factory- 
quality hose 
assemblies ir 
own facilities 
on the job site. 

Call or 
write today 
for your free 
copies of 
Parker 
railroad 

product bulletins. 
Parker Hannifin Corporation 
Hose Products Division, 
30240 Lakeland Blvd., 
Wickliffe, OH 44092. 
(216) 943-5700 
FAX (216) 943-3129 




Parker 



FluidConnectors 



43 



DANELLA RENTAL SYSTEMS, INC. 

Helping Railroaders and Transit Authorities 

NATIONWIDE! 



• Quality Equipment 

• Courteous Service 
> Reasonable Rates 




• Hy-Rail Available 

• Short Term 

• Long Term 

• Rent to Own 



We offer every type of late model light, medium or lieavy duty vehicle and 
construction equipment for railroads and transit authorities including: 

' Pickups • Rotary Dumps • Compressors • Dozers 

Boom Trucks • Tractors • Backhoes • On Track Cable Plow 

' Flatbeds • Low Boys • Front End Loaders • Trenchers 



DANELLA RENTAL SYSTEMS, INC. 

2290 Butler Pike 14101 East Moncrieff Place 

Plymouth Meeting, PA 19462 |i.__|^__ k|q/« anH QCMCA Aurora. CO 80011 

Phone (610) 828-6200 MemOer J^.J^^^ aij^ "an ''^""^ <^°^' 371-7799 

Fax (610) 828-2260 '(204)782-5456 Fax (303) 371-2677 




44 



Proposed Manual Changes 45 



Proposed 1997 Manual Revisions to 
Chapter 16 — Economics of Railway Engineering and Operations 

Part 4 — Railway Operations 

Page 16-4-1. Insert the following Revised Part 4. Railway Operation. 

Table Of Contents 
Section Page 

4. 1 Introduction 

4.2 Car Distribution 

4.3 Trains 

4.4 Train Management 

4.5 Communications 

4.6 Defect Detection 

4.7 Line Capacity 

4.8 Terminals 

4.1 Introduction 

The basic railway operation is to provide a transportation service by moving goods and people 
from one place to another This transportation service should be done safely and efficiently. 

Goods or products are hauled in freight cars. Some cars are specifically designed for particular 
commodities, such as tank cars for liquids, auto-rack cars for automobiles and trucks, and stack cars 
for containers. 

Passenger cars are designed to carry people, and include such variations as coaches, sleeping 
cars, diners and lounge cars. 

For efficiency, cars are assembled into freight or passenger trains so that coupled units of 
motive power can move many individual cars, as required to meet specific transportation needs. 

For a general description of railway operations, see "The Railroad: What It Is and What It 
Does," by John H. Armstrong, .^rd Edition. 1990, Simmons-Boardman Books, Inc., Omaha, NE. 

Also, federal regulations have operational and maintenance impacts on railways. See Title 49 
Code of Federal Regulations, Transportation. Of special interest is Subchapter C- Hazardous 
Materials Regulations, Parts 200-266. 

For information on motive power, see Part 3 — Power, of this Chapter 16 of the AREA Manual 
of Recommended Practices. 

Accordingly, this Part 4 on Railway Operations, focu.ses upon car distribution, train manage- 
ment, communications, defect detection, line capacity, terminals and economic considerations. 

4.2 Car Distribution 

In order to provide good service to shippers, car distribution and control are key elements. 
Modem technology has contributed developments which improve the process of moving cars, both 
empty and loaded, to their destination on time. 

A major technical development has been in the application of computers and communications 
in the field of car distribution. Mainly, this involves the use of computer work stations or PCs con- 
nected to main-frame computers with high-speed digital communications links. 

Major railroads are handling car distribution from customer .service centers, where customers 
make toll-free calls to order empty cars or ask for reports on loads. In many instances, these are com- 



46 Bulletin 760 — American Railway Engineering Association 



puter-to-computer transactions between customers and railroads. Also, railroads and customers con- 
duct transactions over the Internet. 

Car distribution is the process of moving empty cars to fill customers' orders. A major effort is 
to move empty cars the minimum distance to serve customers. 

Empty cars generally show up at interchange points and in areas where traffic is terminated. 
Hence, accurate reporting of loads and empties is essential in planning for moving empties to fill 
shippers' orders in an efficient manner. 

Some railroads schedule individual car movements. A car (whether empty or loaded) is only 
moved when its move is authorized by a movement order. Railroads that use this process typically 
have the customer service center handle this function, rather than local supervision. 

Also, some railroads assign personnel to handle distribution of special cars or cars that handle 
specific freight, such as covered hopper cars, auto rack cars or double-stack cars. 

4.2.1 Work Order Systems 

Some railroads are implementing work order systems in which computer work stations (may be 
lap top computers) in locomotive cabs are linked to a railroad's customer service center. Digital train- 
to-wayside radio and microwave or fiber optic cable provide the communication path between way- 
side radio stations and the customer service center. 

Customer service sends a work order to a train crew indicating cars to be picked up, set out, etc. 
The crews receive their orders on their on-board computer. They respond via the computer, record- 
ing their compliance. Thus, a written record is generated of orders and work accomplished. It is not 
unusual for an industry switch run to pull an empty car from one industry, reporting the pull, and 
receive an order to place the empty car at another industry for loading. Thus, real-time response 
enables the work order system to provide better customer service. 

4.2.2 Automatic Equipment Identification 

The most recent technological development to ensure good customer service is Automatic 
Equipment Identification. AEI provides accurate identification of freight cars and locomotives in 
space and time. 

As of January 1, 1995, freight cars operating in interchange service were to be equipped with 
AEI tags identifying owner and car number. Locomotives were also equipped with AEI tags by Class 
I and regional railroads. 

Railroads are installing wayside scanners to read AEI tags, especially bracketing major yards 
and at many interchange points. 

The AEI wayside scanner provides location, date, time and train direction for all rolling stock, 
allowing trains to be tracked in real-time. Train consist information is stored at scanner sites until after 
the last car of the train has passed the scanner. Then, the tag readings are transmitted to a central office 
where AEI information is used to update the railroad's car reporting data base. This information is peri- 
odically sent to the Association of American Railroads' Washington, DC headquarters for updating the 
AAR's car data base and for informing shippers and other railroads of car movement information. The 
information is available to the railroad's car distribution and/or customer service center. 

Loaded car locations can be reported to shippers and other railroads when interchange is involved. 
Empty car locations can be provided to customer service centers for car distribution planning. 

Some railroads are installing AEI wayside scanners at hot bearing detector locations enabling 
car initial and number to be broadcast on detection of an overheated bearing. 

Communications links are vital for sending AEI data in real time to a railroad's headquarters 
for input to the car reporting data base. Thus, timely information on car location can be automatically 



Proposed Manual Changes 47 



made available. This is especially helpful if AEI scanners are located on the departure side of yards 
to accurately report on cars leaving yards. 

Additionally, AEI tags may contain variable information, such as temperatures of refrigerator 
cars or data indicating health conditions of locomotives. The AEI scanner can read such status of cars 
or locomotives and send the information to headquarters, which in turn can notify the train crew and 
yard forces of any problem. With variable information tag capability, equipment must be added to the 
normally passive tag to program it with this additional data. 

Variable programmable message tags might also include information concerning hazardous 
material carried in a car. 

4.2.2.1 AEI Economics 

Costs associated with installing AEI tags and scanners (readers) include: 

• Initial purchase price and in.stallation. 

• Maintenance of scanners, 

• Tag maintenance, if damaged. For variable message tags, cost of programming and periodic 
testing. 

• Communications for storage and transmission of AEI data should include equipment and .soft- 
ware purcha,se and installation costs, as well as maintenance costs. 

In the case of AEI, tags are usually a mechanical department responsibility. Scanners and com- 
munications equipment are typically a communications department responsibility. 

4.3 Trains 

Trains are a series of cars coupled together and hauled by locomotives. 

4.3.1 Train Consist 

Freight trains are usually made up for specific purposes depending upon the type of service to 
be provided. 

For example, intermodal trains carrying trailers or containers are operated by a .schedule, and 
typically have enough motive power to move the train and meet the schedule by operating at track 
speed. On main lines, these trains usually operate in traffic control system territory where movement 
authority is governed by signal indications. 

Regular freight trains may operate on a specific schedule or on a more general schedule, but 
one which allows trains to make connections. Often, over-the-road speed is not as critical for these 
trains as for intermodal trains. Under current operating agreements, these trains may make set offs 
and pick ups of blocks of cars at yards or at sidings on line. 

Unit trains handle only one commodity, such as coal, grain, or produce, operating between a 
single origin point and a single destination point. Depending upon the commodity, they may operate 
as an intermodal train or on a more relaxed .schedule, as used in hauling coal. Cars for the.se unit trains 
are often unique and may be operated only in specific trains. Unit coal trains usually have the same 
cars operating both loaded and empty. Trains carrying automobile parts and completely assembled 
motor vehicles, although operated as a unit train, will not always have the same cars in each train. 

Local trains operate on main lines, and also on branch lines. Sometimes they operate in one 
direction on one day, lay over at a terminal and return the next day. Some operate daily or on .selected 
days, making a complete round tnp over a portion of a line in one day. Industry runs are sometimes 
called locals, but they usually operate in a metropolitan or terminal area, returning each day to their 
originating terminal or yard. 



4S Bulletin 760 — American Railway Engineering Association 



Work trains are usually operated as extra trains, as required, to serve a specific purpose, such 
as distributing ballast, rail, tics, etc. Or they may be used to haul company material, as well as 
maintenance-of-way house trailers mounted on flat cars, between locations for gangs. 

4.3.1.1 Passenger Trains 

Passenger trains, especially for long distance rail operations, tend to operate in fixed consists of 
numbers and types of cars and locomotives. For long distance service, 20 cars is a typical maximum 
train length. 

Factors to be considered in passenger train make up include: 

• Adequate capacity to provide good service to meet customer requirements, e.g., a comfort- 
able seated ride. 

• Motive power sufficient to provide on-time service to meet .schedules on a consi.stent basis. 

• Train length short enough to require only one stop at station platforms. 

Commuter rail service is substantially different from long distance service. Many of the oper- 
ating parameters are neariy opposite, if an efficient and effective service is desired. 

A pattern of relatively short trains that serve a particular zone, providing fast service to the cen- 
tral business district, and then operate in fast non-revenue service back for a second revenue trip, pro- 
vides far more efficient and effective service than the traditional all stop local trains. 

Commuter service requires a much higher level of on-time performance than either long dis- 
tance passenger or freight service. This on-time performance concept must be included in every 
aspect of design and operations. A 10 or 15 minute delay on a long distance train, although annoy- 
ing, will not cause the customer to use a competitive mode of transportation. A pattern of even minor 
delays will be sufficient reason for a commuter to return to his automobile. Commuter rail service 
requires exacting schedule development and precise operation. 

4.3.2 Freight Train Length 

There is no optimum train length, i.e., that will meet all requirements under all conditions. 
Generally, costs tend to favor long trains with minimum motive power. However, to provide good 
customer service, shorter, faster trains are often required. A railroad's goal is to achieve operational 
balance; to meet customer service requirements at minimum cost. 

Train length is dictated by the length of passing sidings and by dynamic forces while handling 
trains in hilly territory. 

Increasing train length to handle tonnage has both advantages and disadvantages. It is possible 
to estimate the costs and benefits of operating different lengths of trains, arriving at an optimum 
length of train for a particular movement. This cost/benefit analysis of train length is greatly aided by 
computer simulation techniques. Train movement simulation should take into account track, sidings, 
grades, curves and other physical operating characteristics, such as traffic density, tunnel locations, 
destination congestion, etc. 

The primary objective of a railroad in private industry is to maximize profit. However, longer 
trains may work against the maximizing of profit, because they tend to reduce the quality of service 
and increase shippers' total distribution costs. The traditional bias toward longer trains may cause 
railroads to lower prices and reduce the quality of service. The result may be to reduce revenue 
through lower prices to retain a given amount of traffic, or moving less traffic at the same price. 

4.3.2.1 Equipment Limitations 

Car equipment can be a limitation to long mixed trains, particulariy with regard to buff forces 
and drawbar strength. 



Proposed Manual Changes 49 



Slack action occurs partly because of an undulating profile of the rail line. Changes in grade can 
be critical due to the impact forces set up between cars through the couplers, inducing a change of 
speed. This is most noticeable with long trains operating over a line having an undulating profile, 
even when the grades are not severe or limiting. 

Air brake operation can be a limitation. It is usually necessary to limit the speed of long, heavy 
trains, because of greatly increa.sed stopping distances and inadequate signal spacing. In cold weather, 
it is often difficult to get the required air pressure gradient as called for by federal regulations. Recent 
developments (circa 1996) in electronic braking systems could solve these problems. 

It is usual to limit the number of units of motive power on a train by specifying the number of 
powered axles that can be used. This is a function of the coupler .strength of the equipment. 

There are .some elements to consider when several diesel-electric locomotive units are coupled 
together and controlled from the lead unit. Care in operation is required in limiting the number of 
units to prevent jacknifing of the train when braking. There is .some evidence that excessive dynamic 
brake applications on heavy trains may damage the track structure. 

4.3.2.2 Remote Control Power 

The development of control systems for the remote operation of locomotive units has made it 
possible to more effectively place power at other locations in a train consist in addition to the head end. 

Remote locomotives are automatically actuated by a radio system in response to control input 
by the engineman in the lead locomotive unit. Effective train operation can be enhanced if remote 
control power is used to mitigate the difference in speed between relatively heavy tonnage bulk trains 
and high priority trains. 

However, use of remote control power creates added difficulties in terminals, especially in auto- 
mated yards, to switch power out of the middle of trains, service the remote control units, and put 
them into other trains. 

The following factors should be analyzed when considering remote control power utilization: 

• Cutting in remote control power (slave units) at the originating yard is a co.st item as is the 
removal of slave units at the terminating yard. 

• Using remote locomotives to increa.se train length may increase line capacity. 

• Stopping distances decrease through the use of remote control power, as braking can be ini- 
tiated at more than ju.st from the head end of the train. 

• Improved train performance and control from the intermediate spacing of power is especially 
important on heavy grades. 

• Reduction of crew costs due to the reduction of the number of trains, as enginemen are not 
needed on remote control power (slave units). 

• Control of slave units in mountainous territory, especially where there are numerous tunnels, 
may be discontinuous. 

4.3.2.2 Operating and Plant Limitations Terminal Yards: Car inspection forces are staffed to 
handle normal trains within a reasonable period of time. 

Charging trainlines is done in one of three ways: (1) by ground airlines; (2) by road engine crews; 
or (3) by yard engine crews. Where ground lines are available, charging time, and therefore cost is 
somewhat proportional to cut or train length. When the use of a road or yard engine is required, the cost 
of terminal time to the road crew must be considered, as well as lost switching service of the yard crew. 

Terminal air tests are mandatory. Using ground air has numerous advantages, especially if used in 
conjunction with pre-testing. Ground air reduces crew costs and initial terminal delay, but it also reduces 



50 Bulletin 760 — American Railway Engineering Association 



locomotive and car detention costs. With pre-testing to better balance car inspector workloads, overtime 
and perhaps car inspector positions can be reduced as compared to charging with locomotives. 

Doubling in or out of a terminal or a yard would appear to be a most important factor in deter- 
mining optimum train length. In determining the cost of such doubling, where not limited by high- 
way-rail grade crossings, consideration should be given to: 

• Additional wage expense of the road crew if they perform this particular work. 

• Delays to yard or terminal switching to and by other crews. 

• The net income effect of these costs would then have to be weighed against the cost of elim- 
inating such doubling train movements. 

Yard switching may be adversely affected by long trains blocking leads through the necessity 
of having to double into or out of receiving and departure tracks. 

A train of twice the length of another train may require twice as much time to be pulled by a 
given point for inspection. 

Fueling and Serx'icing Locomotives: Most railroads are putting an average of 1200 to 1 500 gal- 
lons into fuel tanks capable of holding more than twice that amount, because many other factors 
besides train length affect the need to fuel. 

Locomotives are placed on service tracks for service, for minor repairs, for turning and consist 
building, as well as fueling. Once on the service track, the increased cost of fueling is almost nil. At 
a fueling rate of about 250 gallons per minute, the actual time to fuel is not as significant as the ben- 
efits of reducing the frequency of fueling. 

Blocking: When a train picks up cars at an intermediate terminal, these cars must be arranged 
"out-of-block" as a unit to be dropped at a subsequent terminal for proper blocking for later train, or 
they may, at the cost of delay, be cut into the proper blocks in the first train. 

On lines where weight restrictions exist over some bridges, a heavy car will have to be imme- 
diately preceded and followed by empty cars. 

When possible, because of dynamic forces that occur on heavy grades, some types of cars such 
as long, light cars, should not be placed immediately behind the locomotive. Detailed treatment may 
be found in AAR Train-Track Dynamics reports. Also, loaded tri-level auto rack cars should not be 
placed next to open top loads such as sand, coal, sulfur, etc. 

Hazardous material when hauled by rail are subject to federal regulations. Cars carrying haz- 
ardous materials shall be placarded to indicate the lading. Also, the position in the train of placarded 
cars is governed by federal regulations. 

Over The Road: The effect of operating longer trains should be to operate fewer trains. 
Therefore, this will result in fewer train meets on line, providing it is physically and economically 
possible to extend the length of passing sidings. Some trains may be too long for passing sidings 
which could limit track capacity. 

An effect of operating longer trains is usually to increase car cycle time. 

Operating longer trains with minimum motive power units, so that they operate at lower speeds, 
could increase conflicts with trains operating at high speeds. The speed differential between the 
slower and the higher speed trains could become so great as to adversely affect scheduling and plan- 
ning for meets and passes. Motive power reliability must be at high levels. 

The number and length of sidings fall into a group of semi-variable costs, which are those that 
occur only when the increase in density of traffic advances sufficiently to require additional sidings 
or lengthening or existing sidings. Simulation of train operations can be helpful in determining sid- 
ing lengths and spacing. AAR simulation models are available. 



Proposed Manual Changes 5 1 



The longer the train, the more likely it will affect both at-grade railroad crossings and highway- 
rail grade crossings, since each such crossing would be occupied by the train for a longer time period. 
Conversely, shorter trains result in more frequent service, which will impact both types of at-grade cross- 
ings, although the time required to traverse each crossing would be relatively brief. As for highway-rail 
grade crossings, the greater frequency of trains may result in higher exposure to accidents. Also, it 
should be noted that motorists become impatient waiting for long trains to clear the crossing. Where 
local governmental ordinances limit train speeds, repeal of such ordinances would permit train speeds 
to be raised which would reduce the time trains would occupy these highway-rail grade crossings. 

Damage to equipment and lading results in large measure from the interchange of inertial 
effects caused by different parts of a train being on various gradients with portions stretched and oth- 
ers bunched and then balancing upon reaching level track. 

Crew costs may be based on the number of diesel units operated, or total locomotive weight. 
Longer trains tend to increase extra crew pay incurred during terminal delays. 

Comparative costs of operation and maintenance of ac versus dc traction motored diesel- 
electric locomotive should be considered. Some railroads are finding that, for a given horsepower, 
the greater adhesion of ac traction locomotives requires fewer locomotive units than dc traction loco- 
motive units. The brushless ac traction motor should reduce maintenance costs. 

Helper Senice: An intermediate breaking up of trains to cut in helpers, or just adding helpers 
at the rear of trains to overcome heavy grades, is a time-consuming and costly process but is the most 
practical and efficient utilization of equipment. 

The desirability of helper service depends on the nature of the route. If heavy grades are con- 
centrated, it is more economical to operate helpers on the heavy grades only than to run extra loco- 
motive units through the entire route, because they would not be needed on most of this territory. 
However, positioning helper locomotives and crews away from terminals, and the time spent cutting 
helpers in and out of trains, are significant costs. It should be noted that recent technology permits 
cutting off helper locomotive units on the fly without stopping the train. 

Track Limitations: The operation of fewer long trains to move a given quantity of traffic 
increases the time available for track maintenance and improves the utilization of m/w forces and 
equipment during schedule work hours. 

The dynamic forces involved with train operation are not thoroughly understood, and has been 
a continuing subject of investigation by individual railroads and the AAR. It is known that train 
length is one of the factors affecting these forces. 

Customer service covers frequency and dependability of service to both consignee and con- 
signor. This element is compri.sed of two categories: (1) customers whose product has a high time 
value (for example, manufactured goods or perishables) for whom frequency is of first importance; 
and (2) those whose product has a lower time value (for example, raw materials) with frequency not 
as significant, though service consistency may be important. It should be kept in mind that length and 
frequency are inverse variables, which both enter into the economics of the railroad. The customer 
who is concerned about a high time-value product, will be interested only in frequency and not in 
train length per se, assuming that lengths are restricted to a maximum limit before undesirable effects 
such as damage to lading from excessive slack action occurs. 

Limitations of plant and track layout at origin and/or destination can incur additional railway 
cost by requiring switching or excessive loading or unloading time. 

Interchange connections with other railroads and trains may be less reliable creating a lower 
quality of service for other traffic. 

The effect of operating longer and fewer trains will normally be to increase the number of cars 
required. 



52 Bulletin 760 — American Railway Engineering Association 



Individual axle loads may be limited by track conditions and the structural capacity of bridges. 
The capacity of bridges is usually the limiting factor for the total weight of a car or series of cars. 
Also, the maximum dimensions of individual loads may be limited by restrictive clearances. 

One effect of the operation of long trains will be the concentration of motive power on these 
trains which affects the flow cycle of motive power, especially if long trains are operated on a spo- 
radic basis. 

The operating costs of long trains compared to short trains must take into account the savings 
such as reduced number of crews as opposed to the increased cost of terminal delay, switching, over- 
the-road time, as well as the effect of equipment and plant investment in the long term. 

4.4 Train Management 

Train management involves moving trains by designated authorization issued by dispatchers 
who use computers, radio and train movement (signal) systems. 

To meet customer logistics requirements, an increasing proportion of cars are moved in scheduled 
trains. Unit trains such as those hauling coal or grain, while not operating on a fixed schedule, are moved 
to meet customer time requirements. Extra trains, often with no specific schedule are also operated. 

Amtrak and commuter railroad operators run passenger trains on fixed schedules. 

4.4.1 Dispatching 

Dispatchers use two methods of issuing orders to move trains: (1) verbal orders to train crews, 
who write the order on a special form; and (2) verbal order to train crews to proceed, but where the 
train movement is governed by signal indications, where installed. 

Computers are used extensively to aid train dispatchers. In Direct Traffic Control (DTC) or 
Track Warrant Systems (TWS), the computer is used to generate train orders, and safety checks are 
built into the software that prevents dispatchers from issuing conflicting orders. Also, the computer 
stores a record of orders issued. 

Computer simulation, an important advance in modem technology, allows railroads to adjust to 
optimize schedules, mofive power assignments and car supply to meet customer service requirements. 

4.4.1.1 Movement by Orders 

In non-signaled territory, trains can be moved by written orders issued by dispatchers directly 
to train crews. 

The Standard Code of Operating Rules of the Association of American Railroads has been the 
basis for operating rules of individual railroads. Variations of the Standard Code may be adopted by 
individual railroads. Additionally, several railroads may agree to a set of operating rules that will 
apply to all of them. 

Dispatchers issue movement orders to train crews via train-to-wayside voice radio. In general, 
a crew member writes the order on a special form and reads the order to the dispatcher, who confirms 
a correct order (dialog between the two will make corrections, if required) with the date, effective 
time and dispatcher identity. 

Also known as the train order system, this method of moving trains in non-signaled territory 
may be called DTC or TWS. In either DTC or TWS, limits are set for a train's movement. The lim- 
its may be mile posts, block names or named locations. 

Under DTC or TWS, train crews report periodically to the dispatcher. Some dispatchers have 
forms on which this train crew reporting information is recorded. If the dispatcher uses a computer, 
he will input the information into his computer. Similar activity occurs when trains are operated in 
automatic block signal territory. 



Proposed Manual Changes 53 



4.4.1.2 Movement by Signal Indication 

Train orders are issued for movement in Automatic Bloctc Signal (ABS) territory for meets and 
passes. However, train operations are normally governed by signal indications. 

Train movements are authorized by signal indications in Centralized Traffic Control (CTC) or 
Traffic Control System (TCS) territory. Under TCS operation, the dispatcher controls the switches 
and signals. Operation through interlockings is also governed by signal indications. 

When a train is ready to leave a yard or teminal, the train crew notifies the dispatcher, who will 
clear a signal authorizing the train to move. From then on, movement authority is by signal indication. 

In TCS or CTC territory, the movement of trains is automatically monitored and OS (on sta- 
tion) time reports are automatically recorded by the dispatch computer or a TCS control machine. 

General practice is to record all voice transmissions of both dispatchers and train crews. 
Recordings are usually kept up to six months or one year. 

4.4.1.3 Dispatcher Territories 

Dispatchers are usually assigned a specific territory. Its size depends on traffic load, on-track 
work by maintenance-of-way forces, signal maintenance, and the number of local trains or industry 
runs. Usually, there is more work for a dispatcher on the first trick (8 am to 4 pm), Monday through 
Friday than on other tricks or shifts. For second and/or third trick, or on weekends, adjacent dis- 
patcher territories may be combined. 

Most railroads have placed all dispatchers at a central office at or near the railroad's headquar- 
ters building. With such centralization, as much as 5,000 to 20,000 route miles of line can be dis- 
patched from one location. 

A few railroads have regional dispatch offices, handling about 5,000 road miles of territory. 

An individual dispatcher may control 200-400 road miles of territory. Where traffic is heavy 
with several local and through trains and extensive m/w activity, an individual dispatcher may have 
a smaller territory, possibly 100-200 road miles. 

Central versus regional dispatching for large railroads is a matter of management preference and 
cost. Advantages of one centralized dispatch center includes having only one of each of the following: 

• Building where dispatching is performed, and its physical and environmental security. 

• Power supply with backup, air conditioning, water supply, etc. 

• Communications and interface to management information systems. 

• Interface to railroad management. 

• Interface to emergency services and communications along line of road. 

Centralized dispatching has some disadvantages, including the co.st of alternate means of power 
supply and communications, realizing they support the entire railroad dispatch function. The.se alter- 
nate power and communications are required at regional dispatch centers, but are not as critical. 
When all dispatching is at one location, a failure of the power supply or communications could dis- 
rupt railroad operations throughout the entire system. In any case, a good disaster recovery plan is 
essential, including back up control at different locations. 

4.4.2 Train Movement Systems 

Train movement systems include automatic block signaling, interiockings, traffic control sys- 
tems and train control systems. Automatic block signaling and interiockings are often part of traffic 
control systems. In fact, TCS can be viewed as ABS with successive interiockings. 

ABS, when included in a TCS system, usually governs train movements between ends of con- 
trolled passing sidings. In multiple track territory, ABS is often installed to govern movements only 



54 Bulletin 760 — American Railway Engineering Association 



in one direction, with a second track utilized for movement in the opposite direction. In single track 
territory, ABS with a variation known as Absolute Permissive Block can provide for following and 
reverse moves, yet this systems prohibits conflicting moves between ends of passing sidings. 

Interlockings are provided where tracks cross, as well as where crossovers permit trains to 
move from one track to another and/or at junctions with other lines. Such track layouts may be com- 
plex, especially in terminal areas. Switches and signals are controlled in such a manner that conflict- 
ing routes cannot be obtained. In some complex interlockings, several simultaneous parallel train 
movements can be made safely, effectively increasing plant capacity. 

The trend is to convert from local control to remote controlling of interlockings, or to incorpo- 
rate interlocking control into existing TCS or CTC systems. 

Under local control, interlocking operators receive requests for train moves from train crews, 
yardmasters or dispatchers. 

Automatic intedockings are usually located where two railroads cross at grade and where traf- 
fic on at least one railroad is light. The first train to occupy an approach track circuit to the inter- 
locking receives a signal indication to proceed, while conflicting routes are halted until the first train 
clears the interiocking. 

4.4.2.1 Traffic Control Systems (TCS) 

Traffic control systems may also be known as centralized traffic control. TCS and CTC are 
operationally indentical. In each, train movements are governed by signal indications. Switches and 
signals are controlled by dispatchers. Train location, switch position and signal indications are dis- 
played on a model board at the dispatcher's office along with a track layout. 

In newer computer aided systems, the dispatcher may operate a PC or a computer work station 
(usually more powerful and with greater storage capacity than a PC). The work station is connected 
to a main frame computer with large storage capacity and high-speed retrieval and reading capacity. 

A large display, such as back-lighted projection screens which shows the entire railroad, may 
be provided. In such ca.ses, the dispatcher can usually bring up a section he wishes to control into his 
work station display unit. 

Some railroads use the work station concept whereby a dispatcher may have as many as three 
monitors to show his territory, but can bring up a small section to perform specific control functions. 
With this concept, the large display of the entire railroad is not provided. 

4.4.2.2 Automatic Train Control Systems (ATC) 

Automatic train control provides an additional safety system ovedaid on ABS orTCS/CTC ter- 
ritories to enforce speed restrictions. Although often classified as train control, cab signal systems do 
not enforce speed restrictions, but only bring signal aspects into the locomotive cab where they can 
be seen by the train crew. This is especially helpful in fog or with heavy snowfall conditions. Also, 
cab signals continuously reflect the track conditions ahead of the train. If a switch is open, for exam- 
ple, it is immediately shown by a cab signal downgrade. The engineman need not wait until he sees 
a wayside signal to take action. Conversely, if a train ahead should clear the main track, the follow- 
ing train's cab signal will upgrade to a more favorable aspect, allowing the engineman to increase 
speed. There is no need to continue at a slower speed until a wayside signal can be seen. 

Brake control can be incorporated into the cab signal system to initiate braking, should the 
engineman fail to acknowledge a change to a more restrictive cab signal aspect. 

Other train control systems include intermittent inductive train stop, and continuous speed control. 

The intermittent inductive system uses wayside inductors or beacons located just in approach 
to wayside signals. Upon a more restrictive wayside signal, an audible indication sounds in the loco- 
motive cab, alerting the engineman that the wayside signal is displaying an aspect to move at a more 



Proposed Manual Changes 55 



restrictive speed. The engineman must acknowledge within 4 to 6 seconds or a penalty brake appli- 
cation is automatically applied, which brings the train to a complete stop. 

With continuous speed control, the speed of the train is monitored in addition to track condi- 
tions. When conditions call for a lower speed (train ahead, for example), an audible indication is 
.sounded. If the engineman responds within 4 to 6 seconds by operating an acknowledgment lever and 
brings the train speed down, he maintains control of the train. If he does not acknowledge, or if he 
acknowledges but does not reduce train speed, a penalty brake application is automatically applied 
that brings the train to a complete stop. 

Generally, if continuous speed control is placed in service, the locomotives are also equipped 
with cab signals. 

On rail transit lines and on a few commuter rail lines, an intermittent train stop system is 
installed with a trip stop arm at each wayside signal location. When the wayside signal indicates Stop 
(a Red aspect), the trip is in the up or raised position. If the train pa.s.ses this Stop signal, the trip arm 
is struck by a car or locomotive mounted brake paddle releasing the air from the brake system and 
the train is brought to a complete stop. 

4.4.3 Economics of Train Management 

Costs associated with improved train management systems depend upon a number of variables. 
These should be considered and may not apply in all cases. Some are added costs, others will pro- 
vide savings. 

The variables to be considered when making a complete economic study of proposed train man- 
agement systems include: 

Payroll: For train crews, road hours may be reduced by improved over-the-road time, and ter- 
minal time may be reduced by the ability to dispatch trains more quickly and efficiently. 

For dispatchers, payroll costs should be reduced by centralized dispatching or consolidation of 
territories resulting in fewer dispatchers required. 

For interlocking operators, payroll costs can be reduced by remotely controlling plants or bring- 
ing interlocking control into TCS or CTC territory. Either of these control changes reduces the num- 
ber of interlocking operators required. 

Signal maintenance forces increase as train movement systems are expanded. Communications 
maintenance forces increase as more radio and data communications facilities are upgraded and 
expanded. This is especially true where defect detectors are equipped with "talkers" or radios to 
report conditions directly to train crews. 

Train Time: Improved performance should reduce origin-to-destination running time. 

Train hours are decrea.sed through shorter standing time for meets and passes, fewer train stops 
and reduced elapsed time when trains move through power and spring switches. With long passing 
sidings in TCS or CTC territory, non-stop meets may be accomplished. 

A reduction in train running time improves the ratio of train miles per train hour. 

Train miles may be reduced due to increased tonnage of trains through improved performance. 
Average train speeds should increase. 

Crew wages may be adjusted for the decreased train time, but savings depend upon labor agree- 
ment provisions. 

There may be reduced locomotive requirements through reduced train miles and reduced loco- 
motive time. 



36 Bulletin 760 — American Railway Engineering Association 



Train Stops: The costs of stopping trains, especially for meets and passes is reduced with 
improved TCS or CTC. With long sidings or sections of double track and single track, non-stop meets 
and passes can occur. 

Fuel savings would result from fewer stops and often less waiting time in sidings in traffic con- 
trol territory. 

Brake shoe wear is reduced through elimination of stops and slowdowns. 

Some small cost reductions may be achieved through less frequent replacements of rails and 
wheels, due to reduced wear as a result of fewer train stops and slowdowns. 

Freight train delay time is reduced. With fewer stops and less delay, damage to lading is reduced 
providing better service to shippers. 

Taxes are affected by changes in plant and can be significant if double track can be replaced 
with single track TCS or CTC with passing sidings. 

Overall maintenance costs with new train movement systems include both increased and 
decreased expenses. For example, if a second main track is removed, there are less track maintenance 
costs. But with more signal equipment required, there are increased signal maintenance costs. Also, 
with second track removed, rate of track geometry changes and track component degradation can 
increase (more trains over remaining track). 

If dispatcher consolidations can be implemented, associated costs are reduced. 

The improvement in motive power utilization may decrease this maintenance cost. 

Car Time: The same amount of traffic is moved at less expense with fewer car hours, due to 
improved over-the-road time. This postpones the necessity of increasing car inventory and reduces 
per diem payments. 

Depreciation changes due to the increased or decreased plant. 

Safety can be expected to improve due to increased mileage of train movement systems. There 
should also be a decrease in the number and severity of accidents. 

Investment changes result from the reduction of double track by the installation of the TCS or 
CTC. Also, if these traffic control systems are applied to single track, they increase its capacity and 
defer the requirement to install a second track. 

For planning expenditures, first consider whether changes in plant or operations can be made 
without the installation of new train movement systems. 

Intangible benefits also should be considered, including better customer service, increased flex- 
ibility in operations and improved productivity. 

Weather conditions affect operations, and costs are associated with weather. For example, one 
should consider snow removal as a cost, as well as costs to recover from floods or other weather 
related occurrences. 

4.5 Communications 

It should be noted that communications facilities and services can be provided by the railroad 
or transportation agency. 

Also, it is possible that all or part of communications facilities and services can be provided by 
other entities, such as communications common carriers and other private communications companies. 

Costs for installation, operation and maintenance have to be analyzed including the condition 
of owning the communications facilities, leasing them, or having the entire communications plant 
and service provided by an outside entity. 



Proposed Manual Changes 57 



One factor to be considered in providing communications by the transportation carrier is that of 
control and recovery after natural disasters. 

4.5.1 Essential Communications 

Digital and voice radio can connect wayside stations to moving trains. For dispatcher-to-train 
crew communications, this can be a voice link and/or a digital voice system. A digital data link is 
needed to transmit data between control centers and the train. The link from the moving train to the 
wayside is via radio; but the link from the wayside radio station to the control center can be fiber 
optic cables, microwave radio systems, UHF or VHF radio station segments, or leased communica- 
tions circuits. The fiber optic cable provides high capacity with room for expansion and is immune 
to electro-magnetic fields created by electrified rail lines or electric power transmission lines. 

Inductive communications technology is often used to transmit data from track to train via bea- 
cons, "'wiggly wires", inductors or transponders. Coded track circuits in the rail can transmit signal- 
ing data, speed commands, etc., to a train. Other systems that do not use coded track circuits can u.se 
digital radio links from the wayside to a train equipped with an on-board computer to handle the con- 
trol and information function. 

4.5.2 On-Board Communications 

On board the train, voice radio may be advisable with portable handsets for crew member com- 
munications. For passenger trains, a suitable location mid-train, such as a dining or lounge car, could 
be a conductor's station equipped with 30 to 50 watt output radio enabling him to contact the control 
center or dispatcher. 

4.5.3 Emergency Communications 

During an emergency when the train is not operating under normal conditions, communication 
is of extreme importance. Battery-powered handheld radio sets are most useful, but standby power 
sources or batteries should be provided for the 30 to 50 watt radio tran.sceivers to enable train crew 
members to contract a control center and local emergency services. The radio communications load 
should be considered when sizing battery power for radio equipment. 

4.5.4 New Technologies 

Cellular and satellite radio communications offer additional wireless links between moving 
trains, m/w forces and railroads' regional offices. Also, some railroads are investigating the use of 
Global Positioning Satellites (GPS) for determining train location, and physical plant mapping. Also, 
being looked into is the use of satellite communications for high volume communications to link rail- 
road headquarters with regional and off-line offices. Satellite communications could reduce the 
requirements for land lines, microwave or cable links. Again, this requires an economic study of costs 
of the various forms of communications covering installation, operations and maintenance. 

4.6 Defect Detection 

See Part 5, Economics and Location of Defect Detector Sy.stems, in Chapter 16 of this Manual. 

4.7 Line Capacity 

The basic operating capacity of a segment of railroad track to handle train movements is depen- 
dent upon the speed of trains and the distance between them. The speed of a train over a specific seg- 
ment of railroad is. in turn, dependent upon the ratio of horsepower to gross tonnage of the train and 
the grades, curves and other such features encountered by the train. In single track, double direction 
operation, the distance between trains is a function of the distance between usable sidings. A usable 
siding is one that must be clear for an opposmg tram to enter and pass through in a normal, straight- 
forward movement, and also long enough for all trains to clear the main track before stopping. The 



58 Bulletin 760 — American Railway Engineering Association 



distance between trains is dependent upon the length and number of signal blocks used to space trains 
within double track territory under smgle direction operation. Signal spacing is a function of the stop- 
ping distance of a maximum tonnage train operating at the maximum authonzed speed; this will pro- 
vide the maximum stopping distance for all trains in this territory. 

Line capacity can be expressed by the formula: 

C, = C,xE 

where 

C|, = Practical line segment capacity 

C, = Theoretical line segment capacity 

E = Dispatching efficiency for line .segment 

The dispatching efficiency of a railroad varies by territory and can be influenced by many fac- 
tors. Some of these are relatively predictable while others are completely random. 

The type of signal system is a major influence on the dispatching efficiency. Traffic control sys- 
tems have a much higher efficiency than dispatching in non-signaled or "dark" territory. Automatic 
block signaling falls somewhere between these two mentioned above. 

Radio communications for train-to-wayside functions can increase dispatching efficiency, as 
well as radio "talkers" at defect detector installations. 

The type of traffic moving over a segment of a railroad influences the line capacity and the dis- 
patching efficiency. A bridge line has a much higher dispatching efficiency than a line with many 
industries and yard limits. The movement of yard engines and local freights can reduce the line 
capacity for through movements. 

The physical characteristics of a line also influence dispatching efficiency and line capacity. A 
line with heavy grades or undulating territory, which may have a higher incidence of train separa- 
tions, has a lower dispatching efficiency than a line with fewer hills and resulting abnormalities in 
train operation. 

Also impacting line capacity are maintenance operations, such as allowing time and track for 
maintenance-of-way work. Some railroads provide "windows" for m/w work and later fleet trains 
after the work is completed. MAV work is a fact of life and should be taken into account when cal- 
culating line capacity. It should be pointed out that on many railroads, such as those in northern cli- 
mates, m/w work is usually done in spring and summer weather. Thus line capacity is impacted dur- 
ing certain times of the year. Of course, one can make the point, that snow clearing also affects line 
capacity. Thus weather and m/w work should be taken into consideration in any line capacity study. 

All of the factors mentioned so far are relatively predictable for a given line segment. Other fac- 
tors which are completely random in their effect include locomotive failures, derailments, overheated 
bearings ("heatboxes"), highway-rail grade crossing accidents, temporary slow orders, etc. Any 
occurrence which prevents a train from making its scheduled running time tends to lower the dis- 
patching efficiency of a line. 

Some factors such as locomotive failures, hotboxes, and highway-rail grade crossing accidents 
may be random in the short term, but can be managed over the longer term. 

Since these factors are random, their statistical prediction may be possible if conditions are sim- 
ilar for a large number of trains. 

The theoretical capacity of the line segments is determined by the formula: 

_ Time x N 
^1 ~ TT 



Proposed Manual Changes 59 



where 

Time = number of units of time in the period for which capacity is being calculated, for exam- 
ple, 1440 minutes in a day 

N = the number of directions run on a single track (either 1 or 2) 

H„ = the maximum gross headway in N directions 

For double-track, single-direction running (N = 1), gross headway (H,) is the average minimum 
time between trains in a minute. 

This can be calculated using the formula: 
„ Db X Bn 

Hi — 

' V 

where 

V = average speed of trains over a line segment 

Db - average signal blocks length 

Bn = number of signal blocks separating trains operating on a proceed signal indication 

For single-track, two-direction operation (N - 2), the gross headway (H,) is defined by the 
formula: 

H, = R, -I- R, -t- T, -t- T, 

where 

R, = minimum unopposed running time in direction 1 for the train with the lowest horse- 
power to gross tonnage ratio on the line segment in direction 

T| = the time necessary for a train in one direction to enter a siding, clear the main track, 
return the turnout to the normal position such that the opposing train may proceed. 

In single-track, two-direction territory, gross headway is the minimum period of time one train 
could operate following a previous train from siding A to the next siding B, if the opposing train at 
siding B is to be moved from B to A between the trains. In using the formula, care must be taken to 
ensure that all values used are in like units. Generally, distance is expressed in feet, time in minutes, 
and speed in feet per minute. 

After calculating the capacity of a line segment, its load factor is calculated. This is done by 
dividing the number of trains currently being operated over the line segment by the practical capac- 
ity of the line segment, and multiplying by 100 to find the percentage. 

The practical maximum line capacity is about 75% of the potential capacity given by these for- 
mulae. The contributing factors include: (1) automatic block signals, (2) quality of dispatcher-to-train 
radio coverage. (3) extent of local business, and (4) track profile characteristics or incidences of train 
separation. 

Operation of a line at too high a load factor is not a desirable characteristic, because little capac- 
ity remains for unplanned movements such as extras or detours. The availability of track time for 
maintenance work becomes limited as the load factor increases. 

To determine the bottleneck segment of a line, the highest load factor is of more importance 
than the lowest capacity segment. Any plans for plant improvements to expedite train movements 
must fall into one of two categories. The speed of trains must be mcrea.sed or the distance between 
trains must be decreased. Line or route modifications or changes in the horsepower to gross tonnage 
ratio will affect the speed of the train. Changes in the location of siding and/or changes to the signal 
system will affect the spacing of trains. Efforts in these areas yield optimum economic benefits when 
directed to those segments with the highest load factors. 



60 



Bulletin 760 — American Railway Engineering Association 



In practice, many railroads employ sophisticated computer simulation models that "run" vari- 
ous segments of the railroad at different loading levels based on actual or projected schedules, to pin- 
point bottlenecks to improve train flow. 

The absolute value of the load factor or comparison to some standard is not really all that impor- 
tant. The importance of the load factor is in enhancing management's capability to rank portions of 
the railroad so as to be able to direct capital improvements where they will have the greatest benefit. 
For this reason, the selection of an exact dispatching efficiency is not all that critical. Theoretical line 
capacity may be used just as easily. One must understand that capital funds are a scarce resource. 
Management requires a quantitative method of ranking demands. Return on investment is of prime 
importance, but in areas such as increases in railroad line capacity, calculation of return on invest- 
ment is rather imprecise. Ranking of line segments by "load factor" becomes even more important. 

The preceding discussion has dealt exclusively with the "physical" capacity of a railroad line 
segment. Additional constraints may limit railroad operation to less than its physical capacity. One 
such constraint is the Hours of Service Law, which limits the on-duty time of train and engine crews. 
As the over-the-road time of trains between crew change points approaches the limit set by this law, 
the railroad is said to be "saturated", although the physical capacity of the line may not have been 
reached. In addition to the over-the-road running times and delay to meets and passes with other 
trains, other factors such as initial and final terminal delay, set-outs and pick-ups on line-of-road, etc., 
may affect the total crew on-duty time, and constrain the operating capacity of the railroad, as 
opposed to its physical capacity. 

4.8 Terminals 

Terminals are facilities dedicated to the performance of one or more specific tasks. These tasks 
may include car classification, car unloading and loading, and interface between various transporta- 
tion modes. Each type of terminal requires specific types of design and equipment to economically 
achieve its purpose. Terminals are covered in Chapter 14 of this Manual. 



Experience has proved you can depend on them. 
WESTERN-CULLEN-HAYES 

Railroad Products' 






For Maintenance-Of-Way 

Hayes Bumping Posts 
Delectric® Operators 
Hayes Derails 



For Signaling 

Crossing Signals 
Gate Mechanisms 
Case & Track Hardware 



For Communications 

Telephone Shelter Boxes 
Lightning Arresters 




WESTERN-CULLEN-HAYES, Inc. 

2700 West 36th Place • Chicago, Illinois 60632 
Telephone: 773/254-9600 
Web site: www.wch.com 



Take advantage of Western-Cullen-Hayes 
service proven equipment and experi- 
enced railway supply personnel to assure 
safe, efficient operation on your line or 
in your yards. Call 773/254-9600. 



Chemelron^s ATV. 

Coming at you with more welding power than ever. 



On track. Off track. Chemetron's latest 
technology All Track Vehicle (ATV) can 
deliver field welding sen/ices fast, and with 
more consistent flash-butt quality than any 
other in-track welder Ever. 

The technology behind the weld quality 
of Chemetron's ATV is 
an improved K-900 
welding head and the 
proprietary software 
our on-board 
computer uses to 
control welding 
cycles. Precisely. 

Chemetron's 
mobile welder was 
engineered to 
exceed AREA 
specs, including 



the "upset to refusal" requirements. Our 
computer system guarantees plant quality 
in-track welds. Continuously. 

Offering optimum production for all 
rail sizes and metallurgies, Chemetron has 
complete mobile welding units for sale or 
lease, for short or long 
term contract welding. 
W\\h full engineering 
and maintenance 
"^^^*T^j support. 

Put Chemetron's 
ATV in-track 
welder to work 
for you by calling 
Larry Taylor at 
847-520-5454. 
Today. 




61 




Will the next rail you buy 
be fully heat-treated, 
head-hardened, or inter- 
mediate strength? 

Will the next turnouts 
you buy be state-of-the- 
art manganese castings, 
vacuum -molded and 
machined for perfect fit? 

The answer is yes, if 
you're out for the best rail 
products the world has to 
offer. And that means 
Foster-Class, from L.B. 
Foster Company. 
World-class. 

Well go aCTOSS the coun- 
try or around the world 
to meet today's standards. 



So you get a double 
advantage: world- 
class technology along 
with superior Foster fin- 
ishing and Foster servic- 
ing right here at home. 

For instance, Foster 
supplied turnouts meet all 
AREA specs, and every 
inch is pre -inspected 
before shipment. 

We go to special lengths 
on relay rail, too. Just as 
we've been doing for 80 
years, we bring you the 
largest stocks in the world. 
And more. Today we take 
up and deliver pre -welded 
lengths up to a quarter- of 



a-miletocutyour 
on-site fabrication costs. 
Go Foster-Class 

for your tallest or 

smallest orders. 

Give us a call and we'll 
ship any rail order — 
including turnouts and 
accessories — on time, 
anywhere, from stocking 
points coast to coast. Plus 
special sections and long 
lengths of new rail, rolled 
to order 

We're also your 
number one source for 
sophisticated track and 
contact rail components 
for transit systems. 



The Foster difference 
is a world of difference. 
Because Foster- Class is 
world-class. Phone or 
write L.B. Foster Com- 
pany, 415 Holiday Drive, 
Pittsburgh, PA 15220. 
(412) 928-3400. 



FOSTER 



L.B.FOSTER 
COMPANY 



62 



Proposed Manual Changes 



63 



Proposed 1997 Manual Revisions to 
Chapter 18 — Light Density and Short Line Railways 

Part 1— Track Components and Track Design 

Page 18-ii. Add the following new Table of Contents. List of Tables and Table 1.1, Dimension and 
Surface Specifications for Relay (Second hand) Rail. 

Table of Contents 

Section/Article Description Page 

1.1 Rail 

List of Tables 

Table Description Page 

1-1 Dimension and Surface Specifications for Relay (Secondhand) Rail 

Rail Sections - In Use Since About 1900 (See Plan 1001 in Portfolio of Trackwork Plans) 



Table 1.1 Dimension and Surface Specifications for Secondhand (Relay) Rail 



Length 


Standard 39 ft lengths. 

Not more than \Q9c of lot between 33 ft and 39 ft. 

No rail shorter than 33 ft. 


Vertical Wear 


Average top wear '/» in. or less with maximum at any one loca- 
tion of Vi: in. (For yard and sidings, average top wear up to 
'a in. with Vi.. in. maximum at a single location). 


Side Wear 


Maximum of 'A in. ('/^ in. for yard and sidings), with wear on 
one side only. 


Lip or Overflow 


Maximum of '/k, in. 


Engine Bums 


Maximum of '/; in. diameter (or 'A in. wide by Vi in. long) and 

Vn in. deep. 

Maximum of six engine bums per rail. 

Engine burns on no more than 8% of the lot. 


End Batter and Chipping 


Maximum of '/k. in. ('/» in. for yards and sidings) when measured 
'/: in. from the rail end with an 18 in. straightedge. 


Running Surface Damage 


Maximum of 'A in. wide by '/: in. long and '/'; in. deep. 
Flat spots are not permitted on the rail head. 


Defects Not Permitted 


Bolt hole cracks or breaks, broken base, crushed head, detail or 
engine bum fractures, head-web separation, piping, horizontal 
or vertical split head, torch cuts or fiame gouges, compound or 
transverse fissures, pitting. 


Condition and Appearances 
Internal Inspection 


Rail must be: free from obvious defects; clean in appearance; 

straight in line and surface and without kinks; and free from 

ba.se defects such as plate wear, spike notches, pitting, and 

fiame-gouging. 

Rail to be ultrasonically inspected before of after installation. 

Defective sections to be rejected and replaced. 



r 



MAGNUM 

CONCRETE GRADE CROSSING 



• Manufactured to fit 
any rail ranging from 
115 lb to 136 1b. 

• Designed for wood 
and concrete ties. 

• Low maintenance — 
long wear. 

• Custom designed for 
switches. 

• Insulated crossings 
available. 

MAGNUM MANUFACTURING CORPORATION 
(801) 785-9700 • FAX (801) 785-9701 



Manufacturing locations in: 

Pleasant Grove, Utah and Everman, Texas 





MAGNUM 



'PoraV^" 



Nolan Rail Products 



Choose from a wide 
range of dependable, 
tested and field 
proven rail products. 
The Nolan Company 
offers the best in long 
lasting, hard working 
maintenance-of-way 
equipment. For more 
information on 
Nolan's full line, ask 
for our full color 
catalog. 




Push Cars 



the 

NOLAN 

company 



1016 NINTH ST S.W. 
CANTON, OH 44707 
(330) 453-7922 
FAX (330) 453-7449 



64 



Proposed Manual Changes 65 



Revisions to the AAR Scale Handbook 
Part 5 — Vehicle Scales 

Page 3-2. Committee 34 has reviewed Part 5, Vehicle Scales, and approved the following revisions, 
effective January 1, 1997. 

5. 1 (c) Vehicle scales may be mechanical, analog, digital, or a combination thereof. 

5.1 (d) All vehicle scales shall meet the specifications and requirements of the National 
Institute of Standards and Technology (NIST), Handbook 44, and State laws for the jurisdiction in 
which the scale is located. 

(Editorial Note: Remove "comma" after the word 'specifications' in the Manual.) 

5. 1 (e) Vehicle scales shall have valid NTEP Certificates of Conformance. 

5.2.1 (a) The minimum pit depth, measured from the bottom of the weighbridge structure to 
the floor of the pit, should be: 

5.2.3 (a) Lever stands of load cell base plates shall be properly leveled, and grouted if neces- 
sary, to provide even distribution of the load over the full surface of the stands or plates. 

5.2.3 (b) Load cell base plates shall be leveled to a tolerance of not more than 0.015 in. per ft., 
with consideration to leveling the weighbridge transversely and on grade longitudinally. 

5.2.4 (a) Piers must support the combined loads applied by the weight of the scale, the weigh- 
bridge, plus the maximum anticipated load on the scale, so that any settlement shall be uniform 
throughout the structure. 

5.2.5 (a) A minimum of four (4) anchor bolts shall be used for load cell base plates where the 
design creates an uplift or shear reaction to the anchor bolts. 

5.3.1 (a) Main girders for weighbridges shall not deflect more than 1/600 of the span at mid- 
point when loaded to the rated concentrated load capacity (Dual Axle Capacity). 

5.6 All scales manufactured after January 1 , 1986, must be marked with the accuracy class des- 
ignation, nominal capacity, concentrated load capacity, scale division "d", and verification scale divi- 
sion "e" if different than "d", clearly on the device. Unless temperature range is 14 degrees to 104 
degrees F (-10 degrees to 40 degrees C), the temperature range must be conspicuously marked. 

5.12 The power requirements of the electronic instrumentation and load cell circuitry for elec- 
tronic scales must conform to applicable regulatory requirements and codes; and the scale must sat- 
isfy the tolerance requirement when scale equipment is subjected to RFl and EMI infiuences which 
may exist during normal scale operation. 



Change 50 ties an hour 
under heavy traffic 




Twenty trains per day use this track. Yet despite this heavy 
traffic, a single MRT-2 Tie Changer can replace 50 ties an 
hour in an average day. The secret is the quick on-off track- 
ability designed into the MRT-2. In less than two minutes, 
any place along the line, it can climb on or off track com- 
pletely under its own power. Old ties - even switch ties - are 
rerrioved whole, with minimal disturbance to track structure. 

If you're now wasting valuable time clearing for trains, switch 
to the MRT-2. Sales and service available throughout North 
America. Contact us for a free demonstration. 



m^SMi 



MODERN TRACK MACHINERY INC. 

1415 Davis Rd., Elgin, IL 60123-1375 

Tel. (847) 697-751 Fax: (847) 697-01 36 



MODERN TRACK MACHINERY CANADA LTD. 

5926 Shawson Drive, Mississauga, Ontario L4W 3W5 

Tel. (905) 564-121 1 Fax: (905) 564-1217 



66 



NORFOLK SOUTHERN TRACKSIDE 
LUBRICATION STUDIES 

1997 Interim Report 

By: D. E. Cregger* 

Abstract 

Interim results from detailed monitoring of rail wear related to trackside lubrication equipment 
and lubricant has revealed significant differences in equipment and lubricant performance. The effect 
of rail profile grinding on high rail gauge face wear was defined for 9-degree, 6-degree, and 4-degree 
test curves. Significant opportunity was found for increasing rail metal wear life and increasing the 
efficiency of lubrication. 



Good afternoon. I am plea.sed and honored to present to you today the interim findings of the 
Norfolk Southern trackside lubncation studies. These on-going studies were initiated due to a loss of 
shunt incident in 1992. The results of our first tests were shared at the 1994 As.sociation of American 
Railroads (AAR) sponsored Town Hall meeting in Kansas City. Those findings of large percentages of 
grease waste were later corroborated through independent testing by the Association of American 
Railroads-Technical Test Center at Pueblo, Colorado. We learned a great deal with our first tests and 
were compelled by the potential benefits to continue the studies. 

Time constraints prevent the presentation of all the procedures, techniques and the many vari- 
ables encountered during the studies of equipment, grease and rail wear. However, it is important to 
understand that much equipment and grease improvement was required before rail lubrication bene- 
fits could be monitored through rail metal loss. Also, note that the test site length grew as we suc- 
cessfully extended grease carry for proper lubrication. Grease carry that deposits on the rail head 
instead of the gauge face and comer does not provide metal wear protection and, therefore, was not 
a factor in determining carry distance. The selected test site is an example of grease deposited on the 
rail in a non-productive manner 

The test site (Exhibit 1) is a single track of 1993 vintage Nippon 136RE premium rail with 65 
MGT traffic that is 95% one direction (eastbound loaded coal trains, mixed freight and grain trains). 
Train movement is from MP-V298.5 through MP-V295.5 with an ascending grade of 0.10%. 
Locations to monitor curvature high rail wear were established at (orange marks): 

Milepost Rail Curvature Description 

V298.5 Tangent Lubricator 

V298.3 North 9.3-deg R Lst curve after lubricator, 0.2 miles 

V297.8 South 6.0-deg L Lst curve after lubricator, 0.7 miles 

V296.0 North 4.0-deg R Last curve in test site, 2.5 miles 

A lubricator installed at V297. 1 was removed from service so that baseline wear rates for unlu- 
bricated rail could be established at V296.0. When we started, grease was being carried intermittently 
from the lubricator into the curve bodies of the: 

1 . 9-degree curve for 0.3 miles on the high rail (aprox. 0.3 miles including tangent on 
north rail). 

2. A similar distance on the high rail of the 6-degree curve (total carry of 0.7 miles including 
tangent and low rail of the 9-degree curve). 



♦A.ss'i. Manager — Chemical Technology, NS. 



67 



68 



Bulletin 760 — American Railway Engineering Association 




in 




O) 




CM 




> 




o 


ai 






in 


<o 


CO 


■♦-' 


O) 


(/) 


^ 





> 


Q) 




r 


"D 


-*-> 


C 


r 


-) 


n) 


o 


n 


to 

CO 


o 


0) 


O 


+ 




in 


o 

1 


c 


CD 


o 


■D 


■«-' 


m 


u 


I- 


a> 


O 


■D 




O 




St 




(0 





z 












o 


0) 




(D 








*i 




*i 




c. 


1- 


(0 




(0 




(U 

lii 


Q. 


0) 


2 


(D 


o 

8 


O 


3 


8 


3 


" i« 


0) 


o 


'k. 


O 




Last 
test; 


HI 
Q 




_2 


2 


^ 

^ 


LU 












t£. 












3 


q: 




^ 




or 


1- 


O) 




O) 




C3) 


< 


0) 




(D 




<D 


ID 


■o 




•u 




T3 


CO 




O 




O 


O 


ai 




CO 




"^ 


_J 






3 






2 


o 




o 




o 


z 




CO 




Z 


H 












(0 












o 












Q. 


CO 




CO 




o 


LU 


00 




h-^ 




(b 


_l 


o^ 




05 




o> 




CM 




CNJ 




CM 


:s 


> 




> 




> 



Exhibit 1 

Trackside Lubrication Study 

Whitethorne District Milepost V295.7 to V298.5 



Paper b\ D. E. Cregger 



W 



Improvements to equipment and grease resulted in consistent liihrication /protection 3.0 miles 
from the lubricator. This surprised us. but even more amazing is that the grease consumption was 
reduced by 30'7c. 

A tribometer was used to a limited extent to confirm presence of lubrication. We quickly found 
that the presence of a low co-eftlcient of friction reading did not indicate that the gauge face was pro- 
tected from wear. Every wheel was not being lubricated while passing the trackside lubricator. Dry 
wheels were found to be wiping away lubrication and wearing away gauge face metal (Exhibit 2). 

Once this mechanism was recognized, we realized that the rail is the working grease reservoir. 
Grease maintained on the gauge face must be adequate to furnish rail gauge face lubrication and sup- 
ply sufficient grease volume for transfer to dry wheel flanges. The volume of grease required on the 
rail gauge face is dependent on the efficiency of the wiping bars and the pumping system. 

The effectiveness of this lubrication method was determined by using a Starrett dial indicator 
and a measuring frame manufactured by Track Renewal Engineering. Inc. to measure rail wear 
(Exhibit 3). The following analysis focuses on the high rail gauge face wear of the three points mon- 
itored for rail wear. 

Rail dimension measurements were collected 42 times in a 678 day interval (Exhibit 4). The 
AAR wear rates (noted at the bottom of the table) for differing lubrication conditions developed by 
Elkins et al-1984 were used as guidelines to describe lubrication conditions as related to gauge face 
rail wear. The semi-dry wear rate (wear in the presence of grease deposited somewhere on the rail 
head and face) was determined for each inspection point. Note that the semi-dry wear rate will vary 
w ith the operational and location conditions of each railroad. 



P ** 



^^^^i^war 




Exhibit 2 
Grease Consumed by Dry Wheels 



70 



Bulletin 760 — American Railway Enjiineerini; Association 




Exhibit 3 
Track Renewal Engineering, Inc. Rail Gage 



TABLE 1; COMPARISON OF CURVES AND SITE WEAR RATES (Basis is Semi-Dry Rail) 



CURVE, 
LOCATION, 

CONDITION 


9.3R 
V298.3, 
.IN 
100-MGT 


9.3R 

% 


6.0L 
V297.8, 
IN 
100-MGT 


6.0L 
% 


4.0R 
V296.0, 
IN 
100-MGT 


4.0R 

% 


SITE 
AVG, 
IN 
100-MGT 


SITE 
AVG 

% 


LUBRICATED 
RAIL 


0.0284 


14 


0.0057 


3 


0.0200 


6 


0.0180 


7 


PROFILE 
GROUND 


0.0348 


17 


0.0207 


10 


0.0396 


11 


0.0317 


13 


SUMMATION 


0.0632 


31 


0.0264 


13 


0.0596 


17 


0.0497 


20 


REDUCTION 


0.1390 


69 


0.1861 


87 


0.2888 


83 


0.2047 


80 


SEMI-DRY 


0.2022 


100 


0.2125 


100 


0.3484 


100 


0.2544 


100 



AAR WEAR RATE DEFINITIONS: (INCHES PER 100 MGT) 

Dry Rail 0.5000 Low Lubrication O.IOOO 

Medium Lubrication 0.0290 High Lubrication 0.0064 



Exhibit 4 



Paper by D. E. Cregger 7 1 



Therefore, for this presentation the wear rate will be referenced as NS Semi-Dry. As we 
improved lubrication performance, wear rates for lubricated wear and the effect of rail profile grind- 
ing on gauge face wear was measured. The difference between semi-dry and the combination of 
lubricated and profile grind wear is the potential improvement. It is interesting to note that we dupli- 
cated AAR High Lube conditions at the 6-degree curve (6L-V297.8) test location. 

Graphically displayed (Exhibit 5), the four tallest bars (NS Semi-Dry Rail) represent the gauge 
face wear of the rail after equipment improvements made consistent operation possible. With further 
improvements in equipment and grease formulations, lubricated gauge face wear in all the test loca- 
tions, the first bar shown for each location, was reduced to at least AAR Medium Lube. Analysis of 
data collected allowed us to define the gauge face wear effect of rail profile grinding, the second bar 
shown at each location. 

The benefit of gauge face lubrication was negated for 30-60 days after rail profile grinding. Gauge 
comer relief removed the rail geometry necessary for grease transfer to the rail face. The grease migrated 
along the wheel profile and deposited at the gauge side of the rail crown. Gauge face lubrication was 
renewed when rail wear created the appropriate rail and wheel geometry for proper grease transfer. 

Additional data processing as shown in Exhibit 6 revealed another segment of gauge face wear 
that is variable to operating conditions, the vertical striped segment of each bar at each location. This 
segment is wear from conditions beyond the response range of the trackside lubricator/grease com- 
bination. These variations include weather, operational limitations of lubricator equipment, grease 
properties, maintenance intervals, and train operations that do not match the track plan. The 22% 
wear due to variable operating conditions in the 9-degree curve (at point 9.5R, V298.3) is heavily 
related to train movements below planned track speed (underbalanced). A signal is positioned 100 
yards prior to the trackside lubricator at V298.5 and STOP signals are common at this location. The 
16% wear due to variable operating conditions in the 4-degree curve (at point 4R, V296.0) is more 
related to factors that influence the grease carry. 

The remaining segment, the top part of each bar at each location, of gauge face wear is perma- 
nent potential savings. The application of proper equipment and compatible grease reduces gauge 
face wear. The average wear reduction for the test track is 63% or 0.1623 inches per 100 MGT. 

The natural question this information generates from anyone dealing with a budget is "How 
much money does it save?" To simplify and bring to a common denominator, this pie chart (Exhibit 7) 
shows wear segment comparisons with lubrication conditions defined as AAR Dry Rail (0.5 inches/ 
lOOMGT). For each dollar of rail gauge face wear cost per 100 MGT, the cost per segment is reflected 
by the related percentage noted. The potential savings, assuming dry rail conditions initially, are 
81.71 cents per dollar of your actual rail gauge face wear cost based on new rail cost FOB plant. Note 
that, if the rail is profile ground more than once per 100 MGT, the additional cost of grind wear must 
be deducted from potential savings. This savings prediction is conservative. AAR Dry Rail is defined 
as a wear range of 0.5 - 0.7 inches per 100 MGT. All calculations were based on 0.5 inches per 100 
MGT. Your dry rail wear may be greater than reflected in this analysis. 

The next quesfion expected is "How did you do it?" We had to learn that our trackside lubrica- 
tion arrangements were not performing to expectations: 

1. Each passing wheel is not lubricated by current designs of trackside lubricators. In cold 
weather, 30-45 cars must pass before the lubrication system is adequately pressurized to sup- 
ply grease through the wiping bar ports. 

2. There are great efficiency differences among actuator/pump designs. The design that best 
provides constant grease pressure results in less waste grease. 

3. Grea.se formulations must be compatible with the equipment operation. The physical char- 
acteristics of the grease after being pumped through the lubricator system must support the 
formation of a grease column at the wiping bar port that transfers the grea,se to the passing 
wheel flange. New grease formulations were developed for NS applications. 



72 



Bulletin 760 — American Railway Engineering Association 




in 
d 



S3H0NI 'dV3/\A 3AllVini/\in00V 



Exhibit 5 

Rail Gauge Face Wear Rate Per 100 MGT 

Test Site MP V298.5-V295.7 



Paper by D. E. Creeger 



73 



•i 

B 

a. 

% 

•8 




CD 






^ 




CO 






^ 


w 




LU 

LU 
> 




q: 


O 




< 


DC 




111 


Cl 


2 


s 


1 

1- 


CJ 


z 


D 






o 


1— 


q: 


o< 


O 


X 




LU 

_l 

u. 
O 


5 

(0 

w 


^\- 


cr: 


o 


U) 


CL 


_l 


O 


^^^B 


^^^H 


GL 


^^^H 


H^^H 


LU 


^^^g 


H|^H 


_l 


^^^M 


HIHI 


^ 




0) 


£J 




O 


2: 




1- 


«< 




Q 


fe IXI 

5q; 




O 
y 


r ) 






< 


OL 


O 

z 


> 


< 


1— 


£t 


UJ 


-) 


$ 


HI 


o 


Q 
LU 


Q. 
O 




1- 


LU 




< 


_l 




o 


OQ 




Qi 


< 


n 


m 


q: 


^ 


_l 


^ 




q: 










in 

O) 






— ^^ 










- 















ioi/M-ooi /S3H0NI 'av3M 3Aiivnni/Mnoov 



Exhibit 6 

NS Semi-Dry Rail 

Gauge Face Wear Components 



74 



Bulletin 760 — American Railway Engineering Association 



POTENTIAL 
SAVINGS 



(81.71%) 



(3.50%) 
tif^^ (8.56%) 




Lubricated Wear 

Profile Grind & Wear 

Variable Operating 
Conditions 

Loss Without 
Improvements 




Exhibit 7 
Wear Cost/100 MGT with AAR Dry Rail 
Curves > 3.5-Deg at Mill Cost Net Scrap 



TABLE 2: *WIPING BAR - PORT DESIGN 



PERFORMANCE: 


LARGE PORT 


SMALL PORT 


BLOCKS 


Grease Carry Distance in Test Site, 
AVG % of Site Protected 


48.2 


100 


69.2 


Waste Grease Collected (2 wheel turns), 
%, range with various arr£ingements 


24.4-38.0 


10.1-32.0 


75.5 


Grease Consumed, 
Pounds/ MGT-Mile 


46.0 


15.2 


92.4 



♦Portec MC-3 wiping bars were modified to small ports and compared to the standard MC-3 
design. 

Exhibit 8 

TABLE 3 ; LUBRICATOR ARRANGEMENT PERFORMANCE COMPARISON 



SUPPLIER 


RATING 


A 


45 


B 


59 


C 


72 


D 


96 



Exhibit 9 



Paper by D. E. Creeger 



75 



4. Wiping bar design is critical to successful track gauge face lubrication (Exhibit 8). Portec 
MC-3 wiping bars were compared with M&S Blocks and modified MC-3 wiping bars. The 
modified bars were equipped with a manifold divider that decreased the port size and dou- 
bled the number of ports. The test site was monitored for the amount of grease consumed, 
waste grease collected at the lubricator site, and grease carry distance. These data were 
processed to include all the parameters of each setup for a direct comparison (Ibs/MGT-mile) 
of grease used per MGT of traffic movement for one mile. The small port design performed 
better than the .standard MC-3 design by a factor of 3X and 6X better than M&S Blocks. 
Note that the grease must be compatible with the lubricator equipment. 

After good lubrication patterns had been established and maintained, various suppliers' equip- 
ment arrangements and mixtures of suppliers' equipment were comparison tested in actual service. A 
rating system (Exhibit 9) was developed from the many measured parameters deemed important to 
NS operations. Equipment arrangement performance was found to have a very large range. 

In conclusion, we at Norfolk Southern view proper trackside lubrication as a continuing source 
of opportunity. The work we have done only improves the performance of old technology. We are 
hopeful that the 21st century brings us lubrication methods representative of current and future tech- 
nology. 

Reference 

Elkins, J. A., Reiff, R.P., and Rhine P.E., "Measurement of Lubrication Effectiveness", 1984. 




• Digs, loads, lifts and grades. • Gets to the job at 55 mph. 
Best-built, hardest working model in the popular 3/4 cu.yd size. 



• Rail gear • Train line air • AAR 
couplers • Axle Vtda for lift 
stability off-rail 



American products built to last 



• Dependable diesel power 

Detroit Diesel, upper structure - 

Cummins, carrier • Erqonomic cab 

design with joystick controls 



P.O. Box 798, Airport Industrial Parit • Winona, MN 55987 • Phone (507) 454-8549 - Fax (507 )454-3326 
Call or fax for detailed specifications, capacities and performance options. 




StarTrmk? 



The railroad problem in your backyard might seem as tough as 
building a line on the moon. 

A&K welcomes your railroad material problems as our challenges. A&K 
provides more than just materials. We offer real "Rail Solutions". A&K problem 
solvers understand that we need to do more than offer quality new and relay 
track materials, welding services, panelizing, accessories and track 
removal. We must also give the best service . 

A&K does it all. A&K has it all. 

Even if your rail problems seem 
as big as the universe ... A&K's "Track Stars" 
will find an on-time solution. 

Call an A&K problem solver today. 

© 1993 A&K Railroad Materials Inc. 



A&K Railroad Materials, Inc. 

Corporate Headquarters 
1 505 South Redwood Road 
P.O. Box 30076 
Salt Lake City, Utah 84130 
Phone: (801) 974-5484 or 
Toll Free: (800) 453-8812 
FAX: (801) 972-2041 

RAIL SOLUTIONS On-Time!. ..On-Track! 



Mi 



(^ 



THE PHILOSOPHY AND DEVELOPMENT OF 
AREA SEISMIC DESIGN CRITERIA 

By: Kenneth L. Wammel*, Zolan Prucz** and Roger S. Boraas*** 



Abstract 



Formulation of seismic design criteria for railroads, as part of the American Railway Engi- 
neering (AREA) Manual of Standard Practice, is a recent and ongoing development. The philosophy 
of railroad managements regarding post seismic event disruptions to operations and acceptable struc- 
tural damage can differ from that of highway departments or other public agencies. The objective of 
AREA Committee 9 is to establish a performance based approach to seismic design of railroad struc- 
tures, that incorporates three levels of ground motion and considers the ability of railroads to control 
post seismic event operations. 

Introduction 

For several decades, earthquake related damage has received significant attention. Seismic 
events can and have occurred in most areas of North America. In particular, the western United States 
continues to experience noticeable seismic events almost on a daily basis. Some highway and build- 
ing failures have been extensive, with many fatalities. This has prompted considerable interest within 
the engineering profession and has triggered political interest and some legislation. 

During past earthquakes, railway structures have generally performed very well with a few 
bridges suffering major damage but most receiving only minor, superficial or no damage. However, 
becau.se of the attention given to highway failures, railroad .structural survival was generally over- 
looked until recently. Public agencies have been pressuring railroads into incorporating seismic 
designs, for new con.struction and retrofit applications, based on highway criteria. In some cases, leg- 
islation has been pas.sed mandating various seismic designs be implemented into railroad structures. 
Railroad bridge engineers have always been vitally interested in maintaining reliability in their infra- 
structure to ensure safety of passengers, customers' goods, employees and the public at large. The dif- 
ference in response of highway and railroad structures to earthquake forces promoted considerable 
debate on the application of engineering principles to designs. But, until recently, almo.st no research 
had been conducted to study railroad structural behavior The fact that railroad structures have his- 
torically survived seismic events, when adjacent highway .structures failed completely, has not been 
adequately considered in addressing new design requirements. 

Typical railroad bridges in North America are very simple in their design and construction. 
They have relatively short spans and large pier or cap top areas that allow for considerable longitu- 
dinal and lateral movements. The design live load to dead load ratio, for railroad bridges, is signifi- 
cantly higher than live load to dead load ratios for other bridge .structures. The longitudinal and lat- 
eral design load requirements provide for a strong and stiff structure. In addition, the track structure 
provides longitudinal continuity over piers and abutments that restrains and dampens superstructure 
longitudinal movements. Railroad bridge designs are usually not restricted by geometric constraints, 
as are highway interchange bridges for example, and their configurations are mainly based on struc- 
tural and functional criteria. 

The performance requirements for railroad bridges, such as the acceptable damage and restora- 
tion of service time, can vary with the bridge type, location, replacement value, inspection and main- 



* Chic!' Engineer .Structures, Union Pacific Railroad, Omaha. NE. 

* A.s.sociatc. Modjcski and Ma,stcr.s. Inc.. New Orlcan.s. LA 
"Diivclor .StiTjctures Design. Union Pacific RaiUoad. Omaha. NE 



77 



78 Bulletin 760 — American Railway Engineering Association 



Icnance procedures, detour availability, type of service and bridge occupancy rate. For example, 
depending on train speed and operating practices of a particular railroad, vertical and/or lateral dis- 
placement of a few inches could be within acceptable limits. 

Most railroad bridges have a very low live load occupancy rate (typically less than 5% for a 
heavy haul line). A centralized signal and communication .system that can control train movements 
allows for an effective railroad post seismic event response. Warning systems such as motion detec- 
tors, inclinometers and other devices may also be tied in to the signal system to assi.st in determining 
the need for in.spections and/or train movement restrictions. Some railroads have connections with 
various .seismic and geological centers that provide early data on seismic activities. The low occu- 
pancy rate and the ability of the railroad to stop or reduce the speed of traffic over a particular seg- 
ment of track until inspections are made are important factors, unique to railroad bridges, that need 
to be incorporated in the seismic design and evaluation criteria. 

In 1994, a group of "volunteers", from the AREA structural committees, met to discuss the issue 
of seismic design of railroad structures. The AREA Manual did not address the issue and there was 
interest among railroad engineers and designers of railroad structures to understand the historical 
record of bridge response to seismic events. Also, there were increasing pressures from some gov- 
ernmental agencies for railroads to accept seismic designs and retrofits based on highway bridge 
characteristics and research. This ad-hoc committee produced a general basis for what is now Chapter 
9. Committee 9 was formed to further develop and refine the ad-hoc committee's work. 

Performance Based Approach to Seismic Design 

The purpose of Chapter 9 is to provide a framework to evaluate the effect of seismic forces on 
railroad bridge structures and help reduce damage to railway facilities. The objectives were estab- 
lished as follows: 

1. Evaluate past experience. 

2. Develop post seismic event operation procedures based on historical precedent and compat- 
ible with structure performance. 

3. Establish well-defined performance requirements that account for the unique features of rail- 
road structures. 

4. Establish analysis requirements consistent with current methods of railroad structure evalu- 
ation and design. 

5. Allow flexibility to address the specific condition, value and importance of a structure. 

6. Provide a basis for more detailed analysis for special projects. 

7. Promote research and testing to validate past experience and support any new design 
requirements. 

8. Provide flexibility for future code updates. 

To achieve these objectives, a three-level ground motion and performance criteria approach, 
consistent with railroad post seismic event response procedures, was developed, as shown in Table 1 . 

Table 1. Three-Level Ground Motion and Performance Criteria Approach to Seismic Design 



Railroad Ground Performance Methods of Analysis 

Response Motion Criteria Limit Methods of Analysis 

Level Level State and Detailing 



II I Serviceability Elastic 

III 2 Ultimate Capacity/Conceptual 
III 3 Survivability Conceptual 



Paper by Kenneth L. Wammel, Zolan Prucz and Roger S. Boraas 79 



Each ground motion level is associated with specific bridge performance requirements and a 
specific railroad response level. Table 1 shows the railroad response level and the performance crite- 
ria limit state in relation to the ground motion level. Guidelines for railroad response after an earth- 
quake, and analysis and design approaches for satisfying each performance criteria limit state, are 
recommended. The railroad response levels consider the earthquake magnitude and distance from 
railroad facilities and are described more fully in Chapter 9. 

Ground Motion Levels 

The ground motion levels reflect the seismic hazard at the site and are directly related to the 
design (or evaluation) earthquake loads. They are defined in terms of expected peak ground acceler- 
ation values associated with a given average return period. Several acceleration coefficient maps may 
be used to interpolate for a selected ground motion return period. The average return period for each 
ground motion level is determined based on acceptable risk criteria and .structure importance classi- 
fication, as shown in Table 2. 



Table 2. Ground Motion Levels for Seismic Design 



Ground Structure 

Motion Acceptable Importance Return Period 

Level Description Risk Criteria Classification (Years) 



1 Moderate Life Safety 1-4 5-100 

2 Large Economics 1-4 200-500 
^ Severe Economics 1-4 1000-2400 



Ground Motion Level 1 represents a moderate earthquake with a reasonable probability of 
being exceeded during the life of the structure. After such an earthquake, trains may proceed at 
restricted speeds until inspections are completed and track released for full speed operation. 
Therefore, the design criteria for Ground Motion Level 1 needs to ensure that the structure is safe and 
serviceable immediately after the earthquake. Ground Motion Level 2 has a low probability of being 
exceeded during the life of the structure, and it represents a larger magnitude earthquake. Ground 
Motion Level 3 has a very low probability of being exceeded during the life of the structure, and it 
represents a rare and severe earthquake. Train traffic is stopped after both Ground Motions Levels 2 
and 3 until property can be inspected, as per Railroad Response Level III. Therefore, for Ground 
Motion Levels 2 and 3, the performance and the acceptable risk criteria can be based mainly on eco- 
nomic considerations. 

Acceptable Risk Criteria 

Earthquakes are extreme events associated with a great amount of uncertainty and therefore risk 
considerations are an integral part of seismic design. The greatest source of uncertainty is associated 
with the regional seismicity and the expected ground motion characteristics at the site. Differences 
between the calculated and the actual seismic response of a structure also add to the general degree 
of uncertainty. To achieve a balance between seismic risk and costs associated with risk reduction, a 
certain amount of risk must be considered as acceptable, unless there is a severe social penalty as.so- 
ciated with structure failure. 

When determining acceptable risk levels, both life safety and economic aspects need to be con- 
sidered. Obviously, the amount of risk that may be acceptable for some bridge structures is greater than 
for others. Factors such as the volume and the type of train traffic, the value and the importance of the 
bridge and the cost of loss of use have to be included. Acceptable seismic risk levels must also be con- 
sistent with the risks due to other extreme events, such as flood, fire and vehicle or vessel collision. 



80 Bulletin 760 — American Railway Engineering Association 



Structure Importance Classification 

The importance classification of a structure along with the acceptable risk considerations deter- 
mines the average ground motion return period to be used for the evaluation of each performance cri- 
teria limit state. It is expressed as a factor that can vary from 1 to 4 and includes the relative contri- 
butions of aspects related to bridge location, traffic over and under the bridge, the value of the bridge, 
and detour availability. Weighting factors for these contributions are assigned to each performance 
limit state. They may vary to represent specific railroad and project requirements. 

Structure Performance Criteria 

The structure performance criteria include the following limit states: 
Sennceability Limit State: 

The serviceability limit state contains restrictions on bridge stresses, deformations, vibrations and 
track misalignments due to a Level 1 Ground Motion. The structure is required to remain in the elastic 
range and only moderate damage, that does not affect the safety of trains at restricted speeds, is allowed. 

Ultimate Limit State: 

The ultimate limit state ensures overall structural integrity during a Level 2 Ground Motion. 
The strength and stability of critical members are of main concern. The structure is allowed to 
respond beyond the elastic range but displacement, ductility and detailing requirements need to be 
satisfied to reduce damage and loss of structure use. The damage should occur as intended in design, 
be readily detectable by visual inspection and accessible for repair. 

Sun'ivability Limit State: 

The survivability limit state is concerned with the survival of the bridge structure after a Level 
3 Ground Motion. Extensive damage, short of bridge collapse, may occur. Structural and geometric 
safety measures, which add redundancy and ductility, are recommended to reduce the likelihood of 
collapse. Other measures designed to prevent collapse in case of serious damage, such as wide bear- 
ing support areas, catcher or backup systems, may be included. Depending on the importance and the 
replacement value of a bridge, an individual railroad may allow irreparable damage for the surviv- 
ability limit state, and opt for new construction. 

Conceptual Approach to Seismic Design 

Elastic analysis methods are recommended for satisfying the serviceability limit state and a 
conceptual approach is recommended for the ultimate and the survivability limit states. The concep- 
tual approach proposed consists of seismic guidelines regarding structure type, foundation type, con- 
figuration and layout, connections, materials, ductility, redundancy, deformation capability and fail- 
ure mode control. By using conceptual seismic design principles along with capacity design methods, 
the engineer can overcome many of the uncertainties involved in the ground motion description, the 
numerical analysis of structure response in the post yield range, and the limited analytical and exper- 
imental seismic research data on railroad bridges currently available. 

Existing Bridges 

The general seismic design criteria approach may be applied to existing bridges. However, 
some aspects are unique to existing bridges and require special consideration. First, to keep the ana- 
lytical efforts to a manageable size, a preliminary seismic screening is recommended to identify the 
most vulnerable bridges. Second, the selection of ground motion return periods, especially for 
Ground Motion Levels 2 and 3, should consider the age, value, load capacity and remaining service 
life of the bridge evaluated. Third, since the current state of knowledge of response analysis of exist- 
ing railroad bridges to severe ground motion is limited, field information and observations from pre- 
vious earthquakes are very important. Bridges selected for retrofit need to undergo a more detailed 



Paper by Kenneth L. Wammel, Zolan Prucz and Roger S. Boraas 



evaluation that includes the effects of the retrofit on the bridge response. The retrofits considered 
should address the cost-benefit of risk reduction, especially regarding the ultimate and the surviv- 
ability limit states. 

Current Committee Assigmnents 

Presently, the subcommittees of Committee 9 are working on the following topics: 

• Expansion and improvement of the Structure Importance Classification section 

• Expansion of design criteria for the serviceability limit state 

• Expansion of conceptual approach criteria for ultimate and survivability limit states 

• Definition and improvement of the preliminary screening to identify the existing bridges 
needing more detailed analysis 

• Development of a post seismic event inspection checklist 

• Development of general guidelines for retrofit designs 

Recently, a new assignment "Track Contribution to Seismic Response" has been undertaken by 
Subcommittee 2. It is a widespread belief that the track plays an important role in the excellent earth- 
quake survival record of railroad bridges. However, no current design criteria allow for that consid- 
eration because it cannot yet be quantified. Therefore, more testing and research of the role of track 
(or other aspects of bridges unique or predominant to railroads) in structural resistance to seismic 
loads is important for further development of detailed design and evaluation criteria. Without hard 
data, no design formula can accurately account for the positive contribution of the track. 

The Future 

The committee has currently focused much of its attention on bridges. The need to understand 
and quantify the differences between railroad and highway bridge response to seismic loading has 
been the driving force behind the committee's work. But other aspects of railroad infrastructure will 
also be addressed. Signal and communication facilities are very susceptible to failure from seismic 
events. Recommended practice for design of these facilities should be included. Building failures, 
subgrade failures and track geometry deviations are also part of the seismic scenario. However, build- 
ings and subgrades are not unique to the railroad and can be covered by many current design speci- 
fications. Track is a very flexible structure, easily restored, and it's response is heavily dependent on 
the subgrade reaction. At this time, it is not anticipated that any new criteria will be developed for 
track design. 

Recent discussions between the FRA and the Ministry of Transport of Japan may lead to an 
agreement for cooperative effort between the two countries. This will provide for information shar- 
ing and the formation of joint, engineering teams for post seismic event investigations into structural 
performance. 

Summary 

The seismic design and evaluation criteria recommended consider the unique structural and 
operating characteristics of railroad structures and the specific needs of railroad bridge owners. The 
approach used is based primarily on detailed performance requirements, and on conceptual design 
and detailing practices. The railroad post seismic event response procedures, along with the type and 
amount of acceptable damage for a given structure, are an integral part of the design and evaluation 
process. A three-level ground motion and performance criteria approach is employed to reduce train 
interruption and ensure structure serviceability after a moderate earthquake, minimize the cost of 
damage and loss of structure use after a large earthquake and prevent structure collapse after a very 
severe earthquake. 




Reliability. Durability. Low maintenance. 



These are key benefits you receive by investing in Miner's high quality 
railcar components. Backed by exhaustive R&D testing, comprehensive 
service analysis and over 100 years of actual rail experience, Miner 
products are components you con count on. 

Draft Gears 
Miner draft gears including the TF-880, Crown SE ', Crown TG " and 
SL-76 provide higher capacity and greater payloads for a wide range 
of service requirements. Certified under various A. A. R. M-901 spec- 
ifications and A. A. R. quality certification M-1003, Miner draft gears 
provide superior cushioning and increased protection of your railcars. 

Side Bearings 
TecsPok'' Constant Contact (TCC) side bearings, featuring Miner's 
patented TecsPok elastomer pads, are designed to reduce truck 



hunting, increase car stability and lower component wear on various 
types of cars. TCC-III side bearings, specifically engineered lor long- 
travel applications, are available in various preloads and maintain 
operation preload at temperature extremes. 

Discharge Devices 
Miner's heavy-duty unloading systems offer reliable, trouble-free 
discharging of coal, groin, chemicals, cement, fxillast and other 
commodities. Whether you need to unload covered hoppers, open-top 
hoppers, coal cars or ballast cars, Miner has the discharge device that 
allows you to do it faster, safer and more effectively. 
Contact your Miner representative today for products that are reliable, 
durable and reduce your maintenance costs. Invest in products from 
W. H. Miner. .roilcor components you can count on! 



^^^ Products Engineered for 21" Century Performance 

hAXHSH W. H. MINER DIVISION 



1200 Eost state street, P.O Box 471. Genera, IL 60134 • 1-630/232-3000 • FAX 1-630/232-3055 

P O Box 178. Lochine, Quebec, ConadaH8S4A6" 1-514/637-0445 • FAX 1-514/637-2126 

3 ch de 10 Procession. B-7870 Lens. Belgium . 32/68/464737 . FAX 32/66/454733 

Ruo Jose Benedettl. 18. 09531-000. Soo Coetono do Sul. Brasll ■ 55/1 1/453-6431 . FAX 55/1 1/442-1787 



© 1996 Miner Enterprises. Inc. 



82 



FRACTURE TOUGHNESS TESTING OF 
AREA GRADE B HAND TOOL STEEL 

By: C. P. Lonsdale* and J. J. Lewandowski** 



Abstract 



Fracture toughness testing was conducted on a number of railroad maintenance striking hand tools 
in order to determine their resistance to crack propagation in the presence of a flaw. Specimens for 
toughness testing were removed from different spike mauls as well as sledge hammers. In the former, 
four of the samples were removed from near the large end striking face of four different spike mauls. 
In the latter, four samples were taken from near each striking face of two different ten pound double 
faced sledge hammers. Samples containing a fine diamond wire saw notch established a notch tough- 
ness of 50.3 MPaVm (45.8 ksiVin) for one sledge sample and 66.0 MPaVm (60.1 k.siVin) for one maul 
sample. Fatigue precracking of the samples to produce a very fine crack prior to testing resulted in sub- 
stantially lower fracture toughness values. The average value for three precracked maul samples was 
45.2 MPaVm (41.1 ksiVin) while the three precracked sledge samples establi.shed an average toughness 
of 31.5 MPaVm (28.7 ksiVin). The fracture toughness of AREA Grade B steel hand tools and some vari- 
ables that affect toughness is discussed. The importance of routine and reliable field tool inspections is 
highlighted due to the relatively low resistance of the steel to fast fracture when a crack is present. 

Introduction 

Sledge hammers and spike mauls are essential hand tools for track maintenance in the railroad 
industry. Although automated maintenance of way equipment has resulted in fewer employees now 
using such striking hand tools, there will always be the need for manual track repairs. Rail replace- 
inents, tie change outs, etc., insure that these tools will be around for many years. However, when 
these tools strike an off center blow, the chance of a chip or spall is greatly increased, since the impact 
contact area is very small and, therefore, the local stress is very high. The striking face contour, or 
comer radius, is particularly vulnerable to such damage. The resistance to spalling has been reported 
to be improved by utilizing a gentle, carefully blended contour (Ref. I). 

Conrail's Technical Services Laboratory examines sledge hammers and spike mauls which fail 
in service. Most often these tools have been in the field for an extended period of time and have 
cracking or deformation evident on the striking face or comer radius. Further impact loading during 
the use of the tool, particularly by an off center blow, often results in the fast fracture of a steel chip. 
Injuries to Conrail Maintenance of Way employees resulted in our investigation of the fracture tough- 
ness of the steel used in North American railroad sledges and mauls. 

Background 

The specifications which govern railroad percussion tools are contained in Chapter 5 (Track) of the 
American Railway Engineering Association's Manual for Railway Engineering (Ref. 2). The steel used 
in striking and stmck tools must be a Grade A or Grade B chemical composition, as shown in Table I. 

The more common alloy used today for sledges and spike mauls is AREA Grade B steel. This 
steel most closely matches the AISI 9260 designation, with vanadium and molybdenum added. The 
striking faces are required to have a hardness of 51 to 55 Rockwell C and this hardness specification 
is maintained to at least one half of an mch below the face. The tools are forged, quenched and tem- 
pered products produced by several manufacturers. Typical tensile test data provided by a manufac- 
turer and obtained on one cylindrical specimen taken from a spike maul and one cylindrical speci- 



* Moiallurgical Engineer. Conrail Technical Services Laboratory. Alloona. PA. 
**Prole.s.sor of Materials Science and Engineering, Case Western Reserve University. Cleveland. OH. 



83 



S4 Bulletin 760 — American Railway Engineering Association 



Table 1. AREA percussion hand tool chemistry specifications. 



Steel 


C 


Mn 


P 


S 


Si 


V 


Mo 


Grade 


Min Max 


Min Max 


Max 


Max 


Min Max 


Min Max 


Min Max 


A 


0.56 0.64 


0.75 1 .00 


0.025 


0.025 


1.80 2.20 


— — 


— — 


B 


0.5 1 0.60 


0.75 1.00 


0.025 


0.025 


1.80 2.20 


— 0.45 


0.35 0.50 



men taken from a sledge hammer are contained in Table 2 (Ref. 3). The threaded end tensile speci- 
mens were electrodischarge machined from the steel between the tool eye and outside wall diameter 
for each type of tool. No comparative evaluations of steel chemistry, cleanliness, microstructure, etc., 
were conducted in order to explain the large differences in sledge and maul sample properties. 
Although our sample size is very limited (one tensile sample for each tool type), it is possible that 
significant differences in raw materials and/or finished tools exist. It appears that an investigation of 
these variables may have benefits for industry. 

Scope of Evaluation 

One way to determine the fracture resistance of Grade B steel involves determining the fracture 
toughness on specimens containing a fatigue precrack in accordance with accepted procedures for 
fracture toughness testing. Subsequent comparisons of these fracture toughness values with those of 
other steels will provide information on the range of strength/hardness/toughness values available in 
other commercially available high strength steels. In addition, knowledge of the fracture toughness 
in combination with some estimate of the applied stresses enables an estimation of the critical crack 
size for catastrophic fracture. 

Experimental Procedure 

Four new spike mauls and two new ten pound sledge hammers were obtained from a group of 
tools on hand at Conrail. All tools were produced by the same manufacturer, Woodings-Verona Tools 
Works. Three point bend fracture specimens were machined from the large end of each spike maul 
and from both ends of the sledges. The samples were taken from just below the center of the striking 
face and parallel to it in each case. The specimens were electrodischarge machined with the crack 
propagation direction away from the striking face center into the body of the forged tool. Examinations 





Table 2. Spike 


maul and sledge hammer tensile test data. 




Tool 
Type 


Tensile Strength 

(psi) 


Yield Strength 
(psi) 


Percent Reduction 
In Area 


Percent 
Elongation 


Maul 
Sledge 


221,000 
256,000 


176,000 
214,000 


25 
18 


8.5 

7 



Table 3. Tool and sample identification data. 



Tool 


Fracture Sample 


I.D. 


Year and Month Tool Produced 


Sledge 1 


lA, IB 




August, 1993 


Sledge 2 


2A, 2B 




December. 1994 


Maul 1 


1 




October, 1995 


Maul 2 


2 




October, 1995 


Maul 3 


3 




October. 1995 


Maul 4 


4 




December, 1995 



Paper by C. P. Lonsdale and J. J. Lewandowski 



of fractured hammers and spike mauls often show that cracks start at the comer radius adjacent to the 
striking face. However, it was felt that the precracked samples would provide an adequate measure- 
ment of the steel's toughness. Table 3 contains tool and sample identification data. The general 
dimensions of the four sledge hammer and four spike maul toughness specimens are shown schemat- 
ically in Figure I . 

Prior to any testing, the sides of the specimens were polished to a mirror finish to aid in crack 
detection. In order to first determine an estimate of the fracture toughness, one (1) toughness speci- 
men of the sledge hammer and the spike maul were tested in the notched condition (i.e. no fatigue 
precrack) in three point bending. The notch was placed in the top surface of the toughness specimen 
using a slow speed diamond impregnated wire saw, producing a notch root radius of approximately 
50 mm. These specimens were notched to a depth aAV of approximately 0.4. A clip gage was placed 
across the notch opening in order to record the notch opening displacement during testing. Tests were 
conducted on a servo-hydraulic testing machine operated under displacement controlled conditions 
at a displacement rate of 0.5 mm/min, while both the load and the notch opening displacement were 
recorded. At the conclusion of the test, the peak load achieved prior to catastrophic fracture was used 
to estimate the toughness in the manner outlined in ASTM E-399. 

The remaining six sledge hammer and spike maul specimens were each notched to an aAV = 
0.25-0.33 prior to fatigue precracking, in an attempt to conduct fracture toughness testing on fatigue 
precracked specimens, in general accordance with the procedures outlined in ASTM E-399. Notching 
was again conducted with the slow speed wire saw, followed by fatigue precracking in accordance 
with ASTM E-399. In all ca.ses, attempts were made to precrack the specimens to an aAV = 
0.40-0.65, in accordance with ASTM E-399. The outside surfaces of the toughness specimens were 
visually monitored during precracking in order to monitor the progression of the fatigue crack. Load 
shedding was conducted after the crack grew a specified amount in accordance with ASTM E-399. 
In some cases, the specimens failed during fatigue precracking. In those cases, the fatigue precrack 
length was noted via optical examination of the fracture surface and the toughness was calculated 
using the aAV and the peak load from the fatigue test. In other cases, the specimens were success- 
fully precracked and the fracture toughness test was conducted in accordance with ASTM E-399. 
Specimens successfully fatigue precracked were tested in three point bending with a clip gage placed 







:^<#ir 





Figure 1. Example dimensions and diagram of fracture toughness specimens. 



86 Bulletin 760 — American Railway Engineering Association 



across the crack mouth in order to monitor crack opening displacement. The loading rates and test 
procedures were as those outlined above. 

Results and Discussion 

The results of the toughness testing are shown in Table 4. The data clearly show that the pres- 
ence of a fatigue crack in the specimens produces a lower toughness than that when a notch is pre- 
sent. The notch produced with the diamond wire saw provided a notch root radius of about 50 mm, 
which is a less severe defect than that produced by a sharp fatigue crack. As expected and as obtained 
on many other materials, the toughness of the sledges dropped from 50.3 MPaVm (45.8 ksiVin) with 
the notch to an average value of 31.5 MPaVm (28.7 ksiVin) with the fatigue precrack. The notch 
toughness of the mauls was 66.0 MPaVm (60.1 ksiVin) while precracked specimens again exhibited 
a reduced toughness value. Although the spike mauls failed during precracking, the estimated tough- 
ness was as low as 34.9 MPaVm (31.8 ksiVin), somewhat higher than that obtained on the sledge 
hammer specimens. Although it is difficult to conduct .standard ASTM tests on such specimens 
because of their small size and difficulty in fatigue precracking, the results obtained on the notched 
specimens provide an estimate of the fracture toughness, while those obtained with the sharp fatigue 
precrack cleady show that the toughness is reduced when a sharp crack is present. 

There also appears to be a difference in both the notch toughness and precracked toughness 
between the sledge hammer and spike maul specimens. Assuming that identical starting material 
compositions are used, one potential source of this difference relates to the differences in manufac- 
ture of the two products. The finished spike maul section diameter on the large end (1% inches) is 
smaller than the finished section diameter of the ten pound sledge (2'/: inches). A review of the man- 
ufacturing process reveals that mauls are forged from a 2 inch diameter round bar whereas ten pound 
sledges are forged from 2Vi inches diameter round stock (Ref. 3). As such the reduction applied to 
the former material during forging is much greater than that applied to the latter. One manufacturer 
reported that a ten pound sledge receives only two forging blows during manufacture while a spike 
maul receives eight to ten forging blows (Ref. 3). Such differences in forging reduction could produce 
beneficial changes to the microstructure in the spike mauls that would affect the strength/toughness of 
the quenched and tempered material. The subsequent heat treatment of such tools and their different 
dimensions also produces a final hardness profile which is somewhat different for the two products, 
while changes to the tempering treatments could also produce changes to strength/toughness combi- 
nations possible in such materials. The spike mauls which contain the smaller section size should 
exhibit a more uniform microstructure and a high hardness should be maintained to a greater percent- 
age of the tool cross-section if both products were given an identical quench, prior to tempering. 

Table 4. Results of fracture toughness testing. 











Toughness 


Tool Type 


Sample I.D. 


Notch/Precrack 


Mpa Vm 


ksi Vin 


Sledge 1 


lA 




Notch 


50.3 


45.8 


Sledge 1 


IB 




Precrack* 


30.5 


27.8 


Sledge 2 


2A 




Precrack 


28.9 


26.3 


Sledge 2 


2B 




Precrack 


35.1 


31.9 


Average Value 


For Precracked Slec 


ge Samples 




31.5 


28.7 


Maul 1 


1 




Notch 


66.0 


60.1 


Maul 2 


2 




Precrack* 


34.9 


31.8 


Maul 3 


3 




Precrack* 


52.6 


47.8 


Maul 4 


4 




Precrack* 


48.1 


43.7 


Average Value 


For Precracked Mai 


1 Samples 




45.2 


41.1 



Note: *Failed during precracking. 



Paper b\ C. P. Lonsdale and J.J. Lewandovvski 



87 



Figure 2 shows the macroetched structure of a commercial Grade B steel railroad spike maul 
which has been longitudinally sectioned (Ret'. 3). The top maul section is From the small end ( 1 % inches 
striking face diameter) while the bottom section is from the large end (1% inches diameter). The 
amount of visible centerline segregation is clearly greater in the larger end section. The smaller end 
that obtained more mechanical working during forging, does not exhibit a pronounced centerline streak. 
Such segregation, which results from solidification of the steel raw stock could then affect the frac- 
ture properties of railroad hand tools if such regions are loaded under high stress conditions. Such sites 
could be preferential sites for both crack initiation and growth. 

The estimates of precracked fracture toughness values reported in Table 4 enable one to esti- 
mate the combinations of stress and crack size which will produce catastrophic fracture in a struc- 
ture. In the case of the materials examined presently, it is assumed that a crack propagates inward 
from the surface of the hammer or spike maul in the manner approximated by the fracture toughness 
specimens tested. The general relation between the fracture toughness and the combinations of stress 
and crack size required to produce catastrophic fracture is summarized in the following equation: 

Kn = geometric factor x ct ( IT a) 

where K^ is the fracture toughness, a is the applied stress, and a is the crack length. Since this 
calculation is provided presently as an estimate, the geometric factor will be set equal to 1 . The equa- 
tion indicates that, for a given level of fracture toughness, an increase in the applied stress will reduce 
the critical flaw size for catastrophic fracture. The equation also indicates that an increase in the crack 
size present in a structure will reduce the stress that can be applied to the structure prior to cata- 
strophic fracture. As an example, assuming that a tensile stress equal to the yield strength of the tool 
material (e.g. 176 ksi) was applied to a tool material with toughness of approximately 33 MPaVm 
(30 ksi\'in). the critical flaw size required for catastrophic fracture would be approximately 0.24 mm 




Figure 2. Macroetched spike maul. 



88 Bulletin 760 — American Railway Engineering Association 



(0.00925 inches). Although the magnitude of the flaw size obtained by such a calculation will be 
affected by a number of variables beyond the scope of this report, it is clear that the combinations of 
stress and toughness cho.sen in the calculation produce a relatively small flaw size that is required for 
catastrophic fracture. This further suggests that frequent and careful inspections for such flaws before 
and after u.se might be beneficial in preventing chip spallation during use. 

The calculation above also indicates that improving the strength/toughness combinations of 
properties of a material will increase the critical flaw size in a given situation. For example, an 
increase in the toughness of the material to 65.9 MPaVm (60 ksiVin) while maintaining the same 
applied stress will produce an increase in the critical flaw size by a factor of 4, while a decrease in 
the toughness to 16.5 MPaVm (15 ksiVin) will reduce the critical flaw size by a factor of 4. Although 
detailed calculations of the stresses present during the use of such tools is necessary to refine the cal- 
culations noted above, the simple calculations provided indicate that inspection of tools should be 
conducted routinely, while the use of materials/heat treatments with higher strength/toughness com- 
binations may also be beneficial in the safe use of such tools. An additional concern with the use of 
such steel tools in cold weather environments, is the reduction in fracture toughness that typically 
accompanies testing a steel at temperatures below room temperature. Such a reduction in fracture 
toughness will reduce the critical flaw size, as summarized above. Finally, the dynamic fracture 
toughness obtained under impact conditions is often lower than that obtained under static conditions 
as that conducted presently. 

Preliminary testing at Conrail using a ten pound sledge hammer and instrumented drift pin has 
shown that very high peak forces are experienced during hand tool impact. The average force 
imparted by a man during a series of blows was approximately 37,000 pounds. When the force of a 
blow is converted to an axial stress at the striking face contact area (e.g. '/« square inches, 0.2 inches 
circular radius) this will lead to severe stresses in excess of the steel's yield stress. This then results 
in high radial tensile stresses which can be estimated using poisson's ratio and the modulus of elas- 
ticity. Provided that the stress is centered and uniformly applied to the striking face, damage will 
likely be of a compressive, peening nature. However, when a glancing blow is struck by a user, par- 
ticularly at the tool's corner radius, the area of contact is smaller, the stresses are greater and the like- 
lihood of crack nucleation and/or growth leading to fast fracture is increased. 

Conrail has recently begun a test program using the Hammer Impact Tester (HIT), an electro- 
pneumatic test apparatus that is able to repeatedly strike a percussion hand tool on various surfaces 
(drift pin, spike, etc.). Initial calibration efforts have focused on making sure that the device imparts 
a blow similar to that of a man. The testing is expected to provide additional information with regard 
to impact stresses, critical allowable flaw size and tool impact fatigue life under various conditions. 

Finally, the factors which control the strength and toughness of steels are quite complex, 
although a listing of some typical properties for some common steels is provided in ASM Handbook - 
Volume 19 (Ref. 4). An inspection of such information indicates that a range of strength/toughness 
combinations are possible in a variety of available steels. A remaining issue involves the relative cost 
of such materials for use in tool applications as well as the effects of changes in test temperature on 
the strength/toughness properties. It may be useful to review the material selection and service expe- 
rience of such tools in the light of the above concepts. 

Concluding Remarks 

It is clear from the fracture toughness results and estimated critical flaw size calculations that 
routine and reliable field inspections of the sledge hammers and spike mauls used by railroad main- 
tenance of way employees is important to the safe use of such tools. Given the relatively low frac- 
ture toughness of currently produced, commercially available, percussion hand tools, and the small 
crack size that can produce fast fracture in a tool under loading, it is apparent that even a small flaw 
can lead to a chip or spall. It is strongly recommended that all railroad hand tools be inspected for 



Paper by C. P. Lonsdale and J. J. Lewandowski 



89 



nicks, dents, chips, spalls, cracks, etc., particularly on the striking face and corner radius. Careful 
inspections of tools before, during and after use by tool users, along with detailed, regular tool inspec- 
tions by supervisors, will help insure that defective tools are removed from .service and replaced. 

Additional efforts to improve railroad hand tools can focus on the many variables which affect 
the fracture toughness of high strength steel. Forging processes, alloy, heat treatment, steel cleanli- 
ness and segregation, etc., are potential areas for improvement. A joint initiative between tool man- 
ufacturers and the railroad industry (perhaps AREA Committee 5) to make tools safer for field 
employees will have lasting benefits. Issues such as alloy commercial availability and cost should 
also be addressed. 

References 

Mclntire, H.O. and Hall, A.M. 1962. The performance of hand-striking tools under simulated service 
conditions. Battelle Memorial Institute, Columbus, Ohio. Unpublished report. 

American Railway Engineering Association. 
AREA 



1996. Specification for track tools. Washington, D.C. 



McShane, J. 1996. Private communication, Woodings- Verona Tool Works, Verona, PA. Tensile tests 
and photography conducted by Industrial Testing Laboratory Services Corporation, 
Pittsburgh, PA. Unpublished results. 

Lampman, S.R. 1996. ASM Handbook. Volume 19 — Fatigue and Fracture, ASM International, 
Materials Park. Ohio. pp. 621-624. 



Canada "^ Canada 
Sydney Rails For 

Quality Track 

• Premium Wear-Resistant Head Hardened Rails 

• Standard Carbon, Intermediate Strength and Pre- 
mium Alloy Grades-all using clean steel practice. 

• All National and International specifications, 
including A.R.E.A., C.N.R., BS-11 and U.I.C. 

• Sections from 50 to 70 kilograms per metre (100 to 
136 pounds per yard). 

• Lengths up to 26 metres (85 feet). 

• More than eighty years of experience supplying 
world markets. 

For additional information, write, wire or call: 

Vice President, Sales 
Sydney Steel Corporation 
Sydney, Nova Scotia, Canada 
B1P6K5 



Telephone: 
(902)564-7910 

Telefax: 

(902) 564-7903 

Telex: 

019-35197 



B 



SYDNEY 

STEEL 

CORPORATION 



IF YOU 
BELIEVE... 








...problems have 
solutions. 

If you value 

performance 

over price. 

And, if you want 

service and 

innovative 

►roducts tliat 

vetrac' 



w-?^P' 



PAIMPROU 

PANDROL INCORPORATED 

501 Sharptown Road 
P.O. Box 367 
Bridgeport, N J 08014 
(609) 467-3227 




90 



THE POTENTIAL OF THE CFS MOW ESTIMATING 
RELATIONSHIPS AS A BASIS FOR COST ALLOCATION 

By: Duncan W. Allen, P.E.* 

Abstract 

Many recent proposals for improved intercity and commuter passenger rail transportation 
include the assumption that the improved service will make use of existing trackage, or improved 
trackage in existing right-of-way. These existing corridors often carry moderate to heavy freight traf- 
fic, and may already be carrying passenger trains. Until 1996, estimates of the likely costs for such 
track use by the passenger train operator were often based on the agreements between Amtrak and 
individual railroad companies which had prevailed since Amtrak's inception in 1971. With the expira- 
tion of the original Amtrak agreement, and the increase in the number of proposed commuter rail oper- 
ations, renewed attention is focusing on both the incremental maintenance of way (MOW) costs for pas- 
senger train operation, and on how to reflect the impact of passenger operations on line capacity. 

The Federal Railroad Administration's Commercial Feasibility Study (CFS) was confronted 
with the task of making complete operating and maintenance (O&M) cost estimates for proposed use 
of freight railroad trackage for "Accelerail" passenger services in the 90-150 mph range, including 
MOW costs. Rather than "borrow" from existing agreements or formulas, the CFS developed an 
independent, resource build-up approach to the estimation of dozens of separate MOW activities. The 
total costs built up from the individual cost estimating relationships exhibit certain characteristic rela- 
tionships to both line density and operating speed, which can be expressed mathematically. 

This paper summarizes the structure and assumptions of the MOW component of the CFS 
model, and compares the general nature of the results to two existing bases which have been used by 
prior operating agreements: speed factored gross tonnage (SFGT) and weighted system average cost 
(WSAC). The CFS formulation employs a more complex form, and uses the maximum authorized 
speed (MAS) for passenger train service as a determinant in addition to tonnage. The proposed for- 
mulation of incremental MOW costs resulting from adding Accelerail-type passenger service to an 
existing traffic base: (I) avoids the shortcomings identified with older formulas; (2) incorporates 
insights based on the CFS research; and (3) offers a modular framework for estimating mixed traffic 
MOW costs. 

Introduction 

Many recent proposals for improved intercity and commuter passenger rail transportation 
assume that the improved service will make use of existing trackage, or of improved trackage in 
existing right-of-way. These existing corridors often carry moderate to heavy freight traffic, and may 
already be carrying passenger trains. Until 1996, estimates of the likely costs for such track use by 
the passenger train operator (PTO) were often based on the agreements between Amtrak and indi- 
vidual railroad companies which had prevailed since Amtrak's inception in 1971. With the expiration 
of the original Amtrak agreement, and the increase in the number of proposed commuter rail opera- 
tions, renewed attention is focusing on both the incremental maintenance of way (MOW) costs for 
pas.senger train operation, and on how to reflect the impact of passenger operations on line capacity. 
This paper does not address the issue of capacity, which is an important issue in its own right. 

The Federal Railroad Administration's Commercial Feasibility Study ( 1 ) (CFS) was confronted 
with the task of making complete operating and maintenance (O&M) cost estimates for proposed use 
of freight railroad trackage for "Accelerail" passenger services in the 90-150 mph range, including 



'Principal Transporlaliiin Enjjinccr. Dc Lcuw. Calhcr & Company 



91 



92 Bulletin 760 — American Railway Engineering Association 



MOW costs. Rather than "borrow" from existing agreements or formulas, the CFS developed an inde- 
pendent, resource build-up approach to the estimation of dozens of separate MOW activities. The total 
costs built up from the individual cost estimating relationships (CERs) exhibit certain characteristic 
relationships to both line density and operating speed, which can be closely approximated by formulas. 

This paper summarizes the structure and assumptions of the MOW component of the CFS model, 
and compares the general nature of the results to two existing bases that have been used by prior oper- 
ating agreements: speed factored gross tonnage (SFGT) and weighted system average cost (WSAC). 

CFS MOW Methodology 

CERs were developed for 21 individual MOW accounts as shown in Table 1. The principal 
sources of information to build each CER were expert opinion and secondary sources specific to each 
account. For those CERs where expert opinion differed significantly, the so-called "Delphi tech- 
nique" was employed to reach a synthesis of these opinions. Aggregate cost data from both the rail- 
roads' Form I reports to the Federal Railroad Administration, and rail transit operators' "Section 15" 
statistical reports to the Federpl Transit Administration (FTA) were used primarily to validate the 
CERs at a higher (i.e., account) level. Although there is considerable scope for more research on 
many of the CER accounts, an effort was made to validate as many individual CERs as possible 
against either available data from specific railroads, or against more detailed transit system data pub- 
lished by the FTA (2). In the specific case of the effort required to maintain track to higher operating 
speed and comfort standards for passenger service, three major secondary sources were used: Pier 
and Campbell's 1993 paper on vertical dynamic forces (3), Senac's 1984 paper on the French national 
railway's HSR experience (4), and the tariff structure (5) of the new German railroad infrastructure 
operating entity, Deutsche Bahn AG. It proved possible to derive a mathematical framework incor- 
porating the effects of the first two sources, which was validated by the third. 

The process for developing HSR MOW CERs was as follows: 

• Break down the item, if appropriate, into component activities. The breakdown used is shown 
in Table 1 . 

• For each component, obtain one or more expert opinions on the frequency of the activity 
under various rail traffic scenarios selected to "bracket" the likely range of CFS corridor sce- 
narios, and on the resources necessary to conduct the activity on a unit basis. 

• Develop a mathematical relationship to express the component activities' frequencies in 
terms of a number of potential explanatory variables, including: MGT (millions of gross tons) 
of freight and passenger train traffic over a track segment; the MASs (maximum authorized 
speeds) at which passenger and freight trains are operated; the fraction of track on tangent 
alignment; and the fraction of track with concrete ties. 

• Apply the mathematized frequencies to the expert-estimated unit labor consumption of each 
activity for a wide range of HSR speed and traffic combinations, summing the results for the 
components within each item. 

• Fit a mathematical expression to the summed results with regression techniques, using one of 
a small set of predefined parameterized forms. 

The most frequently used form of expression was the "MAS/MGT' formula, which took the 
generic form: 

Resources per track-mile = a -t- (b -i- cMAS) (d -i- MGT)*^ 

The parameters a, b, c, d, and e in this expression were estimated separately for each CER. In 
several cases, separate CERs were defined for labor and materials, and it was generally found that a 
separate set of parameters was required for each high-speed rail (HSR) technology option evaluated, 
because of the consists' different weights and configurations. 



Paper by D. Allen 



93 



Table 1. MOW Activities in CFS and their Components 



Item 


Activity 


Component(s) 


1201 


Guidevvay physical inspection (curved track- 
concrete tie) 


physical inspection 


1202 


Guideway physical inspection (curved track- 
non-concrete tie) 


physical inspection 


1203 


Guideway physical inspection (tangent track- 
concrete tie) 


physical inspection 


1204 


Guideway physical inspection (tangent track- 
non-concrete tie) 


physical inspection 


1205 


Guideway routine repair (curved track-concrete tie) 


weld surface defects, replace 
failed weld, spot surface, surface 
out of face, clean shoulder, clean 
under concrete tie, renew pads & 
insulators, renew insulated joints, 
replace failed concrete ties 


1206 


Guideway routine repair (curved track- 


weld surface defects, replace 




non-concrete tie) 


failed weld, spot surface, surface 
out of face, clean shoulder, 
renew insulated joints, replace 
failed wood ties 


1207 


Guideway routine repair (tangent track-concrete tie) 


weld surface defects, replace 
failed weld, spot surface, surface 
out of face, clean shoulder, clean 
under concrete tie, renew pads & 
insulators, renew insulated joints, 
replace failed concrete ties 


1208 


Guideway routine repair (tangent track- 
non-concrete tie) 


physical inspection 


1209 


Guideway physical inspection (grade crossings) 


physical inspection 


1210 


Guideway repair (grade crossings) 


routine repair 


1211 


Guideway physical inspection (high speed switch) 


physical inspection 


1212 


Switch repair (high-speed) 


spot surface high speed turnout 


1213 


Guideway physical inspection (low speed switch) 


physical inspection 


1214 


Switch repair (low-speed) 


spot surface low speed turnout 


1215 


Guideway electronic inspection 


geometry inspections, accelero- 
meter measurements, rail flaw 
inspection 


1301 


Guideway program maintenance 


rail renewal, "high" rail 




("old" curved track-concrete tie) 


renewal, tie renewal 


1302 


Guideway program maintenance 


rail renewal, "high" rail 




("old" curved track-non-concrete tie) 


renewal, tie renewal 


1303 


Guideway program maintenance 


rail renewal, 




("old" tangent track-concrete tie) 


tie renewal 


1304 


Guideway program maintenance 


rail renewal. 




("old" tangent track-non-concrete tie) 


tie renewal 


1305 


Guideway program maintenance 
("old" high speed switch) 


turnout renewal 


1306 


Guideway program maintenance 
("old" low speed switch) 


turnout renewal 



94 



Bulletin 760 — American Railway Engineering Association 



Aggregation of CFS MOW Results 

For the CFS, costs for all CERs were estimated by track segment for the corridors and options 
specified by the FRA. These estimates included situations where HSR was to operate on existing 
freight railroads, others where the HSR operation was on dedicated track owned by the HSR operator, 
and combinations thereof. Several options for contracting out selected MOW and other functions also 
were explored; the specific results of the CFS have been presented and discussed elsewhere ( I ), (6). 

For the purposes of cost allocation, the behavior of the CFS CERs for MOW is of interest for 
a "typical" railroad across a range of traffic densities. Figure 1 shows the total MOW costs for the 
activities shown in Table 1, exclusive of management and "overhead" general and administrative 
(G&A) expenses, for a representative freight-only railroad with between I and 30 annual MGT. This 
figure also shows the differential effect of adding a passenger high-speed rail (HSR) service to the 
corridor, both as a "flat" eight trains per day at a MAS of 125 mph, and as a five percent tonnage 
increment at 150 mph. The representative railroad is 70 percent on tangent, is 70 percent concrete 
ties, and is maintained for freight operation at 60 mph. Growth in the estimated MOW costs is slightly 
less than linear with MGT, i.e. small economies of scale in density are exhibited. This trend contin- 
ues to way high densities (e.g., 100 MGT). 

The CFS formulation makes it possible to distinguish two sources of incremental cost growth 
with the introduction of HSR service, as shown in Figure 2. In this figure, a high-frequency HSR ser- 
vice adding just over 2 MGT is introduced. The total increase in MOW costs can be associated in part 
with the increased tonnage; this "incremental tonnage effect" would be the incremental cost of the 
same increase in MGT if it were composed of the same classes of freight traffic already operating in 
the corridor. A second, larger, increase in MOW costs is related to the increased passenger train oper- 
ating speed, and the requirement to maintain superior ride quality at these speeds on track used by 
freight trains. 




■ 100% Freight® 60 mph 
« w/ 8 Ipd HSR @ 125 mph 
A 5% tonnage HSR @ 150 



Annual Density (MGT) 



Figure 1. Example Aggregate CFS Formula Results 



Paper by D. Allen 



95 



Figures 1 and 2 address total costs as estimated by the CERs for a representative railroad. As 
proposals for new passenger rail services on freight railroads continue to be made, attention has 
focused on the incremental, or marginal, costs of providing for such services, and on the economic 
value of the track capacity foregone by the freight railroad in accommodating them. The issue of 
capacity lies outside the scope of this paper; however, incremental costs can be estimated from the 
CERs by comparing the total costs at intervals along a range of traffic densities. 

Figure 3 shows the estimated incremental MOW costs in 1993 dollars for the representative 
freight-only railroad, including the overhead costs of management, MOW facilities, and G&A 
expense. These costs are presented on a per-thousand gross ton-mile basis. This figure also shows the 
incremental costs if a 1 10-mph HSR service is also operating on the railroad. Economies of .scale with 
density are evident, but become relatively small over 10 MGT, and approach negligibility over 25 
MGT. A close fit to the CFS aggregate results, including the HSR, is represented by the expression: 



0.635 



0.772 



(MGT -Hi)' 



(MGT -(-0.24)' 



An important point to bear in mind is that with the HSR service in place, the higher incremental 
cost applies to all traffic, including freight; with the introduction of HSR to a freight-only corridor, the 
differential between the two cost levels is in effect incurred "retroactively" on the freight traffic that 
was operating prior to the addition of HSR. This means that an incremental cost estimate using the 
CFS methods cannot be simply applied to the HSR traffic; in the CFS, the HSR operator was required 
to reimburse the freight railroad for the total incremental cost incurred by the freight traffic as well. 

Comparison with Previous Formulas 

As discussed by Resor (7) and others, two formulas for estimating incremental MOW costs 
have seen extensive use in the United States: the Speed Factored Gross Tonnage (SFGT) model. 



i 

1 
1 

i 


r-^ 


- 


<- Speed and nde quality effect 
<- incremental tonnage effecl 


1 





Freight Only 
wilh HSR 



Annual Density (MGT) 



Figure 2. Example Effects of Introducing HSR 



96 



Bulletin 760 — American Railway Engineering Association 




■ Freight Only 
« With HSR 



Annual Density (MGT) 



Figure 3. Example Incremental Cost Variation by Density 



based on a 1956 study (8) and since adjusted; and the Weighted System Average Cost (WSAC) 
method, developed by ZETA-TECH Associates for the Association of American Railroads. Both 
methods rely on a small number of equations, which reflect some key assumptions inferred from his- 
torical data: 

• SFGT assumes that the combination of fixed costs and those that vary with traffic can be rep- 
resented as being proportional to the square root of density. This assumption has been criticized 
as having inadequate data to support it beyond 25 MGT, and has been revised. The assumption 
of indefinitely improving economies of scale remains inherent to the method, however. 

• WSAC assumes that the variable portion of MOW costs are entirely variable and are linear 
with tonnage over 25 MGT. Below 25 MGT, fixed costs assume an increasing importance, 
accounting for all costs at 1 MGT. The development of WSAC was supported by the "TOPS 
report" (9), which like CFS, used multiple complex expressions to estimate the maintenance 
cycles and lives of various components. 

By independently constructing relationships for a large number of component activities, and 
using a more complex formulation, the CFS avoids making assumptions on the scale of SFGT or even 
WSAC. The particulars of cost variability for each CER are distinct. It is therefore of interest to 
examine how the CER results compare to these formulas. 

A thorough comparison would require that all three methods be employed for the same railroad 
carrying the same traffic. This was not possible as part of the preparation of this paper, but three 
observations can be made: 

• The level of incremental co.sts in Figure 3 lies between systemwide costs advanced by Amtrak 
and Conrail in their 1995 rate case before the Interstate Commerce Commission (ICC) (10). The 
costs advanced by Amtrak using SFGT appear to be 10 to 25 percent less than the level sug- 
gested by Figure 3, while the costs advanced by Conrail appear to be larger by a factor of 3 to 
4. A spread of this magnitude is not unprecedented in such essentially adversarial proceedings. 



Paper by D. Allen 



97 



• The costs for the "representative" railroad in Figure 3 are significantly lower than costs esti- 
mated by Resor (7) using both the SFGT and WSAC methods for a specific unidentified seg- 
ment of railroad, as shown in Figure 4. 

• Beyond 25 MGT, the behavior of the CFS incremental cost can be approximated by the den- 
sity to the -0.088 power, versus the inverse square root (-0.50 power) for SFGT. in the same 
range, WSAC incremental cost is constant with density (0.0 power). 

A number of factors restrict the conclusions that can be reached from these comparisons. First, 
the railroads in question may not be truly comparable to the "representative" railroad in Figure 3. 
Second, the CERs presume, for any tonnage, a railroad maintained for 60-mph freight operation; this 
means that a number of costs that might not be required at lower densities are nevertheless included, 
and therefore may act to reduce marginal costs. Third, the prior methods include ballast renewal, for 
w hich the CFS model does not vary the life cycle with the addition of passenger traffic. Fourth, as men- 
tioned above, the CFS methods .separate out the effects of maintaining speed and ride quality for pas- 
senger service: the effective combined MOW compensation to freight railroads in the CFS scenarios 
where HSR was operated wa.s comparable to that established by the ICC for Amtrak reimburse to Conrail 
in the 1995 case. Fifth and finally, the CFS data synthesis did not extend below the 5 MGT level. 

Conclusions 

The results obtained for a "representative" hypothetical railroad using the CFS model suggest some 
directions for future development, and support a few conclusions even with the limitations cited above: 

• A direct side-by-side comparison of the CFS-derived and prior methods should be made for 
a set of identical assumptions: unless the CFS methods are extended to lower track classes, 
this would have to include an assumption that the track is maintained to at least FRA Class 5 
regardless of density. This would confirm the apparent "middle ground" status of the CFS for- 
mulas relative to WSAC and SFGT, and would permit more substantive conclusions to be 
drawn. 




■ CFS model (see Nole 1) 
« WSAC (see nole 2) 
* SFGT (see note 2) 



Annual Density (MGT) 



'1: Track ciass loed (5); olhefassMnptPons may differ front SFGT/WSAC example I 
2 Trac* dass may vary etampies from Resor paper (see tool7x)le ir te»l| | 



Figure 4. Examples of Incremental Cost Variation by Density 



98 Bulletin 760 — American Railway Engineering Association 



• The CFS model could be developed further to extend its range and scope; its modular con- 
struction facilitates this. Comparison of replacement cycles for individual activities with the 
TOPS research underlying the WSAC, and ongoing research, could lead to improved CERs 
in selected areas. 

• Economies of scale do appear to exist with respect to traffic density, but they are far more 
modest than the SFGT formulation would suggest, and are very small for densities over 25 
MGT. 

• Consideration should be given to the separation of "tonnage effects" and "speed and ride 
quality effects" (as in Figure 2) in new approaches to costing passenger train use of freight 
railroads. The model developed for the CFS offers a framework for doing this. 

In summary, the CERs developed for the CFS provide a systematic framework for evaluating 
MOW costs in the context of a complete forecast of operating and maintenance costs. The basic 
framework can be developed therefrom to support planning efforts, cost projections, and cost alloca- 
tion studies. 

References 

(1) U.S. Department of Transportation, Federal Railroad Administration, High-Speed Ground 
Transportation for America, August, 1996. 

(2) Federal Transit Administration, Estimation of Operating and Maintenance Costs for Transit 
Systems, DOT-T-93-21, December, 1992. 

(3) Pier, Jerome R., and Campbell, Gordon S., High Speed on Existing Rights of Way: The 
Significance of Vertical Dynamic (P2) Forces, presented at the Transportation Research Board 
Annual Meeting, January 1 1,1993. 

(4) Senac, Guy, Increasing Speeds on Existing Lines, French Railway Review, Volume 2, 
Number 2, 1984. 

(5) informational brochure "Die Deutsche Bahn AG offnet den Fahrweg fur Dritte", c. 1994 

(6) Allen, Duncan W., "Cross-Corridor Comparison of Operating Costs for High-Speed Ground 
Transportation", Transportation Research Board 1997 Annual Meeting, Paper No. 971135, January, 
1997. 

(7) Resor, Randolph R., "Passenger Trains on Freight Railroads: Setting the Price For Access", 
1996 Conference on Passenger Trains on Freight Railroads. 

(8) Proceedings of the American Railway Engineering Association, vol. 58, No. 532, 1956. 

(9) U.S. Department of Transportation, Federal Railroad Administration, "Procedures for 
Analyzing The Economic Costs of Railroad Roadway for Pricing Purposes", January, 1976. 

(10) Interstate Commerce Commission Decision, Finance Docket No. 32467, December 29, 
1995. 



he Hidden 
:nemy . . . 




You can't always see the enemy lying 
beneath the surface -fouled ballast 
waiting to combine with moisture to 
destabilize your track. 

A regular program of shoulder ballast 
cleaning helps keep your ballast 
performing as it should. This stand- 
alone operation provides many 
important benefits including: 

• Extends time between costly 
surfacing cycles. 

• Increases life of track components. 

• Helps eliminate the need for 
expensive undercutting operations. 






r 




The Loram Shoulder Ballast Cleaner, 
along with the Loram Badger Ditcher, 
are your keys to a complete , cost- 
effective drainage maintenance 
system. To learn more about how you 
can maintain the stability of your track 
structure, contact: 



LORMM 



^^m 




Nobody builds it tougher. 

Or services it better. 



Loram Maintenance of Way, Inc. 

3900 Arrowhead Drive 
P.O. 80x188 
Hamel, Minnesota 55340 
Telephone (612) 478-6014 
Telex 29-0391, Cable LORAM 
Fax (612) 478-6916 



MANNESMANN 

DEMAG 

GOTTWALD 



EXPERIENCE 

+ 

INNOVATION 

GS 88.79 TR 




RailquiiB, inc 



The employment of economical operation and modern 
teci^nology is the foundation for a safe cost-efficient and 
time saving operation in track and bridge construction. 
The requirements are increasing. That is why tracl< and 
bridge construction should be equipped to face the 
future. 

The GS 88.79 TR Railroad Crane sets the standard in the fol- 
lowing areas: 

Safety 

Carrying Capacity 

Speed 

Environmental Compliance 

Ease of Operation 

Future requirements are already fulfilled today, new appli- 
cations developed. You can make use of the experience 
of a track and bridge construction specialist and the com- 
petence of a leading crane manufacturer. 

If you take us up on it, you will always be on the right track, 



^^^^^fe==^~^ 




•1 


1 

n nil 


fi-fl 












n^ 




-. -.r 


[_: 


— ■ 


.._._,. 


-■ ■-■ '■ — 


..., .,.-. -Y- 


_,_,_ 


3 


t:^|?P " " 


Sf!STaiSi"~«^ya». : 










"^ 


Li/ W-^-iDcri '•t=^ i-:.or-^- \'r-=t^ '■' " 




^^^^^^ 





r 



RailquiiH inc. 

3731 Nortticrest Road, Suite 6, Atlanta, GA 30340 
(770) 458-4157 • Fax (770) 458-5365 



100 



LONGITUDINAL FORCES IN AN OPEN-DECK 
STEEL DECK PLATE-GIRDER BRIDGE 

By: Duane E. Otter*, Joseph LoPresti**, Douglas A. Foutch*** and Daniel H. Tobias**** 

Abstract 

The Association of American Railroads (A AR) recently tested a 50-foot open-deck steel deck 
plate-girder bridge to measure the longitudinal forces induced by high-adhesion AC locomotives. 
Rail anchoring conditions, both on and off the span, were varied during the test to determine their 
effects on the transmission of longitudinal forces into the bridge. Preliminary results show that cur- 
rent American Railway Engineering Association (AREA) design recommendations for steel bridges 
underestimate these forces. Interestingly, the AREA design guidelines used for timber bridges appear 
to provide a much closer approximation. 

This test is the first in an AAR study of longitudinal forces in bridges under new-technology 
train equipment such as AC locomotives. While the results of this complete study will be used to 
develop new design recommendations, it is too early to make any recommendations based on this sin- 
gle test. Variations in span length and other bridge features may lead to different results. 

Background 

With new AC locomotives being capable of roughly twice the adhesion of older series conven- 
tional DC locomotives, there is some concern that the current longitudinal force design recommen- 
dations should be updated. 

The Association of American Railroads (AAR) recently tested a 50-foot open-deck steel deck 
plate-girder bridge to measure the longitudinal forces induced by high-adhesion AC locomotives. 
Rail anchoring conditions both on and off the span were varied during the test to determine their 
effects on the transmission of longitudinal forces into the bridge. Preliminary results show that cur- 
rent American Railway Engineering Association (AREA) design recommendations for steel bridges 
underestimate these forces. Interestingly, the AREA design guidelines used for timber bridges appear 
to provide a much closer approximation. 

This test is the first in an AAR study of longitudinal forces in bridges under new-technology 
train equipment such as AC locomotives. While the results of this complete study will be used to 
develop new design recommendations, it is too early to make any recommendations based on this sin- 
gle test. Variations in span length and other bridge features may lead to different results. 

Revenue service tests were conducted at Trinidad, Colorado, on the Burlington Northern Santa 
Fe. The test bridge is located on a 1 -percent grade on a line which carries several unit coal trains 
daily, most of them powered by AC locomotives. The bridge is a single span, built in 1908 for a 
Cooper E-55 design loading. Rail is continuously welded 132 RE on the bridge and its approaches. 
The bridge deck has timber ties on 1-foot centers. Every other tie is fastened to the top fiange of the 
girders using an anchor bolt and spring clip. The top flange of the girder has a riveted cover plate 
along its full length. The rivet heads are imbedded into the deck ties. The west approach has seven 
timber ties adjacent to the bridge, then concrete ties on 2-foot centers with Safelok fasteners. The east 
approach has about 55 feet of timber ties, then a turnout which is heavily anchored with Pandrol clips. 



♦Principal. AAR-TTC 
♦Senior Engineer, AAR-TTC 
♦University of Illinois 
♦University of Illinois 



101 



Bulletin 760 — American Railway Engineering Association 



The bridge girders and bracing, rails, and approach trackage were instrumented to measure 
strains, forces, and displacements. Key measurements were longitudinal rail forces in both rails at 
each end of the bridge. The applied tractive effort (whether in traction or dynamic braking) for the 
AC locomotives was obtained from the cab display. Net force into the bridge was calculated by sub- 
tracting the net longitudinal rail force at the ends of the bridge from the tractive effort applied by the 
loconn)tive axles on the bridge. 

During the course of testing, rail anchoring conditions were varied both on and off the bridge to 
study their effects on bridge longitudinal forces. Four different combinations of anchoring were used: 

• Rail on approaches anchored tightly, rail on bridge minimally anchored 

• Rail on approaches minimally anchored, rail on bridge minimally anchored 

• Rail on approaches minimally anchored, rail on bridge anchored tightly 

• Rail on approaches anchored tightly, rail on bridge anchored tightly 

For the approaches, minimal anchoring consisted of timber ties without box anchors, and Safelok 
clips removed from every other concrete tie for 200 feet on the west approach. Tightly anchored 
approaches consisted of box anchors on all timber ties, and Safelok clips on all concrete ties. For the 
bridge, minimal anchoring was supplied by the rail clips and anchor bolts in their as-is condition. 
Tight bridge anchoring consisted of box anchors on each anchored deck tie, as well as tightened rail 
clips and anchor bolts. 

AAR researchers measured forces under revenue service trains in both tractive effort and 
dynamic braking. The highest forces into the bridge were measured in the tractive effort cases as the 
locomotives crossed the bridge. The AC locomotives used in the test (Electro-Motive Division model 
SD70MAC) have a maximum tractive effort per locomotive unit in excess of 150,000 pounds, while 
maximum dynamic brake effort is limited to about 80,000 pounds. All data presented here is for loaded 
unit coal trains of approximately 1 15 cars powered by a three-unit set of six-axle AC locomotives. The 
highest applied force occurs when a total of six powered axles are on the span, as shown in Figure I. 




Figure 1. AC Locomotives Crossing Test Bridge 



Paper by D. E. Otter, J. LoPresti, D. A. Foutch and D. H. Tobias 



103 



Preliminary Test Results 

Figure 2 shows results from train passes when the approaches were tightly anchored. This is the 
most common practice for bridges in continuous welded rail territory. As expected, the force into the 
bridge increases as the tractive effort applied to the rail on the bridge increases. The forces into the 
bridge for a given applied tractive effort appear to be slightly higher, when the rail on the bridge is 
tightly anchored, than when the rail on the bridge is minimally anchored. The amount of applied trac- 
tive effort varied depending on whether the locomotives were in traction or dynamic braking. Applied 
tractive effort also varied from one train pass to another due to differences in train handling, train 
speed, and locomotive performance. 

Figure 3 shows the results from the train passes when the rail on the bridge was anchored 
tightly. The highest forces measured during any of the tests were for this condition and minimal 
approach anchoring. A force into the bridge of nearly 100 kips was measured for an applied tractive 
effort of about 135 kips. The forces into the bridge are noticeably reduced when the rails on the 
approaches are tightly anchored. This allows more of the applied tractive effort to be carried off the 
bridge through the rails. 

Figure 4 shows the results from the train passes when the rail on the bridge was minimally 
anchored. In this case, the effect of anchoring the rails on the approaches is barely discemable. 
Anchoring the rails on the approaches may offer a slight reduction in force reacted by the bridge. 



100 




20 40 60 80 100 120 140 160 

Applied Tractive Effort (Kips) 



o Bridge Anchored Tightly x Minimal Bridge Anchoring 



Figure 2. Force into Bridge with Approaches Anchored Tightly 



104 



Bulletin 760 — American Railway Engineering Association 



100 



(A 


80 








60 


T3 




%— 




CO 




O 

4^ 


40 


c 




8 




o 


20 





X 
X 


1 




o8 

a> 




o 




1 1 1 1 


1 1 1 



20 40 60 80 100 120 140 160 

Applied Tractive Effort (Kips) 



X Minimal Approach Anchoring o Approaches Anchored Tightly 



Figure 3. Force into Bridge with Bridge Anchored Tightly 



100 




20 40 60 80 100 120 140 160 

Applied Tractive Effort (Kips) 

Figure 4. Force into Bridge with Minimal Bridge Anchoring 



Paper by D. E. Otter, J. LoPresti. D. A. Foutch and D. H. Tobias 105 



Implications 

The measured longitudinal forces into the bridge are considerably higher than the design val- 
ues currently recommended by AREA for steel bridges. The current AREA design loading is less than 
4 kips for a 50-foot steel span, compared to measured forces of nearly 100 kips. However, it should 
be noted that this test, on a relatively short steel bridge, arguably represents a worst case in two 
respects: ( 1 ) the bridge is just long enough to accommodate six powered axles, as shown in Figure 1 , 
and (2) the shorter the bridge, the lower the design longitudinal load as per the current AREA for- 
mula for steel bridges. For longer bridges, the concentration of applied longitudinal force will be 
lower and the design load higher. 

The maximum measured longitudinal force into the bridge is roughly 25 percent of the loco- 
motive weight applied to the bridge. This agrees well with the current AREA guidelines used for tim- 
ber bridges, which recommend designing for 25 percent of the weight on locomotive drivers. 
Fortunately, many of the steel bridges designed before 1968 were designed to similar guidelines. 

The test results also show that, as applied tractive effort increases, so does the longitudinal force 
that must be reacted by the bridge. Although the tests were conducted with AC locomotives, it should 
be noted that even for tractive effort levels typical of DC locomotives, the longitudinal force into the 
bridge is higher than the AREA design value. 

The AAR, in conjunction with researchers at the University of Illinois, will use the measure- 
ments from these tests to calibrate an analytical model to extend the results of this test to other 
bridges. 

Further testing is recommended to determine the longitudinal forces in spans of different 
lengths, as they could vary significantly. Differences in bridge construction and anchoring details that 
could reduce the effects of these higher longitudinal forces should be investigated. Longitudinal 
forces induced bv thermal effects in continuous welded rail should also be considered. 



Who in the world can respond 

to your requirements for specific 

track equipment? 




Fairmont Tamper can, and does. 



STz&m&^rai^ ' 



In addition to over 130 different machines we offer 
for railway track maintenance, we also develop new 
ideas, and deliver to your unique specifications. 

We adapt standard equipment to include custom 
options. Design new machines to solve our 
customers' problems. And remanufacture entire units 
or assemblies. 

This requires strong, broad design expertise in 
mechanical, hydraulic, electronic and 
software solutions. Fairmont Tamper 
has it. And our strong customer 
base— over 150 around the world— is 
proof that we deliver. 

In fact, Fairmont Tamper offers 
the most complete line of railway 
maintenance equipment in the wodd. 
Our line includes: • Tampers 
• Ballast regulators, brooms and undercut- 
ters • Rail grinders • Track construction 
and renewal systems • Tie removal and 
insertion machines • Spike drivers and 




pullers • Rail anchor and fastener applicators 
• Hy-Rail® Guide Wheel Attachments 

We also provide track renewal, rail grinding an 
other contract services. 

You'll find full-service Fairmont Tamper faciliti 
on three continents, and export agents around the 
world. Plus worldwide technical/training support fn 
our field service staff. 

In North America— call 
803-822-9160 or fax 803-822-747 
Australia— call 7 2056500 or fax 
7 2057369. U.K.— call 602 3844C 
or fax 602 384821. 



Examples of new ideas developed by 
Fairmont Tamper for track mainte- 
nance needs around the world include 
a transit rail grinder, a new track 
construction machine, the Pony (track 
renewal machine) for japan, section 
gang vehicles for Mexico, and over- 
head maintenance vehicles for China. 



F airmam 
tamper 

|Q) a harsco companq 

Your partner along the way 



IMPROVING INFRASTRUCTURE RELIABILITY 

THROUGH ENGINEERING FACILITY 
MANAGEMENT SYSTEM STANDARDIZATION 

Sponsored by: 

AREA Committee 32 — Engineering Management Systems 

Subcommittee 2 — Engineering iVIanagement Systems 

By: Dale E. Bartholomew, PE* 

Ladies and Gentlemen, I wish to thank the AREA Board and the Program Committee for pro- 
viding this opportunity to AREA Committee 32, Engineering Management Systems, to present this 
paper titled "improving Infrastructure Reliability Through Engineering Facility Management System 
Standardization." I know that title is a real mouthful, but I hope that this presentation can help make 
it clear what Railway Engineering Facility Management Systems are and how they provide an oppor- 
tunity for significant improvement in the effective construction and maintenance of all of our infra- 
structure. I also thank each of you for demonstrating your interest by your presence here this afternoon. 

My name is Dale Bartholomew. This presentation is sponsored by AREA Committee 32 and its 
Subcommittee 2, also titled Engineering Management Systems. In addition to Subcommittee 2, 
Committee 32 has an Education Subcommittee and a Subcommittee dealing strictly with Computer 
Aided Drafting and Design (CADD) Systems, including the interchange of data between these sys- 
tems. Both of these Subcommittees had major involvement in this presentation. The Education 
Subcommittee has sponsored the entire afternoons session and much of the work of the CADD Sub- 
committee is summarized in one of the modules of this presentation. 

Presentation Outline 

This Presentation will begin with an Introduction wherein I will touch on how a comprehensive 
Railway Engineering Facility Management System (REFMS) can improve your railroads infrastruc- 
ture reliability. An even greater goal of this presentation, however, will be to show how Standardiza- 
tion of REFM Systems will facilitate development of enhanced systems yielding an even greater 
improvement in infrastructure reliability than is attainable with current systems. This is because stan- 
dardization will allow faster and more economical development of REFM Systems: should eliminate 
duplication of effort; and will provide systems with greater capability, more compatibility and with 
easier maintenance of the systems themselves and the data that they need to function. 

Next I wish to describe the historical and continuing development of REFM Systems on most 
major freight railroads and transit systems. I will briefly outline early systems and touch on current 
practice. Of course a major intent of this presentation is to outline the work of Committee 32 and its 
efforts to define standards for REFM Systems. The development and publication of the AREA Manual 
Chapter 32 Table of Contents in 1994 was a key first .step. Much of this presentation, however, will 
center on the standard REFM System Functional Sections or Modules and the standard REFM System 
Process and Information Flow that, in reality, are nicely outlined in the Chapter 32 Table of Contents. 

Following the recommendation for all memorable talks, 1 will conclude by "telling you what I 
Just told you." I will summarize how REFM Systems improve infrastructure reliability and reiterate 
how standardization of REFM Systems will provide improved systems and thus, in turn, even greater 



♦Sverdrup Civil. Inc. 



107 



I OS Bulletin 760 — American Railway Engineering Association 



infrastructure reliability. A key conclusion that I hope all of you will come to, however, is that there 
will be significant advantage to you or your railroad to encourage participation and railroad support of 
the AREA Manual chapter development work of AREA Committee 32. This of course should lead to 
the conclusion that you and a number of your people should become active members of Committee 32. 

Introduction 

Railroading has undergone significant change in all methods of operation in the last couple of 
decades. Much of this change has caused all companies, and individuals within them, to reexamine their 
business objectives on a long term basis while also considering short term benefits. Particular emphasis 
has been placed on how to achieve these objectives in the most economical way possible. Automation 
systems and the thrust for a paperless office environment have played a key role in this change. 

Railway Facility Management has not been exempt from this trend. All major railroads and 
many transit agencies now use Railway Engineering Facility Management Systems (REFMS) in one 
form or another to a significant degree. My discussion on how these systems can improve your rail- 
roads infrastructure reliability will be brief, since there have been numerous demonstrations of this 
through presentations at previous AREA Conferences and articles in railway industry publications. 
For example, note the article "Developments in Maintenance Planning" in the February, 1997 issue 
of Railway Track and Structures. 

REFM Systems, thus, are well known to improve infrastructure reliability by facilitating more 
accurate monitoring of facility conditions; and by providing enhanced forecasting, budgeting, sched- 
uling and management of maintenance and capital improvement work. 

REFM Systems are also known to improve infrastructure reliability by providing enhanced 
management, and productivity and/or quality assessments of: personnel, maintenance or production 
gangs, work equipment, material, projects, the engineering design process and maintenance or con- 
struction documents such as plans or specifications. Finally, reliability is improved by minimizing 
manual information acquisition or data entry. Further details of how REFM Systems improve infra- 
structure reliability in the aforementioned ways will also be made apparent as the discussion of each 
basic segment or module of a REFM System is presented. 

The need and benefits of Standardizing REFM Systems has become increasingly apparent to 
AREA Committee 32 as automation systems; computer hardware and software, and communication 
technology; undergo phenomenal change each year. The power of computers continues to increase 
significantly while the price and size are dramatically reduced making it now possible to economi- 
cally place powerful computers or data terminals on virtually every desk or even in the pocket of field 
personnel. These changes have accentuated the potential for REFM Systems and have caused more 
and more railroads to look into their development. 

The Committee's studies have shown, however, that the majority of REFM Systems developed 
to date have been developed as independent endeavors. Each railroad, or possibly even department 
within a railroad or section within an Engineering Department, has taken a significantly different 
approach; utilized different procedures and file structures; used independent data storage; and con- 
tracted with new and/or different system developers. This approach has resulted in excessive devel- 
opment time, high cost and extensive duplication of effort. These "independent endeavor" impedi- 
ments have frequently led to development of systems with limited capability as each railroad's budget 
restraints lead them to direct their major efforts toward a narrow range of Engineering Facility 
Management functions rather than develop comprehensive REFM Systems. 

The Committee has discovered that all railroads perform most functions listed in the Chapter 
32 Table of Contents, but use uniquely differing levels of either manual or automated techniques to 
process and store the information. These further differences often result in internal functional incom- 
patibilities between system modules or segments, and the inability to share data. Recent mergers have 
demonstrated that these systems also have major external incompatibilities with the systems of other 



Paper by D. E. Bartholomew 109 



railroads as well as with outside vendors or governmental agencies. Frequently manual data reentry 
has been required and maintenance of these systems is difficult leading to early system obsolescence. 
A review of presentations at previous AREA Conferences and articles in railway industry publica- 
tions, including the most recent articles, confirms that the trend of "independent endeavor" develop- 
ment is still the norm. 

The maintenance of REFM Systems and their associated data files has been found to be one of 
the most critical problems. Just like the facilities that REFM Systems are intended to help us main- 
tain, REFM Systems and their data files are subject to the Murphy's Law corollary that.... "Anything 
that's been put together will fall apart sooner or later." But, with computer based REFTVI Systems the 
problem is vastly compounded by the extreme rapidity with which new software releases, and higher 
capacity and faster hardware become available. 

Many railroad managers have erroneously assumed that, once the expense of developing a sys- 
tem and converting information (text records, inventory data, drawings, etc.) to electronic form is 
undergone, their data storage, location, retrieval and analysis burdens will be gone forever. Nothing 
is further from the truth. 

Unfortunately, Committee 32 has found that few railroads have an adequate ongoing REFM 
System maintenance budget. Many systems that we reviewed in the 1970's, or even later, have lapsed 
into disuse, in spite of being masterfully conceived and having excellent functionality. 
Redevelopment from scratch has sometimes occurred (independent effort) or is being contemplated. 
The problems: 

• Most new software releases (computer programs) will only read information (data) files cre- 
ated by and saved in the format of the most recent 3 or 4 releases of software, and then only 
the most popular versions. 

• Most new software releases will only work on the latest 2 or 3 generations of hardware (com- 
puters). 

• Repair parts and service are readily available for only the latest 3 or 4 generations of com- 
puter hardware. 

• New software releases are issued approximately every year. 

• New generations of computer hardware are produced approximately every other year. 

The net result is that, if you have not annually updated or translated your information files into 
the format of recent software releases, you will have significantly reduced efficiency, and possibly 
considerable difficulty, accessing the information after just three or four years. If you have not 
updated the computer hardware or software for ten or more years, there will likely be serious com- 
patibility problems with other systems and system updates or information file translations may be 
very difficult. Furthermore, if you can not get parts to repair a hardware malfunction, you may not 
be able to access your information files at all. The system development and information file creation 
work may be totally lost. Most railroads have come to realize that critical documents, such as bridge 
drawings, can not be efficiently stored as permanent electronic files without annual updates. 

It is the intent of the Committee in preparing Chapter 32 to encourage reversal of the "inde- 
pendent endeavor" development trend by providing recommended standards for the design and 
development of REFM Systems. We have come to believe that these standards will facilitate the 
design and development of enhanced REFM Systems in a faster and more economical manner and 
can eliminate duplication of effort. Standardized systems will also have greater capability and more 
compatibility than current systems; and the systems themselves or the data that allows them to func- 
tion will be easier to maintain. 

These benefits are realized since individual railroads will not have to expend the considerable 
time and expense required in the past to define system concepts. That is, they will not have to deter- 



1 10 Bulletin 760 — American Railway Engineering Association 



mine the what, why, how, when and other basic objectives of their proposed REFM Systems. The 
Manual Chapter 32 standards will provide predefined functionality, methodology, appraisal proce- 
dures, data structures, data formats, data entry procedures, terminology, information codes, inter- 
change standards and many other recommended system parameters. Increased utilization of stan- 
dardized systems will encourage software and hardware developers to "build-in" greater 
compatibility and possibly automated updates or translations to new software releases and genera- 
tional hardware improvements. 

Development of REFMS 

Over the last couple of decades, AREA Committee 32 has investigated the REFM Systems on 
most major railroads. This effort has lead us to define the following standard functional components 
or modules of a comprehensive REFM System. 

Project Appraisal and Forecasting Module (PAFM) 

Cost Estimating and Budgeting Module (CEBM) 

Project Scheduling Module (PSM) 

Gang Management Module (GMM) 

Work Equipment Management Module (WEMM) 

Material Management Module (MMM) 

Project Management Module (PMM) 

Fixed Facility Inventory Module (FFIM) 

Engineering Design and Administration Module (EDAM) 

• Computer Aided Drafting and Design (CADD) 

• Document Management (DM) 

• High-Wide-Heavy Load Authorization Module (HWHLAM) 

The above modules of a REFM System are listed in the sequence of primary information or 
process flow and closely parallel the Part Title sequence of the twelve parts of the AREA Manual 
Chapter 32 Table of Contents. Both begin with Project Appraisal and Forecasting since that is the ori- 
gin of any proposed maintenance or capital project; or, in other words, the beginning of the project 
planning function. 

The project process and information flow then progresses down through the list to project com- 
pletion within the Project Management Module, followed by the updating of the Fixed Facility 
Inventory records. These records, which also include field inspection records, provide the informa- 
tion necessary to accomplish new project appraisals and the repetition of the process. Figure 1.3, or 
the stylized representation in Figure 1.4, provide a more pictorial representation of this circular 
process since they show the primary information or process flow with shaded arrows advancing in a 
clockwise direction. A reverse flow; to accommodate project change orders, emergencies or other 
project revisions; will also occur, as depicted in Figure 1.3 with counterclockwise unshaded arrows. 

The Project Appraisal and Forecasting Module (PAFM), upper left in Figure 1 .4, is key to 

the Railway Engineering Facility Management process. This module automates the facility condition 
appraisal and facility degradation analysis processes, and advises Engineering Management of the 
need for a capital improvement or maintenance project. Then, considering manual Engineering 
Management overrides, forecasts of the appropriate time to perform projects is presented. 

The Cost Estimating and Budgeting Module (CEBM) automates project cost estimating and 
budgeting. Combining cost information with the project forecasts, annual budgets and future program 
budgets are developed including consideration of budget limits. This module also forms a flow loop 
with railroad management's budget reviews and the granting of authorities to progress projects. 
Again manual overrides are allowed, to trim budgets by reducing project scale, moving projects to a 
future year, etc. 

The Project Scheduling Module (PSM) automates project scheduling considering the avail- 
ability of design drawings, track time, gangs, material and equipment. The need and time-frame 



Paper by D. E. Bartholomew 1 1 1 



required to publish time-table revision may be considered. The combining of projects that are in close 
proximity and seasonal restrictions are other considerations. 

The Gang Management Module (GMM) automates the management of crews performing 
project work including work location assignments and scheduling, progress reporting, personnel 
schedules, payroll reporting, cost reporting, productivity analysis, etc. 

The Work Equipment Management Module (WEMM) automates the management of equip- 
ment used to perform project work including work location assignments and scheduling, production 
reporting and analysis, cost reporting, equipment maintenance, equipment and parts inventories, etc. 

The Material Management Module (MMM) automates the management of material used in 
a project including requisitioning, long lead time or out-of-stock orders, stock inventories, shipping, 
etc. Sometimes this module is part of or totally incorporated into a system of the Purchasing and 
Materials Department. 

The Project Management Module (PMM) automates the management of project field work. 
Frequently it is a fairly conventional readily available construction Project Management system, but 
needs tailoring to the specific field work needs of a railroad project. 

Fixed Facility Inventory Module (FFIM). A major information management and storage 
module that automates the maintenance of all records of facility type, location, condition, rating, 
clearances, digital photos/videos, etc.; including historical and current inspection records, project 
completion records, etc. 

Computer Aided Drafting and Design (CADD). An important submodule of the Engineering 
Design and Administration Module that automates the production of engineering designs and draw- 
ings. Frequently this submodule overlaps with the Fixed Facility Inventory Module as considerable 
inventory information can be stored within electronic drawing or mapping files; or maps and other 
drawings can be automatically generated from data in facility information files. 

Document Management (DM). A submodule of the Engineering Design and Administration 
Module that automates the management of engineering documents, including electronic text and 
drawing storage and distribution, and maintenance of document inventories. 

The Engineering Design and Administration Module (EDAM) automates engineering 
design and the administration of an Engineering Department, including personnel scheduling, design 
quality and productivity analysis, design drawing availability, etc. 

The High-Wide-Heavy Load Authorization Module (HWHLAM) automates the authoriza- 
tion of high, wide or heavy load passage over various segments of a railroads lines. Usually a sepa- 
rate function relative to the project management loop, but can provide recommendations for facility 
clearance improvements or reinforcements. Utilizes information maintained by the FFIM to a major 
extent. Sometimes this module is part of or totally incorporated into a system of the railroads 
Transportation Department. 

Since all railroads use differing terminology, the above described AREA Committee 32 stan- 
dard REFM System functional component or module names and the related process and information 
flow will form a base of reference for all further REFM System discussions in this presentation. 

Early Systems 

Beginning in the 1970's, AREA Committee 32 began noticing increased interest in the devel- 
opment of segments of a REFM System by many railroads. Frequently this was the vision of one or 
two individuals who worked in a small functional area pretty much alone and on their own initiative. 
To give a few examples, in the early 1970's, the former Southern Railway had a nifty HighAVide 
Load Clearance Approval System on a stand-alone minicomputer This system maintained its own 
clearance data base that was manually digitized from light-trace photographic images produced by 
their new clearance measuring high-rail truck. It was a good start toward a HighAVide/Heavy Load 
Authorization Module in spite of no linkages to other systems. 



Bulletin 760 — American Railway Engineering Association 



By the mid I970's, CONRAILand Southern Pacific had independent Bridge Inventory Systems 
that had no links or common data storage with track or other facility records. Everything about these 
two systems was different but they did form the beginnings of Fixed Facility Inventory Modules. 
Canadian Pacific Railway took a plunge into rail and track management systems with development 
of the following: 

• Rail Failure System (RFS) — Maintained a historical record of rail failures and replacements 
allowing analysis of supplier quality. — A small part of a Fixed Facility Inventory Module. 

• Road Maintenance Management System (RMMS) — Maintained a track inventory and 
made some track condition related project recommendations. — A little more of a Fixed 
Facility Inventory Module and a start on a Project Appraisal and Forecasting Module. 

• Estimated Cost System (ECS) — Estimated Labor, Material and Contract Costs for track 
maintenance projects with little management of the budgeting process. — A start toward a 
Cost Estimating and Budgeting Module. 

• Work Equipment Inventory (WEI) — Work equipment data only such as serial numbers, 
location, condition, age, etc. — A small part of a Work Equipment Management Module. 

By 1980, Canadian National was serious about the development of REFM Systems. They had 
over 100 computer programmers working on separate track and structures management systems. 
During Committee 32's visit to Montreal, we were given a presentation on how their Bridge & 
Structures Systems was being organized. See Figure I.I. For the first time, several REFM System 
functional components were being linked into a comprehensive system that included the following: 

• Work Information System (W.I.S.) — Overall process control. Prioritizes and schedules work 
tasks, including issuing Work Orders. Minimal cost estimating or budgeting consideration. — 
A good start on some functions of the bridge or structures part of both the Project Appraisal 
and Forecasting Module and the Project Scheduling Module. 

• Labor Reporting System (L.R.S.) — Mostly Engineering Department input to an Accounting 
Department System. Reports project labor costs and other accounting information. Also used 
to report labor at Work Equipment Maintenance Shops. Does include personnel availability 
reporting. — A good start on the bridge or structures labor reporting part of a Gang 
Management Module. 

• Material Reporting System (M.R.S.) — Mostly Engineering Department input to a Material 
Department System. Reports project material costs and other accounting information. Also 
used to manage material inventories, catalogs, forecasts, requisitions, usage, etc. — A good 
start on the bridge or structures material management and reporting part of a Material 
Management Module. 

• Workload Analysis System (W.A.S.)— Primarily supports the scheduling function for pro- 
jects and to a lesser extent provides information to the project forecasting function. — A some- 
what "in-between" component that performs additional functions of the bridge or structures 
part of both the Project Appraisal and Forecasting Module and the Project Scheduling 
Module. 

• Plant Inventory System (P.IN.S.) — Maintains structure inventory records of bridges, timber 
trestles, culverts, buildings, miscellaneous structures, etc. — A good start on the bridge or 
structures part of a Fixed Facility Inventory Module. 

In addition, at the 1980 Committee 32 meeting in Montreal, we were shown a process and infor- 
mation flow diagram for the first time. This CN called their "Bridge and Structures System Work 
Environment." See Figure 1.2. This diagram showed the continuous cyclical nature of the facility 
management process and prompted the functional arrangement of modules and the use of shaded 
process or information fiow arrows in our REFM System diagrams. Figures 1 .3 or 1 .4. 



Paper by D. E. Bartholomew 



113 



; L.R.S. 

V LABOUR REP. SYS. / 




P.IN.S. 

PLANT INV. SYS. 

BRIDGES 

TIMBER TRESTLES 

CULVERTS 

BUILDINGS 

MISC. STRUCT. 



,' M.R.S. 

MAT REP. SYS. 




Figure 1.1 Bridges & Structures Systems 



Though of a slightly different arrangement than Committee 32 's diagram, CN's shows how 
inspection information leads to condition priorities which then are both stored as part of the facility 
inventory record. This information is analyzed and a project priority (computer ranking) is estab- 
lished. The importance of engineering or railroad management overrides is also illustrated which then 
results in development of project descriptions (work item lists and work details). This in turn leads to 
project scheduling, material requisitioning and the issuance of work orders. The project work is exe- 
cuted and completion reports are prepared which include labor reportmg. Finally the facility inven- 
tories are updated and with new inspections, the process continues around the circle. 

Unfortunately, once CN's massive REFM Systems development efforts were complete, the 
budgets were drastically cut. System maintenance was nonexistent and, I am told, the 1980 Bridge & 
Structures and Track [Management] Systems are now essentially non-functional. The Committee's 
latest visit to Montreal has revealed that CN has moved on to redevelopment of "newer and better" 
components of a REFM System. 



14 



Bulletin 760 — American Railway Engineering Association 




® 



=WORK ANALYSIS 



Figure 1.2 The B & S Work Environment 



In the late 1980's railroad management began to recognize the benefits of REFM Systems and 
a major reversal of developmental impetus occurred. Instead of Engineering Department underlings 
pleading for a REFM System development budget, top railroad management began insisting that 
REnvi Systems be developed. This caused a significant increase in the REFM System development 
effort of railroads, but also took a lot of the control out of the hands of engineers who's ranks had 
been reduced. More and more development effort was now expended by Information Management 
Departments or Consultants, increasing "independent endeavors." Buriington Northern Railroad 
underwent this process and obtained a track management system and a bridge drawing system with 
development of the following: 

• Roadway Information System (RIS) — Maintained track inventory records. — Part of a 
Fixed Facility Inventory Module. 

• Track Management System (TMS) — Predicts track maintenance requirements. — Part of a 
Project Appraisal and Forecasting Module. 

• Estimating System (ES) — Track maintenance costs and budgets. Also manages schedules 
and material requisitioning. — Parts of a Cost Estimating and Budgeting Module, a Project 
Scheduling Module, and a Material Management Module. 

• Service Maintenance Planner (SMP) — Schedules work to minimize traffic interference, 
including combining work in the same territory. — An additional part of a Project Scheduling 
Module. 



Paper by D. E. Bartholomew 



15 




Figure 1.3 



Bulletin 760 — American Railway Engineering Association 



< ^ 
z >- 

t^ w w 

UJ I- s 

UJ Z il 

Z UJ UJ 

/n S Ql 

y UJ -- 

go 





= 1 

_ -D 

to O 

2 o) 

Q. C 

Q. '^ 

< (0 

? s 

"5* o 

Q. 



"(0 3 

o -a 

-I o 

(D O 

I (Q 

(1> N 

•^ O 



^ 3 
O) < 



T3 





^^H 


^^H 


PI 


w 




3 


C 

c 


3 
O 








c 


^^ 


o 


_0) 


s 








_0) 

(7) 


o 


^ 


(/) 
0) 
Q 


c 






■D 


0) 


(0 


o 


o 




41^ 


O 


Q 


u. 


4-» 

c 








C 




XI 


•n 


(D 


o> 


re 




O 


< 


c 


o 


> 


c 


■4-» 


•*•» 


E 


1- 


re 


X 


c 




'F 


C 

0) 

E 

3 
U 
O 

n 


3 

3 


c 


iZ 




0) 

c 

"5) 

c 


E 

< 


re 

c 
re 

S 


a 

E 
o 
o 


2 

Q 


■ 




HI 

















Figure 1.4 



Paper by D. E. Bartholomew 1 1 7 



• Project Management System (PMS) — Track maintenance project planning and tracking, 
including equipment and personnel management. — Parts of a Project Management Module, 
a Gang Management Module, and a Work Equipment Management Module. 

• Bridge Drawing System (BDS) — Electronic Bridge Drawing storage, retrieval and trans- 
mission system. — A part of the Document Management function. 

Unfortunately, BNRR's Bridge Drawing System used only scanned or translated raster (dot) 
images for drawing storage and transmission. This means that CAD editing of drawings was not pos- 
sible. The system was consultant developed using a unique combination of hardware and tailored 
software that was difficult to maintain and the system rapidly became obsolete. I am told, that the 
system is now essentially non-functional. 

Current Practice 

A 1992 Committee 32 revisit to CP Rail revealed a significantly enhanced rail and track main- 
tenance forecasting system called TMAS for Track Maintenance Advisory System. This portion of 
a Project Appraisal and Forecasting Module contained the most comprehensive evaluation and future 
maintenance prediction algorithms seen to date. Using a "real-time" Track Evaluation Car data acqui- 
sition and recordings system, the facility inventory (or Library System) records are instantly updated. 
A Track Profile Analysis is performed and current and projected Rail Quality, Rail Wear, Track 
Quality and Surface Quality Indexes are computed. These are then used to very accurately forecast 
Programmed Track Maintenance. 

Also in 1992, Chicago Transit Authority's Information Management System (IMS) was 
reviewed. It primarily Maintains Fixed Facility Inventory type records on elevated track structures, 
including location, condition and rating data. The system also, however performs work projections 
and budgeting, or Project Appraisal and Forecasting Module and Cost Estimating and Budgeting 
Module, functions. 

A recent Committee 32 visit to CSX Headquarters in Jacksonville, Florida revealed their use of 
an impressive Windows based graphical interface to the facility inventory and gang management sys- 
tems. This system automatically and instantly updated electronic format graphical mapping every 
time the facility inventory was updated. Project locations were displayed on the electronic mapping 
displays with project progress and gang production reporting information available with the click of 
a mouse. This graphical data retrieval interface was available at CSX offices around the system. 

Just last month the Committee made its first visit to the combined Burlington Northern Santa 
Fe Railway headquarters in Fort Worth, Texas. There we were shown one of BNSF's latest efforts, a 
Corporate Planning and Estimating system that had functionality and process flow closely parallel- 
ing major portions of the Committee's standard REFM System. See Figure 1.5. 

Facility inspection records from BNSF's Inspection System, an equivalent to the Fixed Facility 
Inventory Module, provides critical facility condition information to a functional equivalent of the 
Project Appraisal and Forecasting Module that BNSF calls RCP for Roadway Coordinated 
Planning. The process then flows to their CEPS or Corporate Estimating and Planning System, 
which has functionality equivalence to the Cost Estimating and Budgeting Module. Similar to the 
standard REFM System, CEPS has a side information flow loop to railroad or corporate management 
for budget review and authority. This loop links CAPBUD or BNSF's capital budget Strategic 
Planning System, which provides budget overrides, and a Lotus Notes system which allows 
Electronic Approval of Requests For Authority. Process flow then links a Project Scheduling 
(MSH) component which performs functions equivalent to a Project Scheduling Module such as 
arranging Work Windows. Process flow also supplies information needed to Purchase Material to a 
P & M system in the Purchasing and Material Department. This closely parallels the functionality of 
the Material Management Module. 



Bulletin 760 — American Railway Engineering Association 




Figure 1.5 



Paper by D. E. Bartholomew 1 19 



Like Committee 32's standard REFM System, BNSF shows that accounting information is sup- 
plied to automated systems in the Accounting Department. BNSF represents these systems as the 
Suspense Account (SA) and Capital Projects (CP) components of their Accounting Department's 
Millennium Systems. 

Committee 32 only shows the information flow to an Accounting Department, upper right-hand 
comer of Figure 1.3, rather than illustrate accounting systems as BNSF does, since they are external 
to an Engineering Department and are not a part of an Engineering Management System. 

The fact that BNSF calls the system Corporate Planning and Estimating and includes functional 
components from other departments in their process flow diagram, clearly indicates that the impetus 
for development is no longer from the Engineering Department. When asked by the Committee at the 
Fort Worth meeting, the BNSF system developers stated they have never used, or even heard of, the 
AREA Manual for Railway Engineering. In spite of significant parallelism with portions of the 
Committee's standard REFM System, "independent endeavor" development continues to be normal 
practice. 

Work of Committee 32 

In the early days of Committee 32, then called Systems Engineering or "The Computer 
Committee," REFM Systems had not yet become a concept for Engineering Department considera- 
tion. The Committee busied itself with review of railroad engineering design or analysis computer 
programs, such as rail stress analysis or railroad bridge deck girder design or rating programs that 
individuals had developed, frequently on their own initiative. At that time we had an assignment to 
maintain a Computer Program Digest of reviewed and recommended design or analysis programs. 
Debate ensued as to whether Committee 32 or Committee 15 should maintain the steel truss and 
girder bridge rating programs now maintained by the AAR. None of this work was considered appro- 
priate for inclusion in the AREA Manual. 

As computers began to proliferate, the amount of available software also mushroomed and the 
task of reviewing the new software and maintaining the Computer Program Digest became formida- 
ble. That task was dropped and the Committee began conducting railroad Engineering Department 
computer utilization surveys and holding symposiums to distribute information. Along with the rapid 
increase in computer technology, the potential for managing Engineering Department information 
such as rail flaw detector car data, facility inventory records, material requisitions, and maintenance 
or capital project records was realized; and the concept of developing AREA Manual Chapter 32 
material was proposed. Ill-conceived first efforts, however, did not meet Manual material publication 
standards. At the same time, interest in Committee 32 membership by non-Manual using Computer 
Systems types began to increase. These new members, however, were a great asset in formulating 
Systems concepts, such as functionality and process flow. 

In the I980's and early 1990's, railroads entered the perpetual reorganization mode which also 
caused high Committee membership turnover and further reduction in AREA-Manual-using mem- 
bership. Computer technology advances continued to escalate and realization of the need for REFM 
Systems increased and shifted from the Engineering Department to Corporate Management. 
Engineering Systems Sections were established in many railroad Engineering Departments or the 
number of Computer Systems Department personnel assigned to REFM System development was 
increased. The AREA Board shifted from reticence at publishing Chapter 32 Manual material to 
encouraging Manual material development. 

AREA Committee 32 began an earnest effort at developing AREA Manual matenal by defin- 
ing the unique functionality modules of a REFM System. The next step was development of the 
General Processes and Information Flow Diagram of Figure 1.3. This lead directly to development 
and publication of the AREA Manual Chapter 32 Table of Contents, which is essentially just a linear 
detailed listing of the various elements of a REFM System as depicted graphically in Figure 1.3. 



120 Bulletin 760 — American Railway Engineering Association 



Soon thereafter. Subcommittee 4, Engineering Graphics Systems, assumed responsibility for Part 
2 of AREA Manual Chapter 11, Cartographic Specifications, since almost all railroad mapping had 
begun to be maintained on Computer Aided Drafting (CAD) systems. Chapter 1 1 material was updated 
slightly and republished in 1996 as AREA Manual Chapter 32, Part 11, Section 1 1.3, Mapping. 

Currently, Committee 32 is working on developing AREA Manual Chapter 32, Part 1, Railroad 
Engineering Management Systems Overview. Figure 1.3 will be an important component of this 
material. This effort has taken longer than it should due to the increasing responsibilities of all rail- 
road personnel, and the Committee's declining active membership and meeting attendance. Also it is 
hard to coax volunteer Manual development effort out of railroad or consultant personnel who rarely, 
if ever, use the AREA Manual. Current Committee active membership includes only one or two mem- 
bers that regularly use the AREA Manual. This is why it is hoped that this presentation will encour- 
age many of you to join AREA Committee 32 and become active participants, so we can complete 
Manual Chapter 32, Part 1, and the remaining Parts 2 through 12. 

AREA Manual Chapter 32 Table of Contents 

Since AREA Manual Chapter 32 contains little more than the Table of Contents, it may seem 
that it does not have much value. Committee 32, however, believes that there is significant value to 
the Table of Contents, since, as alluded to above, it does provide a detailed listing of the various ele- 
ments of a standardized REFM System. This in itself could go a long way toward reducing the many 
functional differences that exist between REFM Systems on various railroads. It can also help assure 
system developers that they are not overlooking important functional elements that would provide 
significant added value to a railroad from the REFM Systems development effort. 

If Engineering Departments and, in turn, consultants or other systems developers can be made 
aware of the standard Table of Contents REFM Systems functional element listing, much of the 
"from-scratch," "independent endeavor" developmental expense and system configuration differ- 
ences could be avoided. These system configuration differences, previously illustrated in the Early 
Systems and Current Practice narrative, have been found to be the most difficult and costly roadblock 
to merging REFM Systems. BNSF has abandoned converting some elements of the former BN and 
Santa Fe systems and some of what they are converting is taking significant effort by numerous per- 
sonnel that has continued for over a year 

Standard REFMS Functional Sections or Modules 

Don't try to read the text of Figure 1.3 on the screen. This figure is available as a handout on 
the side tables and will be available in the printed version of this talk. I am showing this slide to you 
because I want to describe the meaning of the various rectangles that you can see on the screen. In 
addition to depicting the General Processes and Information Flow of a REFM System, Figure 1 .3 also 
illustrates the functional organization of each element or module. All of the functional elements or 
modules of a REFM System are contained within the largest and heaviest solid rectangle. This 
includes all modules listed in the simplified representation of Figure 1 .4. This large heavy solid rec- 
tangle can also be viewed as being the walls of a railroad Engineering Department, or encompassing 
the activities performed by an Engineering Department, even though the "real people" of an 
Engineering Department are depicted by the heavy dashed enclosure at the upper left comer of the 
large heavy solid rectangle. 

Around the outside perimeter of the large heavy solid rectangle are a number of dashed rectan- 
gles that represent other offices or departments that share information with the REFM System, but 
are external to the Engineering Department, or their systems are "external" to the REFM System (not 
maintained by Engineering). At the top, a dashed rectangle encircles the Executive and/or Finance 
departments. The smaller loop of arrows represents the "side" information flow loop that illustrates 
the budget or project authority request and approval process and corporate or engineering manage- 
ment budget or project overrides. 



Paper by D. E. Bartholomew 1 2 1 



Continuing at the top right and progressing clockwise, other external departments or offices 
encircled by dashed rectangles include the Accounting Department; Personnel Department; Property 
Management Department; Purchasing and Material Department; Field Offices, Gang or other Field 
Reporting Location; Outside Offices of Contractors, Suppliers, Other Railroads, Governmental Agen- 
cies, or Customers; and the railroad's Transportation Department. The dashed Field Office rectangle 
is shown overlapping the large heavy solid rectangle, since some or portions of all field offices, or their 
REFM System functional interface come under the jurisdiction of the Engineering Department. 

Lighter weight solid rectangles encircle each REFM System functional element. Two of these 
rectangles are shown extending outside the large heavy solid rectangle and overlapping other dashed 
rectangles. This is because the Material Management Module, lower right, or the High-Wide-Heavy 
Load Authorization Module may functionally reside totally within the Engineering Department, par- 
tially within the Engineering Department, or totally within the Purchasing and Material Department 
or Transportation Department, respectively. The Project Management Module's encircling rectangle 
is shown overlapping the dashed Field Office rectangle since a lot of the Project Management 
Module function is performed in the field. All of the rectangle overlaps of the Engineering Design 
and Administration Module, it's Document Management and CADD submodules and the Fixed 
Facility Inventory Module, represent the functionality overlaps of their processes described earlier. 

The large solid drum symbol shown in the center of the diagram represents Data Files that are used 
or maintained by some of all of the REFM System functional modules. The functional modules (smaller 
solid rectangles) are arranged surrounding these files and the processes and information flow is shown 
encircling the Data Files, because they are key to each module being able to perform it's function. 

We don't show the Fixed Facility Inventory Module as the center of the process like some rail- 
road's diagrams, because it is really just another element that performs the Facility Inventory Data 
update function of the overall process loop. It also performs the Facility Condition Data update func- 
tion, incorporating field inspection information (bridge inspections, track geometry car information, 
etc.) into the Facility Inventory Data. The Facility Inventory Data is the most accessed data of the 
system, so is shown as a part of the Data Files at the center of the process. 

Compared to the stylized representation of Figure 1.4, the Gang Management Module, the 
Work Equipment Management Module, and the Material Management Module are more accurately 
shown to have parallel functionality in Figure 1 .3. The process and information flow arrows also take 
three separate parallel paths exiting the Project Scheduling Module. The Material Management 
Module is shown following the other two only because there is subsequent process and information 
flow between the Work Equipment Management Module and the Material Management Module that 
deals with equipment and parts purchases. The three parallel process and information flows converge 
again at the Project Management Module. Another separate process and information flow path is 
shown between the Engineering Design and Administration Module and the Project Scheduling 
Module which depicts the handling of Design and Drafting Time and Engineering Design Schedule 
information and their impacts on individual Project Schedules. 

Taking a closer look at the functionality of each module, we will again begin with the Project 
Appraisal and Forecasting Module (PAFM), upper left in Figure 1 .4, since it is key to the Railway 
Engineering Facility Management process. See Figure 2. 



Specific appraisal functions handled by this module irtclude: 

• Appraisal Function 

• Performing Work Productivity Analysis 

• Performing Work Methodology Effectiveness Analysis 

• Calculating and Analyzing Quality Indexes for: 
+ Track / Rail / Ties / Surface / Bridges / Etc. 

• Developing Degradation Analysis Models for: 

-I- Track Surface / Rail / Ties / Subgrade / Bridges / Etc. 

• Performing Work Prioritization 



Bulletin 760 — American Railway Engineering Association 



<y 



Work Appraisals /- 
Degradation Analysis 
Condition Analysis 



Work forecasts 
Annual Programs 



( Pro/fect Appraisal and 

Forecasting Module (PAFM 

^- /. 

.Condition Indexes 






Figure 2 



Developing and utilizing very sophisticated degradation analysis model algorithms has become 
a common function of this module. These analysis's have proven to provide significant economic 
benefits and improved infrastructure reliability by allowing accurate project forecasting and the opti- 
mization of maintenance programs. The availability of frequent infrastructure condition data from 
inspections or test vehicles such as Rail Flaw Detector Cars. Track Strength Test Cars, or Track 
Geometry Cars; plus the improving capabilities of the automated modeling systems, make this capa- 
bility a reality today. Based on the appraisals performed, specific forecasting functions handled by 
the Project Appraisal and Forecasting Module include: 

• Forecasting Function 

• Developing Annual Programs for: 

+ Surfacing / Rail Renewal / Rail Grinding / Tie Renewal / Bridge Renewal 

• Coordinating Budget Approvals / Revisions, including: 
+ Automated / Manual Overrides / Etc. 

• Developing Multi-Year Program Forecasts for: 

+ Five / Ten Year Maintenance / Capital Improvements 

• Developing/Consolidating Work Program Data 

• Developing/Consolidating Project Data 

The Cost Estimating and Budgeting Module (CEBM) automates the project cost estimating 
and budgeting functions. By combining cost information with project forecasts, annual budgets and 
future program budgets are developed. See Figure 3. 

As previously discussed, this module also forms a flow loop with railroad management for budget 
reviews, authority processing, and manual overrides. Specific functions handled by this module include: 

• Obtaining or Maintaining Cost Data for: 

• Labor / Material / Equipment / Projects / Other Specialized Unit Costs 



Paper by D. E. Bartholomew 123 



• Developing Cost Projections for: 
•5/10 Year Programs 

• Performing Annual Maintenance or Capital Budget Analysis and Processing 

• Budget Estimates / Cash Flow Analysis / Budget Consolidation / Budget Requests and 
Revisions 

• Developing Multi-Year Program Forecasts 

The Project Scheduling Module (PSM) automates project scheduling considering the availabil- 
ity of design drawings, track time, gangs, material and equipment. See Figure 4. 

Specific functions handled by this module include: 

• Performing Gang Workload Analysis 

• Availability / Productivity - Analysis / Forecasts 

• Performing Work Equipment Workload Analysis 

• Availability / Productivity - Analysis / Forecasts 

• Performing Material Delivery Analysis 

• Material Delivery Lead Time Appraisal 

• Advance Order Material Lists / Forecasts 

• Performing Seasonal Restraint Analysis 

• Winter Work Avoidance? 

• Performing Operations Conflict Analysis 

• Track Time Planning 

This analysis could include investigating the possibility of providing long-term track closures 
of busy mainlines such as the 30-hour closure Burlington Northern used a couple of years ago for a 
bridge deck replacement project. See the November, 1995 issue of Railway Track & Structures. The 
analysis can compare the costs of inefficiently performing intermittent work tasks, between trains 
during many short train-free periods, to the cost of rerouting trains for an extended period of time and 
allowing two bridge gangs to quickly and efficiently perform the work. 

The analysis can also include scheduling gang time for the bridge gangs to practice performing 
the work with a model, as BN did, to assure efficiency during the train-free period. The ability to 
coordinate the placement of deck panels and track panels, and still accommodate the reach and lift- 
ing capacity of the locomotive crane, was critical to success of the long-term track closure procedure. 

Another operations conflict minimization option includes performing a: 

• Project Consolidation Analysis 

This analysis also investigates the possibility of providing a long-term track closure of a busy 
mainline, but combines it with the performance of several projects for greater economic benefit and 
improved infrastructure reliability. The analysis again compares the costs of inefficiently performing 
several projects intermittently during many short train-free periods, to the cost of rerouting trains for 
an extended period of time and allowing railway maintenance gangs to quickly and efficiently per- 
form several projects. An excellent recent example of the possible result of this type analysis was the 
six day closure of 162 miles of mainline in Nebraska that Union Pacific utilized to perform "...six 
months worth of work." See the August, 1996 issue of Railway Track & Structures. 

The UP was able to accomplish the installation of 28,200 wood track ties, surface 156.4 miles 
of track, surface 6 No. 30 and 36 No. 20 or No. 14 turnouts, surface 240 grade crossings, renew 67 
road crossings, replace 0.7 miles of curve rail on four curves, weld 390 in-track rail joints, install two 
new concrete bridges, install two new No. 30 turnouts, and renew one No. 14 turnout. 

Other operations conflict minimization functions include performing: 

• Transportation Department Coordination 

• Track Orders / Time Table Revisions 



124 



Bulletin 760 — American Railway Engineering Association 




--\ 



Budgets 
Programs 
Reguests for 
Authority 



Cost Estimates 
Cost Forecosts 



~-^>-- 



Approved Project Lists 
Project Work Descriptions 
Project Forecasts 



Cost Estimating and 
Budgeting Module (CEBM) 



At 



Figure 3 



\ 



Gong, Work Eguipment ond^ Track Occupancy 
Material Availability Analysi^ Analysis and Reguests 



Project Delays 



Project Scheduling 
Module (PSM) 



oterial Reguirements 
M\)^teriol Orders 
DelVery Schedules 



Engineering Design Schedules 



Workv Eguipment 
ReguiXements and 
Schedules 



Ogng Reguirements 
anH Schedules 
V^^ 



Yl 7 



v^ 

Project Schedules 
Project Critical Paths^ 



r< 



>4X A^ 



Figure 4 



Scheduling Time Table Revisions may be an important part of project scheduling if you oper- 
ate a busy passenger mainline. Amtrak showed Committee 32 a sophisticated scheduling system that 
they used when performing work on their busy Northeast Corridor. When Amtrak had to take one 
track out of service to perform bridge maintenance work, for example, that outage would impact train 
schedules enough that a time table revision was required. Their scheduling system would analyze the 
length of the track outage, the proximity of crossovers, etc. to determine the required time table revi- 



Paper by D. E. Bartholomew 125 



sions. The system would then consider the time required to publish the revised time table and get it 
into the hands of the public, when determining the work project schedule. 

A final Project Scheduling Module task is: 

• Developing Project Critical Path Scheduling 

The Gang Management Module (GMM) automates the management of crews performing 
project work. See Figure 5. 

Specific functions handled by this module include: 

• Gang Administration, Including Performing Gang: 

• Availability and Scheduling Analysis 

• Work Reporting 

• Productivity Reporting 

• Unit Cost Reporting 

• Consist Optimization Analysis 

• Personnel Administration, Including Performing Personnel: 

• Availability and Scheduling Analysis 

• Work Reporting 

• Productivity Reporting 

• Roster Maintenance, including updating 

+ Job Assignments / Work History / Seniority / Ratings 
+ Rules Certification / Safety Training / Etc. 

• Personnel Department Linkages 

• Accounting Department Linkages 

The Work Equipment Management Module (WEMM) automates the management of equip- 
ment used to perform project work. See Figure 6. 



Specific functions handled by this module include: 

• Work Equipment Administration, Including Performing WE: 

• Availability and Scheduling Analysis 

• Work Reporting 

• Productivity Reporting 

• Unit Cost Reporting 

• Equipment Consist Optimization Analysis 

• Maintaining a Work Equipment Inventory 

+ Job Assignment / Work History / Reliability Ratings 

• Work Equipment Maintenance Administration, Including: 

• Maintenance Scheduling 

• Maintenance History 

• Maintenance Cost Reporting 

• Maintenance Shop Administration 

• Material / Store Department Linkage 

• Work Equipment Parts Management, Including Maintaining: 
+ Equipment Parts Lists 

+ Parts Inventory 

+ Parts Costs 

+ Parts Delivery Appraisal 

+ Parts Order Forecasting 

+ Parts Condition & Acceptability Reports 

+ Parts Returns Reporting 



126 



Bulletin 760— American Railway Engineering Association 



V 



Gong Productivity 



Gang 
/lanagement 
odule (GMM; 



Personnel MoncgemerTlt 
Labor Reporting / 

Work Histories / 

Productivity Reporting 
Personnel Rotin 



Figure 5 



The Material Management Module (MMM) automates the management of material used in 
a project. See Figure 7. 

Specific functions handled by this module include: 

• Maintaining Material Inventories and Catalogues 

• Material Cost Record Maintenance 

• Current Cost Data 
• Historical Costs 

• Cost Forecasts 

• Producing Material Requisitions 

• Performing Material Delivery Analysis 

• Performing Material Utilization Analysis 

• Purchasing and Material Department Linkage 

• Accounting Department Linkage 



Paper by D. E. Bartholomew 



127 



n\ 



W<t)rk Equipment Productivity 



Work Equipment Availability 




Work Equipment 

Management 
Module (WEMM 



/ 



Work Equipment Inventory, 
Maintenance Scheduling, 
Parts Projections, 
Utilization Reporting, 
Productivity Reporting, 
Failure Reporting, Ratings 



WorK Equipment Requsitions 
Parts Requisitions 



^ 



Figure 6 



The Project Management Module (PMM) automates the management of project field work. 
See Figure 8. 

Specific functions handled by this module include: 

• Managing Schedule Adherence, Including: 

• Preparing Progress Reporting 

• Performing Critical Path Analysis 

• Analyzing Schedule Revisions 

• Performing Resource Allocation Analysis 

• Performing Productivity Analysis and Preparing Reports 

• Managing Material Utilization 

• Preparing Condition and Acceptance Reporting 

• Preparing Return Reporting 

• Preparing Work Equipment Utilization Reporting 



128 



Bulletin 760— American Railway Engineering Association 



Parts Availability 

Parts Delivery Schedules 



Materiel 
Delivery 
Schedules 



Material 
Availability 



Material Management 
Module (MMM) 



Material Costs 



Material Inventory 
Materiel Forecasts 
Vendor Orders 
Material Appraisals 



I 



Figure 7 



• Preparing Project Completion Reporting, Including: 

• Work Documentation 

• As-Built Drawings 

• Fixed Facility Inventory Link 

Fixed Facility Inventory Module (FFIM). A major information management and storage 
module that automates the maintenance of all fixed facility records. See Figure 9. 

Specific functions handled by this module include maintaining current and historical records of: 

• Railroad Location 

• Line Segment 

• Mile Post 

• Track Number 

• Station and Offset 

• Field Survey Records 

• Satellite Geopositioning 

• Video Records 

• Line Service Data 

• Tonnages 

• Line Classifications 

• Ratings 

• Clearances 

• Track Quality Indexes 

• Track Inventory 

• Inspection Reporting 

+ Visual Inspection Records 

* Rail / Tie / Ballast / Subgrade Conditions 
+ Slow Orders 
+ Geometry Car Records 
+ Rail Detector Car Records 
+ Track Loading Vehicle Records 
+ Tie Condition Records 

• Data Elements and Codes 



Paper by D. E. Bartholomew 



129 




Gang and 
Personel 
Reports 




Project 
Schedule 
Revisions 



Work Equipmen 
Productivity 
Reports / 




Schedule Adherence 
Analysis and Reports 

Critical Path Analysis 
and Revisions 

Productivity Anolysis 

Work Orders 



.J- 



Materiel Utilizotion/Reports 
Material Returns/ 



_Z_ 



Project Completion 
Reports 

As Built Reports.. 



Chan-^e^rders 



Project Jvlonagement 
M;xiule (PMM) 



Field Reports 







Figure 8 




Current Facility 

Conditions 
Historicol Facility 

Conditions 
Improvement 

Recommendotions 
Degrodation Rotes 




Generol Inventories 
Global Facility Locotions 
Track Inventory 
Bridge Inventory 
Misc. Facility Inventories 



Clearances 
Ratings 



Detoiled Bridge Inventory 
Misc. Detailed Inventories 
Video Facility Records 



Fixed Facility Inventory 
\ Module (FFIM) 



Figure 9 



Bridge Inventory 

• Inspection Reporting 

• Visual Inspection Records 

• Timber Cruising Records 

• Fatigue Testing / Analysis Records 

• Nondestructive Testing Records 

• Data Elements and Codes 



30 Bulletin 760 — American Railway Engineering Association 



• Other Facility Inventories 

• Stations 

• Buildings 

• Culverts 

• Scales 

• Grade Crossings 

• Signals 

• Data Elements and Codes 

Computer Aided Drafting and Design (CADD). An important submodule of the Engineering 
Design and Administration Module that automates the production of engineering designs and draw- 
ings. Frequently this submodule overlaps with the Fixed Facility Inventory Module as described ear- 
lier. See Figure 10. 

Specific functions handled by this module include: 

• Interactive or Automated Production of Engineering Design Drawings 

• Architectural 

• Bridge 

• Civil / Surveying / Coordinate Geometry 

• Mechanical 

• Signal and Communications 

• Defining Mapping Standards: 

• General Cartographic and Digital Mapping Standards 
-t- Map Classes and Titles 

+ Sheet (Plot) Sizes 

+ Scales 

+ Lettering and Fonts 

-I- Levels and Layers 

+ Cardinal Points 

+ Topographic Detail 

Right-of-Way and Track Maps 

Station Maps 

Profile Maps 

Track Charts (Condensed Profiles) 

Railroad Valuation Detail 

Lease / Tenant Properties and Occupancies 

Zoning / Land Use / Taxation / Assessments 

Deed and Conveyance Rights / Interests 

Aerial and Ortho Photography 

Mapping Database Utilization 

+ Digital Linkages 

Document Management (DM). A submodule of the Engineering Design and Administration 
Module that automates the management of engineering documents, including electronic text and 
drawing storage and distribution, and maintenance of document inventories. See Figure 1 1 . 

The proposed AREA Manual material for this Section will define the following: 

• Interchange Standards 

• Hardware Requirements 

• Software Requirements 

-I- Communications Programs 
+ Access and Security 



Paper by D. E. Bartholomew 



131 




Drawing 
AvoilabI 



Design 

Drawii 

Moppin 

Profiler 





Track Alignnnent 
'l^esign 

StVuctural Diaqrams 
Co6rclinate Geonnetry 
Eartkjwork Quantities 

Computer Aided 

Drafftng and 
Design \CADD) 



General inventories 
Global Facility Locations 
Track Inventory 
Bridge Inventory 
Misc. Facility Inventories 



f\xed Facility Inventory 
' xModule (FFIM) 




Figure 10. 



+ Anti-Virus Programs 
+ Backup Programs 
Storage and Archive Standards 

• Hardware Requirements 

• Software Requirements 

• Data File Formats / Organizations 

• Scheduling and Rotation 

• Compression Techniques 



132 Bulletin 760 — American Railway Engineering Association 



• Translation Standards 

+ New Software Release Translations 
+ System to System Translations 
+ Raster < > Vector Translations 

• Data Availability and Sources 

• Railroad Sources 

+ Digitization Standards 
+ Other Railroads 

• Government Agency Source Standards 

• Manufacturer/ Vender Source Standards 

The Engineering Design and Administration Module (EDAM) automates engineering 
design and the administration of an Engineering Department, including personnel scheduling, design 
quality and productivity analysis, design drawing availability, etc. See Figure 12. 

Specific functions handled by this module include: 

• Structural Analysis 

• Bridge Design 

• Bridge Ratings 

• Building Design 

• Rail Stress Analysis 

• General Engineering Design 

• Trackwork Designs 

• Alignment (Coordinate Geometry) 

• Earthwork Quantities 

• Embankment Stability 

• Engineering Administration 

The High-Wide-Heavy Load Authorization Module (HWHLAM) automates the authoriza- 
tion of high, wide or heavy load passage over various segments of a railroads lines. See Figure 13. 

Specific functions handled by this module include providing 

• Fixed Facility / Load (Shipment) Clearance Assessment 

• Weight Rating / Load (Shipment) Assessment 

• Line / Route Optimization Analysis 

• Authorization Documentation, Including: 

• Restrictions 

• Special Instructions 

• Train Orders 

• Restriction Improvement Analysis, Including 

• Exception Reporting 

• Clearance Improvement Priority Analysis 

• Reinforcement Priority Analysis 

• Client Accommodations 

• Transportation Department Linkages 

• Fixed Facility Inventory Linkages 

Standard REFMS Process and Information Flow 

In the interest of saving time, I will not go over the Process and Information flow in detail here 
since it is well illustrated in Figure 1.3 which was described above. In reality, each module does not 
pass information or data directly on to the next module. In performing their functions, the modules 
having appropriate data management responsibility and authority, update the Data Files listed within 
the large solid drum symbol shown in the center of Figure 1 .3. These Data Files, as well as any oth- 



Paper by D. E. Bartholomew 



133 



r 



Document 

Management 

(DM) 



Digito 

Document 

Storage 
Document 

Inventory, 

Translation, 

and Display 

\ 

growing 
\Transmission 
Efectronic Mai 
Transm's. 





Engineering 

Design and 

Administration 

Module 

EDAM 




Trocl< Alignment 
Design 
StVucturol Diagrams 
Coordinate Geometry 
Eart'Kjwork Quantities 

Comp~uter Aided 

Draftmg and 
Design ''(CADD 



I 



Figure 11 



ers needed, are then accessed by the next module to pass the information along. Thus, in reality, data 
is passed back and forth between all modules, which would be even more confusing to illustrate than 
the primary information flows shown in Figure 1.3. 

It is recommended that all data be stored in one relational "Data Bank" so that data has to be 
entered only once and the amount of data stored is minimized (no duplication). See Figure 14. In real- 
ity, it is not necessary that all data reside at the same physical location, but that it all be relationally 
linked. 

Each REFM System functional Module has data management responsibility and authority to 
update the following Data Files: 

• Project Appraisal and Forecasting Module 

• Quality Index Data 

• Project Data 

• Cost Estimating and Budgeting Module 

• Unit Costs 

• Cost Projections 

• Project Scheduling Module 

• Scheduling Data 

• Seasonal Restrictions 

• Gang Management Module 

• Gang Costs 

• Work Equipment Management Module 

• Work Equipment Inventory 

• Maintenance Data 

• Parts Data 



134 



Bulletin 760 — American Railway Engineering Association 



V^ 



/ V4 




Condition 


Evoluo 


Ma 


teria 


L 


sts 


Est 


imot 


ed 


Work 



Ratings ond Restrictions j 

Document 

Management 

(DM) 

Oigitol 

Document 

Storage 
Document 

nventory, 

Tronslation, 

ond Display 




Drawing 
\Transmission 
Electronic Moil 
F/vX Transmis. 




esign Time 
./Drofting Time 



Structural Anoiysis 

-I : \ ^ 

Engineering 

Design and 

Administration 

Module 

(EDAM) 



Track Alignment 
Design 

Structurol Diagrams 
Coordinole Geometry 
EortXiwork Quantities 

Compljter Aided 

Drafiing and 
Design \CADD) 





General Inventories \ 

Global Facility Locations 
Track Inventory 
Bridge Inventory 
Misc. Facility Inventories 



Figure 12 




Routing Analysis 







High-Wide— Heavy 
Load Authorization 
Module (HWHLAM) 

Clearance and Rating Approisal 
Restrictive Facility Analysis 
Improvement Appraisal 



Facility Restrictions 
Improvement 
Recommendations 




Figure 13 



Paper by D. E. Bartholomew 



135 




Figure 14 



Material Management Module 

• Material Data 

• Cost Projections 
Project Management Module 

• Project Costs 

Fixed Facility Inventory Module 

• Inventory Data 

• Location Data 

• Historical Data 

Computer Aided Drafting and Design 

• Drawing Files 
Document Management 

• Digital Document Storage 

• Document Inventory 



136 Bulletin 760 — American Railway Engineering Association 



Conclusions 

Based on its investigation and review work described in this presentation, AREA Committee 32 
has come to the following conclusions: 

• Railway Engineering Facility Management Systems (REFMS) Help Engineers Improve 
Infrastructure Reliability 

• Standardization of REFMS will Provide Improved Systems 

• There will be Great Benefit from Your Participation in and Your Railroad's Support of the 
AREA Manual Chapter Development Work of AREA Committee 32 

• You should Become an Active Member of AREA Committee 32 




"From sea to shining sea..." Premier Crossing 
Systems are installed at some of the finest 
port and intermodal facilities coast to coast. 
There's a good reason. 



Innovative Premier 
Advanced Panel 
Lag-Down or Lag- 
Free Systems 
where panels go 
right onto wood or 
concrete ties. 





Tieless, Premier 
modular systems 
for LRT installa- 
tions and absolute 

rail isolation. 



Partnering 

Working together to bring 
crossing projects in On 
Time & Under Budget. 

That means 

. working with railroad, contractors, 
and plant maintenance people to 
develop crossing solutions. 
. putting together the right system for 
you to provide low installation costs, 
low maintenance costs, and long 
crossing life. 

. providing on-site assistance from a 
Premier® Field Representative to 
ensure quick and easy installation. 
We call it partnering. 
From the Port Authority of New York 
& New Jersey across the nation to the 
Port of Los Angeles. For intermodal 
facilities, light rail, heavy rail, or 
industrial sites. Premier Crossing 
Systems^"^ are fulfilling industry 
needs. Meeting schedules and 
deadlines. Keeping costs down with 
low installation and maintenance costs. 
And building in long crossing life. 
There are more Premier benefits — 
including salt (Chloride) resistance, the 
non-skid diamond plate crossing 
surface, and rubber flangeway inserts 
to provide a positive shunt-free 
system. 

The original concrete crossing. 
Whatever your next crossing project 
calls for - from panels on concrete or 
wooden ties to a modular, tieless 
crossing system, we can meet your 
needs. Write to us at RO.Box 1 1305; 
Portland, OR 9721 1. FAX 503-240- 
3592, or call to learn more about the 
full line of Premier Concrete Systems... 
1-800-426-5556. 



PREMIER 

CONCRETE RAILROAD CROSSINGS 
Partnering solutions for rail industry challenges. 



137 




....Changing the Law of Gravity. 



138 



MANAGING RAIL RESOURCES 

By: Joe Kalousek* and Eric Magel** 

Abstract 

All railroads, large or small, must replace broken, worn or defective rail, with rail replacement 
costs ranging from 200 to 350 thousand dollars per track mile. This high cost means that any exten- 
sion in rail life will result in appreciable savings and consequent increase the railway profitability. 

Rail life is governed by wear, contact fatigue damage and internal defects. Most of the wear 
occurs at the gauge comer in mild and sharp corners alike. A small amount of wear is cau.sed by 
wheels contacting the top of the rail, but on many railroads most of the vertical wear is due to grind- 
ing. Rail which does not wear will eventually succumb to rolling contact fatigue where microcracks 
at the surface propagate into larger ones until the rail is eventually infested with a network of cracks 
to the depth of 10 to 15 mm. 

Excessive wear results in a shortened wear life, while no wear leads to the development of con- 
tact fatigue defects. Maximizing the rail life requires wear to occur at just the right rate — the "magic 
wear rate." The railway engineer has several options for achieving optimal wear: usage of premium 
rail materials, improvements in truck design or maintenance to enhance their curving ability, lubri- 
cation of the gauge face/wheel flange interface and proper selection and management of the wheel 
and rail profiles. 

Introduction 

Improved materials and maintenance practices have dramatically extended rail and wheel life 
within recent years. Despite the improvements however, rails are often worn and/or damaged and 
replaced before their time. A significant part of the reason is that railroads have not effectively man- 
aged the interaction at the wheel/rail interface. 

Research and "real world" operating experience have shown that the most effective way to 
extend the life (and maintenance cycles) of rail is by treating track/vehicle interaction and the 
wheel/rail interface as a system. Controlling the damage that wheels and rails inflict upon each other 
is an essential element of .safe and cost-effective operations. An optimized .strategy is arguably the 
most valuable and effective means of directly controlling maintenance and capital costs — the cost of 
rail grinding and ultimately rail replacement. 

Extending the life of rail alone, which can account for 60% of a railway's total asset value, can 
add millions of dollars to the bottom line. Taken in conjunction with extended maintenance cycles, the 
savings are significant. While this is true under exi.sting operating conditions, the need to economize — 
to optimize — continues to accelerate as marketing forces push railroads to adopt heavier and heavier 
axle loads. 

While there is no quick or simple fix for problems at the wheel/rail interface, there are ways to 
manage and minimize the damage that occurs there (Figure I). Track engineers have several options 
for achieving long rail life: managing friction, grinding the rail to properiy selected rail profiles (con- 
tact mechanics), adopting premium rail materials, and collaborating with equipment engineers to 
institute improvements in truck design or maintenance to enhance their curving ability. Exploiting 
these options to effectively manage rail resources is the focus of this paper. 



■ Piincipal Research OlTiccr — Ccmrc lor SurCacc Tran.sportalion Technology, Nalional Research Council of Canada 
* As.socialc Research Officer — Centte for Surface Tran.sportalion Technology. Nalional Research Council of Canada 



139 



140 



Bulletin 760 — American Railway Engineering Association 




Figure 1. Tools for controlling damage modes to optimize rail and wheel life. 



Rail Metallurgy 

Life of the rail can be extended by increasing its resistance to wear, surface fatigue damage, and 
fatigue defects. Resistance to these damage modes increases with hardness, cleanliness, and fracture 
toughness. The effect of hardness on gauge face wear resistance has been well demonstrated [1]. 
Increased rail hardness in combination with minimized sulphide inclusions reduces the propensity to 
surface fatigue crack initiation and subsequent development of defects such as head checks, flaking, 
and shelly spots. Oxide inclusion clean rail combined with good fracture toughness reduces the like- 
lihood of deep seated shells formation. Both shelly spots and deep seated shells can initiate transverse 
defects as shown in Figure 2. Frequent occurrence of these transverse defects results in premature 
removal of the rail from the track and can also be a direct cause of costly derailments. 

Most surface fatigue cracks initiate by the mechanism of ratcheting whereby the rail surface 
layer is plastically strained in an incremental manner until its ductility limit is exhausted. At this point 
a microcrack occurs. Ratcheting and sub.sequent crack initiation is caused by a combination of high 
contact stres.ses, low rail hardness, friction, and the possible presence of "coarse" sulphide inclusions. 
The relationship between the first three parameters is shown in Figure 3. The shakedown diagram 
shows that the resistance to ratcheting drops off dramatically when the traction coefficient exceeds 
0.3. At traction coefficients above 0.3, which frequently occur in sharp curves, the increased hard- 
ness combined with rail profiles which reduce contact stress can minimize the load factor (Po/ko) and 
substantially delay the onset of surface damage in the rail. 



Paper by J. Kaolousek and E. Magel 



141 




Figure 2: a) Transverse defect from flake; b) Transverse defect from deep seated shell. 







Surface + Subsurface Flow 






Subsurface Flow 




4 


\ Repealed Plastic Flow 






? — ~ Y 








N. \ 






V \ 






\ \ 






N \ 






Elastic ^ ^ \ 






Shakedown ^ ^ \ 


^ 




■■••■... ^~ ^ \ Surface Flow 


— . 


3 


_ ■ • . . ~^ ^ \ 


L_ 






o 




\ 


^rf 






o 




^v 


(0 




Elastic \. 


LL 




\. 


T3 




^v. 


(0 

o 


2 


. \^^ 


-I 








.. ^^v> 






Elastic Limit 










Elastic-Perfectiy Plastic Shakedown Limit 






1 












1 1 1 1 1 





0.1 0.2 0.3 0.4 0.5 

Traction Coefficient |j 

Figure 3. Johnson's shakedown diagram showing regions of acceptable cyclic stress as a 
function of maximum contact stress {P„), friction (jx), and yield in shear (k ). 



142 Bulletin 760 — American Railway Engineering Association 



Oxide inclusions, especially aluminum oxide stringers, have been associated in aluminum 
deoxidized steels with the initiation of transverse defects from deep seated shells. However, deep 
seated shells have also been demonstrated to initiate at pearlite grain boundaries in clean steels [2]. 
Once a crack is initiated, it will propagate at a rate inversely proportional to fracture toughness. Thus, 
steels with high fracture toughness permit longer defect inspection intervals. 

While manufacturers have significantly increased the hardness and improved the cleanliness of 
rail steels within recent years, the hardness of pearlitic steels appears to have peaked at about 390 
Brinell (BHN) [3]. Researchers have now begun looking at other structures such as bainitic steels. 
Although bainitic steels of the same hardness as pearlitic steels are not as wear resistant, high har- 
ness, low carbon bainite offers wear resistance superior to pearlite. In addition, tests carried out by 
McEvily and Minakawa [4] have shown that increasing the hardness from 370 BHN to 450 BHN can 
improve fatigue life by a factor of 2.6. Internal residual and thermal stresses also significantly affect 
rail failures [5]. Research continues on ways to improve rail steels, thereby reducing the number of 
costly and potentially dangerous transverse defects. 

At present, the best recommendation for heavy haul track is to install clean rail with a hardness 
of about 300 to 320 BHN, 340 to 360 BHN, and 380 to 390 BHN, into tangent track, mild curves, 
and sharp curves, respectively. 

Recent developments in rail welding technology are also producing high quality results in plant 
and field production of welds with long fatigue life [6]. 

Wheel/Rail Dynamics 

Regardless of the quality and hardness of rail and wheel steels, problems can — and do — result 
with inappropriate combinations of wheel and rail profiles, and poody matched truck and track 
designs. Under static loading conditions, each wheel carries an equal portion of the car load. Under 
dynamic conditions, the truck suspension distributes the load and regulates the dynamic impacts. 

Trucks, by virtue of their longitudinal bending and lateral sheer stiffness, govern the wheelset 
curving and hunting behavior [7,8]. While ideal truck designs permit the wheels to roll freely through 
a curve and avoid flange contact, the standard North American three-piece trucks cannot negotiate 
curves without the wheelsets taking an angle of attack. High quality maintenance of existing trucks 
significantly contributes to better rail management, because misaligned trucks are susceptible to 
flanging even under mild curving conditions. Ultimately, poor steering generates high lateral forces 
and contributes to excessive rail and wheel wear. 

Through the use of computer modeling systems, dynamicists are now able to design trucks and 
railway vehicles to optimize curving characteristics and improve overall ride quality. Self-steering 
truck designs, for example, allow the wheelsets to respond to steering moments and align themselves 
more favorably with the rail, thereby reducing the angle of attack. As a result, self-steering trucks can 
reduce the creepage between the wheel and rail, and minimize wear and rolling contact fatigue. 
Body-steered trucks are able to eliminate flange contact in curves of up to 12°. 

For standard three-piece, self-steering, and body-steered trucks, the division between mild and 
sharp curves (i.e., the degree of curvature at which the flange force reaches its maximum) is about 
3°, 6°, and 12°, respectively. 

Hunting, which occurs above a critical speed, is influenced by many parameters, the most 
important being the truck design itself and the effective conicity at the wheel/rail contact. Although 
cross-braced trucks are less prone to hunting than worn standard three-piece trucks, most trucks — 
regardless of their design — will hunt in situations where effective conicity exceeds a certain limit. 
High effective conicity in tangent track occurs when the wheel and rail profiles are conformal, i.e., 
when the rails are "flat" (large crown radius) and the wheel treads are hollow. To avoid this situation 
the track engineer should select rail profiles for tangent track and curves to prevent the wheel tread 



Paper by J. Kaolousek and E. Magel 



143 



from hollowing; i.e., the envelope of all rail profiles a wheel contacts in different sections of track 
should wear the wheel "flat." 

Rail Profiles 

During curving, the contact stresses at the gauge comer are usually twice as large as those 
between the rail crown and wheel tread. To reduce contact stresses at the gauge corner and the gauge 
shoulder in curves, it is important that the wheel/rail profiles be conformal. The wheel profile is con- 
formal to the rail profile if the gap between the profiles is less than 0.5 mm (0.020 inches) at the cen- 
tre of the rail (in one point contact) or at the gauge comer (in two point contact). Figure 4 schemati- 
cally shows three sets of confomial wheel and rail profiles (the gaps are not shown). W and R refer 
to wheel and rail profiles while H refers to the high wheel/rail interface in a curve. The numbers (1, 
2 and 3) matching increase in gauge comer relief. The third set could be said to be more "conical" 
than the second or first sets. 

The table at the bottom of Figure 4 .shows two .sets of conformal combinations of wheel/rail pro- 
files consisting of dimensioned new or worn wheel and rail profiles. In the first set (left column) the 
rolling radii difference (Ar) increases by approximately 0.5 mm. In the second .set, Ar increa.ses by 
I mm to 1.5 mm. 

The question is, which of these sets of profiles represents the optimal combination with respect 
to contact stresses, resistance to formation of contact fatigue defects, and curving? Obviously, the 
third set of profiles offers the highest Ar. With a higher Ar, curving is better and contact stresses are 
lower due to reduced creepage in the wheel/rail interface. Low contact stresses, in turn, contribute 
significantly to the delayed formation of contact fatigue defects. Within the constraints of a particu- 
lar wheel/rail system (axle load, vehicle design, distribution of curves, etc.), opting for a more coni- 
cal geometry results in longer rail life and better management of rail resources. 

On the other hand, the best way to waste valuable rail resources is to artificially introduce a 
non-conformal profile into a wheel/rail system which has wom conformally. About a decade ago, 
136RE-14" head hardened rail was introduced into a system with wom wheels represented by the CN 
Heumann profile. This generated a non-conformal one point contact (the HW3-HRI combination of 
profiles). Although this combination is excellent for .steering, the contact stresses at the gauge comer 
were so high that the rail had to be removed after about 44 million tonnes of traffic due to contact 
fatigue defects [9]. 




AARStd -136RE 14" 
AAR1B-136RE10" 
CN Heumann - NRC H2 



Ar» = Afg = 1 - 1 .5 mm 



AAR1B -136JK8" 

QCM-NRCH2 

NRC-ASW-NRCH4 



Typical Ar gains achieved with various combinations of wheel - rail profiles 
Figure 4. Generic sketch of three distinct sets of conformal wheel-rail profiles. 



144 



Bulletin 760 — American Railway Engineering Association 



In any closed wheel/rail system (mass transit, a mining railroad, a coal line, or the FAST heavy 
axle loop), the wheels and rails wear conformally to each other. How conical the conformal profiles 
will become depends on factors such as the ratio of sharp and mild curves to tangent track, initial 
wheel profile, and the way the track is ground. Figure 5 shows flange root geometries of four differ- 
ent wheel profiles. The QCM profile represents an average worn wheel profile on 100 1 cars operated 
by the Quebec Cartier Mining railroad, while the NRC-ASW profile represents an average worn 
wheel profile on CNR and CPR 100 t coal and grain cars. 

Numerous grinding tests have shown that any time the gauge comer of the rail is over-relieved 
by grinding in a short section of the track, a non-conformal two point contact (the HW1-HR3 com- 
bination of profiles) is introduced. As a result, the contact area supporting the vertical load is trans- 
lated from the gauge shoulder towards the rail center, thereby reducing the available Ar — in some 
instances — to nothing. This combination may leave rust at the gauge comer of the rails, is bad for 
steering, and wastes the rail through accelerated gauge face wear. Figure 6 shows the set of high 
(H1-H4) and low (L1-L3) rail profiles which represent worn rail profiles used by the LORAM/NRC 
template and electronic rail grinding gauges. 

Grinding 

For a rail grinding program to best manage the rail resource, it must be sensitive to the specific 
wheel/rail system. It must strive to grind the appropriate profiles on each curve and tangent track in 
such a manner that the contact stresses with respect to the average worn wheel profile are minimized. 
Rail profiling must also seek to provide the greatest possible contribution to wheelset steering and, 
at the same time, control hunting in tangent track. It must maintain conformal wheel/rail contact in 
curves and non-conformal contact in tangent track. 

Although conformal profiles reduce contact stress in curves, the stress may still exceed the yield 
of the rail — including that of head hardened rail. Excessive gauge face wear or vertical wear due to 
frequent corrective grinding, for example, dramatically shortens the rail life. Insufficient vertical 
wear ensures premature rail failure due to rolling contact fatigue. The mechanism of such failures, 
including crack initiafion and propagation, is described in a number of publicafions. [11,12]. 

The minimum amount of material removed to prevent the development of contact fatigue 
defects is referred to as the "magic wear rate." Any time actual wear exceeds this rate, rail metal is 
uselessly wasted. When the actual wear is less than the magic wear, flaking, spalls, deep seated shells. 



f parallel to wheelset centerline 




NRC-ASW 

QCM 

CN HeiHTionn 

AAR1B 



Toping Line 



Figure 5. Flange root geometry of new AARIB and several worn wheel profiles. 



Paper by J. Kaolousek and E. Magel 



145 



a) 



b) -- 




Figure 6. LORAM/NRC set of eight 8" railhead profiles [10]. 



and other contact fatigue defects develop. Thus, the magic wear rate is an optimal value in which 
fatigue and wear are in balance. Its value changes with many variables including quality of track, traf- 
fic mix, and truck and wheel maintenance. In most heavy haul territories where the rail is lubricated, 
the vertical wear is insufficient and grinding is necessary if the rail life is to be maximized. 

Grinding requirements vary across the many unique rail systems in North America, not only 
with respect to depth of cut (governed by the magic wear rate), but also with respect to rail profiles. 
The LORAM/NRC set of profiles in Figure 6 is designed to allow for selection of a profile which, at 
any location in the track, is conformal to the wheel. In addition to the wheel profile, many other influ- 
encing factors (dynamic wide gauge, type of track fastening system, degree of curvature, hardness of 
the rail, grinding interval) affect the selection of the correct profile to which the rail is to be ground 
[10]. Selection of the most suitable profile cannot be over-emphasized as too much or too little relief 
of the gauge comer will result in increased gauge face wear or accelerated contact fatigue, respec- 
tively. 

The best way to achieve the magic wear rate and maximize the rail life is through frequent 
grinding in which the smallest possible amount of material is removed. This can be achieved by a 
preventive, one pass, rail grinding program in which no more than 0.004"-0.008" is removed from 
the gauge comer of the rail, and no more than 0.002"-0.006" is removed from the rail crown. For 
track containing clean premium rail, the grinding interval should be 10-15 MGT in sharp curves and 
24-30 MGT in mild curves. Tangent track with plain carbon or intermediate grade rail would be 
ground at 30 MGT and 45 MGT intervals. 

Figure 7 shows the relationship between crack growth and preventive grinding carried out to a 
depth of 0.2 mm (0.008") with a grinding interval of 10 MGT. At a vertical wear limit of 20 mm 
(0.8"). this type of preventive grinding would yield a magic wear rate of 0.02 mm/MGT 
(0.0008"/MGT). The rail life with respect to wear and contact fatigue would then be estimated to 
reach 1000 MGT for sharp curves and 2000 MGT for mild curves. 



146 



Bulletin 760 — American Railway Engineering Association 



Accumulated traffic (MGT) 
20 



^^ 


0? 


b 




b. 


0.4 


2 


0.6 


» 




c 


0.8 


1^ 




u 


1 


f! 




«? 


1.2 


s 




in 


1 4 


E 




P 


1 fi 






r 




n 


1.8 


0) 




O 


2.0 




crack growth without preventive grinding 
"magic" wear rate 



Figure 7. Contact fatigue crack growth in rail located in a sharp curve (>3°) as a function of 

accumulated traffic with and without preventive grinding with an assumed "magic" wear 

rate of 0.02 mm/MGT (0.0008"/MGT). 



Friction Management 

The objective of friction management is twofold: a) maintain the coefficient of friction at the 
flange/gauge interface at |x < 0.2 to reduce gauge face wear; and, b) to maintain the coefficient of 
friction at the tread/top interface at 0.2 < |x < 0.4. Lubrication of the gauge face can reduce wear by 
as much as 200 times over the unlubricated condition. At the top of the rail, too low friction is respon- 
sible for skid flats, long braking distances, runaway trains, low dispatchable adhesion, rapid propa- 
gation of contact fatigue cracks, and insufficient wear which promotes the development of contact 
fatigue. Too high friction (0.4 <\x< 0.7) meanwhile produces large flange and L/V forces in curves, 
contributing to tie plate cut-in, rail seat abrasion, increased gauge face wear, and low rail rollover 
derailments. It also causes rapid initiation of contact fatigue spalls and shells, and promotes hunting, 
squealing, and the growth of roll-slip induced corrugations. Managing friction at the wheel/rail con- 
tact patch is arguably the single most powerful tool available to railway engineers. 

Left to itself, the coefficient of friction at the flange/gauge contact zone typically ranges from 
0.3 to 0.6 [13]. At friction levels this high, the gauge face is rough; large wear particles ("dandruff) 
are seen on tie plates; and wear rates, rolling resistance, and fuel consumption are high. 
Implementation of a meticulous lubrication policy results in reduced friction at the gauge face to 0.15 
or lower and reduces wear to a fraction ('/20 to '/200) of wear typical of unlubricated conditions. It also 
reduces the energy needed to pull a train through curves by 10-20% in some cases. Good quality 
flange/gauge lubricant is strongly recommended; but maintaining continuous lubrication is most cru- 
cial. Absence of lubrication does more damage to the rail than continuous lubrication with any lubri- 
cant. Solid lubricants are showing promise since they outperform semi-solid lubricants in mass tran- 
sit systems and may soon do so in the heavy haul industry as well. 

Friction at the tread/top interface is a function of pressure distribution and creepage, but the 
most influential factor is the composition of the "interfacial" layer. This layer is made up of residual 
lubricants; wheel, rail, and brake shoe wear debris; environmental contaminants such as coal dust, 
grain mash, leaves, clay silica, and dust; and moisture. The composition of such a layer changes con- 
tinuously depending on external inputs and outputs as well as changes in its mechanical properties 



Paper by J. Kaolousek and E. Magel 



147 



and chemical composition caused by wheel loading. To visualize these changes we have developed 
the "bathtub" model. This model accounts for the inputs and outputs as well as crushing, mixing, and 
"burning" processes which take place within the layer. The burning process is due to asperity flash 
temperatures, which can reach 6(X)°C over the volume of several jim' (Figure 8). 

With respect to the frictional properties of the tread/top interfacial layer, a railway engineer can 
influence its composition through the brakeshoe tap, the sand tap, and the friction modifier tap. As 
the friction modifier tap is the only input dispensing a substance of known and controlled friction 
value, it must be dispensed in sufficient quantities to overcome discharges from all the other taps to 
stabilize friction fluctuations at the rail. 

Traditional oil-based lubricants are unsuitable as a friction modifier since they may generate 
undesirably low friction in regions requiring adhesion and braking. New generation friction modi- 
fiers have been developed to maintain friction levels between 0.2 and 0.35 within the wheel tread/rail 
top interface. One of these friction modifiers — HPF (High Positive Friction) — has been developed to 
decrease friction from high to intermediate levels. However, the main advantage of this modifier is 
that it can eliminate negative friction which excites stick-slip oscillation. As such, it can be used to 
eliminate squeal or alleviate growth of short pitch corrugations in mass transit systems. There is a 
need to develop more advanced, low cost, tread/top friction modifiers which do not have the limita- 
tions of traditional oil-based lubricants. 

Concluding Remarks 

Improvements in rail metallurgy, truck design (wheelset dynamics), wheel rail profiles (contact 
mechanics) and friction management are the main tools available to railway engineers to remedy 
problems of gauge face wear, contact fatigue damage, and track damage from impacts caused by 
hunting trucks. In most cases, a solution will require a several pronged approach — any one tool in 
isolation will not be as effective as several remedies combined. 

The most economic approach is to take advantage of synergies which exist, for example, within 
a combination of friction management and improved truck design, or hard clean rail and conformal 
(high Ar) wheel/rail profiles. Some of these synergies have been described in this paper; others are 
yet to be discovered. 



STEADY 

WEAR 

TAP 



ERRATIC 
TRAFFIC 
TAP 



ERRATIC 

ENVIRONMENT 

TAP 



ON/OFF 
BRAKESHOE 

TAP 



VpW 



ON/OFF RAIN 
& MOISTURE 



ON/OFF STEADY OR 
SAND INTERMITTEN 

TAP F M. TAP 



^^ /|\V //T\ Z //ix^ /IT 




SOLID PARTICLE 
ATTRITION DRAIN 



SQUEEZE 
DRAIN 



Figure 8. The bathtub model of the wheel tread/top of the rail interfacial layer. 



148 Bulletin 760 — American Railway Engineering Association 



Any synergy which applies to one railway territory will not necessarily work in the same man- 
ner in another territory. Rail lines in the northern part of the North American continent, for example, 
are subject to moisture for a substantial portion of the year. They require different solutions which 
take advantage of synergistic relationships that may not apply in dry southern regions. Synergies 
must be explored for each individual situation (a specific rail line, territory, etc.). 

The more we learn, understand and apply the inter-relationships which positively affect the 
wheel/rail interface, the more economically we can manage a railway's the most valuable track asset — 
the rail. 

References 

1 . Hannafious J., "FAST/HAL rail performance," Proc. Workshop of Heavy Axle Loads, Pueblo, CO, 

October 14-17, A AR, 1990. 

2. Sugino K., Kageyama H., Kuroki T, Urashima C, and Kikucki A., "Metallurgical investigation of 

transverse defects in worn rails in service," WEAR 191 (1996), pp. 141-148. 

3. Steele R., "Next generation rail steels, "Advanced Rail Management Rail Maintenance Seminar, 

ARM, April 1996. 

4. McEvily A.J. and Minakawa K., "Metallurgical evaluation of FAST rail sheets," USDOT Contract 

#DOT-TSC-1551, University of Connecticut, August 1981. 

5. Igwemezie J.O., Kennedy S.L., Xiaodong F., and Rowan W., "Defective rail fracture under 

dynamic, thermal, and residual stresses, "Proceedings of the Fifth International Heavy Haul 
Conference held in Beijing, China, June 6-13, 1993, pp. 264-274. 

6. American Welding Society, "Recommended practices for the welding of rails and related rail com- 

ponents for use by rail vehicles," 1994. 

7. Scales B.T., "Review of freight car bogie design and performance," International Heavy Haul 

Association Conference on Freight Car Trucks/Bogies, June 1996, pp. 1.1-1.14. 

8. Scheffel H., Tournay H.M., and Frohling R.D., "The evolution of the three-piece freight car bogie 

to meet changing demands in heavy haul railroads in South Africa," Intemation Heavy Haul 
Association Conference on Freight Car Trucks/Bogies, June 1996, pp. 1. 15-1.29. 

9. Worth A.W., Homaday J.R., and Richards PR., "Prolonging rail life through rail grinding," 

Preceedings of the Third International Heavy Haul Railway Conference 1986, pp. 106-1 16. 

10. Kalousek J. and Stroba P., "Loram rail grinding gauge manual for corrective, maintenance and 

preventive grinding cycles," LORAM, December 1992. 

11. Grassie S.L. and Kalousek J., "Rolling contact fatigue of rails," to be published in the proceed- 

ings of the International Heavy Haul Conference, Capetown, South Africa, April 1997. 

12. Kalousek J., "Examining rail profiles and contact mechanics," Advanced Rail Management Rail 

Maintenance Seminar, ARM, April 1996. 

13. Kalousek J., Hou K., Magel E., and Chiddick C, "The benefits of friction management— a third 

body approach," Worid Congress on Railway Research, June 1996. 



ON THE BENEFITS OF RAIL MAINTENANCE GRINDING 

By: Allan M. Zarembski Ph.D., P.E.* 

Introduction 

Rail grinding is the process of removal of metal from the top surface of the rail head through the 
use of abrasive grinding materials. Rail grinding has been used by freight railroads and transit systems 
since the late I930's for the elimination of rail surface defects. Those early applications u.sed relatively 
unsophisticated rail grinding cars for the elimination of corrugations, engine bums, and batter at rail ends. 

During the next 4+ decades, grinding techniques improved, using primarily rotating grinding 
stones mounted on dedicated rail grinding cars or sets of cars, referred to as grinding trains. During 
this period, the application of rail grinding was extended to numerous types of rail surface defects to 
include; corrugations, joint batter, weld batter, engine bums, flaking and shelling, as well as for the 
grinding of mill scale from new rail [1]'. This mode of grinding for defect elimination, often referred 
to as "rail rectification", remained the primary use of rail grinding from the early applications in the 
I930's until the 1980's. [2]. 

During the period starting in early 1980, however, this defect elimination or rectification 
approach started to give way to the rail "maintenance" or "preventive" grinding approach. This latter 
approach does not allow surface defects to develop to any significant extent, but rather attempts to 
eliminate the development of these surface defects before they emerge on the rail head. It also makes 
extensive use of rail profile grinding techniques to control the shape of the rail head and the wheel/rail 
contact zones. A key driver to the development and implementation of this new approach was the 
development of a new generation of higher speed fully automatic rail grinding equipment that allowed 
for use of multiple grinding patterns and the real time variation of those pattems while grinding. 

This evolution from traditional grinding to maintenance grinding and the concurrent use of pro- 
file control has resulted in a significant broadening of the use of rail maintenance grinding techniques 
to increasing the service life and reduce the overall cost of rail in track. It has also led to improvements 
in wheel/rail dynamic interaction and the reduction of wheel/rail forces in both the vertical and hori- 
zontal plane (depending on specific profiles used). This reduction in dynamic interaction (and forces) 
results in improved rail life, noise reduction, and reduced damage to both the track stmcture and the 
rolling stock. 

Rail Grinding Applications 

Rail grinding, as noted above, can be divided into two broad categories, based on the specific 
objective and method of achieving that objective. These will be described in the following sections. 

Control of Surface Defects 

Control and/or elimination of defects on the top surface of the rail head represents the tradi- 
tional area of rail grinding [1,2] often referred to as rail rectification. Since these surface defects rep- 
resent locations where vertical wheel/rail dynamic activity is initiated, control of these surface 
defects results in a reduction in vertical dynamics, noise, vibration, and vertical impact forces. While 
this type of grinding has traditionally been one of the remedial type actions, i.e. elimination of defects 
after they appear on the rail head, earlier and more aggressive grinding standards have led to better 
control of this class of defects and the consequential reduction of their adverse impact on overall 
operations and costs. In some cases, the initiation of certain classes of surface defects, e.g. low rail 
freight corrugations, have been completely forestalled by the use of profile grinding techniques. 
Additional benefits have also been obtained by combining profile grinding with the use of improved 
rail steels (cleaner, higher strength, etc.) and/or improved lubrication practices. 



I 1 refer to references al end of paper 
* President, Zeta-Tech Associates 



149 



150 Bulletin 760 — American Railway Engineering Association 



Surface defects normally manifested themselves on the top surface of the rail head. These sur- 
face defect grinding applications include grinding of the following classes of rail surface defects (For 
a complete set of definitions refer to Chapter 2 of the Manual for Railway Engineering of the 
American Railway Engineering Association [3].): 

Corrugations 

Discrete Anomalies 

Engine or Wheel Bums 

Battered and/or Mismatched Joints 

Weld Irregularities 

Rail Head Damage 

Spalling or Flaking (Shelly Spots) 

Shelling (Gage Comer Shelling) 

Surface Batter/Crushed Head 

Plastic Flow (Lip, Flowed Rail) 

Rail Surface Roughness 

Mill Defects 

Mill Scale 

A further benefit that is obtained from the use of a regularly planned grinding activity to con- 
trol these surface defects is that these defects are caught and removed early in their formation cycle, 
while they are relatively "shallow". This results in a reduction in the dynamic loading caused by 
many of these surface defects, and a corresponding reduction to the damage caused not only to the 
rail, but also to the rest of the track structure. 

In current practice, grinding for the control of surface defects is frequently combined with pro- 
file grinding in a consolidated rail grinding activity. 

Profile Grindin}> 

Rail profile grinding refers to the method of controlling and maintaining the shape of the rail 
head (hence the term "profile") through the grinding of the head of the rail [2, 4]. Profile grinding 
goes beyond the basic defect removal approach of conventional grinding and addresses the control 
of the shape of the rail, and the associated interaction between the wheel and the rail, to include 
wheel/rail contact. 

This "shaping" of the rail head and the influencing of the wheel/rail interaction is a major dif- 
ference between traditional defect grinding and profile grinding. Traditional defect elimination tends 
to "flatten" the rail, as illustrated in Figure 1(a) [5]. Profile grinding, on the other hand, grinds a spe- 
cific contour or profile into the rail head (Figure l.b). It should be noted that contour grinding is used 
to restore the original shape or profile of the rail head, while profile grinding is used to give the rail 
head a special profile other than its original rail profile. Through the control of the rail head shape by 
means of profile grinding, the locations of wheel/rail contact, and thus the interaction between the 
wheels and the rail head, can be controlled. 

Elimination of surface defects, if present, is the necessary first step in profile grinding. Thus, 
for rail with surface defects and plastic flow, profile grinding can be a three step process, as illus- 
trated in Figure 2. The initial step consists of one or more grinding passes which eliminate any sur- 
face defects present. The second step also consists of one or more grinding passes which effectively 
reshape the deformed rail head. The third and final step (if necessary) grinds the final rail head pro- 
file. However, as maintenance grinding becomes an ongoing process with frequent grinding passes 
to maintain the desired rail head profile, surface defects generally do not have sufficient time to form, 
and profile maintenance can become a "one pass" process. Furthermore, the overall level of grinding 
required can be reduced through the use of a planned and ongoing grinding program to control these 
surface defects and maintain the profile of the rail head. 



Paper by A.M. Zarembski 



151 





Figure 1. Ground Rail Profile 

(a) "Flat" Profile after defect elimination grinding 

(b) "Contour" profile after profile grinding 

Rail profile grinding encompasses three broad areas of rail maintenance: 

1 . Control of gage face wear and lateral wheel/rail curving forces. 

2. Control of rail surface fatigue. 

3. Control of corrugations. 

While profile grinding can address all three of these maintenance areas, they can not usually be 
addressed simultaneously [2,6,7]. Thus, the profile that is best suited for the control of one of these 
maintenance areas may not be (and in fact is usually not) the best for the other two problem areas. It 
is therefore necessary to define the specific problem or class of problems to be addressed prior to the 
selection of a grinding profile and initiation of profile grinding. 

In all cases, it must be noted that the ground profiles deteriorate with traffic. In one set of tests, 
the profiles tested lasted only 10 MOT (in non-lubricated heavy axle load freight operations) and 
were completely gone after 20 MOT of traffic [8]. This indicates the profiles must be continuously 
maintained and that rail maintenance grinding is an ongoing activity that must be continued and 
maintained in a regular (and defined) basis. Furthermore, this deterioration can vary with geometry 
and track condition. Thus, for example, the rate of profile deterioration appears to increase with cur- 
vature, thus requiring more frequent grinding on sharper curves than on shallower curves and tan- 
gents. This in turn directly affects the nature and schedule of the grinding program. 

Reduction in Rail Wear 

The use of rail profile grinding to control wheel/rail interaction, wheel/rail contact, and (thus) 
rail wear was developed and introduced by the mining railroads of Western Australia during the late 
I970's [3]. It was subsequently introduced in North America in the early 1980s, concurrent with the 
introduction of the first fully automated rail grinding train, RMS-1 [2, 9] The focus of this initial 
application of profile grinding was on the optimization of the "steering" of conventional three piece 
freight car trucks [4]. The results were the development of a set of asymmetric rail head profiles (i.e., 
asymmetric about the center line of the rail head), with a separate profile for the high and low rails 
of the same curve. In addition, for tangent track, where "hunting" wear was noted, special tangent 
profiles were develop to control this form of wheel/rail behavior, and the resulting rail head wear [2]. 

The initial profile grinding concept was designed to make use of the steering of the conven- 
tional three piece freight car truck generated by the conicity of the wheelset [6]. By making use of 
the difference between the wheel radii due to this conicity, which is known as the "rolling radius dif- 
ferential", it is possible to compensate for the difference in length, around the curve, between the high 
rail and the low rail (by having the outer wheel ride on the larger radius portion of its tread, and the 



Bulletin 760 — American Railway Engineering Association 



Profiling 

Step 1: Surface irregularities are ground out 



Low Rail 




Gage 



Step 2: Reshape head deformation 



High Rail 




Gage 



Step 3: Final profiling 

Figure 2. Three Steps of Profile Grinding 



Rail Profile Grinding Concept 




New Profile 
Ground Profile 
<t A.B Show Shift 

After Grinding 

High Rail Low Rail 

Figure 3. Profile Grinding to Improve Car Steering 



Paper by A.M. Zarembski 



153 




^ 



r 



HEAD LOSS = 25% 
ON REMOVAL 



F G 



WITHOUT PROFILE 
GRINDING 





HEAD LOSS - 31% 
IN TRACK 




WITH PROFILE 
GRINDING 



Figure 4. Worn Rail Sections, with and without Profile Grinding 



inner wheel ride on the smaller radius portion of its tread, see Figure 3 [4])-. This results in a shift- 
ing of the wheelset, which in conjunction with the longitudinal "creep" force generated by the rolling 
radius differential (which tends to align the axles into a radial position [4] ) reduces flanging on rel- 
atively sharp curves, and has the potential for eliminating flanging on curves less than 3 degrees 
(based on a 1 :20 wheel conicity) [4], This has been the experience in Australia, as illustrated in Figure 
4, where increased wear life of the order of 70 to 80% has been reported [2]. 

Recent research by the Association of American Railroads [10] suggests that introduction of a 
conformal, single point contact, between the wheel and the gage comer of the high rail (in curves) 
will generate a lower wheel set angle of attack and reduced lateral curving forces as compared to a 
"two point" contact configuration. This is illustrated in Figure 5 for a range of lubrication conditions. 

Control of Rail Fatigue 

A second area of benefit associated with rail profile grinding is in the control of rail surface 
fatigue, and in particular fatigue defects at the gauge comer of the rail head. This includes both sur- 
face fatigue defects, such as spalling, and sub-surface fatigue defects, such as gage comer shelling, 
such as commonly found on heavy axle load freight operations. This is the area of benefit most fre- 
quently reported by North American heavy haul freight railroads [11,12]. 

In this application, profile grinding is used to relieve the very high contact stresses in the region 
of the gage comer of the high rail associated with single point contact in a severe flanging condition 
(such as on a sharp curve) - see Figure 6. These high stresses can result in gage comer fatigue prob- 
lems, including cracking and spalling. By grinding the gage comer of the high rail, the contact is 
shifted away from this comer and into a more central location on the rail head. In the case of sharper 
curves where flanging takes place, a second contact point between the flange of the wheel and the 
gage face of the rail can occur, thus generating "two-point" contact between the wheel and the rail. 
This dividing of the wheel/rail contact site into two points reduces the contact stres.ses at any one 
point and can result in a decrease in both surface fatigue "spalling" and sub-surface fatigue "shelling" 
[2, II]. [Note, it may be possible to obtain a similar effect by broadening the contact point, such as 
through the use of a "conformal" contact profile, provided that a sufficiently large conformal contact 
area is effectively maintained for the broad range of wheel profiles encountered in service. However, 
this may still require a proper grinding program to maintam an effective contact profile and to con- 
trol surface fatigue.] 



This profile is based on new or well maintained wheels. For worn wheels, such as hollow worn wheels wiih false flanges, ihis 
would have to he modified, such as by adding field side relief to the low rail. 



154 



Bulletin 760 — American Railway Engineering Association 




Avarag* Wh*«l sat Anol*-«r-Attaok 
8*otlon 2S. 9-0«gr«« Curv* 



2 Pt Centeet. Lsad Axt« 



Conformal, L«ad Axl* 

2 Rt Contaet. T rail A x l» 

" ^ ~ "^ 1* ConTormafTTrBll Axt< I 



High 
Rail 



Lubad 



Both 
> Ralla - 
Lubad 



a 

8 »r " + 



20 



CO 

e 

m 
o 

5 04- 



6 4- 



Avaraga Lataral Whaal/Rall Foreaa 
Inalda Rail Saetlon 25. e-Dagraa Curva 



2 Pt Contact Laad Axia 



Confbrmal. Laad AxIa 

2 Rt Contact. Trail AxIa 




Conformal, Trail Axl 



Both 

Ralla- 

Diy 



High 
Rail ■ 
Lubad 



Both 
- Ralla- 
Lubad 



Figure 5. Angle-of-Attack and Lateral Forces before (Conformal) and after (2 point contact) 
High Rail Gage Corner Relief Grinding 



HIGH RAIL 



WORN 
WHEEL 

RAILRELO 




Figure 6. Gauge Corner Spalling and Profile Grinding to Relieve it 



Paper by A.M. Zarembski 



155 





Ti 



"^^ 



^ 



SINGLE 

POINT 

CONTACT 



r=\ 



/^ 



TWO 

POINT 

CONTACT 



Figure 7. Single Point Versus Two Point Contact Steering Forces 



Gage comer grinding allows for a "wearing" away of the surface fatigue damaged rail steel, and 
a relocating of the (interior) point of maximum rail stress, before fatigue damage can initiate a fail- 
ure defect. This is particularly important for well lubricated track or for premium hardness rail steels, 
where the rate of wear has been substantially reduced. 

As was noted previously, this change in wheel rail contact, from one point to two-point contact, 
can result in a deterioration in truck curving performance [13], and a corresponding increase in the 
wheel/rail flanging forces (see Figure 7). The result of this can be an increase in gage face wear, if 
no other action is taken. Therefore, this type of gage comer profile grinding should be used primar- 
ily in those areas where rail fatigue and not rail wear is the dominant rail failure mode. 

Control of Corrugations 

The third area where benefit has been derived from profile grinding is in the area of comiga- 
tion control; in particular, the control of the heavy axle load short wave cormgations found on the 
low rail of curves, as commonly observed on North American freight railroads. These corrugations 
generally have wavelengths in the range of 12 to 24 inches on wood tie track [14]. 

These corrugations are often associated with the high contact stresses generated when the false 
flange of a worn wheel runs on the field side of the low rail. This contact, which is counter- formal 
(i.e., the curvature of the two bodies in contact, are opposite to each other), causes significantly 
higher wheel/rail contact stresses than the other (conformal) wheel/rail contact configurations [6]. 
When this high contact stress is located near the field side of the low rail, severe plastic deformations 
and corresponding short wave corrugations can result. 



156 Bulletin 760 — American Railway Engineering Association 



Profile grinding has been used to control these short wave ("freight") corrugations on North 
American freight railroads. By grinding the field side of the low rail to shift the contact point towards 
the center of the rail head, the high stress producing false flange contact is avoided. In tests on North 
American freight railroads, the use of profile grinding to control the regrowth of corrugations was 
found to significantly reduce this regrowth rate as compared to conventional (defect elimination) 
grinding patterns [6,7]. 

While in general it is not possible to effectively combine profiles (i.e. wear and fatigue control 
profiles have significant differences which do not lend themselves to combination into one profile), 
it is possible to control corrugations in conjunction with another profiling activity. Thus, most rail- 
road profiles include corrugation control in addition to control of either wear or fatigue. This has led 
to the significant reduction (and in some cases the elimination) of corrugations in an environment 
where frequent grinding passes are made to control and maintain the required rail head profiles (often 
in conjunction with the use of improved [higher strength] rail steels and/or lubrication). 

Grinding vs. Lubrication 

Rail maintenance grinding takes on increased importance in a overall effective rail maintenance 
program which includes effective lubrication, extensive defect testing, and the desire to obtain the 
maximum life of the rail in its first position. 

This becomes quite apparent in the rail degradation environment found on moderate and heavy 
curvature track under heavy freight loadings. If there is no effective rail lubrication, the result is a high 
rate of gage face wear, with rail (gage face) wear being the predominant cause for rail removal [15]. 
However, with the introduction of effective lubrication, and the corresponding dramatic reduction in rail 
wear, this wear mechanism is significantly reduced so that fatigue emerges as a critical criterion [16]. 

This behavior can be seen in examining the relationship between wheel and rail contact in mod- 
erate to sharp curves. In unlubricated track, rail degradation generally takes the form of severe adhe- 
sive wear on the high rail of the curve. In lubricated track, it takes the form of surface fatigue devel- 
opment (spalling) at the gage comer of the railhead. This build up of surface fatigue is aggravated in 
the case of well lubricated track, where no wear is allowed to occur. Without any significant railhead 
wear, this surface fatigue is allowed to cumulate, with the result that it can cause removal of the rail 
from track, if no corrective action is taken [11]. In addition, in the presence of a high level of lubri- 
cation, and the corresponding significant reduction in rail wear, fatigue can cumulate below the sur- 
face of the railhead, such as the point of maximum shear stress in the railhead. This can result in the 
development of subsurface defects such as rail shells. 

The relationship between rail life and rail lubrication is dramatically illustrated in Figure 8, 
which presents rail life data from a 5 degree curve at the Facility for Accelerated Service Testing, 
FAST [16]. In the unlubricated environment, i.e., in a dry condition, the rail in this curve required 
replacement after 80 to 100 MGT of traffic. (Note, traffic was primarily heavy axle load, 100 Ton car 
traffic.) When the rail was "fully lubricated" the wear rate was reduced by a factor of 10, such that 
the projected wear life of the rail, under the same traffic conditions, was 1000 MGT. This was based 
on gage face wear. However, well before this extended wear life was realized, the rail began to expe- 
rience significant fatigue defects. In fact, the 5 percent defect level (i.e. the 5th percentile which cor- 
responds to the point where many railroads replace rails due to excessive fatigue defects) was reached 
after approximately 180 MGT 

Thus, using standard railroad criterion for the replacement of rail in main line tracks, it was 
found that lubrication of the rail (in this case) extended the rail life from 80-100 MGT (dry) to 180 
MGT (due to fatigue), or approximately double the life. While this represents a significant extension 
of life, it can be seen that the development of fatigue defects resulted in the "failure" of the rail well 
before the rail's potential wear life of 1000 MGT. Further, it can be observed that there was a change 
in failure mechanism, from wear in the dry environment, to fatigue in the lubricated environment. 



Paper by A.M. Zarembski 



157 



Ptrewit 
FaMmd 



200 MGT 300 MGT 
Pr*tflet«dQag«-i I I pPrsdIcUd Qag« 

Fae«W«arUf* I I Faca Waar Ufa for 

for Dry S* Cunra ill 1 Liibdealad S* Curva 



C3J 

10.* 

i.a 
•.1 

0.01 
0.001 



Std C Rail 




— r-1 

5- Curva 








YxZ" 


b' 


Sth .- 
Parcanlll* 






X, 




>•>{ i '^ji 




... 


/'X 




U.S. RR 




l/^ 




/I 




?\\ 




\ i_I l_J___ 





10 



100 



1000 



MGT 

Rail Ufa In MQT'a 



10 000 



W*ar 


FallQu* 1 


Dry 


Lub'd 


5lhP«re«nllla 


10th P«rc«nlU« 


80-100 


1000-t- 


-■^^lao 


/■W300 



Figure 8. Fast Rail Failure Distribution 



In order to allow the rail to more closely approach its wear life potential, it is necessary to control 
the rail fatigue build up. Rail profile grinding offers the potential for extending the fatigue life of the 
rail through the reduction of maximum wheel/rail contact stresses (by shifting point of maximum con- 
tact) and the removal of fatigue damaged metal prior to the development of fatigue defects (by artifi- 
cially creating controlled wear in the well lubricated track environment, and allowing for the removal, 
by wear, of fatigue damage rail metal [15, 16]). Noting that profile grinding has the potential of extend- 
ing the fatigue life of the curve rail to that of rail in tangent track [17], this could translate into a rail life 
of 300 to 400 MGT (based on the same FAST data as presented in Figure 8). This would represent a 
further doubling of the life of the rail, with a corresponding major economic benefit to the railroad. This 
benefit has been shown to yield a ROI benefit of rail grinding of the order of 50 to 90% [17]. 

Benefits of Profile Grinding 

The benefits of rail maintenance grinding in general and rail profile grinding in particular are 
primarily associated with improvements in rail performance and corresponding extension of rail life. 
These improvements/life extensions, which are presented here-in, are generally associated with the 
above defined mechanisms; i.e. wear, fatigue, and/or corrugation control, though in some cases there 
is a combination of effects which contribute to an overall extension of rail life. This rail life exten- 
sion is often seen in conjunction with other improvements, such as improvements in rail metallurgy, 
steel cleanliness, lubrication practices, etc. Since these other improvements can be incremental in 
nature, and continuously ongoing (e.g. improved steel cleanliness ), it is often difficult to isolate the 
benefits associated with rail grinding alone. However, as will be noted in this section, rail grinding 
has been shown to generate measurable extensions of rail life and improvements in rail performance. 

Rail grinding also generates secondary benefits associated with reduced dynamic wheel/rail load- 
ing, such as reduced vertical impact loadings. These reduced levels of loadings (and associated improved 
wheel/rail dynamic interaction) produce benefits in terms of extended component lives (wheels as well 
as rails [18]), reduced maintenance cycles (e.g. surfacing cycles) [19], and reduced fuel consumption 
[20]. In addition, by controlling and eliminating such surface defects as corrugations, rail grinding has 
been shown to reduce noise and vibrations, and improve rider comfort (for passenger operations). 



158 



Bulletin 760 — American Railway Engineering Association 



Wear 

As noted above, rail profile grinding was originally developed by the mining railroads of 
Western Australia to control rail wear, particularly on shallow curves, under heavy axle load opera- 
tions [4]. The result was a reported dramatic decrease in wear for curves of less than 3 degrees and a 
corresponding increase in rail life of the order of 70 to 80% for these curves [4]. This increased life 
is presented in Figure 9 which shows projected system rail requirements for one Western Australian 
railroad (without profile grinding) as compared to actual rail life experienced after the introduction 
of profile grinding [4]. 

This improvement in rail wear behavior was also reported in early FAST data, where field tests 
of the effect of profile grinding measured the reduction in both lateral flanging forces and in gage 
face wear, for several different rail head profiles [8]. In these tests, under 100 Ton car traffic, lateral 
force measurements were taken on a four degree test curve, with three different profiles and a con- 
trol (non-profiled) rail head. In all cases, profile grinding significantly reduced the measured lateral 
forces. This reduction in lateral force translated into a measurable reduction in gage face wear as 
illustrated in Figure 10 [8]. While, recent testing at FAST has raised questions regarding proper rail 
profiles (i.e. one point vs. two point contact) as well as the effect of grinding on rail wear [10], those 
results clearly highlight the importance of defining the proper interaction between the wheel and the 
rail and the corresponding wheel and rail profiles. However, it must be noted that this definition of 
proper rail (and wheel ) profile can have a major influence on the level of improvement in (or degra- 
dation oO rail wear behavior [10]. This effect must be considered in light of any other life limiting 
failure mechanisms that must be addressed by profile grinding. 

Fatigue 

Use of profile grinding, and in particular two point profile grinding, to control rail fatigue 
defects has emerged as a key application for profile grinding in North America, particularly under 
heavy axle load operations. Early applications of profile grinding to control fatigue have been 
reported by Canadian National Railways [11], Canadian Pacific Railway [6,7], BC Rail [21], and 



CO 

E 



LU 



1980-1984 CURVE RERAIL (Projected) 
PLAN. 




ACTUAL^ ^* 



./X 



PREDICTED FUTURE 
CYCLE 



81 



82 



END OF 78 79 80 81 82 83 84 

Figure 9. Effect of Profile Grinding on Curve Relay Requirements (5) 



Paper by A.M. Zarembski 



159 



0UO1O 



RAIL PROFILE WEAR TEST 

Qagt Fm« Wmt BaXt of OwMdc IWI Owring IniUal 12 UOT 
ol Trafflo ailar Ortndbio 



P 

S 0007 



OjOOS 




PROfUEl 



PROFILD 



PROFILES 



CONTROL 



QS? QS? QiTP Qi:? 

ouraoc M«of outssc mmc outsbc msm ouraMcMrtf msm 
ea >»INCHOOWN E3, ••INCHOOWN 

Figure 10. Outside Rail Gage Faceweare — 12 MGT 



UJ 

Ik 
ui 
O 
u. 
o 

i 




1962 

(Start of ProfUt 
Grinding Program) 



1983 
D TO 



1964 
♦ TOTAL DEFECTS 



1985 



1986 



Figure 11. Rail Defects on Major North American Railroad (defects normalized 
by actual miles tested) 



160 



Bulletin 760 — American Railway Engineering Association 



others. In the case of one major North American railroad, profile grinding has been reported to have 
effected a systemwide reduction in fatigue related defects, particularly transverse defects (IDs), 
which in this case includes the detail fracture class of defects (which is most affected by profile grind- 
ing). This reduction in defects is illustrated in Figure 1 1 [17]. Similar effects on defect reduction have 
been reported in Australia where fatigue defects (as detected by ultrasonic rail testing) have been 
reduced by more than half, as illustrated in Figure 12 [18]. 



Rail Grinding 

Mt. Newman Mining Co. Pty. Ltd. 



Total Kilomstars • 1.000 



Eff«otiv« Kilom«tar« • 100 




81 



82 



83 84 85 88 87 88 

Financial Years - Ending May 



89 



90 



Total Kllomatara 



Effactlva Kllomatara 



Ultrasonic Defects 

Mt. Newman Mining Co. Pty. Ltd. 



Dafacts par Million Tonna Rallad 



Numbar of Dafacts 




81 



82 83 84 85 B8 87 88 89 

Financial Years - Ending May 



Figure 12. Rail-Grinding and Rail-Defect Experience 



Paper by A.M. Zarembski 



161 



More recent North American experience have confirmed this effect. This includes recent expe- 
rience on Canadian Pacific [23. 23] and Budington Northern [24, 25]. 

The relationship between effective profile grinding and control of fatigue defects is clearly 
illustrated in Figure 13 [25] which shows the number of detail fracture type fatigue defects experi- 
enced by the Budington Northern Railroad during the period 1984 through 1995. Between 1983 and 
1988, Burlington Northern performed profile grinding, generating a basic "two point" contact con- 
figuration on their rail [25]. As can be seen from Figure 13, the number of detail fractures were con- 
trolled and kept at a relatively low rate. In the period of 1988 through 1990, Budington Northern 
changed to a lighter "conformal" (one point) grinding pattern with a resulting surge in detail fracture 
defects (see Figure 13) as well as a "rash" of defect related broken rail derailments (with a cost in 
excess of S6.5 million) [25]. By 1990 through 1991, Burlington Northern switched back to a more 
aggressive "two point" profile grinding and again experienced a reduction in detail fracture type 
fatigue defects, as illustrated in Figure 13 [25]. Thus, BN reports a strong correlation between effec- 
tive profile grinding and the control of rail fatigue defects. 

Corrugations 

Control of rail corrugations was one of the original priorities in North American profile grind- 
ing practices. In fact, one of the earliest controlled tests of rail grinding carefully examined the effect 
of profile grinding on corrugation development and recurrence [6,7]. The results of these tests, which 
are presented in Figure 14, show that corrugation regrowth was significantly slower using profile 
grinding techniques than it had been using conventional (defect elimination ) grinding patterns. This 
reduced growth rate was equivalent to an extension of the grinding cycle from the previous 6 month 
interval to a 8 month interval, an extension of 33%. In addition, it was observed during that more fre- 
quent maintenance grinding, could reduce the overall amount of grinding by eliminating the corru- 
gations while they were relatively shallow (or even before they begin to emerge). 

More recent testing on CP Rail and BC Rail [21,26] confirmed this benefit and further showed 
that preventive maintenance grinding using appropriate grinding profiles can forestall the develop- 
ment of corrugations and provide for a rail "free of corrugations and service defects" [26]. 



Comparison Passmiles Ground vs. 
Detail Fractures 



PASSMILES 



DETAIL FRACTURES 




1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 

YEARS 



500 



Figure 13. Detail Fracture History on Burlington Northern Railroad 



162 



Bulletin 760 — American Railway Engineering Association 



Average Corrugation Regrowth, All Curves 



Comigatlon .02 
Growth 
(inch6s) 



J03 


M 




JOSB 


• 




JOSA 


— 




.024 


" A. 




.022 


- \ 




.02 


- • \ 


n rlOlM PlllMII 


.018 


— \ \ 


A* ^^^ 


X16 


- \ \ 


^^^,,.^-»-*'^"'^ 


.014 


"* \\ 


^^^ «'*• Pipflto PMflfffi 1 


.012 


— \ \ 


rt****^ B* ^.—''"^^ 


.01 


•* \ \ — 


- ^^^^-"^ 


JOOt 


- \ 




.006 


■• \ 




.004 


., O" ■" 




.002 


-m 







1 1 


A 1 1 1 . 1 _ 



Prv^rind 



Months Post Grinding (4.6 MGT Per Month) 
Figure 14. Corrugation Regrowtti After Profile Grinding 



Maximum Wheel /Rail Contact Force as a 
Function of Corrugation Depth and Frequency 



80 
70 

60 

Contact 
Force 50 
{1000'sof lbs) 

40 



10 



Static 
Load 



30 



20 



40 



20 



35 



50 



65 



Corrugation Depth 
(Inches) 




60 80 100 120 140 160 180 200 
Corrugation Frequency V/X (Hz) 



Figure 15 



162 



Paper by A.M. Zarembski 



163 



Vertical Impact Dynamics 

Corrugations and other rail service defects represent a source of dynamic wheel/rail excitation 
and corresponding impact loading which has the potential for generating significant damage not only 
to the rail, but also to the track structure and vehicles [27,28]. This behavior is illustrated in Figure 
15 which shows a more than doubling of the dynamic wheel/rail forces associated with corrugations 
of the order of 0.050 inches. 

In addition to generating dynamic impact forces, corrugations generate noise and vibration, and 
can cause significant discomfort to passengers on rail vehicles [29]. Effective rail grinding can elim- 
inate the growth and recurrence of corrugations (see Corrugations above) with a corresponding elim- 
ination of wheel/rail dynamic forces, noise and vibration. Noise reduction of the order of 7 dB and 
greater have been reported in conjunction with the elimination of rail corrugations by grinding. 

Overall Improvements in Rail Life 

As has been noted previously, while significant increases in rail life have been reported in con- 
junction with rail grinding, actual field results often commingle several maintenance effects, so that 
the effect of rail grinding is masked somewhat by concurrent improvements in rail steels, in lubrica- 
tion practices, and in inspection practices. In spite of this, railroads who have been able to document 
extensive increases in rail life, have attributed these extensions of rail life, in very large part, to rail 
profile grinding. Such is the case on the Burlington Northern Railroad, where improvements in aver- 
age rail life of the order of 50 to 300-1- '/c have been reported (see Figure 16), with rail grinding being 
credited with being a key factor in this extension [12, 24]. 

Canadian National Railways also reported significant increases in rail life associated with a 
combination of increased lubrication, improved rail steels, and rail profile grinding [11]. This over- 
all increase was reported to be of the order of 5(X)-(-%. CN further reported that if proper grinding was 
not performed on an ongoing basis, the rail could lose "95% of its potential service life" [11]. 

Likewise the case in Western Australia where rail profile grinding, in conjunction with improved 
wheel/rail interaction control has been credited with dramatic increases in rail life, with a current 



1500 ■ 

1250 ■ 

1000 • 

750 . 

Kf\f\ 


?ail Life in M 

J 


QT 


I 


■ 




250. 

0- 






p 






' 



Tangent Rail 



Premium rail steels 



1 d«g. 



11981 



2 d«0.* 



11989 



3 d«g.« 



Figure 16. Average Expectations for Rail Life [12] 



164 



Bulletin 760 — American Railway Engineering Association 



Track Miles Treated by Grinding 

By Year 



IbOUMRll 




1982 1963 1984 1985 1986 198 7 1968 1988 1 990 1991 1992 1993 1994 
uawwu*. Figure 17a 



Rail Replacements - CP Rail 

Track miles of new rail laid 



350 




1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 
|»Newraiil 



Figure 17 a & b 



forecast rail life of the order of 2.4 Billion Gross Tons [18]. This represents a multi-fold increase in 
rail life from that experienced even a decade before. 

Other North American Railroads have likewise reported increases in rail life associated with rail 
grinding to include CP Rail, CSX, and Conrail. In the case of CP Rail, the effect of rail grinding on 
rail life is clearly illustrated in Figures 17a and 17b. [23]. These figures show a dramatic reduction 
in miles of new rail laid during the period 1987 through 1995, which corresponds directly with the 
increase in miles of rail grinding for that same period. While other factors such as improved lubrica- 
tion and metallurgy have contributed to the significant reduction in new rail requirements, CP cred- 
its rail grinding as being a key factor in this dramatic extension of rail service life [22, 23]. 

Issues in Profile Grinding 

While profile grinding has resulted in demonstrable benefits to railroads in terms of extended rail life 
and reduced damage to the track structure and rolling stock, profile grinding does change the wheel/rail 
contact environment with the resulting potential for undesirable behavior if not properly addressed. 



Paper by A.M. Zarembski 



165 



The first such potential area of concern is the change in steering forces associated with the 
change in wheel/rail contact points. As noted previously, changing wheel rail contact from one point 
contact (wear control) to two-point contact (fatigue control) can result in a deterioration in truck curv- 
ing performance, and a corresponding increase in the wheel/rail flanging forces (and associated rail 
gage face wear) [13]. This is illustrated in Figure 7 which shows a small increase in truck turning 
moments, associated with moving the contact point on the high rail towards the center of the rail head 
(and thus increasing the truck turning moment arm). This in turn generates an increase in lateral force 
as shown in Figure 5 which increases rail gage face wear [10] and also the potential for gage widen- 
ing and rail overturning on curves [13]. This is of particular concern in environment where both the 
high and low rails are dry (such as immediately behind the rail grinder) or where the high rail is lubri- 
cated and the low rail is dry, which, as seen in Figure 5, generates the highest level of lateral loads. 

This increase in force is of potential concern, and should be addressed in an overall system 
design, where the wheel and rail profiles must be optimized in order to achieve the optimum steer- 
ing forces while maximizing the lives of the key components. Thus, it may be necessary to "trade- 
off" key track and vehicle performance parameters in order to optimize the overall system. Thus, if 
fatigue is the dominant replacement mode for the rails in curves (in a well lubricated environment), 
additional rail wear may not be an important factor, since the rail life is not limited by the wear. 
Similarly, it may be necessary to balance the strength of the track with the level of traffic loadings, 
such as by careful testing of the gage strength or by the use of high strength elastic fastening systems. 

A second area of potential concern resulting from the shifting of the wheel/rail contact points 
is a change in the location of the applied vertical and lateral forces on the rail, and the corresponding 
overturning moment applied to the rail. This overturning moment is generally defined by an L/V ratio 
(ratio of lateral [L] and vertical [V] forces applied to the rail head) as illustrated in Figure 18. When 
the resultant of the applied lateral and vertical forces passes outside the base of the rail, the rail is no 
longer stable on its own base, but has a net overturning force applied to it. In these cases, the rail is 
restrained by its fastening system, and if this fastening system is weak or inadequate, the rail has the 
potential for rolling (rail head moves out laterally) and even potentially overturning. 




Figure 18. Rail Rollover Diagram 



166 



Bulletin 760 — American Railway Engineering Association 



HIGH RAIL 



LOW RAIL 



LATERAL 

FORCE COhfTACT POINT 




WHEEL 



LATERAL 
FORCE 




VERTICAL 
FORCE 



RESULTANT FORCE RESULTANT FORCE 

Figure 19. Rail Rollover Criterion 

By shifting the contact point of the vertical and lateral forces from the gage comer to the cen- 
ter or field side of the rail head, the overturning resistance is decreased. This is illustrated in Figure 
1 9, which shows that if the lateral and vertical forces are applied to the gauge comer, the L/ V ratio 
required to provide this potential instability is approximately 0.6. If these forces moved to the center 
of the rail head, such as would happen if contact was moved from the gage comer to the center of the 
high rail head, the overtuming L/V ratio is reduced to approximately 0.4. If the forces are moved to 
the field side of the rail head, the overturning L/V ratio can be as low as 0.2. Such a case can occur 
if the false flange of the wheel is riding on the field side of the low. 

In the case of the low rail, several of the ground profiles shift the contact point away from the 
field side of the rail head (and false flange contact) towards the center of the rail head. Thus, in these 
cases, profile grinding to alleviate false flange contact will also serve to increase the overtuming sta- 
bility of the low rail (from an L/V < 0.2 to an L/V > 0.4). 

In the case of the high rail, shifting from one point contact on the gage comer of the rail to a 
two point contact, such as illustrated in Figure 7, will reduce the L/V ratio required to overtum the 
high rail from approximately 0.6 (gage comer contact) to approximately 0.5 for the two point con- 
tact normally introduced by current railroad grinding patterns (fatigue control). This reduction is 
diminished, if the separation between the two contact points is reduced, as is the case for several 
"new generation" grinding pattems [22]. This reduction in overtuming resistance must likewise be 
accounted for by the maintenance of adequate track strength for the load environment on that curve. 

Summary 

Rail grinding has evolved over the past several decades from a relatively simplistic method of 
removing defects from the surface of the rail head to a more complex maintenance technique which 
controls not only the development of defects in the rail, but also the interaction and associated con- 



Paper by A.M. Zarembski 167 



tact between the wheel and the rail. This latter technique, which controls the shape or profile of the 
rail head, allows for the extension of the life of the rail by addressing the dominant rail degradation 
mechanism(s), which can vary from location to location. By combining this profile maintenance tech- 
nique with a preventive grinding philosophy, it is possible to achieve significant extensions in the life 
of the rail. 

Rail profile grinding has been used to address three major modes of degradation; wear, fatigue 
and corrugations (surface defects). By selection of an appropriate rail head profile (or profiles since 
separate profiles are used on opposite rails in curves), reductions in the mode of degradation and its 
associated degradation rate can be achieved. This has been documented on railroads in the United 
States, Canada, and Australia, who have reported significant extensions of rail life. While grinding is 
usually only one of .several simultaneous maintenance strategies employed to improve rail life, it has 
been demonstrated to be a key strategy, the absence of which can result in a rapid degradation in the 
rail condition and corresponding shortening of the rail life. Since rail represents the one of the largest 
(if not the single largest) maintenance of way cost area for most main line freight railroads, the con- 
trol of rail degradation and extension of rail life represents a high priority for most maintenance of 
way departments. 

Rail grinding, like many other technologies, must be used carefully and intelligently. When 
properly used it has been shown to contribute to a significant extension of rail life, reduction in rail 
(and other track maintenance) costs, and improvement in the dynamic wheel/rail loading environment. 
If not used properly, it has the potential for causing increased lateral wheel/rail forces, increased rail 
wear, and increased risk of rail overturning. Therefore, rail grinding must be used with a proper 
understanding of its benefits and limitations. When u.sed effectively, rail grinding represents a valu- 
able tool for the control of rail degradation, and for the reduction in overall track maintenance costs. 

References 

1. Butler, R. W. et al.. Criteria for Rail Grinding, Report of Special Committee No. 3, Proceedings of 

the Roadmasters and Maintenance of Way Association, 93rd Annual Conference, September 
1981. 

2. Zarembski, A. M., "The Evolution and Application of Rail Profile Grinding", Bulletin of the 

American Railway Engineering Association, Bulletin 718, December 1988. 

3. Manual for Railway Engineering , American Railway Engineering A.ssociation, 1995 

4. Lamson, S. T., and Longson, B. H., "Development of Rail Profile Grinding at Hammersley Iron", 

Proceedings of the Second International Heavy Haul Railways Conference, Colorado 
Springs, 1982. 

5. Kalou.sek, J., "Thorough Lubrication and Light Grinding Prevents Rail Corrugation", Proceedings 

of the Rail & Wheel Lubrication Seminar, Memphis TN, June 1987. 

6. Lamson, S. T., " Rail Profile Grinding", Canadian Institute of Guided Ground Transport Report 82- 

7, November 1982. 

7. Lamson, S. T, " Rail Profile Grinding; Phase II Test Report", Canadian Institute of Guided Ground 

Transport Report 84-14, February 1985. 

8. Walker. G. W., "Effects of Rail Profile Variation", Report from the Facility for Accelerated Service 

Testing (FAST), FRA/ORD-86-04, March 1986. 

9. "Rail Grinding Equipment: Traditional and Modem", Speno Rail Services, Technical Notes 

Number 7, 1986. 

10. Hannafious. J., "Rail Grinding at FAST", Proceeding of the First Annual AAR Research Review, 

Volume I: FAST/HAL Test Summaries, Pueblo, CO, November 1995. 



168 Bulletin 760 — American Railway Engineering Association 



1 1 . Worth A. W., Homaday, J. R., Jr., and Richards, P. R., "Prolonging Rail Life Through Grinding", 

Proceedings of the Third International Heavy Haul Railways Conference, Vancouver, B. C, 
October 1986. 

12. Glavin, W. , "Rail Grinding The BN Experience", Bulletin of the American Railway Engineering 

As.sociation, Bulletin 722, October 1989. 

13. "Effects of Rail Gage Comer Grinding and Wheel Tread Hollowing on Gage Widening 

Behavior", Association of American Railroads Vehicle Track Systems Newsletter, Railway 
Age, June 1994. 

14. Zarembski, A. M., "Corrugation Behavior in the Freight Railroad Environment", Bulletin of the 

American Railway Engineering Association, Bulletin 712, Volume 88, October 1987. 

15. Kalousek, J., "Wear and Contact Fatigue Model for Railway Rail", National Research Council 

Canada Report 27491, (Transport Canada Report No. TP 8344E), 1986. 

16. Steele, R. K., "Rail Lubrication: The Relationship of Wear and Fatigue", Transportation Research 

Board, Railroad Maintenance Workshop, Amherst, MA, June 1985. 

17. Zarembski, A. M., "The Relationship Between Rail Grinding and Rail Lubrication", Second 

International Symposium on Wheel/Rail Lubrication, Memphis, TN, June 1987. 

18. Mitchell, R. and Dudek, J., "Operating Experience with Heavy Axles at Mt. Newman Mining Co. 

Pty Limited", Heavy Axle Load Workshop, Pueblo CO, October 1990. 

19. "The Economics of Grinding Part II: Tie and Surfacing Savings", Speno Technical Note Number 

3, 1983. 

20. "The Economics of Grinding Part III: Fuel Savings", Speno Technical Note Number 4, 1984. 

21. Kalou.sek, J. , Sroba, P., and Hegelund, C, "Analysis of Rail grinding Tests and Implications for 

Corrective and Preventive Grinding", Fourth International Heavy Haul Railways Conference, 
Brisbane, Australia, September 1989. 

22. Roney, M. D., Kalousek, J., and Sroba, P., "Management of Rail Profiles Through Rail Grinding", 

Proceedings of the International Heavy Haul Railways Conference, Vancouver, B. C, June 
1991. 

23. Wilson, Allan, "Developing and Managing Rail Maintenance Programs", ARM Rail Maintenance 

Seminar, Chicago, Illinois, April 1996. 

24. Besch, G. O., "BN Grinds Out Longer Rail Life", Progressive Railroading, April 1990. 

25. Tornga, G., "Conformal and Non-conformal Grinding Experiences", ARM Rail Maintenance 

Seminar, Chicago, Illinois, April 1996. 

26. Kalousek, J., "Maintaining Corrugation Free Rail in Heavy Haul Track", Railway Gazette 

International, 1989. 

27. "The Consequences of Corrugations and Surface Defects", Speno Technical Note Number I, 

1982. 

28. Zarembski, A. M., "The Impact of Rail Surface Defects", Railway Track and Structures, 

November 1984. 

29. Grassie, S. L. and Kalousek, J., "Rail Corrugation: Characteri.stics, Causes and Treatments", 

Proceedings of the Institute of Mechanical Engineers, Volume 207, 1993. 



SERVICE LEVEL LIVE LOAD STRESS RANGES 

ON HANGERS AND FLOOR BEAMS OF 

STEEL RAILWAY BRIDGES 

By: Robert A. P. Sweeney*, Hoat Le** and George Oommen*** 

Introduction 

This is the third report on the testing of 70 Railway bridge spans of various length and design. 
The first of the series "A Summary of Seven years of Railway Bridge Testing at Canadian National 
Railway"' provided a summary of field measured static (crawl speed) point stresses compared to the- 
oretical calculated stresses, as used in normal bridge rating practice and several changes to the cur- 
rent American Railway Engineering Association (AREA) manual were also suggested. The second 
of the series, "Impact of Site Measurements on Evaluation of Steel Railway Bridges"" summarized 
and reviewed the test results of some fatigue sensitive components such as bottom flanges of plate 
girders & stringers and bottom chords of through trusses and deck trusses. Reasonable values for the 
a factor (ratio of measured static stresses to theoretical static stresses) and impact factor (as a per- 
centage of the one specified in chapter 15 of the AREA manual) were proposed. 

Hangers in through truss spans and floor beams in all truss spans are the subjects of this paper. 
They are among the most critical members in terms of fatigue sensitivity. This is due to the fact that 
they undergo a very high number of stress cycles compared to members with longer loaded length 
(70' or more), and are loaded in ways that are different from the usual calculations models. 

Furthermore, in terms of hangers, shear lag has been shown to be a major failure instigator in 
riveted hanger connections. This paper does not include the end effects where the influence of shear 
lag is critical. 

This paper explores the extent to which the method used in computing the theoretical stresses 
can influence the value of the a factor in hangers. The paper also propo.ses reasonable impact factors 
that could be used for fatigue life evaluation, depending on which method is used to obtain the the- 
oretical stresses. The paper highlights the imperfections in conventional analytical techniques for 
fatigue life evaluation of hangers and floor beams in truss spans. 

Description of the Tested Spans 

This paper contains test results on 15 hangers and 14 floor beams on various through and deck 
truss spans from 1 15' to 348' long (with one exception of a 640' long span). 

The tested spans had built-up and riveted construction, typical of turn of the century designs. 
Some of the bottom chords and tension members were eye bars with or without pin plates. 

All the trusses tested had open decks, except for one with rail chairs. 

The hangers had loaded lengths between 46' and 70'. The floor beams had loaded lengths 
between 38' and 63'. 

Field Test 

Test data used in this report was based on controlled field tests using a work train. The work 
train generally consisted of one or two locomotives followed by six or more pre-weighed loaded cars. 



* Assi. Chief Engineer Structures, CN Railways 
•'Bridge Testing Engineer, CN Railways 
**• System Engineer Bridge Assessment. CN Railways 



169 



170 Bulletin 760 — American Railway Engineering Association 



Tests were conducted at various speeds ranging from crawl speed, up to 60 mph (depending on 
maximum allowable track speed). The presented data is based on maximum recorded point stresses 
(not averages across a member), and maximum measured impact at the maximum recorded stresses 
(regardless of speed). 

Since the intent of these tests was to establish the dynamic behavior under normal operating 
conditions, no attempts were made to resurface the track or eliminate other imperfections, such as flat 
wheels, out of round wheels, low approaches and bad rail joints). No modifications were done on the 
bearings. All the recordings were carried out when the atmospheric temperature was between 0° C 
and 35° C (32°F and 95°F). 

Discussion on Allowable Stresses 

The present AREA manual specifies two di.stinct ways of computing the theoretical stresses in 
hangers. 

In the Desifin section (1.4) of the manual, the specified basic allowable .stresses of 0.55Fy for 
tension members including bending is further reduced to 71% for floor beam carrying riveted hang- 
ers. This is equivalent to multiplying the theoretical stresses by 1.40, while keeping the allowable 
stresses at 0.55Fy. 

The intent is to compensate for design calculations using simple models, which provide axial 
stresses plus the out of plane bending component due to the floor beam framing into the hangers, 
while ignoring the usually small in plane bending and the larger effect due to shear lag in riveted 
hanger connections. 

In the Rating .section (7.3.4) of the manual, the specified basic allowable stres.ses of O.SOFy for 
tension members (mild steel) is further reduced to 75% for floor beam carrying hangers (riveted con- 
nection) including bending. This is equivalent to multiplying the theoretical stresses by 1.33, while 
keeping the allowable stresses at O.SOFy. 

As in the eariier case, the intent is to compensate for design calculations using simple models, 
which provide axial stresses plus the out of plane bending component due to the floor beam framing 
into the hangers, while ignoring the usually small in plane bending and the larger effect due to shear 
lag in riveted hanger connections. 

The results reported in this paper are free of shear lag effects since the measuring points were 
sufficiently distant from the end connections. 

The reader should refrain from making a direct comparison between the "Design" and 
"Maximum Rating" values, because the effective allowable stresses are very different in each case 
(0.71 X 0.55Fy = 0.39Fy for Design and 0.75 x O.SOFy = 0.6Fy for Rating). 

Discussion on Test Results 

The test results (static & dynamic) are compared against various theoretical values, and pre- 
.sented in a series of 12 figures as follows: 

• Fig. 1-2: Hangers, theoretical stresses based on simple truss models (axial stress only). 

• Fig. 3-8: Hangers, theoretical .stresses based on simple truss models, augmented by a con- 
stants ( 1 .40 & 1.33) and by simple hand style out of plane bending calculations (hanger / floor 
beam frame). 

• Fig. 9-12: Floor beams, theoretical stres.ses based on simply supported beam model. 

Note: The dotted lines in the graphs only show the relative position of the dots with respect to 
the 45° line (above or below it), if the theoretical values were to be multiplied by the given constant. 
Graphs showing the exact absolute position of the dots are not included in this paper. 



Paper by RAP. Sweeney. H. Le and G. Oommcn 



171 



SUMM^UiY OF MEASURED M THBORETKAL STRESSES (static only) 
THROUGH TRUSS HANGISS 





46<L«^ 
































1.0 






12000- 


Alph«-0.»5 
Q Riilchurt 





















^^'^ 


10000- 


-.. j^. 











^^.X"^"'^ 











O 


^^.--^ 




8000' 








n 














a 














^-^^ 












^^ 




1 1 








0- 


1 Theory includes axial only | 
1 



MOO 



4000 6000 8000 

THBOREnCAL STRESSES (PSD 



10000 



12000 



Figure 1 



SUMMARY OF RCASURS) >• THEORETICAL STRESSES fincludins invact) 
THROUGH TRUSS HANGBiS 



46<L<70 



12000 

2 10000 

BC 8000 

a 

g 6000 

3 
V> 

2 4000 



2000-- 



a 
o 


-Rj«io-1.0 
Rj«io-0.«3 

















Rjilcbiin 
Opadadc 
























^y^ fi 

















. 8 














8 












^ 















iThfli 


sry includea axial only [j 
















|Tbeo«v incl. 100S o(th»of. impact | 




1 


1 1 1 



2000 



6000 8000 10000 

THEORflKAL STRESSES (PSQ 



12000 



14000 



16000 



Figure 2 



172 



Bulletin 760 — American Railway Engineering Association 



SUMMARY OF MEASURED v THEORETICAL STRESSES (static only) 
THROUGH TRUSS HANGERS 



14000 
12000 
10000 
8000 



46<L-^?0 



S 6000-- 



2000 



2000 







... 
1.0 














Alpha - 




O Rtildiain 
o Opcnd«ck 










^ — ^ 










--■'^'O 


o 




















W" 










^^>^^^-"' 


CD 
00 










^^^ 


o 










^-^ 


















iTheory -1.40 x axial comp. { 

1 



4000 6000 8000 

THEORETICAL STRESSES (PSI) 



10000 



12000 



Figure 3 



SUMMARY OF MEASURED v TOBOREnCAL STRESSES (patie only) 
THROUm TRUSS HANGERS 





46<L«^ 
















■ 1 
















1.0 




12000- 

2 10000- 




a RAilcfaiin 
O Opadedc 










.^'^'(i 








^^^ 




i 
















.--^ 


r"'" 




2 «x»- 






^^<^ 


. c6 

50 






s 




^^<^" 


o 










r^ 












0- 






|Theo<y =1.33 x axial comp. | 



2000 



4000 6000 8000 

THBORCnCAL STRESSES (PSI) 



10000 



12000 



Figure 4 



Paper by R.A.P. Sweeney, H. Le and G. Oomitien 



173 



14000 
12000 
ICOOO 
8000 

eooo 

4000 
2000 



46<L<e7D 



SWIMARY OF MEASURJDw THEORETICAL STRESSES rincludinr inmacf\ 
THROUGH TRUSS HANGERS 







1 


— 
















-Ruio-I.O 




a 




Op«dedc 










^ 


o 










y^ C 


o 






















.-8 














.y^^ 


.-■■' <? 












^^ 














{Theory 


" 1.33 X axial comp. \- 


















[Theory incl. 35% oftheor. impact 1 

'a 1 V 



2000 



4000 



6000 8000 10000 

THEDRETICAL STRESSES (PSI) 



12000 



14000 



16000 



Figure 5 



SUMMARY OF MEASURED y THBOREnCAL STRESSES (static only) 
THROUGH TRUSS HANGERS 





46<L-OD 
















1 
















1.0 
0.85 
ft 




13000- 
10000- 




□ Rjildiii 








o 
o 


^.^--^ 








< 


.^r^^^ 




















^<^ 










a 














^.^ 


.■■■■'"6 










.^ 














[Theory irKludee out of plar>e t»ndng { 


0- 







1 1 



2000 



4000 6000 8000 

THEORETICAL STRESSES (PSI) 



10000 



120OO 



Figure 6 



174 



Bulletin 760 — American Railway Engineering Association 



SLMVUIlYOr MEASURg)^ THBORgTICALSTRBSeS nneludint mw>aet\ 
THROUGH TRUSS HANGISS 





46<L<70 




















1 












y-^ 


»..■•■■" 






io-1.0 




VXCD' 




Q Riilduin 
O Op«nd«dc 










^ 




b' 










^y^ 


... -6 


i 




















^ 


o 


\ 














< ' \ - 












^ 




a 










2000- 


jThwyy includM od of fiana band ng 1 













1 -I ■ I 1 , ■■ ; 

|Theo(y incl. 100% otthaor. impact |j 


0- 




- 


1 4- 1 1 



2000 



4000 



6000 8000 lOOOO 

THEDRZnCAL STRESSES (PSQ 



12000 



16000 



Figure 7 



SUMMARY OF MEASURED w THBORETKALSIIlESSeS finctudinr iimaM 
THROUGH muss HANGERS 



14000 
12000 

j2 10000 



B( 8000 

2 6000 

3 

^ 4000 



2000 





46<L<I0 





1 


— 










^ 






Rj 


itio-1.0 










' Ratio -0.«3 

a R<adiiin 
- O Opadeck 










y^ 


o 












^ 


o ...■■•■■' 
•■"' o 






















^^ 


i-^ 














SX*^**' 


.6 O 














^^ 

















iThai 


•xi includes out of plana bandog |_| 




^ 








|Th« 


wy incl. BS"* 



fcofthaor. ir 
1 


npact 1 



200O 



4000 



6000 80OO 10000 

THEORETICAL STRESSES (PSD 



12000 



14000 



16000 



Figure 8 



Paper by R.A.P. Sweeney, H. Le and G. Oommen 



175 



SlftMARYOFMEASUUDw THE0RETICALSTRE5SES (Static only) 
3S<1<^ FLOOR BEAM BOTTOM FIANCES 



auuu * 




pha- 1.0 














^ 






£ SOD- 


o 


p«dedc 










/^ 




o 


[9 ffioo' 










/^ 


...■■■" o 


o 














^ 


oo 




o 






a 








^ 


....... 0- k 


> 












^^ 


b 
































0- 


^ 




-^^^ 















1000 2000 3000 4000 SOOO 6000 ?0OO 8000 9000 

THZDRCnCAL STRESSES (PSI) 



Figure 9 



Sl»»fARYOFMEASlRID>» THBORCTiCALSllUSSES (including imp»a) 
38<L<«3 FLOOR BEAM BOTTOM FIANCES 



8000- 

g6000. 


. 1 1 

Rjtio-IO 

Rjtio-O.li 

o Op«od«dc 








f -""^ 




3 














<? 




> o 

o 











^ 






o 
























Q- 






|Th«(xy ind. 100% ofthcor. impact , | 
1 1 



2000 



4000 6000 8000 10000 

THEORETICAL STRESS ES (PSI) 



12000 14000 



Figure 10 



176 



Bulletin 760 — American Railway Engineering Association 



SUMMARY OF ME^lJRED>s THEORETICAL STRESSES fimeluMng ijt^)met) 
3S<L<«3 FLOOR BEAM BOTTOM FIANCES 




Tbaoty ind. 0% oftheor. impact 
I I 



2000 



4000 6000 8000 10000 

THEDRCnCAL STRESSES (PSO 



12000 



14000 



Figure 11 



SUMMARY OF MEASURED w THBOREnCAL STRESSES fincluJing impod) 
38<L<63 FLOOR BEAM BOTTOM RANGES 



8000 



6000 



4000 



2000 



Ratio -1.0 

Ratio -O.tS 

Opandeck 




Thaoty ind. 50% orthcor. impact 



=F 



200O 



4000 6000 8000 10000 

THBOREIKAL STRESSES (PSD 



12000 



14000 



Figure 12 



Paper by R.A.P. Sweeney, H. Le and G. Oommen 177 



Hangers, theoretical stresses based on simple truss models (axial stress only) 

Fig. 1 shows that computed sicitic stresses based on simple truss models (axicil only) do not ade- 
quately represent actual static stresses in the hangers (in other words, the a factors are mostly greater 
than I). However, fig. 2 proves that when full impact (computed according to the AREA manual) is 
applied to those theoretical static stresses, they do compare quite favorably with actual dynamic 
stresses, except for one case. 

Hangers, theoretical stresses based on simple truss models, augmented by constant multipliers 

Fig. 3 shows the correlation with static loading results, assuming the theoretical axial stresses 
are multiplied by 1.40. Fig. 4 shows the case where the theoretical axial stresses are multiplied by 
1.33. The later represent a better correlation to the static test results. Fig. 5 shows that it would be 
reasonable to apply 35% of full impact, after multiplying the axial stresses by 1.33, to coine up with 
a conservative representation of actual dynamic stresses. 

Note that it is just a coincidence that the 1.40 and 1.33 are the same as the manual's assumed 
values for shear lag. It is the writers' opinion that a simple calculation for the bending component is 
better than assuming a constant. 

Hangers, theoretical stresses based on simple truss models, augmented by simple out of plane 
bending calculations 

Fig. 6 shows that when theoretical static stresses based on simple truss models are auf^mented 
by simple out of plane bendin}', they still do not conservatively represent actual static stresses in the 
hangers (some a factors greater than 1 ). On the other hand, fig. 7 reveals that it is generally too con- 
servative to apply full iinpact (computed according to the AREA manual) to those augmented theo- 
retical static stresses when projecting actual dynamic stresses. Fig. 8 shows that it would be reason- 
able to apply 65% of full impact to tho.se augmented theoretical static stresses, and still obtain a 
conservative representation of actual dynamic stresses, except for one case. 

The writers advocate the use of more sophisticated modeling than that used above, combined 
with field testing where more precise values are required. 

Floor Beams, theoretical stresses based on simply supported beam models 

Fig. 9 shows that computed static stresses based on simply supported beam models are a good, 
but not perfect representation of actual static stresses in floor beam bottom flanges (all a factors are 
significantly less than 0.85, except for 1 ca.se). Fig. 10 shows that applying full impact (computed 
according to the AREA manual) to the above computed stresses when forecasting actual dynamic 
stresses is too conservative. Fig. 1 1 illustrates that it is reasonable not to apply any theoretical impact 
to the static stresses, and still obtain a conservative representation of actual dynamic stresses. 

Fig. 12 illustrates that an assumption of an alpha of 0.85 together with 50% of the AREA cal- 
culated impact is also a con.servative representation of actual dynamic stresses for fatigue life pre- 
diction. Nevertheless, given the results shown in Fig. 9, it is not a logical .solution. 

Note on Fatigue Life Calculations 

Since the fatigue life of a member is inversely proportional to the cube of the stress ranges it 
must endure, even a slight rne/estimation of stress range amplitudes can lead to gross (//if/f /estima- 
tion of remaining fatigue life, resulting in premature major capital expenditures. Hence extra effort 
and proper judgment are es.sential in evaluating the remaining life of any structure or component. 

Note on Failure Prediction in Hangers 

Theoretical space frame analysis usually points to the area on the inside face of the hanger, adja- 
cent to the floor beam connection, as a critical location for fatigue crack initiation. However, histor- 



178 Bulletin 760 — American Railway Engineering Association 



ical records show that a large number of failures occurred al the net section through the lowest rivet 
line connectin}> the lumber to the top chord gusset plates'. 

A number of hypothesis explaining this type of failure have been advanced, none of them con- 
clusive. The most likely explanation revolves around shear lag, but much more investigation is required 
to establish the real cause and magnitude of the stresses leading to those failures. Meanwhile that area 
should be subjected to a lot more scrutiny during inspection, despite its relative inaccessibility. 

Conclusion and Recommendations 

For truss hangers and floor beam bottom flanges, the current AREA a factor of 0.85 is non- 
conservative as shown by the correlation with the static stress ranges. Furthermore, the data shows 
that current simple methods of calculation do not represent the actual structural behavior, and in a 
few cases are non-conservative. 

Where the economic expenditure for replacement, retrofit or strengthening is large, a proper 
three dimensional finite element analysis calibrated to field test results is the optimal solution. 
Considerable sums can be saved on the average, but not in every case. 

The paper shows a number of possible solutions for stress ranges to be used for safe fatigue life 
prediction including the dynamic effects. Not one is completely satisfactory. 

For truss hangers: 

• Using a simple truss model with only axial stress ranges and the full AREA calculated impact 
is not always conservative. 

• Using a 1/3 increase in stress ranges to approximate bending effects together with 35% of the 
AREA calculated impact was just conservative on the 15 hangers reported here. 

• Using a simple out of plane bending model together with a simple truss model and 65% 
of the calculated AREA impact covers 14 of the 15 hangers tested and is very close on the 
last one. 

• More investigation is required to deal more effectively with the shear lag phenomena in 
truss hangers. 

For Floor beams of trusses: 

• A simple calculation with no impact is just conservative. 

• Using an a factor of 0.85 and 50% of the AREA calculated impact is also conservative. 
Needless to say 65% of the AREA calculated impact is even more so. Nevertheless, the use 
of an a factor as shown by the static tests is not an appropriate model ( 1 point over the line). 

Acknowledgments 

The authors wishes to thank present and past members of Bridge Rating Group and Testing 
Group at CN, especially, J. Cavaco, R. Scorteanu and R. McNaughton for their contributions, and 
special thanks to R.W. Richardson, Chief Engineer for his continued support. 

Disclaimer 

The data provided in this report is greatly summarized, and of a highly technical nature. 
Extreme caution must be exerci.sed in its use. 

References 

1. Sweeney, Robert A. P., Oommen, George and Le, Hoat, "A Summary of Seven Years of Railway 
Bridge Testing at Canadian National Railway", Bulletin of the AREA #756, May 1996, pp 
333-347, Washington, D.C. 



Proposed Manual Changes 



179 



2. Sweeney, Robert A. P., Oommen, George and Le. Hoat, "Impact of Site Measurements on 

Evaluation of Steel Railway Bridges", lABSE Workshop Lausanne: Evaluation of Existing 
Steel and Composite Bridges, lABSE. Zurich, Switzerland. 

3. American Railway Engineering Association Manual for Railway Engineering, Chapter 15, Steel 

Bridges, 1996, AREA. Washington, D.C. 

4. Report on A.ssignment 4, "Stress Distribution in Bridge Frames-Floor beam Hangers, Iron and 

Steel Structures" Proceedings American Railway Engineering A.s.sociation, Vol. 51, 1950, pp. 
470-501, AREA, Washington. D.C. 




HANSON 
WILSON 



We engineer solutions for railroads. 

Hanson-Wilson Inc. provides engineering and 
architectural services to meet your design and 
construction challenges. We offer the expertise 
of more than 500 professional, technical and 
support personnel. 



3101 Broadway, Suite 900 Kansas City, MO USA 
Tel: (81 6) 561 -9054 Fax; (81 6) 561 -0654 




Engineering Design 
Track and Bridge Inspection & Design 
Construction Engineering 
1 Fueling & Maintenance Facilities 
Intermodal Facilities 



ONTHPIHIWl 

nuNGREASEntoDUcrnnTY 

The Plasser Continuous Action Tamper 09-16 C. A.T. clears the 
way for better working comfort at lower costs. It's innovative design 
am produce a 30% increase in production while reducing stresses 
on both operator and machine. With the machine s working units 
positioned on a separate subframe and indexed from tie to tie 
during the work cycle, a new 
level of working comfort is now 
available for track maintenance 
crews. Compared to conventional 
tamping machines, only 20% of 
the total mass of the Plasser 09-16 
C. A.T. is accelerated and braked 
during the work cycle. The main 
frame of the machine moves 
forward smoothly and continu- 
ously The machine is subject to much less strain and wear. Track 
time can be much more eflFectively utilized. 

For improved cost savings and comfort, rely on the Continuous 
Action Tamper 09-16 C. A.T, exclusively from Plasser 

Call or write today for specifications and frill details. 




Lifting, lining and tamping units are mounted on a 
separately moving satellite frame. 




PLASSER AMERICAN CORPORATION 

2001 Myers Road, P.O. Box 5464 

ChesapeakeM 23324-0464, U.S.A. 

(804)543-3526 



PLASSER CANADA INC. 

2705 Marcel Street 

Montreal, Quebec H4R1A6, Canada 

(514)336-3274 




-^SSai 




DEVELOPMENT OF A RECYCLED PLASTIC/ 
COMPOSITE CROSSTIE 

By: Barry Gillespie*, Mark Lutz**, Dr. Tom Nosker*** and Don Plotkin**** 

Abstract 

Since 1994. Rutgers University. Earth Care Products, Conraii. Norfolic Southern and the US Army 
Corps of Engineers Construction Engineering Research Laboratories have wori^ed jointly to develop 
a recycled plastic/composite crosstie. The group has developed a specification, manufactured cross- 
ties that meet or exceed mechanical property targets, and installed plastic/composite ties in mainline 
service on a Class I railroad and in the Facility for Accelerated Service Testing (FAST) track at the 
AAR Transportation Technology Center in Pueblo. The performance of the ties warrants placing a 
greater number of them in service. Areas of further investigation will include optimizing the manu- 
facturing process and developing a model to predict the long-term mechanical properties of recycled 
plastic/composite crossties. 

Introduction 

In 1994, Rutgers University's Plastics and Composites Group, formerly the Center for Plastics 
Recycling Research, was granted funding by the New Jersey Commission on Science and Tech- 
nology to develop and test composite railroad ties made from recycled plastic. Since the inception of 
this project, the major participants have been Rutgers University, Earth Care Products, Conraii, 
Norfolk Southern and the US Army Corps of Engineers. 

The mixture of plastic that remains after the easily identifiable milk and soda bottles have been 
removed during the sorting process at the recycling center is a suitable feedstock for crosstie fabri- 
cation. Much of the 7.2 billion pounds of this material that is generated annually in the United States 
is landfilled (Ref. 1 ). LandfiUing is recognized as the least favorable disposal strategy by government 
regulatory agencies and the possibility exists that in the future there may be economic incentives for 
products constructed from recycled plastic. 

Alternatives for the wooden crosstie are being considered because of the desire to increase the 
tie service life and also due to changing economic and regulatory conditions which may impact the 
railroad industry's ability to use creosoted wood in the future. Particularly in moist, humid sur- 
roundings, tie life is limited by the activity of biological organisms. Plastic crossties are not subject 
to the attack of these organisms. Economically, as the hardwood market continues to become more 
global in nature, wood will be diverted to uses more profitable than crosstie manufacture. The reduced 
supply of hardwood for crossties will tend to increase the cost of the wood ties. Also, it is apparent 
that over the past twenty-five years environmental regulations have become more stringent. It is in 
the railroad industry's interest to research crosstie materials that do not require a preservative. The 
research into plastic/composite crossties is timely because of the lag time required to prove engi- 
neering design criteria and economic benefits in revenue service. 

It is projected that the full production cost of a plastic/composite tie will be on the order of $70. 
Because this cost is greater than that of a wood crosstie, the plastic tie is not viewed as a one-for-one 
replacement for wood at this time. It is believed that the premium h>enefit of plastic crossties will be 
substantiated in specific applications such as tunnels, curves, switches, bridges, under grade cross- 
ings, paved areas and other locations where tie service life is reduced due to poor drainage. 



* Research Chemist. Norfolk Soulhorn. Roanoke. VA 

* Senior Chemist. Conraii. Alloona, PA 

'Assistant Research Professor. Rutgers University. New Brunswick. NJ 

•Civil and Railway Engineer. U.S. Army Construction Engineering Research laboratories. Champaign. IL 



181 



182 Bulletin 760 — American Railway Engineering Association 



Research Objectives 

The objectives and approach of the cooperative research were established as follows: 

1 . Develop a specification that defines mechanical property targets and performance require- 
ments based on the dimensions and physical and mechanical properties of a conventional 
wood crosstie. 

2. Fabricate prototype plastic/composite ties that conform to the details of the specification in 
item 1. 

3. Conduct laboratory tests to confirm the properties of the plastic/composite tie. 

4. Install the plastic/composite ties in track and monitor their performance. 

Results 
Specification 

The following is the plastic/composite crosstie specification developed by the research group: 

Dimensions/Appearance 

1. Cross Section 7x9 inches +/- 0.125 inches 

2. Length of 102 inches +/- 1.00 inches 

3. Surface flatness to within 0.0625 inches peak-to-peak in the area of the tie plate (11 inches 
from the end and for a distance of 20 inches). 

Performance Requirements 

1. Less than 5% water absorption. 

2. Exposure to diesel fuel, mineral oil and grease affects the mechanical properties less than 10%. 

3. Electrically non-conductive. 

4. Surface degradation due to UV light exposure will not exceed 0.003 inches per year. 

5. Surface deterioration due to abrasion will not exceed that of a wooden tie. 

Mechanical Properties 

Under the following conditions, the track gage will not increase more than 0.125 inches: 

1 . Lateral load of 24,000 pounds. 

2. Static vertical load of 39,000 pounds and dynamic vertical load of 140,000 pounds. 
A Modulus of Elasticity greater than 170,000 psi is necessary to meet this requirement. 

Installation 

1 . Installation can be accomplished using standard materials handling equipment. 

2. Ties are compatible with premium fastening systems. 

Physical Properties of Manufactured Plastic/Composite Crossties 

When considering alternative crosstie materials, it is logical to use wood as a basis for com- 
parison because of its widespread use and proven performance. However, wood was originally 
selected because of its natural abundance and easy machining and its performance has been demon- 
strated empirically over time. It is important to note that optimum material properties for a crosstie 
have not been established and deviations from wood's properties do not necessarily imply that a 
material will not perform adequately. 



Paper by B. Gillespie, M. Lutz, Dr. T. Nosker and D. Plotkin 



183 



Earth Care Products has manufactured four generations of plastic/composite ties. The third gen- 
eration's combination of physical properties exceeded the established targets. 

Temperature Data 

The Coefficient of Thermal Expansion, measured experimentally over the 3°F-123°F tempera- 
ture range, is 0.000027 in/in/°F. 

From June 27 to July 1, 1996, the temperature was measured at fifteen minute intervals on the 
top, middle and bottom of one of the two ties installed at the Association of American Railroads' 
Transportation Technology Center. With one exception, the temperature at each location on the tie 
remained within 10°F of ambient. The exception occurred on the top of the tie during the hottest part 
of the day. When the ambient temperature was 100°F, the top of the tie reached a temperature of 
140°F(Ref. 2). 

Spike Withdrawal Force 

Screw spike holding power in plastic/composite ties is comparable to that of wood crossties. 
Preliminary testing at both the Norfolk Southern and the Corps of Engineers laboratories indicates 
that cut spike holding power is less in plastic/composite ties than in wood crossties. Voids within the 
ties reduce spike holding power. The number of voids can be controlled during the manufacturing 
process. 

Permanent Deformation Under Load 

A one foot section of rail was attached to a standard length plastic tie (S'/z feet) using a Pandrol 
Fastener System. The tie plate was centered 15'/4 inches from the end of the tie. The head of the rail 
was loaded for one hour at each 5,000 pound increment beginning at 20,000 pounds and ending at 
30,000 pounds. The rail was loaded using a 200,000 pound hydraulic Universal Test Machine. At the 
upper end of the test range, the permanent deformation was less than 0.0015 inches (Figure 1). 

Rail Seat Compression Test 

A one foot section of rail was fastened to a 2'/: foot length of plastic tie using a Pandrol Fastener 
System. The tie plate was centered on the plastic section. The Universal Test Machine loaded the 
head of the rail in 10,000 pound increments beginning at 70,000 pounds. The tie failed at 120,000 
pounds (Figure 2). Given the tie plate area of 128 square inches, the failure is equal to approximately 
9(X) pounds per square inch. 











JUWU 


/ 


/ 


29000 




/ 25,000 Ibi. 


/ 

7 






/ / 20.000 lb.. 


V 








15000 




y^ // 




10000 


^ 


^^ // 








'^ / 




coNua 


5000 


^^^^^^ 


^^^^^,,„0.'''^^^l^^^^^ Jinuary 1»6 


- f • 


SCRVKCS 

L»»CWATOWH 



0.0005 0.001 0.0015 0,002 0025 0,003 0035 004 0.0O45 

Oafonnatton In Inchaa 



Figure 1. Recycled Polymer Composite Railroad Crossties 
Plastic Deformation at Extended Holding Time (1 Hour) 



184 



Bulletin 760 — American Railway Engineering Association 




0.120 0140 



Figure 2. Plastic Tie Load to Failure Test 
Recycled Polymer Railroad Crossties 



Bending Strength: Estimated Modulus and Ultimate Strength 

A slow bend test was performed in accordance with Figure 3. 

The Ultimate Strength in pounds per square inch and Young's Modulus in pounds per square inch 
were estimated from the results of the Bending Strength tests using the following equations (Ref. 3): 

Ultimate Strength 

m = S— 
C 

m = 27"- 

2 

F = Breaking Load, lbs = 23,000 lbs. 



bh^ 



9(7)' 



= 73.5 in' 



C 6 6 

S = Ultimate Strength, lb/in ^ 

Young's Modulus 

MEI 
f = deflection, inches = 1 .29 inches 
P = Load, lbs = 23,000 lbs. 
M = Contant = 48 for this geometry 

E = Young's Modulus, lbs/in 



I = Moment of Inertia = 



L = Length = 60 inches 



bh-' 

12 



9af_ 

12 



257.25 in' 



Figure 4 shows the Ultimate Strength and Young's Modulus for generation 2, 3 and 4 ties. The 
generation 3 "pin" designation was used to identify a modification in the fabrication process. 



Paper by B. Gillespie. M. Lutz, Dr. T. Nosker and D. Plotkin 



185 




Figure 3. NS Bending Strength Test 



Bend Strength Comparison 

Composite Crossties 




Figure 4 



Plastic/Composite Tie Installation 

The plastic/composite ties are currently installed in three locations. 

In October of 1995, ten plastic ties were installed at Rose Yard in Altoona, Pa. The ties are non- 
consecutive and are intermingled with twenty wood crossties. Periodically, each tie is examined visu- 
ally and also has six parameters evaluated (Figure 5). The values of the measured parameters have 
not changed since installation. To date, 13 million gross tons (MGT) have passed over this site at 
speeds not exceeding 15 miles per hour. The ties show no signs of weathering even though the win- 
ter of 1995-1996 was particularly severe in the northeast. The ties were covered by snow almost con- 
tinually from mid-November to the beginning of March. 

In April of 1996, two consecutive ties were placed in a 5 degree curve in the FAST track at the 
AAR Transportation Technology Center in Pueblo, Co. In 1996, the ties saw 60 MGT of traffic at a 
speed of 40 miles per hour. In 1997, another 100 MGT are planned. The ties are being monitored 
visually and there has been no noticeable change. 

In October of 1996, six ties were installed in mainline service at Milepost 220.41 on Conrail's 
Pittsburgh Line where track speed is 35 miles per hour. This section of track is in a 6 degree curve 
and typically the traffic level is 30 MGT annually. The ties lie in two .sets of three con.secutive ties 
with six wood ties between them. The track inspector monitors the appearance of the ties and the 
track gage. To date, there has not been any change. 



186 



Bulletin 760 — American Railway Engineering Association 




Figure 5 



Discussion/Future Direction 

The time dependent mechanical properties of plastics and composites will have an effect on 
gage holding in curves due to creep and also when spikes will become loose from stress relaxation. 
Research continues (Ref. 3) to develop a predictive model for polymers in tension that could be 
applied to composite materials in compression. This model would help establish a quick, inexpensive 
technique to predict the long-term mechanical properties of plastic/composite crossties and therefore 
be able to estimate the service life of the tie. The current model is very accurate in the linear portion 
of the stress-strain curve, 3% strain. 

Efforts continue to optimize the manufacturing process to reduce the cost of the ties. Costs asso- 
ciated with the feedstock can be minimized by centralized sourcing and tracking of materials, contracts 
to supply granulated materials, multi-use feedstock packaging into the production line and better uti- 
lization of multi-modal transportation. Operating costs can be reduced by using automated materials 
handling from the loading dock to the extruder. An economy of scale can be derived from utilizing 
larger extrusion and molding systems with more automation (Ref 4). 

The Army railroad system resembles a family of light density short line railroads. About 70 
installations have railroad operations, with a system total of about 2500 miles of track and approxi- 
mately 12,000 turnouts. The US Army Construction Engineering Research Laboratories (USACERL) 
is looking at plastic/composite tie development from this light density line perspective. This approach 
has suggested looking toward two types of ties. One type would be referred to as a "maintenance" 
tie, and the other as a "replacement" tie. 

The maintenance tie is intended to be fully compatible with standard timber ties. Thus, plastic/ 
composite ties could be installed in track next to standard wood ties with no apparent difference in track 
response or behavior under load. And as with standard timber ties, a plastic/composite tie of 6" x 8" 
cross section will be tested to determine the mechanical performance of this more economical size. 

The replacement tie refers to optimizing the plastic/composite tie for the best combination of per- 
formance and economics, without regard to matching characteristics of other ties or tie materials. Like 
the concrete tie, the replacement or optimized plastic tie, might produce a different track response and 
therefore be best applied when completely replacing long, continuous sections of wood ties. 

As the USACERL target is the light density line, it will be looking to develop a tie which works 
well with a standard cut spike or other inexpensive fastener. The intent is to minimize life-cycle 
tie/fastener cost while still maintaining acceptable performance. 

For wood ties, the unit cost is higher for larger and longer pieces and also higher grade pieces. 
An order with pieces of matching thickness and uniform surface charactenstics also has a higher cost. 
Thus, timber switch ties and bridge ties have higher unit costs (per board foot) than do standard track 



Proposed Manual Changes 1 87 



ties. It appears that the unit cost of plastic/composite ties is constant with respect to size. Short line 
and industrial railroads have a high proportion of turnouts and a high labor cost associated with 
switch tie replacement and therefore could benefit from the development of a plastic/composite 
switch tie. 

The results of the physical testing and the performance of the plastic/composite ties in track 
have been promising and they indicate that continued evaluation is warranted. The installation of a 
larger number of con.secutive ties is required to ensure that the pla.stic ties are bearing full in-,service 
loads. In the future, larger consecutive groupings of plastic ties are planned for the FAST track in 
Pueblo, mainline service on Conrail and Army trackage. 

Acknowledgments 

The writers would like to acknowledge the following personnel with their respective organiza- 
tions that have contributed to this project. Conrail: Jim Beyerl, Kurt Gansauer, P. Michael Lovette, 
Sam Luke; US Army Construction Engineering Research Laboratories: Richard Lampo, Principal 
Investigator; Rutgers University: Dr Rich Renfree. 

References 

Modem Plastics, Vol. 1, No. I, January 1997. 

Read, D., AAR Transportation Technology Center, Research and Test Department, July 1996. 

Mark's Standard Handbook for Mechanical Engineers, 8th ed., 1978, pp. 5-26, 30, 33, 34. 

Nosker. T. Renfree, R., Sachan, R., Van Ness, K. E., "Predictive Techniques for Comingled Pla.stic 
Properties". 

Nosker, T, Renfree, R., "Production Scale-Up for Recycled Polymer Composite Railroad (RR) Tics", 
New Jersey Commission on Science and Technology, July 1996. 



POLLUTION Costs. 

Brownie Tank 

Protects. 



Only Brownie's fuel pumping skid 
provides you with a heavy steel 
channel base in a complete, turn- 
key refueling system. All your 
diesel fuel gets into the 
locomotive so there's no costly 
spillage, waste or polluted ground. 

Brownie's fuel pumping skid is 
pre-wired, pre-piped and ready to 
install with no field assembly 
required. You simply hook up the 
inlet and outlet piping and run 
electrical power to it. 

With Brownie Tank, there is no 
sourcing, spec'ing or searching for 
pumps, filters, valves, meters or 
design configurations. 



So whether you're working with an 
outside contractor, architect or in- 
house operation, you can provide 
your locomotive refueling 
operation with the finest, turn-key 
skid in the industry. 

Lube and journal oil skids and 
methanol injection systems are 
also available. 

For a quote or more information 
simply call: 

Brownie Tank Mfg. Co., a 
division of Determan Welding 
& Tank Service, Inc. 
Minneapolis, Minnesota 

(612)571-1744 
Fax(612) 571-1789 




Performance Under Pressure. 3tom^ 



Memoir 

Eldon E. Farris 

1913-1997 

Mr. Eldon E. Farris, age 83, retired Office Engineer, Budington Northern Railroad died January 
2. 1997. Eldon was born in Dawson, but graduated from High School and the University of Nebraska 
in Lincoln, NE. He was very active in his community, church, and local organizations. 

He was an active member of the American Railway Engineering Association, particularly in 
contributing to Committee 1 — Roadway and Ballast. Eldron had been a long time Chariman of 
Subcommittee 5 on pipelines. Although not able to travel to committee meetings in recent years, he 
was still very actively contributing via correspondence. 

Eldon joined AREA in 1946 and was subsequently designated a Life Member after a long and 
distinguished service to the railroad industry. He will be missed by all his associates and friends. He 
is survived by his wife, Virginia, two daughters and four grandchildren. 



189 



190 Bulletin 760 — American Railway Engineering Association 

Memoir 

Edward Q. Johnson 

1917-1996 



Edward Q. Johnson, a former Chief Engineer with the Wabash Railroad, the Norfolk & Western 
Railway and the Chessie System Railroads, died December 13, 1996 at his home in Fredericksburg, 
VA at the age of 79. He is survived by his wife, Elizabeth, two sons, Edward, Jr., of Reston, VA and 
Richard of Bethesda, MD (predeceased by one son, Jeffrey), one daughter, Jill of Columbia, SC, five 
grandchildren and one sister, Helen Cook of Little Rock, AR. 

Mr. Johnson was bom in Globe, AZ but raised mainly in Flint, Ml. He graduated with a B.S. in 
Transportation Engineering from the University of Michigan in 1938 and began a long and distin- 
guished career in the railroad industry as a Chaimman for the Chicago, Rock Island & Pacific 
Railroad. Shortly thereafter, in 1941, he married Elizabeth Barber. 

Between 1938 and 1955, Mr. Johnson worked in different engineering department positions for 
a number of railroads, including the Wabash, The Chicago, Burlington & Quincy, and the Des Moines 
Union Railroad. His positions included Chainman, Rodman, Instrumentman, Assistant Engineer, 
Supervisor of Track, Supervisor of Bridges and Buildings, Terminal Engineer, and Division Engineer. 
In 1955, he was appointed General Manager of the Des Moines Union Railway in Des Moines, Iowa. 

In 1957, Mr. Johnson returned to the Wabash Railroad where he served as Construction 
Engineer, Manager — Operations Research, and Chief Engineer. In 1964, when the Wabash and sev- 
eral other railroads were merged with Norfolk & Western Railway, he was appointed Chief Engineer 
of Norfolk & Western where he stayed until 1971. 

In 1971, Mr. Johnson accepted a position with the Chessie System. During his tenure with the 
Chessie, he served as General Manager — Maintenance, Assistant Senior Chief Engineer, Chief 
Engineer and Assistant Vice President — Engineering. Under his direction, Chessie installed a com- 
plete maintenance-of-way computerized engineering information system to collect, analyze, plan, 
schedule and budget maintenance-of-way data and operations. He retired from Chessie in 1982 after 
43 years of continuous railroad experience. 

After his retirement from Chessie, Mr. Johnson worked with the firm of L.E. Peabody & 
Associates, Inc. in Alexandria, VA, as Senior Engineering Consultant. During his association with 
L.E. Peabody & Associates, Inc., which continued until his death, Mr. Johnson worked on numerous 
special projects and presented expert testimony before the Interstate Commerce Commission. 

Mr. Johnson served as Vice President in 1970 and President in 1971 of the American Railway 
Engineering Association ("AREA"), an organization with worldwide membership composed of rail- 
way engineers, railway supply personnel and individuals interested in solutions to railway engineer- 
ing problems. In addition, he served seven years on the Board of Direction of the AREA. 



One complete service. 
Lowest cost per mile. 




V. 



* A complete, objective test 
of each rail froin end to end. 

^ Simultaneous ultrasonic and 
induction detection methods. 

*Sperry far surpasses every other 
rail testing service in efficiency, 
thoroughness and research. 

* One mileage charge pays 
for everything. 

*The lowest real cost per mile 
and per defect found. 

Details and technical assistance on request. 



SPERRY RAIL SERVICE 

SHELTER ROCK ROAD DANBURY, CONNECTICUT 06810 
(203) 791-4500 



191 



DIRECTORY OF CONSULTING ENGINEERS 



HARDESTY & HANOVER, LLP 



Consulting Engineers Since 1887 



RAIL STRUCTURES & HIGHWAYS 
BRIDGES - MOVABLE & FIXED 

1501 Broadway, New York. NY 10036 

212-944-1150 FAX: 212-391-0297 

other offices: 

New Jersey, Florida , Virginia and Connecticut 




zr 

ZETA-TECH 



ZETA-TECH Associates, Inc. 

900 Kings Highway North 

Cherry Hill, New Jersey 08034 

(609) 779-7795 FAX (609) 779-7436 

e-mail: zetatech@zetatech.com 



Technical and Economic Consulting for Railways 
and Rail Transit 
Technical Consulting • Specialized Software 

- Railway Track and Rail - Track Maintenance Planning 

- Maintenance Management - Operations Simulation 

- Vehicle/Track System - Track Inspection and Analysis 



Economic Analysis 

- System Economics 

- Cost Benefit Analysis 



Costing 

- Maintenance Planning 

- Operations Simulation 



Visit our home page on the World Wide Web: 
http://www. zetatech. com 



IS r/a^sftSys/is/ns 



Providing services for engineering, 
design, planning, construction 
management and operations. 



75 75 Qroad Street, Bloomfield, NJ 07003 
Tel. 20 1-893-6000 • Fax 201-893-3 13 1 



192 



RAILROAD AND RAIL TRANSIT 

Inspection * Planning * Design • Construction Management 
Bridges • Tunnels * Structures • Stations • Yards • Shops 
Trackwork • Electrification • Signals * Communications 



fSi 



Gannett Fleming 

ENGINEERS AND PLANNERS 



P.O. Box 67100 • Harrisfaurg, PA 17106 • [717] 763 7211550 
California Street • San Francisco. CA 94104 • [4151 981-5335 

www.gannettfleming. cam 



^rk ENGINEERS 



CONSOER TOWNSEND ENVIRODYNE ENGINEERS, INC. 

■ Railyards and Shop Facilities ■ Environmental Engineering 

■ Bridge Inspection, Retiabilitation ■ Higti Speed Rail Studies 

and Replacement ■ Construction Management Services 

■ Railroad Planning Studies ■ Grade Crossing Analysis 



Chicago, IL 
(312)938-0300 



Regional Offices: 
Orange, CA New York. NY Nashville, TW 

(714)835-4447 (212)682-6340 (615)244-8864 

24 Offices Nationwide 



Complete Engineering Services for the 
Railroad Industry 



Tracks, Bridges, Structures 
Construction Management 
Shops, Warehouses, Offices 



Fuel Management 
Utilities 
Financial Studies 



BLACK &VEATCH' 



8400 Ward Parkway, Kansas City, MO 64114 (913) 458-2222. 
Offices Worldwide 
http//www. bv.com 



193 



THOMAS K. DYER, INC. 



Rail Transportation 

Consulting Engineers 

Signals Track/Civil Communications Power Electriilcation 

1762 Massachusetts Avenue 

Lexington, MA 02173 

(617) 862-2075 

NEW YORK • PHILADELPHIA • CHICAGO • DALLAS • ST LOUIS 



EIMBINEERIMS IMC. 



Intermodal Yards • Track Design 
Embankment Subgrade Stabilization 

Design Build 
Environmental • Drainage • Surveying 



Corporate 
Heaquarters 

630-858-7050 



Chicago, IL 
Area 

630-434-7050 



Nashville, TN 
Area 

615-552-2525 



Springfield, IL 
Area 

217-525-7050 



E-Mail sheath@patnckengineering com 



Web http://wwwpatrickengineering com 



Serhi\g the Railroad Industry 



STV 



Transportation Leaders 

Complete Planning, 

Engineering and 
Construction Services 



WilliamF. Matts, Sr.V.R 



Tel:212/777-4400 Fax: 212/529-5237 



225 Park Ave. South, New York, NY 10003 



web site: www.stviric.com 



194 



LTK Engineering Services 



Behind the scenes at some of America's 
most successfui rail systems... 

Contact us today. 

Corporate Headquarters 

Two Valley Square, Suite 300, Blue Bell PA 19^22 
215-5^2-0700 215-5^2-7676 FAX 

Regional Offices 

Portland, Los Angeles, Chicago, Philadelphia, Dallas. 
Seattle. San Francisco 




MQDJESKI«"''MASTERS 

CONSULTING ENGINEERS 
FIXED & MOVABLE RAILROAD BRIDGES 



P.O. Box 2345 

Harrisburg, Pennsylvania 17105 

(717) 790-9565 FAX (717) 790-9564 



1055 St. Charles Avenue 

Nev\/ Orleans, Louisiana 70130 

(504) 524-4344 FAX (504) 561-1229 



http://www2.epix.net/~modjeski/ 

Harrisburg • New Orleans • Poughkeepsie • Bordentown • St. Louis 




PARSONS 
BRINCKERHOFF 

Complete Rail Engineering, Planning and 
Construction Management Services 



' Railroads 

■ Rapid Transit 

■ Bridges 

' Systems 



• Tunnels 

• Shops & Yards 

• Track 

• Vehicles 



Washington, DC — Paul Reistrup (202) 783-0241 
San Francisco — Tony Daniels (415) 243-4600 

100+ Offices Worldwide 



195 



Providing Mationwide Railroad 
Engineering Services 

'.'y Yard and Terminal 
•jx/^Track and SJrif »«»" 
f^ iPassenger i^l 
. I^ignalizatioil 



'""Surging r , .. „ 

Envirenniental Services fi^ 

'Construction IManageinent ,*^ 

^S Carter -Bti 

Consultants in Enuinc^i „.„. 
Planning and the Environment 




Railroad Engineering Services 
Since 1910 



• Railroad Facilities & Building Designs 

• Track & Bridge Engineering 
Construction Administration & Surveys 
• Fueling Systems 



TKDA 



ENGINEERS • ARCHITECTS • PLANNERS 



1500 Piper Jaff ray Plaza • 444 Cedar Street • Sf. Paul. MN 55101-2140 



(612) 292-4400 



KR 



HDR Engineering, inc. 



We offer complete railroad 
engineering services. 

■ Tracks "Tunnels 

■ Bridges ■Facilities 

■ Water Resources ■Environmental 

8404 Indian Hills Drive 
Omaha, NE 68114-4049 
1-800-366-4411 
http://www.hdrinc.com/ 



Call for employment information 



196 



New York Pennsylvania Florida Ohio Michigan 



itm. 



Complete Rail Engineering 
and Management Services 



BERGMANN ASSOCIATES 

Engineers, Architects, Surveyors, PC 

j i. 

44 Hudson Place M (201)653-2898 

Hoboken, NJ 07030 f Fax (201) 653-3464 



Hazelet & Erdal 



Dames & Moore 

Successors to the 
Scherzer Rolling Lift Bridge Company 

serving the railroad industry since 1897 

FIXED and MOVABLE RAILROAD BRIDGES 

Design - New and Rehabilitation 

Inspection - Structural, Mechanical, & Electrical 

Rating and Analysis 

547 W. Jackson Boulevard, Suite 1500, Chicago, Dlinois 60661-5717 
Phone (312)461-0267 • FAX (312)461-0373 

CorjKirate offices nationwide providing a broad range of engineering services 



SHANNON & WILSON. INC. 

GEOTECHNICAL AND ENVIRONMENTAL CONSULTANTS 

• Landslide Evaluation & Correction 

* Embankment & Subgrade Stabilization 

•Tunnel Design & Maintenance 

• Environmental Management 

• Bridge Foundation Engineering 

Seattle • Richland • Fairbanks • Anchorage • St. Louis • Boston 

Corporate Headquarters, Seattle: (206) 632-8020 
400 N. 34th, Suite 100, P.O. Box 300303, Seattle, WA 98103 



197 



ESCA 



CONSULTANTS. INC. 

1606 WiaOW VIEW RO PO BOX 159 
URBANA. ILLINOIS (217)384-0505 



RAILROAD & HIGHWAY BRIDGES • TRACKWORK 
INDUSTRIAL FACILITIES • SPECIAL STRUCTURES 



INSPECTION & RATING 

REPORTS & STUDIES 

DESIGN & PLANS 

CONSTRUCTION SUPERVISION 



BO\^\IAN. BARRETT & ASSOC 



CONSULTING 
ENGINEERS 



3 1 2 



2 2 8.0100 
FAX 
312.228.0706 

DESIGNING A BETTER I N FRASTRVCTU RE 
TO LAUNCH US INTO 
THE 21ST CENTURY 



WE OFFER THE FOLLOWING RAILROAD DESIGN EXPERTISE: 

• MASS TRANSIT • INTERMODAL FACILITY 

• TRACKWORK • BRIDGES 
• INSPECTION AND RATING • DRAINAGE SYSTEMS 

• CONSTRUCTION MANAGEMENT • SURVEY 



130 E. RANDOLPH STREET SUITE 26.50 CHICAGO. ILLINOIS 60601 




NOLTE and ASSOCIATES, Inc. 

Engineers / Planners / Surveyors 

Serving Clients Ttiroughout tlie Western United States 



2950 Buskirk Ave., Suite 225, Walnut Creek, CA 94596 
Tel: (510) 934-8060 FAX No. (510) 939-5451 



HERZOG 



Herzog Contracting Corp. 

P.O. Box 1089 

St. Joseph, Mo. 64502 



816-233-9001 



Railroad Services 

• Railroad Construction & 
Maintenance 

t Commuter Rail Operations 

• Material Handling 

• Ultrasonic Rail Testing 

• Rail Car Leasing 



Comprehensive Transportation Engineering Services 



Kllam 



ConsLiltinL' En 



49 West 37th St., New York, NY 10018 • 212-869-7800 
1 University Plaza, Hackensack, NJ 07601 • 201-489-8080 

Other Offices: NY«NJ»PA*CT*MA»NH»MD«FL 



PARSONS TRANSPORTATION GROUP 

BARTON-ASCHMAN • DE LEUW, GATHER • STEINMAN 



Comprehensive Railroad 

and 

Transit Engineering Sen/ices 

Parsons Transportation Group Inc. 

1133 15th Street, NW, Washington, DC 20005 

(202) 775-3300 • Fax: (202) 775-3422 



http://www.parsons.com 



199 



Gary A. Gordon, P.E. 

President 




GORDON, BUA & READ, INC 



CONSULTING ENGINEERS 

34 SALEM STREET 

READING, MASSACHUSETTS 01867 

(617)944-7110 

Fax (617) 944-6708 



Civil 

Railroad 

Structural 

Transportation 



<TL 



serving the railroad industry 
Call (800) 522-2CTL 



• testing of ties, fastener systems, rails, rail joints 

• state-of-the-art dynamic testing equipment 

• million-lb static & dynamic capacity 

• track design/construction problem solving 

• onsite track system testing & instrumentation 

• vehicle component testing 

Claire G. Ball 

Construction Technology Laboratories, Inc. 
5420 Old Orchard Road, Skokie, IL 60077-1030 



Trackwork* Facilities* Operations'Structures* Environmental •« 



ffl?H 





R 



BRW INC. 

700 N.E. Multnomah 
Suite 1000 
Portland, OR 97232 
503/232-5787 
503/232-6373 FAX 



/ \ §■■ imi^mm^ aHmmmmt ^^miKA>'« . 

'Trackwork* Facilities* Operations'Structures'Environtnental 



200 



Index of Advertisers 



A&K Railroad Materials, Inc 76 

Association of American Railroads. . . . cover 3 
Brownie Tank Manufacturing Company. . . 188 

Burke-Parsons-Bowlby Corporation 34 

Burro Crane 75 

Cattron Incorporated 7 

Chemetron True Temper 61 

Danella Rental Systems, Inc 44 

Fairmont Tamper 106 

HDR, Inc 44 

Hanson & Wilson 179 

Kerr-McGee Chemical Corporation . . back cover 

Koppers 6 

L. B. Foster Company 62 

Loram Maintenance of Way, Inc 99 

Magnum Manufacturing Corporation 64 

Modem Track Machinery Inc./ 

Modem Track Machinery Canada Ltd. ... 66 



The Nolan Company 64 

Nordco ii 

Omni Products 7 

Osmose 23 

Pandrol Incorporated 90 

Pandrol Jackson 8 

Parker Hannifin Corporation 43 

Plasser American Corporation/ 

Plasser Canada Inc 180 

Premier Concrete Railroad Crossings 137 

Railquip, Inc 100 

Railway Tie Association 24 

Rocla Concrete Tie, Inc 23 

Sperry Rail Service 191 

Surety Manufacturing & Testing, Inc 138 

Sydney Steel Corporation 89 

W. H. Miner Division 82 

Westem-Cullen-Hayes, Inc 60 



Index of Consulting 

BRW, Inc 200 

BAC Killam Consulting Engineers 199 

Bergmann Associates 197 

Black & Veatch 193 

Bowman, Barrett & Associates 198 

CTE Engineers, Inc 193 

Carter & Burgess, Inc 196 

Constmction Technology 

Laboratories, Inc 200 

ESCA Consultants 198 

Gannett Fleming, Inc 193 

Gordon, Bua & Read, Inc 200 

HDR Engineering, Inc 196 

Hardesty & Hanover 192 



Engineers Directory 

Hazlett & Erdal/Dames & Moore 197 

Herzog Contracting 199 

L.S. Transit Systems 192 

LTK Engineering 195 

Modjeski & Masters 195 

Nolte & Associates, Inc 198 

Parsons Transportation Group 199 

Parsons Brinckerhoff 195 

Patrick Engineering 1 94 

STV 194 

Shannon & Wilson 197 

TKDA 196 

Thomas K. Dyer 194 

ZETA-TECH Associates, Inc 192 



201 



NOTES 



202 



WHEN IT COMES TO KEEPING 
TRACK OF CURRENT 
MR PUBUCATIONS, 
WE WROTE THE BOOK. 

Field Manual of Interchange Rules 

Office Manual of Interchange Rules 

Rules of Interchange for TOFC & COFC Interchange Service^ 

Rules Governing the Loading of Commodities on Open Top] 
Cars & Trailers 

Manual of Standards & Recommended Practices 

(Can Be Purchased as a Complete Set or Individually) 




Miscellaneous Specifications 
Specifications for Design 

Fabrication & Const. 
Car Construction — 

Fundamentals & Details 

Appendices 
Brakes & Brake 

Equipment 
Wheels & Axles 
Journal Bearings & 

Lubrication 
Trucks & Truck Details 
Lubrication (Shop) Manual 
Lettering & Marking of Cars 
Wheels & Axles (Shop) Manual 
Roller Bearing (Shop) Manual 

For Ordering Information, Call (202) 639-2211 



Side Frames & Truck Bolsters 
Locomotives & Locomotive 

Equipment 
Maint. Req. Brake 

Control Valve & 

Equipment 
Couplers & Draft 

Gears 
Quality Assurance 
Specification for 

Tank Cars 
M-lOOl 




Association of American Railroads 



Visit our 
Web Site for 



upcoming 

Special Events, 

Seminars and 

Exhibits: 



www.aar.org 

For More Information, 
Call (202) 639-2230 

50 F Street, N.W 
Washington, D.C. 20001 










FOREST PRODUCTS DIVISION 

Staying Ahead 
Qflhe 

ive. 



OSHA STAR 



Responsible Core™ 

IS09000 
CERTIFIED 

Ar Kerr-McGee, some of rhe roughesr 

srondords we ser ore our own. 

So when orgonizorions like ISO, OSHA and 

Responsible Core exonnine companies for quality, 

safety and environmenrol stewardship, you 

can rrusr rhar we will always meet their standards. 

Because keeping rhe qualir/ of our company 

OS high as rhe quality of our products is just 

one more way we're staying ahead 




KERR-MCGEE 

CHEMICAL 
CORPORATION 

TJfe^CKING QUALTIY FbRi%IERIG\. 

(405) 270-2424 • P.O. Box 25861 • Oklahoma City, OK 731 2f 




American Railway Engineering Association 

October 1997 

Volume 98, Bulletin 761 



BOARD OF DIRECTION 
1997-1998 

President 
Mr. p. R. Ogden, Norfolk Southern, Vice President-Engineering, 99 Spring Street, 
Room 801. Atlanta, GA 30303 

Vice Presidents 
Mr. D. C. Kelly, Illinois Central Railroad, Vice President-Maintenance, 17641 S. 
Ashland, Homewood, IL 60430 

Mr. Rick Richardson, Jr., Canadian National Railways, Chief Engineer, 935 
de la Gauchetiere Street. West Montreal. PQ-H3B 2M9 

Past Presidents 
Mr. B. G. Willbrant, Amtrak, Assistant Chief Engineer. 21 Berkshire Drive, Wayne, 
PA 19087 

Mr. J. R. Beran, Union Pacific Railroad, Chief Engineer-Maintenance of Way 
Structures, 1416 Dodge Street, Room 1000, Omaha. NE 68179 

Treasurer 
Mr. W. B. Dwinnell, III, SEPTA. Railroad Division. Chief Line Maintenance Officer, 
1234 Market Street, Philadelphia, PA 19107 

Directors 
Mr. J. R. Clark, Jr., CSX. Assistant Chief Engineer, 500 Water Street, Jacksonville, 
PL 32202 

Mr. E. p. Reilly, Union Pacific Railroad, Chief Engineer-MAV Central, 1860 Lincoln 
Street, 14th Floor, Denver, CO 80295 

Mr. L. Anderson, Illinois Central, Superintendent Engineering, P.O. Box 2600, 
Jackson, MS 39103 

Ing. Lorenzo Reyes R., FNM. Director Ferrocarril Sureste, Estacion Terminal, 
Montesinos S/N Colonia Altos. Veracruz. Mexico 91700 

Mr. W. C. Thompson, Union Pacific Railroad, Director Engineering Research. 1416 
Dodge Street, MC 3300. Omaha, NE 68179 

Ms. C. D. Wylder, MARTA, Executive VP Operations and Development. 2424 
Piedmont Road, N.E., Atlanta. GA 30324-3330 

Mr. R. L. Keller, Montana Rail Link. Chief Engineer, PO. Box 8779. 210 
International Way, Missoula. MT 59807 

Mr. John Cunningham, Amtrak, Assistant Chief Engineer-Track. 3rd Floor. South 
Tower 30th Street Station. Philadelphia, PA 19104 

Mr. Michael Roney, CP Rail System, General Manager-Engineering Services and 
Systems. Suite 500. Gulf Canada Square. 401-9th Avenue, S.E., Calgary, Alberta T2P 
4Z4 Canada 

Mr. Walter Heide, Conrail, Assistant Chief Engineer-MAV, 2001 Market Street, Room 
10-B, Philadelphia, PA 19101-1410 

Mr. Gary Woods, Norfolk Southern. AVP-Maintenance of Way Structures, 99 Spring 
Street, Box 142, Atlanta, GA 30303 

Mr. Michael Armstrong, BNSF, AVP-Maintenance Planning. 2600 Lou Menk Drive. 
Ft. Worth. TX 76131-2830 

Executive Director 

David E. Staplin 

50 F Street. N.W., Washington. DC 20001 

(202) 639-2190 



American Railway 
Engineering Association 

BULLETIN 
No. 761 

OCTOBER 1997 

Proceedings Volume 98 (1997) 

D. E. Staplin, Editor 

CONTENTS 

Cover Story: Stampede Pass Reopens 203 

Proposed 1998 AREA Manual and Portfolio Revisions 

(Chapters 7, 9, 15, 19 and Portfolio of Trackwork Plans) 209 

Presentations at the 1997 Annual AREA Technical Conference — March 1997 

1. Implementation of High Speed Rail in an Existing Corridor 265 

2. Rail Wear Management at Conrail 276 

3. The Changing Role of the Consultant 280 

4. Norfolk Southern's Engineering Department Action Plan for Safety . 285 

5. Effects of Track Maintenance on the Reliability of a Single Track 
Railroad Line as a Function of Axle Load 297 

6. Efficient Load Testing and Evaluation of Railroad Bridges 321 

7. Manufacturing In-House Machinery Projects 331 

8. Wayside System for Measuring Rail Longitudinal Force Due to 
Thermal Expansion of Continuous Welded Rail 337 

9. Conrail's Inifrastructure Reliability Optimization 347 

10. Florida's High Speed Rail Project: Building a Public- 
Private Partnership 353 

Technical Papers: 

1. Rail Travel (Creep) Caused by Moving Wheel Loads 365 

Memoirs: 

Edward M. Cummings 389 

Howard W. Lichius 390 

Richard K. Pullem 391 

Jack P Shedd 392 

Index of Advertisers 403 

Index of Consulting Engineers 403 

Front Cover: Rehabilitating the BNSF's Stampede Tunnel 

Published by the American Railway Engineenng Association. March. May. Oclobcr and December al 

50 F St.. N.W.. Washington. D.C. 20001 

Second class postage at Washington. D.C. and at additional mailing offices 

Subscription $81.(K) per annum 

Copyright© 1997 

AMERICAN RAILWAY ENGINEERING ASSOCIATION (ISSN (XK)3-0694) 

All rights reserved 

POSTMASTER: Send address changes to Amencan Railway Engineering Association Bulletin. 

50 F Street. N.W.. Washington. DC. 2(X)0I 

No part of this publication may be reproduced, stored in an information or data retrieval system, or transmitted, in any form, or by 

any means — electronic, mechanical, photocopying, recording, or otherwise — without the prior written permission of the publisher. 

Published by the 

American Railway Engineering Association 

50 F St., N.W. 

Washington, D.C. 20001 









"an Count O 




Reliability. Durability. Low maintenance. 



These are key benefits you receive by investing in Miner's high quality 
railcar components. Bocked by exhaustive R&D testing, comprehensive 
service analysis and over 100 years of actual rail experience, Miner 
products are components you can count on. 

Draft Gears 
Miner draft gears including the TF-880, Crov/n SE ', Crown TG" and 
SL-76 provide higher capacity and greater payloods for a wide range 
of service requirements. Certified under various A. A. R. M-901 spec- 
ifications and A. A. R. quality certification M-1003, Miner draft gears 
provide superior cushioning and increased protection of your railcars. 

Side Bearings 
TecsPak' Constant Contact (TCC| side bearings, featuring Miner's 
patented TecsPok elastomer pads, are designed to reduce truck 



hunting, increase car stability and lower component wear on various 
types of cars. TCC-III side bearings, specifically engineered for long- 
travel applications, are available in various preloads and maintain 
operation preload ot temperature extremes. 

Discharge Devices 
Miner's heavy-duty unloading systems offer reliable, trouble-free 
discharging of coal, grain, chemicals, cement, ballast and other 
commodities. Whether you need to unload covered hoppers, open-top 
hoppers, coal cars or ballast cars. Miner has the discharge device that 
allows you to do it faster, safer and more effectively. 
Contact your Miner representative today for products that are reliable, 
durable and reduce your mointenonce costs. Invest in products from 
W. H. Miner.. .roilcor components you can count on! 



^^^ Products Engineered for 21" Century Performance 

^^INER W. H. MINER DIVISION 



1200 East State Street. PO Bon "171 , Geneva. IL iMIM . 1-630/232-3000 • FAX 1-630/232-3055 

P, O, Box 178. Lachine. Quebec. Canada H8S '1A6 • l-51'l/637-0445 • FAX 1-514/637-2126 

3 ch.de la Procession. B-7870 Lens. Belgium . 32/68/454737 . FAX 32/68/454733 

Duo Jose eenedetli. 18. 09531-000. Soo Coetono do Sul. Brosll ■ 55/1 1/453-6431 • FAX 56/1 1/442-1787 



D 1996 Miner Enterprises, Inc. 



STAMPEDE PASS REOPENS 

By: Bob Powers, P.E, and Ron Poulsen* 

In 1983, the Burlington Northern Railroad's Stampede Pass Route was determined to be a 
redundant asset. At that time, it was thought that the Stevens Pass Line, with the 8 mile (13 km) long 
Cascade Tunnel, and the Columbia River Gorge Line would be able to provide sufficient east-west 
capacity over the Cascade Mountains well into the twenty-first century. As a result, approximately 
153 miles (246 km) of the Stampede Line was sold to the Washington Central Railroad, an indepen- 
dent short line operating between Pasco and Cle Elum. Wash. The Burlington Northern Railroad 
retained the remaining 97 mile (156 km) segment but removed it from active service. 

Ten short years later, Burlington Northern was reevaluating this decision. Traffic on the exist- 
ing east-west routes for the now Burlington Northern Santa Fe was at or near capacity. Bolstered by 
the booming Pacific Rim import-export trade, the decision was made by Burlington Northern Santa 
Fe to resurrect the Stampede Pass Line, thus easing the congestion in the region and allowing the rail- 
road greater flexibility in freight handling. 

With the decision made to reopen Stampede Pass, the emphasis shifted to how quickly the line 
could be made operafional. Starting with preliminary engineering in October 1993, many complex 
issues had to be analyzed and resolved to complete the project. HDR assisted Burlington Northern 
Santa Fe in developing this project and solving these complex issues by providing engineering, envi- 
ronmental and construction services. 

Elements of the design work consisted of rehabilitating the 2 mile (3 km) long Stampede Tunnel. 
This included snowshed replacement at the portals of the tunnel, bridge rehabilitations, culvert replace- 
ments, sidings at Kanaskat, Lester and Easton, Wash., as well as wye tracks at Lester and Easton for 
turning snow removal equipment. Design work also included site layout in addition to structural and 
utilities design for maintenance-of-way facilities at Kanaskat, Lester, Easton and Martin. 

Construction of the project began in July 1996 with substantial completion scheduled for early 
December 1996. Substantial completion for the project was defined as the owner's ability to operate 
trains 24 hours a day without restricting train speeds. Almost immediately, unforeseen and unique 
construction obstacles began to present themselves. These obstacles required immediate attention if 
trains were to run the first week in December 1996. 

Rehabilitating Stampede Tunnel 

The single biggest challenge in the resurrection of Stampede Pass was the rehabilitation of 
Stampede Tunnel, and the prevailing theme throughout this challenge was constructability. The 
Stampede Pass Tunnel is located deep in the Cascade Range and access is extremely difficult. The 
nearest concrete plant is more than 35 miles (56 km) away over rough, hilly terrain. Steep side slopes 
on the approach cuts at the portals severely limit the fiexibility of a construction operation. The tim- 
ber snowsheds that once protected the east and west portals of the tunnel were destroyed by fire sev- 
eral years ago and are located in avalanche runout zones, or areas in which avalanches decelerate, 
stop and form a deep dense snow deposit. Magnitudes of the avalanche loads, submitted by the 
avalanche subconsultant, were on the order of 2000 psf (96 kPa) vertical and 500 psf (24 kPa) lat- 
eral. The proposed replacement structure would have to address these concerns as well as be water- 
tight and maintenance friendly. 

The resulting snowshed design by HDR was a precast segment shaped like an inverted L. The 
erection process consisted of setting two of these segments in place and then bolting the segments 



♦HDR Engineering. Inc. 



203 



204 



Bulletin 761 — American Railway Engineering Association 




Work In Progress 

in the Stampede 

Pass Project 



'v..>ivf m ."* 11--' r^ / I *i 



'*! 




Paper by Bob Powers, PE. and Ron Poulsen 205 



together at the top, with the final configuration resembling a three-legged box. In all, 208 segments 
were set in 20 days at the east portal and 92 segments were set in 16 days at the west portal, for a 
total of approximately 600 ft (183 m) of new snowshed. The segments were transported from the pre- 
cast supplier, via a flat-bed truck, to a temporary holding zone adjacent to the mainline track and then 
shipped by rail, two at a time, to the project site. Once erected, the segments were capped with a 
structural concrete roof slab and n extensive water-proofing system. The snowsheds were then back- 
filled to reduce possible avalanche forces acting on the snowsheds, as well as to eliminate any poten- 
tial safety hazard. 

Replacing Bridge 46.6 

Another unique challenge in the reopening of Stampede Pass was the replacement of Bridge 
46.6 at the east portal of the Stampede Tunnel. As the east portal. Stampede Creek flows parallel to 
the track for approximately 1 10 ft (34 m) and then turns north and crosses the tracks. Winter storms 
and spring thaws caused frequent ice jams and drainage problems at the single span structure and 
were responsible not only for the deterioration of the structure but often for the temporary closing of 
the line. Bridge 46.6 was considered to be the most significant maintenance concern, and therefore 
the weakest link, on the original Stampede Pass Line. 

The solution was to replace Bridge 46.6 with a heated box culvert. Six precast, hydraulically- 
sized culvert sections were set into place and coupled together carrying Stampede Creek under the 
tracks. Heat cables were cast into the bottom slab and side walls of each precast culvert section. The 
culvert heating control system consisted of an automatic and manual control mode, allowing for a 
continuous flow of water in through the culvert and preventing ice jams from accumulating. Record 
snow fall accumulation and sub-zero temperatures have proven the design to be extremely success- 
ful to date. 

Nature Complicates Construction 

In August 1996, unsuitable soil for the snowshed foundation was encountered at both the east 
and west portals of the Stampede Pass Tunnel. Within 48 hours, the site was investigated by HDR 
along with Shannon and Wilson, the geotechnical subconsultant, and a revised design was submitted 
to the contractor. 

Nature further proved to be the most significant obstacle during the construction of this project. 
On Sept. 10, 1996, after heavy rains saturated the project site, the hillside immediately above and to 
the east of Stampede Creek (East Portal of Stampede Tunnel) slid into the creek and into the snow- 
shed construction boundaries. Subsequent slides dumped soil, rock and debris onto foundation exca- 
vations and the tracks to the east of the creek. The site was immediately investigated by HDR and 
Shannon and Wilson. Substantial engineering decisions and design changes were made in the field 
on remedial work measures to stabilize the slide and maintain project schedules. Such measures 
included rock bolting, slope stability and stabilization, and the design of a soldier pile tie-back wall. 

With heavy rains continuing through October, it became apparent that if work continued as 
planned, substantial completion by December 1996 was not possible. Understanding the importance 
to Burlington Northern Santa Fe of having an operational line by the first week in December, HDR 
and the contractor reevaluated the work activities. Work activities essential to the running of trains 
were accelerated and the rest of the activities were delayed. As a result, the contractor redirected and 
rescheduled his work forces. On Dec. 7, 1996, Burlington Northern Santa Fe ran the first commer- 
cial train over the newly reopened Stampede Pass Route. 

The reopening of Stampede Pass proved to be a truly unique project in both design and con- 
struction. Clearly, if ever a project resisted completion by presenting challenge after challenge, the 



206 



Bulletin 761 — American Railway Engineering Association 



reopening of Stampede Pass was that project. Quick response and teamwork were the prevailing fac- 
tors throughout the life of the project 1996. 

Stampede Pass Line Remains Open During Landslides 

Record snow fall, ice storms, wind storms, landslides and sub-zero temperatures tested the 
design early this past winter In December 1996 and January 1997, record snow and rain fall, cou- 
pled with a brief warming trend, resulted in numerous landslides throughout Washington State and 
the Pacific Northwest, which closed the Stevens Pass Line and Columbia River Line. The only 
Burlington Northern Santa Fe line to remain operational in the region was the newly reopened 
Stampede Pass Line. All other lines were forced to temporarily close as a result of the landslides. 
Freight operations in the region would have come to an abrupt halt if Burlington Northern Santa Fe 
had not decided to reopen Stampede Pass. 



The Single-Component Solution 

OMNI's Standard Concrete-Rubber (SCR) System, 

concrete panels with built-in rubber flangeway fillers: 

Streamlines handling 

and installation 

• 

Eliminates the hassles 

associated with 

top-down rubber 

• 

Saves time and money 

1-800-203-8034 

OMBdl^ 

GRADE CROSSING SYSTEMS ^_^ 




■ r 1 A 1 Gauge Panel 

I Steel Angle Rubber Flange . 



-0 ■ u ■ 



Optional Lag Design 



Meet the family of track experts 



Tie Tampers in six sizes, Tie Handiers, Tie insertefs, Gmding Modules and 
Trains, Uitmonic Rail Flaw Detection and a Ml mge of Rail Measument 
Services. Qual'itf Equipment n^anufactured to meet or exceed the demand- 
ing standards of ISO 9001. Wfien tfie t)ottom line counts, count on us. 






•j(2 Pandrol 
Jackson 

Pandrol Jackson, Incorporated 
200 South Jackson Road 
Ludington Michigan 49431 
Telephone 616-843-3431 
FAX 616-843-4830 



fKoppers inausxnes... solving trooiems ror Mmenca s Mauroaas 



Why Koppers Industries? 

Thirteen Strategically Located Wood 

Treating Plants • 920 Dedicated 

Employees • Over Eighty Years of 

Service to Railroads • A National Tie 

Procurement Network • Wood and 

Concrete Crossties • Wood and Concrete 

Switch Ties • Used Tie Disposal • Car 

Cleaning • Tie Pre-Plating • Track Panels 

• Concrete Turnouts • Wood and 

Concrete Grade Crossings 

Enough Said? 




Tie Pre-Plating 



Track Panels 



Tie Disposal 



Turnouts 



Railroads around the world know 
Koppers Industries, Inc. for high 
quality crossties. But there is more... 
much more. Today, KM is helping 
America's railroads with a wide array 
of innovative products and services... 
all designed to solve problems and 
improve efficiency. The Railroad 
Products and Services Division offers 
track panels, concrete turnouts, 



wood and concrete ties, switch ties 
and wood and concrete grade cross- 
ings. Our cost cutting services 
include tie pre-plating, tie disposal, 
warehousing, crosstie rehabilitation, 
car cleaning and special packaging. 
Call us today. Let's talk about our 
solutions to your problems. 
Koppers Industries. ..in partnerstiip 
witti America's Railroads. rpsdi 



KOPPERS 

INDUSTRIES 



Railroad Products 
and Services Division 

Koppers Building, 

436 Seventh Avenue 

Pittsburgh, PA 15219 

Phone 1-888-567-8437 

Fax 412-227-2328 



Proposed 1998 AREA Manual 
and Portfolio Revisions 

The following proposed Revisions of the AREA Manual for Railway Engineering and Portfolio 
ofTrackwork Plans have been recommended to the Association by the Technical Committee respon- 
sible for each after a letter ballot is approved by: ( 1) a two-thirds majority of the eligible members 
voting, and (2) by at least fifty percent of the total eligible voting members on the committee. They 
are being published here for comment by the general AREA membership and any other interested 
parties. Comments should be sent to AREA headquarters by December 1, 1997. The.se comments will 
be considered by the AREA Board of Direction in deciding whether to give final approval for inclu- 
sion of the proposed changes in the Manual and Portfolio Revisions, which if approved, go into effect 
August 1. 1998. 



Proposed 1998 Manual Revisions to 
Chapter 7 — Timber Structures 

Part 1 — Material Specifications for Lumber, Piles, Glued Laminated 
Timber and Fasteners 

Page 7-1-25. Add the following new Section 1.14 Specifications for Timber Bridge Ties. 

1.14 SPECIFICATIONS FOR TIMBER BRIDGE TIES 

1.14.1 MATERIAL 

1.14.1.0 Kinds of Wood 

Before manufacturing ties, producers shall determine which species of wood is suitable for 
bridge ties and which species will be acceptable by the railway. 

1.14.2 PHYSICAL REQUIREMENTS 

1.14.2.1 General Quality 

The general quality of bridge ties shall conform to the appropriate grading rules. All ties shall 
be free from any defects that may impair their strength or durability as bridge ties; such as decay, 
splits, shakes, excessive slope of grains, or numerous holes or knots, bark, wanes, etc. 

1.14.3 DESIGN 

1.14.3.1 Support Conditions 

Depending on the intended .service conditions, bridge ties may be classified as structural or 
bearing ties. Structural ties are normally used for open deck bridges having steel girder spans. Under 
these conditions the strength of the ties is governed by flexure or horizontal shear. Bearing ties are 
normally used for open decks of timber trestle spans or on open decks of steel beam spans having 
multiple beams where the strength of ties is governed by bearing on the top of the stringer flange. 

1.14.3.2 Dimensions 

a) The minimum cross-section for structural and bearing type bridge ties shall be based on the 
applicable clauses of Chapter 7, Part 2 and Chapter 15, Part 1 of the latest revisions of this 
manual. 



209 



210 Bulletin 761 — American Railway Engineering Association 



b) The minimum width of bridge ties shall be eight (8) inches nominal. 

c) When the ties are dapped, the minimum depth of the tie shall be the one obtained by 1.14.3.2 a) 
above. 

d) The minimum length of bridge ties shall be ten (10) feet nominal or center-to-center of outer 
supports plus three (3) feet whichever is greater. 

1.14.4 INSPECTION 

1.14.4.1 Place 

Before accepting ties for installation, the bridge ties will be inspected at locations specified by 
the railway. 

1.14.4.2 Manner 

Prior to treatment, inspectors will make a reasonably close examination of the top, bottom, sides 
and ends of each bridge tie with regard to its manufacture and compliance with respect to the grad- 
ing rules. Each bridge tie will be judged independently, without regard to decisions on other ties in 
the same lot. 

1.14.4.3 Handling 

Bridge ties are to be handled with care to prevent damage. Damaged ties will not be accepted. 

1.14.4.4 Decay 

Bridge ties with signs of incipient or active decay will not be accepted. 

1.14.4.5 Knots 

Bridge ties with a large knot or numerous knots will not be accepted. A large knot is one in 
which its average diameter exceeds one-fourth of the surface on which it appears. Such a knot may 
be allowed if it occurs outside of the bridge tie supports in an overhang area. Numerous knots are any 
number equaling one-half of a large knot in damage effect. 

1.14.4.6 Shake 

Bridge ties which contain a shake of more than one-fourth the width or depth of the bridge tie 
will not be accepted. A shake is a separation along the grains, most of which occurs between the rings 
of annual growth. 

1.14.4.7 Split 

Bridge ties split more than three inches will not be accepted. A split is a separation of wood 
extending from one surface to an opposite or adjacent side. 

1.14.4.8 Slope of Grain 

Except in woods with interlocking grain, a slope in grain in excess of 1 in 15 will not be permitted. 

1.14.4.9 Manufacture 

a) Except as provided in the grading rules of the respective species, all bridge ties shall be 
straight, well sawn, cut square at the ends, have bottom and top parallel (except for tapered 
ties) and sized surfaces when specified. 

b) The surface of a bridge tie will be considered straight: (1) when along the top and bottom 
surfaces, a string line stretched from mid-width at one end of the tie to mid-width at the 
other end of the tie is within ± 1/8 inches of being truly straight and (2) when along each 
side a string line stretched from mid-depth of one end of the tie to the mid-depth of the other 
end of the tie is within ± 1/8 inches of being truly straight. 



Proposed Manual Changes 2 1 1 



c) A bridge tie is not well sawn when its surfaces are cut with score marks more than 1/8 inch 
in depth or when its surfaces are uneven. Such ties shall not be accepted. 

d) The top and bottom surfaces or opposite sides of a bridge tie will not be considered parallel 
if any difference in tie thickness measured along the depth or along the sides exceeds 1/4 
inch. On sized or dapped .sections the difference in the thickness cannot exceed 1/8 inch. 

1.14.4.10 Dimensions 

The following tolerances in sawn or machined dimensions of bridge ties shown on the railway's 
plan will be accepted, unless otherwise specified by the railway. 

Width: ± 1/8" 

Depth: Sized or dapped areas: ± 1/16", Others: ± 1/8" 

Length: ± 1/4" 

Fabricated holes: ± 1/16" 

1.14.5 DELIVERY 

1.14.5.1 Location 

Bridge ties delivered for acceptance shall be stacked at suitable and convenient locations as 
approved by the railway. Bridge ties delivered on the premises of a railway for inspection shall be 
stacked on blocking placed on firm ground, not less than ten (10) feet from the nearest track but not 
at or near a public crossing, nor where they will interfere with view of trainmen or other vehicles 
approaching the railway. 

1.14.5.2 Risk, Rejection 

All bridge ties are at the supplier's risk until accepted. All rejected ties shall be removed from rail- 
way premises by the supplier at his expense within thirty (30) days after the date of the inspection. All 
rejected material not removed by the end of this period will become the property of the railway. 

1.14.6 SHIPMENT 

Bridge ties shipped by railway cars or other means shall be separated into bundles therein 
according to bridge locations for which they are intended, and also according to the location on the 
bridge spans, unless otherwise stipulated in the contract, on the railway order form or on the accom- 
panying plans for the ties. 

1.14.7 DAPPING OR SIZING BRIDGE TIES 

1.14.7.1 Dapped or sized ties may be used on bridges. Dappingor sizing of ties should be per- 
formed in a framing mill properly equipped to perform such work. Dapping or sizing 
should be performed before treatment. If performed after treatment, application of the 
preservative to exposed surfaces should be made in the field before installation. 

a) When dapped bridge ties are used, the width of dap shall be the width of flange 
plus 1/2 inch and the depth of dap shall be 1/2 inch or such that the undapped 
portion will not bear on gusset plates, bracing, etc. 

b) When sized ties are used, the railway may specify SIS, S2S, SIE, S2E or S4S. 

c) On curved tracks, superelevation may be provided by tapered ties, which may 
be dapped or sized. Superelevation may also be provided by raising blocks 
securely fastened to the bottom of ties. An approved tie plan must be provided 
to the framing mill and the ties should be uniquely and individually numbered 
to identify ties having different dapping dimensions. The method of numbering 
shall comply with the requirements of the railway. 



Bulletin 761 — American Railway Engineering Association 



1.14.8 BRIDGE TIE INSTALLATION 

1.14.8.1 Bridge Tie Spacing and Spacers 

a) The recommended nominal clear distance between the bridge ties is four (4) inches. 

Also refer to Part I of Chapter 1 5 of the latest revision of this manual for tie spacing require- 
ments. 

b) Bridge tie spacers may be a minimum 4" x 8" wood or 3" x 5/8" steel bar having predrilled 
holes for fasteners or of other design as specified by the railway. 

c) A tie spacer shall be fastened to each bridge tie with 5/8" diameter drive spikes, lag screws 
or lag bolts. To avoid splitting, it is recommended to pre-bore holes in the ties. 

1.14.8.2 Rail Fastening 

The type of rail fasteners to be used will be determined by the railway. 

a) For spikes refer to Chapter 5, Part 2 of the latest revision of this manual. 

b) For spiking refer to Chapter 5, Part 4 of the latest revision of this manual. 

c) For other fastening systems refer to manufacturers specifications. 

1.14.8.3 Tie Plates 

a) For tie plates refer to Chapter 5, Part 1 of the latest revision of this manual. 

b) To prevent plate cutting of bridge ties, a suitable size of double shouldered tie plate shall be 
used taking into consideration species of wood, axle loads, predominant train speeds, track 
curvature, etc. 

c) The minimum recommended size of tie plates is: 

Main line bridges decks: TVi" x 14" 
For other bridge decks: 7" x 11" 

d) The railway may use tie-plates of special design (i.e., for direct fixation of rails) to prevent 
ties from being spike-killed. 

1.14.8.4 Bridge Tie Pads 

a) Tie pads may be used to minimize plate cutting and to reduce impact and vibration effects 
on the bridge structures. 

b) Tie pads may be made of a plain or reinforced elastomeric material, impregnated fibrous 
material or any other suitable product, provided they are strong enough for the loading, are 
water repellent and stay firm in shape during service. 

c) The size of tie pad shall conform to the tie plate used and shall be of suitable thickness. 

d) Refer to Chapter 10 Part 1 of the latest revision of this manual for material requirements and 
testing. 

1.14.8.5 Bridge Tie Fastening 

a) For fastening bridge ties to timber stringers, one of the following anchoring systems may be used: 
i) Bolts or drive spikes. 

ii) Machine bolts with adequate washers and nuts, 
iii) A combination of i) and ii). 

b) For fastening bridge ties to steel beams and girders, one of the following anchoring systems 
may be used: 

i) Machine bolts with a plate or spring washer and standard or lock type nut. 



Proposed Manual Changes 213 



ii) Hook bolts with a plate or spring washer and standard or lock type nut. 

iii) Machine bolts with a clip and plate or spring washer and standard or lock type nut. 

iv) Other systems may be used if approved by the railway. 

v) Ties installed on the rivet or bolt heads of built-up girders should have the fasteners re- 
tightened after traffic has set the new deck down on the girder flange. 

c) The size and the spacing of the anchoring system should be such as to provide adequate sta- 
bility for the open deck considering the loads and forces as described in Chapters 7 and 1 5 
of the latest revisions of this manual. 

d) Refer to Chapter 7, Part 1 and Chapter 15, Part 8.3 of the latest revision of this manual for 
additional guidelines. 

1.14.9 PRESERVATIVE TREATMENT OF BRIDGE TIES 

Refer to Chapter 3, Parts 6 and 7 of the latest revision of this manual. 

1.14.10 SPIKE OR BOLT HOLES 

Refer to Chapter 3 of the latest revision of this manual. 

1.14.11 TIE PLUGS 

Refer to Chapter 3, Part 1.6 of the latest revision of this manual. 

1.14.12 TIE BRANDING 

Refer to Chapter 3, Part 1.4.5 of the latest revision of this manual. 

1.14.13 END SPLITTING CONTROL DEVICES 

Refer to Chapter 3, Part 1 of the latest revision of this manual. 



Part 2 — Design of Wood Railway Bridges and Trestles 
for Railway Loading 

Page 7-2-14. Revisions to Pile Load Distribution. Replace present Figure 2-5, Figure 2-6, Figure 2- 
7, Figure 2-8, Figure 2-9 and Figure 2-10 with new Figure 2-5 plus accompanying graphs and tables. 



214 



Bulletin 761 — American Railway Engineering Association 



Figure 2-5 
Distribution of Load to Piles of Timber Trestles 



R - live load ukcn by one bcni (one nil ), Kips 

X. - loid uken by one middle rile. Kipi 

Xk • load Uken by one inlennediau pile. Kips 

X, - laid uken by one pcnullinuie pile, kipl 

Ci. C). Ct- disuncc of inner pile from ccnurlinc of bent, inches 

b - disuncx between out£r piles, inches 

a - width or strinfcf be vlnt on cap, inches (assumed unifonn ). For biMast floor br4d|es may be taken as 

60" 

I " momem of iitcrtia of cap. in* 

A- cross section area of cap. in' 

L ~ elTcctive Icnfih of pile, inches. Taken as exposed len{ih * 1/2 penelialion. 

D > depth of cap, inches 

0- breadth of cap. inches 

E - modulus of eluticiiy of cap material, in bendini. Kips/in' 

Et ' modulus of elasticity of cap material, Ijansversc, (say I/SO E for wood). 

G " modulus of elasticity of cap material, shear, (say 1/16 E for wtjod ). 

Er ■ modulus of elasticity of cap maieriaJ, in axial compression, Kips/in* 

q,. q^q.- DericclionofcapduclaR- l,X, X«,X,nalactJn| 

r^ r«. r,> deflection of cap due 10 X, • 1, acting alone 

s^s«,s,- deriectJon of cap due 10 X^ - I. acting alone 

(,t^t,- deflection ofcap due 10 X, -I, acting alone 

u • Compression of cap andshoruningofpilcduetounitloadonpilc, inches (pile assumed to be 1 4" dia at 
top tapered to 10" dia at 30* A.ILE.A. Spec.). Cap compftssion assumed distributed over length - B -* 
2*, where a ■ distance above the bottom of cap. 



7PIIFBF>fr 



lf^l% - ■ ■ ■ 






X. X, X.X, X, X. 



7 PILE BENT 



8 PILE BENT 



Dtfltction Equtlioni ( for consistent defonnalions of cap and piles ) 



(r.■►'/,u)X.•►(s.»u)X.+ (^♦u)X,-(ll.♦ u) R 
(r,»'/,u)X.t(i,*«)X.»(l,*u)X.-((i,+ u)R 
(r,-> '/,u) X.> (s.* u)X>t (I.* u) X,-(<i.-m)R 



* 2u) X.* (S.+ u) X,* (I.+ u) X, - (q.+ u) R 

♦ u) X,+ (S.+ 2u) X.* (t.* u) X. -(q,* «) R 
*u)X.*(s.*u)X>*(l.*2u)X.-(q.*u)R 



Deflection Coefncients ( defledions of cap, due to unit loads ) 



Due 10 R - I 



_L[...^.6(30,'bOO.'.4(30,'].*ii^ .... 
3±.[<3,.....„„o,'(|-C.)-4(|-C.]\i(l.30- 



+ -^-^ b-60-at 

ag[ 

q, = substitute C for C> m equation (S) above 



J^[(3b'-a'-.2(30,»(|-C.]-4(|-C.)'4[l. 



B subsiitttie Ci for C* in cqiution (4) above 
* sufaftmitc C> for Ci in cquMion (4) above 



Due to X. - 1 




,,._L_.b'*Jii.b 

4tEI AG 




'-iaii-]-' 


-<l-)' -S 


r. - substitute C for Ci i 


n equaiion (1) above 


Due to Xk - 1 





a x.-it Tx.-i ^ 



t. • substitute Ci for G in equaiion (I) above 



= ^(|-C.][3<|-C.].<|-C.)'-(|.C.)'] 



..(10) 

..(11) 



AGI2" 



-■] 



.^(|-C|3<|-C^-<|.C^(|.C 



..(!•> 
..(II) 



DurloX.-! 



t.'2r. 

k = S. 



.._L(,..c.(|-C.)\i|(|-C.] 



..(13) 
..(14) 



Atir^'^xVrA 



6E1^ \1 J AGU J 



(14) 



Pile Shortening and Compression of Cap ( due to unit load on pile ) 



Application I. Substitute design dimensioru in equations (4) to (16) inclusive. 

2. Substitute these coelTicienu in equations ( I ). (2) and (}). 

3. Solve equations ( I ), (2) and (3) for X, X, aiuJ X, giving loads on all inner piles, outer piles take remaining load 
HqIC: For 3-pile bent and 4 pile bent, use equaiion ( I ) omilling X, and X^ 

For S-pile bent and 6 pile bent use equations (I) AND (2) omilling Xf. 



Proposed Manual Changes 2 1 3 



SPECIFICATIONS FOR DESIGN OF WOOD BRIDGES 
AND TRESTLES 

Distribution of Load on Piles 

(Figure 2.5 plus accompanying graphs and tables) 

Example No. 1 

The 5 pile-bent of a trestle which carries a chord of bunched stringers under each rail, has a 14" x 
14" timber cap. The spacing of the piles is 36" and the effective length of piles (i.e. the exposed length 
plus one-half of the penetration) is 30 feet. Each chord possesses four 8" x 16" stringers. Using graphs 
or tables, find out the distribution of the wheel load (assumed as one or axle assumed as two) on the piles. 

Given: 

a = 3 1", c2 = 36", b = 144" and L = 30 feet 

Select the appropriate graph or table. 

Intermediate pile (2) = 0.562 

Outside pile (3) = 0.133 

Centre pile (1) = 2x (1 -(0.562-1-0. 133)) = 0.610 

Answer: Pile # 12 3 4 5 

Load 0.133 0.562 0.610 0.562 0.133 

Note: The middle pile takes the maximum load, then the intermediate piles and the load carried 
by the outside piles is the smallest. 

Example No. 2 

Data same as in Example No. 1, except that the chords now consist of five 8" x 16" stringers. 
Find out the distribution of wheel load on piles of the bent. 

Given: 

a = 39" and the rest of the data is same as in Example No. 1 

Select the appropriate graph or table. 

Intermediate pile (2) = 0.550 

Outside pile (3) = 0.143 

Centre pile (1) = 2x( 1-0.550-1-0.1 43)) = 0.614 

Answer: Pile # 12 3 4 5 

Load 0.143 0.550 0.614 0.550 0.143 

Note: Increase in the value of "a" has resulted in decrease of load on the intermediate piles and 
a corresponding increase of load on the outside and the centre pile. 

Example No. 3 

Data same as for Example No. 1, except that there is a ballast deck now in lieu of an open deck. 

Given: 

a = 60" and the rest of the data is same as for Example No. 1 

Select the appropriate graph or table. 

Intermediate pile (2) = 0.514 

Outside pile (3) = 0.177 

Centre pile (1) = 2x( 1 -(0.5 14-h0.1 77)) = 0.618 



216 Bulletin 761 — American Railway Engineering Association 



Answer: Pile # 12 3 4 5 

Load 0.177 0.514 0.618 0.514 0.177 

Note: Compared to the Example No. 2, the ballast deck causes further decrease of load on the 
intermediate piles and corresponding increase of load on the outside and centre piles. 

Example No. 4 

Substitute 15" x 15" concrete cap in lieu of the 14" x 14" timber cap in Example No. 1 and then 
find the distribution of wheel load on piles of the bent. 

Given: 

a = 31", c2 = 36", b = 144" and L = 30 feet 

Select the appropriate graph or table. 

Intermediate pile (2) = 0.505 

Outside pile (3) = 0.197 

Centre pile (1) = 2 x (l-(0.505+0.197)) = 0.596 

Answer: Pile# 12 3 4 5 

Load 0.197 0.505 0.596 0.505 0.197 

Note: The 15" x 15" concrete cap being stiffer than the 14" x 14" timber cap of Example No. 
1, provides a better distribution of wheel load on piles i.e. lower value for the intermediate and the 
middle piles and higher value for the outer piles. 

Example No. 5 

Same as Example No. 1, except that the bent is shallow now with an effective length of piles 
L= 10 feet. 

Given: 

a = 31,c2 = 36", b= 144" and L= 10 feet 

Choose the appropriate graph or table 

Intermediate pile (2) = 0.598 

Outside pile (3) = 0.099 

Centre pile (1) = 2 x (l-(0.598+0.099)) = 0.606 

Answer: Pile # 12 3 4 5 

Load 0.099 0.598 0.607 0.598 0.099 

Note: Compared to the Example No. 1, the piles are a lot stiffer in this case and as such there 
is more load on the intermediate piles and less load on the middle and the outer piles. 

Example No. 6 

The 6 pile-bent of a trestle which carries a ballast deck and has a 14" x 14" timber cap. The 
spacing of the piles is 30" and the effective length of piles (i.e. the exposed length plus one-half of 
the penetration) is 30 feet. The deck possesses ten 8" x 16" stringers. Using graphs or tables, find out 
the distribution of the wheel load (assumed as one) on piles of the bent. 

Given: 

A = 60", cl = 15", c2 = 45", b = 150" and L = 30 feet. 

Select the appropriate graph or table. 



Proposed Manual Changes 



217 



Intermediate pile (2) = 0.380 

Outside pile (3) = 0.114 

Middle pile (1) = 1 -(0.380+0.1 14) = 0.506 

Answer: Pile # 12 3 4 5 6 

Load 0.114 0.380 0.506 0.506 0.380 0.114 

Note: The outside piles are carrying a smaller amount of wheel load even when compared to a 
5 - pile bent of Example No. 3. 

Example No. 7 

Same as the Example No. 6, except that the timber cap is now substituted with a 15" x 15" 
concrete cap. 

Given: 

Other data remains the same as for Example No. 6. 

Select the appropriate graph or table. 

Intermediate pile (2) = 0.363 

Outside pile (3) = 0.159 

Middle pile (1) = 1 -(0.363+0. 159) = 0.478 

Note: The 15" x 15" concrete cap being stiffer than the 14" x 14" timber cap of the Example 
No. 6 provides a better distribution of wheel load on piles. 



4 PILE BENT 

:i t 




CHART LEGEND 



EFFECTIVE PILE LENGTH OF Iff 
EFFECTIVE PILE LENGTH OF 30' 



218 



Bulletin 761 — American Railway Engineering Association 



AREA Pile Load Distribution Tables and Graphs 



4 PILE BENTS 
( 9 charts total ) 

b = 80. 90. 100. 110. 120. 130. 132, 140 & 144 Inches 



Pile Cap 


Eff. Pile 
Length 


a 


Cl 


C2 


C3 


Table 


Chart 
No. 


12" X 14" Timber 


10' 


23 


12.18,24 


- 




Yes 


1 


29 


12,18.24 


- 




Yes 


2 


31 


12,18.24 


- 




Yes 


3 


30- 


23 


12,18.24 


- 




Yes 


1 


29 


12.18.24 


- 




Yes 


2 


31 


12,18,24 


- 




Yes 


3 


14" X 14" Timber 


10' 


23 


12,18,24 


- 




Yes 


4 


29 


12,18,24 


- 




Yes 


5 


31 


12,18,24 


- 




Yes 


6 


30" 


23 


12,18,24 


- 




Yes 


4 


29 


12.18.24 


- 




Yes 


5 


31 


12.18,24 


- 




Yes 


6 


1 5" X 15" Concrete 


10- 


23 


12.18.24 


- 


- 


Yes 


7 


29 


12.18,24 


- 


- 


Yes 


8 


31 


12.18,24 


- 


- 


Yes 


9 


30- 


23 


12.18.24 


- 


- 


Yes 


7 


29 


12,18,24 


- 


- 


Yes 


8 


31 


12,18,24 


- 


- 


Yes 


9 



CHART 1 

4-PILE BENT 

12" X 14" TIMBER CAP 

a=23" 























































^ 


^ 














01 


■W's 


V 


^ 




... 


■;^ 


^ 


in. 








Cf 


"■\ 


s ^ 


^ 


>y. 


,> 




"r-y 


< 


-^ 


i 




CI. 


''■x 


S f 


/ 


^\ 


'< 


> 


>. 


^ 








i 






y 


■^ 


■^ 


'^'. 




y^.'- 














x' 


y.. 


'y 


K- 


'j' 


-fj 






INNER f 


ILE 


Xa) 










^ 


^' 




















i 






s 


S-; 




















ps 


s 


s 


S'; 


^ 


•<.•■ 






otr 


rER 


>ILE 


(X) 




i 






K 


>^ 


s 


n"; 


^ 


■^•^ 












K 




CI. 


r^ 




\ 


«? 


f^ 


->v" 


s 


s 
















C1. 


ir^ 




\ 


^ 


<. 




■^ 


t<. 


.^^ 














CI 


•ar' 




^ 




■•• 


> 


-~ 


























■^ 


•^ 


0.1 ■ 



























3 

= 0.7 



2 0.8 

t 

3 

Q 0.5 



O 0.4 

g 

P 



CHART 2 

4-PILE BENT 

12" X 14- TIMBER CAP 

a»29" 





















































-■ 












Cl 


■ws 


^^. 




^ 


^ 


;^ 


^ 








Cl- 


"•"n 


S 


^ 


X 


^■^ 


'^ 


"^ 




tt 




CI. 


"n 


s 


^ 


A 


.- 


,• 




> 


•■ 










y. 


Xi 




y. 




>.'■ 












vv 


<< 


V. 


y. 




^.• 






INh 


ERP 


ILE 


X«) 




^ 
s^ 


^ 


> 


X' 




















^ 


**? 


s. 


■Si 






















S; 


s 


^• 


5^ 


^•- 






oir 


tR 


>ILE 


PO 








s 


x: 


S 


N" 


^ 


■S" 














Cl. 


r^ 




V 


x" 


*^ 


>> 




<; 














Cl- 


^r'' 




^ 


> 




>> 


■^ 




.:^ 












Cl. 


<tr- 






■^ 


^" 


■rt^ 


t^ 


























^ 





























so so 100 110 120 130 140 

SPACING "b" OF OUTER PILES (InclMt) 



80 to 100 110 120 130 140 

8PACIN0 -b- OF OUTER PILES (Inehu) 



Proposed Manual Changes 



219 



CHART 3 

4-PILE BENT 

12" X 14" TIMBER CAP 

a*31" 









i 












































^ 


^ 














Cl 


■«"n 


\ 


t^ 




^^ 


^ 


:^ 


?., 








Cl- 


ir y 


s 


X 


^ 


^;^ 




rr!. 




r^ 


f 




CI- 


"•s 




^ 


X 


i»< 


■^. 




^ 








i 








>< 




y. 




<:, 
















^ 


' y^'' 


^ 


<^'- 






INN 


ERF 


ILE 


X.) 




a 


^ 

V /- 


K^ 


> 


<^ 




















u. 




*^ 


s 


S-; 






















■N 


'■"S 


N 


b 


S"; 






(XT 


■ER 


»ILE 


(X) 




i 






S 


y 


^N 


V 


■j- 


■>s"' 












K 
O 




c. 


\r/ 




< 


V 


^ 




s 




... 






S 0.3 

s 








CI- 


ir/ 




^ 


V 


k^ 


r^ 


^ 


^ 


.;: 














Cl 


If 


/ 




^^ 


••• 


•^ 


^ 


























"^ 


^ 


0.1 - 





























CHART 4 

4-PILE BENT 

14" X 14" TIMBER CAP 

a»23" 




80 90 100 110 120 130 140 

SPACINO -b- OF OUTER PILES (InclMS) 



80 SO 100 110 120 130 140 

SPACINO "V OF OUTER PILES (kiclws) 



CHART 5 

4-PILE BENT 

14" X 14" TIMBER CAP 

a«29" 




CHART 6 

4-PILE BENT 

14" X 14" TIMBER CAP 

a«31" 

























































-- 














Cl 


■J4-S 


N 


, — 




-^ 




:> 


s 








Cl> 


""•\ 




X 


X 


»■> 




> 




rr 






Cl- 


"•\ 




^ 


A 


K 


^ 


.-;. 


>r 


■■■ 






i; 






y 


><, 




>. 




>. 












% ■ 




A 


]'^ 


^ 


;^ 








INN 


ERF 


ILE 


X.) 








x^ 


]^ 


i^' 




















ft 01 




^ 


N 


S--, 






















S 


'S 


S;- 


"^ 


s- 






oir 


■ER 


>ILE 


(X) 




1 






s 


v: 


^ 


X" 


•ir 


X" 














Cl- 


r ^ 




V 


X 


*s 


^ 






-. 














Cl- 


r^ 




^ 


> 


*>^ 


>v 


■^^ 




^ 














Cl 


ij«-- 


/ 






































■^ 


0.1 ■ 





























80 so 100 110 120 130 140 

SPACING -b~ OF OUTER PILES pnchat) 



80 SO 100 110 120 130 140 

SPACING -b- OF OUTER PILES (Inchn) 



220 



Bulletin 761 — American Railway Engineering Association 



Q 0.5 



5 0.4 

z 
o 
p 

a: 
O 

2 



CHART 7 

4-PILE BENT 

IS" X 15" CONCRETE CAP 

a«23" 



















































^ 


— 












CI 


■>«-\ 


\, 






^ 


^ 


-' 








CI. 


"■\ 






>s 




^ 


^ 


-^ 






CI- 


""n 




x 


K 


< 




> 


!^-' 


^• 










^ 


X 




^ 
^ 


^ 


i-- 












^ 


y 


y 






i'-- 






INN 


=RP 


ILE( 


Xi) 




% 




)y. 






















< 


3»> 


^ 






















^ 


\ 


<s 


•V 


> 


f ;; 






Oil 


ER 


>ILE 


(X) 








s 


V 


<^ 






i\- 














CI" 


yr^ 


r ■" 


\ 


V 








»■» 






•- 








CI- 


r/ 






V 




■^ 


^ 


^ 


•« 












CI 


j«-' 


A' 








^ 


■^ 
























^ 


--. 





























O 0.4 

p 



CHART 8 

4-PILE BENT 

15" X 15" CONCRETE CAP 

a-2«" 





















































" 












CI 


J<s| 


s 


^ 








-- 








Cl- 


"•n 






>< 


>T^ 








.. 




CI- 


'=-s 




^ 


X 


^ 


^ 


> 


?*•■ 


„■ 












X 




"^^^: 


:: 


••■ 










/> 


y 

■^ 


>: 


*:• 






INNER PILE (X«) 




•^ 




^r 






















•p 




% 


S-; 






















F'^ 




> 


jr; 






I I 
OUTER PILE (X) 










X 


"^ 




s 


'"■'.' 


•:: 


■.- 










CI- 


\T^ 




X 


^^ISk- 


':'. 




:~ 








CI- 


ir/ 




^ 


>< 


X 








•> 












Cl-J<-' 


/ 


"^ 








>^ 


























^ 





























BO W 100 110 120 130 140 

SPACING -b- OF OUTER PILES (Inch**) 



80 SO 100 110 120 130 140 

SPACING ■%- OF OUTER PILES (Inchat) 



CHART 9 

4-PILE BENT 

15" X 15" CONCRETE CAP 

a«31" 




80 80 100 110 120 130 140 

SPACING "b* OF OUTER PILES (InchM) 



Proposed Manual Changes 



221 



5 PILE BENT 

i t 




CHART LEGEND 



EFFECTIVE PILE LENGTH OF Iff 
EFFECTIVE PILE LENGTH OF 3ff 



5 PILE BENTS 

( 16 charts total) 

b = 90. 100, 110, 120, 130, 132, 140, 144 & 150 inches 



Pile Cap 


Eff. Pile 
Length 


a 


C1 


C2 


C3 


Table 


Chart 


12" X 14" Timber 


10- 


29 


- 


24,30,36,42 




Yes 


1 


31 


- 


24,30,36,42 




Yes 


2 


39 




24.30,36,42 




Yes 


3 


60 




24,30.36.42 




Yes 


4 


30' 


29 




24.30.36.42 




Yes 


1 


31 




24,30.36.42 




Yes 


2 


39 




24.30.36.42 




Yes 


3 


60 


- 


24.30.36.42 




Yes 


4 


1 4'X 14" Timber 


10' 


29 




24.30.36,42 




Yes 


5 


31 




24.30.36.42 




Yes 


6 


39 




24.30.36,42 




Yes 


7 


60 




24.30.36.42 




Yes 


8 


30- 


29 




24.30.36.42 




Yes 


5 


31 




24.30.36.42 




Yes 


■6 


39 




24.30.36.42 




Yes 


7 


60 




24.30.36.42 




Yes 


8 


16" X 16" Timber 


10' 


29 




24.30,36.42 




Yes 


9 


31 




24.30,36.42 




Yes 


10 


39 




24.30.36.42 




Yes 


11 


60 




24.30.36.42 




Yes 


12 


30- 


29 




24,30.36,42 




Yes 


9 


31 




24,30.36,42 




Yes 


10 


39 




24,30,36,42 




Yes 


11 


60 




24,30.36,42 




Yes 


12 


15" X 15" Concrete 


10- 


29 




24.30.36.42 




Yes 


13 


31 




24.30.36.42 




Yes 


14 


39 




24.30.36.42 




Yes 


15 


60 




24.30,36.42 




Yes 


16 


30" 


29 




24.30.36.42 




Yes 


13 


31 




24.30.36.42 




Yes 


14 


39 




24.30.36.42 




Yes 


15 


60 




24.30.36,42 




Yes 


16 



Bulletin 761 — American Railway Engineering Association 



CHART 1 

S-PILEBENT 

irX 14" TIMBER CAP 

■-29" 




90 100 110 120 130 140 150 

SPAaNO -b- OF OUTER PLES (hwhm) 

CCNTBI NX (X4 CMOm THC HDMMOei OPTM UNn- LOAD 

FKOM EACH RAIL 



CHART 2 

S-PILE BENT 

12" X 14" TIMBER CAP 

■-31" 




90 100 110 120 130 140 150 



SPACMO *b* OF OUTER P1£S Onchw) 

! PO) CAMaEl TIB NBUMOtR or TMl UMT UMO 
PKINIIACHIIAI. 



CHARTS 

6^ILE BENT 

12" X 14" TIMBER CAP 

■-39" 




90 100 110 120 130 140 150 

SPACINO -b" OF OUTER PILE* (Ineim) 

CaoCR HLI (Xa) CAWUCI TMC naiAliaCR or TIC UMT UlAO 

PROMfACHIIAl. 



CHART 4 

5-PILE BENT 

12" X 14" TIMBER CAP 

a-60" 




100 110 120 130 140 ISO 



SPACINO "b" OF OUTER PILES (bichM) 

CCNTCK PiLC (Xa) CAMOn TMC fUMAMOOl Of TMC UMT LOAO FROM 
EACH RAM. 



Proposed Manual Changes 



223 



CHARTS 

6-PILE BENT 

14" X 14" TIMBER CAP 

a-29" 




CHART 6 

S-PILE BENT 

14" X 14" TIMBER CAP 

a»31" 




SPACINO -b' OF OUTER PI^S (inctici) 
ccNTcn pu pui uioae* tmi muHoai or tmc umt umo from 

CACHKML 



SPACINO -b- OF OUTER PILES (Inctwi) 

CCMTXR RLE <Xa) CAMOeS TME ROUWaaPI or THC UMT LOAD FROM 
lACHRU. 



CHART? 



CHART 8 



5-PILE BENT 

14" X 14" TIMBER CAP 

a-39" 




100 110 120 130 140 ISO 



SPACINO -b- OF OUTER PILES (InchM) 

CENTER Pl^ (XI) CARRKS TW RCM«M)CR Of TK UMT LOAD FROM 
EACMRML 



S-PILE BENT 

14" X 14" TIMBER CAP 

a»60" 




100 110 120 130 140 ISO 



SPACINO "b' OF OUTER PILES (Inclwt) 

CENTER nLE |Xa) CiWREI T>C REMiUNOER Of TME UMT LOU) FROM 
EACHRJUL 



224 



Bulletin 761 — American Railway Engineering Association 



CHART 9 

6^LE BENT 

16" X 16" TIMBER CAP 

a-29" 




90 100 110 120 130 140 150 

SPACmO -b- OF OUTER PILES OnclMS) 
CCNTCK nu (Xa| CMWC* TW RCMMNOCR or TIC INT LOM> mOH 

ttemua. 



CHART 10 

8-PILE BENT 

16" X 16" TIMBER CAP 

a-31" 




so 100 110 120 130 140 ISO 

SPACING ■%" OF OUTER PILES (InclMS) 
ecNTCii nu pui cjuDun na RUUUNCMii or TMi mar UMO FROM 

nCMRM. 



CHART 11 



CHART 12 



S-PILE BENT 

16" X 16" TIMBER CAP 

a»39" 



SPILE BENT 

16" X 16" TIMBER CAP 

a»60" 




BO 100 110 120 130 140 ISO 

8PACIN0 1' OF OUTER PILES (InchM) 

CCHTDI PU (Xa| OMMU TW RMUaaSR W TM* a«T LOAD FMOM 
UCMIIM. 



S O.S 

s 

I 0.4 





IKT 


FERK 


lEDL 


»TEI 


>ILE 


<Xb) 




























C2> 


■^\ 


















ca- 


s^s 






\ 












C2> 


■"'s 






\ 


^< 




h 






C3 


•2€ 


N 


^ 


>^ 


=^ 


5? 


^ 






>^. 




*** 


^ 


<^ 














^ 


s 


t>< 


'^ 




















^ 


:^ 






b 


^ 




Ol 


TER 


PIU 


(X) 








^ 


^ 






^ 


>^ 


■>«,' 








CI- 


M-/ 






> 




<^ 


^■ 


r~«^ 




■^ 








CJ- 


rr'' 






y 


A' 


^ 




^ 












o 


ar/ 






/ 


^ 


^ 
















C2> 


«•/ 

































so 100 110 120 130 140 ISO 

SPACINO -b* OF OUTER PILES (InchM) 
ccMTiii rut ixa| CAMon tmc RCiuuNoeiar the UMTLOionioM 

eACNRM. 



Proposed Manual Changes 



225 



CHART 13 

5-PILE BENT 

15" X 15" CONCRETE CAP 

a-29" 




CHART 14 

6-PILE BENT 

IS" X 15" CONCRETE CAP 

a-31" 




SPACING "h" OF OUTER PILES (Inchm) 

CEMTei nu (lU) CAMOCS DC RDUWOBt or DC UMT UMD mOM 

fACHIUUL 



90 100 110 120 130 140 ISO 

SPACINO f OF OUTER PILES (Inchm) 

CCNTm nu |X>| CMM» Tia ROUaOai or 1W UMT LOAD FROM 
tikCHIUUL 



CHART 15 

6-PILE BENT 

15" X 15" CONCRETE CAP 

a-31" 



3 

a. 
ui 
i 0.5 

> 
a 
z 

UJ 

to. 
9 



i 0.2 

s 





IN- 


ERh 


ESM 


TEF 


ILE(Xb) 






















O 


'""n 


s 


o 


""■s 












CJ- 


"■s 






\ 


^ 


-?5 


^ 


^ 


C2. 


'•^N 


s 






:> 




«^ 


^ 


\ 




\ 






^ 


d 


:i>- 

^ 


,11 


1 1 • 


..- 


::' 




''* 




^ 


■^ 


^ 


>> 


















^ 


■s- 






















\ 

s 


n] 


^ 


'a^ 


r-. 


••-. 






3irrE 


.RPI 


.E(X 


1 


^ 


S 




^ 


'^ 


^ 


:- 


-.' 










o 


^/ 




^ 


^ 

N 


h^ 




^ 


^ 










' 


■7-jcr 


/" 


f\ 


s,^ 






-^ 




^ 


'>>^ 








c 


2-3r 


> 




■> 


^ 






■^ 














»Mr 


/* 


"^ 




^ 


_^ 



























CHART 16 

5-PILE BENT 

15" X 15" CONCRETE CAP 

a«60" 



80 100 110 120 130 140 ISI 

SPACMO -to- OF OUTER PILES (InchM) 
c eNrai pu ix<i cMwsi THC imuaaoi or THt uwT lOM) FROM 

lACHMI. 




SPACmO -b- OF OUTER PILES (inchas) 

CCNTm FU |Xa| CAIWCS T«e RIMAMOBI OF THE UNn- LOAD FROM 
CACHRAI. 



226 



Bulletin 761 — American Railway Engineering Association 



6 PILE BENT 

'A t 




CHART LEGEND 



EFFECTIVE PILE LENGTH OF Iff 
EFFECTIVE PILE LENGTH OF 3a 



AREA Pile Load Distribution Tables and Graphs 

6 PILE BENTS 

( 20 charts total } 

b = 100. 110. 120. 130. 132. 140. 144 & 150 inches 



Pile Cap 


Eff. Pile 
Length 


a 


Cl 


C2 


C3 


Table 


Chart 


12" X 14" TImtjer 


10- 


28 


12. 15. 15 


36, 39, 45 


- 


Yes 


1 


31 


12,15, 15 


36. 39. 45 


- 


Yes 


2 


39 


12, 15, 15 


36. 39. 45 




Yes 


3 


60 


12, 15, 15 


36. 39, 45 




Yes 


4 


30- 


29 


12,15,15 


36, 39. 45 


- 


Yes 


1 


31 


12,15, 15 


36. 39, 45 


- 


Yes 


2 


39 


12,15,15 


36, 39, 45 


- 


Yes 


3 


60 


12, 15, 15 


36, 39, 45 


- 


Yes 


4 


14"X14"Timl)er 


10' 


29 


12,15,15 


36, 39. 45 


- 


Yes 


5 


31 


12,15,15 


36. 39. 45 


- 


Yes 


6 


39 


12, 15, 15 


36. 39, 45 




Yes 


7 


60 


12,15,15 


36, 39, 45 




Yes 


8 


30- 


29 


12, 15, 15 


36. 39, 45 


- 


Yes 


5 


31 


12.15,15 


36, 39, 45 


- 


Yes 


6 


39 


12, 15,15 


36, 39, 45 


- 


Yes 


7 


60 


12, 15. 15 


36. 39, 45 


- 


Yes 


8 


16" X 16" Timber 


10' 


29 


12, 15. 15 


36. 39, 45 


- 


Yes 


9 


31 


12, 15. 15 


36. 39, 45 


- 


Yes 


10 


39 


12, 15. 15 


36. 39. 45 


- 


Yes 


11 


60 


12, 15, 15 


36, 39, 45 


- 


Yes 


12 


30' 


29 


12, 15, 15 


36, 39, 45 


- 


Yes 


9 


31 


12, 15, 15 


36, 39, 45 


- 


Yes 


10 


39 


12,15,15 


36, 39, 45 


- 


Yes 


11 


60 


12.15,15 


36, 39, 45 


- 


Yes 


12 


15" X 15" Concrete 


10' 


29 


12. 15,15 


36, 39, 45 




Yes 


13 


31 


12.15.15 


36. 39, 45 


- 


Yes 


14 


39 


12. 15, 15 


36, 39. 45 




Yes 


15 


60 


12,15,15 


36, 39, 45 


- 


Yes 


16 


30- 


29 


12,15,15 


36. 39. 45 


. 


Yes 


13 


31 


12. 15, 15 


36. 39. 45 




Yes 


14 


39 


12,15,15 


36. 39, 45 


- 


Yes 


15 


60 


12, 15. 15 


36, 39, 45 


- 


Yes 


16 


15" X 18" Concrete 


Iff 


29 


12. 15, 15 


36, 39, 45 


- 


Yes 


17 


31 


12,15.15 


36, 39. 45 




Yes 


18 


39 


12.15,15 


36. 39. 45 


- 


Yes 


19 


60 


12, 15. 15 


36. 39, 45 


- 


Yes 


20 


30- 


29 


12.15.15 


36, 39, 45 


- 


Yes 


17 


31 


12.15.15 


36, 39, 45 


- 


Yes 


18 


39 


12. 15. 15 


36, 39. 45 


. 


Yes 


19 


60 


12, 15, 15 


36, 39, 45 


- 


Yes 


20 



Proposed Manual Changes 



227 



CHART 1 

6-PILE BENT 

12" X 14" TIMBER CAP 

a-29" 




CHART 2 

6-PILE BENT 

12" X 14" TIMBER CAP 

a«31" 




SPACING -b- OF OUTER PI^S (InchM) 

«MD( HLE PUI CAmaei TIC HCMMBQI or TIC UMT LOAD moil 
ttCMIUM. 



SPACtNQ "b' OF OUTER PILES (lnch«) 

fWEK nte pia) CiUWC* nc RCMAMDOt or nc UNn- LOW FROM 
EACH RAO. 



CHART 3 



CHART 4 



6-PILE BENT 

12" X 14" TIMBER CAP 

a»39" 





1 1 1 1— I 

IKfTERMEDIATE PILE pQ>) , 

1 1 1 1 1 


fir. 


caMT 


^L 






1 1 c^•^r.C3^^r 


V 






^ 










r-^ 


— -■ 


~~ 


^•: 












'^. 






\- 




\i 


\ 






— 


a 








^ 


rrr- 






"n 


V- 








... 




O 

& at 








z 

1 


■r^ 








































O 0.3- 

i 


























^^ 
















O 




F^ 


^ 


f^ 


^ 












a 


C1-1 


r.o 


J^ 


"^ 


-^ 


V 


2:: 




^ 


^ 










Ei-ir. 


c2-ar 


v 




"7 


^ 


"^ 





c 


UTER 


PiLEpq 

— 1_.. 




CI" 

— 1 


r.c3- 

1 1 


^/^ 







6-PILE BENT 

12" X 14" TIMBER CAP 

a«60" 




SPACMO -b* OF OUTER PILES (InchM) 
— ctwmMrAimgnwgmMWiiomo— TLOnomoM 

CACMRML 



SPACINO "b- OF OUTER PILES (InchM) 

»«CR nU |XO CMKCt -nc REMMOCH OF TW IMT LOAD FROM 
CACHRAI. 



228 



Bulletin 761 — American Railway Engineering Association 



CHART 5 

e-PILE BENT 

14" X 14" TIMBER CAP 

a-29" 




CHART 6 

S^LE BENT 
14" X 14" TIMBER CAP 

a»31" 




I I I I I 

INTERMEDIATE PILE (Xb) 
-ci-ir. 



ci-ir.04r 




SPACINO "b" OF OUTER PILES (Inclwi) 

MNER nu (X4 CiUmCS nc RQUMDei OF TIC UNR- UMO ntOH 

EACHRML 



SPACING 'b- OF OUTER PILES (InclMt) 

■NNCR MLE (Jte) CAimet THC REtUMDOt or TIC UNn' UMO FROM 

CACHRAI. 



CHART? 

6-PILE BENT 

14" X 14" TIMBER CAP 

a-39" 




CHARTS 

6-PtLE BENT 

14" X 14" TIMBER CAP 

a«60" 



1 1 1 1 1 

INTERMEDIATE PILE (Xb) 
1 1 r 1 1 


1 1 

C1«ir.C3»4S" V 






1 


ci-ir.ojBT. 

1 1 K 






V 


\, 




2-.c2or^ 














.a . 


. 




rr 


N, 




" V 


^ 






rrrr 


1^ 


rTTT 


^ 




^ 


-rm 




-" 


- — 





"■' 


^^ 










































">^- 
















^ 


^ 


^ 


^ 














Cf 


r.c2- 


\:v 






^ 


^►'^ 




^ 









c 


-1«*.C 


2.3«r 


7"^ 




*•■!> 

^ 


F 


^ 












ct> 


r.o 


^v 






OUTER PILE (X) 

_i__J L_ 















SPACING -b- OF OUTER PILES (InchM) 

■•Cll M< (bl CAMOEt THE mUMDOl or TIC UMT UMO nKNi 

EACHRM. 



SPACINO -b- OF OUTER PILES (Inchm) 

MNEK fU |X>) CARRCS THE REMAMDCR Of THE UWT UMD FROM 
EACHRAL 



Proposed Manual Changes 



229 



Q 0.3 



CHART 9 

6-PILE BENT 

16" X 16" TIMBER CAP 

a«29" 



■I I 

INTERMEOW 

1 1 


r— 1 

TC PILE (Xb) c 
1 1 


1 1 






1 1 


ci.ir.ei-3i 


fr 






N 




Ct 1 


■ 


K 










...^ 




^ 











1 


\ 








■-^ 


. . • • 




-^ 


:rT 






\' 


_ 




■ 




-^ 


^ 


rrrr- 


























































^ 






















^^ 


~ > 


S 




**^' 


*■'. 










T 


Cl«1 


TO 


./^ 




1 




"■^^ 


:::: 


■^ 


^ 






ci-ir.cj 

1 


^l 


r 

Cl- 


ir.cj" 


5 


"^ 


~-~ 




. ..1 


LTTER 

1 


PILE(X) 

1 















CHART 10 

6-PILE BENT 

16" X 16" TIMBER CAP 

a«31" 




SPACING "b' OF OUTER PILES (Ineh**) 

t PKC |X4| CARXeS THE REMAMIEK CF THE IMT LOU) ntOH 
EACHRML 



SPACING -b- OF OUTER PILES (InchM) 

MNEIt PU IXal C«M» THE RUUnOai OF THE UNIT LOAD n«MI 
EACHRM. 



CHART 11 

6-PILE BENT 

16" X 16" TIMBER CAP 

a«39" 




100 110 120 130 140 ISO 

SPACmO -b- OF OUTER P&£S (kichM) 

MNCR FU |Xs) CAMRC* THE ROUWOER or THI UMT LOAD fWMI 
EACHRAI. 



CHART 12 

6-PILE BENT 

16" X 16" TIMBER CAP 

a«60" 




SPACING -b' OF OUTER PILES (InChM) 

MNER RLE (U| CARRIES THE REMAHOER OF THE UMT LOAD FROM 

EACHRAR. 



230 



Bulletin 761 — American Railway Engineering Association 



CHART 13 

6-PILE BENT 

15" X 15" CONCRETE CAP 

a-29" 




CHART 14 

6-PILE BENT 

15" X 15" CONCRETE CAP 

a-31" 





Till 

INTERMEDIATE PILE (Xb) 

1 1 1 1 




I I ■ 
ci-ir,c2-«- . 

1 1 Vs 










— - 


ci-ir.oaCk^ 






\ 




"■i:::+k 






i 


s 












^, 




.... 






— -- 




;ier 


^ 




•" 















































































^ 








>^- 


















^ 


2 






■>>'•- 


• -I; 






C1.1 


r.o; 


rP' 






fe 


7^ 






t 










CI>l$-.C2-3Br 

1 


V^ 




> 


? 




0' 


Ol 


ITER 

1 


»ILE(X) 

1 




CI- 
1 


ir.cxp ' 

1 







SPACING -b- OF OUTER PILES (InchU) 

MNO« PU (X4 CAMiaet nc ROMHOCK V THC UMT LCMD raOM 

EACH Ml. 



110 120 130 140 150 

SPACINO "b" OF OUTER PILES (InchM) 

NNER HLE (Xa) CJUUUeS THeimUMOei OF THE UNR' |JQ«0 FROM 
eACMRiUL 



CHART 15 

6-PILE BENT 

15" X 15" CONCRETE CAP 

a»39" 



1 1 

INTERMEDIATE 

1 1 


PILE 
01 


(Xb) 

■ip.ca 


^''v- 


c,.,P.c«r^| 






^ 


! ■ *\ 


J^.^ 









"\ 


[\ 






\ 








^ 


-rr 








.. 








^ 









.•-• 








■^ 


"^_ 










•••• 






r 




















































■*V 

^ 


r-' . 


^ 




&" 














^ 


^ 


s 


/ 




•••• 


P^-' 


..._ 




CI 


■ir.c 


.J 


^^ 


^ 


.^^ 


^ 








CI-IP.O 


.^ 




^ 


57?= 


O 


nER 


31LE( 

1 


<■) 




c 


1-1P. 


a-4r 


r ! 



CHART 16 

6-PILE BENT 

15" X 15" CONCRETE CAP 

a>60" 




SPACING ■%- OF OUTER PILES {fnehm) 
MCHRML 



SPAaNO "b* OF OUTER PILES (InelM*) 

■Wm nu (Xa| CAmSI THE NCMMNOei OF TME UMT UMID FROM 
EACHRM. 



Proposed Manual Changes 



231 



CHART 17 

6-PILE BENT 

15" X 18" CONCRETE CAP 

a«29" 




CHART 18 

6-PILE BENT 

15" X 18" CONCRETE CAP 

a-31" 



1 1 1 

INTER»«ED1ATE PILE (») cl-ir.CJ.«-J 




SPACING *b- OF OUTER PILES (kiclM*) 
« nu (U) CARioe* itc wiuuNocii or nc UMT LOAD fhom 

UCHRM. 



SPACINO 'b- OF OUTER PILES (InclMt) 

MHOI PU (1U| CAIWBt T>C IttlUMJCII or THE UWT LOikO raOM 
EACH WW. 



CHART 19 



CHART 20 



6-PILE BENT 

15" X 18" CONCRETE CAP 

a-39" 




6-PILE BENT 

15" X 18" CONCRETE CAP 

a-60" 





1 

r^TER 


1 II 
MEDIATE PILE (Xt 
1 1 


1 

) 














c 


•ir.c 


"'\ 


CI- 


r.c2- 








CI- 


r. c»jF . 


^ 


^ 


S 






^ 




^" 




a 


^ 


^ 




^ 


^ 


TTT 




^ 




jj^ 




:V: 









































^; 


















^ 




^ 


-^ / 


X:: 


:::: 


::- 


C1-1 


r.cj- 


w 1^ 




^ 


^ 


7^ 


^ 




/; 








:i-ir. 


ej-ar 


1/ 




7 




^ 












CI- 


s-.cj- 


^i" 






o 


JTER PILE ( 

\ 1 


X) 















SPACING "V OF OUTER PILES (IncllM) 

■MEII rU put CMMU nC IVlMMm or THE UNT LOW FROM 

EACHIUUL 



100 110 120 130 140 150 

SPACINO -b- OF OUTER PILES (inclMS) 

»MEII PU (Xa| CAMOEt THE REMAMOER or TME UMT LOAD FROM 
EACHRAL 



232 



Bulletin 761 — American Railway Engineering Association 



7 PILE BENT 



i t 




CHART LEGEND 

C2=24". C3=48" 

02=27". C3=51 

C2=27", C3=57 

C2=30". 03=54" 

C2=30", 03=60" 



AREA Pile Load Distribution Tables and Graphs 

7 PILE BENTS 
( 16 Charts total ) 

b = 120, 130, 132, 140, 144, ISO, 156, 160 & 168 inches 



Pile Cap 


Eff. Pile 
Length 


a 


Cl 


C2 


C3 


Table 


Cttart 


14" X 14" Timber 


10- 


39 


- 


24, 27, 27, 
30,30 


48.51.57. 
54.60 


Yes 


1 


60 


■ 


24, 27, 27, 
30,30 


48.51.57. 
54,60 


Yes 


2 


ac 


39 


- 


24, 27, 27, 
30.30 


48.51.57. 
54,60 


Yes 


3 


60 


- 


24, 27, 27, 
30,30 


48. 51. 57. 
54.60 


Yes 


4 


16" X 16" Timber 


Iff 


39 


- 


24, 27, 27. 
30,30 


48.51.57. 
54.60 


Yes 


5 


60 


- 


24, 27, 27, 
30,30 


48. 51. 57. 
54,60 


Yes 


6 


30- 


39 


- 


24. 27. 27, 
30,30 


«. 51, 57, 
54,60 


Yes 


7 


60 


- 


24. 27, 27, 
30.30 


48, 51. 57, 
54,60 


Yes 


8 


15" XI 5" Concrete 


Iff 


39 


- 


24. 27. 27. 
30.30 


48.51.57, 
54,60 


Yes 


9 


60 


- 


24, 27. 27. 
30.30 


48,51.57. 
54.60 


Yes 


10 


3ff 


39 


- 


24. 27. 27, 
30.30 


48.51,57. 
54.60 


Yes 


11 


60 


- 


24. 27. 27. 
30.30 


48.51.57. 
54.60 


Yes 


12 


15" XI 8" Concrete 


Iff 


39 


- 


24. 27. 27. 
30.30 


48.51.57. 
54.60 


Yes 


13 


60 


- 


24, 27, 27, 
30.30 


48.51.57, 
54.60 


Yes 


14 


3ff 


39 


- 


24, 27. 27. 
30.30 


48.51.57. 
54.60 


Yes 


15 


60 


- 


24, 27, 27. 
30.30 


48.51.57. 
54.60 


Yes 


16 



Proposed Manual Changes 



233 



CHART 1 



CHART 2 



7-PtLE BENT 

14" X 14" TIMBER CAP 

a-39" L-10' 



7-PILE BENT 

14" X 14" TIMBER CAP 

a«39" L-30' 















HTC 


mcou 


I 


I 

e(»i 




r 


... - 


--- 


--- 


-■- 

















■- 








— 


__. 


--- 


-- 


--- 





- — 


- 


a 04 












































►- 




IHT 


'MKa 


tTE n 


ip<e) 










— 


t 


^ 


-— 


— 


__- 


-- 


--- 





' 


- . — ' 


_. 


o 


--' 


" 


^^- 




■ 












P 


-;.:.■ 


.--• 






,..- 


-•-• 


-■■" 








2o, 




.--■■ 


-"^ 




















^ 


^ 


^ 






JUTfH 


PiKX 






0.0- 








"■^' 




^^ 


-■'y- 


sn. 


■■~iZ 


- 



































NT«MEOIATC ni£(»| 

, 1 A I-.. 






r 


-.:•.: 


• ■- 





:.-■ 




'.-. 


.'.'1 


.'_'.'! 


.'.r 


'-. 


ii; 


-— 


— 





— 





"■ 










& 04- 






































H 




IMTl 




TC nufXe) 










- 




^ 







__ 


- — 


— 


-- 


--- 


— 


~ 


^ 02- 


--"■ 


,., 


-'- 


•-; 


b-"^ 













2 ""* 
1 






^ 


..'■ 


---- 












r*g 


?^ 
























- ^r 


.~.*^ 


^ 


. OUTfR WIfpC) 


















-r 


— 


— 


0.0- 




"■~" 


2l 



SPACING -b- OF OUTER PtLEt (btchM) 

MCK PU PU| CAMBE* THi MCHiUNDOl or TIC UMT LOtO r 
EACH Ml. 



SPACING "b- OF OUTER PILES (InehM) 

MWa RLC (Xa) CiUnsS TME RCMAMDOI or -nc UMT LOAD FROM 
EACHRAI. 



CHART 3 

7-PILE BENT 

14" X 14" TIMBER CAP 

a»60" L-10' 

































WT 


DIMEOI 


ATE n 


JOb) 








3 °° 

^ 04. 









_.. 


.... 


— 


..._ 


— 


v.:: 


■- 


::- 


ri" 


:i: 


iz. 


Ill 





::."_ 


ZL" 


:i"_ 


-'■ 


z 






















s 




OERU 


IDIATE 


pupc 


4 











- 


? 








_-- 


-- 


--■ 


r.V 


r.-.r 


-..- 


■r.- 


S 02. 


---' 


-■- 


-r.r 


• ^.* 


'■' 


..-■ 


,-- 


— 


■" ' * 




e 
o 




,•-" 


.--' 
















^."^ 


^ 








ov 


W PtL 


EPO 












^ss? 


J",T.- 




rr: 


— - 


. 


, ^ 




0.0- 














'■^- 


t^-* 


r:.-, 


~ 



CHART 4 

7-PILE BENT 

14" X 14" TIMBER CAP 

a=60" L=30' 

































WTEW 


kCOIATI 


1 PHE 


O) 






























04 • 


:::: 


'"^' 


.'— - 


-■.T 


zJz 


L-J. 


li-i 


lH— " 


is. 


Z'. 




^^ 
























INT 


WMEd 


KTt n 


*P«:) 










_ 







— 


L- 


rr. 


r^ 


--• 


- 


^~' 


-■-: 


r.- 




:."-" 


-■' 


--: 


■'■' 


'-■- 


._■• 


-- 


-■■ 









c";^ 




'" 




















>^ 


^ 


^ 






OU' 


•ER PIL 


EOT 












■ 


'•■"^ 


'"ivl 


:"^ 


:rr: 


:-- 


— 


0.0- 




















•"^ 



SPACING -fa- OF OUTER PILES (InclM*) 

PMEX n.E PU) CARMEl THE REHAMDER or nC UMT UMO F 
EACHRAL 



120 130 140 150 160 

SPACING -b- OF OUTER PILES (Inchai) 

MWR FILE Ote) CARRSt T>« RSIAMOei OF TIC UMTT LOAD FROM 
EACHRAI. 



234 



Bulletin 761 — American Railway Engineering Association 



CHART 5 

7-PILE BENT 

16" X 16" TIMBER CAP 

a»39" L«10' 













IMTE 


RMCDW 


TI PttJ 


T 

Epa» 
















.... 














• - 


t°' 


::i- 


1'-- 


:'-'. 


;;:". 


-•- 


•-• 


-- 


-- 


— 


- 


. — 


















— 


■ 04- 














S 
< 






















l^ 


TERME 


51ATE 


•IU(Xe 


i 




- 






- 












^ _ 




--■ 








.... 


_. 


S 02- 


--■" 


_., 


,^- 


_.. 


-••^ 












.. 


2 "■■' 
o 




-•-■ 




_^. 


-- ■' 





' 










^ 


;^ 
























^^ 


?2^ 




■r- 


OUT 


ER PKJ 


m 




0.0- 














-^ 


.*7V 


5ru; 


^ 



CHART 6 

7-PILE BENT 

16" X 16- TIMBER CAP 

a-39" L-30* 



9 

O 0.3 



g 0.2- 



































1 1 1 

INTERMEOUTE PU |»l) 

1 1 1 















_.- 




.-■- 


r.-.r 


-■• 


— - 


■- 




'•-' 


ri- 


III 


■_r: 


:_~ 


"• 





- — 


-■ 






































INTERMEDIATE PILE (Xc) 
1 1 










— 










fc- 





— • 


_ 


-~ 





-• 




- ■"" 





"■ 


— 


' 






_.- 





- 




iii* 





--- 


• ■- 


— 
















^ 


= ?t 








OUTER PIL£(X) 
1 1 














"^-= 


1 


SK 


. 


— 

























SPACINO 'b' OF OUTER PILES (rnchn) 

MNER PILE (1U| CARRCI THE REMAMOER OF THE UNIT LOAD FROM 
EACHRML 



SPACINO -b- OF OUTER PILES (Inchtt) 

MNER PILE (Xl) URRC* THE RCIUMOai or im UMT L0«0 FROM 

EACHRML 



CHART 7 

7-PILE BENT 

16" X 16" TIMBER CAP 

a-60" L-IO" 

































INTI 


RMEOi 


ITE PU 


t(») 






















.... 








._ 




jjl' 


L'.- 




■^z: 


iiz 


1"!.'. 


ZIT- 


Zl' 











-■ 




























INTEf 


MEDIA 


I PIU 


(Xt) 











— 








— 


^ 


^ _ 


"_. 


"__. 


r'- 


v.r. 


-• 




^ . 


--- 


-■;-.: 


'■■■■ 




::,. 




-■" 


r" 






?-<; 


.-'• 


-'■" 
















0.1 • 




-•^ 


^?~ 








OUT 


ER PIL 


EM 












~r:S 


Tj-v: 


rr^ 


^r 


:rr 


— 


00- 




















2: 



CHART 8 

7-PILE BENT 

16" X 16" TIMBER CAP 

a-60" L"30' 

































INTE 


RMEDU 


TE PIL 


;(Xb) 






































'-.: 


-•.-. 


-.•.:: 


■-•_ 


::L- 


'J.'.'- 


I'.-. 


-"• 




l~~. 


ZS— 


— 1 




1 — 


























INTE 


IMEDIA 


TE PIU 


(Xc) 










_ 




^ :_ 








'_^. 


■;^^ 


'^■ 


' .v. 


-^~ 


-.-.- 


-- 




' — 


- - 


-~: 


k-.-- 





-•■ 


-•■- 


— 








^ 




— — . 






















'ia-r 


-::-: 


^ 


-^ 




OUTER 


PILEp 


) 














■>•.', 


="f=? 


^^.V^ 


^^^ 


^ 


0.0- 























SPAaNO "b- OF OUTER PILES (InelMs) 

MNER PU (Xa) CARRES THE REMAHDER OF THE UWr LOAD PROW 

EACHRAI. 



SPACING -b" OF OUTER PILES (InchU) 

*MER PILE |Xa| CARRES THE REMAMOER OF THE UMT LOAD FROM 

EACHRAL 



Proposed Manual Changes 



235 



CHART 9 

7-PILE BENT 

15" X 15" CONCRETE CAP 

a-39" L-iO' 













iwn 


JtMEOU 


ITi «l 


I — 




























1" 


-\'~ 


!'.'_ 


-l' 


.■-T. 


-"."1. 


•-• 


-■- 


--■-■; 


-•- 


-■ 


$ 


1^ 





















o 


















Z 

1 
9 
























INTE 


UHEDU 


TE ni 


ipte) 




1 






- 


I 
It. 

o 

* 3. 


— -^ 


_^ 




_^ _ 


- — 


'-■ 


' — " 





-•-. 


-: 




-'' 


.-- 




-■"T 




,., 


-•• 


-•- 




O 0-2 
1 


-■' 


• •-• 




..-■ 


U--' 












2 0, 


iT- 


"' 






















irj 






. 


oir 


rcH PI 


£(X) 






0.0- 








"" 




"-■: 


-r::c 


oI'Ti 


■V?\T 


— 



CHART 10 

7-PILE BENT 

15" X 15" CONCRETE CAP 

a»39" L«30* 





























1 1 1 

INTERMEDIATE PILEpOi) 

1 1 1 






























. . -■ 


-r.-. 


-•••-■ 


J.'— 


:•:•; 


'.'.'- 


ii- 


:L-J 


r; 


-•^J. 


=^ 


— -- 




■ 
















IMTCRMCDUTC PILE (Xc) 
1 1 
















rr: 


-^ 


■Z-. 


--:. 


— -_ 


"f-T. 


-■ 


-— 


-•- 


--/. 


'""7. 


_'.., 


'.-■■ 


■■- 


'■" 






Sp^ 


5;^ 


























CXfTER PILE(X) 
















■•~=^ 


"*''=rij 


r: 























SPACINO -b- OF OUTER PILES (inelMt) 

M«EK PLE PU) CAIOaet TMC REHAfOCK VnC UMT LOAD FROM 
EACH HAL 



SPACINO 'b- OF OUTER PILES (Inchli) 

MNER PU (X^ CAMUU THE REMAMOCR OP THE IMT LOAD PROM 
EACHRAA. 



CHART 11 



CHART 12 



7-PILE BENT 

15" X 15" CONCRETE CAP 

a«60" L-10' 



7-PILE BENT 

15" X 15" CONCRETE CAP 

a=60" L=30* 

































iffn 


:rmeo 


ATE PI 


E(>a» 








1°' 

III 






._. 








._. 


.... 


- - 


— 7 


■- 


,' 


-■- 


--■_ 


'. 


-• — 








'-S 


--"- 


-'■ 




— 




















i 




wn 


LRMEO 


ATE PI 


E(XC) 











- 


-1 

i 

* 03 . 






■ 


^ 


.-- 


--■ 


--- 


— "" 


::.'■ 


-■• 


--' 


'S.l 


-'- 


••-■. 


-•- 


.._■ 


--' 


'•■ 






U 0.^ 

K 





.,- 


-'- 


- - "^ 














2 0, 


\^ 


^r^ 






ou 


ER Pll 


Epq 
















-"'i: 


^.'^ 


rr: 


rrr 




k- 




0.0' 
















•^rr 


::r. 


^ 



O 0.3 

3 

K 

0-2 

s 

a. 
a. 

2 .. 

















































INT 


RMED 


ATE PI 


Jptt>) 



















-.--.■ 


r.:: 


r.--: 


-■.-.r 


•- 


I'^-l: 


r:!-. 


£5i 


^ 


--H^ 





— 






— 






INT 


ERMED 


ATE PI 


.Epic) 
















— 





^: 


r^rti 


— r. 


'-7.^. 


-• 


:.'.. 


T. ~ 


•:: 


:;;, 





— 


-•■" 








'K^ 


^ 


c^? 


-^ 




















~"~-'> 


-■~i:; 


' "■■•■] 




:rr 


. 










OUT 


ER PA. 


EPO 




^ * ■ 


--M 


^- 























SPACING -b- OF OUTER PILES (inclMil 

MNCR PILE IXaI CARR» THE RIMAMOCR Of TME UMT LOAD FROM 
EACHRAL 



SPACING -b- OF OUTER PILES (Inclwi) 

HNER PU PU) CARRIE< THE REMAINDER or THE UMT LOAD FROM 
EACH RAIL 



236 



Bulletin 761 — American Railway Engineering Association 



CHART 13 

7-PILE BENT 

15" X 18" CONCRETE CAP 

a«39" L-10" 



i 0.4 

X 

ui 

t 

a 

cj 0.3 































IHTI 


nmsu 


TE TO. 


!PQ>) 








..-- 


-.'■. 


-■■. 


-.■.■.: 


■"'. 


".'.'.7 








•~ 


:_- 


Z'-- 


:i: 


--• 


— 













— 
























NT 


JUEO 


ATT ri 


£(Xe) 











— 






^ 


— 


-;;;^ 


--• 




.. 


-•-. 


-: 


"' 


— "" ' 


.'- 


•-; 


"■ _ 




,..- 


■'•• 







■•■■ 




.-- 


.-- 


'■' 












=^? 








ou- 


'ER Ft 


ipo 












-^^ 


T^x- 




:::: 


. 




















"■--■ 




=':K] 


^ 



CHART 14 

7-PILE BENT 

15" X 18" CONCRETE CAP 

a-39" L-30' 















































& 04. 










INTERMEDIATC n 
1 1 


LEpm) 






















— •.: 


-•.-.: 


-.7 


z 

% 


si 


'^--\ 


;i"i 


^ 


E^ 


'.'_. 


; — "^ 


^^ 





— 


""III 

INTEKMEOIATE PIU (Xc) 
1 1 










§ 






rrt 


rzt 


--- 


rT: 


zrj" 


r.': 


^r.j'. 


-: 


K;^^ 


':'.':. 


[•': 


I'.'.' 


:::: 


■-• 











p 


^^ 


•^ 


^^5^^ 
























■*"~'".-: 


-■~c 


^Kfe 


— 














OUTm PllEpt) 

1 






0.0- 























SPACING ■%" OF OUTER PILES (InchM) 

MNEK PU (X<| aUOOU THE RClUMm OF THE UWr UMO FROM 
EACHRAI. 



SPACINO 'b' OF OUTER PBXS (InchM) 

VMER PU OU) CARIUES TME REMAMOER W THE UNH' LOAD FROM 
EACH RAIL 



CHART 15 

7-PILE BENT 

15" X 18" CONCRETE CAP 

a»60" L-10' 

































INTE 


RMEDU 


T£ Pit 


EPO)) 








3" 
























-::.-. 


• --" 


"ir".' 


.".-- 


■i~- 


sl. 


11.1 


~-~ 


:i--- 


r: 


Z 

I 


-— — 








~~ 
















INTE 


IMEOl* 


TE ML 


:(Xe) 








, 


t 










:r:: 


' ^ 


--- 


•'"_ 


-.-. 


--■ 


' — 


-•- 


'^^- 


■""■ 


•-•- 


._•■ 


-' 


-•• 


"■' 






"-^ 




--' 
















I 




~-'V- 


j^; 


^^ 


^_^ 




OUT 


ER Pll 


(X) 














■•"^ 


^iv; 


:^ 




-__ 


_ 


0.0 




















- 



SPACING -b" OF OUTER PLES (InchM) 

WNCR Pt^ pm CARRE* THE REIIUUNOai OF THE UNIT UMD FROM 
EACHRAI. 



CHART 16 

7-PILE BENT 

15" X 18" CONCRETE CAP 

a=60" L»30* 



9 , 

q 0.3- 



































































INTE 


RMEOU 


Tl PIL 


i(Xb) 












...--• 


-•": 


-v.; 


'■."-' 


'}^S. 


-y-i 


-- 


i-"^ 


=-: 


" ""^ 








INTE 


IMEOV 


TE PIL 


(Xe) 




^ 











E^r 


r^:- 


zr3 


n:r- 


^r*-^- 


^ 




::.■■ 


-••" 


-••- 





"" 












^ul: 


55; 


^ 


















OUTl 


R PILE 


(X) 


''*=■ 


*?V 


=~^ 


r: 











































SPACING "b" OF OUTER PILES (InchM) 
MNER PU |Xa) CARRU THE REMAINOCR OF TMC UMT LOAD FROM 

EACHRAK. 



Change 50 ties an hour 
under heavy traffic 




Twenty trains per day use this track. Yet despite this heavy 
traffic, a single MRT-2 Tie Changer can replace 50 ties an 
hour in an average day. The secret is the quick on-off track- 
ability designed into the MRT-2. In less than two minutes, 
any place along the line, it can climb on or off track com- 
pletely under its own power. Old ties - even switch ties - are 
renioved whole, with minimal disturbance to track structure. 

If you're now wasting valuable time clearing for trains, switch 
to the MRT-2. Sales and service available throughout North 
America. Contact us for a free demonstration. 



MODERN TRACK MACHINERY INC. 

1415 Davis Rd., Elgin, IL 60123-1375 

Tel. (847) 697-751 Fax: (847) 697-01 36 




©EISM 




MODERN TRACK MACHINERY CANADA LTD. 

5926 Shawson Drive, Mississauga, Ontario L4W 3W5 

Tel. (905) 564-1211 Fax: (905) 564-1217 



237 



238 



Bulletin 761 — American Railway Engineering Association 




Proposed 1998 Manual Revisions to 
Chapter 7 — Timber Structures 

Part 2 (Continued) 

Page 7-2-73. Immediately following Table 2-13, add the following new Tables "Comparison of Unit 
Stresses in Timber Trestles for Cooper E 80 Loading, No Impact," charts for open and ballast deck 
trestles. 



Proposed Manual Changes 



239 



W 



Area Manual for Railway Engineering - Timber Structures 



Assumptions: 

L = Distance C to C bents for bearing on caps In feel 
= Distance face to face of caps plus .5 ft for stringer bending 
and shear. (Assume 14" cap) 
W = Total Dead Load per lin. ft of track: 
Rail and fastenings = 200 pounds per lin. fL 
Tie 8"x8"xlO' @ 12" ctrss = 267 pounds per lin. ft 
R = Total Reaction. 



h = height of stringer in feel 

b = breadth of stringer in feet 

P = weight on one driving axle = 800(K) pounds 

a = distance from load P to support in feet 

Dressed size = Nominal size less 1/2" in depth 

In Calculating bearing, bending, and shear stresses, outer 

stringers are considered as carrying no load. 



240 



Bulletin 761 — American Railway Engineering Association 



u 



il 



11 



fS 



s ? » 



i e = 

If 
If 



s t 



la 






ll 



Area Manual for Railway Engineering - Timber Structures 



Assumptions: 

L = Distance C to C bents for bearing on caps in feet 
= Distance face to face of caps plus .5 fL for stringer bending 
and shear. (Assume 14" cap) 
W = Total Dead Load per lin. ft. of track: 
Rail and fastenings = 200 pounds per lin. ft. 
Ballast = 120 pounds per cu. ft. 
R = Total Reaction. 



h = height of stringer in feet 

b = breadth of stringer in feet 

P = weight on one driving axle = 80000 pounds 

a = distance from load P to support in feet 

Dressed size = Nominal size less 1/2" in depth 

In Calculating bearing, bending, and shear stresses, outer 

stringers are considered as carrying no load. 



Experience has proved you can depend on them. 
WESTERN-CULLEN-HAYES 

Railroad Products' 






For Maintenance-Of-Way 

Hayes Bumping Posts 
Delectric® Operators 
Hayes Derails 



For Signaling 

Crossing Signals 
Gate Mechanisms 
Case & Track Hardware 



For Communications 

Telephone Shelter Boxes 
Lightning Arresters 




WESTERN-CULLEN-HAYES, Inc. 

2700 West 36th Place • Chicago, Illinois 60632 
Telephone: 773/254-9600 
Web site: www.wch.com 



Take advantage of Western-Cullen-Hayes 
service proven equipment and experi- 
enced railway supply personnel to assure 
safe, efficient operation on your line or 
in your yards. Call 773/254-9600. 



DANELLA RENTAL SYSTEMS, INC. 

Helping Railroaders and Transit Authorities 

NATIONWIDE! 



' Quality Equipment 
' Courteous Service 
' Reasonable Rates 




• Hy-Rail Available 

• Short Term 

• Long Term 

• Rent to Own 



We offer every type of late model light, medium or heavy duty vehicle and 
construction equipment for railroads and transit authorities including: 



Piclcups 


• Rotary Dumps 


• Compressors 


• Dozers 


Boom Trucks 


• Tractors 


• Backhoes 


• On Track Cable Plow 


Flatbeds 


• Low Boys 


• Front End Loaders 


• Trenchers 



DANELLA RENTAL SYSTEMS, INC. 

2290 Butler Pike 14101 East Moncrieff Place 

Plymouth Meeting, PA 19462 |u|__i-__ f^of* __ j DCMCA '^"'O'^' CO 80011 

Phone (610) 828-6200 MemDCr JJf.Jj^'' ^iJO "an ''*'°"^ l^"^* ^^^-"SS 

Fax (610) 828-2260 (204)782-5456 Fax (303) 371-2677 



241 




Connect with 

Parker for 
all your 
hose and 

fittings needs. 



For your air brake applications, fuel 
and oil lines or hydraulic system 
connections, Parker has the hose 
and fittings you need. 

Including skive-type hose and fittings 
or no-skive hose with permanent or 
reuseable fittings. 

Or our patented Parkrimp system 
that allows you to make 
permanent, factory- 
quality hose 
assemblies in \ 
own facilities o 
on the job site. 

Call or 
write today 
for your free 
copies of 
Parker 
railroad 

product bulletins. 
Parker Hannifin Corporation, 
Hose Products Division, 
30240 Lakeland Blvd 
Wickliffe, OH 44092. 
(216)943-5700 
FAX (216) 943-3129 




Parker 



FluidConnectors 



242 



One complete service. 
Lowest cost per mile. 




* A complete, objective test 
of each rail from end to end. 

ir Simultaneous ultrasonic and 
induction detection methods. 

*Sperry far surpasses every other 
rail testing service in efficiency, 
thoroughness and research. 

*One mileage charge pays 
for everything. 

*The lowest real cost per mile 
and per defect found. 

Details and technical assistance on request. 



SPERRY RAIL SERVICE 

SHELTER ROCK ROAD DANBURY, CONNECTICUT 06810 
(203) 791-4500 



243 



Proposed 1998 Manual Revisions to 
Chapter 9-Seismic Design for Railway Structures 

Section 1 .5 Existing Structures 

Page 9-1-22. Insert the following language at the beginning of Article 1.5.7.5 Retrofit Designs. 

Railroads may decide to retrofit bridges to protect against human casualty resulting from col- 
lapse in an earthquake, or to expedite restoration to service following an earthquake. The following 
factors should be considered in any retrofit design: 

1 . Retrofit design must be site specific and must consider the condition and stability of the exist- 
ing structure, including soils and foundation. 

2. Attachments of substructure to superstructure must permit normal movement of the structure. 

3. Behavior of the retrofit system shall not cause damage to the primary structure which would 
preclude promptly returning the structure to service after a seismic event. 

4. Retrofits must permit both routine and post-seismic inspection, repair and component 
replacement. 

C-Section 1.3 Basic Concepts and Nomenclature 

Page 9-2-5. Insert the following addition to the Commentary on Post Seismic Event Operation 
Guidelines. 

C-Section 1.5 Post Seismic Event Operation Guidelines 

1.5.1 Scope (1998) 

The post-seismic event operation guidelines are intended for use where experience or adequate 
knowledge of regional attenuation rates is not available. They provide the basis for a policy for areas 
where attenuation rates are relatively high, such as California. A more conservative policy would be 
appropriate in areas where seismic experience is limited and/or attenuation rates are relatively low. 
These conditions exist in most of central and eastern North America. Where justified by adequate 
experience and/or analysis, a less conservative policy may be appropriate. 

1.5.2 Presumption of Seismic Resistant Design (1998) 

It may be desirable to have an arrangement with a technically and legally qualified engineer to 
inspect essential buildings immediately after an earthquake so that their safety can be determined and 
certified to avoid unnecessary evacuations and/or restrictions on building use. Essential buildings 
would include, among others, dispatching centers, yardmaster's towers, shop facilities, fueling facil- 
ities, buildings containing certain communications facilities and, for lines with commuter service, 
passenger stations. 

1.5.2.1 Track Structure 

Although the track structure, with the possible exception of the ballast, is rarely affected by 
shaking, distortion of the underlying ground may severely impact track geometry. Longitudinal dis- 
tortion can cau.se track buckling, or high tensile stresses in the rails with the resulting risk of pull 
aparts. Lateral movements and/or settlement due to liquefaction or embankment failure can cause 
.serious defects in line, surface and cross level. 



244 



Proposed Manual Changes 



245 



1.5.2.2 Other Structures, Buildings and Facilities 

Staictures located near the fault rupture are likely to suffer serious damage in a major earth- 
quake. Safety of operation ultimately depends on post-event inspection of facilities in areas subjected 
to major ground movements and/or severe shaking. Proper design can reduce, but not totally elimi- 
nate the probability of significant damage. 

1.53 Inventory (1998) 

Most railroads have a good inventory of their own facilities. However, in several earthquakes 
where damage to railroad facilities was minor, structures owned by others, including structures over 
or adjacent to tracks, have collapsed. The presence of any structures whose collapse could adversely 
affect operations should be determined and recorded. Underground structures subject to seismic fail- 
ure and buried utilities, including pipelines, should also be identified. 

1.5.4 History (1998) 

Areas with frequent significant seismic activity are more appropriate for historical analysis than 
areas that have rare, but severe, earthquakes, such as parts of central and eastern North America. 

1.5.6.1 Tbnnels 

Tunnels usually are subjected to less severe loading from earthquakes than structures on the sur- 
face of the ground. However, they have been damaged by shaking and .severely damaged by dis- 
placements at locations where they were intersected by fault ruptures. 





(ATmON KIPS YOU 

aiNMiiutfm 

MUNDSItfETl 

— wm{\ 



W7SX 



S 




USA: Tel: 412/962-3571 • Fax: 412/962-4310 
CANADA: Tel: 905/873-9440 • Fax: 905/873-9449 
EUROPE: Tel: +44(0)1932-247511 • Fax: +44(0)1932-220937 
CATTRON '" Inio. via World Wide Web: www.cattron.com 




FRH BROCHURE Leorn how CATTRON' 



improve your boltom line. 



CJontrctl S\rstGxins 



246 Bulletin 761 — American Railway Engineering Association 



Proposed 1998 Manual Revisions to 
Chapter 15-Steel Structures 

Part 1 -Design 

Page 15-1-48. Substitute the following for Article 1.7.3. Thickness of Web Plates. 
1.7.3 Thickness of Web Plates 

a. The thickness of the webs of plate girders without longitudinal stiffeners shall not be less 
than: 



30,500 



of the clear distance between the flanges, except that if the extreme fiber stress in the com- 
pression flange is less than that allowable, the denominator 30,500 may be multiplied by the factor: 

where: 

P^ = allowable stress in the compression flange, as determined by the applicable formula of 

Article 1.4.1, psi 
/= the actual extreme fiber stress in the compression flange, psi 
F^. - yield point as specified in Table 1-1 for the material 

b. The thickness of the webs of plate girders with longitudinal stiffeners, proportioned in accor- 
dance with Article 1.7.8, shall not be less than 1/2 that determined in paragraph a. 

c. The thickness of the webs of plate girders with or without longitudinal stiffeners shall not be 
less than 1/6 the thickness of the flange. 



Page 15-1-50. Substitute the following for Article 1.7.8 Intermediate Stiffeners. 

1.7.8 Web Plate Stiffeners (Intermediate Transverse and Longitudinal) 

a. Where the depth of the web between the flanges or side plates of a riveted, bolted or welded 
plate girder exceeds 1 1400/ Vf^ times its thickness, it shall be transversely stiffened by pairs (except 
as noted in paragraph c) of angles riveted or bolted, or of plates welded, to the web. The clear dis- 
tance, d, between intermediate transverse stiffeners shall not exceed 96 inches, nor the clear distance 
between the flanges or side plates, nor that given by the formula: 

, 10500r 

where d = clear distance between intermediate transverse stiffeners, inch 
t - thickness of web, inch 
S - calculated shear stress in the gross section of the web at the point under 

consideration, psi 
F^ - yield point as specified in Table 1-1 for the web material, psi 



Proposed Manual Changes 247 



The moment of inertia of the intermediate transverse stiffeners shall not be less than: 

/ = 2.5^,. rl-^-0.7 

taken about the centerline of the web plate in the case of stiffeners furnished in pairs (on each side of 
web plate) and taken about the face of the web plate in contact with the stiffener in the case of sin- 
gle stiffeners. 

where I = moment of inertia, /Vu/rrs"* 

d . = actual clear distance between intermediate transverse stiffeners, inch 
D = depth of web between flanges or side plates, inch 

b. For intermediate transverse stiffeners, the width of the outstanding leg of each angle, or the 
width of the welded stiffener plate, shall not be more than 16 times its thickness nor less than 2 inches 
plus 1/30 of the depth of the girder. 

c. Intermediate transverse stiffeners used on one side of the web plate only (single stiffeners) 
shall be connected to the outstanding portion of the compression flange. 

d. Where the depth of the web between the flanges or side plates of a riveted, bolted or welded 
plate girder exceeds 22500/v7^times its thickness (where / - the calculated compressive bending 
stress in the flange, psi), it shall be stiffened by intermediate transverse stiffeners in accordance with 
paragraphs a, b and c; and by a longitudinal stiffener. Longitudinal stiffeners may be discontinuous 
at their intersections with intermediate transverse stiffeners. Longitudinal stiffeners are usually 
placed on one side of the web plate. The stress in the stiffener (from participation in the girder stress) 
shall not be greater than the basic allowable bending stress for the material used in the stiffener. 

e. The centerline of a plate longitudinal stiffener or the gage line of an angle longitudinal stiff- 
ener shall be D/5 from the inner surface or leg of the compression flange component. 

f. The longitudinal stiffener shall be proportioned so that: 

.2 

Ir = Dt'\ 2.4-^-0.13 

where I^ = minimum required moment of inertia of longitudinal stiffeners about the edge in 
contact with the web plate, inches'*, for stiffener used on one side of the web and 
about the centerline of the web plate for stiffeners used on each side of the web 



g. The thickness of the longitudinal stiffener shall not be less than: 

b'4f 
2250 

where b' = width of outstanding leg of longitudinal stiffener, inch 
/= calculated compressive bending stress in the flange, psi 



248 Bulletin 761 — American Railway Engineering Association 

Part 9-Commentary 

Page 15-9-24. Substitute the following for Article 9.1 .7.3 Thickness of Web Plates. 
9.1.73 Thickness of Web Plates 

a. The specified thickness of web plates for flexural members is based on work done by Hovey 
(Bibliography 32). Hovey showed that the buckling of the web of a flexural member on the com- 
pression side of the neutral axis can be prevented either by the use of horizontal (longitudinal) stiff- 
eners or by making the web of such thickness that stability against buckling is ensured. Vertical 
(intermediate transverse) stiffeners are not effective in resisting buckling caused by bending. 
Assuming the actual extreme fiber stress in the compression flange is 0.55F , and that the compres- 
sion stress in the web adjacent to the flange is less than this by an assumed percentage, the ratio of 
the thickness of the web to the clear distance between flanges, for a web without horizontal stiffen- 
ers, to ensure the stability of the web against flexural buckling may be expressed by the formula 
V^/30, 500. 

Where the extreme fiber stress in compression is less than the allowable, then the ratio may be 
modified as specified. 

b. For web plates stiffened by a horizontal (longitudinal) stiffener located at 0.20 of the web 
depth from the compression flange, work by Rockey and Leggett (Bibliography 87) has shown that, 
to ensure the stability of the web against flexural buckling, the web plate thickness required is only 
43% of that required without a horizontal stiffener. It is specified that the web plate thickness shall 
not be less than 1/2 that determined for a web plate without a horizontal stiffener. 

Page 15-9-25. Substitute the following article 9.1.7.8 Intermediate Stiffeners 

9.1.7.8 Web Plate Stiffeners (Intermediate Transverse and Longitudinal) 

a. Hovey showed that the ratio of web clear depth to thickness for which stiffeners are not needed 
is determined by the formula VA^SiE/F^, where F^^ , is the yield point in shear of the web material 
(Bibliography 32). With F^^ = 0.636 /=;, the formula became 14,800/"/^;, The formula 1I,400/V^, 
used in Art. 1.7.8, makes allowance for lack of flatness in the web plate. 

b. Where stiffeners are required, their spacing is dependent on the web thickness and the shear- 
ing stress in the web. The development of the specification formula is based on work by Moisseiff 
and Leinhard and is based on a factor of safety of 1.5 against buckling of the web (Bibliography 40). 
This factor of safety is lower than the basic factor of safety generally used throughout these specifi- 
cations, but is considered adequate because elastic buckling of the web does not cause failure. When 
elastic buckling of the web occurs, its share of additional diagonal compression is transferred to the 
flange and vertical (intermediate transverse) stiffeners. The specification requirements for size of 
stiffeners meet the requirements developed by Bleich (Bibliography 8). 

c. The 96 inches maximum spacing of the stiffeners is specified in order to provide stiffeners at 
reasonably close intervals so as to aid in eliminating the effect of any small out-of-flatness that may 
exist in the web. The 96 inch maximum spacing is based on work done by Basler indicating that for 
fabrication, handling and erection purposes the maximum stiffener spacing should not exceed 260^, 
where / is the web thickness in inches (Bibliography 86). The distance between vertical (intermedi- 
ate transverse) stiffeners shall not exceed the distance between the flanges (web depth) because the 
formula for stiffener spacing, 10,500f/VX is developed from the theory of elastic stability with this 
assumption (i.e. critical buckling coefficient in shear always less than 9.35). 



Proposed Manual Changes 



249 



d. The web plate depth criteria of (22500/ V/Tr specified in paragraph d) relates to the web plate 
depth required to preclude flexural elastic buckling of the web plate without horizontal (longitudinal) 
stiffeners. 

e. The specification that the centeriine of the longitudinal stiffener be placed at 1/5 the web depth 
from the compression flange is from work by Rockey and Leggett (Bibliography 87) showing this to 
be near optimum location to resist flexural buckling of a simply supported plate. 

f. The specification for longitudinal stiffener size is taken as a reasonable upper bound for gird- 
ers of practical proportions based on the work by Dubas (Bibliography 88). 

g. The specification for the thickness requirements for longitudinal stiffeners is based on the 
local buckling behaviour of the stem of a tee section. 

Page 15-B Bibliography Additions in Conjunction with Proposed Revisions to Commentary Articles 
9.1.7.3 and 9.1.7.8 

86. Easier, K., New Provisions for Plate Girder Design, Proceedings, A ISC National 
Engineering Conference, New York, 1961, pages 65-74. 

87. Rockey, K.C., and Leggett, D.M.A., The Buckling of a Plate Girder Web under Pure 
Bending When Reinforced by a Single Longitudinal Stiffener, Proc. Inst. Civil Eng., Vol. 21, 1962. 

88. Dubas, C, A Contribution to the Buckling of Stiffened Plates, lABSE 3rd Congress 
Preliminary Publication, Liege, 1948. 



Specialists in Iceeping 
Bridges in service 

•TIMBER •STEEL •CONCRETE 

Over 40 years of railroad experience. 
Inspect. . . Repair. . . Treat. . . Strengthen 




RAILROAD DIVISION 
P O Box 8276 • Madison, Wisconsin 53708 
608/221-2292 • 800/356-5952 



WE'RE THE BRIDGE PRESERVERS 




Will the next rail you buy 
be fully heat-treated, 
head-hardened, or inter- 
mediate strength? 

Will the next turnouts 
you buy be state-of-the- 
art manganese castings, 
vacuum -molded and 
machined for perfect fit? 

The answer is yes, if 
you're out for the best rail 
products the world has to 
offer. And that means 
Foster-Class, from L.B. 
Foster Company. 
World-class. 

Well go aaoss the coun 
try or around the world 
to meet today's standards. 



So you get a double 
advantage: world- 
class technology along 
with superior Foster fin- 
ishing and Foster servic- 
ing right here at home. 

For instance, Foster 
supplied turnouts meet all 
AREA specs, and every 
inch is pre -inspected 
before shipment. 

We go to special lengths 
on relay rail, too. Just as 
we've been doing for 80 
years, we bring you the 
largest stocks in the world. 
And more. Today we take 
up and deliver pre -welded 
lengths up to a quarter- of 



a-mile to cut your 
on-site fabrication costs. 
Go Foster-Class 

for your tallest or 
smallest orders. 

Give us a call and we'll 
ship any rail order — 
including turnouts and 
accessories — on time, 
anywhere, from stocking 
points coast to coast. Plus 
special sections and long 
lengths of new rail, rolled 
to order. 

We're also your 
number one source for 
sophisticated track and 
contact rail components 
for transit systems. 



The Foster difference 
is a world of difference. 
Because Foster- Class is 
world-class. Phone or 
write L.B. Foster Com- 
pany, 415 Holiday Drive, 
Pittsburgh, PA 15220. 
(412) 928-3400. 



FOSTER 



L.B.FOSTER 
COMPANY 



250 



Proposed Manual Changes 25 1 



Proposed 1998 Manual Revisions to 
Chapter 19 — Bridge Bearings 

Page 19-1 Table of Contents - Delete the existing Chapter 19 Table of Contents for Part 2 and 
replace with the following: 

2 Construction 19-2-1 

2.1 Introduction 19-2-3 

2.2 Steel Bearing Components 19-2-4 

2.3 Bronze or Copper-Alloy Sliding Expansion Bearings . . (Under Development). . . 19-2-7 

2.4 TFE Bearing Surface (Under Development). . . 19-2-7 

2.5 Elastomeric Bearings 19-2-7 

2.6 Multi-Rotational Bearings (Under Development). . 19-2-12 

2.7 Seismic Isolation Bearings and Devices (Under Development). . 19-2-12 

Part 2 - Construction 

Page 19-2-1. Delete the existing Part 2 in its entirety and replace with the following. 

AMERICAN RAILWAY ENGINEERING ASSOCIATION AREA 



Part 2 
Construction^ 



— 1998 — 
FOREWORD 

The purpose of this part is to formulate specific and detailed requirements for the construction of 
bearings for nonmovable railway bridges. 

TABLE OF CONTENTS 

Section/Article Description Page 

2.1 Introduction 19-2-3 

2.1.1 General 19-2-3 

2.1.2 Shop Drawings 19-2-3 

2.1.3 Packaging, Handling and Storage 19-2-3 

2.1.4 Manufacture or Fabrication 19-2-3 

2.1.5 Construction and Installation 19-2-3 

2.2 Steel Bearing Components 19-2-4 

2.2.1 General 19-2-4 

2.2.2 Shoes and Pedestals 19-2-4 

2.2.3 Rockers, Rollers and Sliding Bearings 19-2-4 

2.2.4 Sole, Base and Masonry Plates 19-2-5 

2.2.5 Inclined Bearings (Under Development) 19-2-6 

2.2.6 Anchor Bolts 19-2-6 

2.3 Bronze or Copper-Alloy Sliding Expansion Bearings . . . (Under Development) 19-2-7 



References. Vol. 98, p. 251. 



252 Bulletin 761 — American Railway Engineering Association 



TABLE OF CONTENTS (CONT) 

Section/Article Description Page 

2.4 TFE Bearing Surface (Under Development) 19-2-7 

2.5 Elastomeric Bearings 19-2-7 

2.5.1 General 19-2-7 

2.5.2 Materials 19-2-7 

2.5.3 Plain Elastomeric Bearings 19-2-8 

2.5.4 Reinforced Elastomeric Bearings 19-2-8 

2.5.5 External Steel Load Plates 19-2-8 

2.5.6 Tolerances 19-2-8 

2.5.7 Marking 19-2-9 

2.5.8 Acceptance Criteria 19-2-10 

2.5.9 Test Criteria I 19-2-10 

2.5.10 Test Criteria II 19-2-11 

2.5.11 Certification 19-2-11 

2.5.12 Installation 19-2-11 

2.6 Multi-Rotational Bearings (Under Development) 19-2-12 

2.6.1 Pot Bearings (Under Development) 19-2-12 

2.6.2 Disc Bearings (Under Development) 19-2-12 

2.6.3 Spherical Bearings (Under Development) 19-2-12 

2.7 Seismic Isolation Bearings and Devices (Under Development) 19-2-12 

SECTION 2.1 INTRODUCTION 

2.1.1 GENERAL 

a. The work covered by this Part consists of furnishing and installing bridge bearings and 
bridge bearing components including shim plates, anchor bolts, lubricants, adhesives and the bedding 
materials used under masonry plates. 

b. Bearings shall be constructed as specified and in accordance with the details shown on the 
plans. Whenever complete details for bearings and their anchorages are not shown on the plans, bear- 
ings shall be furnished to conform with the limited details shown on the plans and shall provide the 
design capacities for loads and movements shown or specified and the performance characteristics 
specified. 

2.1.2 SHOP DRAWINGS 

The Contractor shall prepare and submit shop drawings for bridge bearings in accordance with 
Section 1.1 of AREA Manual Chapter 15. Such shop drawings shall show complete details of the 
bearings and of the materials proposed for use and must be reviewed by the Engineer. The Engineer's 
written approval of shop drawings must be received before fabrication of the bearings is begun. 

2.1.3 PACKAGING, HANDLING AND STORAGE 

a. Prior to shipment from the point of manufacture, bearings shall be packaged in such a man- 
ner that during shipment and storage the bearings will be protected against damage from handling, 
weather, or any normal hazard. Each completed bearing shall have its components clearly identified, 
be securely bolted, strapped or otherwise fastened to prevent any relative movement, and marked on 
its top as to location and orientation in each structure in the project in conformity with the plans and 
approved shop drawings. Dismantling at the site is not permitted unless absolutely necessary for 
inspection or installation if directed by the Engineer. 



Proposed Manual Changes 253 



b. Bearing devices and components shall be stored at the work site in an area that provides pro- 
tection from environmental and physical damage. When installed, bearings shall be clean and free of 
all foreign substances. 

2.1.4 MANUFACTURE OR FABRICATION 

a. Bearing devices or assemblies shall consi.st of components meeting the material require- 
ments of Part I of this Chapter. 

b. Unless otherwise stipulated, dimensional tolerances of bearing components shall conform to 
the requirements of Article 3.1.7 of AREA Manual Chapter 15. 

c. Bearing assemblies shall be pre-assembled in the shop by the supplier and checked for com- 
pleteness and geometry before shipping to the site. 

2.1.5 CONSTRUCTION AND INSTALLATION 

a. Bearings shall be installed by qualified personnel in the positions shown on the plans. 
Bearings shall be set at time of installation to the dimensions prescribed by the manufacturer, the 
Engineer, or as shown on the plans and adjusted as necessary to take into account the temperature 
and future movements of the bridge. 

b. Bearings shall be set level, to the alignment and elevations established by the Engineer, and 
must have full and even bearing on all bearing planes. 

c. Bearing surfaces located at improper elevations or set not level and true to plane shall require 
either grinding of the surface, grouting of the bearing seats or modification of the bearing such that 
intended bearing placement is as originally designed with the least amount of bearing modification, 

d. Whenever bearings are designed by the Manufacturer and/or a Manufacturer's Warranty is 
required by the Contract, installation shall be performed under the Manufacturer's supervision. 



SECTION 2.2 STEEL BEARING COMPONENTS 

2.2.1 GENERAL 

The surface finish of bearing and base plates and other bearing surfaces that are to be in contact shall 
have a maximum surface roughness value less than or equal to 125 (ANSI B46.I, Surface Texture). 

2.2.2 SHOES AND PEDESTALS 

2.2.2.1 Materials 

Steel used in shoes or pedestals shall be of the types and grades shown on the plans or otherwise spec- 
ified. 

2.2.2.2 Fabrication 

a. Bearing surfaces of cast pedestals that are to be in contact with steel or masonry shall be 
planed. 

b. Structural members which are indicated in the contract drawings or specifications to be 
annealed or normalized shall have finished machining, boring, and straightening done subsequent to 
heat treatment. Normalizing and annealing (full annealing) shall be as specified in ASTM A919. The 
temperatures shall be maintained uniformly throughout the furnace during the heating and cooling so 
that the temperature at no two points on the member will differ by more than 38°C (100°F) at any 
one time. 

c. Members of Grades 690/690W (100/lOOW) or Grade 485W (70W) steels shall not be 
annealed or normalized and shall be stress relieved only with the approval of the Engineer. 



254 Bulletin 761 — American Railway Engineering Association 



d. A record of each furnace charge shall identify the pieces in the charge and show the tem- 
peratures and schedule actually used. Proper instruments, including recording pyrometers, shall be 
provided for determining at any time the temperatures of members in the furnace. The records of the 
treatment operation shall be available to and meet the approval of the Engineer. The holding temper- 
ature for stress relieving Grades 690/690W (100/lOOW) and Grade 485W (70W) steels shall not 
exceed 610°C (1 125°F) and 580°C (I075°F) respectively. 

e. When called for by the contract plans or specifications, members such as bridge shoes, 
pedestals, or other parts that are built up by welding sections of plate together shall be stress relieved 
in accordance with the requirements of Section 4.4 of ANSI/AASHTO/AWS Bridge Welding Code 
D1.5. 

2.2.3 ROCKERS, ROLLERS AND SLIDING BEARINGS 

2.2.3.1 Materials 

Steel used in rocker, roller and sliding bearings or bearing components shall be of the types and 
grades shown on the plans or otherwise specified. 

2.2.3.2 Fabrication 

a. Burrs, rough and sharp edges, and other flaws shall be removed. 

b. Pins and rollers shall be accurately turned to the dimensions shown on the drawings and 
shall be straight, smooth, and free from flaws. Pins and rollers more than 230 (9 in.) in diameter shall 
be forged and annealed. Pins and rollers 230 (9 in.) or less in diameter may be either forged and 
annealed or cold-finished carbon-steel shafting. 

c. In pins larger than 230 (9 in.) in diameter, a hole not less than 50 (2 in.) in diameter shall be 
bored full length along the axis after the forging has been allowed to cool to a temperature below the 
critical range, under suitable conditions to prevent damage by too rapid cooling. The hole shall be 
bored before the pin is annealed. 

d. Pin holes shall be bored true to the specified diameter, smooth and straight, at right angles 
with the axis of the member and parallel with each other unless otherwise required. The final surface 
shall be produced by a finishing cut. 

e. The diameter of the pin hole shall not exceed that of the pin by more than 0.50 (1/50 in.) for 
pins 130 (5 in.) or less in diameter, or by 0.80 (1/32 in.) for larger pins. 

f. The distance outside to outside of end holes in tension members and inside to inside of end 
holes in compression members shall not vary from that specified by more than 0.80 (1/32 in.). Boring 
of pin holes in built-up members shall be done after the member has been assembled. 

2.23.3 Installation 

a. Setting of rocker, roller and sliding bearings shall take into account any variation from mean 
temperature of the supported span at time of setting and any other anticipated changes in length of 
the supported span. At mean temperature, after release of falsework and any shortening due to pre- 
stressing forces, the rockers and rollers shall be vertical or the sliding components shall be in proper 
alignment. Care shall be taken that full and free movement of the superstructure at movable bearings 
is not restricted by improper settings or adjustment of bearings. 

b. The Contractor shall coat contact surfaces thoroughly with oil and graphite just before plac- 
ing roller bearings. 

c. Cylindrical bearings shall be carefully positioned so that their axes of rotation are in align- 
ment and coincide with the axis of rotation of the superstructure. 



Proposed Manual Changes 255 



2.2.4 SOLE, BASE AND MASONRY PLATES 

2.2.4.1 Materials 

Steel plates used in or on masonry, sole plates, base plates and shim plates, unless otherwise speci- 
fied, shall conform to ASTM A36/A36M. 

2.2.4.2 Fabrication 

a. Holes in bearing plates may be formed by drilling, punching, or accurately controlled oxy- 
gen cutting. Burrs shall be removed by grinding. 

b. Sole plates of plate girders shall be in full contact with the girder flanges. Sole plates and 
masonry plates shall be planed or straightened. 

2.2.43 Installation 

a. Bearing plates shall be accurately set in level position as shown on the plans and shall have 
a uniform bearing over the whole area. They may be set on shims or on leveling .screws, with non- 
shrink grout so placed as to fill completely the space between the steel and the masonry. 

b. When plates are to be embedded in concrete, provision shall be made to keep the plates in 
correct position as the concrete is being placed. 

2.2.4.4 Bedding of Masonry Plates 

a. Filler, fabric, or elastomeric sheet materials shall be placed as bedding material under 
masonry plates when shown on the plans or specified. Such material shall be of the type specified or 
as ordered or approved by the Engineer and shall be installed to provide full bearing on contact areas. 
Immediately before placing the bedding material and installing bearings or masonry plates, the con- 
tact surfaces of the concrete and steel shall be thoroughly cleaned. 

b. Preformed fabric pads used as bedding shall be composed of multiple layers of 225 g (8- 
ounce) cotton duck impregnated and bonded with high quality natural rubber or of equivalent and 
equally suitable materials compres.sed into resilient pads of uniform thickness. The number of plies 
shall be such as to produce the specified thickness, after compression and vulcanizing. The finished 
pads shall withstand compression loads perpendicular to the plane of the laminations of not less than 
70 MPa (10,000 psi) without detrimental reduction in thickness or extrusion. 

c. Sheet lead used as bedding shall be common desilverized lead conforming to ASTM B29. 
The sheets shall be of uniform thickness and shall be free from cracks, seams, slivers, scale, and other 
defects. Unless otherwise specified, lead sheets shall be 3 ( 1/8 in.) in thickness with a permissible tol- 
erance of 0.80 (0.03 in.) plus or minus. 

d. Mortar used for filling under masonry plates shall conform to ASTM C270. 

e. Elastomeric bearing pads used as bedding shall be plain elastomeric bearing pads (unrein- 
forced) meeting the requirements of Section 1 .6 and Section 2.5. 

2.2.5 INCLINED BEARINGS 

(UNDER DEVELOPMENT) 

2.2.6 ANCHOR BOLTS 

2.2.6.1 Materials 

Anchor bolts shall meet the requirements of ASTM F1554 or as shown on the contract plans or spec- 
ifications. 



256 Bulletin 761 — American Railway Engineering Association 



2.2.6.2 Fabrication 

Anchor bolts shall be swedged or threaded to secure a satisfactory grip upon the material used to 
embed them in the holes. 

2.2.63 Installation 

a. The contractor shall drill holes for anchor bolts and set them in portland cement grout, or 
preset them as shown on the plans or as specified or directed by the Engineer. 

b. Location of anchor bolts shall take into account any variation from mean temperature of the 
superstructure at time of setting and anticipated lengthening of bottom chord or bottom flange due to 
dead load after setting, the intention being that, as near as practicable, at mean temperature and under 
dead load, the anchor bolts at expansion bearings will be centered in their slots. Care shall be taken 
that full and free movement of the superstructure at movable bearings is not restricted by anchor bolts 
or nuts. 



SECTION 2.3 BRONZE OR COPPER-ALLOY 
SLIDING EXPANSION BEARINGS 

(UNDER DEVELOPMENT) 



SECTION 2.4 TFE BEARING SURFACE 

(UNDER DEVELOPMENT) 



SECTION 2.5 ELASTOMERIC BEARINGS 



2.5.1 GENERAL 



a. Elastomeric bearings which are designed to act as a single unit with a given shape factor 
must be manufactured and vulcanized as a single unit and shall not be revulcanized after manufac- 
ture. 

b. Elastomeric bearings described herein shall include plain bearings (unreinforced pads con- 
sisting of elastomer only) and reinforced bearings with elastomer and steel laminates. 

c. Elastomeric bearings over 12(1/2 in.) thick, unless otherwise detailed or specified, shall be 
reinforced with steel laminates every 12 (1/2 in.) through the entire thickness. 

d. Bearings shall be furnished to the dimensions indicated in the contract plans or approved 
shop drawings. They shall be composed of elastomer of the specified type, grade, and shear modu- 
lus (or hardness); shall be adequate for the specified design load; shall meet the required test criteria; 
and shall satisfy any special requirements of the contract. In the absence of more specific informa- 
tion, elastomer shall be 60-durometer, shall be adequate for 7 MPa (1,000 psi) design compressive 
stress, and shall meet Test Criteria I. 

2.5.2 MATERIALS 

The materials for elastomeric bearings shall conform to the requirements of Section 1.6. 

2.5.3 PLAIN ELASTOMERIC BEARINGS 

a. Plain elastomeric bearings shall be cast in molds under pressure and heat, and may be 
molded individually, cut from previously molded strips, or slabs molded to the full thickness of the 
finished bearing. Plain bearings shall be fully vulcanized, uniform and integral units of such con- 



Proposed Manual Changes 257 



struction that the bearing cannot be separated by any mechanical means into separate well-defined 
elastomer layers. Evidence of layered construction shall be cause for rejection. 

b. Cutting of plain bearings from previously molded strips or slabs shall be performed in a 
manner to avoid heating of the material, and to produce an edge with no tears or other jagged areas. 
The surface roughness shall not exceed 250 (ANSI B46. 1, Surface Texture). 

c. Molds shall have a finish that provides a smooth undamaged surface for the bearing. 

2.5.4 REINFORCED ELASTOMERIC BEARINGS 

a. The supplier shall submit detailed shop drawings as defined in Article 2.1.2 before any fab- 
rication of reinforced elastomeric bearings is started. The manufacturer shall note on the shop draw- 
ings the shape factor, effective elastomer thickness, compressive area, shear area, width to height 
ratio, and length to height ratio. 

b. Reinforced elastomeric bearings shall have alternate layers of elastomer and steel reinforce- 
ment as shown on the design drawings, and shall be cast in individual molds under heat and pressure 
to form an integral unit of such construction that the bearing cannot be separated by any mechanical 
means into separate well-defined elastomer layers. Evidence of layered construction shall be cause 
for rejection. 

c. Molds shall have a finish that provides a smooth undamaged surface for the bearing. 

d. Steel reinforcement shall be abrasive blast cleaned to remove all rust, mill scale, and other 
contaminates, and shall be free of sharp edges and burrs. 

e. Steel reinforcement shall be covered by a minimum of 3 (1/8 in.) of elastomer on all faces. 
No surface of steel reinforcement shall be left exposed. 

2.5.5 EXTERNAL STEEL LOAD PLATES 

a. External steel load plates shall be abrasive blast cleaned to remove all rust, mill scale or other 
contaminates, and shall be hot bonded to the bearing during vulcanization. 

b. The external load plates shall be protected in accordance with the contract documents. 
Unless otherwise specified, they shall be given a shop coat of primer. No shop primer shall be used 
on external load plates which are to be field welded. 

2.5.6 TOLERANCES 

a. Flash tolerance, finish, and appearance shall meet the requirements of the latest edition of 
the Rubber Handbook as published by the Rubber Manufacturers Association Inc., RMA F3 and 
T 1.60mm (0.063 in.) for molded bearings. 

b. For both plain and reinforced bearings, the permissible variation from the dimensions and 
configuration required by the plans and these specifications shall be as follows: 

( 1 ) Overall vertical dimensions mm (in.) 

Design thickness 32 (1-1/4 in.) 
or less 

Design thickness over 32 (1-1/4 in.) 

(2) Overall horizontal dimensions 
900 (36 in.) and less 
over 900 (36 in.) 

(3) Thickness of individual layers of 
elastomer (laminated bearings only) 
at any point within the bearing 



-0,-t-2 


(-0,-1-3/32) 


0,-h4 


(-0,+3/I6) 


-0,-1-6 


(-0,-1-1/4) 


-0,+ I2 


(-0,-1-1/2) 


±20% of design 


value but no 


more than: 




±0.5 


(± 0.02) 



•0,+3 


(-0, +1/8) 


±3 


(± 1/8) 


±3 


(± 1/8) 



258 Bulletin 761 — American Railway Engineering Association 



(4) Variations from a plane parallel to 
the theoretical surface (as 
determined by measurements at the 
edge of the bearings) 

Top slope relative to the bottom of 

no more than 0.005 radians 

Side 6 (1/4) 

(5) Position of exposed 

connection members 3 ( 1/8) 

(6) Edge cover of embedded 
steel laminates at laminate 
restraining devices and 
around holes and slots 

(7) Size of holes, slots, or inserts 

(8) Position of holes, slots, or inserts 

2.5.7 MARKING 

Each bearing shall be marked in indelible ink or flexible paint. The marking shall consist of the order 
number, lot number, bearing identification number, elastomer type and grade number. Unless other- 
wise specified in the contract documents, or impossible due to the pad thickness, the marking shall 
be on a face which is visible after erection of the bridge. 

2.5.8 ACCEPTANCE CRITERIA 

a. The acceptance criteria for the bearing shall be specified by the Engineer and shall meet the 
requirements of Test Criteria I or Test Criteria II. Test Criteria I acceptance shall be applied to all 
bearings. Test Criteria II acceptance shall, at the discretion of the Engineer, be required for more crit- 
ical or unusual bearings. 

b. The supplier shall give written notice 30 days prior to the start of bearing fabrication. This 
notification shall include number, quantity, size, manufacturer's name, location, and the name of the 
plant coordinator where the bearings are being produced. The Engineer's representative shall choose, 
or direct to be chosen, the sample bearings for testing. Should the bearings chosen for testing have 
an integrally bonded tapered external steel load plate, the supplier shall supply another tapered steel 
plate so that parallel top and bottom surfaces are provided for testing. The Engineer shall decide if 
he wishes to be present during the testing. 

c. All testing shall be performed by, and at the expense of the supplier, and shall be conducted 
according to the requirements of this chapter. 

2.5.9 TEST CRITERIA I 

a. The manufacturer shall test and report the verification of the location and parallelism 
requirements of Article 2.5.6 (b) by measurements under a proof load of 3.4 MPa (500 psi). 
Measurements shall be taken at angle intervals of 90 degrees and the largest and smallest measure- 
ment for each reinforcement layer shall be reported. 

b. The manufacturer shall proof load each reinforced bearing with a compressive load of 10.3 
MPa (1500 psi), or 1.5 times the design load if this load is given. If bulging patterns indicate steel 
placement which does not satisfy design criteria and manufacturing tolerances or if bulging suggests 
poor reinforcement bond, the bearing shall be rejected. If there are 3 or more separate surface cracks 
which are greater than 2 (0.08 in.) wide and 2 (0.08 in.) deep, the bearings shall be rejected. 



Proposed Manual Changes 259 



c. The elastomer shall satisfy the minimum properties of Table 1.6.2. Other material tests shall 
be performed whenever there is a change in the type or source of raw materials, elastomer formula- 
tion or production procedures, or as required by the Engineer. 

d. A Cold Temperature Shear test shall not be required unless indicated in the contract docu- 
ments. If required, the test shall be conducted in the manner listed in ( 1 ) through (5) below. One com- 
plete set of all performance tests shall be performed and reported on each production run. 

(1) Unless otherwise specified, a test temperature of -29 ± 0.5 Degrees C (-20 ± 1 Degrees 
F) is to be used for determination of low temperature properties. Should a lower tem- 
perature test be required, the requirements of this test shall be set forth in the contract 
documents with due consideration of the ability of the manufacturer to perform the test 
at a lower temperature. 

(2) The bearing shall be conditioned at the test temperature for 96 hours. 

(3) The total time lapse between removal of the bearing from the -29±0.5 Degrees C (-20±1 
Degrees F) environment and completion of the cold weather test shall not exceed 30 
minutes. Bearings shall be insulated from any heat conducting surface of the testing 
apparatus with a suitable material, having a thermal conductivity of not more than 0.3 
W/m' (0. 1 BTU/hr/sqft.). During removal of the bearing and positioning for the test, the 
bearing shall be completely covered with a 50 (2 in.) minimum thickness insulating 
blanket having a thermal conductivity of not more than 0.13 W/m^ (0.04 BTU/hr/sqft.). 
During the actual testing, the exposed sides of the bearing shall be covered by the blan- 
ket. 

(4) After the bearing is conditioned at the test temperature and placed into position for the 
testing, the bearing shall be subjected to a vertical load of 3.4 MPa (500 psi), and then 
sheared to a total strain equivalent to 25 percent of the effective original rubber thick- 
ness. Shear stresses, based upon the plan area of the rubber, shall be recorded at and 
15 minutes after the total strain has been reached. The shear stress, measured 15 min- 
utes after the ultimate strain has been reached shall not exceed 0.34 MPa (50 psi) for 
bearings constructed of neoprene, nor 0.21 MPa (30 psi) for bearings constructed with 
natural rubber. 

e. To establish conformance with the requirements of Table 1 .6.2, one complete set of tests 
shall be conducted on each production run. 

2.5.10 TEST CRITERIA II 

a. Provisions of Test Criteria I shall be satisfied. The shear modulus of the material in the fin- 
ished bearing shall be evaluated by testing a specimen cut from it using the apparatus and procedure 
described in the Annex of ASTM D4014 or, at the discretion of the Engineer, a comparable nonde- 
structive shear stiffness test may be conducted on a pair of finished bearings. A test temperature for 
the shear modulus test shall be specified by the Engineer. More than one temperature may be 
requested by the Engineer. The shear modulus shall not vary by more than ±15% from the specified 
value in the contract documents, as determined by the requirements of ASTM D4014. 

b. The compressive stiffness, as determined by the requirements of ASTM D4014, shall vary 
by no more than ±10% from the median value of all bearings, nor ± 20% from the design value. The 
compressive stiffness test shall be performed on a completed bearing. 

2.5.11 CERTIFICATION 

The manufacturer shall certify that each bearing meets the requirements of this chapter, and shall sup- 
ply a certified copy of the test results. Where actual test values can be obtained, they shall be reported 
and not listed only as "Passed." 



260 



Bulletin 761 — American Railway Engineering Association 



2.5.12 INSTALLATION 

a. Elastomeric bearings shall be installed in accordance with the design plans. Substructure 
bearing surfaces to receive the bearings shall be level, smooth, and finished to the correct elevation. 

b. Top and bottom elastomer surfaces shall be level under dead load only. Tapered steel load 
plates bonded to the bearing, or tapered steel sole plates on the bridge span shall compensate for span 
grade, rotation, or camber. 

c. Bearings which are to be attached to the bridge span and/or substructure shall use a positive 
attachment detail. Adhesive bonding is not permitted. 

d. Welding of bridge span members to the bearing load plates is not permitted unless there is 
more than 40 (1 1/2 in.) of steel between the weld and the elastomer. The temperature of the steel 
plate in contact with the elastomer shall not exceed 200 degrees C (400 degrees F) during the weld- 
ing process. 

SECTION 2.6 MULTI-ROTATIONAL BEARINGS 

(UNDER DEVELOPMENT) 

2.6.1 POT BEARINGS (UNDER DEVELOPMENT) 

2.6.2 DISC BEARINGS (UNDER DEVELOPMENT) 
2.63 SPHERICAL BEARINGS (UNDER DEVELOPMENT) 

SECTION 2.7 SEISMIC ISOLATION BEARINGS AND DEVICES 

(UNDER DEVELOPMENT) 




HANSON 
WILSON 



We engineer solutions for railroads. 

Hanson-Wilson Inc. provides engineering and 
architectural services to meet your design and 
construction challenges. We offer the expertise 
of more than 500 professional, technical and 
support personnel. 



3101 Broadway, Suite 900 Kansas City, MO USA 
Td: (81 6) 561 -9054Fax: (81 6) 561-0654 




Engineering Design 
Track and Bridge Inspection & Design 
Construction Engineering 
Fueling & Maintenance Facilities 
Intermodal Facilities 



Proposed Manual Changes 



261 



Proposed 1998 Manual Revisions to 
Portfolio of Trackwork Plans 

Add the following Plan No. 1003-98, Data for Currently Produced Rail Sections, to the Portfolio. 

PORTFOLIO OF TRACKWORK PLANS 




ON THE PROWL 
10 INCREASE PRODOCmniY 

The Plasser Continuous Action Tamper 09-16 C.A.T. clears the 
way for better working comfort at lower costs. It's innovative design 
can produce a 30% increase in production while reducing stresses 
on both operator and machine. With the machine's working units 
positioned on a separate subframe and indexed from tie to tie 
during the work cycle, a new 
level of working comfort is now 
available for track maintenance 
crews. Compared to conventional 
tamping machines, only 20% of 
the total mass of the Plasser 09-16 
C.A.T. is accelerated and braked 
during the work cycle. The main 
frame of the machine moves 
forward smoothly and continu- 
ously The machine is subject to much less strain and wear. Track 
time can be much more effectively utilized. 

For improved cost savings and comfort, rely on the Continuous 
Action Tamper 09-16 C.A.T, exclusively from Plasser 

Call or write today for specifications and frill details. 




Lifting, lining and tamping units are mounted on a 
separately moving satellite frame. 




nASSBMBICMCMNUIIIII PIASSEI CUUM INC. 



2001 Myers Road, P.O. Box 5464 

Chesapeake,VA 23324-0464, US. A. 

(804)543-3526 



2705 Marcel Street 
Montreal, Quebec H4R1A6, Canada 

(514)336-3274 




POLLUTION Costs. 

Brownie Tank 

Protects. 



Only Brownie's fuel pumping skid 
provides you with a heavy steel 
channel base in a complete, turn- 
key refueling system. All your 
diesel fuel gets into the 
locomotive so there's no costly 
spillage, waste or polluted ground. 

Brownie's fuel pumping skid is 
pre-wired, pre-piped and ready to 
install with no field assembly 
required. You simply hook up the 
inlet and outlet piping and run 
electrical power to it. 

With Brownie Tank, there is no 
sourcing, spec'ing or searching for 
pumps, filters, valves, meters or 
design configurations. 



So whether you're working with an 
outside contractor, architect or in- 
house operation, you can provide 
your locomotive refueling 
operation with the finest, turn-key 
skid in the industry. 

Lube and journal oil skids and 
methanol injection systems are 
also available. 

For a quote or more information 
simply call: 

Brownie Tank Mfg. Co., a 
division of Determan Welding 
& Tank Service, Inc. 
Minneapolis, Minnesota 

(612)571-1744 
Fax(612) 571-1789 




Performance Under Pressure. 3t^^me 



263 




• Digs, loads, lifts and grades. • Gets to the job at 55 mph. 
Best-built, hardest working model in the popular 3/4 cu.yd size. 



• Ray gear • Train line air • AAR 
couplers • Axle leeks for Hfl 
stability oH-rail 



mmm mk 



• Dependable diesei power 
Detroit Diesei, upper stnicture - 
Cummins, cairier • Eigonomic cab 
design with joysticfc controb 



American products built to last 

P.O. Box 798, Airport Industrial Paric • Winona, MN 55987 • Phone (507) 454-8549 - Fax (507 )454-3326 

Call or fax for detailed specifications, capacities and performance options. 



Canada "^ Canada 
Sydney Rails For 

Quality Track 

• Premium Wear-Resistant Head Hardened Rails 

• Standard Carbon, Intermediate Strength and Pre- 
mium Alloy Grades-ail using clean steel practice. 

• All National and International specifications, 
including A.R.E.A., C.N.R., BS-11 and U.I.C. 

• Sections from 50 to 70 kilograms per metre (100 to 
136 pounds per yard). 

• Lengths up to 26 metres (85 feet). 

• More than eighty years of experience supplying 
world markets. 

For additional information, write, wire or call: 

Vice President, Sales 
Sydney Steel Corporation 
Sydney, Nova Scotia, Canada 
B1P6K5 



Telephone: 
(902) 564-7910 

Telefax: 

(902) 564-7903 

Telex: 

019-35197 



B 



SYDNEY 

STEEL 

CORPORATION 



264 



IMPLEMENTATION OF HIGH SPEED RAIL 
IN AN EXISTING CORRIDOR 

By: Charles C. De Weese* 



Abstract 



Environmental impacts and the expense of developing new high speed rail corridors between 
urban areas have made dedicated high speed rail systems impossible to implement in many locations. 
Use of existing rights of way of freight railroads appears to offer significant advantages. 
Unfortunately, increasing train speeds creates significant impacts to existing corridors, especially for 
grade crossings and intermediate communities. 

Earlier analysis in Illinois Department of Transportation's Chicago-St. Louis high speed sug- 
gested closing many grade crossings in rural areas and in small communities. Community and gov- 
ernment analysis of the proposed potential crossing closures anticipated restricted mobility, undesir- 
able mixing of diverse traffic types, and traffic rerouting distances in excess of Illinois Commerce 
Commission guidelines. Strong resistance from affected residents, business and local leaders pro- 
duced intense political opposition. 

A thorough grade crossing analysis was conducted, including meetings with county engineers, 
city staff farm bureau leaders, and other local interested parties. A wide spectrum of grade crossing 
safety improvement ideas were considered, including lower-cost grade separations, and conventional 
warning devices. Positive protection achieved through the use of Vehicle Arresting Barriers, was 
introduced and is now being tested. Fencing was proposed to limit pedestrian exposure to harm. This 
analysis resulted in a set of revised ideas that were presented workshop-format meetings in selected 
locations along the 300 mile corridor The result has been an acceptance of the grade crossing ideas 
for the project by the public that allows continuing further project development. 

Background 

Environmental impacts and the raw expense of developing new high-speed rail corridors 
between urban areas have made high speed rail systems with dedicated, grade separated rights-of- 
way socially and economically impossible to implement in many U. S. applications. U.se of existing 
rights-of-way along freight corridors appears to offer significant advantages. Unfortunately, increas- 
ing train speeds creates significant impacts on existing corridors, especially for grade crossings and 
intermediate communities. 

Nationally, the Federal Railroad Administration (ERA) has encouraged the Incremental devel- 
opment of existing corridor rights-of-way for several years. Where train speeds are 110 miles per 
hour or less, current FRA policy allows properly warned grade crossings to remain open. Above 1 10 
miles per hour, current guidelines suggest providing "positive protection", frequently grade separa- 
tion but also potentially including creative devices that have the effect of guaranteeing non-intrusion 
of a motor vehicle into the path of an oncoming train. 

Typical operating and engineering problems encountered in implementing high speed rail pas- 
senger operations include on-line or at-grade crossing freight interference, line geometry, grades, and 
acquisition of additional right-of-way to add tracks. The Chicago-St. Louis corridor that Illinois 
Department of Transportation (IDOT) has been developing for several years suffers very little from 
these problems. The geometry of the line is excellent, the grades are insignificant, and the line was 
constructed with double track for the majority of its length. 



'Railroads Technology Secior Manager. De Leuw Gather & Co.. Parsons Transponalion Group 

265 



266 Bulletin 761 — American Railway Engineering Association 



Statement of the Problem 

Previous Chicago-St. Louis high speed corridor development analyses concluded that the travel 
times possible with high-speed rail would produce a viable system, able to cover operating costs from 
the farebox. In order to achieve these travel times, that work suggested closing many grade crossings 
in rural areas and in small communities. Community and local government analysis of the potential 
crossing closures anticipated restricted mobility, undesirable mixing of diverse traffic types, and traf- 
fic rerouting distances in excess of Illinois Commerce Commission (ICC) guidelines. These impacts 
generated intense project opposition from residents outside the major population centers. 

A combination of vagaries in quality and quantity of Amtrak services over time in the corridor 
have had a negative effect on ridership and public support. The central part of the line (Joliet to Alton) 
has had five owners since Amtrak began operating service on the line. (Gulf, Mobile and Ohio, 
Illinois Central Gulf, Chicago, Missouri and Western, Southern Pacific, and Union Pacific). 
Maintenance levels and speed limits have varied with these changes, thus trip times have been unre- 
liable, and the result has been suspicion of the traveling public as to the viability of passenger travel. 

The earlier work proposed closing over half of the crossings in the corridor, and IDOT received 
strong criticism from elected leaders in the communities along the route. Strong resistance from 
affected residents, business, and local leaders produced intense political opposition. An unfortunate 
result of high speed rail development is that residents between stations tend to bear the brunt of 
impacts in terms of construction and any adverse impacts (perceived or real) from fast trains operat- 
ing on a line where slower trains have been the norm for some time. On the other hand, these per- 
sons living and working between stations perceive themselves as the least likely to derive benefits 
from High Speed Rail. The result in central Illinois was a significant group of people not living in 
cities who believed that High Speed Rail represented an effort by persons living in cities (Chicago, 
Bloomington, Springfield and St. Louis) to have a faster way to travel on a mode that had proven 
unreliable and slow in the past and would result in many grade crossings being closed. The general 
view of these people was that closing grade crossings was a reduction in mobility, resulted in unde- 
sirable mixing of traffic types, caused increased exposure to other various safety hazards, and was in 
support of a project to promote a mode of travel mainly to be used by people living in cities that had 
been proven unsatisfactory in the past. 

In order to progress the project, it was determined by IDOT that the preparation of Draft 
Environmental Impact Statement (DEIS) should include a thorough review and evaluation of the 
entire grade crossing plan, in addition to the other work required with respect to siting passing track 
and double track capacity improvements for high-speed trains and freight trains. That review was to 
be carried to local elected officials before the DEIS work was completed, documented, and carried 
to the hearing process. IDOT's concern was that until a set of grade crossing safety improvement 
ideas which would be acceptable to local communities could be prepared, there would be no objec- 
tive analysis of the project in the DEIS process. Intense opposition to the project would continue 
without an approved set of new crossing improvement ideas. 

Applied Methodology 

The major engineering elements of the DEIS preparation process were 

• determining the locations where a second main track, or passing loops, could be located to 
allow high-speed trains on two hour headways to meet with minimal delay, 

• identifying the connections and new construction that would be required for each of four 
alignment alternatives to access Chicago, 

• determining proposed train speed at each grade crossing location, and 

• identifying the locations where additional freight train meeting and passing capability would 
be needed. 



Paper by Charles C. De Weese 267 



The locations where the second main track would be required to allow high speed trains to meet 
with minimal delay, the proposed high-speed train speeds at each grade crossing, and the additional 
freight train siding and second main track locations were analyzed using a train performance calcu- 
lator and train dispatch simulation model. That analysis provided sufficient information to proceed 
with the detailed determination of grade crossing requirements. 

A thorough grade crossing analysis was conducted, using detailed information gathered from 
meetings with county engineers, city staff, farm bureau leaders, and other local interested parties. A 
wide spectrum of grade crossing safety improvement ideas were considered, including lower-cost 
grade separations, and conventional warning devices equipped with flashing lights, gates, and median 
barriers. More creative solutions, such as the positive protection provided by Vehicle Arresting 
Barriers were also considered. IDOT is currently preparing for an experimental test of "Dragnet" 
VAB equipment at three existing crossings in the Chicago-St. Louis corridor. The Dragnet demon- 
stration is expected to commence late spring or early summer this year. 

Introduction 

The Chicago to St. Louis High Speed Rail Corridor is a portion of the Midwest Hub, one of five 
Intermodal Surface Transportation Efficiency Act (ISTEA) Section 1010 designated high speed rail 
corridors. As train speeds increase in high speed rail corridors, traditional crossing warning systems, 
which require vehicle operators to evaluate their ability to proceed across a crossing safely, becomes 
less effective. As a result, the potential severity of an accident to the vehicle, the train and the occu- 
pants increases. Along these corridors, the FRA has recommended that all existing crossings either 
be closed, grade separated, or equipped with special signing and active warning devices and/or gates 
with constant warning time, or provided with positive protection. The cost of grade separating, pro- 
viding upgraded warning devices, or even installing positive protection devices at all crossings along 
prospective high-speed rail corridors is prohibitive. Closure of nonessential crossings on high-speed 
rail corridors will enhance the safety of railroad passengers and highway users. 

Increases in train speeds along high-speed rail corridors warrant an increase in the level of grade 
crossing safety. Yet, train operations do not always provide for increased train speeds over the entire 
corridor due to lengthy deceleration and acceleration at stations or alignment restrictions. The FRA 
has suggested a national goal of reducing the number of the nation's public and private crossings by 
25 percent by the year 2000. Therefore, consistent with FRA guidelines and good engineering prac- 
tice, the number of crossings should be minimized and the remaining ones equipped with appropri- 
ate warning devices. Whether a crossing is eliminated or improved is generally based on the number 
and speed of trains, vehicular usage and nearby availability of alternate, appropriate existing cross- 
ings or crossings that will be improved as part of the corridor project. 

Approach 

The study team developed a comprehensive and systematic corridor approach to evaluating rail 
crossings for the Chicago to St. Louis High Speed Rail Corridor. A corridor approach, which is rec- 
ommended for ISTEA Section 1010 corridors, evaluates multiple rail crossings in a rail segment 
rather than individual crossings. The approach is a multi modal effort to improve rail efficiency while 
also improving pedestrian, vehicular and rail safety. The approach for evaluating the rail crossings 
consists of the following steps: 

• Review of previous public comments; 

• Acquisition of new data; 

• Discussions with local officials, county engineers and planners; 

• Estimates of train speeds; 

• Evaluation of vehicular traffic demand, roadway/crossing capacity and accident data; 



268 Bulletin 761 — American Railway Engineering Association 



• Classification of crossings by function and use; 

• Identification of travel markets served by the crossings; 

• Estimates of adverse travel; and 

• Development of preliminary grade crossing treatment ideas. 

Review of Previous Public Comments 

A previous grade crossing study, the Chicago-St. Louis High Speed Rail Study Grade Crossing 
Safety Analysis had been prepared. Public comments regarding this study were reviewed and pub- 
lished in 1994 to familiarize the study team with relevant corridor issues. A database was created to 
centralize all available information on the existing public and private crossings in the corridor. New 
information on some of the existing crossing conditions identified since the previous study was pub- 
lished was used to supplement the database. 

Acquisition of New Data 

The study team collected available ambulance, police and other emergency services district 
boundaries and routes; locations of hospitals; locations of schools and bus routes; locations of 
regional grain elevators and other major agribusinesses; and additional secondary source data. 

Discussions with Local Officials, County Engineers and Planners 

One significant source of secondary data were local county engineers and county and regional 
planners along the corridor. The study team met with representatives from each county and, in some 
cases cities, villages, and regional development agencies during the data collection process. The pre- 
vious public comments regarding prior proposed grade crossing treatments were summarized on a 
corridor map and discussed with these local jurisdictions. 

Train Speed Estimates 

Grade crossing protection and warning devices on high speed rail corridors are dependent on 
predicted train speeds. The study team analyzed the potential operating speeds of the corridor using 
a simulation model to identify areas with the following speed ranges: less than 80 miles per hour, 80 
to 110 miles per hour, and 110 to 125 miles per hour. Suggested crossing treatments for locations 
where train speed is 1 10 mph or less than generally conform to the rules and requirements of the ICC, 
the Manual on Uniform Traffic Control Devices for Streets and Highways and the Requirements for 
Railroad — Highway Grade Crossing Protection of IDOT 

For train operation between 80 and 1 10 mph, the guidelines recommend closure of redundant 
crossings, improved signing, and upgraded warning devices equipped with constant warning time 
equipment. 

Train operations between 1 10 and 125 mph require positive protection, which not only protect 
the motor vehicle operator but also protect the train from intrusion of a vehicle onto the crossing. 
IDOT has successfully crash tested a prototype Vehicle Arresting Barrier (VAB), which when acti- 
vated by an approaching train will lower a net across the roadway on each side of the tracks to inter- 
cept an approaching vehicle on the highway. The VAB system also will have a vehicle intrusion 
detection system to detect vehicles that have been arrested by the net. Where closure or grade sepa- 
ration of crossings is not practical, VABs are considered an alternative. As previously mentioned, 
IDOT will be conducting experimental tests of the VAB system at three existing crossings in the 
Chicago-St. Louis corridor. 

Although train speeds of 140 mph were evaluated in previous studies for the corridor, top 
speeds of 125 mph were considered in these evaluations. 



Paper by Charles C. De Weese 269 



Evaluation of Vehicular IVaffic Demand, Roadway/Crossing Capacity and Accident Data 

The preliminary sources of vehicular traffic demand — Average Daily Traffic (ADT) — were 
obtained from the 1993 and 1994 grade crossing safety analysis and 1995 24-hour machine traffic 
counts. A minimum ADT of one was used for occasional use facilities, such as farm entrances, and 
zero was used for crossings limited to emergency use only and for crossings that did not appear to be 
used at all. Many ADTs were updated using available IDOT county traffic maps. For some low to 
moderate volume public roads and many private crossings, vehicular volumes were counted in the 
Fail of 1995. 

In accordance with ICC rules, crossing treatments for locations where the train speeds are less 
than 80 miles per hour are dependent upon the product of the ADT times the average number of daily 
train movements and safe stopping sight distances. Locations where the crossing treatment may be 
sensitive to the ADT were identified and reviewed prior to finalizing the rail crossing treatment pro- 
posed. 

Accident data also was used to identify candidate crossings for closure and improvements. 

Classification of Crossings by Function and Use 

The grade crossing locations were categorized by county. Consistent with a corridor approach, 
the study team categorized each crossing within each county by function and use. Crossings given 
the highest consideration for retention were those on state, county, township or municipal routes 
which are important for maintaining interregional route continuity, those in contiguous central busi- 
ness districts, neighborhoods, industrial areas or agricultural areas, or those on the only available 
route between probable destination pairs were given highest consideration for retention as remaining 
open. Crossings given primary consideration for consolidation were those on other, less critical, state, 
county, township, and municipal routes and private crossings. In many cases consolidation ideas for 
these types of crossing also included the provision for a frontage road or alternate access. 

Identification of Travel Markets Served by the Crossings 

At-grade crossings on routes that serve special travel markets were identified and given high- 
est consideration for remaining in service. These travel markets included public services, such as 
school buses and emergency service providers, transportation-dependent businesses or commerce, 
and hazardous waste carriers. 

Estimate of Adverse Travel 

The ICC has established a .set of rules (Title 92, Chapter III, Subchapter c. Part 1536) for clo- 
sure of existing grade crossings and the construction of new crossings. According to these rules, a 
crossing should not be eliminated if the closure would result in unduly burdensome adverse travel for 
the users of the crossing. The ICC considers the amount of adverse distance which the closure will 
cau.se to not be considered unduly burdensome if 

(1) it is equal to or less than four miles in unincorporated areas; or 

(2) it is equal to or less than 0.75' miles in incorporated towns, villages, and cities. 

The Commission also requires consent of the affected highway agency(s) if two adjacent cross- 
ings are closed and the distance between the two crossings — as measured along the railroad tracks — 
is equal to or greater than one mile. 

The study team determined the amount of adverse travel for each at-grade crossing to identify 
candidates for consolidation. The presence of adequate alternate routes also was determined for each 
crossing identified as a candidate for consolidation. The alternate railroad crossing roadway was con- 
sidered adequate if it had an all-weather surface; sufficient width (capacity) to accommodate its exist- 



270 



Bulletin 761 — American Railway Engineering Association 



ing and rerouted vehicular traffic; the same type(s) of vehicular traffic; and an appropriate crossing 
warning treatment. In addition alternate crossings should be on routes that have adequate vertical and 
horizontal clearances, particularly in agricultural areas. 

Development of Preliminary Concepts and Grade Crossing Treatments 

The preliminary concepts and grade crossing treatment ideas were developed jointly by a multi- 
disciplined study team using video of the corridor and field reconnaissance of critical locations. Site 
specific analyses were documented and considerations for each public or private use of at- grade cor- 
ridor crossings were also documented. The suggested action included the proposed status of the 
crossing (opened or closed) and the conceptual crossing treatment ideas (e.g., warning devices, vehi- 
cle arresting barrier, or separation). 

For any at-grade crossing that was proposed to remain open and provided with active warning 
devices — primarily flashing lights and gates — constant warning time (CWT) capabilities were sug- 
gested. This would included upgrades, if necessary, of existing warning devices to provide this fea- 
ture where it is not currently present. 

Results 

The preliminary concepts and grade crossing treatment ideas, which are a result of a systematic 
and comprehensive corridor approach, were consistent with FRA guidelines, previous public com- 
ments and recommendations received from local officials. Where a crossing is proposed for closure, 
traffic can be rerouted to another crossing which would provide equal or better warning and/or pro- 
tection or a grade separation. This greatly reduces or eliminates the potential for train-vehicle acci- 
dents. Traffic will increase on the crossings where traffic rerouting is proposed. In all instances where 
crossings are proposed to be closed, adequate reserve capacity existed on the adjacent crossings pro- 
posed to remain open to accommodate this rerouted traffic. The results of the initial analysis are sum- 
marized in Table 1 . 









Table 1 














Summary Ideas for All Counties 
Rail Crossing Evaluation, High Speed 
1996 SUGGESTED ACTION 


Rail 














Totals for Alternative Alignments 










Alternative 
Alignment 


Close 


Close w/ No No 
Frontage Action Ciiange 
Road 


Ped. Conventional 
Bell and Gates (1) 
Flashers 


Electric 

Lock 

Gates 


Vehicle 

Arresting 

Barrier 


Grade 
Separate 


Total 
(2) 


Chgo-Jol-St. Louis 


53 


36 


6 118 


5 82 


12 


10 


6 


328 


Con rail 


57 


40 


5 105 


1 77 


14 


10 


6 


315 


Green Grass 


46 


42 


6 96 


1 73 


12 


10 


17 


303 


Rock Island District 


62 


32 


6 128 


22 80 


12 


10 


4 


356 



( 1 ) Conventional gates are recommended a.s an initial safely device. At many locations, more sophisticated enhanced warning 
devices could be in.slalled depending on local conditions. 

(2) Total Cro.ssings; Total number of crossings for which there are recommendations as of December .11, 1996. Includes highway 
overpasses and underpasses only when traffic will be diverted to these crossings as a result of a suggested at-grade crossing 
closure. 



Paper by Charles C. De Weese 271 



Safety Analysis and Considerations 

There are certain crossings where vehicle detection circuitry is recommended and where traffic sig- 
nals should be interconnected with grade crossing signals. Vehicle detection circuitry is recommended at 
crossings where there is school bus traffic and the proposed train speed is greater than 79 mph. 

Quad gates, gates that lower on all roadway approach and departure lanes (four quadrants) when 
trains approach, also were considered as part of this analysis. Although quad gates could still be rec- 
ommended later in the study, they are not suggested for any crossing at this time. They would be sig- 
nificantly more expensive to install than conventional gates, and they would delay vehicular traffic for 
a longer time. The additional delay stems from the dual requirements, as currently envisioned by the 
ICC, to ( 1 ) have the crossing exit gates lower sufficiently after the entrance gates are in place to allow 
vehicles to exit the crossing without being trapped, and (2) for all the gates to be in place at least 20 
seconds and more usually 30 seconds before the train occupies the crossing. Further, quad gates pro- 
vide no guarantee that a vehicle could be prevented from entering crossings. Trapped vehicle detec- 
tion circuitry would have to be included in any quad gate application. This requirement would result 
in an extended pre-waming time, thereby creating additional delay for highway vehicles. 

Positive protection, a new grade crossing safety improvement technique, could be achieved 
through the use of Vehicle Arresting Barriers. VABs have been introduced to the public, and are being 
tested in Illinois. Through towns, fencing was proposed to limit pedestrian exposure to harm. 

This analysis resulted in a .set of revi.sed ideas that were presented to local elected officials in 
workshop-format meetings in 14 .selected locations along the 300 mile corridor. Invitees for the meet- 
ings included mayors, council members, farm bureau leaders, and city engineers. The meeting for- 
mat consisted of an approximately 30 minute presentation by an experienced public information pro- 
fessional, followed by up to 90 minutes of group and individual questions and answers. Changes in 
crossing treatment ideas were propo.sed and included in the DEIS as a result of the meetings. Where 
questions could not be resolved at the workshop, follow-on meetings and analy.ses were completed. 

Results 

At-Grade Crossings 

Implementation of the HSR will impact vehicular traffic throughout the corridor. However, 
impact will be limited mostly to low volume roads becau.se almost all major, high volume roads that 
were built or substantially upgraded over the years have included grade separations with the existing 
railroads. Also, only grade crossings on lower volume roads have been suggested for closure. While 
approximately 30 percent of the at-grade crossings are suggested for closure, these crossings only 
accommodate around two to three percent of the ADT crossing the alternative alignments throughout 
the HSR corridor. 

Average daily traffic on the existing at-grade crossings range from one to 41,000 vehicles (93th 
Street-Rock Island District alignment, MP 10.80). Of the crossings suggested for closure, none have 
an ADT greater than 2,210 vehicles. In all instances where crossing closures are suggested, adequate 
reserve capacity exi.sts on the adjacent crossings to handle the diverted traffic. 

Most of the higher volume crossings are located within the urbanized areas of the corridor. 
Springfield has 14 grade crossings where ADT exceeds 5,000 vehicles, while Bloomington/Normal has 
five such crossings. A number of the smaller towns also have one or two crossings in the higher volume 
range, but none of these are proposed to be closed under the HSR alternative. Closure of these streets 
would cau.se undue disruption to local traffic circulation and negatively impact land use access. 



MANNESMANN 

DEMAG 

GOTTWALD 




EXPERIENCE 

+ 

INNOVATION 

GS 88.79 TR 



Railquimiiic 



The employment of economical operation and modern 
tectinology is the foundation for a safe cost-efficient and 
time saving operation in track and bridge construction. 
The requirements are increasing. That is why track and 
bridge construction should be equipped to face the 
future. 

The GS 88.79 TR Railroad Crane sets the standard in the fol- 
lowing areas: 

• Safety 

• Carrying Capacity 

• Speed 

• Environmental Compliance 

• Ease of Operation 

Future requirements are already fulfilled today, new appli- 
cations developed. You can make use of the experience 
of a track and bridge construction specialist and the com- 
petence of a leading crane manufacturer 

If you take us up on It, you will always be on the right track. 





^ RailquiiH inc. 

■^ 3731 Northcrest Road, Suite 6, Atlanta, GA 30340 

I (770) 458-41 57 • Fax (770) 458-5365 



272 




"From sea to shining sea..." Premier Crossing 
Systems are installed at some of the finest 
port and intermodal facilities coast to coast. 
There's a good reason. 



Innovative Premier 
Advanced Panel 
Lag-Down or Lag- 
Free Systems 
where panels go 
right onto wood or 
concrete ties. 





Tieless, Premier 
modular systems 
for LRT installa- 
tions and absolute 
rail isolation. 



Partnering 

Working together to bring 
crossing projects in On 
Time & Under Budget. 

That means 

. working with railroad, contractors, 
and plant maintenance people to 
develop crossing solutions. 
. putting together the right system for 
you to provide low installation costs, 
low maintenance costs, and long 
crossing life. 

. providing on-site assistance from a 
Premier® Field Representative to 
ensure quick and easy installation. 
We call it partnering. 
From the Port Authority of New York 
& New Jersey across the nation to the 
Port of Los Angeles. For intermodal 
facilities, light rail, heavy rail, or 
industrial sites. Premier Crossing 
Systems^"^ are fulfilling industry 
needs. Meeting schedules and 
deadlines. Keeping costs down with 
low installation and maintenance costs. 
And building in long crossing life. 
There are more Premier benefits — 
including salt (Chloride) resistance, the 
non-skid diamond plate crossing 
surface, and rubber flangeway inserts 
to provide a positive shunt-free 
system. 

The original concrete crossing. 

Whatever your next crossing project 
calls for - from panels on concrete or 
wooden ties to a modular, tieless 
crossing system, we can meet your 
needs. Write to us at PO.Box 1 1305; 
Portland, OR 9721 1. FAX 503-240- 
3592, or call to learn more about the 
full line of Premier Concrete Systems... 
1-800-426-5556. 



PREMIER 

CONCRETE RAILROAD CROSSINGS 
Partnering solutions for rail industry challenges. 



273 



274 Bulletin 761 — American Railway Engineering Association 



Because of the low volumes noted above, alternative access, rather than capacity, was the pri- 
mary consideration in determining which crossings could be closed. In this regard, the ICC regulation 
governing the maximum allowable adverse travel became one of the key criterion used for evaluation 
of potential closures. This criteria specifies four miles of adverse travel as maxiitium allowable in 
unincorporated areas and 0.75 miles as the maximum in incorporated areas. The travel distance is mea- 
sured as the shortest, usable path from one side of the closed crossing to the other. Average adverse 
travel as a result of the proposed at -grade crossing closures on the alternative alignments ranges from 
0.87 miles on the IC/SPCSL alignment to 1 .09 miles on the Green Grass alignment. 

The use of grade separations was limited to locations where grades are conducive to separation 
and where land use impacts will be minimal. These opportunities typically occur on the edge of urban 
areas where development is less dense than in the center of the town. Of the seven crossings sug- 
gested for grade separations in the HSR corridor, only 135th Street/Romeoville Road with ADT of 
7,000 could be considered as having moderately high volume. This crossing is located along the 
IC/SPCSL alignment near Romeoville. No more than six existing crossings are proposed for separa- 
tion along any one alignment. 

Along the new track that will be constructed for the Green Grass alternative alignment, 13 grade 
.separations are proposed. The other 10 potential crossings along this new alignment are suggested for 
closure. 

Safety 

Accident estimates were developed for highway-railroad at-grade crossings in the HSR corri- 
dor to evaluate the potential effect of the proposed warning and protection device improvements for 
at-grade crossings that are included as part of the HSR alternative. The U.S. Department of 
Transportation's procedures, as presented in the Railroad-Highway Grade Crossing Handbook were 
used to develop these estimates. These procedures evaluate the vehicular and train traffic using the 
crossing, train speed at the crossing, highway type, and other crossing characteristics to estimate the 
number of accidents that will occur at a single crossing. The accident estimates were developed for 
existing conditions, the No-Build alternative, and each of the HSR alternative alignments. 

The results of this analysis indicate that implementation of HSR service with suggested grade 
crossing treatments would substantially reduce the projected number of accidents occurring at high- 
way-railroad at-grade crossings. The estimated number of accidents at grade crossings per year under 
the HSR alternative is projected to decline regardless of which alignment is selected. When con- 
ducting this analysis, future traffic growth was not considered when developing estimates for the No- 
Build and HSR alternatives. As a result, the estimated number of accidents along the entire IC/SPCSL 
alignment is lower under the No-Build alternative than under existing conditions because the num- 
ber of freight trains operating along this alignment in the future could be reduced as a result of the 
UP/SP merger. If traffic growth were considered in this analysis, the estimated accidents for the No- 
Build and HSR alternatives would be higher. However, estimated accidents along the HSR alterna- 
tive alignments would still be lower than existing and No-Build conditions. 

The result has been an acceptance of the grade crossing ideas for the project by the public that 
allows continuing further project development. 



Rya-A-Raii 



'® 




Road-to-Rail Conversion Systems 

y Mounting versatility; only unit able to 

mount behind front axle 
/ Three (3) models available; front cind rear 

units sold separately or together 
v^ Two types of braking systems: Cobra or 

internal shoe 
y Safety locks manually lock cylinders up or 

down 
/ Adapts to most standard industrial vehicles 
Z Factory-available replacement parts 






'■^-'-^' 



The Most Efficient Rail Car Mover 

y No complicated weight transfer 
/ Eight (8) models: 24,500 - 55,000 lbs. 

draw bar pull (from one coupler) 
/ Better traction and braking 
v^ Large capacity compressor available 
y Optional remote control 
y Free on-site application survey 
y Best factory support 
/ Lease options available 



Phone: Cciitral MaRuf acturfng, Inc. f^- 

(816) 767-0300 4116 Dr. Greaves Road Grandview, Missouri 64030 (816) 763-0705 



Tamping^ 
Tools 

• Wear Skids 

• Adzer Bits 

• Undercutter 
Parts 

► Regulator 
Wear Parts 



weab£ 



Manufacturer of 

Tungsten Carbide 

Wear Protected/ 




i 34 Goodwin Drive 
Crystal Cin^ Missouri 
I 63019 USA 

FAX 31lb937-338« 

314-937»3326 



Q^yffiWSk NEVER WEARS OUT. 



275 



Rail Wear Management at Conrail 

By: W. L. Heide* 

It is a pleasure and privilege to be here to present to you some of the methods of rail wear man- 
agement at Conrail. To begin, there have been some people referring to the phrase that relates to using 
a material or a production process to the fullest extent; that is, do not replace or do before its time. 
That is in essence the philosophy of our rail program at Conrail at this point in time. That is not to 
say that this is the only way to look at a rail management program, because there are other concerns 
such as need for cascade rail or relay rail for different situations. This philosophy is not the same as 
saying don't fix it until it's broken. The whole purpose of rail wear management is to extend the life 
of the asset. 

Rail is the most expensive maintenance of way asset that we have to maintain and, therefore, a 
lot of effort is put into making sure that we use our dollars wisely. To begin with, it is important that 
rail be maintained, not just left to chance in various operational and environmental situations. What 
I will present to you today are some of the methods by which we maintain our rail, and then I will 
discuss how we determine what rail should be replaced. This will include our policy for use of rails, 
rail grinding, lubrication, fastening systems, wear limits, rail defects, rail wear measuring, use of 
models and rail maintenance planning. 

There are many ways at Conrail in which we strive to get the most from our rail. One of these 
is the use of good quality steel rail according to our specifications. We install it where we believe we 
can optimize the life cycle. At Conrail, there have been many different rail sections in main and 
tracks. These have consisted of 127Dudley, 130RE, 133RE, 136NYC, 140RE, 152PS and 155PS. 
From the beginning of Conrail in 1976, the rail section chosen for main line use was 132RE. In 1991, 
it was decided to change the rail section to 136-10, to be more in line with most of the other Class 1 
railroads who were adopting this section and were proposing the change in the AREA manual. The 
standard 1 36RE rail had a 1 4" head radius and the newer 1 36RE rail has a 1 0" radius to be more com- 
patible with the wheel/rail interface. At Conrail, our policy is to use new rail in main track locations 
where the tonnage is normally greater than 15 MGT. Standard 300 Brinell Hardness rail is laid in tan- 
gent and curved track up to 2 degrees. Premium, fully heat treated or head hardened rail at 350 Brinell 
Hardness or more is used in curves over 2 degrees. Fit rail is used in branch lines, secondary lines, 
sidings and yards. There are, of course, exceptions to this policy and those decisions are made at the 
system Chief Engineer level. 

An important maintenance requirement is to utilize rail grinding to remove surface imperfec- 
tions and provide a smooth riding surface, profile the rail head to improve the wheel steering capa- 
bility, and remove possible defect causing conditions that may later cause premature failures. Rail 
grinding has been used for many years and has been shown to be a real benefit in maintaining the rid- 
ing condition of the rail and, in turn, lowers the stress on the other track structure components. By 
having a smoother running surface, the ties and ballast last longer as well as the equipment compo- 
nents. The utilization of the rail grinder on Conrail has been important to us and we take pride in get- 
ting about a 70% average utilization for the year. Our grinder normally does about 10,000 pass miles 
a year. Udlization is defined as the number of production hours divided by the total hours worked. 
We work one grinder 6 days a week on 12 hour shifts throughout the year Another interesting thing 
to note on our rail grinding is that the pass to track mile ratio has decreased from 2.0 in 1988 to 1.1 
in 1996. Our average grinding speed on the other hand has increased from 2.5 MPH in 1998 to 5.4 
MPH in 1996. This tells you that we are getting more in the maintenance mode of grinding, rather 
than a corrective mode. 



* Ass't. Chief Engineer - M/W, Conrail 



276 



Paper by W. L. Heide 277 



Another means of maintaining our rail is in the use of switch and crossing grinders. We have 
found that we can extend the life of our turnout components, such as stock rails, switch points and 
frogs by the use of these grinders. In the past, we relied on our welders to do the grinding on the 
switches and frogs in our interlockings and major yard areas. However, the amount of grinding nec- 
essary to truly maintain those components was more than the welder and his grinder helper could 
manage. The addition of the switch grinder has helped us greatly in this area. The grinder is also used 
to supplement the big grinder in the area of grade crossings where the big grinder may have prob- 
lems with various types of highway grade crossing surface materials. We also closely monitor our 
performance in the utilization of the switch and crossing grinder, and have been able to maintain 
about 70% with this machine, just as we do with the big grinder. About 4,000 pass miles of grinding 
was done in 1996 using 2 switch & crossing grinders. 

Rail wear is also closely related to lubrication, especially in curved track. This is another 
method to extend rail life and has been a major topic of discussion for many years. Well, at Conrail, 
we believe in rail lubrication and we have done testing to show that, not only is rail and wheel wear 
reduction a benefit, but fuel savings is a real plus. We did some locomotive lubrication testing at the 
Transportation Technology Center about 10 years ago, which resulted in a fuel savings in the range 
of 1 1% for a coefficient of friction of 0.20. Dry rail has a coefficient of friction in the range of 0.40 
to 0.50 and well lubricated rail is in the range of 0. 15 to 0.20. We set a goal a few years ago to reduce 
our coefficient of friction to a value of 0.20 in order to get the benefits of fuel savings, wheel wear 
and rail wear. Our track wayside lubricators were checked closely and many were updated to the 
newer hydraulic style, or replaced as necessary, to provide the required amount of grease to reduce 
our coefficient of friction. We purcha.sed for each of our 5 Divisions a tribometer device, which mea- 
sures coefficient of friction in order to be able to monitor our conditions and take corrective actions 
where necessary. This device is a small portable unit which sets on the rail and is manually pushed 
along, providing a coefficient of friction reading as you activate the measuring wheel. It can take 
readings on the gage comer or top surface of the rail, as desired. We are in the process of testing a 
unit which can be towed with a hi-rail truck to measure coefficient of friction on both rails and record 
the data on a computer. Along with out wayside lubricators, we equipped most of our road locomo- 
tives with wheel flange lubricators to supplement our lubrication efforts. They key is to have both 
wayside and locomotive lubricators working in unison to provide the desired lubrication. The type of 
grease to use has always been controversial, because it is very difficult to determine the effects in the 
field. We are presently using both a calcium based and a lithium based grease at various locations. 
Lubrication is important and must be monitored to make sure that enough is being distributed, but 
not too much to cause locomotive wheel slip. The benefits in rail wear improvement have been shown 
in field tests and at FAST. 

As mentioned earlier, 1 want to touch upon the fastening systems that we are presently using at 
Conrail. Most of our rail is laid on standard tie plates with cut spikes. In territory where the tonnage 
is greater that 5 MGT and curvature is greater than 1 degree, we use 3 rail holding and 1 place hold- 
ing spikes per plate. In locations where we install new rail in curves greater that 2 degrees, we use a 
resilient fastening system with screw spikes on wood ties. The types of fasteners may or may not have 
an effect on the life of the rail. We have been using the resilient system for about 9 years now, and it 
is difficult to say that rail life has either improved or diminished. However, we have had good suc- 
cess in reducing rotation of the rail and maintaining gage. 

As we reclaim rail from rail renewal and retirement projects, this rail may be brought back to 
our cropping and reclamation plant to be inspected and classified into different grades to be used for 
main track, branch line or siding and yard rail projects. Following is a chart showing our rail classi- 
fication based on vertical and horizontal wear for the various types of trace usage: 



278 Bulletin 761 — American Railway Engineering Association 



Rail Sections 


Main Track 


Brancli Line 


Siding & Yard 


(1) Vertical Head Wear 








155.140,136,133,119 


5/16" 


8/16" 


10/16" 


152, 132, 130 


4/16" 


6/16" 


8/16" 


131, 127, 115, 100 


3/16" 


5/16" 


6/16" 


112, 107, 105 


2/16" 


3/16" 


4/16" 


(2) Horizontal Head Wear 








155, 152, 140, 136, 








133, 132, 130 


6/16" 


9/16" 


12/16" 



131, 127. 119, 115, 112, 

107, 105, 100 4/16" 7/16" 10/16" 

The chart is used for classification of rail in the reclamation plant and not to condemn rail in 
track. It has always been difficult to determine exactly what vertical or horizontal wear limit should 
be used to remove rail from track. Our experience over the last several years has led us to conclude 
that the rail can stay in track longer than previously thought. Many of the older recommendations for 
removal suggested that when the head loss was greater than 25%, it was time to remove the rail from 
main track. We have found that it can actually wear safely to as much as 50%, before having to be 
replaced. This can amount to more than 3/4" vertical wear and 5/8" horizontal wear. 

Rail wear management not only has to be concerned with wear, but also with internal defects. 
We utilize inspection cars and trucks to inspect rails for internal defects and, of course, we do this on 
at least an annual basis as required by the FRA. Rail defects can tell you something about the condi- 
tion of the existing rail and, as part of our rail strategy, we look at the number and types of rail defects 
that are found in different types of tracks. Based upon these findings, we use this to help predict and 
plan rail replacement. 

In order to determine what the life of the rail is in track, there are difference ways of measuring 
the wear. You can rely on going out in the field and doing a hand measurement of the rail using a tem- 
plate and taper gauge. This can give you a vertical and horizontal measure of the rail wear, however, 
you must be careful to make sure the rail you are measuring is typical of the condition of the rail in a 
curve or location. There are many devices that can be used in the field to get spot readings of the rail 
wear, and this information can be very useful to the field engineer to confirm data he may have 
received from other measuring systems. At Conrail, our main measuring system for rail wear is 
mounted on our Rail Analyzer Car. The Rail Analyzer Car is part of our track geometry car consist 
and, therefore, gets over most of our main tracks on a quarterly basis. The system we are not using 
consists of lasers and cameras, which is capable of measuring the rail profile at any speed that we oper- 
ate. Data is collected on board the car and brought back to the office for processing. After the data is 
processed, we then have a good reference to look at to determine those locations that are exhibiting 
the most rail wear. Rail wear profiles and plots are used by field people to keep watch on those loca- 
tions and provide feedback to our planners for possible rail renewal projects. The rail measuring sys- 
tem has been very beneficial in enhancing our rail replacement strategy, because of the enormous 
amount of data that can be generated and utilized to develop an intelligent planning process. 

That brings us to how we plan our rail renewal program. Throughout the year, we are constantly 
watching our rail conditions. Our field people monitor their territories closely and prioritize what rail 
they recommend on the program for the following year. Their basis for decision is the existing con- 
dition of the rail surface, the rail wear, the types and number of rail defects and, in some cases, the 
weight of existing rail. From a system level, we utilize more scientific methods to determine the rail 
program. Presently, we have available to us a rail model which can predict the life of the existing rail 
based on tonnage, wear and defects. From the model, we can then compare the proposed rail renewal 



Paper by W. L. Heide 



279 



requests from the field to finalize our annual rail program. The model can work very well in provid- 
ing a program which can span many years, however, it still requires some feedback from the field to 
make sure we have not overlooked a condition that may not be seen by the measuring systems. 
Conditions such as corrugation, bent joint areas, weld conditions, etc., cannot always be measured. 
Utilizing all the data input and field priorities, a program is then finalized to meet the capital expen- 
ditures that are allocated for rail. This final rail program is then made a part of what we call our 
Yellow Book, and this is used throughout the year to identify all of the capital production work which 
will be done on the Conrail system. 

To conclude, you can see that there are many functions that are involved in managing rail wear and 
an annual rail program. 1 have tried to give you some insight as to how we handle rail wear management 
at Conrail by looking at our policies for use, rail grinding, lubrication, fastening systems, wear limits, rail 
defects, rail wear measuring, rail models and planning. All of these factors are important in determining 
when and where rail should be replaced to maximize the life cycle cost of the rail asset. 



MAGNUM 

CONCRETE GRADE CROSSING 

• Manufactured to fit 
any rail ranging from 
115 lb to 136 1b. 

• Designed for wood 
and concrete ties. 

• Low maintenance — 
long wear. 

• Custom designed for 
switches. 

• Insulated crossings 
available. 

MAGNUM MANUFACTURING CORPORATION 

(801) 785-9700 • FAX (801) 785-9701 

Manufacturing locations in: / \ 

Pleasant Grove, Utah and Everman, Texas MAGNUM 




THE CHANGING ROLE OF THE CONSULTANT 

By: Gary Griggs* 

Opening 

Deregulation, devolution, privatization, global competition, downsizing . . . 

Design/build, public/private partnerships, contracting-out, outsourcing . . . 

A new way of doing business. ... A changing role for the consultant . . . 

I have spent my entire career in the consulting engineering business — starting with a high school 
summer job as a surveyor relocating railroads along the Columbia River — to today — heading up the 
largest transportation engineering company in the U.S., Parsons Brinckerhoff's Infrastructure Company. 

For most of my professional career I have been called upon to provide quote "traditional ser- 
vices" . . . i.e. planning, design, and construction support services to a client for a specific project. It 
has always been a trusted role of owner's engineer . . . working closely with the client's engineering 
staff — often quite sizable — ultimately resulting in a facility to be owned and operated by the client. 

Today, we in the consulting engineering field are asked more and more to work outside of that 
box — to apply new paradigms — to take on new roles, to assume more risk, to provide more cost 
effective services. . . Today we own projects, we finance them, we build them and even operate and 
maintain them. A very different role from the "traditional" consultant. 

There are several factors which I think are bringing about this change. They are: 

• The tremendous unmet infrastructure needs in this country and a serious disinvestment in that 
infrastructure. 

We have all seen the numbers. Our investment in infrastructure as a percentage of GDP is one 
of the lowest of the developed countries, and continues to decline in major market sectors. Our 
railroads are in serious need of upgrading, forty percent of the highway system is in need of 
repair, and 60% of the bridges in this country are substandard. Our safety is at risk and our 
national and global competitiveness are at stake. We must find ways to solve these problems, 
and we in the consulting business are being asked to help find solutions. 

Other factors include: 

• A new federalism with the goal of reducing the federal role in funding programs, passing the 
responsibilities to the local level (so-called devolution). 

• A desire to streamline and downsize government at all levels. 

• A revolution of anti-tax sentiment at all levels and a loss of faith on the part of the public in 
government's ability to respond to their needs and spend their tax dollars wisely. 

• A new transportation funding approach which emphasizes intermodality, efficiency, innova- 
tive finance, and local decision making (ISTEA and NEXTEA). 

In the late seventies and early eighties I was actively involved in a number of federally-funded 
railroad programs, and I'm sure many of you were also: The multi-billion dollar Northeast 
Corridor Improvement Project (NECIP) upgrading the vital rail link between Washington, DC 
and New Haven, Connecticut; and the Pueblo Test facility upgrade for testing of AEM-7s and 
other equipment to be used on the NECIP. What a tremendous project, introducing 125 mph 
operations. At one time the Pueblo facility was a hot bed of advanced research pursuing the lat- 
est technology, technology which was to keep the U.S. in the lead — for example: air cushion 
and linear induction technology. 



* President, Parsons Brinckerhoff Quade & Douglas, Inc. 

280 



Paper by Gary Griggs 28 1 



But it was the new federalism of the 80's driven by the anti-tax revolution of the public that 
changed the public sector programs. Fortunately, the Northeast corridor upgrade continued, 
albeit to a more limited extent than originally planned; and in fact continues today on the north- 
end electrification from New Haven to Boston — always subject to the yearly appropriation bat- 
tles and commitment to Amtrak as a necessary service provider. 

The revolution was driven somewhat by the premise of applying private sector approaches to 
what has traditionally been public works. It was argued that if private sector can't do it, it's not 
worth doing. I am not here to pass judgment on whether the premise is right or wrong, but rather 
to report this as a significant external factor that continues to change the way we do business. 
It may well be driven by broader market forces, and it can certainly be argued that there have 
been many benefits in terms of productivity and global competitiveness. But it has also brought 
about a new way of doing business. 

In addition, funding responsibility is being passed to state and local governments. A good exam- 
ple is California where a dozen so-called "self-help" counties voted to tax themselves to fund 
vital programs. And even that is becoming more difficult to do in light of the anti-tax revolu- 
tion manifested in super-majority approval requirements. 

Other factors affecting change in our business include: 

• An increasing need to do more with less focusing on management of existing infrastructure 
with greater efficiencies and longer life cycles. 

• Deregulation resulting in increased competitiveness in all sectors of our economy — the rail- 
road industry is an excellent example and a great success story, and 

• Overall increasing global competitiveness. 

These factors result in a need to do business differently in order to address the infrastructure 
needs of the country. 

New Project Delivery Approaches 

The lack of traditional funding sources has resulted in a search for alternative means of project 
delivery and, in particular, funding. The idea of integrating the project funding and the project deliv- 
ery system is leading to numerous opportunities in public/private partnerships and privatization of 
infrastructure. With public funds diminishing, the relatively new practice of government working 
with the private sector to forge new business relationships is emerging at a very fast pace. In the past 
when we were called in, we were specifically limited to providing planning, design, and other sup- 
port services. Now we are being asked to finance, design, build, operate and maintain projects. 

Design/build is one of the delivery systems which is replacing the traditional design bid/build 
model. Ten years ago de.sign/build was applied to just 3% of the total U.S. infrastructure market; but 
today it's approaching 30%. The railroad industry has, of course, used it for years. Globally, it has 
long been a more accepted form of contracting and amounts to just over 50% of the infrastructure 
market and is climbing. Interestingly enough, one of my first non-U. S. assignments back in the late 
70's was a design/build iron ore railroad project in Africa. 

Today my firm is involved in many design/build projects including multi-billion dollar pro- 
grams in Southern California, the $1 billion 1-15 toll road in Utah, and the Ford Island Bridge in 
Hawaii. The Federal Transit Administration has promoted design/build through its Turnkey 
Demonstration program, including the Hudson-Berger Light Rail project in New Jersey, the Tren 
Urbano rail transit project in Puerto Rico, and the extension of BART to San Francisco International 
Airport. And, of course, Amtrak's Northend Electrification Project. 

The role of the consultant in design/build is quite different from the "traditional" design bid/ 
build role. First of all, it can be one of several roles: Owner's engineer, contractor's engineer, or engi- 



282 Bulletin 761 — American Railway Engineering Association 



near responsible for design and construction. The latter two roles represent the greatest change in 
responsibilities for the consultant. For example, working as a subcontractor to a prime contractor 
under what is usually a low bid arrangement — very much in violation of the QBX/Brook's Bill envi- 
ronment of consulting engineering; or, as a prime contractor responsible for both design and con- 
struction. We now have a construction division. 

Privatization 

"Show Us the Money": was the headline from a recent ENR article: "On Road, Rail and Air 
Jobs, It's 'Show Us the Money.'" Perhaps the greatest change to the way we do business is in the area 
of finance. Funding constraints in the infrastructure marketplace continue to produce increased client 
interest in innovative finance and project development approaches that promise new revenue sources, 
cost savings and faster project delivery. 

For example: 

• In the last 4 years, 10 states passed legislation to permit new forms of public private partner- 
ships for innovative project delivery, and; 

• 8 states are now establishing federally authorized state infrastructure banks 

• And, it is estimated that there are presently 185 privately financed infrastructure projects 
worth $64 billion dollars presently underway in the U.S. 

Many of the programs are directed toward toll roads, which often have the best chance at pro- 
viding the revenue stream required to cover the long-term debt. In some ways, this approach 
addresses the anti-tax sentiment by making the user pay. However, we have also encountered con- 
siderable resistance to tolls as well. In Washington State the opposition group was called TRUST, an 
acronym for Tolls Represent Unfair State Taxes. 

The consultant's role here can cover all aspects of the project from financing to operations. 
There have, of course, been failures and successes (for example: the Dulles Toll Road, which is strug- 
gling, and the California SR 91, a great success). Perhaps the most notable example in the railroad 
environment is the Channel Tunnel, and we all know of the challenges faced there. 

Taking on the role of developer or project facilitator is something new to most of us in the con- 
sulting business, and it's a role that is not easily assumed. But it does appear to be the way of the 
future and if we want to stay in business, we better join the team. It requires a close relationship with 
the owner and is often a joint undertaking with the owner or public sector working together in struc- 
turing appropriate financings. And like the consultant, some public institutions do well and others are 
finding the arrangement difficult. Owners are finding it hard to give up the control they have tradi- 
tionally held and there are tremendous political risks to the developers. 

It has introduced a whole new vocabulary for the consultant: 63-20 tax-exempt regulations, 
non-recourse financing, senior and subordinated debt, sweat equity, etc. 

Operations and Maintenance 

Operations and maintenance is also becoming more and more a part of our business. My firm 
is presently operating portions of the Oriando-Orange County Expressway Authority toll road in 
Florida — an assignment we would not have envisioned several years back. Again, owners or public 
sector clients are asking us to expand our role, and again it means a new role for the consultant. Often 
the O&M is tied to a broader project role, such as the Hudson Bergen Light Rail Project in New 
Jersey, where in addition to utilizing a design/build contracting approach, the contractor is also 
required to operate and maintain the system for 15 years. This will help to insure that the owner gets 
a system which works. 



Paper by Gary Griggs 283 



Outsourcing 

It appears that the increasing competitiveness in the marketplace is driving many of our clients 
to downsize their engineering staffs and contract out more. It is often argued that there are greater 
efficiencies to be effected by using private sector. These include reducing the need to maintain large 
staffs during busy times and slow times. Consultants can level their demands by virtue of having 
many projects. Consultants also often have a larger base to draw from for bringing resources and 
expertise to the client, perhaps better than what can be made available through the client's organiza- 
tion. It also puts a great responsibility on the consultant to provide the same level ofexperti.se previ- 
ously available. This has been especially true with the railroads given the deregulation and consoli- 
dation impacts, and the downsizing of the railroad engineering forces. Much of the downsizing has 
resulted in retirement of some of the most experienced among us which further emphasizes the need 
for training and education of those who follow. I know AREA is actively involved in assuring that 
this potential gap is filled. We in the consulting business do not take lightly the responsibility we 
assume in this new relationship. 

And it is not a trend that is without debate. While we have seen the trend in many client orga- 
nizations across the country, there are some that are resisting. Perhaps the most notorious being the 
Professional Engineers in California Government who have been resisting the contracting out of ser- 
vices to private sector for many years. In fact, there will more than likely be a public vote this 
November on a PECG initiative titled "Government Cost Savings and Taxpayer Protection 
Amendment" which, if passed in favor of PECG, could result in the virtual elimination of the con- 
tracting out program by the State of California. We also have unions to contend with. 

Increased Competitiveness 

The original "Canons of Ethics" adopted by the ASCE Board of Directors years ago read: 

• "The engineer will not advertise his work in a self-laudatory manner. . ." 

• "The engineer will not try to supplant another engineer in a particular employment. . ." 

• "The engineer will not compete with another engineer on the basis of charges for work. . ." 

Today we in the consulting field operate in a tremendously competitive environment — we 
advertise, we openly compete against each other, and contrary to the Brooks Bill requiring quality- 
based selections we are often required to bid our services. What has brought about this change? I 
don't think, by and large, that we in the consulting engineering field have changed our basic beliefs — 
and I am happy to say that most of the original canons still stand. But our owners and clients are being 
faced with limited resources to undertake tremendous tasks, and we must also do everything we can 
to be as cost effective as we can. 

Further, under the pressure of greater efficiencies and cost reductions, the relationship with our 
clients is far more businesslike, based on tighter constraints, greater risks, and perhaps, just perhaps, 
a little less trust. 

It seems to me that we in the consulting business are experiencing some of the same changes 
as the railroads. We are consolidating, expanding, and increasing our competitiveness and produc- 
tivity. We also outsource those services that we feel can be more efficiently done by others. 

Close 

We are experiencing a "new way of doing business"! It is competitive, it is global, and it is 
implementing innovative project delivery approaches in order to respond to our tremendous unmet 
infrastructure needs. I, for one, am exhilarated by the changes and challenges facing us all. 



Who in the world can respond 

to your requirements for specific 

track equipment? 




In addition to over 130 different machines we offer 
for railway track maintenance, we also develop new 
ideas, and deliver to your unique specifications. 

We adapt standard equipment to include custom 
options. Design new machines to solve our 
customers' problems. And remanufacture entire units 
or assemblies. 

This requires strong, broad design expertise in 
mechanical, hydraulic, electronic and 
software solutions. Fairmont Tamper 
has it. And our strong customer 
base— over 150 around the world— is 
proof that we deliver. 

In fact, Fairmont Tamper offers 
the most complete line of railway 
maintenance equipment in the world. 
Our line includes: • Tampers 
• Ballast regulators, brooms and undercut 
ters • Rail grinders • Track construction 
and renewal systems • Tie removal and 
insertion machines • Spike drivers and 




pullers • Rail anchor and fastener applicators 
• Hy-RaiP Guide Wheel Attachments 

We also provide track renewal, rail grinding and 
other contract services. 

You'll find full-service Fairmont Tamper facilities 
on three continents, and export agents around the 
world. Plus worldwide technical/training support from 
our field service staff. 

In North America— call 
803-822-9160 or fax 803-822-7471. 
Australia— call 7 2056500 or fax 
7 2057369. U.K.-call 602 3844004 
or fax 602 384821. 



Examples of new ideas developed by 
Fairmont Tamper for track mainte- 
nance needs around the world include 
a transit rail grinder, a new track 
construction machine, the Pony (track 
renewal machine) for japan, section 
gang vehicles for Mexico, and over- 
head maintenance vehicles for China. 



F airmam 
tamper 

l») a harsco companq 

Your partner along the way 



NORFOLK SOUTHERN'S ENGINEERING DEPARTMENT 
ACTION PLAN FOR SAFETY 

By: P. R. Ogden* 

We have just heard a very interesting and firsthand recap of what it was Hke for United Airlines' 
Captain Haynes and his crew on Flight 232. 1 suspect many here remember the heroic feats of Captain 
Haynes, his crew, and many other people involved in the work after the crash. 

When we talk about safety of operations on the railroad we don't have a story to tell with heroes 
like Captain Haynes and his crew but we do have a story to tell about how we have, can, and will 
make the rail industry a work place free of personal injuries and fatalities. The challenge before us is 
not an easy one but the returns for our efforts and investments are enormous. 

Today I am here representing some 24,000 Norfolk Southern employees, and more specifically 
5,750 Engineering employees. We have made safety on Norfolk Southern our No. 1 priority and, in 
the next few minutes, I will discuss with you why and how we changed our corporate culture as it 
pertains to safety of operations. 

To best understand and describe to you NS's safety process and performance in recent years we 
have to go back to 1987. 

Safety Ratio — No of FRA Injuries x 2000.000 M. Hours 
No. Of Manhours Worked 

One of the measurements we have for safety in the Rail Industry is the FRA injury rate per 
200,000 man hours. For an example, a group with one injury and 200,000 man hours would have a 
safety ratio of 1 .0. An injury as defined by FRA occurs when one or more of the following conditions 
exist after an accident: 

• Employee is given a prescription for medication 

• Employee is placed on restricted activity 

• Employee cannot work (lost time) 

In 1987 the employee injury rate on NS was 6.14 and for the Engineering Department it was 
worse, as we had a safety ratio of 9.68. For the System, overall performance was flat compared to 
1986; No Improvement, and the Engineering side of the house was 11% worse than the previous year. 
We were going in the wrong direction and as you might imagine Management was not happy with 
this type of performance; consequently, no one was satisfied. 

In the Harriman Award competition, which recognizes the top three railroads of each class, we 
were not even close to any recognition. 

We have always considered safety of operations an important part of our overall operating plan. 
After our poor performance in 1987, a decision was made to go outside our company for help. 

Our safety process as we know it today began in 1988 when we contracted with the Dupont 
Safety Services for an assessment of our safety performance, safety process and recommendations 
for improvements. Dupont historically has had, and continues to have one of the best safety records 
in the world with a safety ratio consistently in the range of .50 (one injury per 4(X),000 m.h. worked). 

Dupont sent us a team of 3 people. After interviewing several of our top management people, 
including the CEO, the team spent about 4 or 5 months on the property in the field. 

Their initial assessment identified a number of areas where we had deficiencies and obstacles 
that prevented any real improvements, such as: 



*Vice President Engineering. Norfolk Soulhem 



285 



286 



Bulletin 761 — American Railway Engineering Association 



• Employees were committing too many unsafe acts 

• Unsafe conditions that could cause serious injuries 

• Rule violations 

• Lack of Safety awareness 

• Willingness to take chances 

• Insufficient safety training 

• Production apparently took precedence over safety 

• Supervision generally took a passive role when it came to safety 

• Operating and Safety rules were perceived by employees to be a primary tool for discipline 
and not guides for good safety performance 

• Rules were not uniformly applied across the system 

In the eyes of a very safety-focused group like Dupont we had a lot of problems, but with their 
assessment, we also had many issues identified and defined to target for improvements. 

It was also obvious after receiving their assessment that we had to change the way we managed 
our safety process. Some of the changes would be easy; others would be very difficult. 

Change we did, though, and the results were immediate. 

In 1988 we reduced our reportable injuries by almost 10% and finished second in the Harriman 
award competition winning the Silver certificate. In 1989 NS won the Harriman Gold certificate with 
the lowest number of employee injuries for all Group A railroads, finishing the year with a safety ratio 
of 4.89. 



NOnFOLK 

^S System Injury History 
FRA Reportable Injuries 




1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 



Exhibit 1 — System Injury History 

NS now has won the Harriman Gold for seven consecutive years, 1989 through 1995. The 
results for 1996 are not in but we hope to make it eight consecutive years. We finished 1996 with a 
safety ratio of 1.25. From 1987 through 1996, we have reduced our safety ratio from a 6.14 to 1.25, 
an 80% improvement. 



Paper by P. R. Ogden 



287 



As for the Engineering Department, which includes Maintenance of Way, Communications & 
Signals, Design & Construction, and Maintenance of Equipment we have experienced an even bet- 
ter rate of improvement and success. 



NSCngineering Department History 
FRA Rep ort able I nj uries 



S 600-H 



^ 400. 




1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 



Exhibit 2 — Engineering Injury History 

In 1988 we achieved a 15% reduction in reportable injuries compared to 1987 and have had 
continuous improvement every year since. In 1996, we finished with a safety ratio of 1.22, which is 
an 87% improvement over 1987, when we had a safety ratio of 9.68. 

We have also had the best safety ratio of all the Group A Railroad Operating Departments — this 
being Mechanical, Transportation, and Engineering — four of the last five years. 




Exhibit 3 — NS Engineering vs. Other Engineering 



288 



Bulletin 761 — American Railway Engineering Association 



When compared to just the Engineering Departments of all Group A railroads, NS has had the best 
safety ratio since 1991, which was the first year the AAR kept the safety statistics by departments. 

The lowest bar represents NS's safety ratio for each year 1991-1996. 

These statistics are not presented to show we are better than anyone else, and we are not boast- 
ing or bragging. Yes, we are proud of these achievements as I believe we should be. These statistics 
mean two things much more important than numbers and awards. 

1. Our employees are getting injured less today than before, and we have a much safer work 
place because of our efforts to improve. The real winners are our employees who stay 
healthy and on the job, the communities that we serve and our customers. 

2. The statistics also demonstrate what one can achieve when you focus on an issue, develop a 
plan of attack, devote the resources and have the determination to make the plan work. 
Safety is manageable. 

I know many of the other railroads can also stand here and cite a good rate of improvement, and 
so can the rail industry. 




Exhibit 4 

The above chart shows the improvements in personal injuries for the Rail Industry. We some- 
times as an Industry take our lumps regarding rail safety when in fact we have a good story to tell. 
At times though, it is not portrayed as such in the media. Yes, we have room for improvements but 
we have come a long way and we are headed in the right direction. The rail industry, as you can see 
from this chart, has better than a 50% reduction in employee personal injuries from I991-I996. 



Paper by P.R. Ogden 



289 



Injury Rate - US Industries 




Dupont Ess? 


1 
1 




" 


i 
1 




Railroads -( r- 

- ■ - - 1 1 1 


- 1 • 1 i 1 1 






9 














2 4 6 8 10 12 14 16 
Injury Rute per 200,000 Manhoim 



Exhibit 5 

Also look at the chart comparing the rail industry ratio for occupational injuries against other 
U.S. Industries. These statistics were taken from the U.S. Bureau of Labor Statistics. As you can see 
the rail industry has a better safety record for personal injuries than many other industries. Again, as 
an industry, we have made improvements and are moving in the right direction. 

After we received the initial assessments from Dupont in 1988, NS made many changes. But 
most of the changes can be summed up in two basic changes. 

• Corporate Safety Philosophy 

• Safety Action Plans 

Let's first talk about the corporate safety philosophy. 

The first block in the foundation for a successful safety effort begins with a commitment to the 
process throughout all levels of the organization. This commitment must begin at the very top. with 
the CEO. It must be a visible commitment and one that is communicated to all employees for the 
process to work effectively. You must have commitment from the top to the bottom. It has to become 
a way of life for everyone. You sell it every day throughout all phases of the work activity by setting 
the right example. 

The commitment by top management establishes the importance of safety and assures support 
for the individual component of the whole process. For the best results throughout your organization, 
management must believe that safety is equal in importance with all other business parameters. It 
must be just as important as cost, productivity, quality, employee relations, and customer .satisfaction. 
Not only must this philosophy be stated — it must be demonstrated. 

For NS to improve, Dupont recommended that management establish a safety policy that would 
spell out all the principles that would govern all decisions regarding safety. Without such a policy, 
safety tends to be pushed aside when other business concerns become pressing. 

After reviewing Dupont's assessment and their recommendations, establishing a .safety policy 
statement seemed to be the first step we had to take to improve our process. 



290 Bulletin 761 — American Railway Engineering Association 



Dupont had a philosophy in which their safety process was built around six main principles. It 
was their recommendation that we use the same six principles and we did. 

We took their advice, made a few minor changes, and added some language that best fit our 
management style, and established a safety policy with a statement that reads as follows: 

Statement of Policy Safety 

"The Norfolk Southern Corporation is committed to the principle that safety is good business. 
No one should be exposed to unnecessary hazards and risks. 

Responsibility for safety and environmental stewardship cannot be transferred. Each employee 
of this Corporation, therefore, is held personally accountable for his/her actions on the job." 

1 . All injuries can be prevented. 

2. All exposures can be safeguarded. 

3. Prevention of injuries and accidents is the responsibility of each employee. 

4. Training is essential for good safety performance. 

5. Safety is a condition of employment. 

6. Safety is good business. 

We call these principles our six tenets of safety. 

The first tenet of safety "All Injuries can be Prevented" is the toughest point to get everyone to 
buy into. We are going into our tenth year with this as one of our basic points in the safety process 
and we still have not convinced everyone that this is a realistic belief. However, more and more of 
our employees believe this each day. Why? There are several reasons: 

• Our employees, for the most part, will tell you they are now convinced that NS is sincerely 
committed to this belief. 

• The numbers are also more convincing today than they were ten years ago. Two examples are: 

i.e. — First, in the Engineering Department, we have reduced our yeariy FRA reportable 
injuries over the last decade by 89% — from 62 1 to 7 1 . Only 7 1 injuries last year. With 
such a low number it is more realistic to believe we can eliminate the final 71. 

— Second, when putting our action plan together for 1997 we noticed that over the last 
three years all but five of our supervisors had gone at least one calendar year with 
zero injuries. 

Also in 1993, we experienced a 500 year flood in the Midwest as did other railroads. NS had 60 
miles of track under water at one time and over 70,000 feet of track washed out. We had over 500 peo- 
ple working in the worst of conditions. The repair work was completed without any injury incidents. 

So yes. Zero Injuries may be a difficult principle to believe, but it is the cornerstone of our 
approach to safety. It governs our attitude and is the foundation which we use to support all the other 
points in our policy statement. 

There are five other safety principles in the policy statement, each critical to the process. For 
the purpose of today's discussion, I will comment on two more of the remaining five points. 

Safety is a Condition of Employment 

Safety is an important aspect of our assessment of an employee's work performance. This 
counts in a person's chances for promotion and merit raises. Also, when there is persistent dis- 
regard for one's safety performance and when all other means have failed, as a last resort dis- 
cipline may be necessary. The last of the six tenets is: 



Paper by PR. Ogden 291 



Safety is Good Business 

First, prevention of injuries is a part of our business plan because it is morally the right thing to 
do. Second, there are many benefits to a work force that works together as a team to improve their 
safety performance. This team concept spills over into all other phases of our jobs and will pro- 
duce many positive results, including an increase in productivity and efficiency of operations. 

We have learned that when this change in safety philosophy is documented, we have a clear, 
unchanging guideline which acts as a standard reference for everyone. This establishes an effective 
working safety policy that is known, understood, and accepted by all employees. 

As I previously mentioned, this policy must place safety at least an equal with all others pieces 
of your business plan. 

This policy must be: 

• Set by top management 

• Written — simply, and in general terms 

• Understood by all employees 

• Actively promoted throughout the organization 

Our safety policy, as you have seen, is written in clear and concise terms. We believe it appropri- 
ately addresses the key elements of accountability, preventability, training; and that safety is a condition 
of employment. The policy also expresses management's belief that all injuries can be prevented and 
that safety is good business. 

Safety Action Plan 

The second part of turning our safety program around was to develop a plan of action. Dupont 
in the course of developing their process over many years identified twelve essential elements that 
must be in the foundation of any effective safety management program. Dupont's twelve essential 
elements were as follows: 

1 . Management commitment 

2. Documented safety philosophy 

3. Safety goals and objectives 

4. Committee organization for safety 

5. Line responsibility for safety 

6. Supportive safety staff 

7. Rules and procedures 

8. Audits 

9. Safety communications 

10. Safety Training 

1 1 . Accident Investigations 

12. Motivation 

As we worked to develop a plan of action and discussed each of the elements with Dupont, we 
found we already had most of them in place over the system, but not with any consistency. This 
inconsistency also made it clear, that to achieve the level of improvements desired, we had to develop 
a plan that would give the NS team the goals to manage safety the same as we manage production 
and costs. Safety must be a line responsibility, the same as operations. It is not just the responsibility 
of the Safety Department. 



292 Bulletin 761 — American Railway Engineering Association 



From this review process and with Dupont's recommendations the team put together a plan 
titled "Norfolk Southern's Six-point Action Plan for Safety of Operations." 

This is a corporate plan to be used as a guideline, for all facilities and departments to manage 
safety the same as any other part of their business. It not only included the twelve elements recom- 
mended by Dupont but included some other issues that were unique and important to us and which 
we felt needed to be included in an NS plan. 

The six points of the NS plan area: 

1 . Safety Policy and Goals 

2. Education 

3. Communication 

4. Recognition 

5. Enforcement 

6. Accident & Injury Investigation 

Using the NS six-point action plan as a guide, each major Operating Department put together 
an action plan for safety that included these six key points and other issues which were unique for 
each Department. 

This plans gives our Supervision an excellent guideline to manage the safety process. 

I will go through the outline of the plan quickly to give you a feel for the subjects covered under 
each point. 

The first point is: 

1 . Safety Policy and Goals 

.01 — Personal Policy Statement 

.02 — Safety Performance Goals 

.03 — Safety Performance Evaluation 

One of the keys to our safety process is that all supervisors must issue a personal policy state- 
ment annually by January 15. This statement reaffirms our commitment to a safety operation, out- 
lines job performance, and expresses a genuine concern for the safety and health of all employees. 
The statement will also convey the safety goals for individual groups. 

The second point is Education. 

2. Education 

.01 — Safety Training 
.02 — Safety Audits 

.03 — Safety and Operating Rules Examinations 
.04 — Job Briefings/Job Assessments 
Here we have guidelines to cover: 

• Safety Training 

• Safety Audits 

• Safety and Operating Rules Examination 

• Job Briefing/Job Assessments 



Paper by P.R. Ogden 293 



I don't think there is anything new here, or unique, that you are not familiar with. We ail know 
the importance of safety audit, job briefings, training, and educating our work force. 

Training is more important than ever. We have an extensive training program at our 
McDonough Training Center in Atlanta which covers a variety of subjects with the Maintenance of 
Way, Signal, and Train and Engine employees. Here you see a class being conducted for our new sig- 
nal maintainers. About 500 Engineering employees attend one of the workshops each year. 

In addition to McDonough, training on the job is an ongoing process through safety audits, 
safety meetings, and individual job performance evaluation as needed. 

The third point is Communications. Here you will find guidelines for: 

3. Communications 

.01 — Safety Promotion 

.02 — Safety Committees/Meetings 

.03 — Job Briefings 

.04 — Individual Involvement 

.05 — Quarterly Personal Contact 

As with Education, we all understand the importance of good communication. Safety has to be 
sold every day the same as any other part of your business plan until it becomes a frame of mind. 

The two most important items are: First to get each employee involved in the process of pre- 
vention; and the second is the necessity for personal contacts between supervision and each employee 
under their jurisdiction. This type of contact provides positive reinforcement and an opportunity to 
check for a clear understanding of our safety practices. 

The personal contacts are done in a number of different ways: 

1. Hy-Rail Inspection Trips 

• 6 per year over our system 

• Hy-Rail in lieu of train and office cars so we can stop and talk with our employees on the 
job site. 

• 15 to 20 of our top managers 

2. Hy-Rail Track Inspection Trips 

• Stop and talk with all the employees we meet or pass on the right-of-way. 

3. Walk through inspections of our shops, plants and other facilities. 

• We use these opportunities to talk with the employees individually or in groups. 
The fourth point is: 

4. Recognition 

.01 — Recognition of Safety Performance 

.02 — Family Recognition 

This is simply a means to involve families in the workers' community when possible and to give 
recognition in many ways to individuals and groups when their performances have contributed to the 
improvements and the success of the process. 



294 Bulletin 761 — American Railway Engineering Association 



The fifth point is: 

5. Enforcement 

.01 — Safety Accountability 

.02 — Safety Discipline 

We hold supervisors responsible for the prevention of injuries and accidents. But, each 
employee is also held personally accountable for safety performance, regardless of position. 

In most cases with the right leadership, training, audits, job briefings and on-the-job counsel- 
ing as needed, discipline is never necessary. Unfortunately, sometimes discipline has to be used as a 
last resort when an employee fails to respond to all other training and educational efforts. 

The last point in the plan is: Accident and Injury Investigation. 

6. Accident and Injury Investigation 
.01 — Medical Attention 

.02 — Drug and Alcohol Testing 

.03 — Notification 

.04 — Cause Analysis 
The points covered here are several. 

First, when an employee is injured the primary concern is to ascertain the need for medical 
attention and obtain any needed care promptly. 

Once this is taken care of a thorough investigation of the accident is conducted in order to deter- 
mine the root cause and initiate corrective action. 

To sum up our safety process: We view safety as an equal with other parts of our business plan 
such as efficiency and customer satisfaction. Safety is manageable. It is like managing anything 
else — you have to have a policy and an action plan in place. You have to be committed to the process. 
You have to take care of the details. 

Managing safety successfully involves changing the way employees and management think. 
Effective safety management must be an ongoing process, the same as managing our maintenance 
programs. 

We have found that the effort devoted to safety has many good returns, not only in terms of a 
reduction in cost of medical and employee lost time, but also in productivity, quality, employee rela- 
tions, efficiency, and just a better overall operation. In other words. Safety is Good Business. 

Norfolk Southern's Vision is: 

Be the Safest 

Most Customer-Focused and 

Successful Transportation Company in the world. 

There are three pieces to this vision and I suspect some may question whether realistically we 
would or could put safety as our No. 1 business priority ahead of customer focus and an efficient 
operation. 

Well, safety does have top priority on NS and 1 submit that you really cannot fit all of the busi- 
ness pieces of an operating plan together successfully without an effective safety process. The better 
job we do in preventing personal injuries and train accidents, the better position we place ourselves 
to serve our customers and operate an efficient railroad. 



Paper by PR. Ogden 



295 



We have achieved a lot with our safety process since 1987. The immediate goal for the 
Engineering group is to complete 1997 with a safety ratio of less than I.O Somewhere in the near 
future, certainly within the next five years, we want to be at the Dupont level which is in the range 
of .50. A ratio of .50 is one injury per 400,000 man hours or 200 men working one full year with only 
one injury. Our safety ratio in the Engineering Department for the first V/i months of 1997 is .49. 
Long-term, our goal is an injury-free work place. Zero. 

In every company, department, shop, or gang, some attitude toward safety exists. How you 
value safety may not be spelled out, or even conscious, but you do have to work with some assess- 
ment of safety's importance. 

I will end with this thought. 

You will achieve the level of safety that you demonstrate you want to achieve. 



Nolan Rail Products 



Choose from a wide 
range of dependable, 
tested and field 
proven rail products. 
The Nolan Company 
offers the best in long 
lasting, hard working 
maintenance-of-way 
equipment. For more 
information on 
Nolan's full line, ask 
for our full color 
catalog. 




Push Cars 



(he 

NOLAN 

company 



1016 NINTH ST S.W. 
CANTON, OH 44707 
(330) 453-7922 
FAX (330) 453-7449 



64 



IF YOU 
BELIEVE... 







...problems have 
solutions. 

If you value 

performance 

over price. 

And, if you want 

service and 

innovative 

products that 

improve track. 

call: 



PANDROL' 

PANDROl. INCORPORATED 

501 Sharptown Road 
P.O. Box 367 
Bridgeport, NJ 08014 
(609) 467-3227 







296 



EFFECTS OF TRACK MAINTENANCE ON THE 

RELIABILITY OF A SINGLE TRACK RAILROAD LINE 

AS A FUNCTION OF AXLE LOAD 

By: Carl D. Martland* and William E. Robert** 

Abstract: 

This study demonstrates that heavy axle load operations can reduce train delay and improve 
reliability for a single-track railroad line. Compared to the 33 ton base case, operations with 36 ton 
or 39 ton axle loads result in additional track maintenance, but require fewer trains for the equivalent 
net traffic. For the high-tonnage route (annual traffic of 80 million gross tons) examined in this study, 
heavy axle loads result in a slight reduction in train delay and increase in reliability, as the additional 
delay caused by additional track maintenance is more than offset by the reduction in delay from the 
reduction in number of trains. On this route, heavy axle loads cause a reduction in trip time of 2 to 
3% and a reduction in standard deviation of trip time of 5 to 9%. However, for the medium-tonnage 
route (annual traffic of 30 million gross tons) examined in this study, heavy axle loads result in a 
increase of approximately to 1% in annual average trip time, and an increase of approximately 18% 
in standard deviation of trip time. In both routes the analysis indicates more favorable results for 36 
than for 39 ton axle loads in terms of train delay, trip time and reliability. 



1. INTRODUCTION AND SUMMARY 

This study demonstrates the heavy axle load effect on train delay and reliability for a single- 
track railroad line. Compared to the base case with 33 ton axle loads, heavy axle load operations (36 
ton or 39 ton axle loads) result in additional track maintenance, but require fewer trains for the equiv- 
alent net traffic. For the high-tonnage route examined, heavy axle loads result in a slight reduction in 
train delay and increase in reliability, as the additional delay caused by additional track maintenance 
is more than offset by the reduction in delay from the reduction in number of trains. However, for the 
medium-tonnage route examined in this study, heavy axle loads result in a slight increase in train 
delay and reduction in reliability. 

This study was conducted as a part of the AAR HAL Phase II Economic Analysis'^', which 
updates the previous HAL Phase I Analysis''^' based on recent research and revised assumptions rec- 
ognizing increased use of newer track components and improved track maintenance models. The pre- 
sent study addresses the net effect of heavy axle loads on train performance, measured in terms of 
train delay and reliability. 

HAL traffic has two contradictory effects on capacity and delay. To the extent that heavy axle 
load operations lead to fewer trains, heavy axle loads result in fewer delays due to meets and passes, 
and thus allow for higher capacity. However, to the extent that more track maintenance is required 
for heavy axle load operations, more train delays are expected from increased track closure hours, 
thus reducing capacity. 

The approach followed was to update Romps' analysis of effects of track maintenance on train 
reliability'"'. The present study differs from Romps' analysis in its use of track maintenance require- 
ments from the AAR HAL Pha.se II Economic Analysis (Romps' analysis was based on HAL Phase 
I maintenance requirements), and in its use of updated track maintenance productivity rates. The sim- 
ulation model and train operating assumptions are the same as those used previously. 



*Senior Research Associale, Ma.s.sachu.selLs Institute of Technology 
**Graduate Student, Massachusetts Institute of Technology 



297 



298 Bulletin 761 — American Railway Engineering Association 



The study results are expressed in terms of the average trip time per train, the average delay per 
train, and the standard deviation of trip time, a measure of reliability. These results are reported for 
axle loads of 33 tons, 36 tons and 39 tons for a high-tonnage route carrying an annual total of 80 mil- 
lion gross tons (MGT) and a medium-tonnage route carrying an annual total of 30 MGT. For the high- 
tonnage route heavy axle loads result in a reduction of approximately 2 to 3% in annual average trip 
time, a reduction of approximately 8% in annual average train delay, and a reduction of approxi- 
mately 5 to 9% in standard deviation of trip time. For the medium-tonnage route heavy axle loads 
result in a increase of approximately to 1% in annual average trip time, an increase of approxi- 
mately to 9% in annual average train delay, and an increase of approximately 18% in .standard devi- 
ation of trip time. In both routes the trip time and reliability results are more favorable for 36 ton axle 
loads than for 39 ton axle loads. 

The following sections summarize this study of the effects of track maintenance on train relia- 
bility. Section 1 . 1 provides background information relating to the study and Section 1 .2 summarizes 
the study results. 

1.1 Background 

Train delay and reliability are critical factors in the efficient operation of freight railroads. Train 
delay may be defined as excess time required for a train to reach its destination beyond the scheduled 
time. Reliability may be defined as the ability to meet a particular train schedule, variability in train 
arrival times or variability in train travel times'^'. As the traffic on a line increases, delays tend to increase 
and reliability drops. Ultimately, track capacity is limited by the maximum acceptable delay to trains. 

The AAR HAL Phase I Economic Analysis recognized, but did not investigate, track capacity 
or train delay issues. Subsequent to the analysis the AAR sponsored several relevant studies on line 
performance. As part of the research on freight service reliability, Dontula''"'-'^' developed a simple 
line simulation model and used it to examine train performance on a single-track line as a function 
of siding spacing, speed limits and other factors, including the probability of delays related to engi- 
neering or mechanical problems. This study showed that engineering and mechanical problems could 
reduce the capacity of a single-track line by 5 to 10%. 

In general, the interaction between track maintenance and line performance is highly line-spe- 
cific. Dontula provided an example of the effect of engineering and mechanical failures on line 
capacity. He examined a 100-mile, 45 mph, single-track line with six sidings. Under ideal conditions 
(i.e. no engineering or mechanical failures), the operating capacity of this line is 40 trains per day 
(approximately 70 MGT assuming 5,000-ton trains). He calculated the capacity by determining the 
maximum number of trains that could be dispatched reliably within a 5-week period. When he 
assumed one failure per 1,500 train-miles (with an average expected delay of one hour), the capacity 
of the line was 37 trains per day with 92% of the trains arriving on time, assuming that the scheduled 
arrival time was the mean arrival time plus one hour. With a failure rate of one per 1,000 train miles, 
the capacity dropped to 35 trains per day with an on-time performance of 87%. A 50% increase in the 
failure rate led to a 5% reduction in capacity. 

As part of the effort in developing the Total Right-of-way Analysis and Costing System 
(TRACS)I**I, Romps' 'll-l improved the line simulation model and focused on the interaction between 
track maintenance requirements and train performance. A major part of his research addressed the 30 
MGT and 80 MGT coal routes used in the Phase I HAL studies. Romps used TRACS to predict the 
amount of track maintenance and extent of track maintenance windows required on these routes 
assuming steady state conditions. He then used the simulation model to examine the effects on train 
performance for a 292-mile route with 2-mile sidings separated by 12 miles of single track. 

For the 80 MGT route Romps determined that the number of trains per week dropped from 1 63 
for 33 ton axle loads to 146 for 36 ton axle loads and 134 for 39 ton axle loads. Compared to the base 



Paper by Carl D. Martland and William E. Robert 299 



case of 33 ton axle loads, the amount of track time required for maintenance increased 7% for 36 ton 
axle loads and 30% for 39 ton axle loads. For this line, the reduction in train meets more than com- 
pensated for the increased time required for maintenance; average train travel times declined by 
approximately 2% and the standard deviation of travel times decreased by approximately 12%, 
assuming typical rail defect rates. The average train travel times were similar for 36 ton and 39 ton 
axle loads, with slightly better reliability for 36 ton axle loads. Romps did not calculate how many 
additional HAL trains could be operated without degrading performance. 

For the 30 MGT route there were 62, 56 and 52 trains per week in the three ca.ses, and mainte- 
nance hours increa.sed by 10 and 15% for the 36 ton and 39 ton axle load cases, respectively. Average 
train travel times decreased by less than 0.5%, while the standard deviation of travel time decreased 
by 30% for 36 ton axle loads and 15% for 39 ton axle loads. 

These earlier results demonstrated that operating with heavier axle loads can produce modest 
improvements in train performance, despite substantial additional track maintenance. As demon- 
strated in Dontula's work,, improvements in track performance can be translated into increases in 
capacity. However, capacity analysis is highly line-specific. These results are illu.strative only, and 
the results for a particular line would depend upon train volumes, track condition, maintenance prac- 
tices, and route condition. 

1.2 Summary of Results 

The AAR HAL Phase II Economic Analysis was based on two prototypical coal routes: an 80 
MGT high-tonnage route and a 30 MGT medium-tonnage route. Except for route length and distrib- 
ution of sidings, the routes examined in this study were identical to the HAL Phase II routes. For this 
study, the length of the routes and distribution of sidings were kept consistent with the assumptions 
of the previous analysis performed by Romps: both the high-tonnage and medium-tonnage routes are 
292 miles long with 2-mile sidings separated by 12 miles of single track. 

Based on the route characteristics described in Section 2, maintenance requirements were deter- 
mined for each of six cases: 

• 80 MGT, 33 ton axle loads 

• 80 MGT, 36 ton axle loads 

• 80 MGT, 39 ton axle loads 

• 30 MGT, 33 ton axle loads 

• 30 MGT, 36 ton axle loads 

• 30 MGT, 39 ton axle loads. 

Track maintenance requirements were based on the HAL Pha,se II Economic Analysis, in turn 
developed based on state-of-the-art maintenance models such as TRACS, along with other models 
used as appropriate. Section 3 summarizes the annual maintenance requirements determined for each 
of the six cases. 

The line-haul simulation model dispatches trains over a specified route based on the train sched- 
ule, defect rates and maintenance scenario. Based on the arrival times for each train simulated, the 
model calculates the average trip time per train, average train delay per train, and standard deviation 
of trip time. Separate runs were performed for each of the six cases for different levels of unscheduled 
maintenance (resulting from train or infrastructure failure) and scheduled maintenance. For unsched- 
uled maintenance, runs were performed to explore the effects of zero failures, train failures only (no 
infrastructure failures), average rail defect rates, summer (low) defect rates, and winter (high) defect 
rates. For scheduled maintenance, runs were performed to explore the effects of 4-hour, 6-hour and 8- 
hour maintenance windows. Section 4 discusses the application of the simulation model. 



300 



Bulletin 761 — American Railway Engineering Association 



Based on the maintenance requirements and simulation model results for summer and winter 
months, average annual values were calculated for delay per train, trip time per train and standard 
deviation of trip time. Assumptions were developed concerning the maintenance windows for each 
type of activity, the percentage of work occurring in summer months, and the percentage of work 
occurring in the "shadow" of other maintenance. The results are described in Section 5, and summa- 
rized below in Table 1 . The results for train delay and trip time standard deviation are shown graph- 
ically in Figure 1 and Figure 2. 

For the 80 MGT route, the annual average trip time, average delay per train, and standard devia- 
tion of trip time are lower for heavy axle loads than for the base case of 33 ton axle loads. The average 
trip time is 9.8 hours for the base case, 9.5 hours for 36 ton axle loads and 9.6 hours for 39 ton axle 
loads; this represents a reduction of approximately 2 to 3% for heavy axle loads. The average train delay 
is 2.5 hours per train for the base case, and 2.3 hours for 36 and 39 ton axle loads; this represents a 
reduction of 8% for heavy axle loads. The standard deviation of trip time is 2.2 hours for the base case, 

2.0 for 36 ton axle loads and 2.1 for 39 ton axle loads, a reduction of approximately 5 to 9%. 

For the 30 MGT route the annual average trip time, average delay per train and standard devi- 
ation of trip time are the same or higher for heavy axle loads than for the base case. The annual aver- 
age trip time is 8.4 hours for the base case, 8.4 hours for 36 ton axle loads, and 8.5 hours for 39 ton 
axle loads; heavy axle loads cause an increase of approximately to 1% in trip time. The average 
train delay is 1 . 1 hours per train for 33 and 36 ton axle loads and 1 .2 hours for 39 ton axle loads, indi- 
cating a to 9% increase in train delay for heavy axle loads. The standard deviation of trip time is 

1.1 hours for the base case of 33 ton axle loads, and 1.3 hours for heavier axle loads, an increase of 
approximately 18%. 

The results of the study for high-tonnage route examined indicate that heavy axle loads are 
expected to result in improvements in trip time and reliability, and reductions in train delay relative 
to the base case. For this route, the reduction in number of trains for heavy axle load operations more 
than offsets the effects of increased track maintenance. However, for the medium-tonnage route 
heavy axle loads are expected to result in slight increases in trip time and train delay, and a reduction 
in reliability. Consistent with the results of previous work described in Section 1.1, this study shows 
that heavy axle load operations can lead to improvements in train performance and track capacity, but 
that the results are highly route-specific. 



Table 1. Annual Average Train Performance Results by Route and Axle Load 



Description 


Annual Average Results 


Percetitage 




(hours per train) 


Change Relative 








to Base Case 


33 ton 


36 ton 


39 ton 


38 ton 


39 ton 


High-Tonnage Route (80 MGT) 












Total Delay 


2.5 


2.3 


2.3 


-8% 


-8% 


Trip Time 


9.8 


9.5 


9.6 


-3% 


-2% 


Trip Time Standard Deviation 


2.2 


2.0 


2.1 


-9% 


-5% 


Medium-Tonnage Route (30 MGT) 












Total Delay 


1.1 


1.1 


1.2 


0% 


9% 


Trip Time 


8.4 


8.4 


8.5 


0% 


1% 


Trip Time Standard Deviation 


1.1 


1.3 


1.3 


18% 


18% 



Paper by Carl D. Martland and William E. Robert 



301 




33 ton 



36 ton 
Axle Load 



1 80 MGT H 30 MGT 



39 ton 



Figure 1. Annual Average Train Delay Results by Route and Axle Load 



3.0 



-= 2.5 



^ 2.0 



^ 1.5 



0) 1.0 



:= 0.5 



0.0 




33 ton 



36 ton 
Axle Load 



1 80 MGT m 30 MGT 



39 ton 



Figure 2. Annual Average Standard Deviation of Trip Time Results by Route and Axle Load 

2. DESCRIPTION OF THE CASE STUDY 

This section describes the case study used to investigate the effects of track maintenance on 
train reliability. Section 2.1 summarizes the route characteristics for the case study. Section 2.2 sum- 
marizes the traffic characteristics, and Section 2.3 discusses important maintenance assumptions 
made during the analysis. 



302 



Bulletin 761— American Railway Engineering Association 



2.1 Route Characteristics 

The route chosen for analysis was a 292-mile long route. The route consisted of single track 
with 20 2-mile long sidings distributed along the line. The maximum speed on the mainline was 40 
mph, and the normal operating speed of all trains also was 40 mph. The maximum speed on sidings 
was 30 mph, and the slow order speed was 10 mph. 

Other route characteristics were the same as those used for the AAR HAL Phase II Economic 
Analysis'-*'. These route characteristics include the following: 

• percentage of track of different degrees of curvature, 

• spacing of turnouts 

• the quality of track components (including factors such as the hardness of the rail), and 

• the initial condition of components. 

Different values for curvature, turnout spacing, quality of track components, and initial condi- 
tion of track components were assumed for a high-tonnage coal route with annual traffic of 80 MGT 
and a medium-tonnage coal route with annual traffic of 30 MGT. The reader is referred to the AAR 
HAL Phase II Economic Analysis for details regarding route characteristics'^'. 

2.2 Traffic Characteristics 

Traffic characteristics were determined for six cases: 3 different axle loads for the 30 MGT and 
80 MGT routes. As for the route characteristics, traffic characteristics were those used for the AAR 
HAL Phase II Economic Analysis'-^'. The base case for each route was for an axle load of 33 tons; 
axle loads of 36 tons and 39 tons were examined in addition. 

The lines examined were assumed to be coal lines with loaded trains traveling in one direction 
and empty trains returning in the opposite direction. The number of cars per train was assumed to be 
constant regardless of axle load, so the number of trains required decreases as the axle load increases. 
The number of trains per week for each case was used as input for the simulation model. The simu- 
lation further allowed for dispatching a different number of trains per day, so 67% of the weekly traf- 
fic was scheduled for 5 off-peak days, and 33% was scheduled for 2 peak days of the week. This dis- 
tribution was based on typical traffic loads and is consistent with the previous analysis'''. Table 2 lists 
the number of trains scheduled for each day of the week for each of the six cases. 

2.3 Maintenance Assumptions 

The AAR HAL Phase II Economic Analysis was used as appropriate to determine maintenance 
requirements for each of the six cases. However, certain additional assumptions were necessary for 



Table 2. Scheduled Trains per Day 



^"""■^' 


Trains Scheduled per Day by Route and Axle toad ' 




80 MGT 






30 MGT 




33 ton 


30 ton 


39 ton 


33 ton 


36 ton 


39 ton 


Sunday 

Monday 

Tuesday 

Wednesday 

Thursday 

Friday 

Saturday 


22 
22 
22 
22 
26 
28 
20 


20 
20 
20 
20 
24 
24 
18 


18 
18 
18 
18 
22 
22 
18 


8 
8 
8 
8 
10 
12 
8 


6 
8 
8 
8 
10 
10 
6 


6 
6 
6 
8 
10 
10 
6 


Total Trains per Week 


162 


146 


134 


62 


56 


52 



Paper by Carl D. Martland and William E. Robert 



303 



translating the maintenance requirements into annual track closure hours. For instance, for each 
maintenance activity labor productivity rates were required. For many activities these rates could be 
determined based on the HAL Analysis or based on inspection of models used for the HAL Analysis. 
However, for other activities a productivity rate could not be derived from the analysis and reason- 
able assumptions were required based on additional research performed at MIT to determine appro- 
priate TRACS defaults'^l. Table 3 summarizes the labor productivity rates used for this analysis, with 
references to the sources of the rates. 

Listed below are other major assumptions made regarding maintenance activities: 

• Maintenance windows may be either 2, 4, 6, or 8 hours long. 

• Most scheduled maintenance, such as tie replacement and rail relaying, occurs in summer months. 

• Maintenance windows of 2 hours or less do not cause significant train delays — these win- 
dows may be inserted between trains without adding to delay. 

• Multiple work crews do not add to train delay. This includes crews performing two different 
activities in the same area, as well as crews operating at the same time at different points 
along the line. 

• 65% of rail defects occur in winter months (assumed to be the six coldest months of the year). 
This value was based on the results of a recent study of rail defect rates'"''. 

• 80% of rail defects are found using an inspection car; the remaining 20% result in .service fail- 
ures. This value was based on a range of values reported in a recent study of rail defect rates''"'. 

• 40% of service failures result in track closures; 60% result in slow orders. 

• Spot tie replacement is performed in 2-hour maintenance windows and does not cause either 
slow orders or track closures. 

• Unless otherwise specified, other assumptions made in determining maintenance require- 
ments are equivalent to those used in the previous analysis. 



Table 3. Labor Productivity Rates for Track Maintenance Activities 



Maintenance Activity 


Rate 


Source 


Rail Relay (mi/hr) 


0085 


TRACS Defaults 


Rail Defect Repair by Inspection (defects/hr) 


1 


TRACS Defaults 


Rail Defect Repair (defects/hr) 


0.25 


TRACS Defaults 


Production Tie (ties/hr) 


150 


TRACS Defaults 


Spot Tie (ties/hr) 


2 


TRACS Defaults 


Ballast Tamping (ft/hr) 


1,500 


AAR HAL Analysis 


Ballast Undercutting (ft/hr) 


1,000 


AAR HAL Analysis 


Ballast Plowing (ft/hr) 


1,200 


AAR HAL Analysis 


Time to Replace Turnout Switch Point (hrs) 


4 


AAR HAL Analysis 


Time to Replace Turnout Frog (hrs) 


8 


AAR HAL Analysis 


Time for Turnout Undercutting (hrs) 


8 


AAR HAL Analysis 


Time for Turnout Production Grinding (hrs) 


2 


AAR HAL Analysis 


Time for Turnout Surfacing (hrs) 


4 


AAR HAL Analysis 


Time to Replace Turnout Switch Ties (hrs) 


2 


AAR HAL Analysis 


Time for Turnout Installation (hrs) 


11 


AAR HAL Analysis 



304 



Bulletin 761 — American Railway Engineering Association 



3. MAINTENANCE REQUIREMENTS 

This section summarizes the maintenance requirements determined for the case study. 
Maintenance requirements were determined based on the AAR HAL Phase II Economic Study'-'' in 
a manner consistent with the previous analysis'*'. For the purpose of this analysis, maintenance is 
described either as unscheduled maintenance or scheduled maintenance. Unscheduled maintenance 
refers to that maintenance that cannot be anticipated ahead of time and results either in an unsched- 
uled track closure or slow order to be put in effect until the maintenance is complete. Unscheduled 
maintenance is discussed in Section 3.1. Scheduled maintenance refers to maintenance that can be 
anticipated far enough in advance to schedule track closure hours before the maintenance occurs. 
Scheduled maintenance is discussed in Section 3.2. The distinction between unscheduled and sched- 
uled maintenance is made because in applying the simulation model the two different types of main- 
tenance have different effects on train delay, as discussed in Section 4. 

Annual hours of track maintenance, expressed in terms of track hours (not man hours), are sum- 
marized in Table 4 for each of the six cases studied. The analysis assumed that activities that require 
2-hour maintenance windows, including spot tie replacement and some turnout maintenance activi- 
ties, are conducted between trains with at most only minor effects on train delay. Thus, although a 
significant number of hours are required for spot tie replacement, turnout tie replacement and turnout 
grinding, these activities do not significantly affect train delay or reliability. Figure 3 illustrates the 
distribution of track maintenance hours for those activities expected to have a significant effect on 
train performance. Separate graphs are shown for 33 ton axle loads for the 80 MGT and 30 MGT 
routes. Maintenance requirements for specific activities are discussed in the following sections. 

3.1 Unscheduled Maintenance 

Unscheduled maintenance activities included in the analysis consist of train failures, rail 
defects, and spot tie replacement. Train failures, discussed below in Section 3.1.1, are different from 
other maintenance activities included in the analysis in that they are not a form of track maintenance. 
Rail defects, discussed in Section 3.1.2, may result in unscheduled maintenance, but most may be 
detected by insf)ection and are instead repaired through scheduled maintenance. Spot tie replacement 
is discussed in Section 3.1.3. 



Table 4. Annual Track Maintenance Hours by Route and Axle Load 



Maintenance 


80 MGT 


30 MGT ] 


xxfMmMifi'isx 


36 ton 


39 ton 


_iatin„> 


;::l:M.Jt0J!?,;-,. 


39 ton 


Rail Relaying 


262 


262 


284 


98 


109 


115 


Rail Defect Repair 


84 


140 


136 


53 


66 


68 


(through inspection) 














Rail Defect Repair 


84 


140 


136 


53 


66 


68 


(service failures) 














Ties - Production 


181 


188 


190 


144 


152 


153 


Ties - Spot 


1,255 


1,279 


1,273 


1,134 


1,153 


1,176 


Ballast Renewal 


98 


97 


103 


37 


37 


39 


Ballast Resurfacing 


274 


282 


290 


105 


108 


111 


Turnouts - 2-hr Window/s 


292 


223 


263 


219 


168 


198 


(ties, grinding) 














Turnouts - 4-hr Windows 


210 


179 


229 


158 


134 


172 


(points, surfacing) 














Turnouts - 8-hr Windows 


298 


285 


337 


223 


214 


254 


(install, frogs, undercutting) 














Total Hours 


3,038 


3,075 


3,241 


2,224 


2,207 


2,354 



Paper by Carl D. Martland and William E. Robert 



305 



80MGT 



30MGT 





■ Rail Relaying 


■ Rail Defects - Inspection 


D Production Ties 


M Ballast Resurfacing 


D Ballast Renewal 


■ Turnout - 4 hr win. 


E Turnout ■ 8 hr win. 







Figure 3. Annual Scheduled Track Maintenance Hours for 33 Ton Axle Loads 



3.1.1 Train Failures 

The simulation model allows the user to specify the mean miles between train failures and the 
mean repair time for a train failure. A train failure may result from either a locomotive or car failure. 
For the purpose of this analysis, the mean miles between train failures was assumed to be 25,000 
miles, and the mean repair time was assumed to be 65 minutes for all cases. Based on these values 
the simulation model randomly generates train failures, with the actual miles between failures and 
repair time for a particular failure exponentially distributed. These values are the same as those used 
for the previous analysis'''. 

3.1.2 Rail Defects 

Rail defect rates are those used for the AAR HAL Phase 11 Economic Analysis, which were 
determined using TRACS. Rail defects are caused as a function of annual traffic, axle loads, rail met- 
allurgy, degree of curvature, and rail relay, lubrication, and grinding policies. The rail defect rates 
used for the analysis, expressed in terms of rail defects per mile per year, are summarized for each of 
the six cases below in Table 5. As discussed in Section 2.3, 65% of rail defects are assumed to occur 
during the six coldest months of the year. Thus, different defect rates are listed for average, summer, 
and winter conditions. 

Consistent with the AAR HAL Phase II Economic Analysis, Table 5 indicates significantly 
higher defect rates for heavy axle loads than for 33 ton axle loads. For the 80 MGT route the average 
defect rate increases from 0.359 defects per mile per year for 33 ton axle loads to 0.583 to 0.599 for 
heavy axle loads. As noted in the HAL Phase II Economic Analysis, for the 39 ton axle load case, extra 
grinding is required to prevent a further increase in rail defect rates, resulting in lower defect rates for 
39 ton axle loads than for 36 ton axle loads. The results for the 30 MGT route show generally lower 
defect rates, and a more subtle heavy axle load effect. For this route, the defect rate increases from 
0.227 defects per mile per year for 33 ton axle loads to 0.283 to 0.292 for heavy axle loads. 

Rail defects found and repaired in concert with periodic inspections using detector cars were 
treated as scheduled maintenance. Rail defects that cause service failures (not detected and repaired 
following an inspection) were treated as unscheduled maintenance. As discussed in Section 2.3, 80% 
of rail defects were assumed to be detected by inspection before a service failure occurs. The annual 
hours of scheduled rail defect repair were based on a productivity rate of one defect repaired per hour. 



306 



Bulletin 761 — American Railway Engineering Association 



Table 5. Rail Defect Rates by Route and Axle Load 



Description 


Rail Defect Rate by Route and Axie toad 


80MGT 


30 MGT 1 


33 ton 


36 ton 


39 ton 


33 ton 


36 ton 


39 ton 


Average Rates 
(defects/mile/year) 


0.359 


0.599 


0.583 


0.227 


0.283 


0.292 


Summer Rates 
(defects/mile/year) 


0.251 


0.419 


0.408 


0.159 


0.198 


0.205 


Winter Rates 
(defects/mile/year) 


0.467 


0.778 


0.758 


0.295 


0.368 


0.380 



Of the remaining 20% of rail defects that cause service failures, 40% were assumed to result in 
track closure and 60% in slow orders. In general, defects on curves or very severe defects cause track 
closure, but for most defects on tangent track, trains can still safely pass over the damaged section of 
rail, albeit at a lower speed. The time of a track closure for defect repair was assumed to be 5 hours 
plus or minus 2 hours (this includes time for transportation to the repair site), and the time for a slow 
order was assumed to be 24 hours plus or minus 4 hours. A uniform probability distribution function 
was used to determine the actual track closure or slow order time for repair of a particular defect. 

The values for rail defects per mile leading to service failures used in this study are significantly 
lower than the values in the previous analysis for two reasons. First, compared to the AAR HAL 
Phase I Economic Analysis, the Phase II analysis assumed increased use of harder rail (compared to 
Phase 1, the Phase II analysis assumed more rail with Brinell hardness 270 has been replaced with 
harder rail). Second, this analysis assumed that 80% of rail defects would be repaired with minimal 
train delay as part of the routine defect inspection process prior to causing a service failure, whereas 
the previous analysis assumed that all rail defects would result in delays similar to those experienced 
in service failures. 

3.1.3 Spot Tie Replacement 

The HAL Phase II Economic Analysis used the TRACS tie model to determine annual mainte- 
nance requirements for spot and production tie replacement. Spot tie replacement is required when a 
cluster of three or more ties in a row fails. Spot tie replacement generally is performed by small teams 
working with a productivity rate of approximately 2 ties per hour. 

Results from TRACS runs for the HAL Phase II Economic Analysis were examined to deter- 
mine the annual number of spot and production ties. Where the runs were not "steady state," but 
showed a significant increase in the rate of replacement over a 30 year period, the results for the last 
20 years of the 30 year run were used to determine average replacement rates. 

Annual hours of spot tie maintenance required for each case are summarized above in Table 4. 
Based on these results, more hours are required for spot tie replacement than for any other mainte- 
nance activity. However, the results indicate little dependence on axle load or annual traffic. For the 
80 MGT route, the results indicate that 1,255 to 1,279 hours of maintenance are required, depending 
on axle loads. For the 30 MGT route 1,134 to 1,176 hours are required. Regardless of the large num- 
ber of hours required for spot tie replacement, this maintenance activity has no effect on train delay 
in this analysis. In the previous analysis spot tie replacement was assumed to require track closure or 
a slow order, but this analysis assumed that spot tie maintenance can be performed between trains 
without causing track closure or slow orders. 

3.2 Scheduled Maintenance 

Scheduled maintenance activities included in the analysis consist of rail relay, rail defect repair, 
production tie replacement, ballast resurfacing and renewal, and turnout maintenance. These activi- 
ties are described in the following sections. 



Paper by Carl D. Martland and William E. Robert 



307 



3.2.1 Rail Relaying and Defect Repair 

The amount of rail relayed each year for each of the six cases is the same as that determined in 
the AAR HAL Phase II Economic Analysis (scaled for the different lengths of routes in this analy- 
sis), which uses TRACS for rail maintenance modeling. Assuming a productivity rate of 0.085 miles 
relayed per hour. Table 4 summarizes the annual maintenance required for each case. 

The results indicate very little heavy axle load effect. For the 80 MGT route, 262 hours of main- 
tenance are required annually for 33 ton axle loads, 262 hours for 36 ton axle loads, and 284 hours 
for 39 ton axle loads. For the 30 MGT route 98 hours are required for 33 ton axle loads, 109 hours 
are required for the 36 ton case, and 1 1 5 hours of annual maintenance are required for the 39 ton axle 
load case. 

Rail defects require scheduled maintenance, al,so. As discus.sed in Section 3.1.2 the analysis 
assumed that 80% of defects would be found by inspection. For this 80%, a productivity rate of one 
defect repaired per hour was applied, leading to the annual hours for defect repair for each ca.se as 
shown in Table 4. Both rail relay and defect repair were assumed to occur in 8-hour work windows. 

3.2.2 Production Tie Replacement 

Production tie replacement differs from spot tie replacement in that large numbers of ties are 
marked during inspection and then replaced. The annual number of production ties was determined 
from TRACS output as described in Section 3.1.3 for spot ties. Production tie replacement is exe- 
cuted by large production gangs, possibly using specialized machines such as the P811 superma- 
chine. As such, the productivity rate of 1 50 ties replaced per hour is significantly greater than the spot 
tie replacement rate of 2 ties per hour. 

Projected annual production tie maintenance requirements are shown in Table 4. The annual 
hours of maintenance ranges from 181 to 1 90 hours for the 80 MGT route and from 144 to 1 53 hours 
for the 30 MGT route. The results indicate a slight heavy axle load effect. Production tie replacement 
was assumed to occur in 8-hour work windows. 

3.2.3 Ballast Resurfacing and Renewal 

The AAR HAL Phase II Economic Analysis derived the surfacing cycle (in years) and renewal 
life (in MGT) for "good" and "bad" ballast subgrade conditions for prototypical 80 MGT and 30 
MGT routes. The HAL analysis further assumed that for each route subgrade conditions were "good" 
for 90% of the route and "bad" for 10% of the route. The ballast surfacing cycles and renewal lives 
used in the AAR HAL Phase 11 Economic Analysis are listed below in Table 6. 

To determine annual hours of track maintenance hours for ballast resurfacmg and renewal, one 
must supplement the values in Table 6 with productivity rates. Productivity rates were estimated 
based on default values u.sed in the HAL Economic Analysis for the AAR Ballast Model'"'. In this 
model, a rate of 1,500 feet per hour was assumed for ballast tamping, 1,000 feet per hour was 
assumed for undercutting, and 1,200 feet per hour was assumed for ballast plowing. Ballast resur- 
facing involves plowing and tamping; a rate of 1 ,200 feet per hour, the rate of the slower activity, was 
estimated for resurfacing. Ballast renewal involves undercutting, plowing, and tamping, so a rate of 
1,000 feet per hour, the rate for the slowest activity, was estimated for renewal. 

Table 6. Ballast Surfacing Cycle and Renewal Life by Route and Axle Load 



Description 


80 MGT 


30 MGT 1 


33 ton 


3$ ton 


39 ton 


j^Mmm 


36 ton 


39 ton 


Good Conditions 

Surfacing Cycle (years) 
Renewal Life (MGT) 


5.0 
1410 


50 
1410 


5.0 
1410 


13.0 
1410 


13.0 
1410 


13.0 
1410 


Bad Conditions 

Surfacing Cycle (years) 
Renewal Life (MGT) 


30 
630 


2.5 
532 


22 
463 


8.0 
630 


6.7 
530 


5.9 
461 



308 



Bulletin 761 — American Railway Engineering Association 



The resulting annual maintenance requirements for each case are shown in Table 4. There is a 
heavy axle load effect for "bad" ballast/subgrade conditions, but the conditions are "bad" for only 
10% of each route, so there is no more than a slight heavy axle load effect overall. For the 80 MGT 
route, annual hours for resurfacing range from 274 to 290, and annual hours for renewal range from 
98 to 103. For the 30 MGT route, annual hours for resurfacing range from 105 to 111 and annual 
hours for renewal range from 37 to 39. The analysis assumed that ballast maintenance occurs in 6- 
hour work windows. 

3.2.4 Turnout Maintenance 

The AAR/MIT Turnout Model' '^1 used for the AAR HAL Phase II Economic Analysis was used 
to determine annual track maintenance hours required for turnout replacement and repairs. Based on 
axle loads, annual traffic, and details concerning the turnout components, the model projects turnout 
life (in MGT) and the number of times over the turnout life that different maintenance activities will 
be required. The model further allows for specification of the number of track closure hours required 
for each type of turnout maintenance activity and provides a first approximation of train delay costs. 

Maintenance activities such as lubrication, bolt tightening, and point adjustment were included 
in the turnout cost model, but were not assumed to cause track closure. Production grinding and 
switch tie replacement were assumed to require 2 hours of track closure, point replacement and sur- 
facing were assumed to require 4 hours, and major activities such as frog replacement, undercutting, 
and turnout installation were assumed to require 8 hours or more. Turnout installation was expected 
to require 1 1 hours, but since the maximum track closure modeled for this study was 8 hours, track 
closure hours for turnouts were artificially divided into 8 hour shifts. The number of times each main- 
tenance activity is projected to occur over the life of a turnout is listed for each case in Table 7. Total 
hours of turnout maintenance for each case are listed in Table 4. 

The results indicate that for the less time consuming activities such as surfacing and tie replace- 
ment, there is little heavy axle load effect, or even a negative effect (with heavy axle loads requiring 
less maintenance). However, for major activities such as frog replacement and turnout installation, 
there is a clear heavy axle load effect between the 33 ton and 39 ton axle load cases. For the 80 MGT 
route 298 hours are required annually for major turnout maintenance activities for the base case, 285 
(13 fewer hours) for 36 ton axle loads, and 337 for 39 ton axle loads. For the 30 MGT route 223 hours 
are required annually for major turnout maintenance activities for the base case, compared to 214 (9 
fewer hours) for 36 ton axle loads, and 254 for 39 ton axle loads. 

4. SIMULATION MODEL ANALYSIS 

This section describes how the simulation model was used to predict average trip time, standard 
deviation of trip time (a measure of reliability) and train delay for each of the six cases studied. 
Section 4.1 describes the inputs required for using the model, Section 4.2 discusses effects of 
unscheduled maintenance (due to train failures and service failures from rail defects), and Section 4.3 
discusses the effects of scheduled track maintenance. 



Table 7. Maintenance Events 


per Turnout over Turnout Life by Route and Axle Load 


-Maintenance 
Activity 


80 MGT 


30 MGT 1 


33 ton 


36 ton 


39 ton 


33 ton 


36 ton 


39 ton 


Replace Switch Point 


2 


2 


3 


2 


2 


3 


Replace Frog 


3 


3 


4 


3 


3 


4 


Undercutting 


2 


2 


2 


2 


2 


2 


Production Grinding 


11 


9 


10 


11 


9 


10 


Surfacing 


7 


6 


7 


7 


6 


7 


Replace Sw/itch Ties 


14 


11 


13 


14 


11 


13 


Installation 


1 


1 


1 


1 


1 


1 



Paper by Carl D. Martland and William E. Robert 309 



4.1 Model Inputs 

The simulation model used is a line-haul simulation model identical to the model used for the 
previous analysis''', except for minor changes described below. The model requires a series of input 
files, including a route file, inbound and outbound schedule files, an unscheduled maintenance file, 
and a scheduled maintenance file. Using data from the input files, the model simulates three weeks 
of train operations and provides results for trip time, standard deviation of trip time and train delay 
for any time period within the three weeks. 

The route file specifies the length of the route and the length and location of all sidings. As dis- 
cussed in Section 2.1 the route studied is 292 miles long with 20 2-mile sidings spaced every 12 
miles. The route file also lists the maximum speed on the mainline and sidings, which may vary for 
each segment of track. 

The schedule files determine the number of inbound and outbound trains dispatched each day 
of the simulation. For each day the user specifies the identification number of the first train, the time 
at which the first train is dispatched, and the interval between trains for the rest of the day. To fine- 
tune the schedule, such as for holding trains at the terminal during scheduled maintenance, individ- 
ual trains may be delayed or canceled. The number of trains scheduled per day for each of the six 
ca.ses is discussed in Section 2.2. 

The unscheduled or random maintenance file specifies parameters for train failures and service 
failures due to rail defects. For train failures the user specifies the mean miles between train failures, 
and the mean time required for repair. For service failures from rail defects, the user specifies the 
number of defects per mile per year, the probability that a defect results in track closure (versus a 
slow order), the mean time for track closure, the mean time for a slow order, and the maximum range 
(plus or minus) on the track closure and slow order times. Unscheduled maintenance events are ran- 
domly generated based on the data in the input file, and upon their generation immediately cause 
either track closure or a slow order. Train schedules may be modified to best accommodate sched- 
uled maintenance, but unscheduled maintenance cannot be anticipated and though infrequent, may 
have a significant affect on average train delay and reliability. The values used in the simulation 
model for unscheduled maintenance are discussed in Section 3.1. 

The scheduled maintenance file allows the user to specify maintenance windows. For each 
maintenance window the user specifies the location of the maintenance, the locations where inbound 
and outbound trains must be held, the duration of the track closure and/or slow order, and the loca- 
tion of the slow order within the segment. 

The previous version of the model required the user to type in the name of each input file each 
time the model is run. The model was modified to facilitate use of batch files to speed the process of 
running the model. Also, a Lotus 1-2-3 spreadsheet was developed to allow the user to organize dif- 
ferent runs, to execute all desired runs in a keystroke through use of spreadsheet macros, and to 
import the text files with the results of each run. These changes do not affect the model results, but 
expedite the process of running the model. 

4.2 Analysis of Effects of Unscheduled Maintenance 

The simulation was used to explore the effects of a series of different defect rates, independent 
of any scheduled maintenance. For each case runs were performed for no defects (neither train fail- 
ures nor rail defects), train failures only, train failures with average rail defect rates, train failures with 
summer defect rates, and train failures with winter rail defect rates. Detailed results are presented in 
Appendix A for average train delay, trip time and trip time standard deviation. These results are sum- 
marized below. 

Even without unscheduled and scheduled maintenance, there is train delay and variation in trip 
time due to delays from meets and passes of inbound and outbound trains. Without any delay the trip 
is 7.3 hours (for a train traveling 292 miles at a speed of 40 miles per hour); delay from meets and 



310 



Bulletin 761 — American Railway Engineering Association 



passes adds an average of 0.9 hours per trip for the 80 MGT route with 33 ton axle loads, and an aver- 
age of 0.5 hours per trip for the 30 MGT route with 33 ton axle loads. When fewer trains are dis- 
patched there is less delay from meets and passes; accordingly there is slightly less delay from heavy 
axle loads in the "no defect" case (up to O.I hours less). 

For the 80 MGT route, train failures add an additional 0.3 to 0.4 hours of train delay per train, 
and cause a significant increase in standard deviation of trip time (from 0.4 to 1 .2 hours per train for 
33 ton axle loads and 0.5 to 1 .0 hours per train for 39 ton axle loads). However, accounting for rail 
defects adds little to average train delay, as in service defect rates (even in winter) were quite low. 
For the 30 MGT route, including train failures adds up to 0.2 hours per train to the average train delay 
and up to 0.2 hours to the standard deviation of trip time. Adding rail defects increases average train 
delay and standard deviation of trip time by no more than 0. 1 hours per train. Average total delay for 
33 ton axle loads is shown for each case in Figure 4. The standard deviation of trip time is shown in 
Figure 5. 

In summary, the analysis indicates that train meets cause significant train delay, regardless of 
any unscheduled or scheduled maintenance. For the 80 MGT route, train failures add an average of 
0.3 to 0.4 hours of delay per train and cause a significant increase in the standard deviation of trip 
time. For the 30 MGT route, train failures do not appear to cause a significant increase in average 
train delay. Service failures resulting from rail defects do not appear to cause a significant increase 
in average train delay either for the 80 MGT or 30 MGT route. This result is different from that of 
the previous analysis; the previous analysis assumed more service failures due to rail defects and 
allowed for service failures for spot tie replacement. 

4.3 Analysis of Effects of Scheduled Maintenance 

The effects of scheduled maintenance were examined through a series of runs with 4, 6 and 8-hour 
work windows. For each run, a segment of track was closed once a day for five days in the second week 
of the three-week simulation. The track was closed for either 4, 6, or 8 hours each day. In addition a 24- 
hour slow order was placed on a 2-mile length of track within the segment. Also, the schedule was 
altered slightly to delay trains at the terminal during maintenance periods (such scheduled delays are 
not factored into the train delay calculations). Results were obtained for the second week of the simu- 
lation only, including the two days of the second week in which no maintenance is conducted. 

1.50 



•jo 1.25 




No Defects Train Only 



Summer 
Defect Rate 



Average 



Winter 



1 80 IVIGT m 30 IVIGT 



Figure 4. Average Total Delay per Train for 33 Ton Axle Loads 



Paper by Carl D. Martland and William E. Robert 



311 



1.50 
1.25 
1.00 
0.75 



V) 

I 0.50 
•=■ 0.25 



0.00 



1.2 1.2 1.2 1.2 




1 


■ o.7 Ho.7 Ho.7 


0.7 


0.4 

\ 


iiii 

iil:: 

Ilii 






A 


iijii 


A 




A 


iiiji 





No Defects Train Only 



Summer 
Defect Rate 



Average 



Winter 



1 80 MGT 1! 30 MGT 



Figure 5. Standard Deviation of Trip Time for 33 Ton Axle Loads 



The results for the three different lengths of maintenance windows for each of the six cases are 
provided in Appendix A. The results demonstrate that scheduled maintenance has a significant affect 
on train delay and on trip time standard deviation. For the 80 MGT route with 33 ton axle loads and 
summer defect rates (most scheduled maintenance occurs in the summer) 4-hour maintenance win- 
dows increase the average delay per train from 1 .3 hours to 2.7 hours. For 6-hour windows the aver- 
age delay is 3.2 hours and for 8-hour windows the average delay is 4.7 hours. The trend is similar for 
the 30 MGT route with 33 ton axle loads and summer defect rates; average delays per train for no 
maintenance, 4-hour windows, 6-hours windows and 8-hour windows are 0.6, 1 .4, 2.2 and 3. 1 hours, 
respectively. These results are shown graphically in Figure 6. Corresponding results for standard 
deviation of trip time are shown in Figure 7. 

The average delay per train and standard deviation of trip time tend to decrease as axle loads 
increase, as there are fewer trains dispatched for heavier axle loads. Figure 8 illustrates the reduction 
in average delay as a function of axle load for 8-hour maintenance windows and summer defect rates. 
Figure 9 illustrates the corresponding results for standard deviation of trip time. All results shown are 
for summer defect rates. Runs were performed for winter defect rates, but the results were similar. In 
some cases, delay and standard deviation of trip time were lower for runs using winter defect rates, 
an indication that the random variation between runs is a more significant factor than the difference 
between summer and winter defect rates. 

5. RESULTS 

The track maintenance requirements discussed in Section 3 and results from the simulation 
model discussed in Section 4 were used to determine the effects of track maintenance on train delay 
and reliability. Section 5.1 compares the present study with the previous analysis. Section 3.2 sum- 
marizes the results of this study. Section 5.3 presents the conclusions, and Section 5.4 discusses 
issues of interest in further research. 

5.1 Comparison with Previous Results 

Although broadly consistent, the two studies nonetheless differ in several details. The primary 
differences are as follows: 



312 



Bulletin 761 — American Railway Engineering Association 




Maintenance 



4-hour 
Window 



6-liour 
Window 



IVIaintenance Window 



1 80 MGT m 30 IVIGT 



8-liour 
Window 



Figure 6. Average Total Delay per Train versus Maintenance Window for 33 Ton Axle Loads 

• The two studies come to very different conclusions regarding annual maintenance require- 
ments. Relative to the previous analysis this study concludes that significantly more hours are 
required for spot tie replacement (not a factor in the results) and for turnout maintenance. 
Significantly less time is required for ballast resurfacing and renewal. These differences are 
primarily a result of use of different productivity rates, but also are a function of differences 




No 4-hour 6-hour 8-hour 

IVIaintenance Window Window Window 

Maintenance Window 



80 MGT m 30 MGT 



Figure 7. Standard Deviation of Trip Time versus Maintenance Window for 33 Ton Axle Loads 



Paper by Carl D. Martland and William E. Robert 



313 




33 ton 



36 ton 
Axle Load 



39 ton 



1 80 MGT m 30 MGT 



Figure 8. Average Total Delay per Train for 8-hour Maintenance Windows 




33 ton 



36 ton 
Axle Load 



39 ton 



1 80 MGT H 30 MGT 



Figure 9. Standard Deviation of Trip Time for 8-hour Maintenance Windows 



314 Bulletin 761 — American Railway Engineering Association 



between Phase I and Phase II of the HAL Economic Analysis. With regard to ballast mainte- 
nance, the Phase II analysis assumed longer ballast lives and surfacing cycles, with less of a 
HAL effect than the Phase I analysis; also, since the Phase I analysis, the ballast maintenance 
model was revised and calibrated more accurately. 

• The numbers of rail defects per mile per year resulting in service failures were projected to 
be significantly lower than found previously. This is a result of differences between Phase I 
and Phase II of the HAL Economic Analysis (especially the assumption in Phase II of less rail 
of Brinell hardness 270 — this rail suffers many more defects under base case and HAL traf- 
fic than does harder rail), and is a result of the assumption made in this study that 80% of rail 
defects are repaired before leading to service failures. 

• This study assumed that spot tie replacement generally does not cause either track closure or 
slow orders. The previous analysis assumed that when spot tie replacement is required, there 
is a 90% chance that a slow order is issued, and a 1 0% chance of a track closure. 

• Due to the lower rates of unscheduled maintenance assumed in this study, as discussed above, 
service failures due to rail defects are a less significant factor in train delay and reliability 
than in the previous study. 

5.2 Train Delay and Reliability as a Function of Maintenance and Axle Load 

Section 3 presents the annual maintenance requirements for each of the six cases; the results 
indicate that track maintenance tends to increase with heavier axle loads (the 36 ton axle load case 
for the 30 MGT route is an exception — maintenance hours are slightly lower for this case than for 
the base case). However, the results of Section 4 indicate that maintenance windows for cases with 
heavy axle loads cause less additional delay than maintenance windows for 33 ton axle loads. Thus, 
the increased delay caused by additional maintenance for heavy axle loads is at least partially offset 
by the reduction in delay for a particular maintenance schedule. 

Based on the maintenance requirements and simulation model results, average annual values 
were calculated for delay per train, trip time per train and standard deviation of trip time. The main- 
tenance requirements were used to determine for each case how many weeks of the summer and win- 
ter months involve no maintenance, how many weeks involve 8-hour work windows, how many 
involve 6-hour work windows, and how many involve 4-hour work windows. As noted in Section 2.3 
2-hour work windows were assumed to cause no additional delay. Also, although it would be rea- 
sonable to assume that there would be weeks with combinations of 2, 4, 6 and 8-hour windows, such 
a level of detail was beyond the scope of the study. 

The assumption that nearly all maintenance occurs in summer months leads to the conclusion 
that different maintenance activities must occur simultaneously, or there are not enough hours in the 
summer to complete the required maintenance. Table 8 lists the assumptions made concerning the 
maintenance windows for each type of activity, the percentage of work occurring in summer months, 
and the percentage of work occurring in the "shadow" of other maintenance. Simultaneous mainte- 
nance activities occur either when two activities are performed by the same crew at the same time, 
such as through the use of a P8 1 1 supermachine, or when there are two maintenance crews on dif- 
ferent parts of the line at the same time. Table 9 shows the resulting distribution of weeks in the sum- 
mer and winter months with different levels of maintenance activities. Note that even assuming a fair 
amount of maintenance occurs simultaneously, for the 80 MGT route there is very little time during 
the summer without some level of track maintenance activity. 

The overall results are summarized in Table 1. Figure 1 summarizes the results for annual aver- 
age train delay, and Figure 2 summarizes the results for annual average trip time standard deviation. 

For the 80 MGT route, the annual average trip time, average delay per train, and standard devia- 
tion of trip time are lower for heavy axle loads than for the base case of 33 ton axle loads. The average 



Paper by Carl D. Martland and William E. Robert 



315 



Maintenance Activity 


DaUy Maint 
Window 
(hours) 


Percentage 

ofWorltJn 

Summer 


Percentage 

of Work In 

Shadow of 

Other Maint. 


Rail Relaying 
Rail Defect Repair 
(through inspection) 


8 
8 


100% 
50% 


50% 
0% 


Ties - Production 
Ties - Spot 


8 
2 


100% 
50% 


100% 
100% 


Ballast Renewal 
Ballast Resurfacing 


6 
6 


1 00% 
100% 


50% 
50% 


Turnouts - Ties & Grinding 
Turnouts - Points & Surfacing 
Turnouts - Frogs, Under- 
cutting & Installation 


2 

4 
8 


100% 
100% 
100% 


100% 

50% 

0% 



Table 8. Maintenance Scheduling Scenario 



Daily 

Maintenance Window 


Percentage of Weeks in Year with Window |5 days/wk) j 


80MGT 


30 MGT 1 


33 ton 


36 ton 


33 ton 


mMMmm 


sK:J§;:|y5ibs; 


39 ton 


Summer 














None 


5% 


6% 


0% 


23% 


24% 


20% 


4 hours 


10% 


9% 


11% 


8% 


6% 


8% 


6 hours 


12% 


12% 


13% 


5% 


5% 


5% 


8 hours 


23% 


23% 


26% 


14% 


14% 


17% 


Winter 














None 


48% 


47% 


47% 


49% 


48% 


48% 


4 hours 


0% 


0% 


0% 


0% 


0% 


0% 


6 hours 


0% 


0% 


0% 


0% 


0% 


0% 


8 hours 


2% 


3% 


3% 


1% 


2% 


2% 


Total 


100% 


100% 


100% 


100% 


100% 


100% 



Table 9. Distribution of Maintenance Windows by Route and Axle Load 



trip time is 9.8 hours for the base case, 9.5 hours for 36 ton axle loads and 9.6 hours for 39 ton axle 
loads; this represents a reduction of approximately 2 to 3% for heavy axle loads. The average train delay 
is 2.5 hours per train for the base case, and 2.3 hours for 36 and 39 ton axle loads; this represents a 
reduction of 8% for heavy axle loads. The standard deviation of trip time is 2.2 hours for the base case, 

2.0 for 36 ton axle loads and 2. 1 for 39 ton axle loads, a reduction of approximately 5 to 9%. 

For the 30 MGT route, the annual average trip time, average delay per train and standard devi- 
ation of trip time are the same or higher for heavy axle loads than for the base case. The annual aver- 
age trip time is 8.4 hours for the base case, 8.4 hours for 36 ton axle loads, and 8.5 hours for 39 ton 
axle loads; heavy axle loads cause an increase of approximately to 1% in trip time. The average 
train delay is 1 . 1 hours per tram for 33 and 36 ton axle loads and 1 .2 hours for 39 ton axle loads, indi- 
cating a to 9% increase in train delay for heavy axle loads. The standard deviation of trip time is 

1 . 1 hours for the base case of 33 ton axle loads, and 1 .3 hours for heavier axle loads, an increase of 
approximately 18%. 



3 16 Bulletin 761 — American Railway Engineering Association 



5.3 Conclusions 

This study shows that heavy axle load operations can lead to improvements in train performance 
and track capacity, but the results are highly route-specific. Although heavy axle loads tend to lead to 
increased track maintenance, the delay and loss of reliability from increased maintenance is generally 
offset by the reduction in the number of trains. For the high-tonnage route examined, heavy axle loads 
are expected to result in improvements in trip time and reliability, and reductions in train delay rela- 
tive to the base case. For this route the reduction in number of trains for heavy axle load operations 
more than offsets the effects of increased track maintenance. However, for the medium-tonnage route, 
heavy axle loads are expected to result in slight increases in trip time and train delay, and a reduction 
in reliability. For both the high and medium-tonnage routes, the trip time, train delay and trip time stan- 
dard deviation results are more favorable for 36 ton axle loads than for 39 ton axle loads. 

Overall the results of this study are consistent with the previous analysis described in Section 
1.1, especially with regard to heavy axle load effects. However, unlike Romps' analysis, this study 
indicates a reduction in reliability for heavy axle load operations for the medium-tonnage route. As 
shown in Table 1, the trip time standard deviation was projected to increase by 18% for heavy axle 
load operations, compared to a 15 to 30% decrease in Romps' analysis. 

The results for the high-tonnage route are more significant for evaluating the effect of heavy 
axle loads on track capacity. Even with the projected increase, the trip time standard deviation is sig- 
nificantly lower for the medium-tonnage route than for the high-tonnage route (1.3 hours for the 
medium-tonnage route, compared to 2.0 to 2.1 hours for the high-tonnage route), an indication that 
for the medium-tonnage route studied there is significant additional track capacity, regardless of 
heavy axle load effects. For the high-tonnage route studied, heavy axle load operations increase track 
capacity by reducing trip times and improving reliability. 

5.4 Recommendations for Further Research 

The line-haul simulation model used in this study is a robust model with potential for use in 
researching a number of issues related to the effects of track maintenance on train delay and reliabil- 
ity of a single-track rail line. Following are potential areas for future research: 

• Use of the model to explore capacity issues more fully 

• Comparison of model runs with different maintenance scheduling strategies regarding how 
many crews are scheduled simultaneously, the effects of spacing maintenance along the line, 
and other related issues 

• Accounting for clustering of service failures and including this effect in the calculations of 
effects of maintenance on train delay (this effect is explored in the previous analysis, but is not 
factored into the calculations of effects of maintenance on reliability as a function of axle load) 

Other issues raised by this study cannot be explored with only the simulation model, but may 
nonetheless merit further research, including: 

• Comparison of the defect rates and maintenance requirements projected to requirements of a 
comparable line currently in operation, especially with regard to seasonal effects 

• Investigation and quantification of maintenance scheduling methodology, including study of 
degree to which different types of maintenance activities are coordinated 

• Quantifying the costs and benefits (in terms of improved productivity as well as reduced train 
delay) of improved maintenance practices and high technology maintenance equipment 

• Examination of double-track operations using a new or revised simulation model 

• Exploration of the effects of premium components, such as premium rail for tangent track, 
advanced fastener designs and longer-life concrete ties rather than wood ties 



Paper by Carl D. Martland and William E. Robert 



317 



APPENDIX A. DETAILED SIMULATION MODEL RESULTS 



TOTAL DELAY 


(hrs/train] 




















BOMGT 






30MGT 




Defect Rate 


Maintenance 


33 ton 


36 ton 


39 ton 


33 ton 


36 ton 


39 ton 


No Defects 


None 


0.9 


0.9 


0.8 


0.5 


0.5 


0.4 


Train Only 


None 


1.3 


1.2 


1.1 


0.5 


0.6 


0.6 


Average 


None 


1.3 


1.1 


1.1 


0.6 


0.6 


0.6 


Summer 


None 


1.3 


1.1 


1.1 


0.6 


0.6 


0.6 




4hr 


2.7 


2.4 


2.4 


1.4 


1.4 


15 




6hr 


3.2 


3.2 


2.9 


2.2 


2.2 


2.0 




8hr 


4.7 


4.0 


3.9 


3.1 


3.1 


3.0 


Winter 


None 


1.3 


1.1 


1.1 


0.6 


0.6 


0.6 




4tir 


2.6 


2.4 


2.4 


1.4 


1.3 


1.4 




6hr 


3.4 


3.2 


2.8 


2.2 


2.1 


2.0 




8hr 


4.5 


4.0 


3.9 


3.1 


3.1 


2.9 



TRIP TIME 


(hrs/train) 




















BOMGT 






30MGT 




Defect Rate 


Maintenance 


33 ton 


36 ton 


39 ton 


33 ton 


36 ton 


39 ton 


No Defects 


None 


8.2 


8.2 


8.1 


7.8 


7.8 


7.7 


Train Only 


None 


8.5 


8.4 


8.3 


7.8 


7.9 


79 


Average 


None 


8.6 


8.4 


8.3 


7.9 


7.9 


7.9 


Summer 


None 


8.6 


8.4 


8.3 


7.9 


7.9 


7.9 




4hr 


10.0 


9.6 


9.7 


8.7 


8.7 


8.8 




6hr 


10.5 


10.5 


10.2 


9.4 


9.5 


9.3 




8hr 


12.0 


11.3 


11.2 


10.4 


10.4 


10.2 


Winter 


None 


8.6 


8.4 


8.4 


7.9 


7.9 


7.9 




4hr 


9.9 


9.6 


9.7 


8.7 


8.6 


8.7 




6hr 


10.7 


10.5 


10.1 


9.4 


9.4 


9.3 




8hr 


11.7 


11.3 


11.2 


10.4 


10.3 


10.2 



TRIP TIME STD DEV 


(hrs/train) 




















BOMGT 






30MGT 




Defect Rate 


Maintenance 


33 ton 


36 ton 


39 ton 


33 ton 


36 ton 


39 ton 


No Defects 


None 


0.4 


0.4 


0.5 


0.7 


0.6 


0.6 


Train Only 


None 


1.2 


1.1 


1.0 


0.7 


0.8 


0.8 


Average 


None 


1.2 


0.9 


1.0 


0.7 


0.8 


0.8 


Summer 


None 


1.2 


0.9 


1.0 


0.7 


0.8 


0.8 




4hr 


1.9 


1.9 


2.0 


1.2 


1.5 


1.4 




6hr 


2.8 


2.7 


2.7 


1.8 


2.2 


1.9 




8hr 


4.1 


3.7 


3.6 


2.8 


3.1 


2.7 


Winter 


None 


1.2 


0.9 


1.0 


0.7 


0.8 


0.8 




4hr 


1.8 


1.9 


2.0 


1.2 


1.4 


1.3 




6hr 


2.8 


2.7 


2.7 


1.8 


2.2 


1.9 



318 Bulletin 761 — American Railway Engineering Association 



References 

1. John F. Romps, Modelling Track Maintenance and Its Effects on the Reliability of a Single Track 

Railroad Line, MST Thesis, MIT Department of Civil and Environmental Engineering, 
May 1993. 

2. John F. Romps and Carl D. Martland, "The Effects of Track Maintenance on the Reliability of a 

Single Track Railroad Line," AAR Affiliated Lab Working Paper 94-6, prepared by the MIT 
Center for Transportation Studies, October 1994. 

3. Carl D. Martland, "FAST/HAL Evaluation — Phase II: Modelling the Implications of Phase II 

Results for the Generic Case Studies," AAR Affiliated Lab Working Paper 95-3, prepared by 
the MIT Center for Transportation Studies, December 1995. 

4. Cad D. Martland, "Economic Interpretation of the FAST/HAL Test Results," MIT Rail Group 

Working Paper, prepared by the MIT Center for Transportation Studies, July 1991. 

5. C. D. Martland, P Little, O.K. Kwon and R. Dontula, "Background on Railroad Reliability," AAR 

R&T Report R-803, prepared by the MIT Center for Transportation Studies, March 1992. 

6. Rajesh V. Dontula, A Model for Evaluating Railroad Line-Haul Performance, MST Thesis, MIT 

Department of Civil and Environmental Engineering, May 1991. 

7. Rajesh V. Dontula, and Carl D. Martland, "The Effects of Engineering and Mechanical Reliability 

on the Line Haul Performance of a Single Track Railroad," Proceedings of the Canadian 
Transportation Research Forum, June 1992. 

8. Alvaro R. Auzmendi, Total Right-of-way Analysis and Costing System (TRACS Version 4.3) User's 

Manual, Association of American Railroads, June 1994. 

9. CD. Martland, "Defaults for TRACS Based on Published Data Sources," Technical Note 94-2, 

prepared by the MIT Center for Transportation Studies, December 1994. 

10. D.D. Davis, "Rail Defects and Their Effects on Rail Maintenance Planning," AAR Research and 

Test Department Report No. WP-I54, April 1992. 

1 1. S.M. Chrismer and E.T Selig, "Mechanics Based Model to Predict Ballast Related Maintenance 

Timing and Costs," AAR Report R-863, July 1994. 

12. E.W. Smith, CD. Martland, S.N. Handal, D.R. Acharya, K.F Chiang and M.R McGovern, 

"Development of a Comprehensive Life-Cycle Costing Model for Railroad Turnouts," AAR 
Report R-854, 1993. 



met 




Corporate Office: 100 North Conahan Drive Phone: (800) 360-WEED 
Hazleton Commerce Center Fax: (717)459-5500 
Hazleton, PA 18201 



INCOMPOnATIO 

Hanwniac Tcckmiocy «>d fracTNiin the Ea^i w ii' kj. i 



RAILROAD VEGETATION MANAGEMENT 

♦ Chemical Weed & Brush Control Including: 

Roadbed 
Yards 
Crossings 
Off-Track 

♦ Tree Trimming & Brush Cutting 



SPECIAL SERVICES 

♦ Graffiti Removal 

♦ High-Pressure Washing with State- 
of-the-Art Water Recovery Systems 

♦ Under Bridge Crane Rentals 

♦ Liquid Calcium Chloride Sales 
Applications 



Afl work completed by experienced, licensed personnel to assure maximum efficiency. 

OFFICES: 



Hazleton, PA - (717) 459-5800 

Kansas City, MO - (816) 483-4478 



Chicago, IL - (815)464-9862 
Denton, TX- (817) 382-8883 




Railway Track Maintenance Equipment: 



Anchor Applicator 

Anchor Remover 

Auto - Lift Rail Lifter 

Grabber Dual Rail Spike Puller 

Super CLAW Single Rail Spike Puller 

MS Modular Maintenance Machine 

With Modules For: 

Cut Spike Feeding And Driving 

Nut Tightener 

Screw Spike Feeding And Driving 

Screw Spike Removing 

Spike Pulling 

Tie Boring 



Model "C" High Production Spike Driver 

Gauging Attachment For Model "C" 

Rail Drill 

Ride On Adzer 

Tie Plate Pre Gager 

Two Tie High Production Screw Spiker 

Tie Spacer 

Track Inspector 

FOR MORE INFORMATION CALL 

NORDCO 

RO. Box 1562 • Milwaukee, Wl 53201 
(414)769-4600 



319 




....Changing the Law of Gravity. 



320 



EFFICIENT LOAD TESTING AND EVALUATION 
OF RAILROAD BRIDGES 

By: Jeffrey L. Schulz, P.E.* and Brett C. Commander, P.E.** 

Abstract 

This paper describes a method that has been developed for efficient load testing and evaluation 
of short- and medium-span railroad bridges. Through the use of a specially-designed Structural 
Testing System (STS) and efficient load test procedures, a typical bridge can be instrumented and 
tested at 64 points in less than one working day and with minimum impact on rail traffic. With field 
data, a simple frame or finite element model is "calibrated" and its accuracy is verified. Appropriate 
design and rating loads can then be applied to the resulting model and stress predictions made. This 
"Integrated" technique has been performed on approximately 125 different structures located in the 
U.S. and in Asia. Both the merits and limitations of this approach are discussed, and several brief 
examples are provided. 



1. Instrumentation 

The key to this approach is the ability to instrument and load test the structure in a very short 
time compared to that required for standard static load testing methods. Rather than attempting to 
adhere foil strain gages to structural members, strain transducers are attached quickly with special 
tabs and adhesive or with C-clamps. Average installation time for the transducer is three to five min- 
utes compared to 15 to 30 minutes (or longer) for a standard foil gage installation. In addition to being 
easier to attach, the transducers are roughly 2-1/2 times more sensitive than foil gages configured in 
a typical quarter-arm bridge. The transducer's accuracy is approximately (3%. Furthermore, the 
longer gage length averages the strain over a three-inch length, making it much less susceptible than 
a foil gage to odd strain variations caused by the rivets or irregular surfaces. 

There are, of course, many data acquisition systems available that can be used for running 
bridge load tests. However, most of them are either designed for use in the laboratory and/or they are 
so application-general that often either a computer specialist or electrical engineer is required to con- 
figure them. The Structural Testing System (STS) discussed here was designed specifically for the 
task of running structural load tests, and therefore, the configuration has been "hardwired", dramati- 
cally increasing its ease of use. For example, as shown in , each STS unit accepts four strain trans- 
ducers and is connected in series with other STS units. This means that for many tests, only one cable 
runs from the Power Supply (located with the PC down on the ground) up to the bridge. This saves 
a significant amount of time in stringing cables. The STS units, which perform all of the signal con- 
ditioning and A/D conversion, are small enough to be placed on the structure. Many of the noise 
problems associated with data acquisition are eliminated since the data is transmitted digitally. 

Simplifying the field procedures even further is the built-in tracking system for the sensors. 
Each transducer is configured to automatically identify itself to the STS which in turn records their 
channel number and retrieves their calibration factor. This procedure is kept in the background, 
greatly reducing the possibility of note-taking errors in the field. The result is that only the transducer 
ID number (written clearly on each transducer), and its location on the structure must be recorded in 
the field. There is no need to track cables, junction boxes, or channel numbers. In addition, all of the 
connections between STS units themselves and between the STS units and the transducers are accom- 
plished through heavy-duty twist lock, military-style connectors. There is no soldering or "screwing 
in" for the wiring. 



"ChierTesi Engineer. Bridge Diagnostics. Inc. 
"Tesl Engineer. Bridge Diagnostics, Inc. 



321 



322 



Bulletin 761 — American Railway Engineering Association 



IZ V-BC 
ItD V-IC 



A 
11 11 



INTELLIDUCERS 
C4 PER STS) 




■^ - - - Up to 64 Channels 



Noiebook 
PC 



BDI - STS 

- Signal Condi'tioning 

- Anpllflca'blon 

- Balancing 

- A/I 

- Analog and Digltol Filtering 

- Signal Pass Through and Transnission 

Figure 1. Schematic of Structural Testing System. 



For effective use of the strain data, the position of the test load on the structure must be moni- 
tored. With railroad structures, this is exceptionally easy since the lateral positions are pre-defined. 
The longitudinal position is monitored by placing marks along the deck surface at convenient inter- 
vals and manually depressing a switch as the locomotive reaches each mark. This is accomplished 
through the STS with a Remote Position Indicator (RPI) which is based on two-way communication 
radios (walkie-talkies) that allow the personnel riding with the vehicle to depress a switch each time 
the vehicle's front axle crosses the designated mark on the bridge deck. This action sends a signal 
down to the receiving radio which is then read by the STS. The data, which is recorded as a function 
of time, can now be converted to a function of vehicle load position. 

It should be mentioned that most of the testing performed with the STS has been to record 
strains. However, other types of sensors can be used (full-wheatstone circuit-type transducers) such 
as accelerometers and LVDT's. 

2, Testing Procedures 

Another key to completing the test quickly is to use a moving the load to induce live load 
responses. A locomotive, with known axle weights, is driven across the structure at crawl speed 
(hence "semi-static") and data is recorded continuously rather than collecting data only when the load 
is placed at discreet locations. This has the benefit of greatly reducing testing time, thus minimizing 
the impact on traffic, but more importantly the quality of the test is greatly improved. The continu- 
ous data allows a good qualitative evaluation of the structure's live-load response since discontinu- 
ities and unusual responses, often associated with damage or non-linear behavior, can be easily 
detected simply by graphing the response histories. 

It is typical to instrument, load test, and remove the equipment in five to eight hours for most 
bridges. A variety of trusses and steel girder (boxes, plates, rolled sections) and pre-stressed (concrete 
boxes, I-sections, bulb tees, etc.) bridges have been tested and evaluated. Also, the approach has been 
used successfully on reinforced concrete structures (a large box girder, T-beams, and several contin- 
uous slab bridges). One note should be mentioned concerning R/C bridges. Surface strains on R/C 
structures have little value due to the presence of cracks unless the gage length is sufficient to aver- 
age out the effect of the cracks. Rather than attempting to remove the concrete cover and expose rein- 
forcement steel (a formidable task), special extenders are used in conjunction with the strain trans- 
ducers to average the surface strain due to the live-load. The required gage length is a function of the 
depth of the member and the length of the span. 



Paper by Jeffrey L. Schulz, P.E. and Brett C. Commander, RE. 323 



3. Analysis and Model Calibration 

In general, the primary function of the load test data in the Integrated approach is to aid in the 
development of an accurate structural analysis. Since a comparison of measured and computed 
responses is performed, it is necessary that the analysis be able to represent the actual response 
behavior. This requires that the actual geometry, boundary conditions, and load situations be realisti- 
cally represented. When end restraints are determined to be present, elastic spring elements having 
both translational and rotational stiffness terms are inserted at the support locations. In more compli- 
cated structures, such as trusses, a 3-dimensional frame model is generated. The reason that an analy- 
sis is required is that the loads applied during a test are usually not the critical load situations. An 
analysis is also required to completely identify the load paths associated with a structure. 

Loads are applied to the model in a manner similar to how they were applied in the field. A model 
of the loading vehicle, defined by a two-dimensional group of point loads, is placed on the structure 
model at discrete locations along the same path that the test vehicle followed during the field test. Gage 
locations identical to those in the field are also defined on the structure model so that strains can be 
computed at the same locations under the same loading conditions. All of the data input is performed 
visually with aid of computer graphics to ensure that the geometry, loads and gage locations are 
defined correctly. A structural analysis is performed on the model for all of the specified load positions 
and strains are computed at all of the gage locations. The accuracy of the model is determined by com- 
paring the computed strains with those measured in the field. Strains are compared visually with the 
use of various data plots and numerically using a variety of statistical relationships. 

In many situations, correct stiffness terms for many structural parameters cannot be determined 
through standard procedures. This occurs when design plans are not available, when significant dete- 
rioration is present, component interaction is complex, or when support details cannot be considered 
either simple or completely rigid. Various structural parameters can be evaluated by determining 
what values provide the best correlation between the experimental and analytical data. This process 
is performed with the aid of an structural identification process built into the analysis program. 
Iterations of analysis, data comparison, and model refinement are performed until the "error" 
between the computed and measured responses is minimized. 

The engineer's responsibility is to determine which parameters should be modified and to 
define reasonable lower and upper limits. The selection of adjustable parameters is defined by the 
unknowns and the visual comparisons of respon.ses. Experience in examining the data comparisons 
is helpful, however, two general rules apply concerning model refinement: When the shapes of the 
computed response histories are similar to the measured strain records but the magnitudes are incor- 
rect, this implies that member stiffnesses must be adjusted. Alternatively, when the shapes of the 
computed and measured response histories are not very similar at all, then the boundary conditions 
or the structural geometry are not well represented and must be refined. It has been observed many 
times that even relatively slight differences in boundary conditions will have large effects on the 
model's live-load response. Therefore, an analytical model that has not been calibrated with field 
measurements is highly suspect in its ability to predict the structure's actual behavior. In some cases, 
an accurate model cannot be obtained, particularly when the responses are observed to be non-linear 
with relation to load position. Even then, a great deal can be learned about the structure and intelli- 
gent evaluation decisions can be made. 

Once a "calibrated" model has been developed to the engineer's satisfaction, it can be used for 
a variety of purposes, most notably for predicting stresses due to a design vehicle or an overload. The 
specified load (or combination of loads) is applied to the calibrated model, and stress envelopes are 
generated for all the load conditions, allowing the maximum stresses to be obtained for each struc- 
tural component. The appropriate load rating equations, depending on the type of structure and con- 
trolling agency, are applied to obtain a rating factor for the structure or members in question. 



324 Bulletin 761 — American Railway Engineering Association 



4. Advantages and Limitations 

As discussed, the primary advantage of the approach described here is a great savings in time 
in both the field and the office, which of course, translates into a savings in money. In general, the 
field testing time with the STS is about one-quarter to one-third of that required for other systems 
based on foil strain gages (either bonded or weldable). 

However, it should be noted that this approach is used to evaluate the structure's overall live- 
load response, not for determining maximum stresses at connections and a stress concentrations 
(however, this could be accomplished with foil strain gages attached to the STS). In fact, these areas 
are specifically avoided during the instrumentation process so that overall bending and axial behav- 
ior can be determined. 

It should also be noted that this type testing is not meant as a replacement for inspection, but is only 
another tool for the evaluation engineer to determine the actual behavior of a structure. The final deter- 
minations obtained through this method are always a function of sound engineering judgment. This is not 
a "black box" approach that provides all the answers. In fact, sometimes unexpected field results lead to 
even further questions to be asked than were at the beginning! In addition, just because a load test has 
been performed, the load ratings will not always increase, in fact, sometimes they decrease. 

It is important to remember that linear-elastic response behavior is assumed when ratings are 
performed on loads greater than the applied test load. Since the stress levels are known this is gener- 
ally not a problem, however, when the strength of the structure is effected by secondary components, 
care must be taken. For instance, if unintended composite action is present or if the end restraints are 
substantial enough to significantly reduce mid-span stress levels, the engineer must decide if the sec- 
ondary component will remain in effect at higher load levels. Usually there are many ways to address 
these issues, the questions that must be considered are the age of the structure, the past load history 
(has the structure been overloaded in the past?), the projected load history (will it be significantly dif- 
ferent than the past?), structural redundancy (will failure of a secondary component have a signifi- 
cant effect on the structural stability?), etc. 

5. Examples 

A) Royal Gorge, Colorado: 

In some cases the goals of a load test are very specific and the test procedures can be designed 
to address the question at hand. For example, a hanging girder bridge in the Royal Gorge of the 
Colorado River, shown in Figure 2, had been modified with the addition of a concrete wall under the 
girder. The hangers needed to be modified or removed to increase the overhead clearance and it was 
necessary to determine what loads, if any, where still being carried by the hangers. The hangers and 
the girder were instrumented to determine how the loads were distributed. The field test was per- 
formed in under four hours with 20 channels of data recorded without disrupting the regular train 
schedule.' It was determined that the hangers were indeed carrying load, and it was recommended by 
the consulting engineer that the hangers be modified to obtain the desired clearance. 

B) Perak River, Malaysia 

This seven-span structure crosses the Perak River near Kuala Kangsar in Malaysia and was 
built in the eady 1900's (Figure 3). The Malaysian national railroad (KTM), was concerned about the 
significant side-sway that occurred as trains crossed the structure at speeds higher than 20 mph (35 
kph). The goal of this load test was to determine what mechanisms were causing the lateral motion. - 
Only one span was tested since they were all identical and behaving similarly. All instrumentation 
installation for 64 channels, testing, and removal was completed in approximately seven hours. 



Paper by Jeffrey L. Schulz, P.E. and Brett C. Commander, P.E. 



325 







Figure 2. Hanging girder bridge in the Royal Gorge of tiie Colorado River- 
Denver Rio Grand R.R. 



326 



Bulletin 761 — American Railway Engineering Association 



'l^^. 




Figure 3. Victoria Railroad Bridge over the Perali River, Kuala Kangsar, Malaysia. 



The actual load testing consisted of the train crossing the structure at three different speeds. A 
slow speed pass (5 kph) was first made and its position tracked for use in the modeling and analysis 
procedures. A medium-speed (35 kph) and a high-speed (50 kph) pass were also made to determine 
the structure's behavior under dynamic loads. 

A linear-elastic model of the structure was developed to obtain another view of the live load 
responses. The three-dimensional frame model, shown in Figure 4, was assembled with the aid of a 
graphic pre-processor. The structural geometry and member definitions were defined as realistically 
as possible. The analysis was performed for 20 different train positions so that strain histories simi- 
lar to those obtained from the field test were generated by the analysis program. 

Based on the observed data and correlation with the analysis results, several aspects of the 
structural behavior were identified. First, it was determined that asymmetric support conditions were 
present causing lateral displacement from vertical loading. The most likely scenario was that one of 
the roller supports was frozen or that the bridge was "twisted" due to uneven settlement of the sup- 
ports. It is likely that both conditions exist to some extent. Uneven resistance of the roller supports 
was verified by the analysis by placing spring elements, with variable stiffnesses, at the support loca- 
tions. The optimization procedure was implemented to obtain stiffness values that produced the best 
correlation between the computed and measured responses. The asymmetric support conditions, 
along with the geometry of the structure, caused an imbalance in the relative stiffnesses of the trusses 
and thereby eliminating the symmetry of the load transfer. Lateral deflections were then induced by 
the vertical loading. This effect was a static condition since the magnitude of the lateral deflection 
was dependent only on the magnitude of the load on the bridge. The magnitude of the static lateral 
deflections were small compared to the deflections felt during the high-speed train passes so it was 
apparent that support conditions were not the only problem. It was likely, though, that the imbalance 
of loads on this structure amplified the dynamic responses. 

Further investigation of the strain histories indicated that localized torsion deformations were 
occurring on the primary truss members throughout the structure (Figure 5). It was determined that tor- 



Paper by Jeffrey L. Schulz, P.E. and Brett C. Commander, P.E. 327 




Figure 4. 3-D frame model of railroad truss bridge. 

sional distortions in the bottom truss chords originated from the flexural response of the lateral floor 
beams in conjunction with the rigid connection between the floor beams and the bottom chords. Since 
the chord members of the truss have effectively no torsional stiffness, the end rotations of the floor 
beams induced a localized torsion distortion in the bottom chord, which was also seen as uneven axial 
forces in the vertical and diagonal truss members, which in turn caused torsional loading in the top 
chords. These distortions occurred at all of the floor beam locations and were cyclic with each axle 
crossing. The frequency of the distortions was directly related to the train's speed making it possible to 
excite the natural frequencies of the structure. Since the truss structures are stiff in the vertical plane and 
very flexible transversely, the fundamental mode of vibration occurred in the lateral direction. Based on 
these observations, it was determined that the observed lateral deflections were caused by resonant 
behavior. Because of the relatively large mass of the structure and train, combined with the lateral flex- 
ibility of the truss, the fundamental frequency was observed to be approximately 1/3 Hz. 

The importance of the this investigation was that many aspects of structural behavior become 
apparent with the availability of field measurements, which were relatively easy and inexpensive to 
obtain. Some of the same conclusions could be obtained in a purely analytical manner, but it would 
not be as obvious, and the conclusions would be difficult to verify. 

C) Norwalk, Connecticut. 

The Washington & Main Street Bridge was designed in 1 895 and currently carries the Metro 
North Railroad, Amtrak, and other local railroads on four commuter tracks. The bridge is a single 
span structure consisting of three parallel trusses, each separated by two tracks. The bridge geome- 
try was sharply skewed with variable angles at each support making the structure completely asym- 
metric. Because of the lack of symmetry and the era in which it was built, nearly every member was 
unique. All truss connections were pinned, the top chord members were riveted plate members, and 
the bottom chord and diagonal members consisted of multiple eye-bars with up to 6 parallel bars. 
Based on the age of the structure and analysis alone, the bridge was given a substandard rating. 

The main objectives of the load tests were to determine whether the rating results were accu- 
rate and to evaluate the consistency of bar forces in multiple-bar tension members. This was accom- 



328 



Bulletin 761 — American Railway Engineering Association 





AVERAGE STRAIN HISTORIES 
TRUSS MEMBER 2 (N-N) © MIDSPAN OF WEST TRUSS 




'c" 80. 
'6 

7 60. 


A I 




O 


awPWEo srMio V / \ \ 




1 '°- 


\. 1 u \ 




MAGNITUDE 
P p 


A Tl A A / \ 






2 -20. 
< 

t— 

m -40. 

LiJ 

o 

^ -60. 

UJ 

> 

< 


V ^ iCASURtD SIRWN 




-300. -250. -200. -150. -100. -50. 0. 50. 100. 150 200 




TRAIN POSITION (ft.) 





Figure 5. Measured and Computed Axial Strains From a Vertical Member near Midspan. 

plished by instrumenting several primary members on each truss. All of the bars on selected tension 
members were instrumented resulting in 160 gage locations. Because of the high volume of traffic, 
altering train schedules was not possible, so all instrumentation was attached between scheduled 
crossings. Testing was accomplished by recording strains during several normal train crossings. The 
type of train was noted with each recorded crossing so that axle weights could be determined later. 
Five instrumentation setups were completed, each consisting of 32 strain transducers, and several 
train passes were recorded during each setup. The entire instrumentation, testing, and take-down 
process was completed in five days without any track delays. 

By direct observation of the field data, it was determined that the consistency of stress among 
parallel eye-bars was generally within 30 percent. However, three members were found to contain 
bars which carried essentially no load at all. Upon inspection of these members, no defects in the 
members or pins were detected, therefore it was likely that the variation in bar force was due to dif- 
ferences in effective bar lengths. It was likely that the bars would eventually experience loading if 
enough displacement occurred. It is important to note that the loads present during the tests were con- 
siderably less than the design load. 

A 3-D frame model of the entire superstructure was generated and calibrated with the measured 
strains (Figure 6). From the calibration process, two observations were made. First it was determined 
that there was significant restraint at two of the three roller supports, which significandy reduced the 
tensile stresses in the bottom chord members. It was also determined that the floor system of the 
trusses was absorbing a significant portion of the bottom chord tensile force. The overall effect was 
that the measured tensile stresses along the bottom chord members were typically much less than ini- 
tially predicted. As would be expected the contribution of the floor was most apparent at the center 
truss. However, due to inconsistencies in the additional tensile resistance, it was recommended that 
the floor contribution be conservatively ignored in capacity calculations. 



Paper by Jeffrey L. Schuiz, P.E. and Brett C. Commander, P.E. 



329 




Figure 6. Computer Representation of Truss Structure. 



6. Conclusions 

Traditionally, load testing has been expensive, time consuming, and often resulting in ques- 
tionable results. The problems of cost and difficulty have been effectively eliminated with advanced 
technology and efficient field testing procedures. With today's experimental techniques, load testing 
can be performed quicker and more cost effective than a detailed finite element analysis. The ability 
to interpret load test results has also improved greatly because of the ability to process large quanti- 
ties of data and systematically compare analysis results with measured results. Significant efforts 
have also been made to keep the engineer involved in the evaluation process. The goal is to provide 
the engineer with additional information so that informed decisions can be made. Visual data com- 
parisons provide the engineer with valuable data and the ability to evaluate various analysis tech- 
niques. Since the accuracy of an analysis is known, a greater degree of confidence is obtained in the 
structural evaluation. 

The primary effect of the Integrated approach is a higher degree of accuracy in load rating. This 
is particularly beneficial when dealing with structures that are difficult to evaluate or when making 
decisions on borderline structures. 



References 

1. G.G. Coble, B.C. Commander, J.L.Schulz, "Load Tests on Hanging Bridge: Rio Grand Railroad, 

Royal Gorge, Colorado", Report submitted to Modjeski and Masters Consulting Engineers, 
July 1989. 

2. Bridge Diagnostics, Inc., "Load Testing and Evaluation: Victoria Railroad Bridge Over the Perak 

River, Kuala Kangsar. Malaysia", Report submitted to KTM Rail, Kuala Lumpur, Malaysia, 
December 1994. 




URVE 

CONQUEROR 

Rocia prestressed concrete ties conquer the curves. They 
provide a stiffer track modulus, improved lateral stability 
and gauge control, increased rail life, greater locomotive 
fuel economy, and significantly lower maintenance costs. 

Two North American manufacturing plants to serve you. 
Unsurpassed quality and durability. Go with the leader in 
North American prestressed concrete ties. Call: 



uaii: ^^ 



Concrete Tie, inc. 






f*»S'3 



'iga-MM^v 



701 West 48th Avenue • Denver, Colorado 80216 
(303) 296-3500 • Fax (303) 297-2255 

268 East Scotland Drive • Bear, Delaware 19701 
(302) 836-5304 • Fax (302) 836-5458 



330 



MANUFACTURING IN-HOUSE MACHINERY PROJECTS 

By: John H. Blanchfield'' 

I would like to thank everyone for this opportunity to present a brief overview of Norfolk 
Southern's approach to innovative design and manufacture of in-house machinery projects. 

During this portion of the briefing, I am going to take you on a tour of our processes for in- 
house design. How we begin the process, the origin. 

How we base our decision, the analysis. 

How we implement the design, and finally the construction of the machine. 

"Now we begin our tour" 

Our criteria for originating any project, whether in-house or outside, is based on four measures, 
"safety, productivity improvements, quality, and cost savings." 

Is the project a safety necessary one? Does the project create a safety problem? Are there 
ergonomics problems involved? 

Next we address the productivity improvement issue. Does the project improve our productiv- 
ity and by how much? Will the project cause a reduction in productivity in another associated area? 
The third question to address is quality. Does the project improve the quality of the affected area or 
item? Will the project enhance or diminish quality in some manner? And finally, what is the return 
on investment? Is the project cost effective? Is there an outside market? 

After determining that the necessary criteria has been met, we must answer several more questions. 

The first of which is, is the machine or device available from the commercial marketplace? 

The second question is, does Norfolk Southern have the engineering time to design the project? 

The third question is, will the project fit the shop production criteria? 

The fourth question is, does Norfolk Southern have the expertise for the proposed project? If 
all the.se criteria can be met successfully, we move to the next step in the process "The Design 
Parameters". 

From the outset, input into the project must be obtained from as many available resources as 
possible. At Norfolk Southern we solicit information from all phases of our operation; from the oper- 
ators, from the repair person, from supervision, and from our engineering section. 

We believe that this type of collective input enables Norfolk Southern to develop a compre- 
hensive design for the best possible equipment. 

First, we seek input from our operators, who play a very important role in the design process. 
Their input into safety, ergonomics, and operation are invaluable. 

Who better understands the daily obstacles that must be overcome to insure a successful outcome? 

We ask our operators "what if questions. 

"What if," we place this item in this location, would the machine be safe to operate? 

"What if," we placed this filter or grease fitting in this location, could you maintain these 
items easily? 

In the area of preventive maintenance, operators will be the first to tell you, the easier an item 
is to service, the more likely the item will be serviced. 



' Supcnntcndenl Field I^uipmenl, Norfolk Soulhcrn 



331 



332 Bulletin 761 — American Railway Engineering Association 



"What if," we placed this apparatus in this location, would the design result in a comfortable, 
easily operated piece of equipment? 

In the area of reliability and repair accessibility, the repair personnel are our best source of 
information. 

They are the people who know if an item can stand the vigorous test of time to last the intended 
life cycle. 

Our repair people also assist in properly locating machine components to allow the most effi- 
cient position for changing and maintaining such components. 

We then solicit information from our various areas of supervision. We rely heavily on supervi- 
sion for their contribution to safety, and what they believe can be done to improve the safe operation 
of the proposed machinery. 

We inquire into what changes are needed to improve quality. Additionally, we solicit produc- 
tion requirements for the intended operation. 

Production requirements which are attainable and, when applied, produce significant reductions 
in the cost of operations. 

After the gathering of all the information pertaining to the proposed project, our engineering 
section tabulates and analyzes all the available information. Feasibility studies are performed. 

The engineers seek solutions. They make inquiries into the latest state of the art technology and 
if, or how, it is applicable to the proposed project. 

After all of the above exercises have been successfully completed, we enter the next phase, 
which 1 will call the "action" phase. This phase consists of the actual design of the machine, appara- 
tus, or facility, the construction of the project and the final testing. 

The design portion starts with the actual layout of the project on a computer, utilizing a state of 
the art, three dimensional solid modeling computer design program or CAD. 

Following the initial design, a mechanics analysis is performed to check for interferences of 
moving parts and their stability. If all criteria have been met and satisfied, we then run a stress analy- 
sis on the individual components. At this point, if design changes are necessary, they are made and 
re-analyzed. 

The next step is for engineering to produce all the necessary computer programs for the shop 
computer numerical control (CNC) parts producing machinery. 

Our design computers have a direct data link to the applicable CNC machine whether the 
machine is a lathe, milling machine, or oxy-graph. This process ensures parts are made correctly the 
first time according to the design. 

The next step in the design phase is producing parts lists for ordering and construction. 

A machine manual is produced for shop and field use, and finally, a construction schedule is 
created to assist in the project construction. 

All the drawings can then be down loaded, via a data link, to the shop for either viewing from 
a terminal, or production of a hard copy of individual parts, or assemblies for shop floor use. 

The next production phase is construction. This phase is the culmination of all the preceding 
processes. This is the product coming into existence. 

In the first step of this phase, our engineering department, our shop, and our material section 
coordinate the purchasing of project materials. 

Deliveries are established and checked against the construction schedule, to ensure parts are 
available when they are needed and where they are needed. 



Paper by John H. Blanchfields 



333 



The shop superintendent will then coordinate the construction of the project between the vari- 
ous areas of the shop. 

Parts will be fabricated in the respective specialty areas for the specific part. 

Parts fabricated will be assembled into sub-components and, finally, sub-components will be 
assembled forming the overall machine. 

After all the components have been assembled, wiring and electrical work completed, pneu- 
matic and hydraulic components installed, and their associated hoses and pipes fitted, we are ready 
for testing. 

Testing consists of the setting of pressures for pneumatics and hydraulics, working each mech- 
anism separately or in related groups, and finally testing the entire system under as close to actual 
working conditions as possible, utilizing the operator, who will be assigned to working the machine 
whenever possible. 

After we are satisfied with the performance of the machine, we will paint it and apply all decals. 

Upon completion of painting and stenciling, we re-test the machine to assure no problems exist 
prior to shipment. 

The final step in the process is the placing of the machine in service. The engineers, who have 
designed the machine, are responsible for this step. This step serves two purposes. First of all. the 
engineers are the most familiar with the systems and their functions, which expedites operator and 
repair person training or solving any problems that may be encountered at start up and second, this 
is invaluable hands on experience for our engineers, providing practical experience and a sense of 
ownership. 

Now that I have described the process, 1 would like to present a brief overview of a project, 
which was recently completed utilizing the steps above. 

The project is our new welded rail pulling/pushing machine, WRPM. This machine is utilized 
in our dual rail laying operation. 

The original machine required two (2) people to operate, it was awkward to set up and we 
believed exposed personnel to a possible safety situation (Figure 1). 




. '^trtusi^mti^' 



Figure 1 



334 Bulletin 761 — American Railway Engineering Association 



The old machine utilized wire rope cable attached to a heavy pulling block, which was placed 
on the rail by a laborer. 

The cable was then pulled by one or both 60 ton winches thus pulling slack out of one or both rails 
in the narrow gauge cradles. As you can see, this machine was not conductive to an efficient safe opera- 
tion. Input from our field forces indicated that this was a potential candidate for in-house design. 

A meeting was held between field and engineering personnel. From this meeting, a consensus was 
reached on the parameters for the new machine. We needed a machine which was quicker and easier to 
attach to the rail, with minimum exposure to personnel and could push as well as pull the rail. 

From this meeting, a concept was developed. 

The concept was a machine which could clamp the rail, hydraulically move the rail in either 
direction by pulling or pushing with good control, and reducing exposure to personnel to a minimum. 

The next step was application of qualifying parameters. 

Will the concept provide for a safer operation? 

Yes, cables and heavy pulling blocks will be eliminated. This will reduce lifting of heavy 
objects and eliminate the possibility of broken wire rope cables. 

Does the project improve productivity? 

Yes, the new clamping arrangement is more efficient, reducing staging time. This will save 
approximately two (2) hours in a typical work day. 

Will the project improve quality? 

Yes, utilizing hydraulic clamping will allow the gang the ability to remove or place slack back 
in the rail as necessary while laying rail. 

Will the project provide cost savings? 

Yes, the project will result in a reduction of personnel. 

The next series of questions apply to in-house vs. out-sourcing projects. 

Was the machine available from the present marketplace? No, there are no machines of this 
variety available from the open market. If we contracted out the project, would there be a market for 
product by other railroads? 

No. Dual rail laying is not practiced on most railroads and this does not provide for an after 
market for the product. At this point, we will usually contact outside manufacturers to see if there is 
interest in manufacturing the product. 

The next question to be answered is, do we have the design time available for the project? 
Design time was available and schedules interfaced well. 

We then scrutinized the shop production schedule, to determine if any part of the project was 
not within our ability to manufacture. We found no problems in these areas either. 

Our final question to be answered was in the area of justification. We gave all figures to our 
accounting department, and positive return on investment was attainable. 

Design of the machine was begun. Engineering assembled all information gathered previously 
and set about their initial design. The initial design was conceptualized and presented to field per- 
sonnel for comments. 

The concept drawings were then formalized into working shop drawings. 

Stress tests were made on critical components and computer programs were written for CNC 
shop machinery. 



Paper by John H. Blanchfield 



335 



Engineering met v%'ith shop personnel to review the project and establish parts ordering and 
shop scheduling. A manufacture critical path was made and the construction phase was begun. 

After all construction, electrical wiring, and hydraulic and pneumatic pipe fitting had been com- 
pleted, the testing phase was started. 

Following the completion of testing, the machine was cleaned and painted. 

After painting all systems were re-tested to insure no problems had evolved as a result of paint 
application. 

And here you see the final product ready to perform in the rail laying consist (Figures 2, 3, 
4 and 5 ). 




^v_ ' a 



Figure 2 



Figure 3 



^^ 


' ll^^^^H 


— .,,. hw 


^^ 



Figure 4 



,^-r^ 



Figure 5 



I hope that everyone has found this presentation enlightening and informative. I have attempted 
to give you an overview of our in-house design procedure from concept through design, manufacture 
and testing, and finally as the machine appears in service. We at Norfolk Southern believe, through 
the utilization of our unique capabilities and talent, we produce safe, top quality, and efficient proto- 
typical roadway equipment at reasonable cost with good return on our investment. Again, thank you 
for this opportunity to present Norfolk Southern's approach to innovative design and manufacture of 
in-house machinery projects. 




336 



WAYSIDE SYSTEM FOR MEASURING RAIL 

LONGITUDINAL FORCE DUE TO THERMAL 

EXPANSION OF CONTINUOUS WELDED RAIL 

By: Larry L. Doll*, Harold D. Harrison and Thomas O. McCanney** 

1. Abstract 

Salient Systems, Inc. is developing the technology and the deployment strategy for wayside 
measurement and reporting of Rail Longitudinal Force due to Thermal Expansion of CWR. A sum- 
mary of this work is presented including principles of operation, test sites, and observations. Data are 
presented showing daily rail stress cycles, trends over weeks and months, and some unique events 
such as rail movement and winter snows. A perspective is given on the utility of the technology and 
future plans. 

2. Objectives and Approach 

Salient Systems, Inc. is developing the technology and the deployment strategy for wayside 
measurement and reporting of Rail Longitudinal Force due to Thermal Expansion of CWR. The 
objective is to develop a practical means to monitor rail longitudinal loads and temperatures at fixed, 
widely distributed locations. The approach is based on a low cost data collection module that can be 
mounted on the rail at widely spaced locations allowing for long term monitoring of the rail neutral 
temperature. The data collection modules, or rail stress monitors, consist of battery power, stress and 
temperature .sensors, signal conditioning, microcontroller, data buffer memory and communication 
means to transmit the data to a reader. Readers have been implemented using both hardwired RS-485 
links and RF links. To date, all readers have been fixed at the wayside and sites and have been lim- 
ited to the range of the transmission medium. Future readers will use RF links transported by a high- 
railer to enable wider geographical coverage in a cost effective manner. 

3. Principles of Operation 

The measurement is made using a circuit propcsed by Harrison'-^. As shown in Figure 1, lon- 
gitudinal and vertical stain gauges are placed on both sides of the web of the rail at the neutral axis. 
The gauges are configured in a bridge circuit so that longitudinal and vertical strains cause positive 
reinforcement of the bridge output voltage while compensating for temperature induced elongations 
in both directions. The result is a sensor which measures longitudinal .stress without being con- 
founded by either temperature variations or the degree of partial or full constraint of the rail with 
respect to longitudinal strain^. 

Equation ( 1 ) shows that the rail stress monitor measures longitudinal stress, even in the pres- 
ence of strains composed of thermal expansion and any degree of partially to fully constrained lon- 
gitudinal rail expansion. We will see examples in the data of rails which are fully constrained as well 
as examples of rails which slide under load. The sliding exhibits the most general ca.se where some 
movement accompanies both thermal expansion and internal longitudinal stress. 



— = — 2-((T, -vcTx -Oj + va, )- {a, =ar) 

V 2E 2£ 

e K^(\ + v) 

— CT/ when (Tt =0 

V 2E 



(1) 



• Manager Engineering Methods & Research. Union Pacific Railroad 

* Salient Systems. Inc. 

337 



338 



Bulletin 761 — American Railway Engineering Association 



-Longitudinal G«ges 




NOTE: The physUil ;•?• 
ler.gth Is l/l". 



Gag; lead Color Code 
Black - B 
Uhltt - M 
Aed • R 




LONGITUDtNAL RAIL STRESS CIRCUIT 



Figure 1. Rail and Circuit 

The rail stress monitor also records measurements of the rail temperature based on a thermis- 
tor. Equation (2) predicts the elongation of the rail in response to the combination of stress and tem- 
perature change. If the constraint is complete, then this elongation is zero, and the stress is propor- 
tional to the temperature difference relative to the rail's neutral temperature. 



^ E 



(2) 



For fully constrained rail : Ci - -EoAT 



p (steel) = 0.283 Ib/cu in 
a = 6.5 ppm/F 
£a = 195 psi/F 
V = 0.30 



A (133 lb/yd) = 13.05 sq in 
E = 30,000,000 psi = 30 psi/ppm 
AEa = 2.546 kip/F 
K„ = 2.03 



nV 



_d_e__ ^ 

da V Vpsi 



Temperature range = 1 50 F 
Stress range - 29,250 psi 



Signal range = 1 .3 mV/I 
Force range - 382 kip 



(3) 



Practical calculations require the physical constants and ranges of the variables involved. These 
values are important in understanding the data presented below. 



Paper by Larry I. Doll, Harold D. Harrison and Thomas O. McCanney 



339 




Figure 2. Block Diagram of Rail Stress Monitor and RF Interrogator. 



Figure 2 shows the layout for a system incorporating RF interrogation. The analog and signal 
processing hardware and software have been thoroughly checked out in a number of sites. The design 
incorporates low power CMOS integrated circuits and provides a .self-calibrating microprocessor. 
The communication capability is serial in nature, thereby making it easy to interface to RS-485 or RF 
transmitters. At hardwired sites, the host supplies power to the modules. At RF sites, lithium batter- 
ies supply power. Due to the low duty cycle, a battery life of several years can be expected. The 
design philosophy applied to this circuit includes the following elements. 

• Resolution of one degree Fahrenheit with corresponding resolution in stress and force. 

• Auto calibration to maintain zero stability, which is a key factor due to the nature of the strain 
gauge circuit's application. Since electronic drift is negated through the auto calibration pro- 
cedure, the zero point does not vary more than ±1 degree F (2500 lb) over the temperature 
range. The maximum aging rate is less than 0. 1 degree F (20 psi) per year. 

• Ability to withstand shock and vibration of direct attachment to the rail (approximately lOOOg 
worst case). Mechanical isolation between the module and the enclosure is the key here. 

• Ability to endure and measure over a temperature range of -40 to +160 F 

• Isolation of modules from all other circuits to provide both lightning protection and insure 
track signaling circuit reliability. 

• Provision of a "SLEEP" mode that shuts down all but a micro-power timekeeper circuit. 

4. Calibration and Hysteresis 

Installation consists of attaching the strain and temperature sensors to the rails, wiring the mod- 
ules to the sensors, and attaching them to the rails by means of their covers. It is also necessary to 
measure the offset of the stress measurement at zero rail stress in order to account for the tolerance 
in the sensors and the installation of the gauges. Consequently, installation is normally done on rail 
that is lying free in the ballast, before it is mounted on the ties. 



340 



Bulletin 761 — American Railway Engineering Association 



NEW SOUTH RAIL LYING FREE IN BALLAST 
SaBent Systems, Inc. Rail Stress Test at UP/Caliente 




20 30 40 50 60 70 80 90 100 110 120 
RAIL TEMPERATURE (F) 



Figure 3. Calibration on Free Rail. 

The hysteresis loop observed on the new rail, lying free in the ballast awaiting installation on 
the high side of the curve in Caliente Canyon, typifies the behavior of unconstrained rail (Figure 3). 
This curve represents the time from Monday at 21:00 hours (M21) through Tuesday at 23:00 hours 
(T23). The loop progresses counter clockwise. The sensor is mounted near the center of a rail approx- 
imately 800 feet in length. 

The vertical extent of the hysteresis loop in kips can be estimated by accounting for the force 
necessary to slide the rail back and forth on the ground. Thermal expansion causes the motion and 
friction between the rail and the ground resists it. 

The flat top and bottom of the hysteresis loop represent sliding at constant resistance. At each 
end of the day when the temperature reverses, the rail stress also reverses. Until the stress is large 
enough to cause sliding, the compression is restrained to a large degree. The steep sides have slopes 
which approach the theoretical 2.5 kip/F. (Note that the rail laid in a snake-like form along the slope 
of the ballast. It was neither straight nor level. For this case, only an approximation is relevant.) 

It is interesting to compare the free rail hysteresis loop to one for the same module taken after 
the rail was installed and welded into service (Figure 4). The depth of the hysteresis loop remains, at 
about the same level as in the ballast. 

5. Sites 

The rail stress monitor has been actively pursued over the past several years. Table 1 presents 
a summary of the test sites. The Gothenburg site was installed in Spring of 1990 and remains active 
today. The data from the Gothenburg and Donaldson monitors are being databased along with other 
train and environmental measurements to provide advanced capabilities for both railroad operations 
and new methods development. 

6. Thermal Cycles Observed 

The daily variations in temperature and rail stress are particularly striking in Caliente Canyon, 
Nevada. Figure 4 shows the daily cycle in hysteresis loop format. Additionally daily cycles retrace 



Paper by Larry I. Doll, Harold D. Harrison and Thomas O. McCanney 



341 



Table 1. Rail Stress Monitor Test Sites 



Site 


Railroad 


Temp 


Bridge 


Link 


Reader 


Gothenburg 


UP 


Yes 


Yes 


RS-485 


Fixed 


Pueblo 


TTC 


Yes 


Yes 


RF 49 MHz 


Fixed 


Bolivar 


CSX 


Yes 


Yes 


RS-485 


Fixed 


Caliente 


UP 


Yes 


Yes 


RF49MHZ 


Fixed 


Donaldson 


UP 


Yes 


Yes 


RS-485 


Fixed 


Blue Mts 


UP 


Yes 


Yes 


RF 900 MHz 


Hi-Railer 



approximately the same path, yielding a plot which shows the similarities of the cycles but makes the 
differences difficult to see. Plotting force and temperature versus time, as in Figure 5, opens up the 
daily cycles to comparison. 

Figure 5 trends the measured rail force and the independently measured temperature. These two 
plots are mirror images of one another. The temperature was fitted to the stress data using Equation 
(2) and the force calculated from temperature is displayed. It nearly duplicates the measured force 
data points. The apparent and instantaneous rail neutral temperature was calculated from each 
moment's temperature and stress values. The results appear in Figure 5 as a collection of points from 
about 80 F to 90 F. The best fit constants yielded a rail neutral temperature of 87.6 F The best fit slop 
was -2.06 kip/F, somewhat smaller than the theorefical slope of 2.55 kip/F for this 133 lb/yd rail. Also 
of note, the regression was done using the first 500 points of the data (about half of that shown in 
Figure 5) and resulted in a standard error of estimate of 4.7 kip with a r-squared of 0.990. 

At Caliente Canyon, a total of 20 modules were placed in 1000 ft. of the curve. It was only pos- 
sible to install three modules on the free (new south) rail in the ballast. The modules were placed at 
the middle and ends of the string. The rail gang used fire to heat the rail to its target neutral temper- 
ature so more modules would only have been burned. As a result, not all modules were calibrated to 
find the unknown offset caused by manufacturing tolerances in the strain measurement channel. The 
force/stress axes for the module at 5N is labeled accordingly. 



NEW SOUTH RAIL INSTALLED 

Salient Systems, Inc. Rail Stress Test at UP^Caliente 



6AM May 2 (M6) Ihru 8AM May 3 (Te). 1 994 




20 30 40 50 60 70 80 90 100 110 120 

PAILTEMPEnATUBE(F) 



Figure 4. Hysteresis Loop on CWR. 



34: 



Bulletin 761 — American Railway Engineering Association 



Salient Systems, Inc. Rail Stress Test 

UP/Cafiente, NV; 4AM May 24 • 1 1 AM June 1 , 1 994 

120i 




05/24 05/25 06/26 05/27 05/28 05/29 05/90 05/31 06/01 06/02 
TIME 



Figure 5. Daily Cycles. 



Salient Systems, Inc. Rail Stress Test 

UP/Caliente, NV: 941028:7AM - 941 105:4PM 




10/2810/2910/3010/31 11/01 11/02 11/03 11/04 11/05 
TIME 



Figure 6. Winter Events. 



Probably the most interesting feature of the daily cycles, shown in the May time frame, is the 
wide range of temperature endured by the rail. The sun warms the rail to 120 F while the air tem- 
perature may be 80 F. At night, the rail radiates into the clear desert sky and drops to freezing tem- 
peratures well below ambient. 



Paper by Larry I. Doll, Harold D. Harrison and Thomas O. McCanney 



343 



Figure 6 shows the effects surrounding an early winter event. It would appear that the rails were 
immersed in snow for a few hours on November 3. An interesting spike occurs in rail stress follow- 
ing the thaw. This is a very large spike in stress. It is well substantiated by the rail stress monitor 
which observed several monotonically increasing readings on the rising edge and several more 
monotonically decreasing readings on the falling edge. 

Location 6N is located on the north rail near a bolted joint. The joint has been visually and pho- 
tographically verified opening and closing each day in response to the rail stress cycle. The corre- 
sponding measured values are shown in Figure 7 for a few days in July. One of the cycles is expanded 
in Figure 8. The 6N location was actually calibrated on free rail during the process of transferring the 
rail from its original high-side location to the low-side of the curved section. 

Vertical lines mark the boundary between falling temperature and subsequent temperature rising. It 
would appear that stress is being relieved throughout the falling temperature/rising stress zone as indi- 
cated by the numerous discontinuities in the stress data. The temperature curve is notably smoother 

It has been possible to carry out long term monitoring at the UP Gothenburg wayside site. 
Figure 9 shows the rail neutral temperature measured by three rail stress modules on three of the four 
rails at this two track site. The rail neutral temperature is calculated assuming zero elongation. The 
small residual daily variation suggests that small movements occur in response to the daily cycling 
of thermal stress. 

Figures 10 and 1 1 show the longer term variation of rail neutral temperature ob.served over a year. 
A quantity of data was recorded on six occasions during the year. Both plots exhibit the same general 
trend. The vertical extent of the data scatter is due to the residual daily variations introduced above. 

Figure 12 shows monthly behavior in 1996 during November and December. The density 
exhibited in this plot is due to the fact that new databasing techniques are now automating the col- 
lection of the data. Each day's cycle is recorded with a resolution of about four points per hour. Closer 
study of the rail neutral temperature behavior is made possible by the introduction of such advances 
in data acquisition and data retrieval. 



Salient Systems. Inc. Rail Stress Test 

UP/Caliente, NV: 8PM July 1 6 • 9AM July 25, 1 994 




■100 



12:00 OO-OO 12:00 00.00 12:00 00:00 12:00 OttOO 12:00 
TIME 



Figure 7. Sliding at a Bolted Joint. 



344 



Bulletin 761 — American Railway Engineering Association 



Salient Systems, Inc. Rail Stress Test 

UP/Caliente, NV: QAM July 17 - 6PM July 18, 1994 



.A BoKed Joint Sliding. 




Figure 8. Sliding, Expanded View. 



Gothenburg Rail Stress Measurements 




12-20 12-22 12-24 12-26 12-28 12-30 
Date 



Figure 9. Monthly Neutral Temperatures. 



7. Perspective and Utility 

The rail stress project has produced at least three major accomplishments to date. 

• It has demonstrated the viability of the basic approach to measurement of rail thermodynamics. 

• It has developed a module with long life on a small battery and RF communications to a 
remote receiver. 

• It has developed a module capable of reporting rail variables to a host which supplies DC 
power and communicates over an RS-485 link. 



Paper by Larry I. Doll, Harold D. Harrison and Thomas O. McCanney 



345 



Gothenburg 



130 

£l25f 
a 

i 120 \ 
|115 
1 110 
1105 



H 



H 1 h- 



100 
Sep-94 Jan-95 Apr-95 Jul-95 Oct-95 Feb-9€ 



TrINrail 



Figure 10. Yearly Neutral Temperatures. 




Sep-94 Jan-95 Apr-95 Jul-95 Oct-95 Feb-9€ 



TrI.Srail 



Figure 11. Yearly Neutral Temperatures. 



The utility of this device to the railroads is expected to be: 

• Reduced risk of track buckling 

• Improved productivity of track maintenance crews 

• Track quality monitor 

• Monitoring other problems such as washouts and sink holes. 



346 



Bulletin 761 — American Railway Engineering Association 




North Rail 



11-01 11-11 11-21 12-01 12-11 12-21 
Gothenburg Track 1 



Figure 12. Monthly Behavior. 



8. Future Plans 

Salient is presently installing Rail Stress Monitors at a UP site near Donaldson, Arkansas. This 
system, like the one at Gothenburg, is hosted by the Salient SiteMaster system. The rail stress and tem- 
perature measurements will be collected along with train data, impact data and weather data into a data- 
base. The database will serve UP's Research and Methods Department. The database provides the tools 
necessary to automate management of the stream of train data generated by wayside equipment. 

Salient is working with the RF community to develop a robust yet cost-effective final design 
for the RF link that will allow isolated rail stress monitors to be interrogated, reliably, by a high- 
railer-borne reader. This last step will allow railroads to begin installing the monitors on rails for 
deployment throughout their regions of operations. 

9. Acknowledgments 

This work was supported in part by a grant from DOT/SBIR Program Office DTS-22 of the 
DOT/RSPA/Volpe National Transportation Systems Center under contract number DTRS-59-91- 
C-0008. Many thanks are extended to Dr. Andy Kish for his support and encouragement throughout 
the program. 

10. References 

1. H. Harrison, "Evaluation of methods for measurement of longitudinal rail force in unloaded 

track," Battelle Columbus Laboratories, Contract No. DOT-FRA-9162, Task 2, July 1981. 

2. H. Harrison, "Strain gage based longitudinal rail stress measurements at FAST," Technical Note, 

Battelle Columbus Laboratories, Contract No. DOT-FR-9162, October 12, 1983. 

3. H. D. Harrison, Thomas O. McCanney, and Larry L. Doll, "Wayside systems for measuring rail 

longitudinal force due to thermal expansion of continuous welded rail," SPIE Proceedings 
Series Volume 2458, 7-8 June 1995. 



CONRAIL'S INFRASTRUCTURE 
RELIABILITY OPTIMIZATION 

By: R. J. Rumsey* and R. Shiloh** 

Abstract 

All Railroad Engineering Departments are currently struggling with the question of how to 
maintain a reliable infrastructure system, while simultaneously reducing total maintenance costs. 
Historically, these reductions have been achieved through increased productivity, new materials and 
operational methods improvements. In the last 10 years, the introduction of new computer systems 
and new measurement systems including the geometry car, rail profile and defects detection, and split 
axle technology has created a new paradigm for measuring infrastructure reliability. Using this infor- 
mation will help to find the relationship between a "desired" level of reliability and total cost. 

Introduction 

This paper will cover the challenges facing the Maintenance Engineer concerning the complex 
balancing act between reliability and cost. 

Webster's dictionary defines reliability as "consistently dependable performance or result". 
This sounds very simple but at what cost is reliability achieved? Figure 1 shows what can happen 
when a large deviation from reliability occurs. This is something that we all want to avoid! 




Figure 1. Deviation From Reliability 



"Chief Engineer, Conrail 

' Asst. Chief Engineer. Planning & Support, Conrail 



347 



348 Bulletin 761 — American Railway Engineering Association 




Capita! improvement 

Long Term Investment 

to Acliieve Steady State infrastructure 

Figure 2. Railroad Infrastructure Fundamentals 

The Railroad Infrastructure Fundamentals can be illustrated as a pyramid (Figure 2). At the base 
of the pyramid is the area of Long-Term Capital Improvements, which is the investment required to 
achieve a steady state infrastructure. Steady state is the investment amount (Units) equal to the dete- 
rioration level. The second and third levels are all the activities associated with short-term infra- 
structure operation: Daily Maintenance and inspection. On the top of the pyramid is a small area that 
covers human errors associated with the aforementioned activities. We have an impact on this level 
both through continuous training, and through monitoring areas of negative behavior However, as 
many resources are allocated to minimize this area, it would not be eliminated. 

Measures and Drivers of RR Infrastructure 

What are the measures and drivers of the Railroad Infrastructure? The answer to this question 
can be summarized in one sentence: "Track and Signal Caused Train Delays". 

We are in the business of moving trains, and the level of transparency is our reliability measure. 
All other items such as slow orders, derailments, FRA (Federal Railroad Administration) and 
Geometry car defects, and other measures that are in use on most railroads, are sub components. That 
is, tactical measures, that in the aggregate are correlated to the ultimate infrastructure reliability. 

Reliability as a Function of Cost 

Each reliability level is a function of cost. Figure 3 shows a typical graph of the relationship 
between a Reliability Index (Q) and Cost. As in any other industry, as you approach the highest level of 
reliability (97-99%, three sigma) cost increases exponentially. Presently, most of us do not know exactly 
where we stand on this curve. This issue has historically been approached through the budgeting process: 
As more money is invested in the pyramid, it is assumed that reliability increases. This is probably a cor- 
rect assumption ! However, with the development in computer technology (data collection and analysis) 
in the Engineering Departments, a process can be developed to achieve some kind of optimization. The 
objective of this optimization will be: "Minimize Cost for Accepted Level of Reliability". 



Paper by R.J. Rumsey and R. Shiloh 



349 




Figure 3. Reliability as a Function of Cost 



1 

1 










UQL - Upper Quality Limit 




















LQL - Lower Quality Limit (Minimum) 








Time 





Figure 4. Deflning Reliability Limits — AQL — Acceptable Quality Limits 



Acceptable Quality Limits (AQL) 

Acceptable Quality Limits have two limits, a Lower (LQL) and a Upper Level (UQL) (Figure 4). 
Conceptually our concern is with the LQL-Lower Quality Limits-this is the acceptable minimum. 
The UQL generally is an outcome of the investment strategies (behaviors) that are implemented. 

Optimized Investment Cycles 

There are three potential investment .strategies. Figure 5 shows Strategy A. This is most likely 
a strategy that we all used in the past. At certain intervals (cycles), investment is initiated in order to 



350 



Bulletin 761 — American Railway Engineering Association 




Figure 5. Optimized Investment Cycles — Strategy "A' 



improve reliability. Due to a general desire to extend the work cycle, the investment is continued and 
it is done on a higher level to "last" longer. What can be concluded from this strategy? 

( 1 ) There are periods of time spent below LQL. These are areas of Reduced Reliability-below 
the accepted minimum. 

(2) There are also periods of time spent above the UQL. In general, being above UQL is not 
"bad". However, these are periods of over investment. These moneys could have been used 
elsewhere to improve reliability, or to support corporate projects whose return on invest- 
ment is higher than the cost of capital. 

An improved approach is Strategy B (Figure 6). Under this strategy, a minimum amount is 
invested to stay above the LQL. Assuming that the physical deterioration slopes for Strategies A & 
B are the same. Strategy B will create shorter maintenance cycles. This cycle length reduction can be 
viewed as a deterrent to productivity and cost. This should not be a major problem, if the additional 
cost of lower gang productivity and train delays is taken into consideration. With today's tools, this 
strategy (B) can be further improved by incorporating short-term, precisely allocated, maintenance 
cycles in order to achieve total optimization (Strategy C, Figure 7). 

To achieve this high level of optimization requires very close integration of Capital Investment 
(Long-Term) and Routine Maintenance. By combining these two different maintenance horizons, the 
costly Capital Investment cycle can be extended and reduced. 

Examples of Intermediate Maintenance are: 

• Surfacing (smoothing) of spots detected by the Geometry Car. 

• Rail life extension by preventive grinding. 

• Precise maintenance of weak spots. 

With the latest introduction of the Gage Restraint Measuring System (GRMS), this tech- 
nology will help to pinpoint weak infrastructure spots. With the development of software 
systems, the deterioration of weak spots can be projected, and improve the planning 
process that should help in better resources allocation. 

A major component of the optimization equation is to incorporate train delay cost. Usually this 
is a number that is difficult to enumerate. Using network simulations, this number can be developed 



Paper by R.J. Rumsey and R. Shiloh 



351 





V Strategy "B" 


K 1 
1 \ 1 




N \ 


1 S 1 






V _ _ ^_ „ 


1 \ 






V \ 




N, UOL 1 S 


.2 

a; 

1 


\ 


\ 




/\ 


1 


o 


. j| 


_ _ — _ 31 


^ _ 1- _ _N 


L -/ — _ ^ al 


. 


LQLI 


\ 1 / N 

\ I / s 

\ 1 / \ 








TiJe Strategy "A" 



Figure 6. Optimized Investment Cycles — Strategy "B" 



.5 

I 



Strategy "C 




Time Strategy -A> 



Figure 7. Optimized Investment Cycles — Strategy "C 



using sensitivity analysis. What is the impact on the network (locomotives, crews, cars and cus- 
tomers) if a train is delayed? This number will not be accurate, but will give us a rough approxima- 
tion. On specific routes this number can be calculated more accurately. 

Another major component in developing and executing optimization strategy is a commitment 
by all departments and management levels to the selected strategy. For instance, some changes are 
needed in the Accountmg System to allow flexibility in funds usage, in order to avoid higher costs in 
other categories (Capitalization vs Expense). 



352 Bulletin 761 — American Railway Engineering Association 



The final measure of our cost effectiveness is the total cost (Capital and Expense) of maintain- 
ing the infrastructure as a function of either Million Gross Tons (MGT) or an alternative measure as 
the denominator. 

Measuring Results 

Based on the objective of minimizing the cost for accepted level of reliability, the optimal strat- 
egy can be determined by calculating the Net Present Value (NPV) of all our projected activities and 
costs. However, these calculations require extensive analysis. With the approach of the 2l.st century 
and facing more competition, limited budgets, and increased traffic, we cannot afford not to do it. 
Since billions of dollars are spent on infrastructure maintenance, implementation of some of the tech- 
niques outlined above could potentially yield large savings with improved reliability. 

What is Done at Conraii 

Tools are currently being developed that will facilitate an in-depth analysis of our maintenance 
strategies. For example: 

• Improved Track Analysis System 

— GRMS outputs 

— Software improvement for production models 

— Failed tie detection system 

• Data Processing to Facilitate Development of Long-Range Plans 

• Changes in Accounting Systems 

— Continuously working with the Finance Department to review our practices 

• Introduction of Intermediate Maintenance cycles in response to LQL Deviation 

— Surfacing cycles 
— Crossing Renewals 
— Spot Undercutting 

• Advance Materials for Increased Reliability and Reduced Life Cycle Cost 

— Selective use of Pandrol Fastening Systems 
— Rubber crossing components 
— Switch turnouts components 

Summary 

Railroad Engineering Departments are facing big challenges ahead. In order to achieve a 
Reliable Infrastructure for Minimum Cost (Optimization), the following strategy should be adopted: 

• Current policies and practices have to be reevaluated for their contribution to the total 
optimization. 

• The key component of this evaluation is to not be afraid of breaking historical practices. 
Old policies and practices have to be adapted to recent developments in technology, mate- 
rials, computer science and network analysis. 

• To succeed, we have to work very closely with the Transportation Department to under- 
stand what the track reliability demand is. It is important to conduct trade-off analyses 
between engineering track requirements and train operations. The same close relationship 
is required with the Mechanical Department in order to develop technologies and practices 
that will reduce the overall stress (deterioration) of the infrastructure. 

• Continuous "What if questions have to be asked and their impact calculated. 
These questions include, but are not limited to: 

— Larger vs shorter maintenance cycles 

— High vs low tie density replacement 

— Impact of intermediate maintenance cycle 
Once a strategy is developed, it cannot be put away and forgotten. We have to constantly be aware 
of new developments in order to refine our working strategy. This will bring the Engineering Department 
increasingly closer to the optimization goal of minimum cost for accepted reliability level. 



FLORIDA'S HIGH SPEED RAIL PROJECT 
BUILDING A PUBLIC-PRIVATE PARTNERSHIP 

By: Claudio Dallavalle'" and Jack E. Heiss""* 

Introduction 

This paper examines the development of Florida's high speed rail project within the context of 
the Request for Proposals issued by the Florida Department of Transportation (FDOT) and the suc- 
cessful response submitted by Florida Overland eXpress (FOX). 

Development of the Request for Proposals 

Under the pressure of rapid population growth that has continued since the close of World War 
II, Florida is now the fourth most populous state and is projected to be third within 10 years. Eighty 
percent of the population inhabits the portion of the state in or south of a line connecting Daytona 
Beach and Tampa, almost exclusively along the coasts and the Interstate 4 corridor This concentra- 
tion of residents, along with a tourist population expected to be 60 million visitors by 2006 has 
prompted the state to seek transportation options other than air and highway. 

Expansion of existing transportation facilities has become either impossible or prohibitively 
expensive. Major international airports at Tampa and Miami are landlocked. The attempt to build a 
new airport at Miami met with vigorous environmental opposition. Expansion of highway capacity 
has reached the point of diminishing returns, with some new construction costing nearly $100 mil- 
lion per mile. The Florida DOT has now implemented a policy limiting interstate highway expansion 
to six general purpose lanes and four high-capacity vehicle lanes. 

In the early 1980's, development of high speed rail systems in Europe and Japan, prompted the 
State of Florida to investigate this form of transport and ultimately enacted legislation to implement 
a high speed rail system through innovative financing mechanisms. 

The Florida High Speed Rail Transportation Commission Act was made law in 1984. This act 
established a commission appointed by the Governor with the authority to grant a franchise to build 
a high speed rail system in the state. The 1980's was a period of rapid expansion in Florida with easy 
money being made in real estate development. This situation led to the belief that high speed rail 
transportation could be built with "excess" profits from such development, with no reliance on pub- 
lic financing. 

In early 1988, the commission issued a Request For Proposals and in late 1988 received two 
applications. After two years of review, it became apparent that further pursuit of a franchise under 
the conditions established was futile. Both applicants abandoned the process. 

Florida was left with a failed scheme and a shattered dream. However, the needs of the trans- 
portation system still loomed. It became obvious that there were lessons to be learned from this ini- 
tial attempt which failed quickly and for multiple reasons. 

To start a recovery, Florida's 1991 legislature abolished the High Speed Rail Commission and 
assigned responsibility of the project to the Florida Department of Transportation. By the 1992 leg- 
islative session, the FDOT had evaluated the original 1984 process and made recommendations on a 
revised process. The legislature incorporated the changes into the highly- revised Florida High Speed 
Transportation Act. 

Major changes resulting from the previous failed effort are: (I) the expenditure of public funds is 
specifically acknowledged as a legitimate option; (2) the franchise is to be awarded early in the process. 



♦Engineering Manager, Florida Overland eXprts.s 

♦Railway Engineering Adminisiraior, Florida Deparlmcnl ol'Transporialion 



353 



354 Bulletin 761 — American Railway Engineering Association 



eliminating the need for excessive expenditures by the applicants in developing permit-quality informa- 
tion while still in competition; and, (3) FDOT is granted the authority to develop innovative financing 
mechanisms, including Public-Private partnerships, to implement the high speed rail system. 

1. The expenditure of public funds is specifically acknowledged as a legitimate financing tool. 

Florida's initial attempt suffered from a 'high speed rail is free' myth. While 1984 Act did 
not prohibit the expenditure of public funds, one applicant touted that public finance was not 
needed since real estate development rights awarded with the franchise would build the sys- 
tem. This concept was embraced by the commission and the popular press. The other fran- 
chise competitor stated that approximately $500 million in public funds would be needed for 
their system. Constant pressure over these differences led the competitor seeking public 
funds to withdraw. Not surprisingly, however, the remaining competitor revised its applica- 
tions to require significant public participation, well more than $500 million. The collapse 
of the process soon followed. 

Analysis revealed that to generate enough 'excess profits' to finance a high speed rail sys- 
tem, the franchisee would need to have exclusive development rights for a lion's share of 
the entire Florida market over many years. Also, real estate development is slow to gener- 
ate profits, requiring years to accumulate enough capital to construct a system. The alterna- 
tive of selling the development rights to third parties left advocates of growth management 
pondering a future of unrestrained expansion of suburban tracts and strip malls, the antithe- 
sis of a high speed rail transportation system. 

Analysis shows that no transportation system exists without public participation. While 
today's freight rail system operates without ongoing public participation, many railroads 
were helped through capitalization. Today, both the highway system and the aviation system 
would not exist or even operate without capital investments and on-going subsidies beyond 
those provided by user fees. 

The lesson learned was if Florida wanted a high speed rail system, Florida would need to be 
a financial participant in the system. 

2. The franchise is to be awarded early in the process, eliminating the need for excessive 
expenditures by the applicants in developing permit quality information while still in 
competition. 

Florida's original franchise concept combined franchising and permitting into a single 
award. While envisioned to give the state the best possible system honed in a competitive 
atmosphere, the process required the franchise applicants to develop permit-quality infor- 
mation while still risking a zero outcome. 

Besides the high level of at-risk expenditure, a secondary effect was a reluctance by com- 
petitors to reveal their plans lest they be incorporated into the competition's plans. The with- 
holding of information until the last possible moment delayed progress, making a bad risk 
situation worse. 

The lesson learned is that the private sector is very careful about placing capital at-risk. 

Also, the process should not require private capital to be committed before it is actually 
needed. Likewise, information should not be developed until it is needed. 

3. FDOT is granted the authority to develop innovative financing mechanisms, including 
Public-Private partnerships. 

Public-private partnerships is a much used and abused term. It is often used to bring the 
mystique of entrepreneurship to a project but when translated in action it usually means the 
private sector pays and the public sector regulates. 



Paper by Claudio Dallavalle and Jack E. Heiss 355 



A partnership, to be successful must play to the strengths of the partners. With large infra- 
structure projects such as high speed rail, the strength of the public sector is its ability to 
raise capital. The private sector's strength lies in its ability to adapt to market conditions. 
These strengths can be brought to bear through structuring a business arrangement with the 
proper incentives and well-defined common goals. 

The lesson learned is that a partnership requires both participants to share the risks and 
rewards, doing what each partner does best to reach the common goal. 

Based on these lessons, FDOT formulated a new Request for Proposals in 1995. This RFP 
attempted to structure an arrangement built on the strengths of the public and private sectors. In the 
state's view, the public sector's strength was its experience in permitting large transportation projects 
and to raise capital. The strength the state sought from the private sector was entrepreneurial spirit 
and the ability to develop a marketable transportation product required for commercial success. 

To implement these concepts, the RFP asked the private sector applicants to describe their own 
concept of the transportation product within a framework limited only by three factors: (1) the sys- 
tem must connect the southeast Florida metropolitan area with the Orlando and Tampa Bay regions; 
(2) the system must operate more than 120 miles per hour; and, (3) its technology must be proven or 
commercially available. Additionally, the RFP, while offering public financing for most of the pro- 
ject, also required the franchise applicant to be a financial participant. 

Within this framework, the basic strengths of the partners are engaged. The public sector raises 
capital and the private sector designs a marketable product and takes financial risk in selling that 
product. 

Figure 1 is included to provide the reader with an overview of the magnitude of the project. 
The map depicts the alternative alignments which the preliminary and EIS development phases will 
evaluate before selecting the optimal route. 

Development of the "FOX" Proposal Response 

General 

The goal of any proposal effort should be to construct such compelling arguments that the client 
has an overwhelming desire to make you the preferred contractor. The keys to a successful proposal 
are not a result of luck, or achieved by accident, rather, to win requires hard work, commitment and 
a good proposal strategy. 

This section will examine the Florida Overland Express's (FOX's) response to the State of 
Florida's High Speed Rail Request For Proposal {Figure 2). This examination will be divided into 
the three phases of the proposal development process: PRE-PROPOSAL, PROPOSAL, and POST 
PROPOSAL. Furthermore, each stage will be presented in two parts. First, key general guidelines in 
producing a winning proposal, followed by how these principles were applied in preparing FOX's 
response to the State of Florida's High Speed Rail Request For Proposal. 

Pre-Proposal Activity 

General 

The old adage "The Early Bird Catches the Worm" is analogous to "The company who does 
their homework -wins". Generally, if a company waits until a proposal is issued before it informs 
it.self about the pending project, it is at a distinct disadvantage in winning of the project. The issuance 
of a formal Request For Proposals is NOT the beginning of the development process. Rather, it is the 
culmination of the first phase of the project development process. 



356 



Bulletin 761 — American Railway Engineering Association 




Figure 1 



The pre-proposal activities that companies engage in should focus on POSITIONING and 
INFORMATION GATHERING as well as laying the groundwork for the proposal itself. The POSI- 
TIONING process is an internal educating and evaluation process whereby the potential responder 
sensitizes itself to the impending project. It is a self analysis of strengths, weaknesses, resources, and 
so on, of the firms success potential. INFORMATION GATHERING focuses on potential clients, 
their organization, market trends, competitor evaluations, commercial risk, etc. This early assessment 
tells you whether a firm has a reasonable chance of winning the contract if it bids. 

In conclusion, the main reason for a company to conduct a complete pre-proposal exercise is to 
position itself to intelligently assess the opportunity in a complete and systematic manner. 

Florida High Speed Rail Pre-Proposal Activities 

After the collapse of the first high speed rail effort in 1 990, the State of Florida solicited comments 
and inputs from private industry on their proposed new effort to bring high speed rail to Florida. 

The informational package issued by the State asked the industry to review and comment on the 
plan primarily focused on the "WHAT' and "HOW" of what the state wanted to do with respect to 
high speed rail. The document touched on such diverse issues as routes, technology, financing, project 
structure, operations, franchise duration, and public private partnership concepts. Review of the State's 
document inevitably lead readers to surmise that if a company chose to pursue this effort it could only 
do so as part of a diverse world class team. It was obvious that any one company could not meet all 
the proposed project's needs from either a delivery, or a risk perspective. For example, engineering and 
construction firms would not be in a position to manufacture locomotives or rail cars, nor would rail 
and car manufacturers be able to design and build an infrastructure project of this magnitude. As a 
result it became apparent that it would require a diverse team to pursue this project. 



Paper by Claudio Dallavalle and Jack E. Heiss 



357 



FLORIDA DEPARTMENT OF TRANSPORTATION 
REQUEST FOR PROPOSAL 



TABLE OF CONTENTS 



Part A INTRODUCTION and INSTRUCTIONS 

To Applicants 

Section 1 INTRODUCTION 

Background 

Project Prospective 

Section 2 STATUTORY OVERVIEW 

Legislative Intent 

Franchise 

Post-franciiise Agreements 

Certification 

Section 3 PRE-APPLICATION MATTERS 

Information Sources 

Letters of Interest 

Requests for Notice 

Voluntary Pre-application Conference 

Process for Classification of RFP 

Limitation on Communications with Department 

Personnel and CPEAC Members 

Section 4 PREPARATION OF APPLICATION 

Scope of Response to the RFP 

Application Format 

Executive Summary 

Location and Time for Submission of Applications 

Copies to Other Agencies 

Confidentiality and Proprietary Information 

Application Fees 

Public Entity Crime Statement 

Waiver of Irregularities 

Right to Reject Application 

Section 5 REVIEW AND SELECTION PROCESS 

Schedule 

Evaluation Criteria for Franchisee Selection 

Department's Review, Evaluation and Selection Process 

Part B THE APPLICATION 

Chapter I BUSINESS ENTITY INFORMATION 

AND APPLICANT QUALIFICATIONS 

Ownership and Form of Business Entity 

Business Entity Financial Information 

Applicant's Prior Experience 

Assignment of Responsibilities 



Chapter II MARKETING PLAN 

System Goals 

Service Areas 

Marketing 

Florida's Travel and Transportation Market 

Market Approaches 

Service Plan 

Chapter III SYSTEM DEVELOPMENT PLAN 

Routes 

Technology and Performance 

Overall System Plan and Service Concept 

General Technical Information 

Rolling Stock 

Track/Guideway 

Signal and Communications/Train Control 

Research or Development Requirements 

Pre-revenue Service Testing Program 

Maintenance Systems 

Management Information System 

Regulation Compliance, Waivers, Certificates of 

Particular Applicability 

Workforce 

Labor Contracts 

Chapter IV ENVIRONMENTAL AND 

COMMUNITY EFFECTS 

Environmental Compliance Issues 

Areas with Unique Environmental or Cultural Qualities 

Growth Management Issues 

Governmental Coordination 

Chapter VI PROPOSED LEGISLATION AND 

FRANCHISE DOCUMENTS 

Terms and Conditions of Franchise 

Post Franchise Agreement Proposals 

Legislative Requirements/Constitutional Revision 

Chapter VII BENEFITS TO FLORIDA 

Transportation Benefits 

Environmental Benefits 

Energy Savings 

Economic Benefits 

Other Topics 



Figure 2 



358 Bulletin 761 — American Railway Engineering Association 



From our company's perspective (Fluor Daniel) an engineering company, our first course of 
action was to make a "GO or NO GO" decision in the pursuit of this project. Armed with the State's 
informational package, knowledge of the State's previous high speed rail efforts, along with an in- 
depth working knowledge of high speed rail, a POSITIONING or opportunity assessment was con- 
ducted. An internal self analysis of our strengths, weaknesses, resources and so on was made. At the 
end of this first step, we concluded that we were internally strategically well positioned and could 
provide a team with the following competitive edge: becau.se we (Fluor Daniel) are the largest pub- 
licly traded engineering and construction company in the United States we had the required available 
technical resources to do the work, as well as, having the financial stability and market credibility. 



KEY ISSUES MATRIX 

INTERNAL SELF ANALYSIS 

Weighted 
EVALUATION FACTORS YES NO Factors 

1. Does the project comply with Company objectives 

2. Do we meet the project's requirements 

3. Do we have the resources necessary to support this effort: 

People 
Financial 
Local Presence 

4. Are we price Comjjetitive 

5. What is our strategic position 

1. Are we a leader/follower 

2. Team Positioning Lead/Join 

6. What are our strengths 

1. 

2. 

TOTAL INTERNAL Go/NO-Go SCORE 



EXTERNAL SYSTEM ANALYSIS 



Weighted Company Competitor 
YES NO Factor Rating Assessment 



1 . Is there an incumbent contractor 

2. Prioritize the available technologies 

1. 

2. 

3. Is the rail technology committed and with who 

1. 

2. 

4. Competitor Team(s) 

1 . Strong/NeutralAVeak 

2. Strong/NeutralAVeak 

5. Project Analysis 

1. Go/Nogo Probability Strong/NeutralAVeak 

2. Project Support Strong/Neutral/Weak 

3. 



TOTAL EXTERNAL SCORE 



Paper by Claudio Dallavalle and Jack E. Heiss 359 



Next, an assessment of the client organization, market trends, competitor evaluations, commer- 
cial risk, etc. led us to believe that if we were teamed with the right technology, we had an excellent 
chance of winning the contract. As part of our market analysis, an in-depth key-issues matrix analy- 
sis of the different rail technologies and car manufacturers was developed and conducted, and the 
strengths and weakness of each system were identified and charted. The matrix included reviewing 
each system provider's technology (each technology was rated based on functional applicability), the 
vendors quality of product, financial well being, and delivery history. In addition, potential major 
engineering and construction firms were evaluated as both potential partners, as well as competitors. 

As a result of the aforementioned analysis a decision was made to continue to monitor the pro- 
ject's development and open preliminary dialogues with the various rail manufacturing firms. As one 
can see, the assembling of the FOX team came about not by chance but as a direct result of hard work 
and many discussions between the marketing, technical and management staffs of a variety of firms. 
A excerpt of our Key Issues Matrix analysis is included below as for reference purposes. 

Once the State officially released the Request For Proposal in February of 1995, the FOX con- 
sortium was officially formalized and a final consortium membership of four global companies form- 
ing a limited partnership was settled on. Because the project required that a safe and proven high 
speed rail technology be used, G.E.C. Alsthom, developers and manufacturers of the world's fastest 
and safest rail equipment and one of the leaders in power generation and systems was selected (the 
TGV technology ). Bombardier, was selected because they are the largest rail car manufacturer in 
North America and own the license to build TGV equipment in North America. Fluor Daniel, the 
largest publicly traded engineering and construction firm was selected because of its design and man- 
agement expertise, and Odebrecht Contractors of Florida were selected because of their construction 
expertise. The key factors in determining final FOX consortium members were: 



TECHNICAL COMPETENCY 

FINANCIAL SOLVENCY and STRENGTH, and 

WORLD CLASS STATUS 



PROPOSAL ACTIVITY 

General 

A successful proposal response is one which is perceived to satisfy the client's requirements. In 
other words, to be successful a respondent must communicate their response in such a way as to be 
judged best in satisfying the client's needs. The five keys to preparing a winning proposal response are: 

1 . Listen to the client, and find out what the client's real issues, concerns and desires are. This 
is done by reading and understanding all written communiques regarding the proposal (espe- 
cially if the client asks for comments during the proposal development process). 

2. Understand the Client, review past client proposal behavior, understand the client's proposal 
and review staff. 

3. Ask intelligent questions aimed at learning the client, obtaining information, but never ask 
strategic proposal development questions which could become a matter of public record. 

4. Identify your competitive strengths and weaknesses and then devise an effective proposal 
strategy. 

5. Plan the effort; there is no substitute for good plannmg. Good plannmg ensures that the 
resulting proposal will be focused, responsive, professional and timely. 



360 Bulletin 761 — American Railway Engineering Association 



Florida High Speed Rail Proposal Planning 

The Request for Proposal was issued by the State of Florida on February 28, 1996. The docu- 
ment itself was less than three-eighths of an inch thick. However, the RFP was comprehensive and 
clear as to its intent, it was direct with regards to form, substance and timeliness, and yet it gave 
potential bidders room for creativity. The Figure on the following page is the RFP's "Table of 
Contents" it is included for illustrative purposes. 

The key to winning is good planning . And good planning leads to the development of a good 
proposal strategy. Developing a proposal response without focus or strategy is analogous to Lewis 
Carol's story "Alice in Wonderland" as described below: 



At one point in the story, Alice asks the Cheshire Cat if it would please 
tell her which way she ought to go. To which the Cat replied, that it 
depended a good deal on where Alice wanted to go. Alice then replied, 
that she didn't much care where she went. To which the Cat replied, then 
it doesn't much matter which road she took. 



Like Alice, if a proposal response lacks direction (FOCUS and STRATEGY) then it does not 
matter how one develops it, for it will surely miss its mark. Therefore, a good strategy identifies to 
the writer the issues they must address to satisfy the client's needs. For it is these issues that gener- 
ally form the basis for the client's selection. 

The steps used in developing our winning proposal consisted of: 

/. Issues Identification: It was our belief that the two underlying key issues for the state were: 
( 1 ) improve Florida's statewide transportation network by developing a cost competitive high-speed 
rail transportation alternative, and (2) create a public-private partnership which together would invest 
time and resources into the development of the high speed rail transportation system for Florida. 

2. Differentiating Feature Identification: Our focus here was to emphasize the technological 
and managerial advantages of the FOX team. FOX's proposed technology consisted of using the 
fastest and most proven high speed rail system in the world today, the TGV. Highlighted were such 
facts as: (1) over I billion passenger miles without a fatality, (2) state of the art technology, (3) fif- 
teen year track record, and (4) the fastest trains available in the world. Managerially the focus was 
on the team's record of: (1) meeting project schedules, (2) being cost competitive, (3) having on staff 
world class rail, engineering and construction expertise, and (4) clearly identifying the benefits for 
Floridians from the project. 

3. Completeness of Effort: Key to any successful proposal development effort is complete- 
ness of effort. The more specific and concise one is with their responses the greater the likelihood of 
success. An example of FOX's approach to completeness of effort is illustrated below. Within the 
time parameters allotted by the proposal, a comprehensive alignment fatal flaw field survey was con- 
ducted to ascertain and assess first hand actual in-situ site conditions which were used as the basis 
for cost and financial proforma estimate development. Figure 3 graphically depicts FOX's fatal flaw 
approach to developing its capital cost and proforma estimates. 

4. Strengths and Weaknesses: FOX's strengths were identified and expanded upon through- 
out the proposal. However, key to our success was our truthful approach to weakness evaluation. For 
example, within the time limitations allowed for response FOX made a concerted effort in both weak- 
ness identification as well as formulating potential remedial recommendations. For example, ( 1 ) rail 



Paper by Claudio Dallavalle and Jack E. Heiss 



361 



Development of Capital Cost Estimate 



STEP II 
Fatal Flaw Analysis 



HDR 

Tampa ■ Polk Co 



Carter Burgess 

Polk Co. ■ Okeechok-e Co. 



STEP I 

Develops Concept 

Alignments 



FH^ 



HNTB 

Okcechohi'c Co. - Miami 



Universal F-ield Services 
ROW (Total Line) 



GEC 

Preliminary Gcolcchnical 
Roule .Soil [ivalualion 



Woodward Clyde 

Preliminary Environmenlal 
Reviews 



Bermello & Ajamil 

.Station Coneept 
Development 



FOX 



STEP III 

Independent 

Criteria & 

Quantity Check 

SYSTRA/SNCF 




STEP IV 

Independent 

Audit of Capital 

Construction 

Estimate 



> 



Independent 
Audit 



FOX 



Specialty Estimators 

— Civil 

— Structural 

— Rail 

— Architectural 

— System 



Procedures & 
Documantation 

— Quantities 

— Unit Prices 



Field Veriflcation & 

Preliminary Unit Prices 

& Project Costs 



7 



Figure 3. Fatal Flaw Capital Cost Estimate 



noise could be made into an issue if not addressed. Therefore in our proposal a detailed discussion 
and overview of rail noise was presented as well as the potential means of dealing with noise. (2) 
EMF would also potentially be construed as an issue by the public at large, here too, a detailed in- 
depth discussion was presented alleviating many misconceptions of EMF; (3) environmentally, once 
a corridor was identified, a great deal of time was spent on assessing the environmental impacts of 
the systems and route. Though no final solutions were adopted, FOX let the reader know it was aware 
of potential issues and that it was ready to deal with them if selected. By FOX identifying and 
addressing the issues it was able to take a proactive lead thereby minimizing criticism. 

5. Competitor Identification: Since we did not know what our competitors were proposing we 
did not focus on this issue. Rather, our focus was internal emphasizing our strengths. 



362 Bulletin 761 — American Railway Engineering Association 



Florida High Speed Rail Proposal Preparation 

The final step in the proposal development process was the creation of the document itself. The 
steps followed in the FOX proposal development process were: 

1 . Set up the proposal development story board war room 

2. Brainstorm the section themes and content 

3. Define proposal format 

4. Outline each section, and prepare first drafts 

5. Write, review and polish the draft 

6. Conduct and independent review and critique 

7. Prepare final draft 

8. Managerial review 

9. Produce and submit 

It is our belief that key to our success was that as the various sections were being developed 
each proposal contributor was asked to develop their section by asking themselves the following 
Questions: 



What is the point of this section 

What's in it for the state 

Why should the State select FOX. 

What makes FOX different and unique and, why that is important to the State, and 

What added value does FOX provide the State. 



In conclusion, it is important to remember that a good proposal is not just writing a winning 
document but it's one that is able to lay the foundation for a successful project execution and client 
relationship. 

Post-Proposal Activity 

General 

Whether you win or lose it's always important to know why. Once the contract has been awarded 
and regardless of the outcome-feedback from the client on all aspects of your proposal is important 
and can be used on the next effort. Remember, a loss is still a transaction, it is an opportunity to get 
to know the client and vice versa and new market penetrations often involve losses initially. One 
should not get discouraged. 

The Next Step 

Building a Public-Private Partnership 

Even though the project is currently in its very early developmental phase, the Public-Private 
Partnership tone was set early during the contractual negotiations. Once selected, FOX and the State 
agreed to enter into the negotiation process not as adversaries but rather as two halves working 
towards a common goal. Using "Partnering and Teaming " techniques to establish both the behav- 
ioral ground rules and structured organizational working formats, both side entered into the negotia- 
tions with a "Win-Win" attitude. 

During the contractual work plan development and negotiation process it became evident that 
each entity brought to this effort a set of strengths and capabilities which complemented and com- 
pleted each other. These strength characteristics were identified and performance roles evolved as a 
result. A list of some of the key strengths of each is presented below: 



Paper by Claudio Dallavalle and Jack E. Heiss 363 



SECTOR STRENGTHS 

PUBLIC PRIVATE 

Better permitting capabilities More technical resources 

Ability to issue Tax Exempt Bonds Rapid project delivery focus 

ROW acquisition Management geared towards efficiency 

Eminent domain rights 

Technical support Construction/operation capabilities 

Seed money Private capital 

Conceptual design Complete design capabilities 

To date, the "Public-Private" Partnership process is developing into a mutually beneficial working 
relationship. However, like any other process there are positive features and there are some negative 
drawbacks, highlighted below are what we have identified as being the key advantages and disad- 
vantages to engaging into a "Public-Private" partnership. 



KEY ADVANTAGES 

Rapid Project Delivery 

Access to Larger Capital Markets 

Public Funds stretched over more projects 

DRAWBACKS 

Some reduction in direct control 

Break in traditional procurement processes 

Establishing a reliable funding stream 



Summary 

In conclusion, both the State of Florida and FOX's belief that the framework for building a true 
"Public-Private" Partnership has been formulated, and that the Florida High Speed Rail project could 
not be financed and developed solely by either entity and, that its successful fruition is dependent on 
our partnership. 




Star Track? 



The railroad problem In your backyard might seem as tough as 
building a line on the moon. 

A&K welcomes your railroad material problems as our challenges. A&K 
provides more than just materials. We offer real "Rail Solutions". A&K problem 
solvers understand that we need to do more than offer quality new and relay 
track materials, welding services, panelizing, accessories and track 
removal. We must also give the best service . 

A&K does it all. A&K has it all. 

Even if your rail problems seem 
as big as the universe ... A&K's "Track Stars" 
will find an on-time solution. 

Call an A&K problem solver today. 

© 1993 A&K Railroad Materials Inc. 



A&K Railroad Materials, Inc. 

Corporate Headquarters 
1 505 South Redwood Road 
P.O. Box 30076 
Salt Lake City, Utah 84130 
Phone: (801) 974-5484 or 
Toll Free: (800)453-8812 
FAX: (801) 972-2041 

RAIL SOLUTIONS On-Time!. ..On-Track! 



Motrv nctiiiud 



(^ 



364 



RAIL TRAVEL (CREEP) CAUSED BY 
MOVING WHEEL LOADS^ 

By Arnold D. Kerr* and Alexander Babinski** 

Summary 

Shortly after railways were introduced, it was noticed that when a train moves at constant speed 
the rails have a tendency to permanently displace in the direction of the moving train. This phenom- 
enon, is referred to in the railway literature as rail travel or rail creep. When the continuously welded 
rails (CWRs) of a track are prevented from moving axially at a specific track location, like at a road 
crossing or at a bridge abutment, rail creep may cause the accumulation of compression forces at 
these locations, that will contribute to track buckling. Therefore, for over a century, attempts have 
been made by various railway engineers to clarify this phenomenon and to devise analytical methods 
for its prediction. But, in spite of this effort, to date there is no generally accepted analysis, nor a con- 
ceptual mechanism, for explaining rail creep. 

One aim of the present paper is to briefly describe the early encounter with the rail creep phe- 
nomenon and the development of anti-creep devices, as a practical measure of coping with this prob- 
lem. Then, a number of hypotheses are discussed that attempted to explain the basic mechanism of the 
rail creep phenomenon. Two of these approaches — the Johnson hypothesis that is based on the plastic 
axial response at the interface of rail and base, and the vertical viscoelastic response of the rail sup- 
port — for determining the creep driving force are analyzed in more detail, since they contain elements 
that may enter a general theory of rail creep. A conceptual shortcoming that is contained in the analy- 
ses that are based on the Johnson hypothesis, is pointed out. It is hoped that this paper will familiarize 
the railway engineer with the rail creep phenomenon and the various attempts to explain it, and at the 
same time will form a basis for the future development of a general theory of rail travel (creep). 

Introduction and Statement of Problem 

Since the early evolution of the railways, it was observed that the rails have a tendency to move 
axially, as well as deflect vertically, when subjected to moving trains. The obvious causes of the per- 
manent axial rail displacements are: ( I ) The friction forces generated between the wheels and rails 
during acceleration and braking of trains; especially noticeable at stations, and (2) The temperature 
changes in the rails; especially in the vicinity of expansion joints. These causes are simple conceptu- 
ally and hence are easy to analyze. 

Not so easily explainable are the permanent axial displacements of rails that are caused by trains 
that move at constant speed. This phenomenon of rail travel is referred to as "rail creep". For example, 
for jointed tracks, according to Camp (1903) ". . . measurements have been taken of the distance the 
track crept under a moving train and these show that a movement occurred in the track of from 2 to 37 
ins. depending on the temperature, weight of engine and train, and softness of bottom . . .". For jointed 
tracks, the initiation of creep generally causes the rail sections between joints to move in the direction 
of the moving train, tending to reduce the gaps at the front ends and increase them at the rear ends. Since 
each rail section in a jointed track undergoes similar movements, the entire track moves forward. 

With the introduction of continuously welded rails (CWR) after Worid War II, the tendency of 
rails to creep may lead to an accumulation of axial compression forces in the vicinity of locations 
where the rails are constrained axially. Like, for example, at road or rail crossings, at bridge abut- 
ments, and at turnouts. These compression forces may contribute to track instability, since in these 
cases a smaller temperature increase is sufficient to cause buckling. Therefore, when considering the 



' Research supported in part (A.B.) by a Fellowship in Railway Engineering of the A.s.sociation of American Railroads 
* Professor, Department of Civil Engineering, University of Delaware, Newark, DE 197 16 USA 
"Graduate Student. Department of Civil Engineering. University of Delaware. Newark, DE 19716 USA 

365 



366 



Bulletin 761 — American Railway Engineering Association 



possibility of thermal track buckling at such locations, forces caused by creep have to be included. 
Another possible consequence of rail creep are displaced and skewed cross-ties, that cause gauge 
tightening. In recent decades, the negative effects of rail creep have been intensifying due to the ever 
increasing wheel loads, train speeds, and the density of traffic. 

Since the middle of the 1800's, to reduce rail creep various devices were invented and used on 
European and North American railways. They were originally referred to as "anti-creepers". In more 
recent decades they are called "anchors". 

An early measure, was to provide the rail base with holes through which cut spikes were driven 
into the wood ties, as shown in Fig. 1. Since these rectangular holes weakened the base and caused 



iti^tH^lamtmOlM 



Leipzig-Dresden (1836) 

Fig. 1 Early Measure for Preventing Rail Travel 
(Haarmann 1902, p. 24) 

the initiation of cracks, this method was discontinued. Later, in another approach, a separate angle 
bar was attached to the rail at a joint and locked to the adjoining ties, as shown in Fig. 2. 




' ^ ■y ^- L--^-'^^- 53 





Osnabriick-Brackwede (1886) 



Fig. 2 Anti-creep Device of the 

Osnabriick-Brackwede RR, 1886. 

(Haarmann, 1902, pp. 54-55) 



During the following decades various anti-creep devices were developed in Europe and North 
America. Numerous examples of early developments on the Austrian railroads are described by Wirth 
(1909b). For many other anti-creeper inventions, refer to the entries and descriptions at the U.S. Patent 
Office in Washington, D.C. Anchors currently used in North America are shown in Fig. 3. 




Fig. 3 Examples of Anchors Currently Used in North America. 

Note, however, that when anchors produce a strong axial connection between rails and ties, pre- 
venting the rails from sliding over the ties, then the entire rail-tie structure and the crib-ballast tend 
to displace axially. 



Paper by Arnold D. Kerr and Alexander Babinski 



367 



Proposed Concepts for Explaining Rail Travel (Creep) 

Because of the adverse effect rail travel has been having on track, since the late 1 800's various 
attempts were made to explain how rail creep is generated by trains that move at constant speed, and 
to provide analyses for estimating their effect. During the following decades a number of hypotheses 
were postulated. 

As an example, according to Engerth (1900, p. 8), Kriiger in 1 886 claimed that because of "the 
coning of (wheel) tires and the inclination of the rail heads (caused by canting), a resistance to move- 
ment is produced, which has the effect of giving the rails a tendency to movement in the direction of 
running of the trains". Engerth analyzed this hypothesis (pp. 8-10) and found, that contrary to the 
claim by Kriiger, the generated axial forces are relatively small, "and therefore devoid of any impor- 
tance" for rail creep considerations. The paper by Engerth also contains a description of other 
hypotheses, like the effect of axial forces generated by the interaction of the wheel flanges and the 
rails on rail creep. Also this one was found by Engerth to be of no importance for explaining the 
observed rail creep (1900, p. 21). 

In the following, three other hypotheses which attracted the attention of a number of 
researchers, are described in more detail. They are: (I) Rail travel caused by wheel impact at joints, 
(II) The Johnson creep hypothesis based on the moving bent rail, and (III) Rail travel caused by the 
axial wheel-force component, when vertical rail support is viscoelastic. 

(I) Rail Travel (Creep) Caused by Wheel Impact at Joints 

In a well maintained track, away from joints, rails deflect (nearly) uniformly under the wheels 
of a moving train. However, since the bending stiffness of two joint bars is generally much lower than 
the bending stiffness of the connected rails (currently less than 50%), the rail deflections increase as 
the wheel approaches the joint region. This is shown in Fig. 4{a). For a poorly maintained joint with 
loosely attached joint bars, the rails may additionally be offset vertically, as shown in Fig. Mb). 





(a) Well maintained joint (b) Poorly maintained joint 

Fig. 4. Effect of Moving Wheel Load at Joint. 

The concept of rail travel caused by wheel impact at joints is based on the notion that when the 
wheel approaches a joint it sinks and then strikes the adjoining rail with a force F. This impact force F 
may be resolved into two components: F, and F^^. The vertical component F^, causes elastic and non-elas- 
tic deformations in the rail ends, ballast, and subgrade. The horizontal component F^ "pushes" the rail 
forward, contributing to rail creep. Intuition suggests that the wheel force F, and hence F,^, increases with 
increasing static wheel load P, increa.sed train speed, and deteriorating condition of the rail joint. 

Rail travel caused by wheel impact at joints where suggested by Koyuar in 1887, assummg the 
situation in Fig. 4{a), and by Wasiutynski ( 1 896), assuming the rail deformation shown \n Fig. 4(b). 
This approach was described by Engerth (1900, p. 22). The effect of wheel impact at joints was also 
di.scus.sed by Wirth (1909a, p. 319), who concluded that the horizontal component of the wheel 
impact force (F^ in Fig. 4b), is the main cause of rail creep. Note also the closely related presentation 



368 



Bulletin 761 — American Railway Engineering Association 



by Skibinski (1913) who was a proponent of a joint impact hypothesis. A similar approach based on 
Fig. 4{b) was presented by Kyuner (1925), and based on Fig. 4(«) by Sailer (1928). Reservations 
against the joint creep driving concept by Wirth (1909a), were expressed by Weikard (1909). The 
effect of joints on creep was also discussed by Frishman (1942). 

When considering the effect of F^, it should be noted that a moving train has many wheel sets and 
a jointed track contains many joints; thus, a train that moves over a jointed track will cause many impacts. 
Therefore, even a relatively small creep driving force at each joint, may be sufficient to produce perma- 
nent axial displacements (creep) of the short rail sections; especially in this vibrating environment. 

However, in more recent decades, rail travel (creep) was also observed in continuously welded 
rails, that do not contain joints. Thus, another mechanism must be at work for generating rail creep. 
This will be discussed next. 

(11.) The Johnson Creep Hypothesis Based on the Moving Bent Rail 

In 1887 J. B. Johnson postulated that rail creep is caused by bending of the rail when subjected 
to a moving wheel load, and the resulting frictional slip at the interface of the rail and its supporting 
base [as reported by Z. (1888)]. 

The first reported attempt to prove the validity of this hypothesis is due to Zimmermann, who 
based his analysis on a model consisting of a finite elastic beam simply supported at both ends, and 
subjected to a moving load. For related analytical details refer to Engerth ( 1 900, Sec. I). 

Another attempt was made by den Tex (1910, 1913), who represented the rail-in-track by the 
more realistic model of a long elastic beam attached to a Winkler base (closely spaced springs). The 
corresponding analytical model is shown in Fig. 5(a). Note, that in this model the axial resistance is 
assumed to be of sliding friction type (Coulomb friction). 

The analyses of den Tex were greatly extended and generalized by Albrekht (1958a, 1958b). 
The corresponding track model is shown in Fig. 5{b). It includes also the horizontal elastic response 
of the tie-ballast support. 

The analyses by Albrekht were generalized further by Kogan (1967), Menshikova (1972), 
Kogan (1981), and Verigo and Kogan (1986). They based their analyses, essentially, on the rail-in- 
track models shown in Fig. 5(fl), and the use of the differential equations for the rail to determine the 
permanent rail displacements (creep). For a comparative study of the analyses by Albrekht, Kogan, 
and Menshikova from a unified point of view refer to Babinski and Kerr (1995). 

In the following, the phenomenon of rail travel (creep) caused by a wheel that moves at a con- 
stant speed, in accordance with the Johnson hypothesis, is presented. It is based on the simple track 



EI. 



X 



i 



3SX5I53SSI 



(a) A Simple Rail-in-Track Model. 



EI. 



\ 



pVS^ 



'i ■j'T fj t jj^ i y^^. 



(b ) A Generalized Rail-in-Track Model. 
Fig. 5 Models for Rail Creep Analyses According to the Johnson Hypothesis. 



Paper by Arnold D. Kerr and Alexander Babinski 



369 



model shown in Fig. 5(a) and it utilizes some of the analytical features presented by den Tex, 
Albrekht, Kogan and Menshikova. The purpose of this section is to explain the rail creep mechanism, 
to present the essential features of the analysis for continuously welded rails (CWR), and to point out 
shortcomings of these analyses. 

The general differential equations for a continuously supported elastic beam (rail) subjected to 
a di.stributed vertical load q, an axial load n, and an axial tension forces Si, are: 



EI 



dx' 



i^f 



9N 

ax 



H-n-r = mw 



dx 



3'w 



+ p = q 



(1) 



where 



N = EA 



9x^ll aT 



, M = -EI- 



ax^ 



(2) 



w(x,t) = w(x,o,t) and u(x,t) = u(x,o,t) are the vertical and axial displacements of point x on the refer- 
ence axis at time t, EI is the constant bending stiffness of the rail in the vertical plane, A is the rail 
cross-sectional area, m,^ is the mass of rail per unit length, m^ is the mass of the rail and the added 
apparent vertical mass of the ballast and subgrade, r(x,t) is the axial resistance force along the inter- 
face of rail and its support, and jl = -rh^^ is the distributed bending moment along the rail caused by 
r at the rail base, h^^ is the distance between the rail bottom and the reference axis. The used sign con- 
vention is shown in Fig. 6. For the derivation of these equations, and the range of their validity, refer 
to Marguerre (1938) and Kerr and El-Aini (1978). 

For the problem under consideration, n s 0. Next, it is assumed that the effect of the axial forces 
^ is negligible onw by dropping the second term in the first equation in (1), that dyi/dx is small of 
higher order, and that the pressure that the rail exerts on its support is 

p(x,t) = kw (x,t) + q,. , (3) 

where k is the track modulus and q^ is the own weight of rail per unit length. Eq. (3) represents the 
response of the Winkler base. Since the wheel load P is moving with a constant speed v^, that is 
smaller than the critical speed v^^., it may be shown that a steady state will exist. For a justification 
refer to Kerr (1981). Thus, to an observer who moves with P, the deflections of the rail will appear 
static. This suggests the possibility of transforming the partial differential equations in (1) into ordi- 
nary differential equations in the moving reference frame 



u(x+Ax.z,t) 





V(X+AX,1) 



Fig. 6 Positive Sign Convention Utilized in Eq.'s (1) and (2). 



370 



Bulletin 761 — American Railway Engineering Association 



^ = X - \j^t ; C = z . (4) 

as shown in Fig. 7, noting that 3()/9x = d()/d^ and 3()/3t = -v^^d()/dt,. With this transformation, and 
the made simplifying assumptions, the first equation in (1) reduces to 



^.dw -)dw.. „c-K 

dq d^- 



(5) 



where 5^, is the Dirac delta function. 

The rail creep researchers, den Tex, Albrekht, Kogan, and Menshikova, worked with the expression 



^(^) = |^e"Pl^l(cosP|^| + sinP|^|) 



<^< 



(6) 



where (3 = \ ky(4El). This implies that they tacitly neglected the velocity term in (5). This simplify- 
ing assumption is not necessary. In the following analysis this velocity term is retained. The solution 
of (5), when P acts at ^ = 0, is [Kerr (1981) p. 410] 



w(£) = — t— e 
^ 2kA(o 



' ((0 cosco|^| -I- X sinco|^|) 



.<^< 



where 



(7) 



(7') 



The corresponding w (^)-profiles are plotted in Fig. 7 for the UIC 60 rail with I - 3,055 cm'* (73.4 
in^), E = 2.07 x 10^ N/mm^ (30 x 10^ Ib/in^), k =20.7 N/mm^ (3,000 Ib/in^), m^ = m,.^;, = 60.3 kg/m, m^ 
- ' ^ "^raii ^^" estimate of the apparent vertical mass of the tracks), and P = 111 kN (25,000 lb), for two 
speeds v^. Note the small effect speed v^ has on the moving deflection profiles in the range <v^^< 
200 km/hr. 

It should be noted, however, that in a CWR track the rails may be subjected to tension or com- 
pression forces, due to temperature changes. For a discussion of the effect of these forces on the rail 
response, especially on the critical speed v^^, refer to Kerr (1972). 

The expressions for w (^) in eqs. (6) and (7) imply that a rail is continuously attached to the 
base; thus, that the vertical pressure may be positive or negative. However, in actual tracks with loose 
cut spikes, the own weight of a rail is generally not sufficient to prevent its separation from the base. 
Therefore, Albrekht (1958a) made the simplifying assumption that when the rail deflections are neg- 
ative according to eq. 6, the rail lifts off the base and the vertical contact pressure in this region is 
zero.* Albrekht also assumed that the vertical deflections and contact pressures beyond these lift-off 
regions are negligible. These assumptions are retained. From Fig. 7 it follows that for the speed range 
shown, "separation" starts at ^ = ± 37t/4p. 



-3JI/4P 



3n/4P 



V.=0 




v„-200lm/h 



Fig. 7 Rail Profiles w (^) Caused by a Wheel Load P Moving with Speed v^ 



*Thi,s a.ssumplion violates venical equilibrium. The rigorous solution when the rail rests on the Winkler base, with the possibil- 
ity of lift-olT, is more involved. It was presented by Kcit and Bassler (1982). 



Paper by Arnold D. Kerr and Alexander Babinski 



371 



According to Coulomb's law and eq. (3), the axial resistance at the interface of rail and its tie- 
ballast support is stated as follows: 



r(^) = HP(^)^ 



H[kw(^) + q,. 




when 



w > 
w<0 



(8) 



where \i is the friction coefficient. 

Next, we determine the axial movements of a point on the rail base due to rail bending, caused 
by a moving wheel load P; a major "player" in the Johnson hypothesis. The deformed rail is shown 
in Fig. 7. For the study of the axial rail base displacements consider the section of rail shown in Fig. 
8. The cross-section AB at t,, that was initially vertical, rotates due to bending clockwise by an angle 
G, in accordance with the positive sign convention. Since the anticipated 9 is very small, it may be 
expressed as 



e.4.4. 



(9) 



Because of this rotation, points of the rail-base experience displacements in the ^ direction. 
According to Fig. 8, the total axial displacement of the rail-base at I, is 

u^(^) = u{^) + Ue(^) . (10) 

Denoting by h the vertical distance between rail reference axis and the bottom of rail, Ug can 
be expressed as 



dw 

UgC^) = -h„ sine(^) = -h„0(^) = -h, — . 



(11) 



The "minus" sign indicates that for a positive 0, the displacement Ug is in the negative direction, as 
shown in Fig. 8. Noting the w -expression in eq. (6), and performing the differentiation in ( 1 1 ), yields 



U9(^) = -hfl-^-h„ 



dw_, PP^a^-HCO^)„_,,,^.X|^|^._|,|^ (11.) 



IkXat 



-5^n(0[e"^l'^' sinco|^|]. 



The 5^/7(^)-function is shown in Fig. 9. 




Fig. 8 Displacement of a Point on the Rail Base. 



372 



Bulletin 761 — American Railway Engineering Association 



■ 


.'Sn(i) 


'1 


::_^____ 


^:_ 


1' 5 







Fig. 9 The sg«(^)-Function 



It may be shown [Albrekht (1958 b), pp. 9-14] that, because the vertical rail deflections are 
very small, the u of the bent rail caused by a passing wheel load P is negligible compared to Ug. 

Thus, 



u, =u + Uo =Uft =h. 






sgn(^)[e' ' ' sinco 



m 



(12) 



Next, the expression for the axial velocities of points on the rail base, \)/^), is determined by 
differentiating the above equation with respect to time. The result is 



^^^^^ = ir = ir = -ari ''^J--^ft'''^U^t• 



^w'j^^ 



Since ^ = x - vjt, and hence 9^9t = -v^, it follows that 

a'w(^) 



V, =X)oh« 



d^' 



Substituting w(^) from eq. (7) into the above expression results in 






' (00 coscol^l - X sina)|^|). 



(13) 



(14) 



(15) 



The graphs of w(^), u^(^), v^{t,), and sign[v^] are presented in Fig. 10. Note that the upper curve, 
w vs. 4, is the moving rail deflection profile, as shown in Fig. 7. The 5/^n-function shows the direc- 
tion of movement of points on the rail base. It is defined as 



1-1 when \)r < 
" X)r =0. 
+1 " ■0, >0 



(16) 



Note the difference between the sign and the sgn function shown in Fig. 9. 

According to Fig. 10, points in the immediate vicinity of P, zone [2], have negative velocities 
(hence they displace backward), whereas points in front and behind this region, zones [1] and [3], 
have positive velocities (hence they displace forward). Therefore, as P moves along the rail, each 
point on the rail base is undergoing forward-backward axial movements. 

Having established the kinematics of rail deformation due to the moving load P, next we deter- 
mine the "creep driving force" according to the Johnson hypothesis. 

Because the positive direction of the resistance force r(^) that acts on the rail base is opposite 
to the positive velocity of the points on the rail base it follows, noting eq. (8), that 



r(^) = 



^[kw(^)-l-qr]5i^n[\)r(^)] when w(^) > 
" w(^) < 



(17) 



Paper by Arnold D. Kerr and Alexander Babinski 



373 



Therefore the axial resistance force that acts on the rail base is 



r(q) = n<j^^e •"'■"(cocosco|q + Xsina)|q) + qr}-.v/>n[\Jr] 



(18) 



where v^~ e ^'^' ((ocoscol^l - Xsincol^l) . 

Assuming the effect of q^. to be negligible compared to kw, the graph of this force distribution, that 
acts on the rail bottom, is shown in Fig. 10 (V). The area 03~, represents the resultant of the distributed 
forces r(^) that resist the movement to the left. It is larger than the areas ((0| + to,) that represent the 
resultant of the forces r(^) that resist the movement to the right. Summing these forces, by integrating 
r(^) in (18) over the range - 371/4(5 < ^ < 37t/4p*, the resulting resistance force is obtained as 

3n/4p U/Af, 37t/4p 

Rcr^.p= I r(^)d^ = 2 j n[kwK?«[tjJd^+ j ^[kw].v,>![u,]d^ 

-^j:/4P I o it/4P 



= -HP<^l + e 



,-37tX/4p 



-7:A./4P 



( KiO^ ((i)~ -X') . f 71(0 

-2 COS + sm 

V 4P J Xco 1^ 4(5 



/'37rco'l (co"-X") . ( 3n(ii 

COS sm 

I, 4p ) 2Xco [ 4(i 



(19) 



H- (-) 



Vertical Rail 

A 

Displacement, w. 



Axial Displacement of 
Rail Base, ur . 



Velocity Distribution of 




Distribution of Resistance 
Forces Acting at Rail Base 
Caused by Rail Base Slippage 



Fig. 10 Generated Displacements, Rail Base Velocities, and Resistance at Rail Base. 



'According to Albrekht this is the dominant range. Note comments above eq. (8). 



374 



Bulletin 761 — American Railway Engineering Association 



where X, (o, (3, and a are given in (7'). This is the force that is caused by the slippage between the rail 
bottom and the tie-ballast base. 

den Tex (1913, p. 372) referred to this force, as the force that drives the rail in the direction of 
the moving wheel ("Die Kraft die die Schiene in derselben Richtung treiben will, . . ."). This notion, 
that -R^-reo = f^creep '^ ^^^ "creep driving force", was also adapted by Albrekht (1958a, p. 19), Verigo 
and Kogan (1986, p. 402) and Menshikova (1972, p. 121). The minus sign in front of R^ccp '^ nec- 
essary because, according to the used sign convention, r(^) are positive when they act in the opposite 
direction to ^. 

The dependence of F^,^^^ on the wheel velocity v^ is shown in Fig. 1 1 . Note, that for a fixed 
wheel velocity x> , F, ,. increases linearly with increasing friction coefficient ^l and with increasing 
wheel load R 

Next, the permanent axial displacements caused by the moving "creep driving force", Fj.^^.^p, are 
determined using the second differential equation in (1). 

Before solving this equation, it is instructive to study the anticipated axial rail response caused 
by F. , , by considering the corresponding discrete rail model shown in Fig. 12, for the case of a 
short or long rail that is free to move axially at both ends. To simplify the demonstration, it is assumed 
that r^, is constant; thus, independent of the wheel load P. 

When F arrives at ® , the element (D slips in the direct of F by u^, compressing the elastic 
spring between ® and @. When F moves on to @, element @ slips by the displacement u^^, com- 
pressing the spring ahead of it and relieving the spring behind it. A similar situation occurs when F 
moves to ® , then to ^ , and finally to ^ . Note, that the localized compression zone moves ahead 
of F. When F leaves ® the rail-model is stressless, but is has moved axially by a permanent uniform 
displacement u^,. This is the creep displacement caused by the pass of one wheel. The pass by the fol- 
lowing wheel load will produce another uniform displacement u^,, etc. 

Next, we demonstrate that when the axial rail displacements are blocked at a fixed track loca- 
tion, (like at a road or rail crossing), axial compression forces will accumulate in the rails in the vicin- 
ity of this location. These forces in turn, may contribute to track buckling and to other undesirable 
lateral track deformations at these locations. 

For this purpose we modify the axial rail-model shown in Fig. 12 by preventing the axial move- 
ment of element © . The resulting rail-model and the various stages for the moving load F, are shown 
in Fig. 13. 

on 




Fig. 11 Dependence of F^^^ on u^, for UIC 60 rail and k = 20 N/mm^ (3,000 Ib/in^). 



Paper by Arnold D. Kerr and Alexander Babinski 



375 



FannamQ).AjiaU} 
Daplod^KaU 



C»mzpcmtlu>i 
AamlFtnt 



© ® ® © © 



■^ 



I I I 



I I I I 



F antra mQ).Ajially 
DiiplaciJ Kait 




F arrirtj m Q). AxiaUy 
DtiplaadKae 



Comspandmg 
Anal Font 




//}//7///7/y7///}/ 



F arrntt al ®. AxieOf 
Displactd Kaa 



Corrtspoadmg 
Axial Fan* 




////////////////// 



Farrtns at © AilaUy 
DiiplactJ Kaa 



Comtpamdmt 
AiialFanr 



w 



\w 



t^ 



////////////////// 



Fig. 12 Simple Model for Demonstrating the Generation of Rail Creep by the Moving Force F. 

From Fig. 13(fl) it follows that during the pass of the first wheel, the model response is the same 
as described previously, except that when F reaches ^4) then Co) and leaves the model, a residual 
compression force remains in the last spring. Fig. 13(/7) shows the situation during the passage of the 
next wheel. This pass adds to the residual compression force between w and ^ , as indicated. The 
largest force this spring can accumulate will depend on F, on the number of passes and \>^, on the 
axial stiffness of the rail-model, and on the sliding frictional resistance along the interface of rail and 
base. Once the largest spring force is reached, additional wheel passes will start the accumulation of 
compression forces in the neighboring spring (between ^ and W ), etc. Thus, the region of gener- 
ated rail compression forces increases with the increasing number of passing wheels. 

The purpose of the above discussion has been to show, on a simple model, the mechanism for 
generating rail travel (creep) and, in the case of an axial constraint, the accumulation of rail com- 
pression forces by the passing wheels of a moving train. Next, it is shown how results of this type 
may be obtained analytically, by solving the second differential equation given in (1), 



aw . - a-Q 

— — i-n - r = m — —. 
ox 9t' 



(20) 



Generally, N is the nonlinear expression given in (2), n = 0, and r (^) is given in (17), noting that the 
r(^) in (17) acts at the rail base. The resuhing differential equation is non-linear and its solution 
requires a numerical method which is beyond the scope of this expositional paper 



376 



Bulletin 761 — American Railway Engineering Association 



a) Pass of the First Wheel Load 



<D 



® 



® 



® 



Farrlvtsat (T).AjUiUy 
Displaced KaU 



Conapatdlnt 
Axial Fone 



/ 7> / y> / /v> / ^7/ / / 



I ! I I I I 

(El 



FaiTira at U). Aiiatty 
DlMplactd Rail 



CorrespemJIni 
Axial Fore* 






////////////// /y 



pq 



I I 
I I 



Dltttattdnaa 



Corrtspoa^i 
Axial Font 




Farrtveiat (j). Axially I 

CHipUKetl RaU | 




//////////////// 



Correxpambng 
Axial Force 



b) Pass of the Second Wheel Load 



FarriMtiat Q).AxiaUy 
Disptactd Kail 



'iWS^^^^^ 



CorrespondiHt 
Axial Forc€ 



Z7 



'^ 

m 



F arrives ai Q).Ailalty I 
Displaced Kail I 

Z7 






Corresponding 
Axial Force 



•2H.' i:2ii. 



:;2«. 



:2«.: 



Fig. 13 Simple Model for Demonstrating the Accumulation of Rail Compression 
Forces Caused by the Moving Force V^ree^- 

For the purpose of demonstrating the rail creep solution and its characteristic features, eq. (20) 
is simplified by linearizing I^I and by replacing the creep driving force shown in Fig. 10 with the con- 
centrated resultant F, given in eq. (19). It is also assumed that the interface resistance r = r^ = con- 
stant. This simplifying assumption is justified since, at each instant, the region of the generated r is 
very small (Fig. 10) and r is to be considered as an average value. The corresponding analytical 
model is shown in Fig. 14. 



With these simplifying assumptions, differential equation (20) reduces to 



dK 



a^G 



EA— ^-r^ .y/g«[u(x,t)] = m— ^. 
dx ot 



(21) 



Paper by Arnold D. Kerr and Alexander Babinski 



377 



m^ 




^(V 



Yi ^"^.\ ^i "n ^ 



t; 



Fig. 14 Simplified Model for the Analysis of Rail Travel (Creep). 



Next, this equation is transformed using the moving coordinates system given in eq. (4). 
The result is 



i2.- 



EA— Y-r„5(j?n 



-\) 



9u 

"3^ 



m\)^ 



d-G 



or rewntten 



. xm\ L. d^u 



= 0. 



For the tracks currently in use on main lines mt)y(EA) s 10'^ « 1. Thus eq. (23) reduces to 

i2.-. 



EA— 2- + r„,9/A'n 
d^ 



= 0. 



Noting that EA(du /d^) = I^(£,), equation (24) may be written as 

dN 

-— + r„5/^n[N] = 0. 
dq 



(22) 



(23) 



(24) 



(25) 



Therefore, the analytical formulation for the rail problem shown in Fig. 14 consists of the two 
differential equations 



(26) 



dq 


-co < ^ < 


dq 


<^ <cx. 


ling conditions 




lim N/ -^ 


finite 


^^-^ 




lim Nr -> 


finite 



Nr(0)-N/(0) + F = 0. 



(27) 
(28) 
(29) 



This is a boundary value problem. By inspection, it follows from the first differential equation in (26) 
that 

(30) 



N,(^) = 0, 



< ^ < 



since sign[0] = 0. When considering the second differential equation in (26) it is expected that where 
54|.(^) is 9i it will be a compression force. Thus, in this region .9j,gn[N|.] = -1 and the corresponding 
equation in (26) reduces to 



378 



Bulletin 761 — American Railway Engineering Association 






-r =0 



with the general solution 



^,(^) = r„^ + C 



(31) 



(32) 

The integration constant C is determined from matching condition (29). Since 5l,(0) = 0, it follows 
that 

Ivi^(O) = -F and C = -F . (33) 

Thus, 

l^r(^) = r^^ - F < ^ < / (34) 

where / - F/r^^. For the domain / < ^ < oo, the second differential equation in (26) is satisfied for 

^S^) = 0, / < ^ < - (35) 

noting that sign[0] - 0. 

Thus, the solution to the boundary value problem stated in (26) to (29) is 

f in -oo < ^ < 

N(^)= -(F-r„^) " < ^ < / (36) 

[ " / < ^ < oo 

where / = Fk^. The graphical presentation of this solution is shown in Fig. 15(1). The corresponding 
resistances between rail and base is shown as (II). 

The associated residual rail displacements (creep) caused by the moving wheel load P (i.e. by 
F), are obtained by integrating the expression forlvl(^) = EA du/d^ over the domain / < ^ < oo. 



<l) 



(in 



pp^ 






=const 4 



(111) 



J." 



l=F/ro ^ 



Fig. 15 Solution of Formulation (26) to (29). 



Paper by Arnold D. Kerr and Alexander Babinski 



379 



According to eq. (36), N(^) = in / < ^ < oo. Also, since u (°°) = 0, it follows that 

G(^) = / i ^ < - . 

For the domain < ^ < / 

[du=-^jN(^)d^ = -^j(F-r,^)d^. 



(37) 



0(/)-G(5) 
According to eq. (37), ii(/) = and therefore 



u{^)-^j (F-r,^)d^ = 



EA 



2EAr„ 



For the domain - oo < ^ < 






0<^ < /. 



(38) 



According to (38), u(0) = F-/(2EAr^,). Noting that in -oo < ^ < the axial force N(^) = 0, it follows 
that 



G(^) = - 



2EAr, 
Thus, the obtained creep displacement expressions are 

F^ 



-co < ^ < 0. 



(39) 



u(^)^ 



2EAr„ 

(F-r,^)2 



2EAr^, 




-oo < ^ < 
< ^ < / 

/ < ^ < - 



(40) 



These creep displacements are shown in Fig. 15 (III). 

Note, that the analytical results presented in Fig. 15 are similar in nature to the results of the 
simplified discrete rail model shown in Fig. 12. Namely, that the moving wheel is generating m front 
of it a small compression region and leaving behind it a residually displaced rail that is free of axial 
forces. These resuhs, although based on a simplified analytical formulation, demonstrate the analyt- 
ical approach for determining rail travel (creep) and its characteristic features, based on the Johnson 
hypothesis. As an example, according to Bochenkov (1962, p. 192) field observations revealed that 
a moving loaded train with wheel loads that were about 4 times larger than those of a lightly loaded 
train, produced permanent axial rail displacements (creep) that were 15-20 times larger. This agrees 
with eq. (40), since u (0) is proportional to F-^, and F is proportional to the wheel load P. Therefore, 
the creep displacement u^, is proportional to P- and when P = 4Pj, the resulting u^, is proportional to 
(4P^J- = 16 P^; which agrees with Bochenkov's field ob.servations. 

As reported by Weikard (1909), in double track territory it was noted that the outer rails were 
creeping more than the inner rails, apparently because the track was "softer" under the outer rails. 
This implies that the smaller the track modulus the larger will be the creep displacements. To check 
the validity of this conclusion, the u -expression given in eq. (40) was evaluated at ^ = 0, for a range 
of k-values, a wheel load of P = 1 1 1.2 kN (25,000 lb), and velocities \j^, = 100 and 200 km/h. The 
results are shown in Fig. 16. They are in agreement with the statement by Weikard. 



380 



Bulletin 761 — American Railway Engineering Association 



mm 

0.0018 



>■ 0.0014 

I 

S 0.0012 



m 

0.00007 

















— 


— I00kmfl\ 




urcs4 ': 




^^ ^^~^^-~ 


-~-,._W^«> 


r-r^-^T^-^ 




UK 54 


• ■■ ■■■ ;•• 


UIC60 


■ : ;• 



1.000 2.000 3,000 4.000 



5.000 6.000 IbAn' 



10 15 20 25 30 35 40 N/ntm' 

TRACK MODULUS, k 

Fig. 16 Dependence of Rail Creep on the Track Modulus for One Wheel Pass 

Note that with increasing vertical track stiffness and rail flexural stiffness, rail creep decreases. 
But, increased velocities increase rail travel (creep). 

For more extensive analyses refer to Kogan (1967), Menshikova (1973), Verigo and Kogan 
(1986), and Babinski (1995). However, it should be noted that these presentations are based on the 
moving rail deflection profile shown in Fig. 7, for v^ = 0. Also, the profiles for < \)_, < t) ., presented 
above, are symmetrical with respect to the moving load P and therefore exhibit a horizontal tangent 
at P. This, in turn, implies that as P moves with a constant speed v^, it does not produce work. But, 
as shown above, work is being dissipated continuously by the friction mechanism at the rail support. 
Thus, the above analyses violate the principle of conservation of energy. To avoid this situation, the 
solution of the first equation in ( 1), w (^), should be such that P does not act at a point of symmetry 
with respect to P. One such approach is utilized in the following section. 

(ni) Rail Travel (Creep) Caused by the Axial Component of the Moving Wheel Force 

In this approach for analyzing rail creep, a wheel load that moves at constant speed d^ < v^^, 
generates, in addition to the vertical load, also a horizontal force component. This requires a non- 
symmetrical rail deflection profile with respect to P. It may be achieved by including in the first equa- 
don in (1) a dissipation mechanism in the rail support. 

An analytically simple base model that will produce a result of this type, may be obtained by 
including a base damping force that is proportional to the vertical rail velocity (viscous damping). 
Then, the first equation in ( 1 ) becomes 



^, 3 w 3 w 3w , . 
EI — ^ + m — ^ + r\ + kw = q 

dt^ at 



9x^ 



(41) 



where r\ is the damping coefficient and m = m^. 

This differential equation for an infinite beam subjected to a constant vertical load P, moving at 
constant speed v^, was solved by Dorr (1948), Rzhanitsyn (1949, 1968), Kenney (1954), 
Shakhunyantz (1959), Achenbach and Sun (1965), Steele (1967), and others. For a discussion of 
these and of closely related solutions refer to Kerr (1981). 



Paper by Arnold D. Kerr and Alexander Babinski 



381 



The solution method is often simplified by utilizing the fact that after a time a steady state 
exists. Then, using the transformation variable ^ = x - vji, eq. (41) reduces to 



EI 



d^ 



d^w 



dw 



+ m\)o —-Y - r|u„ -— + kw = P6(^) 



d^ 



d^ 



(42) 



a linear ordinary differential equation with constant coefficients. Its solution may be obtained in the 
form w = Ae'^'', as done by Kenney (1954) and Shakhunyantz (1959), or by using the Fourier trans- 
form method, as done by Dorr (1948) and Steele (1967). 



According to Shakhunyantz (1959, Part I, Section 4) the solution may be presented as 

W/(^) = eP^A, cosK,^ + Aj sinK,^) -«> < ^ < ] 

w^(^) = e"^'^ (A3 cosK,^ + A4 sinK,^) < ^ < 0° j 

where p is determined from the cubic equation for p^ 



/ 2 >2 

2p^+i^ 
2E1 



ri'-u- 



16(EI)-p- 



EI 



and 



= Jp-+ ^ + -J-^ 

K2J V 2EI 4EIp 



(43) 



(44) 



(45) 



The integration constants are obtained from the matching conditions at P(^ = 0) and the regularity 
conditions at ± °°. These constants are: 



A, =A,-w(0) 

A, 



4p^ -K? +Kt , „ 

^ ' ' w(0) 



4pK, 
— ! ^w(O) 

4pK2 



(46) 



where 



w(0) = 



H, = 



El4p(H2-Hf)' 



' Sp^EI 



H,=J— + 



2 2 

T1 ^0 



EI 16p'(EI)- 



(47) 



The solution in (43) to (47) was evaluated for a UIC 50 rail, k = 14 N/mm^ (2,000 Ib/in^), u^ 
= 250 km/h (155 mph), and \\ = 0.7 ri„ = 0.7x2 Vlcm [Kenney (1954), eq. 11] where m = 10 m,.^^, = 
10x54.4 kg/m. The result is shown in Fig. 17. 

Note, that the deflection curve is not symmetric with respect to P, as anticipated, and that the 
largest deflection occurs behind P at ^ = - /, where 



/ = — arctg— * 

K| pS + K, 



and 



4p'^+K2 

4pK, 



(48) 



(48') 



38: 



Bulletin 761 — American Railway Engineering Association 




<3! 



-400 -300 -200 -100 100 200 300 400 

i [cmj 

Fig. 17 Deflection Profile Caused by Moving Wheel Load P 

According to Fig. 17, since 9 is very small, the vertical component of the force the wheel exerts 
on the rail is P^ = P. The corresponding horizontal component is P,^ = P9 = PId w/d^l. The expression 
for 9 is, according to Shakhunyantz (1959, p. 24), 



dw 



= -(K,S-p)w(0) 



(49) 



^=0 



Therefore 



Pk =P 



= P(K,S-p)w(0), 



(50) 



5=0 



where w(0) is given in eq. (47). For the track parameters used in Fig. 17 

P,^ = 120 N (27 lb) . 



(51) 



The force P,,, which a moving wheel exerts longitudinally along the rail, may be considered as the 
"creep driving force" for the problem under consideration. Note the corresponding force in Fig. 12. Once 
P|^ is known, the corresponding permanent rail displacements (creep) may be explained using the model 
shown in Fig.'s 12 and 13 and calculated using the second differential equation in (1), as done in the pre- 
vious sub-section II. Although P^ is relatively small at a wheel, each train has many wheels, and a track 
is subjected to many trains, and these forces may be sufficient to produce noticeable rail travel effects. 

An attempt to use this concept for determining the rail creep force was presented by 
Lyashchenko (1966). 

Conclusions 

The published analyses based on the Johnson hypothesis for explaining rail travel (creep), although 
on the right track, show a major shortcoming. Namely, they violate the principle of conservation of energy. 
In this connection, note the related problems by Anscombe and Johnson (1974) and by Chang, Comninou 
and Barber (1983), who studied the occurrence of slip caused by a moving load, using elasticity theory. 

The analysis based on the Johnson hypothesis presented in the present paper shows, by includ- 
ing the speed of the moving wheel v^, that its effect may be significant for the range 100 km/h < \>^ 
< 250 km/h encountered on main lines (Figs. 11 and 16) and should not be neglected, as done by 
Albrekht, Kogan, and Menshikova. The analysis also established that with decreasing vertical track 
modulus k, rail travel (creep) increases (Fig. 16). Also, according to eq. (40), an increase of the wheel 
load increases rail travel, whereas an increase of the axial resistance r^, decreases it. 

The presented analysis for determining the "creep driving force", which includes energy dissi- 
pation in the base by viscous damping, eliminated the violation of energy conservation. However, the 
included energy dissipation mechanism is of a different kind than the one used to date in conjunction 



Paper by Arnold D. Kerr and Alexander Babinski 383 



with the Johnson hypothesis, shown in Fig. 5. Therefore, there is a need for a creep analysis that will 
combine these two approaches by using a consistent energy dissipation mechanism. 

It is hoped that this paper will contribute to a better understanding of the mechanics of rail creep 
by railway engineers, and will form a basis for the future development of a general theory for this 
phenomenon. 

References 

Achenbach, J. D. and Sun, S.-t (1965) "Dynamic Response of Beam on Viscoelastic Subgrade", J. 
Engineenn)> Mechanics Division, ASCE, pp. 61-76. 

Albrekht, V. G. (1958a) "Ugon Zheleznodorozhnogo Puti i Borba s Nim" (Creep of railway tracks 
and counter measures. In Russian), Gos. Transp. Zhel.-Dor. Izdatelstvo. 

Albrekht, V. G. (1958b) "O Prodolnykh Silakli, Voznikayushchikh po Poverklinosti Soprikasaniya 
Podoshvy Relsa i Osnovaniya pri Prokhode Koles Podvizhnogo Sostava " (On longitudinal forces 
which occur at the interface of rail and base during the passage of the wheels of a moving train. 
In Russian), Trudy Moskovskogo Instituta Inzhenerov Zh/D Transporta, Vypusk 80/1, 

Anscombe, H. and Johnson, K. L. (1974) "Slip of a Thin Solid Tyre Press-Fitted on a Wheel" 
International Journal of Mechanical Sciences, Vol. 16, pp. 329-334. 

Babinski, A. A. (1995) "Rail Creep in Railway Tracks", Master's Thesis, Department of Civil 
Engineering, University of Delaware. 

Babinski, A. and Kerr, A. D. (1995) "Rail Creep According to the Johnson Hypothesis" Research 
Report No. 95-2, Department of Civil Engineering, University of Delaware. 

Bochenkov, M. S. ( 1 962) "K Voprosu o Vliyanii Progiba Relsa na Ugon Puti" (The effect of rail deflec- 
tion on rail creep. In Russian), Trudy Novosibirskogo Instituta Inzhenerov Zh/D Transporta, 
Vypusk 31, pp. 184-193. 

Camp, W. M. (1903) "Notes on Track — Construction and Maintenance", Auburn Park Publ. Chicago. 

Chang, F. K., Comninou, M. and Barber, J. R. (1983) "Slip Between a layer and a Substrate Caused 
by a Normal Force Moving Steadily Over the Surface", International Journal of Mechanical 
Sciences, Vol. 25, No. 11, pp. 803-809. 

den Tex, K. (1910) "Die Schienenwanderung in der Richtung des Verkehres" Organ fiir die 
Fortschritte des Eisenbahnwesens, pp. 334-335. 

den Tex, K. (1913) "Die Schienenwanderung in der Richtung des Verkehres" Organ fiir die 
Fortschritte des Eisenbahnwesens, pp. 372-373. 

Dorr, J. (1948) "Das Schwingungsverhalten eines fedemd gebetteten unendlich langen Balkens," 
Ingenieur Archiv, Vol. 16, pp. 287-298. 

Engerth, J. (1900) "Creeping of Rails", Proc. International Railway Congress, 1st Section, Question 
X, pp. 1-80. 

Frishman, M. A. (1942) "Stabilizatsiya Puti ot Ugona" (Creep stabilization of tracks. In Russian), 
Trudy NIVIT, Vypusk 4, Transzheldorizdat, Moscow. 

Haarmann, A. (1902), "Das Eisenbahngleis" , Kritischer Teil, Verlag von Wilhelm Engelmann, 
Leipzig, Germany. 

Kenney Jr., J. T. (1954) "Steady State Vibrations of Beam on Elastic Foundation for Moving Load", 
Journal of Applied Mechanics, pp. 359-364. 

Kerr, A. D. (1972) "The Continuously Supported Rail Subjected to an Axial Force and a Moving Load" 
International Journal of Mechanical Sciences, Vol. 14, pp. 71-78. 

Kerr, A. D. ( 1 98 1 ) "Continuously Supported Beams and Plates Subjected to Moving Loads — A Survey" 
Solid Mechanics Archives, Vol. 6, Issue 4, pp. 401-449. 



384 Bulletin 76! — American Railway Engineering Association 



Kerr, A. D. and Bassler. S. B. (1982) "Effect of Rail Lift-Off on the Analysis of Railroad Tracks" Rail 
International, October, pp. 38-48. 

Kerr, A. D. and El-Aini, Y. M. (1978) "Determination of Admissible Temperature Increases to Prevent 
Vertical Track Buckling" Journal of Applied Mechanics, Vol. 45, No. 3, pp. 565-573. 

Kogan, A. Ya. (1967) "Prodolnye Sily v Zhelezodorozhnom Piiti" (Axial forces in a railroad track. In 
Russian), Trudy VN-IIZhT, Vypusk 332, Izd. Transport, Moscow. 

Kogan, A. Ya. (1981) "Metod Osredneniya v Reshenii Zadachi Ugona Puti" (Method of averaging 
for solving problems of rail creep. In Russian) Vestnik Vsesoyuznogo Nauchno-Issled. Inst. 
Zh/D Transporta, Nr. 3, pp. 51-57. 

Kyuner, K. E. (1925) "Problemy Usileniya Verkhnogo Stroeniya Puti v Svyazi s Yavleniem Ugona i 
Deistviem Temperatury" (Problem of track reinforcement in connection with the appearance of 
creep and the effect of temperature, in Russian), Zheleznodorozhnoe Delo, Nr. 4, pp. 19-29. 

Lyashchenko, V. N. (1966) "O Prodolnykh Dinamicheskikh Gorizontalnykh Silakh, Voznikayushchikh 
Pri Vzaimodeistvii Kolesa s Relsom" (On longitudinal horizontal dynamic forces that occur 
during interaction of wheel and rail. In Russian) Trudy KhllT, Vypusk 81, Izd. Transport, 
Moscow, pp. 3-33. 

Marguerre, K. (1938) "Ober die Behandlung von Stabilitatsproblemen mit Hilfe der energetischen 
Methode" Zeitschrift fiir Angewandte Mathematik iind Mechanik, Vol. 18, Heft 1, pp. 57-73. 

Menshikova, V. I. (1972) "Dinamicheskie Prodolnye Sily i Peremeshcheniya Relsov Zheleznodorozhnogo 
Puti (Ugon Relsov)" (Dynamical axial rail forces and the displacement of rails of a railroad 
track [Rail creep]. In Russian), Trudy CNII MPS, Vypusk 466, pp. 83-189. 

Rzhanitsyn, R. A. (1949) "Nokotorye Voprosy Meklianiki System, Deformiruyushchikhsya vo Vremeni" 
(Problems of mechanical systems that deform with time. In Russian), Gosteoretizdat, Moscow. 

Rzhanitsyn, R. A. (1968) "Teoriya Polzucliesti" (Theory of creep. In Russian) Izdatelstvo Literatury 
po Stroitelstvu, Moscow. 

Sailer, H. (1928) "Der Eisenbahnoberbau im Deutschen Reich" Veriag der Verkehrswissenschaftlichen 
Lehrmittelgesellschaft m.b.H. bei der Deutschen Reichsbahn, Berlin. 

Shakhunyants, G. M. (1959) "Raschety Verkhnogo Stroeniya PutC (Analysis of Railroad Tracks. In 
Russian), Gos. Transp. Zhel.-Dor. Izdatelstvo, Moscow. 

Skibinski (1913) "Uber Schienenstoss-Verbindung" Organ fUr die Fortschritte des Eisenbahnwesens. 

Steele, C. R. (1967) "The Finite Beam with a Moving Load" Journal of Applied Mechanics, Vol. 34, 
pp. 111-118. 

Verigo, M. F. and Kogan, A. Ya. (1986) "'Vzaimodeistvie Puti i Podvizhnogo Sostava" (Interaction of 
track and the moving train. In Russian), Chapter 5, pp. 399-417. 

Wasiutynskii, A. A. (1896) "Usilenie Relsovykh Stykov" (Reinforcement of rail joints. In Russian), 
Trudy XIII Soveshchatelnogo Siezda Inzhenerov Sluzhby Puti Russkikh Zheleznykh Dorog. 

Weikard (1909) "Zur Frage der Schienenwanderung" Organ fUr die Fortschritte des Eisenbahnwesens, 
Nr. 20, pp. 361-363. 

Wirth, A. (1909a) "Die Schienenwanderung und ihre Verhutung", Zeitschrift des Oesterreichischen 
Ingenieur—und Architekten-Vereines, LXI Jahrgang, Nr. 20, pp. 317-322. 

Wirth A. (1909b) "Die Schienenwanderung und ihre Verhutung", Zeitschrift des Oesterreichischen 
Ingenieur—und Architekten-Vereines, LXI Jahrgang, Nr. 21, pp. 333-340. 

Z. ( 1 888) "Fur das Wandern der Schienen" Centralblatt der Bauverwaltung" , Nr. 32, August, p. 347. 




Recognized throughout the industry. . . 

For all the wood products and service you depend on. 



Companies like yours — whether 
Class I, regional, short line or transit 
railroads or contractors — have been 
ordering wood products from 
Burke-Parsons-Bowlby for more than 
30 years. That's a good sign BPB 
delivers the quality products and 
flexible service you can depend on. 
BPB offers the broadest wood 
product lines in the industry, in- 
cluding: crossties; switch ties; tie 
plugs; highway grade crossings; 



bridge timbers; paneled bridge 
decks; and more. 

With plants located in Virginia, 
Pennsylvania, West Virginia and 
Kentucky, BPB has the capacity to 
fill your daily and emergency product 
needs. Our own trucking subsidiary. 
Timber Trucking, allows us to deliver 
your order right on schedule. Today 
we're providing all the wood prod- 
ucts you depend on and developing 
the ones you'll need for tomorrow. 



Quality products, sorvice, inno- 
vation, our commitment to you and 
the future of railroading. 

Get BPB's wood products and 
prompt delivery on track for your 
operation. Call 1-800-BPB-TIES 
(1-800-272-8437). 






The Burke-PBTSons-Bowlby 
Corporation 

P.O. Box 231, Ripley, WV 25271 
(304) 372-2211 



385 



The Hidden 
Enemy 




You can't always see the enemy lying 
beneath the surface - fouled ballast 
waiting to combine with moisture to 
destabilize your track. 

A regular program of shoulder ballast 
cleaning helps keep your ballast 
performing as it should. This stand- 
alone operation provides many 
important benefits including: 

• Extends time between costly 
surfacing cycles. 

• Increases life of track components. 

• Helps eliminate the need for 
expensive undercutting operations. 







w^ 


jUHHHHHpHMI 


Unv^slluvUE 



The Loram Shoulder Ballast Cleaner, 
along with the Loram Badger Ditcher, 
are your keys to a complete , cost- 
effective drainage maintenance 
system. To learn more about how you 
can maintain the stability of your track 
structure, contact: 



MV W # WJrU^MWK 




Nobody builds it tougher. 

Or services it better. 

Loram Maintenance of Way, Inc. 

3900 Arrowhead Drive 
P.O. Box 188 
Hamel, Minnesota 55340 
Telepiione (612) 478-6014 
Telex29-0391. Cable LORAM 
Fax (612) 478-6916 



Engineered To Handle Higher Speeds, 
Tighter Turns And Heavier Loads, 
Today's Wood Crossties Offer You Superior Cost 
And Performance Advantages. 





Maybe not. Maybe even tougher when you consider the 
tremendous variables that track engineers contend with. 
Materials with differing physical properties and maintenance 
'M*!*!^! needs. Track structures that must stand up to an enormous 
range of loadings and speeds, for long periods of time, in all 
types of weather, over all kinds of terrain. Enter yet another 
factor— high-speed passenger traffic— and the complexity 
increases. 

Fortunately, there is a proven performer you can count on 
to handle your toughest demands. The treated wood crosstie. 

Pre-engineered by nature and enhanced by man, the wood 
crosstie is a marvel of natural science and applied technology. 

Concrete Euidence That UJood Is Vour Best Choice. 

For 150 years, the wood tie has taken everything 
that man and nature have dished out. Steep 
grades. Brutal environments. High speeds. 
Heavier axle loads. Tighter curves. 

Through it all, resilient wood has been and 
continues to be the material of choice for durability, 
economy and strength. No other material matches 
wood's value on freight and passenger lines. 

And innovations in hardware, installation 
methods and wood preservation have taken a 
good thing and made it even better By improving gage 
retention. Minimizing maintenance. And increasing tie life. 




Building track structures is decidedly 
complex. But choosing the best crosstie 
isn't exactly "rocket science", 
just say wood. 



m 

Railway Tie 
Association 



Wood crossties. Something to build on. 

115 Commerce Drive • Suite C 
Fayetteville.GA 30214 

Phone (770) 460-5553 
Fax (770) 460-5573 



Chemetrott^s ATV. 

Coming at you with more welding power than ever. 



On track. Off track. Chemetron's latest 
technology All Track Vehicle (ATV) can 
deliver field welding services fast, and with 
nnore consistent flash-butt (|Lrality than any 
other in-track welder. Ever. 

The technology behind the weld quality 
of Chemetron's ATV is 
an improved K-900 
welding head and the 
proprietor/ software 
our on-board 
computer uses to 
control welding 
cycles. Precisely 

Chemetron's 
mobile welder was 
engineered to 
exceed AREA 
specs, including 



the "upset to refusal" requirements. Our 
computer system guarantees plant quality 
in-track welds. Continuously. 

Offering optimum production for all 
roil sizes and metallurgies, Chemetron has 
complete mobile welding units for sale or 
ease, for short or long 
term contract welding. 
With full engineering 
and maintenance 
support. 

Put Chemetron's 
ATV in-track 
welder to work 
for you by calling 
Larry Taylor at 
847-520-5454. 
Today 




388 



Memoir 

Edward M. Cummings 

1920-1996 



Bom in 1920, Ed Cummings served with the U.S. Army Air Corps during World War II. After 
the War he attended Northeastern University in Boston under the GI Bill, receiving a Bachelor of 
Science in Civil Engineering degree. In 1947, following graduation, he entered railroad service with 
the Boston & Maine. He joined the Baltimore & Ohio Railroad in 1950 as Asst. Division Engineer 
at Connellsville, Pennsylvania. Over the next 33 years Ed held positions of Division Engineer, 
Regional Engineer, General Manager Engineering and Regional Asst. Chief Engineer with the B&O 
and its successor, the Chessie System Railroads. In 1983 he retired from Chessie at Detroit, 
Michigan, but continued to work as an independent railroad engineering consultant until shortly 
before his death on September 5, 1996. 

Ed was an active member of the AREA Conference Operating Committee and Committee 3, 
serving as Chairman of the latter from 1977 to 1980. He was also an active member of Roadmasters 
and the B&B Association. 

An avid railroad modeler, Ed was a life member of the NMRA, serving as President of the 
North Central Region of NMRA from 1969 to 1973. 

Edward M. Cummings is survived by his wife Lee and their two sons William and Thomas. 



390 Bulletin 761 — American Railway Engineering Association 

Memoir 

Howard W. Lichius 

1936-1997 



Howard Lichius (a bridge engineer in the Kansas City Office of Howard, Needles, Tammen & 
BergendofO had a severe stroke on Easter Sunday while visiting his brothers and sisters in St. Louis. 
He was taken to the hospital where his condition remained serious all day Sunday and Sunday night. 
Early Monday, March 31, 1997, Howard suffered a second stroke and passed away. 

Howard Lichius was a long-time HNTB employee. He graduated from the University of 
Missouri at Rolla and joined HNTB in 1958, where he spent his entire career He was nationally rec- 
ognized as an expert in the design of bridge structures, especially movable bridges. His experience 
included 40 bascule bridges, 22 vertical lift bridges, several swing spans, and two floating retractable 
bridges. Howard served on the American Railway Engineering Association's Committee 15 for the 
design of steel railroad structures, and assisted in the preparation of the movable bridge specifications 
for the railroad industry. Also, as a member of the National Cooperative Highway Research Council, 
he assisted in the preparation of movable bridge specifications for the AASHTO Bridge 
Subcommittee. 

Among the many notable projects to which Howard made a significant contribution are: the 
BNRR Williamette River bridge in Portland, OR, the longest double-track lift bridge in North 
America; the Badger bridge in Los Angeles, CA; the UPRR lift bridge at Lewiston, ID; the KCSRR 
bridge over the Red River at Alexandria, LA; the BNRR bridge over the Mississippi River at 
Hastings, MN; Walnut Street bascule bridge in Green Bay, WI; the award winning UPRR bridges 
over Grasshopper Creek and Little Saline Creek near Marion, IL; and the Latah Creek bridges for the 
BNRR in Spokane, WA. 

Other notable accomplishments included pioneering work related to HNTB's design and plan 
preparation for segmental concrete bridges in the early 1970's. Howard also developed a software 
program for the design of the operating and drive machinery for movable bridges of all types, includ- 
ing bascule, vertical lift and swing spans. 

Not only was Howard an outstanding technical engineer and a great teacher and mentor, he also 
demonstrated extreme dedication, integrity, honesty, and teamwork. 



Memoir 391 



Memoir 

Richard K. Pullem 

1933-1997 



Bom in Point Pleasant, West Virginia, Dick Pullem attended Marshall University and started his 
railroad career with the Chesapeake & Ohio Railway in Huntington, West Virginia. Beginning in 
1955 as a Division Roadman, he subsequently held positions of Instrumentman, Asst. Cost Engineer, 
Asst. Track Supervisor, Track Supervisor, Asst. Division Engineer, and Engineer Work Equipment. 
In 1970, Dick was promoted to Division Engineer at Saginaw, Michigan and two years later was 
moved to Akron, Ohio as Manager — Engineering. While at Akron, he left Engineering for a position 
in the Transportation Department, Division Manager, then moved to Cincinnati to head up the 
Chessie System's Western Division Business Unit. In 1983 he returned to Engineering and to 
Huntington as Asst. Chief Engineer — Maintenance and two years later became Chief Engineer of 
Chessie System Railroads, the last person to hold that title. With the merger of Seaboard Coast Line 
and Chessie System in 1986 Dick became the first Chief Engineer of CSX Transportation. He retired 
from CSX at the young age of 54, but continued to work as an independent railroad engineering con- 
sultant for several southeastern short lines. 

Dick was President of the Roadmasters and Maintenance of Way Association, 1977-78, and 
served on the AREA Board of Direction from 1985 to 1987. He also was on the National Defense 
Executive Board. 

Richard Kirby Pullem is survived by his wife Barbara and two sons, Richard, Jr. and Gary. 



392 Bulletin 761 — Atncrican Railway Engineering Association 

Memoir 

Jack P. Shedd 

1922-1997 



Jack P. Shedd, a member of Committee 8 for 2! years and AREA member for 24 years, died in 
Overland Park, Kansas, on January 19, 1997. He was 74 years old and resided in Eudora, Kansas. 

Bom in Laramie, Wyoming, on May 8, 1922, he married Vivian Casey in 1947. He is survived 
by his wife, a son, and three grandchildren. 

Mr. Shedd, a registered professional engineer, earned his undergraduate and graduate degrees 
in civil engineering from the University of Wyoming. He was a member of Sigma Tau Honorary. As 
a commissioned officer in the U.S. Navy from 1943 to 1946, Mr. Shedd received the Purple Heart. 
Mr. Shedd then joined the Department of Civil Engineering at Kansas State University, Manhattan, 
Kansas, where he taught from 1947 to 1952. Mr. Shedd was then hired by Howard, Needles, Tammen 
and Bergendoff (now HNTB Corporation), Kansas City, Missouri, where for 35 years he worked on 
bridge projects all over the United States. Following his retirement from HNTB, Mr. Shedd remained 
an active professional engineer as a consultant. Mr. Shedd also belonged to the American Society of 
Civil Engineers, the National Society of Professional Engineers, and the American Welding Society. 
Mr. Shedd was a valuable member of Committee 8 and his expertise in bridge design and construc- 
tion was very helpful in Committee 8's work. 



DIRECTORY OF CONSULTING ENGINEERS 



HARDESTY & HANOVER, LLP 



Consulting Engineers Since 1887 



RAIL STRUCTURES & HIGHWAYS 
BRIDGES - MOVABLE & FIXED 

1501 Broadway, New York, NY 10036, 

212-944-1150 FAX: 212-391-0297 

other offices: 

New Jersey, Florida , Virginia and Connecticut 




zr 

ZETA-TECH 



ZETA-TECH Associates, Inc. 

900 Kings Highway North 

Cherry Hill, New Jersey 08034 

(609) 779-7795 FAX (609) 779-7436 

e-mail: zetatech@)zetatech.com 



Technical and Economic Consulting for Railways 
and Rail Transit 
Technical Consulting • Specialized Software 

- Railway Track and Rail - Track Maintenance Planning 

- Maintenance Management - Operations Simulation 

- Vehicle/Track System - Track Inspection and Analysis 



Economic Analysis 

- System Economics 

- Cost Benefit Analysis 



Costing 

- Maintenance Planning 

- Operations Simulation 



Visit our home page on the World Wide Web: 
http://www.zetatech.com 



MS 



Providing services for engineering, 
design, planning, construction 
management and operations. 



/5 /5 Qroad Street, Bloomfield, NJ 07003 
Tel. 20 1-893-6000 • Fax 20 1 -893-3 13 1 



393 



RAILROAn AND RAIL TRANSIT 

Inspection • Planning • Design • Construction Management 

Bridges • Tunnels • Structures * Stations • Yards • Shops 

Trackwork • Electrification • Signals * Communications 



GannBtt FiBming 

ENGINEERS AND PLANNERS 



P.O. Box 67100 • Harrisburg. PA 17106 • (717) 763-7211550 
California Street • San Francisco, CA 94104 * [415] 981-5335 

vifww.gannettneming.com 



^Tm\ ENGINEERS 



CONSOER TOWNSEND ENVIRODYNE ENGINEERS, INC. 



Railyards and Shop Facilities 
Bridge Inspection, Retiabiiitation 

and Replacement 
Railroad Planning Studies 



Chicago, IL 
(312)938-0300 



■ Environmental Engineering 

■ Higti Speed Rail Studies 

■ Construction Management Services 

■ Grade Crossing Analysis 
Regional Offices: 

Orange, CA New York, NY Nashville, TN 

|7 1 4| 835-4447 (2 1 2| 682-6340 (6 1 5| 244-8864 

24 Offices Nationwide 



Complete Engineering Services for the 
Railroad Industry 



Tracks, Bridges, Structures 
Construction Management 
Shops, Warehouses, Offices 



Fuel Management 
Utilities 
Financial Studies 



E3 BLACK &VEATCH* 



8400 Ward Parl<way, Kansas City, MO 64114 (913) 458-2222. 
Offices Worldwide 
tittp//w/ww. bv.com 



394 



THOMAS K. DYER, INC. 



Rail Transportation 
Consulting Engineers 

Signals Track/Civil Communications Power Electrification 

1762 Massachusetts Avenue 

Lexington, MA 02173 

(617) 862-2075 

NEW YORK • PHILADELPHIA • CHICAGO • DALLAS • ST. LOUIS 



ElMGIIMEERilMB INC. 



Intermodal Yards • Track Design 
Embankment Subgrade Stabilization 

Design Build 
Environmental • Drainage • Surveying 



corpor-ate 
Heaqiarters 

630-858-7050 



Chicago, IL 
Area 

630-434-7050 



Nashville, TN 
Area 

615-552-2525 



Springfield, IL 
Area 

217-525-7050 



E-Mail sheathfuipatnckengineering com 



Web http //www patrickengineering com 



Serii\g the Railroad Industry 



STV 



engineers / architects / planners / construction managers 



Transportation Leaders 

Complete Planning, 

Engineering and 
Construction Services 



William F.Matts.Sr.VR 



Tel:2 12/777-4400 Fax: 2)2/529-5237 



225 Pork Ave. South. New York. NY 10003 



web site: www.stvinc.com 



395 



LTK Engineering Services 



Behind the scenes at some of America's 
most successful rail systems... 

Contact us today. 

Corporate Headquarters: 

Two Valley Square. Suite 300, Blue Bell PA 19^22 
215-5^2-0700 215-5^2-7676 FAX 

Regional Offices 

Portland, Los Angeles, Chicago. Philadelphia. Dallas. 
Seattle. San Francisco 




MODJESRHMASTERS 

CONSULTING ENGINEERS 
FIXED & MOVABLE RAILROAD BRIDGES 



P.O. Box 2345 

Harrisburg, Pennsylvania 17105 

(717) 790-9565 FAX (717) 790-9564 



1055 St. Charles Avenue 

Nevj Orleans, Louisiana 70130 

(504)524-4344 FAX (504) 561-1229 



http://www2.epix.net/-modjeski/ 

Harrisburg • New Orleans • Poughkeepsie • Bordentown • St. Louis 




PARSONS 
BRINCKERHOFF 

Complete Rail Engineering, Planning and 
Construction Management Services 



• Railroads 

• Rapid Transit 

• Bridges 

• Systems 



' Tunnels 
• Shops & Yards 
' Track 
' Vehicles 



Washington, DC — PaulReistrup (202) 783-0241 
San Francisco — Tony Daniels (4 15) 243-4600 

100+ Offices Worldwide 



396 



Providing Natioiiwide Railroad 
Engineering |(^-' 

Yard-aniii^im|al PUm^ 




T^SSmS 



Con$truction IVtana^ement 




Railroad Engineering Services 
Since 1910 



• Railroad Facilities & Building Designs 

• Track & Bridge Engineering 

Construction Administration & Surveys 

• Fueling Systems 



TKDA 



ENGINEERS • ARCHITECTS • PLANNERS 



1500 Piper Jaffray Plaza • 444 Cedar Street • St. Paul. MN 55101-2140 



(612) 292-4400 




397 



New York Pennsylvania Florida Ohio Michigan 



A. 



Complete Rail Engineering 
and Management Services 



BERGMANN ASSOCIATES 

Engineers, Architects, Surveyors, PC M 



44 Hudson Place 
Hoboken, NJ 07030 



(201)653-2898 
Fax (201) 653-3464 



HAZELET & ERDAL 



DAMES & MOORE 

Successors to the 
Scherzer Rolling Lift Bridge Company 

serving the railroad industry since 1 897 

FIXED and MOVABLE RAILROAD BRIDGES 

Design - New and Rehabilitation 

Inspection - Structural, Mechanical, & Electrical 

Rating and Analysis 

547 W. Jackson Boulevard, Suite 1500, Chicago, Ulinois 60661-5717 
Phone (312)461-0267 • FAX (312)461-0373 

Corporate offices nationwide providing a broad range of engineering services 



SHANNON &WILSON. INC. 

GEOTECHNICAL AND ENVIRONMENTAL CONSULTANTS 

• Landslide Evaluation & Correction 

• Embankment & Subgrade Stabilization 

•Tunnel Design & Maintenance 

• Environmental Management 

• Bridge Foundation Engineering 

Seattle • Richland • Fairbanks • Anchorage • St. Louis • Boston 

Corporate Headquarters, Seattle: (206) 632-8020 
400 N. 34th, Suite 100, P.O. Box 300303, Seattle. WA 98103 



398 



ESCA 



CONSULTANTS. INC. 

1606 WILLOW VIEW RO P O BOX 159 
URBANA. ILUNOIS (2171 38«-0505 



RAILROAD & HIGHWAY BRIDGES • TRACKWORK 
INDUSTRIAL FACILITIES • SPECIAL STRUCTURES 



INSPECTION & RATING 

REPORTS & STUDIES 

DESIGN & PLANS 

CONSTRUCTION SUPERVISION 



B \* AI A \ . BARRETT & 



m 



CONSULTING 
ENGINEERS 



;W 2 . 2 2 8 



100 
FAX 
312.228.0706 
DESIGMiyC A BETTER INFRASTRUCTURE 
TO LAUNCH US INTO 
THE 2IST CENTURY 



WE OFFER THE FOLLOWING RAILROAD DESIGN EXPERTISE: 

• MASS TRANSIT • INTERMODAL FACILITY 

• TRACKWORK • BRIDGES 
• INSPECTION AND RATING • DRAINAGE SYSTEMS 

• CONSTRUCTION MANAGEMENT • SURVEY 



1.30 E. RANDOLPH STREET SlITE 26.50 CHICAGO, ILLINOIS 60601 




NOLTE and ASSOCIATES, Inc. 

Engineers / Planners / Surveyors 

Serving Clients Tliroughout the Western United States 

2950 Buskirk Ave., Suite 225, Walnut Creek, CA 94596 
Tel: (510) 934-8060 FAX No. (510) 939-5451 



399 



HERZOG 



Herzog Contracting Corp. 

P.O. Box 1089 

St. Joseph, Mo. 64502 



816-233-9001 



Railroad Services 

• Railroad Construction & 
Maintenance 

• Commuter Rail Operations 

• Material Handling 

• Ultrasonic Roil Testing 

• Rail Car Leasing 



Comprehensive Transportation Engineering Services 



I iCillam 



Consulting: En 



49 West 37th St., New York, NY 10018 • 212-869-7800 
1 University Plaza, Hackensack, NJ 07601 • 201-489-8080 

Other Offices: NY«NJ»PA«CT«MA«NH«MD«FL 



PARSONS TRANSPORTATION GROUP 

BARTON-ASCHMAN • DE LEUW, GATHER • STEINMAN 



Comprehensive Railroad 

and 

Transit Engineering Services 

Parsons Transportation Group Inc. 

1133 15th Street, NW, Washington, DC 20005 

(202) 775-3300 • Fax: (202) 775-3422 



http://www.parsons.com 



400 



Gary A. Gordon, P.E. 

President 




GORDON, BUA & READ, INC 



CONSULTING ENGINEERS 

34 SALEM STREET 

READING, MASSACHUSETTS 01867 

(617)944-7110 

Fax (617) 944-6708 



Civil 

Railroad 

Structural 

Transportation 



<TL 



serving the railroad industry 
Call (800) 522-2CTL 



• testing of ties, fastener systems, rails, rail joints 

• state-of-the-art dynamic testing equipment 

• million-lb static & dynamic capacity 

• track design/construction problem solving 

• onsite track system testing & instrumentation 

• vehicle component testing 

Claire G. Ball 

Construction Technology Laboratories, Inc. 
5420 Old Orchard Road, Skokie, IL 60077-1030 



'Tra ckwork * Faciliti es * Operations » S tructures* Environmental « y 




K 



BRW INC. 

700 N.E. Multnomah 
Suite 1000 
Portland, OR 97232 
503/232-5787 
503/232-6373 FAX 



Trackwork* Facilities »Operations»Structures* Environmental 



401 



NOTES 



402 



Index of Advertisers 



A&K Railroad Materials, Inc 364 

Association of American Railroads. . . . cover 3 

Ballast Tools, Inc 275 

Brownie Tank Manufacturing 

Company 263 

Burke-Parsons-Bowlby Corporation 385 

Burro Crane 264 

Cattron Incorporated 245 

Central Manufacturing, Inc 275 

Chemetron True Temper 388 

Danella Rental Systems, Inc 241 

De Angelo Brothers, Inc 319 

Fairmont Tamper 284 

HDR, Inc 238 

Hanson & Wilson 260 

Hougen Manufacturing Inc 336 

Kerr-McGee Chemical Corporation . . back cover 

Koppers 208 

L. B. Foster Company 250 

Loram Maintenance of Way, Inc 386 

Magnum Manufacturing Corporation 279 



Modem Track Machinery Inc./ 

Modem Track Machinery Canada Ltd. . . . 237 

The Nolan Company 295 

Nordco 319 

Omni Products 206 

Osmose 249 

Pandrol Incorporated 296 

Pandrol Jackson 207 

Parker Hannifin Corporation 242 

Plasser American Corporation/ 

Plasser Canada Inc 262 

Premier Concrete Railroad Crossings 273 

Railquip, Inc 272 

Railway Tie Association 387 

Rocla Concrete Tie, Inc 330 

Sperry Rail Service 243 

Surety Manufacturing & Testing, Inc 320 

Sydney Steel Corporation 264 

W. H. Miner Division ii 

Westem-Cullen-Hayes, Inc 241 



Index of Consulting Engineers Directory 



BRW, Inc 401 

BAC Killam Consulting Engineers 400 

Bergmann Associates 398 

Black & Veatch 394 

Bowman, Barrett & Associates 399 

CTE Engineers, Inc 394 

Carter & Burgess, Inc 397 

Constmction Technology 

Laboratories, Inc 401 

ESCA Consultants 399 

Gannett Fleming, Inc 394 

Gordon, Bua & Read, Inc 401 

HDR Engineering, Inc 397 

Hardesty & Hanover 393 



Hazlett & Erdal/Dames & Moore 398 

Herzog Contracting 400 

L.S. Transit Systems 393 

LTK Engineering 396 

Modjeski & Masters 396 

Nolte & Associates, Inc 399 

Parsons Transportation Group 400 

Parsons Brinckerhoff 396 

Patrick Engineering 395 

STV 395 

Shannon & Wilson 398 

TKDA 397 

Thomas K. Dyer 395 

ZETA-TECH Associates, Inc 393 



403 



NOTES 



404 




HOW MUCH HORSEPOWER WOULD IT 
TAKE TO GIVE YOU THIS MUCH PUU.? 



ulling together this much information 
30ut the rail industry just got easier. The 
997/ 98 i\i\R publications catalog is now 
.ailable. VVe have compiled a single source 
f information about, what has long been, 
ne of the most dsnamic industries. 



jialysis of Class I Railroads 
.ailroad Ten -Year Trends 
.ailroad Facts 

Lail Cost Adjustment Factor 
lAR Railroad Cost Indexes 
"erritorial Distribution of Rail Traffic 
by Commodity Class 
'rofUes of U.S. Railroads 
I'eekly Railroad Traffic 
Weekly of Cars Loaded and Unloaded 
Lailroad Equipment Report 



Field Manual of Interchange Rules 
Office Manual of Interchange Rules 
TOFC/COFC Interchange Rules 
Rules Governing the Loading of 
Commodities on Open Top 
Cars and Trailers 
Manual of Standards and 
Recommended Practices 
Miscellaneous Specifications 
Specifications for Design Fabrication 

and Construction 
Car Construction — Fundamentals 

and Details Appendices 
Brakes and Brake Equipment 
Wheels and Axles 

Journal of Bearings and Lubrication 
Trucks and Truck Detail 
Lubrication (Shop Manual) 
Lettering and Marking of Cars 
Wheel and Axles (Shop Manual] 



Roller Bearings (Shop Manual) 
Side Frames and Truck Bolsters 
Locomotives and Locomotive 

Equipment 
Maint. Req. Brake C^ontrol Vahe 

and Equipment 
Couplers and Draft Gears 
Quality Assurance 
Specification for Tank Cars 
Ml 001 



We've pulled It all together! 




Association of American Railroads 

50 F Street N.W., Washington, D.C. 20001 



For more ordering information, call (202) 639-2211. 
Can't wait to receive your catalog? — Visit us on the web at www.aar.org. 




FOREST PRODUCTS DIVISION 

Setting 

TheRiture 

' In J t 

otoi 



At Kerr-McGee, we don't believe 

in woiring for the future. Insreod, 

we're nnoking ir happen today. 

Through everyday investnnents like 

ennployee education progronns, 

evironnnentally-sound manufacturing plants, 

produa innovation, rigid safety standards, 

and nnore. We're nnaking sure that 

the future of technology 

never stands still. In fact, we're 

^^"""Motion. 




KERR-MCGEE 

CHEMICAL 
CORPORATION 

TfeKCKING QUUITY F5Rj/ViERICA. 



ISO SON 

mm 

(405) 270-2424 • P.O. Box 25861 • Oklohomo Cily, OK 73125