(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 "Transactions"

Digitized by the Internet Archive 

in 2011 with funding from 

University of Toronto 



http://www.archive.org/details/transactions32engi 



Volume XXXII Part I 



Transactions 
(of 



3nrorporaicb 1887 



a* 



die (Arabian .-Sbriftg of (Etoil (Bngincn? 

being a 
Series of Papers Presented to the Institute 
in 1917, 1918 and 1919 
descriptive of 

THE QUEBEC BRIDGE 



Montreal Z/l 
1919 



The Right of Publication and Translation it Reserved 



H 



'S 



PREFACE 

THIS volume of the Transactions of The Engineering Institute 
of Canada consists of a series of papers descriptive of the 
Quebec Bridge, presented during the years 1917, 1918 and 
1919. These papers are being published together, as it is felt 
by the Council that it is advantageous to have the full description 
pf this important structure contained within one volume. 

The Quebec Bridge has been fully described in illustrated addresses 
and in papers before Headquarters in Montreal as well as before the 
great majority of the Branch 

On November 22nd. 1917. Mr. C. N. Monsarrat, M.E.I.C, 
Past Vice-President, gave an illustrated address at Headquarters, dealing 
generally with the whole structure and particularly with the details of 
the substructure. The text of this address is incorporated in the 
present volume on pages 1 to 16 inclusive. 

On December 6th, 1917, Mr. George F. Porter, M.E.I.C, delivered 
a comprehensive, illustrated address at Headquarters, on the super- 
structure of the Bridge and its erection. 

On December 20th, 1917, Mr. Phelps Johnson, M.E.I.C, Past 
President, presented before Headquarters a series of views descriptive of 
the shopwork of the superstructure. 

On January 10th, 1918, Mr. G. H. Duggan, M.E.I.C, Past President, 
read an exhaustive discussion on the design of the Bridge before 
Headquarters, which is to be found on pages 17 to 62 of this volume. 

On April 17th and April 24th, 1919, there was presented before the 
Montreal Branch of the Institute an exhaustive paper by Mr. Johnson, 
Mr. Duggan and Mr. Porter, describing the design, the manufacture and 
the erection of the superstructure, to be found on pages 63 to 162 of the 
present volume. 

In addition to the above presentations, in 1917 Mr. Monsarrat 
delivered his address before the Ottawa Branch and the Toronto Branch, 
while during the Spring of 1918 Mr. Porter delivered his address before 
all the Branches from Winnipeg West, as well as at the Professional 
Meeting in Halifax. 

Montreal, May 1919. The Editor 






CONTENTS 

Page 

The Substructure of The Quebec Bridge — C. N. Monsarrat 

M.E.I.C 1 

Notes on The Work of The St. Lawrence Bridge Company in 
preparing the Accepted Design of the Superstructure of 
The Quebec Bridge — G. H. Duggan, M.E.I.C 17 

The Design, Manufacture and Erection of the Superstructure of 
The Quebec Bridge — Phelps Johnson, M.E.I.C, G. H. 
Duggan, M.E.I.C, and George F. Porter M.E.I.C 63 

Appendix A —Summary of Tests — Quebec Bridge 163 

Appendix B — Notes on Quebec Bridge Stresses — A. L. 

Harkness, B. A. Sc, A.M.E.I.C 262 

Bibliography 304 



^nrorporalci) ISS7 

a* 

£l)c (Eaiiaiiian^ofifly of dM( (BoqinctnJ 



THE SUBSTRUCTURE 

of 
THE QUEBEC BRIDGE 

by 
C. N. MONSARRAT, M.E.I.C. 

IX August 1908, a Board of Engineers was appointed by the Dominion 
Government to prepare a new* design, specifications, and working draw- 
ings, and to supervise the rebuilding of the Quebec Bridge across the St. 
Lawrence, following the failure of the partly erected superstructure on 
August 29th, 1907 This Board was composed of a Canadian, II. E. 
Yautelet, M.E.I.C, former Asst. Chief Engineer, C. P. Ry., Montreal, 
as Chairman and Chief Engineer; an engineer from Great Britain, Maurice 
Fitzmaurice, C.M.G., M.E.I.C, Chief Engineer of the London County 
Council, who had been on the staff of the Forth Bridge; and an American, 
Ralph Modjeski, M.E.I.C, Consulting Engineer of Chicago. 

One of the first steps to be taken in connection with this work was the 
removal of the debris of the fallen structure which filled the space between 
the anchor and main piers on the south side of the river and extended into 
deep water north of the main pier. Some 9,000 tons of this material had to 
be cut up and removed, the work extending over a period of two years. 
The oxy-acetylene torch and dynamite played an equal part in the cutting 
of this material, and no great difficulty was experienced, except just south 
of the main pier, where some of the larger members were driven some 10 
to 15 feet into the bed of the river. 

The possibility of utilizing the old masonry for the new bridge was a 
matter that required considerable study and investigation. As the Board's 
studies of the superstructure developed, it was found that it would be d< 
able to widen the bridge from 67 to 88 ft. centre to centre of trusses. In 
order to accommodate the necessary increase in live load and also the addi- 
tional weight of steel required for the new design of this structure, it was 
found that the reactions on the piers would be nearly 100% greater than 
those of the original structure. As a consequence, the old piers as they 
existed, would not be large enough to accommodate the new bridge. 



It. was considered impracticable to enlarge or reinforce the north main 
pier owing to the fact that it was comparatively shallow and a reinforced 
pier would not lend itself to a proper distribution of the loads coming upon 
it. 

The south main pier, being considerably deeper, offered a better 
opportunity for enlargement. A scheme of reconstruction was worked out 
by which the old pier was to be taken down to the top of the original caisson 
and two new caissons 30 ft. wide were to be sunk, one' along the south side 
and the other across the West end, and a new pier built distributed 
over the old and two new caissons. In order to properly distribute the 
reactions from the superstructure it would be necessary to move the centre 
of the pier 15 ft. to the west and south. 

On the north shore it would be necessary to build an entirely new 
pier which was to be placed 57 ft. south of the existing pier, or towards the 
river. The difference between the final location of these two piers would 
result in reducing the length of the main span from 1800 ft. to 1758 ft., 
centre to centre of piers. 

Before entering into a contract for the construction of the substructure, 
a series of borings were made at and about the location of the main and 
anchor piers, which were driven to bed rock in each case and extending 
15 ft. into solid rock in order to make sure that it was bed rock rather 
than a boulder that had been struck. These borings showed that bed rock 
would be encountered at approximately the same elevation on both sides 
of the river, at a depth of about 101 ft. below extreme high water, or about 
85 ft. below the bed of the river at the location of the main piers. The for- 
mation of the bed of the river on the two sides, however, was entirely 
different. On the north shore heavy boulder formation was encountered for 
the entire depth, while on the south shore sand was encountered with only 
a sprinkling of boulders at various depths. 

The contract for the substructure was awarded to Messrs. M. P. & 
J. T. Davis, of Quebec, in February, 1910, and work was started 
immediately. 

The caisson for the north main pier was built at Sillery, about three 
miles below the bridge site, and was 180 ft. long and 55 ft. wide. It was 
constructed of 12" x 12" Southern Pine with a cutting edge of the same 
material 30" square. The caisson had a working, or air, chamber under 
the roof 8 ft. high in the clear and divided by longitudinal and transverse 
bulkheads into eighteen compartments. It was built on launchways laid 
with a 10% grade leading out into deep water, and when the walls of the 
caisson had been built up about 40 ft. it was lowered into its inclined posi- 
tion by means of jacks. When ready for launching an impetus was given 
by means of horizontal jacks at the rear as well as vertical jacks to reduce 
the friction on the launchways, the whole operation working smoothly and 
without mishap. 

2 



The eaisBOQ was (owed to the bridge site August 1, 1910, and, without 
difficulty, was placed in its proper position. It was held in place against 
the heavy current by steel cables attached to " dead men " about 500 ft. 
up and down the river. These wire ropes were supported by floats about 
100 ft. apart, and were provided with a tackle to take up the slack and keep 
the caisson in its correct position at all stages of the tide. 

When concreting started the caisson was drawing about 19 ft. of water, 
and weighed about 2,000 tons. The caisson first began to touch bottom at 
low tide after some 2,000 cubic yards of concrete had been deposited. It 
had been leaking to a certain extent but was easily kept dry by means of 
t wo steam pumps installed on the top of the caisson. At this time, however, 
an accident happened to the boiler equipment, and before it could be 
repaired, the caisson had filled with water, causing it to ground on the 
uneven bottom. As a result, the caisson was seriously strained, the seams 
opening up to such an extent that it was impossible to keep air in the work- 
ing chamber. It was, therefore, decided that it would be necessary to 
remove the concrete from the caisson and tow it to the dry-dock at Levis 
and attempt to repair it during the coming winter. 

This accident emphasized the difficulties of landing such a large caisson 
and there was an entire reconsideration of the design of the caissons and 
masonry, resulting in the abandonment of the scheme of enlarging the 
foundations of the old south pier. Having in mind the difficulties already 
encountered on the north shore, it was considered advisable to use the 
repaired caisson on the south shore where sinking operations would be 
much easier owing to the sandy formation of the river bottom. This 
entailed the sinking of an entirely new pier south of the old location. 
On the north shore it was considered advisable, in view of the experience 
gained, to sink two smaller caissons with a 10 ft. space between the final 
location of both caissons being fixed 65 ft. south of the existing masonry, 
thus restoring the length of the main span to 1800 ft., the same as that of 
the original bridge, and keeping the centre line coincident with that of the 
old structure. 

An examination of the large caisson, or caisson No. 1, showed that the 
longitudinal joints had opened up near the centre from 1 to 2 inches, and 
that the total length of the caisson had increased about 53^ inches. During 
the fall of 1910 and spring of 1911, this caisson was thoroughly over- 
hauled and repaired, the seams being caulked and the outside of the 
caisson and roof of the working chamber re-sheathed. On May 28th, 
1911, it was floated out and towed up the river a distance of about nine 
miles to the site on the south side, which, being exposed at low water, had 
been carefully prepared and levelled off. Before leaving dry-dock the 
caisson had been provided with a false bottom which reduced the draught 
froml9toll feet. As there was 15 ft. of water over the site of the pier at 
extreme high water, the caisson was floated in and placed in exact position 
for sinking without serious difficulty. A portion of the false bottom was 

3 



then removed, and the various shafts were left unobstructed in order that 
the rise and fall of the tide would not lift it from its bed. No further work 
was done on this caisson during the season of 1911, the contractor directing 
all his efforts towards the sinking of the caissons on the north side of the 
river. 

Caissons No. 2 and No. 3, for the north main pier, were constructed at 
Sillery, at the same location as caisson No. 1, the same details of construc- 
tion being followed throughout. Each of these caissons was 85 ft. long and 
CO ft. wide. Caisson No. 2 was towed to the bridge site on June 10, 1911, 
and caisson No. 3 on July 11 of the same year. In the sinking of these two 
caissons the contractor did not meet the same difficulty as experienced 
with caisson No. 1. The walls and roof of the working chamber were tight, 
preventing the leakage of water and escape of air. Some difficulty was, 
however, encountered in sinking owing to large boulders fouling the cutting 
edge, which resulted in this edge being forced inward from 6" to 10" in both 
caissons. In order to prevent further difficulties of this kind, which might 
seriously handicap sinking operations, the method of sinking was changed. 

Timber blocking was placed beneath the bulkheads, after which a 
trench was excavated all around and below the cutting edge, allowing the 
full weight of the caisson to be carried on this blocking. This trench was 
excavated to a depth of about 2 ft., after which it was filled with blue clay 
in bags. When all was ready the blocking was undersecured with water 
jets and the caisson lowered on a cushion of clay, which acted as a lubricant 
and also prevented considerable air leakage. This method was found to 
work very advantageously and further damage to the cutting edge was 
prevented. 

It was the original intention to sink both these caissons to rock, but, 
as the work progressed, the sinking became more difficult owing to the 
heavy boulder formation, and finally, when the caisson had reached 
elevation 20.0, the Board decided that the foundations at this point should 
be quite satisfactory for the loads which they would be called upon to sup- 
port. Bearing tests were made and it was found that a load of 59 tons per 
sq. ft. showed a settlement of x /% in., practically no settlement being noted 
at 20 tons per sq. ft. As the average working load at the foot of this pier 
was only eight tons, the Board considered that there would be no justi- 
fication in carrying the foundations to a lower level. Caisson No. 2 was 
sunk a total distance of 41.6 ft. at an average rate of 4.9 in. per day. 
Caisson No. 3 was sunk 38.3 ft. with an average rate of 5-% in. per day. 

After the caissons had reached their final location the working chambers 
were filled with concrete composed of one part of cement, two and a half 
parts of sand and four parts of small crushed stone. This concrete was 
made much drier than the concrete used in the main caisson, it being found 
that concrete deposited under compressed air gave better results when very 
dry than in a more or less liquid state. 

4 



Concrete was deposited in terraces, the men working towards the 
centre from the sides and ends. Great care was taken to ram the concrete 
thoroughly round the roof timbers so that a bearing would be assured under 
the roof of the working chamber. After the working chamber was filled 
as carefully as possible by hand the shafts were filled with concrete. As a 
still further precaution, a rich grout was forced in through the 4 in. blow 
pipes by compressed air under a pressure of 100 lbs. per sq. inch. After 
caissons No. 2 and No. 3 had reached their final depth, the material in the 
10 ft. space between them was excavated to a depth of 38 ft. below high 
water this space being then enclosed by means of timbers connected to 
the outside walls of both caissons. The area between the caissons was then 
filled with concrete up to a point 7 ft. below low water. The water was 
then pumped out and six steel girders 6 ft. deep were placed through the 
ends of the caissons and into cavities left in the adjacent concrete. The 
entire space was then filled with concrete in which the steel girders were 
embedded, thus forming a monolith upon which the masonry shaft of the 
pier could be built. 

The sinking of the large caisson for the south main pier was started 
July 28th, 1912, and was completed October 24th, 1912, or at the average 
rate of 0.75 feet per day during the entire period. The material encountered 
at this point was, as indicated by the borings, chiefly sand, and the caisson 
was carried down to rock, which was reached at 101 ft. below high water or 
86 ft. below the bed of the river. 

The difficulty experienced on the north side in keeping the cutting 
edge intact and also the fact that the caisson had previously been over- 
strained, led the contractors to take unusual precautions to prevent the 
possibility of any accident happening to the caisson during the sinking 
operations. For this reason, special appliances were devised for relieving 
the cutting edge from carrying all the load, and by the use of sand jacks 
the total weight of the caisson was distributed over the entire bottom 
area. The manner of using these sand-jacks was one of the most interesting 
features connected with the sinking of this caisson, and possibly merits 
special description. 

The jacks themselves were of very simple construction. The cylinders 
had an internal diameter of 31 inches, and were 36 inches long, constructed 
of \4 inch steel plate with 4-inch lap joint; two angles 1H" x 1}^" x 24'- 
reinforced the cylinder at top and bottom. The piston was a block of yellow 
pine 2' 6" square and 5 ft. long. Four feet at one end was round with a 
diameter of 29 in., thereby allowing 1 inch play in the cylinder. The lower 
end of the piston was reinforced by a 23^" x Y%' welded iron band. During 
operation the piston was attached rigidly to the roof of the working chamber 
by long screw bolts, and remained there permanently during the entire 
period of sinking. 

5 



In preparing for a drop, the first step was to excavate a hole under the 
piston. The cylinder was filled about two-thirds full of sand and placed in 
position under the piston and blocked up hard against it by means of 
timbers. While this was being done the caisson was supported on timber 
blocking under the bulkheads and other points. At the bottom of the 
sand jack was a 2" iron pipe extending entirely across the cylinder, the side 
of which was split and opened up to allow the sand to escape. This type 
had no bottom to the cylinder, the timbers acting as a support for the 
sand. Another type used had a steel bottom and two 3" holes with sliding 
cover at each side at the foot of the cylinder. The operation in both cases 
was the same. 

When everything was ready for a drop, the timber blocking supporting 
the caisson was undermined by a water jet and the full load taken by the 
sand jacks. A man was stationed at every jack, and at a given signal, 
afforded by the flashing of the electric lights, each man turned a hydraulic 
jet with 60 lbs. pressure into the hole at the bottom of the cylinder, thus 
washing the sand out. The sand was caught in canvas bags of uniform 
size. When the canvas bag was full the lights flashed again and the water 
jet was turned off. Another bag was then obtained, and at the signal the 
jet was again turned on and the bags filled. Each cylinder contained in the 
neighborhood of 16 bags of sand, and this operation was continued until 
the required settlement was obtained. By adopting the signal system and 
emptying the sand into bags, it was possible to ensure that the whole 
caisson was being sunk at a uniform rate, and that there was no reasonable 
possibility of any part of the caisson being strained by being sunk more 
rapidly than another portion. As a rule a drop of from 18 in. to 2 feet 
could be effected at each operation, the recurrence of the operations 
depending entirely on the nature of the material to be removed. When 
the drop had been finished the blocking was again placed under the bulk- 
heads to take the load of the caisson, and the holes under the sand jacks 
deepened in order that the operation might be repeated. The greater 
part of the material excavated in this caisson, being sand, was forced 
out through blow pipes. 

Practically no problems were encountered in the construction of the 
north and south anchor piers and the north intermediate pier. Both anchor 
piers were constructed south of the existing anchor piers. For the north 
anchor pier a coffer dam had to be constructed around the foundations 
since the foot of the pier was below high water mark. The south anchor 
pier was well above high water mark so that all excavation was in the dry. 

The anchorage girders were embedded in concrete in the base of these 
piers and the first length of anchorage eyebars connected thereto. Above 
the top of these eyebars a shaft was left at each end of the anchor pier to 
allow the remaining anchorage eyebars to be connected at the proper time. 
When the full dead load stresses were applied to these anchorage bars, the 
shafts were filled with concrete to the top of the first length of eyebars. 

6 



For the main piers entirely new stone was used, but for the anchor and 
intermediate piers the specification allowed the use of the stone from the 
old masonry. The greater portion of the old stone was consequently used 
in the ('(instruction of these piers. The abutments were not radically 
changed, it being only necessary to raise the ballast walls and make minor 
alterations to suit the new design. 

The masonry in the pier shaft consisted of grey granite rock faced 
ashlar obtained from quarries situated about 60 miles north of the bridge 
site. This facing was laid with alternate headers and stretchers and backed 
with concrete in which were embedded displacer stones having an approxi- 
mate volume of about one cubic yard. Headers were required to have a 
length of at least 23^ times their breadth with a minimum length of 7 ft. Bed 
joints were Yi inch throughout and vertical joints % in. for a distance of 
12 in. back from the face and not exceeding 4 in. at any point. All face 
stones of the top course were doweled to the second course with two 1 ]/i in. 
dia. steel dowels to each stone and extending 6 in. into each course. All 
face stones of the second and third courses were clamped together hori- 
zontally with steel clamps 16" in length x 1", with a 3 in. vertical projection 
at each end, and were also doweled to the course below, as described above. 
All face stones of the rounded end of the piers below the third course were 
both clamped and doweled. The upper 18 ft. of the main piers were built 
with cut granite backing and about 40% of these backing stones were made 
to project up into the course above in order to give a good horizontal bond. 

The concrete used in the caisson and the backing of the piers was 1 of 
cement, 23^ sand and 5 parts broken stone by volume, except the concrete 
in the working chamber, which was 1 of cement, 2J/£ sand, and 4 parts 
broken stone. The cement was required to pass a tensile test for neat 
cement of 450 and 540 lbs. for 7 and 28 days respectively, and for 1 part of 
cement and 3 parts of sand, 140 and 220 lbs., respectively. 

The power plant, dining-room for " sand-hogs " and a two-storey 
bunk house were located at the water's edge, at the foot of the cliff, as were 
also the mechanical plants which furnished the power for the various 
operations. All supplies and materials were delivered at the top of the 
cliff, some 160 ft. above high water level, and thence by gravity through 
chutes or elevator to the concrete plant and service tracks below. 

The air compressors discharged into a 12-inch main from which 7- 
inch branches with flexible rubber connections led into the two caissons, 
each branch being fitted with a gate valve so that the air could be cut out 
of either caisson at will. The main 12-inch pipe was about 500 ft. long 
and was laid in a sluice of running water in order to keep the temperature 
of the air down to about 75 degrees Fah. As a consequence the tempera- 
ture in the working chamber rarely exceeded 90 degrees Fah., although in 
the service shaft and man-locks the temperature exceeded 100 degrees Fah. 

7 



The concrete mixing plant was placed just at the foot of the cliff. 
Half-way up the slope was the rock crushing plant. The rock used for the 
concrete was obtained from an adjoining cut and was brought to the brow 
of the hill in cars which dumped into a chute leading to the crusher plant. 
The stone was fed into two gyratory crushers which were capable of dealing 
with about 500 cubic yards in twelve hours. After passing through the 
crushers the stone was led over an inclined screen of 2-in. mesh, and thence 
into a storage hopper bin of about 200 yards capacity. Three chutes led 
from this to the concrete mixing platform below, the mouth of each chute 
being directly over a mixer. From this platform the sand, stone, cement 
and water, were fed in the proper proportions to the mixers underneath the 
platform, which in turn dumped into self -discharging buckets on trucks, 
which were hauled to the caissons by horses. Three mixers were used on 
the work, two having a capacity of 2/3 cubic yards and the other 1-1/3 
cubic yards. Owing to the conditions under which the work was carried on 
the mixers never had a chance to work to their full capacity; their best 
da3''s work being 450 cubic yards for the 24 hours. 

The sand and coal were delivered from the upper level to the lower 
through chutes. On top of the coal chute was a double line of rails with 
balanced trucks which conveyed the cement from cars at the upper level 
to the storage shed at the level of the concrete mixing platform. 

As the " sand-hogs " employed on the excavation in the working 
chamber of the caisson were compelled to work in shifts during the whole 
24 hours, sleeping and dining accommodation was provided on the lower 
level capable of accommodating about 100 men. Similar accommodation 
was provided on the upper level for about half this number. 

On the dock the contractor erected a number of buildings, which 
included an office and bath room for the accommodation of the inspectors 
and a hospital with a doctor in continual attendance where first aid might 
be administered in case of serious accident, or regular treatment in case of 
minor troubles. There was also provided a coffee-house, kept at a high 
temperature, where the " sand-hogs " could change their clothes and receive 
hot coffee at the end of their shift in the working chamber. In addition 
to the above there were the usual stores, offices, etc., for the contractor's 
own use. In connection with the hospital arrangements there was also 
provided a steel hospital lock connected with the compressed air system, 
to which men suffering from the " bends " could be immediately trans- 
ferred and treated. 

For serving each caisson, four 30-inch shafts for material and two 30- 
inch ladder shafts were employed. For ejecting the sand and smaller 
stones, four 4-inch blow pipes were used. The larger boulders were broken 
up and hoisted through the material shaft in buckets having a capacity of 
2/3 cubic yards. Four 7-in. compressed air pipes supplied air to the work- 
ing chamber and served the blow pipes. Two 6-inch pipes supplied the 



water for " washing " the sand. One 2-inch pipe supplied high-pressure 
air for drilling, etc., and a second 2-inch pipe carried the wires for the 
electric lighting of the working chamber and ladder shaft. 

As soon as the sinking was completed on the north shore as much of the 
plant as could be spared was moved to the south side. The men's dressing 
rooms and sleeping quarters were placed on skids, launched into the river, 
floated across, and placed in position on the other side. The layout for the 
mixing plant, sand chute, coal chute, etc., was practically the same as on 
the north side of the river, all the materials being led to the lower level by 
gravity. The stone for the crushers was quarried directly from the top 
of the cliff so that one derrick could pick up the stone in the quarry and 
deposit it in the hopper leading to the crushing plant half-way down the 
cliff. 

The mechanical plants used on this side of the river were much the 
same as on the north side, although increased about 50%. Water at 100 lbs. 
pressure was led into the working chamber and distributed around all four 
sides in a horizontal main from 4 in. to 6 in. in diameter. Each of the 18 
compartments was provided with a valved outlet and jet pipe with 1 in. 
nozzle which was used to loosen the sand and gravel and facilitate excava- 
tion. Each chamber was also provided with a 6-inch vertical blow-out 
pipe and lighted throughout with electricity. 

The caisson was fitted with six 3-ft. material shafts, each having a 
Moran air lock, and with four 3-ft. ladder shafts having simple air locks 
composed of short upper sections with top and bottom diaphragms. In 
order to enable the maximum number of men to leave the caisson at one 
time, a 6-ft. horizontal steel man-lock, about 30 ft. long, was located on 
the deck of the caisson and built permanently into the solid concrete of the 
pier. This man-lock was reached from above by means of a 4-ft. vertical 
stair shaft. This lock was large enough to accommodate 30 or 40 men at 
once, thus greatly expediting the entrance and exit of each successive shift 
and effecting an economy of air consumption. When the men wished to 
leave the working chamber they assembled in this air lock and the air 
pressure was very gradually reduced untU normal pressure was reached. 
When nearing the final stage the average time that elapsed from the mo- 
ment of entering the man-lock until the pressure was sufficiently reduced to 
enable the men to pass into the outer air was about 35 minutes. The same 
process was repeated on entering the working chamber, although this could 
be done much more rapidly, the total length of time required being only 
about five minutes. 

A hospital lock was also established on the shore. If a man should 
pass from the working chamber to the outer air too quickly he was affected 
by what is known as the " bends." When a man was so affected, he was 
admitted to the steel hospital lock and pressure turned on until the pain 
ceased, after which the pressure was very gradually reduced until normal 

9 



pressure was reached. Under moderate pressures 100 men were employed 
in eight-hour shifts. As the caisson went down and the pressures increased, 
the lengths of the shifts were diminished gradually until the minimum of 
two 1-hour shifts in the 24 was reached. As the length of the shifts de- 
creased the rate of pay of the men increased. 

The total amount of masonry included in this contract is as follows: — 

North Abutment (Alterations) 404.5 cubic yards. 

North Intermediate Pier 1665.6 " 

North Anchor Pier 17736. " 

North Main Pier 31870.4 " 

South Main Pier 38279.4 " 

South Anchor Pier 16073. " 

South Abutment (Alterations) . . , 61.1 " 

Total 106,090.0 " 

After the piers had been constructed and pointed they were all thor- 
oughly cleaned by sand blast. The work of dressing the bridge seats, 
especially on the two main piers, was very important as it was necessary 
to ensure an absolutely level bearing for the large steel shoes carrying the 
structure. In order to provide datum points for the guidance of the stone 
cutters in dressing these bridge seats, grooves were first cut along the edges 
of each dress stone and points established at each intersection which were 
absolutely true in elevation. The stone cutters then cut away the stone 
between these points. When they got close to the finishing cut the surface 
was gauged by means of long steel straight edges manufactured from small 
eyebeams. The bottom surface of this straight edge was coated with 
paint and swept over the surface. The high spots on the masonry, there- 
fore was indicated by the paint scraped off the straight edge. These were 
further dressed and the process continued until the whole surface was 
absolutely level. As far as can be judged this bed was finished with a 
variation of not more than 1/100 of an inch at any point. About six weeks 
steady work with stone cutting machines and masons was required on each 
of the four bridge seats of the main piers. 

Owing to the great length of span it was impossible to make actual 
measurements across the river for the location of the piers. Base lines 
over 1300 ft. long were established on each side of the river, and by an 
elaborate system of triangulation the piers were located. After the main 
piers had been constructed a further series of triangulations was made, 
and it was found that the centre to centre distance was 1800 ft. 2-7/16 
inches) — about 2-7/16 inches more than originally planned. This extra 
distance was provided for in the design of the steel work. Actual measure- 
ments of the length centre to centre of main piers, made after the bridge 
was erected, checked the calculated lengths very closely. 

10 



S. H. Woodard, M.E.I.C., was engineer for the contractors and was 
responsible for the design of the caissons and for the many unusual features 
employed in sinking. 

During the progress of the work, several changes occurred in the per- 
sonnel of the Board, Mr. Fitzmaurice resigned in June, 1910, and was 
replaced by Charles MacDonald of Gananoque, Past President, Am. Soc. 
C.E. In February, 1911, Mr. Vautelet resigned. On April 4th, 1911, when 
the contract for the superstructure was signed, Mr. MacDonald, acting 
Chairman and Chief Engineer, retired, leaving Mr. Modjeski the sole 
remaining member of the Board as originally appointed. On May 1st, 
1911, C. N. Monsarrat, M.E.I. C, Engineer of Bridges, Canadian Pacific 
Railway, was appointed to succeed Mr. Vautelet as Chairman and Chief 
Engineer, and on May 17th, 1911, C. C. Schneider, past President, Am. 
Soc. C.E., was appointed as third member, replacing Mr. MacDonald. 
On the death of Mr. Schneider in January, 1916, H. P. Borden, M.E.I.C., 
was appointed to succeed him. 

The final report of the Board, in which a more detailed description of 
the substructure will appear, is in course of preparation and will be pub- 
lished at an early date. 

C. N. Monsarrat, M.E.I.C. 



II 




Floating Caisson No. 1 into position on Soutb Shore after repairs had been made. 
Debris of old Structure on foreground. 




Caisson No. 1 in position behind old Main South Pier which is in course of 
demolition. 



12 




Smaller Caissons Nos. 2 and 3 ready for launching at Sillery. 




View showing Sinking of Caisson No. 2 and Operation of Material Locks. 

13 




Steel Grillage and first length of Anchor Bars, South Anchor Pier. 



14 




South Main Pier in Course of Construction. Note the Alternate Headers and 

Stretchers in Granite Face and Vertical Bond Stones projecting through 

the Courses. 




Completed Main and Anchor Piers in North Shore. 
15 




Complete Main and Anchor Piers on South Shore. 




General View of Crossing Showing Completed Masonry. 

1G 



n n n 




£/ev-75-o' 



£/ei/ OQ4 



Section b-b 




DETAILS OF CAISSON 

' - pi No - !•— Originally Constructed for the North Main 
/' • > t£ Pier but u'«mately used on the South Shore. 



Ma/er/a/ Loc/r 





Section B-B 



DETAILS OF CAISSON 



No, 1. — Originally Constructed for the North Vain 
Pier but ultimately used on the South Shore. 



SECTION E-E 



SECTION C — C 



Ifc 




General Diagram of Crossing showing location of Old and New Masonry 



NOTES ON THE WORK OF THE 
ST. LAWRENCE BRIDGE COMPANY 

in preparing 

THE ACCEPTED DESIGN OF THE SUPERSTRUCTURE 

of 

THE QUEBEC BRIDGE 

by 

G. H. DUGGAN, M.E.I.C. 

A PAPER describing the superstructure of the Quebec Bridge is in course 
of preparation by the Engineers of the St. Lawrence Bridge Company. 
It will however be some time before the paper is presented as many cal- 
culations must be abbreviated, and the working drawings, and many of 
the drawings for the construction equipment, as well as those issued for 
instruction to the shop and the erection force, must be condensed into 
forms that will come within the compass of an ordinary paper. 

The paper is being written in order that the members of the Society 
may have for reference a technical description of the structure as built, 
and of the special plant and methods employed in its manufacture and 
erection, but it will not come within the scope of the paper to discuss 
the designs prepared for tender or the considerations that led to the adop- 
tion of the final design. 

There are however some engineering considerations in connection 
with the preliminary designs and trial work leading up to the final design 
which, while beyond the scope of the paper, may be interesting to the 
members of the Society. Moreover, I feel that the Society should have 
some record of those who actively contributed to the success of the under- 
taking because so many engineers have been employed upon the work, 
and it has been so unusual, and of such long duration, that a mere state- 
ment of those in charge of various Departments at the finish of the work 
would omit the names and responsibilities of some who should be recorded 
in the history of the undertaking. 

As it is probable that I have the most intimate knowledge of all phases 
of the work from the time our designs for tenders were started until its 
completion, the duty of giving this information and record seems to 

17 



devolve upon me, and no better way of presenting it occurs to me than to 
tell the story of our work, with a discussion of our designs up to the point 
where the description will be taken up by the forth-coming paper. 

A brief history of the preliminaries, and the award of the contract 
to the St. Lawrence Bridge Company, will assist in an understanding of 
the competition and the designs. 

The project of bridging the St. Lawrence at Quebec is an old one, but 
its early history has no bearing on the present Quebec Bridge, and may 
be neglected. Tenders were called by the Quebec Bridge & Railway 
Company in 1899, which later resulted in the award of the contract to the 
Phoenix Bridge Company for the bridge which failed on August 29th, 
1907. 

Immediately after the accident the Dominion Government appointed 
a Royal Commission to report upon it, and after receiving the report, 
appointed a Board of Engineers to prepare a new design for the bridge. 
I was told by the then Minister of Railways and Canals that in appointing 
this Board it had been the desire of the Government to select Engineers 
best qualified by their experience in long span bridges to deal with this 
unusual problem, and to appoint one engineer from Canada, one from 
Great Britain, and one from the United States. After advisement he had 
appointed Mr. H. E. Vautelet, M. Can. Soc. C. E., for a long time Bridge 
Engineer of the Canadian Pacific Railway, as Chairman and Chief 
Engineer; Mr.(now Sir Maurice) FitzMaurice, C.M.G.,M.I.C.E., of London, 
England, who had been one of the engineers on the staff of the FoTth 
Bridge; and from the United States, Mr. Ralph Modjeski, M. Am. Soc. 
C.E., who had been connected with many long span bridges. 

The report of the Royal Commission appointed to investigate the 
failure of the Phoenix Bridge in 1907 is very comprehensive, and goes 
beyond the mere taking of evidence and the investigation of the faults of 
the bridge, as the Commission assembled most of the available data on other 
long span bridges, illustrated their important features, recorded the tests 
on large size compression members that had any bearing upon the work, 
and made a number of tests to supply some lacking experimental data of 
the behavior of large compression members under stress. 

The Board of Engineers continued the investigations of the Royal 
Commission and made a number of trial designs. Mr. Vautelet selected 
one of these trial designs and brought it to the condition of a working design 
about the end of 1909. It was his intention to make this the Official Design 
on which the bridge was to be built, and to call tenders for the construction 
on it only. The other members of the Board did not consider the design 
to be in all respects a desirable one and consented to tenders being called 
upon it only on the condition that contractors would be allowed to submit 
tenders on their own designs if they so desired and, probably due to these 

18 



disagreements, tenders were not actually advertised until the 17th June 

1910. Before tenders were called and during the discussion on the design, 
Mr. FitzMaurice resigned from the Board. Mr. Chas. MacDonald, 
M. Can. Soc. C.E., a noted bridge builder, a Canadian by birth and perma- 
nent resilience, but then retired from active practice and spending much of 
his time in the United States, consented to become a member of the 
Board until the Contract could be awarded. 

After tenders were received, Mr. Vautelet still strongly contended 
for his design while his colleagues, Mr. Modjeski and Mr. MacDonald, 
favored the design of the St. Lawrence Bridge Company. To settle this 
dispute the Minister, as provided for in the original Order-in-Council, 
called in, with Mr. Vautelet's consent, two additional Engineers, Messrs. 
M. J. Butler, C.M.G., Past President, Can.Soc.C.E. and Henry Hodge, 
Ain.Soc.C.E. of New York, to assist the Board in coming to a decision. 
Four engineers of the Advisory Board recommended the design of 
the St. Lawrence Bridge Company, Mr. Vautelet only dissenting. 
Mr. Vautelet resigned when his colleagues' recommendation was accepted 
about the end of February 1911. 

Messrs. Butler and Hodge were appointed to the Board only to 
assist in deciding upon a design and specifications and their duties ceased 
with the signing of the Contract. Mr. MacDonald had also made it a 
condition that he should be relieved when the design was arranged and 
the contract awarded, and after the Contract was signed, on April 4th, 

1911, Mr. Modjeski was the only member left of the Board. About a 
month later, Mr. C. N. Monsarrat, M. Can.Soc.C.E., was appointed to 
the position made vacant by Mr. Vautelet's resignation, and shortly after 
Mr. C. C. Schneider, Past President of the Am.Soc.C.E., joined the 
Board. Mr. Schneider was regarded as the Dean of Bridge Engineers 
in America and was a valued member of the Board until his death in 1916. 
When Mr. Schneider died, Mr. H. P. Borden, M.Can.Soc.C.E., who had 
been Secretary of the Board, was appointed to fill the vacancy. 

Sometime before this Mr. Phelps Johnson, President of the Dominion 
Bridge Company, Past President of the Society, had arranged with 
Mr. F. C. McMath, M.Can.Soc.C.E., President of The Canadian Bridge 
Company, that in view of the magnitude of the work, and the interest of 
Canadians in making the work a Canadian enterprise, the Dominion and 
Canadian Bridge Companies should combine their forces in the organiza- 
tion of a special Company to tender on the bridge, and, if successful, to 
carry out its construction, each Company taking a half interest in the 
venture and contributing such of its staff as might be necessary to make 
an organization for the new Company. This Company was later incor- 
porated as the St. Lawrence Bridge Company. Prior to this Mr. G. F. 
Porter, M. Can. Soc. C.E., Chief Draftsman of the Canadian Bridge Com- 
pany, and Mr. P. L. Pratley, M. Can. Soc. C.E., of the Engineering Staff 

19 



of the Dominion Bridge Company, had been released to the Board of 
Engineers to assist in the preparation of the official design. 

The magnitude of the disaster to the bridge being erected by the 
Phoenix Bridge Company, with its lamentable loss of life and serious 
financial loss, coupled with the fact that the bridge was larger than any- 
thing that had heretofore been attempted and the probable very heavy 
cost of constructing the bridge in a proper manner, had caused serious 
misgivings in the minds of the public and the Government as to the 
practicability of the undertaking, and from the outset the Government 
had safeguarded itself in every possible way. 

A prominent clause of the contract read as follows: 
"The Contractor must satisfy himself as to the sufficiency and 
"suitability of the design, plans and specifications upon which the bridge 
"is to be built, as the Contractor will be required to guarantee the satisfactory 
"erection and completion of the bridge, and it is to be expressly understood 
"that he undertakes the entire responsibility not only for the materials 
"and construction of the bridge, but also for the design, calculations, 
"plans and specifications, and for the sufficiency of the bridge for the 
"loads therein specified. And the enforcement of any part, or all parts, 
"of the specifications shall not in any way relieve the Contractor from 
"such responsibility". 

To implement the above guarantee the S't. Lawrence Bridge Company 
was obliged to make a cash deposit of $1,297,500., and in addition both 
the Canadian and Dominion Bridge Companies signed guarantees for the 
completion of the bridge putting their whole assets at stake. 

The discussion of the designs will be addressed to those who have not 
made special study of long span bridges, and it is hoped those who are 
conversant with the subject will forgive if taken over familiar ground, or 
if the subject seems to be treated in too elementary a manner. 

Plate I. — Shows in outline the elevation of — ■ 

(1) The design of the Phoenix Bridge Company winch failed in 1907; 

(2) The Official Design when tenders were advertised, but afterwards 
known as Design 1, and supplemented by Designs up to V before tenders 
closed. 

Plate II. — Shows in outline the elevation of the great cantilever 
bridges that had been built up to that time. 

It will be seen that all existing bridges, except the Forth, were of so 
much less span than the Quebec Bridge that the chief guides or precedents 
for the problems in hand were the Forth Bridge, and the Phoenix Bridge 
which had failed. Many good features of the Forth Bridge were not 
applicable for reasons that will be given later. 

20 



The report of the Royal Commission which investigated the failure 
of the Phoenix Bridge, brought out clearly many faults of that design 
that could easily be corrected in a new design, some of .these being the 
small width centre to centre of truss, the very high unit stresses, the 
curved compression chords, the inadequate lacing of compression members 
and the poor splice connections of the members. 

The report also disclosed the use of open joints during erection. In 
our view the failure of the Phoenix Bridge may be ascribed chiefly to these 
open joints, and as open joints are difficult to avoid in designs requiring 
sub-divided main panels, it may be well to discuss this subject briefly. 

Plate III. — Taken from the report of the Royal Commission, shows 
the deformation diagram of the anchor arm of the Phoenix truss, and 
the open joints in erection. 

In ali framed structures the lengths of the members as manufactured 
or before entering into the structure, differ from the lengths they will 
have in the completed structure, due to the elongation or compression, 
as the case may be, induced by the stresses to which they are subjected. 
It is customary to make provision for this in the "framed lengths" of the 
members, by calculating the amount of the extension or compression, 
and diminishing or increasing the length of the member so that it may 
have its goemetrical length in the structure after the stresses are imposed. 
Thus, if the members were assembled on their sides on a flat surface with- 
out any stresses in them, the whole truss would be considerably distorted 
from the form it would have when erected, as shown by the heavy lines 
of the drawing. 

In cantilever erection the members are necessarily placed without 
load upon them and until the work has proceeded a considerable distance 
the truss will approximate to the form which it would adopt if laid down 
flat. As the load comes on, the truss is gradually brought toward the 
form it will finally have, when the weight of the centre span is attached 
to the end of the cantilever. 

In certain forms of trusses with sub-divided panels, the "framed 
lengths" which will bring the truss to its proper form when under full 
load, cause some of the members to take a considerable bend during 
erection although they will eventually be straight when under load. As 
erection requirements cause these members to be put up in sections and 
then spliced at the joints, they cannot be assembled easily without allowing 
all the bend to come where they are spliced, thus causing a wedge shaped 
opening at the joint with the sections of the member only bearing at one 
edge. As the load comes on and the structure deflects the members 
straighten and the joints gradually close, but before the joint comes to 
a true bearing over its whole surface, the intensity of the pressure on one 
edge is very great, and may cause failure, as in the case of the Phoenix 
Bridge. 

21 



Diag. 2 of Plate I shows the original design prepared by Mr. Vautelet. 

Plate IV. — Copied from the Official Drawing, shows the anchoT arm 
to a larger scale. 

Plate V. — Illustrates the cantilever arm and suspended span to the 
same scale. 

The truss had a main web system of the single warren or triangular 
type, each truss panel being sub-divided to give a point of support for an 
intermediate floor beam, and thus make the floor system panels of 
moderate length. This sub-division was accomplished by means of a 
triangular frame suspended from the mid-heights of the main diagonals. 
This system of bracing overcame in large measure the local deformations 
and secondary stresses of the ordinary sub-divided panels, but it required 
an adjustment in the top chord of the triangular frame, as the framed 
length of this chord during erection would be different from its final length 
when the whole truss was completed and had taken its normal deflection. 

Figures 1-la-lb-lc show in outline the progressive erection of the 
cantilever arm of the Official Design. 

Mr. Vautelet had worked out two schemes of erection for this bridge, 
the two schemes differing only in the method of erecting the anchor arm. 

In the first it was intended to erect the anchor arm complete on 
false-work; while in the second it was intended to place a temporary 
supporting pier at the first main panel point shorewards from the main 
pier, erect this panel on false-work and then erect both the River and 
anchor arms as cantilevers — each arm to be carried out at such a rate 
that they would always balance over the base formed by the main and 
temporary piers. 

In either case, the scheme of cantilever erection was to use a top 
chord traveller starting at the main pier and working out panel by panel. 

The general form of the truss prevented any panel from being self- 
supporting until a main panel point had been reached, and it was intended 
to overcome this, by making use of the compression chord of the triangular 
sub-truss frame as the upper tension member of a cantilever that would 
hold up the lower half of the second main panel until the long 
tension diagonal could be connected, this small cantilever being anchored 
by means of tie-bars to the corresponding point of a completed panel 
in the anchor arm. Similarly, when the third panel was being erected, 
temporary tie-bars were to be placed between the chords of the first and 
second sub-trusses; in the first instance to hold the long compression diag- 
onal during erection, and afterwards to make use of the sub-truss as a 
cantilever as in the first panel. The principle was to be continued through- 
out the cantilever arm. 

22 




Figure 1 




Figure la 



23 




Figure lb 




Figure lc 



24 



This method of erection required that the upper chords of the sub- 
frames should be designed to carry their compression stresses when the 
bridge was completed, and should in addition be capable of carrying the 
heavy tension Btresses imposed during erection. Moreover, and herein 
lay a considerable difficulty, they should be capable of being adjusted 
with nicety to varying lengths while under heavy stresses. 

A consideration of the Official Design in the light of the investigations 
and the data assembled by the Royal Commission and by the Board of 
Engineers assured us that Mr. Vautelet had put on paper a bridge in 
which, if built, every confidence could be placed that it would perform 
the work for which it was designed. It was, however, manifest that many 
of the members would be much too large to manufacture with any existing 
equipment, and that the manufacture must be carried out with a degree 
of accuracy hitherto unattained to assure that the parts would go together 
in the structure and perform their intended functions properly. This 
requirement could, however, be fulfilled as it only needed proper equip- 
ment and exceptionally good workmanship, the serious difficulty being 
the erection or the assembling of the members to construct the Bridge. 

When I returned to my old position as Chief Engineer of the Dominion 
Bridge Company in January 1910, all the essential details of the official 
design had been worked up, inquiries were being made as to the largest 
sizes of material obtainable and possible methods of erection were being 
considered. Mr. Johnson had followed the work of the Board, and when 
it was decided that the Dominion and Canadian Bridge Companies would 
tender jointly for the construction of the Bridge, he had given much thought 
to the problems of its construction. He had early foreseen such great 
difficulties in manufacture and erection that he considered desirable some 
radical departures from Mr. Vautelet's design or, indeed, from any design 
for a long span bridge that had heretofore been illustrated. Mr. Johnson's 
studies led him to an appreciation of the possibilities of what has since 
come to be known as the "K" form of bracing, never before used for an 
important structure and as he and his staff were at the time fully 
occupied with other important work, he asked me to take up the 
development of this design and generally supervise the preparation 
of tenders. 

Mr. F. P. Shearwood, M. Can. Soc. C.E., Assistant Chief Engineer 
of the Dominion Bridge Company, had been collaborating with 
Mr. Johnson in considering the erection of the Official Design and the pos- 
sibilities of the alternative "K" design. He had already made a number 
of preliminary sketches and outline strain sheets, and it was unfortunate 
that Mr. Shearwood's regular duties prevented him from following up 
this work. He was, however, available for consultation and was of much 
assistance throughout the development of the "K" design. 

25 



The Designing staff was built up as rapidly as possible, Messrs. 
Harkness, Wilson and Linton of the Dominion Bridge Company Staff 
early being transferred to it. As the work developed the staff was added 
to until when Mr. Porter and Mr. Pratley returned from the Board of 
Engineers, about May 1910, the staff had outgrown the available 
accommodation and a new office was built in which the preparation of 
our designs could be carried out. The Staff consisted of Messrs. G. F. 
Porter, M. Can. Soc. C.E.; P. L. Pratley, M. Can. Soc. C.E., A. L. Harkness, 
A.M. Can. Soc. C.E., L. R. Wilson, A. M. Can. Soc. C.E. (now Major); 
E. C. Kerrigan, A.M., Can. Soc. C.E., H.M. Lamb, A.M. Can. Soc. C.E. 
(now Professor): A. P. Linton, A.M. Can. Soc. C.E., (now Major): Jas. 
McNiven, A.M. Can. Soc. C.E. (also overseas) and F. C. MacDonald, 
with frequent assistance of others from the Drawing Office in the making 
of tracings and the computation of weights and ordinary stresses. 

The work was so unusual, with so little precedent to guide, that it 
required a large staff of engineers to carry it on. 

In preparing competing designs for ordinary bridges there is so 
much data at hand as to the cost of shop manufacture, erection costs, and 
the weight of details, that it is generally unnecessary to do more than 
prepare strain sheets with a few details from which weights can be very 
closely estimated; but, the magnitude of this bridge required, in addition 
to calculations for dead and live load stresses, extended calculations of 
wind, temperature and traction stresses and also of the bending stresses 
due to the weight of the unsupported length of the members, and to the 
elastic deformation of the structure or what are usually termed secondary 
stresses. The make-up of the members, as well as the details for connect- 
ing the several members together, also necessarily differed much from 
previous practice with smaller members. 

The above considerations made it impracticable to follow the ordinary 
course of balancing the merits of trial designs by comparing outline strain 
sheets, and for every design which was considered worth a real trial it 
was necessary, after fixing the length of the suspended span and deciding 
how it might be erected, to work back from the ends of the cantilever 
arms to the piers, designing and detailing each member with sufficient 
accuracy to obtain a very close approximation of the final weight of the 
member in the structure, and particularly to consider how the larger 
members could be manufactured and erected in the bridge. 

Concurrently with the design of the bridge itself it was therefore 
necessary to keep in view the design and cost of the plant and equipment 
for manufacturing the work, the transportation of the members to the 
site and the design and cost of the erection plant necessary for the design 
in hand. Included in the latter are the special erection travellers, steel 
erection staging, pontoons and storage yards with heavy cranes, in addition 
to the ordinary erection equipment required for a large bridge. 

26 



Our preparations are evidence that we were most anxious to obtain 
the contract, but we were equally anxious, if successful in our tender, 
that the Bridge should have a pleasing appearance, that none of the errors 
of the Phoenix design should be repeated, that the make-up of the members, 
the details and the connections should be of the most efficient character 
and, indeed, that we might be confident the bridge would be in all respects 
a credit to us and to Canadian bridge building. 

A superficial examination of the Official Design had revealed that it 
was the result of much careful work, and that it had many excellent 
features. It was therefore felt that before making, or concurrently with 
the preparation of, our alternative designs, a careful detailed study of 
Mr. Vautelet's design would be of much assistance in developing our 
own work. The critical examination confirmed the opinion that in many 
respects, little or no improvement could be made. The principal features 
that were adopted, to be incorporated in the alternative designs were: 

(a) The general form of the compression members with abutting 
joints fully spliced before being subjected to stress; 

(b) The compression chords without bends in their length but 
increasing in depth towards the piers to keep a proper ratio of thickness 
of material as the stresses increased. 

The curved bottom chords of the Phoenix design while perhaps tending 
to economy of material, presented several very objectionable features, 
the principal being that the horizontal wind forces, and these are very 
considerable, cause heavy vertical components at the joints, reversing 
in direction as the chord is in tension or compression. These do not 
exist with the chords lying in one plane from end to end. It is most 
difficult if not impracticable to fully or efficiently splice chords deflected 
at panel points. 

(c) All of the main compression members in the Official Design 
were made up of four webs, shop connected in pairs, so that each truss 
was virtually two trusses placed close together and connected by tie- 
plates and lattice bars in the field. This assured a better distribution 
of stress throughout the members of the truss, permitted the heavy members 
to be shipped in practical lengths and greatly facilitated the erection. 

(d) The construction of the shoe and transfer of the load from the 
steel work to the masonry. The total vertical load was estimated at 
30,500 tons; the transverse wind load at 790 tons, and the longitudinal 
wind load at 3,730 tons; a horizontal compression load of 15,200 tons. 
The loads required a bearing area on the masonry of 700 square feet, 
and there were many difficulties in the way of evenly distributing such 
large forces over so great an area while at the same time providing for the 

27 



transverse forces. The pins were so placed that each loaded the area of 
the bed-plate tributary to it and the design amply provided for all the 
forces as well as the necessary stability. 

(e) The sleeves on the pins to reduce the friction and thus provide 
for the necessary deflection during erection were also good. 

The displacement diagrams were good as well as the secondary 
stresses resulting from deformation. 

The objectionable features from the theoretical view point and from 
that of construction will be referred to in comparing the Official Design 
with the present Bridge, where it will be shown that we were able to depart 
from the make up of some members, and from the details and the connec- 
tions between web members and chords in a manner muteh to our advantage 
in shop and erection, and we think also to the betterment of the final 
structure. 

Plate VI. — Gives a comparison of the larger choVd sections of 
the present Quebec, Forth and other great cantilever bridges. The 
section of the Quebec chords is much larger than that of any of the 
other bridges, having 1940 square inches against the next largest of 853 
square inches, and the section of the Official Design was still larger, 
having 2,038 square inches. The Hell Gate Arch at New York, since 
constructed, has a large section, 1,392 inches, but that is of different 
type of construction and is not really comparable for our purpose. 
Similarly, the circular section of the Forth Bridge is not comparable for 
reasons to be given. 

Plate VII. — ShoVs the cross sections of all the important members 
in the cantilever arm of the bridge as constructed. The bottom chords 
between panel points were about 86 feet long, and the heaviest section 
of bottom chord weighed, with its details, about 380 tons. It would 
be impracticable to place so large a member with its center of gravity 
some 45 feet from its connection to the work already built. By splitting 
it down the longitudinal medial line and splicing it about the middle of 
its length, each panel of lower chord could be manufactured, shipped and 
erected in four pieces, none exceeding 93 tons in weight and of which the 
centre of gravity was only some 25 feet away from the point of its 
connection or where the erection traveller could stand. The large com- 
pression diagonals, while not so great in sectional area, were considerably 
longer and were really more difficult to handle. The load of these heavy 
members and the reach necessary to place them, thus became the 
measure of capacity of the erection traveller. 

Plate VIII. — Shows the general character of the erection operations 
of the Official Design, the temporary members for holding up the 
permanent members of the bridge already placed but not finally con- 
nected, and the members that the traveller will place when in the position 

28 



shown. Although the amount of permanent bridge to be supported by 
temporary members would necessarily vary for different positions of 
the traveller, the principle remains the same throughout the whole 
erection of the cantilever arm. 

Plate IX. — Shows the top chord traveller designed for the above 
method of erection. 

Travellers having a reach of one panel, one and one-half panels, and 
two panels, were each designed and compared, the one illustrated with 
a two panel reach being adopted as having the minimum of objectionable 
features. 

Work of this magnitude cannot be handled without risk in spite of 
every precaution, and much consideration was given to adopting methods 
and plant that promised a maximum of safety. In a general way the 
short reach traveller is much lighter and the traveller itself is thus safer 
to handle, but it requires more temporary material for holding up portions 
of the structure that cannot be permanently connected up, with the tra- 
veller in the position for placing these members. It was therefore con- 
sidered that while the long traveller was more difficult to handle, its use 
involved less risk than having so much of the permanent structure hang- 
ing on temporary adjustable members as would be required with the 
short traveller. The long traveller also permitted much more rapid work. 

It was intended to carry the traveller upon temporary stringers set 
by the traveller itself on the top of the vertical posts as these were erected, 
those over which the traveller had passed being picked up and shifted 
to the forward position as required. 

The traveller alone was estimated to weigh about 1,000,000 pounds, 
with a moving load on its front wheels of nearly 900,000 pounds, and when 
lifting its maximum loads the reaction of the front supports would be over 
1,500,000 lbs. These heavy loads necessitated heavy stringers to carry 
them, and very heavy stools to rest on the top of the posts on which the 
traveller could be blocked when at rest and working. They also required 
considerable increase in the sections of the vertical posts of the truss and 
in the sway bracing connecting these posts, the normal functions of these 
posts in the completed structure being merely to carry the weight of the 
top chord. 

It was difficult to plan how the stringers already passed and left to 
the rear could be picked up, transferred to the front end of the traveller, 
and set in the position required to supply a track on which the traveller 
could proceed. There was a decided risk in passing this very heavy trav- 
eller down the sloping chord and stopping it at exactly the right point, 
as any positive stops used could only be of a temporary nature, it being 
necessary to remove them to allow the traveller to pass on after performing 

29 



its work at the point in question. It was difficult to give lateral stability 
to the stringers carrying the traveller with its great reach and large wind 
surface. 

In addition to the above, the operation of the traveller when erecting 
the members of the bridge presented many serious problems. It has been 
stated that for the sake of reducing the weight of the members to pieces 
that could be handled, every main member of each truss was field connected 
down a medial plane. The general width of the important compression 
members was about 10 feet out to out, and the width centre to centre 
about 5'6". The heavy members to be erected could only be brought 
out on the floor of the bridge as far as the forward point of support of the 
traveller. They must then be picked up by the tackles, eased forward 
and transversely before being hoisted into their exact position. To do 
this conveniently the tackles should be hung over the centre line of the 
piece being hoisted, but, if over the centre line of the outside half, the 
tackles would be 5'6" too far out from the inside half. Generally this 
placing is performed by a guy tackle, but with such heavy weights and 
short drift it would add appreciably to the stresses, and consequent neces- 
sary capacity of the hoisting equipment. 

The above difficulties were not due to the form of truss or to the 
general design of the structure beyond its magnitude, the problems arising 
from the large dimensions of the members both in length and width and 
the very heavy weights to be handled at a long distance from the point of 
support. There were, however, other difficulties due to the form of truss; 
the principal being the large quantity of temporary adjustable members 
and their connection to the permanent members, together with the exact 
adjustment under very heavy stress necessary in order to make connections 
of main panel points. Some of these members would have a stress of 
3,300,000 lbs., a heavy force under which to make delicate adjustments. 

Other difficulties presented themselves in carrying out the erection 
step by step and, while none of them were insurmountable, the sum was 
so great that it seemed worth making every effort to avoid them. I think 
it was the knowledge that these problems would arise long before the 
erection was worked out in detail that led Mr. Johnson to the "K" form of 
bracing. 

The Official Design I. was estimated to weigh 72,700 tons, of which 
52,000 tons was nickel steel. In very long span bridges a limiting length exists 
where the structure can only carry its own weight, and is unable to carry 
any superimposed load. The length naturally depends on the relation 
existing between the strength of the material under stress and its weight. 
For ordinary carbon steel construction and the loading specified, the length 
of the Quebec Bridge span begins to approach this limit. It was therefore 
of great importance that there should be no unnecessary dead-weight, 

30 



particularly in the suspended span and at the ends of the cantilevers, 
where every pound of load, either dead or live, requires several pounds of 
material in the cantilevers to carry its effect to the supporting piers. 

Nickel steel, which w r as known to be about 40% stronger in tension 
and was expected to be 40% stronger in compression, although not proved 
by experiment at that time, was hence a necessity in portions of the struct- 
ure. It cost, however, at the time the estimates were being made, nearly 
two and one-half times as much as carbon steel, and it was apparent that 
if carbon steel were used where it would not affect the sections of the 
members in the long channel span, — namely: in those members supported 
directly by the piers and throughout the anchor arm, — -while it would 
considerably increase the total weight of the structure, much economy in 
cost would result. It was also apparent that there was much unnecessary 
weight in the long vertical redundant members inserted only to support 
the weight of the top chord and the erection traveller, but which could not 
carry any live load nor serve any other useful purpose in the completed 
bridge. There was promise of economy and improvement in appearance 
in reducing the depth at the ends of the cantilevers of both the river and 
anchor arms, and in reducing the lengths of the anchor arms. 

Plate X. — Taken from Mr. Modjeski's paper on long span bridges, 
shows a graphic comparison of the weights of the Quebec and Forth 
Bridges. While this does not enter into the present discussion it shows the 
very great difference in weight per foot of the centre portions of the 
bridge, and the portions near the main supporting piers and serves to 
illustrate the importance of reducing the weight towards the centre of the 
channel span. 

Tenders were advertised on the 17th June 1910, to be submitted on 
the 1st of September, this date later being extended to the 1st of October 
1910. The time allowed would have been altogether too short to prepare 
alternative designs, but the Board had issued a preliminary specification 
at the end of 1909 and the designs and information of the Board were 
available to contractors who expected to tender, so that in reality there was 
about eight months in which to prepare tenders. As already stated, the 
Board had evolved satisfactory forms of compression members and in 
many instances details that could be used and the Official Design supplied 
data from which close approximations of weight could be made for alter- 
native designs — indeed the work of the Board was so thorough that prac- 
tically all existing knowledge on the subject was reviewed and put into 
workable form. 

While Mr. Yautelet and I did not agree on the design of the Bridge, 
I wish to acknowledge the large contribution made by him to the know- 
ledge of the subject and his readiness to explain his designs and the results 
of his studies. A history of the undertaking would be incomplete if 
Mr. Vautelet's part therein were not included. 

31 



We are also assisted by the return of the members of our staff, Messrs, 
Porter and Pratley. Mr. Porter had been in charge of the design of the 
details for the Board of Engineers and, while the make-up of the members 
and the details finally used differed considerably from those of the official 
designs, the familiarity with these large and unusual details acquired by 
Mr. Porter during his service with the Board, no doubt enabled him to lay 
out this work much more rapidly than if it had been new to him. 
Mr. Pratley too, had much experience in the intricate calculations of the 
secondary stresses for the Board's designs, and was thus enabled to 
encompass a very large amount of work in calculating our designs, and 
because of the experience of these two men with the Board of Engineers 
our work was very much lightened. 

The specifications issued, fixed the distance centre to centre of piers, 
1758 feet; a clear head room of 150 feet for a distance of at least 600 feet 
in the centre of the span; the elevation of grade; the maximum height of 
steel work over the pier, 290 feet; the cross section of the roadway and the 
width, centre to centre of trusses, 88 feet; the loading and the unit stresses. 

The requirements for material, testing and workmanship called for 
the very highest standard. 

It was required that tenders on Contractors' designs be accompanied 
by complete stress sheets, deformation diagrams, details and make-up of 
all members, splices, lacing, and all information necessary to judge the 
adequacy and agreement with the specification of the proposed design. 

The loading and unit stresses were briefly as follows: — 
LIVE LOADS. 

Trusses: Railway Load — on one or two tracks, 5000 lbs. per lin. ft. 
with 2-Cooper's E-50 Engines placed to give maximum stress. 



Floorbeams & Stringers and 
Members receiving Max- 
imum Stress for a length of 
moving load covering 2 pan- 
els or less 



Highway Load — 920 lbs. per lin. ft. of each roadway for trusses. 

^ 100 lbs. per sq. ft. or 4,600 per lin. ft. of 
bridge. 
2-53 Ton cars each 60 ft. long and 12 ft. 

wide on each track. 
A concentrated load of 24,000 lbs. on 2 axles 
10 feet apart. 
DEAD LOAD. 

Railway: Floor 670 lbs. per lin. ft. of each track. 
Highway: Floor above stringers 2300 lbs. per lin. ft. each roadway. 
Snow load of 1500 lbs. per lin. ft. of bridge. 

WIND LOAD. 30 lbs. per sq. ft. on surface of two trusses and train 14 
feet high. 

TRACTION. 750 lbs. per lin. ft. of one track. 

32 



UNIT STRESSES (CARBON STEEL) 



Tension Main 
Trusses 

Suspenders and 
Members liable 
tc sudden load- 
ing 

Railway String- 
ers 

Floor beams & 
Highway string- 
ers 

Compression 
Members i n 
Main Trusses. . 



Live Load 



lbs. per 
sq. in. 

10,000 



7,000 
8,000 

9,000 
10,000-407 



Dead Load 
& Snow 



lbs. per 
sq. in. 

20,000 



14,000 
16,000 

1S,000 
20,000-80-, 



All Coexist- 
in k stresses 
except secon- 
dary strains 



lbs. per 
sq. in. 

20,000 



14,000 
16,000 

18,000 
20,000-80- 



All Coexist- 
inn stresses 
including se- 
condary strains 



lbs. per 
sq. in. 

22,000 



15,400 
17,600 

19,800 
22,000-88 J 



NICKEL STEEL increase Units given for Carbon Steel as follows: 

Tension 40% 

Compression & Pins 25% 

Plate XI. — Shows in outline all the designs submitted with the 
tenders. Official No. 1 was the only design exhibited at the time tenders 
were called, but it was later supplemented by Official No. V and tenders 
were asked on various modifications of these two designs. 

The study of the Official Design and its erection had shown that in 
these great bridges the dominating factor of the design roust be the practi- 
cability and safety of its erection, and this was emphasized in the develop- 
ment of our alternative designs. Economy of material is also an important 
factor because, in addition to its first cost, the effect of unnecessary weight 
may be cumulative, considerably increasing the sections of the heaviest 
members which are the measure of the shop equipment and of the erection 
appliances. 

It was known that long panels with diagonals approximating an incli- 
nation of 45 degrees to the horizontal would give the most economical 
form of Web System, but a diagonal inclined at 45 degrees would require 
that the deep panels near the pier should be about 200 feet long. It was 
seen at once that these long panels could not be constructed by the erection 
methods we had decided to adopt, and that some compromise was 
necessary either by shortening the panels or adopting a system of sub- 
trussing. 

33 



A consideration of the Forth Bridge construction will perhaps assist 
in the further discussion of the other designs. 

A length of the Forth Bridge equal to the projected Bridge at Quebec 
weighed about 35,000 tons — all carbon steel. The Official Design was 
estimated to weigh 72,700 tons, 52,000 tons of which was nickel steel. 
The Forth Bridge had successfully performed its duties for some 28 years 
and we were naturally not unmindful of this precedent — indeed, serious 
consideration was given to its design, and to the possibilities of adopting all 
or some of the special features giving it its remarkable economy in weight. 

Referring to Drawings Nos. 2 and 7 it will be seen that the Forth 
Bridge differs in several particulars from the design of any of the other 
great bridges, and from any of the designs prepared for the Quebec Bridge. 

The compression members are all circular in section and of large diam- 
eter. The circular section gives the largest radius of gyration for the 
metal employed, but this in itself was not an important factor as nearly 
every member in the bridge required sufficient area to give an effective 
radius with any of the ordinary forms of cross section. It results, however, 
in economy in the details because latticebars, tie-plates and other material 
which does not carry direct stress but is merely introduced to stiffen the 
rectangular member may be almost entirely omitted. 

The trusses of all the bridges illustrated, including those of the designs 
submitted with tenders, lie in vertical parallel planes, while the trusses 
of the Forth Bridge are battered from a width of 120 feet at the shoes to 
33 feet at the top. 

There is important economy in this construction. The floor beams 
may be made only long enough to accommodate the traffic; the weight of 
the top lateral bracing and sway bracing is very much reduced and the 
section of the compression chords is also reduced. The total stress in 
the first panel of the bottom chord of the Quebec Bridge is about 15,000 
tons, requiring a section of 1628 square inches; of this stress 4,000 tons 
is due to wind, requiring a section of 450 sq. inches, and had it been prac- 
ticable to increase the width at the shoes to 120 feet, as in the Forth Bridge, 
the wind stress woidd have been reduced by about one-quarter and the 
section of the chord by 112 sq. inches. 

The drawings do not show it, but if a horizontal section were taken 
at the floor level it would be seen that this section of all other bridges is 
rectangular, while in the Forth Bridge, the cantilevers taper towards 
each other, until at their ends they are only 32 feet wide. By making the 
suspended span only wide enough to accommodate the traffic and for its 
own lateral stability considered as a simple span, much economy is realized 
due to the shortening of the lateral bracing and floor beams. As already 
pointed out, any saving in the weight of this portion of the bridge is multi- 
plied several times in the weight of the cantilevers. 

34 



The diagonals have an effective inclination throughout the trusses. 

Members of such large dimensions as those used at the Forth could 
not be fabricated in shops at a distance, transported to the site and erected, 
and in the case of the Forth Bridge the manufacturing plant was actually 
built at the site of the Bridge. 

Moreover, many of the details for connecting the circular compression 
web members were of necessity so complicated that it would have been 
next to impossible to manufacture them, except by the method of laying 
out each piece as the work was built up. The shapes and plates entering 
into the large members of the Forth Bridge were therefore marked off, 
bent, fitted and drilled or punched at the site, and the work was erected 
plate by plate and section by section — the whole system of laying off, 
fitting up, and riveting, being closely analagous to steel shipbuilding. 

This method of erection not calling for the handling of heavy pieces 
permitted each large member to be projected out plate by plate, carrying 
itself as a cantilever, until it reached a point of intersection with some other 
member, where they would give mutual support, or where in the case of a 
chord, a tie or strut could support the member from the intersection of 
the diagonals. 

To carry out the work in the case of the Forth Bridge a large force of 
men was required, the number being at times over four thousand. The 
work proceeded continuously throughout the year for over four j - ears. 
There were plenty of men trained in steel shipbuilding on the Clyde, 
and climatic conditions permitted the work to be carried on at all seasons 
of the year. 

A sufficient number of skilled hand mechanics was not available in 
Canada, it is doubtful if they could have been assembled even for continuous 
work, and when it is realized that erection work at Quebec could only be 
carried on for some six months in the year, it will be seen that it was quite 
impracticable to adopt this type of construction. 

We therefore decided early that the only practical method of erecting 
the bridge in a reasonable time would be so to design it that the field force 
would be a minimum. This necessitated manufacturing the members into 
as large pieces as could be transported and erected, and in order that these 
pieces might be manufactured economically it was necessary to install 
special plant capable of handling unusually large members, and requiring 
special machinery and mechanical appliances. 

The use of this machinery also reduced the manufacturing force to 
the smallest possible limit. Our shop forces at no time exceeded five hund- 
red men, and our field force working at the same time on the cantilever 
arms on both sides of the River, and on the erection of the suspended 
span at Sillery Cove, did not exceed, exclusive of field painters, 500 men, 

35 



or about 1,000 all told for shop and erection. The average field force 
when only one cantilever was being erected, or before the erection of the 
suspended span, was very much less. 

Although it was thus found impracticable to adopt circular compression 
sections or the field construction of the Forth Bridge, consideration was 
given to adapting the method of construction we proposed to battered 
trusses and to vertical trusses tapering in plan. 

It will be realized from a consideration of the erection traveller and 
erection methods discussed, that members of the dimensions actually 
handled, or of any practicable size, could not be erected in an inclined 
plane without a very large amount of temporary supporting material,' 
and that the field connections of inclined members would be most difficult. 

With regard to the plan of vertical trussses with cantilever arms 
tapering in a horizontal plane, this also had ultimately to be abandoned 
on account of the difficulties of erection. A top chord traveller would 
have been impossible, due to the constant change of gauge of the tracks 
on which it would run and a through traveller could not be made wide 
enough for stability. Indeed, the only way of erecting either plan, seemed 
to be by means of an outside traveller supported on temporary cantilever 
beams such as was used on the Phoenix Bridge. A few preliminary 
sketches and estimates quickly demonstrated that the cost of erection 
and temporary material would far out-weigh the saving in weight to be 
gained by the narrower centre span. Even had this not been the case, 
the bent detail for connecting the members where the planes of the tapered 
trusses intersected, offered many difficulties, and it is doubtful if it could 
have been made with sufficient accuracy to insure the calculated distribu- 
tion of stress throughout the intersecting members. 

In a cantilever bridge of large dimensions the suspended span is the 
first element to be considered, as it may be designed and estimated without 
regard to the rest of the structure except as to its width, while the weight 
of the suspended span and the method of its erection must be considered 
in designing the trusses of the cantilever and anchor arms. 

There was sufficient data on long simple spans to permit an economical 
arrangement of panels, outline of trusses and form of bracing to be readily 
chosen, and little trial designing was necessary here. It was desirable, 
however, to determine the method of placing the suspended span in position 
before fixing upon its length, its depth at the ends of the trusses or the outline 
of the trusses. If it were determined to erect it on falsework and float it 
into position, the depth at the ends would be kept low enough to give only 
sufficient head room for efficient portal bracing, so that the form of the 
chords might approximate to a parabola which was known by experience 
to give much the most economical outline for very long simple spans. 

36 



If it were to be erected by cantilevering out from each side and joining in 
the centre, the end depth would of necessity be increased to something 
approximating the depth at the centre, otherwise it would be difficult to 
provide material for the heavy moment caused by the weight of the canti- 
levered portion, together with the weight of the erection travellers and 
material at the ends of the half span cantilevers, when the final connection 
was being made. 

Even with the greatest practicable depth at the ends of the trusses, 
the cantilevering method of erecting the centre span called for a very 
considerable increase of material in the cantilevers and the span to provide 
for the erection stresses, and this in itself seemed a sufficient reason for 
early determining to float the centre span into position, thus permitting 
the most economical outline to be adopted. 

There were, however, other considerations in addition to that of econ- 
omy leading to this conclusion. When a suspended span is cantilevered 
out from each side, in order that the connections at the centre which 
convert it to an ordinary span may be made, the two halves must meet 
exactly in the centre and the upper and lower chords must be in alignment 
both vertically and horizontally. The length of the steel work varies with 
the temperature and from change of load, and the horizontal alignment 
may also vary from changes in temperature and wind loads. Adjustments 
must therefore be provided where the span connects to the cantilever in 
order that these conditions may be met. Satisfactory appliances for these 
adjustments have been devised, and successfully used on all the great 
cantilever bridges heretofore built, the centre spans of these having been 
erected by cantilevering, but in some instances the final connection at the 
centre of the span was only made after great trouble and with some risk. 
The element of danger has lain in the difficulty of dealing sufficiently 
rapidly with the existing heavy forces, when making the delicate adjust- 
ments necessary to follow the changes in length and alignment that take 
place through the variations of temperature. Furthermore, when the 
centre connection is made, if the chord connections to the cantilever arm 
are not immediately released a change of temperature may cause the centre 
span to so connect up the cantilevers as to complete an arch from pier 
to pier, or at the other extreme to develop an undue amount of tension 
in the top chords. Tentative designs for the adjustment mechanism 
required for the very heavy stresses found here, showed that it would be 
cumbersome.and costly. 

. Having decided to float the centre span into position, its length 
became a question of balancing the increased difficulty of floating and 
placing a very long span against the saving in weight that might be effected 
by approaching the theoretical length for economy. 

Theoretically, the least weight of material in the finished structure 
would result from making the suspended span about 1103 feet long, but 

37 



this is a quite impracticable length for the span itself, and 668 feet was 
the longest simple span hitherto built. The Official Design divided the 
total span into three equal portions, making the cantilever arms and the 
suspended span each 586 feet. This arrangement was adopted as it 
fairly balanced the above considerations, and it facilitated the use of the 
official data in the new designs. 

The types of trusses from which to choose may be enumerated as 
follows: 

(1) The Official Design, Figures 1 and 2, Plate XI a Warren 
truss with each main panel divided in two and floor stringers of 
moderate length, forcing an uneconomical inclination to the diagonals. 

(2) "M" Design, Figure 5; a Warren truss with each main panel divided 

in four, permitting the most economical inclination of the web members 
to be chosen while retaining short stringer lengths. 

(3) A Warren truss having the diagonals at a favorable inclination without 

redundant members to support the top chord, each long panel being 
sub-divided in the middle to give a- stringer length about twice that 
adopted for any of the other systems. See Figure 6. 

(4) Double intersection trusses with and without sub-divisions. 

(5) The "K" form of bracing. 

The Phoenix Bridge with all diagonals in tension and sub-divided to 
give two stringer panels in each main panel, gained a favorable inclination 
of the diagonals and an economical form through the curvature of the 
bottom chord, but this form of truss was naturally not considered. 

It was at once apparent that type No. 3 would give the least weight 
in the finished structure and have a pleasing appearance, but unless it 
were fabricated and erected in the same manner as the Forth Bridge, 
there seemed so many obstacles in the way of erecting it with safety and 
at reasonable expense, that merely outline sketches were prepared and 
serious consideration was not given to this type. 

Some preliminary designs were made for double intersection trusses, 
but although the stresses and sections of the web members are only 
half those of the single intersection and the web members thus lighter 
and easier to handle, there seemed no way of getting favorable inclination 
of the web members without encountering similar erection difficulties 
to those met with in the single intersection design. A tentative design 
had been made by the Board on this system, but Mr. Vautelet had not 
found advantages in it to compensate for the stresses being statically 
indeterminate, and he was strongly opposed to its use. 

Mr. Johnson and I early pinned our faith to the "K" form of bracing, 
but some of our associates wished to compare the merits of a bridge with 

38 



B more conventional and more economical arrangement of bracing by 
preparing a complete design. Mr. Emil Larsson, Assistant Chief Engineer 
of the American Bridge Company, was asked to make this design, which 
is illustrated on Plate XI as "M-N". 

Anticipating the more detailed discussion of the "M" design which 
will follow, it was at once realize 1 that while it would make a very economical 
arrangement of the members actually carrying stresses there would be 
objections to its employment, in that the sub-truss system would probably 
introduce heavy secondary stresses from local panel loads; the top chord 
must be supported from the panel points as in the Official Design by 
long vertical members not useful for carrying live loads. During erection, 
very heavy temporary members would be necessary to hold up the inclined 
struts and the bottom chords until the main panel points were reached. 

Plate XII. — Is a general elevation of the "K" truss design submitted 
with the tenders. 

The "K" form of bracing is quite different from the trusses of any 
of the other great cantilever bridges or from any form of trussing that 
had been previously illustrated in British or American practice. 

It may be likened to a double intersection Warren truss with every 
half panel reversed, and a vertical member inserted, thus retaining the 
advantages of the double intersection truss in halving the shears between 
two members in each panel. It is statically determinate as to stresses, 
the shear being positively divided at each panel point, whereas in the 
double intersection truss, however great the care in calculating the stresses, 
one system or the other may accumulate more than its share of stress 
due to errors in manufacture and erection. The top chord length of 
the double intersection truss is halved without redundant members. 
It will be demonstrated later that no temporary work is required in erection 
and that it is much the safest and easiest form of trussing to erect. 

Most of the difficulties anticipated in manufacturing and erecting 
the Official Design had their principal source in the steep and uneconomical 
inclination of the web members, the large section of these members, and 
the necessity of either having the erection traveller extend out over two 
panels or having a very large quantity of temporary holding-up material, 
expensive to supply and very difficult to place and adjust with the required 
nicety. With the "K" form of truss the main panel could be halved 
for the same angle of web bracing; or, conversely, for the full panel the 
angle of the bracing would be nearly twice as favorable. The early 
sketches were for the full panel of 84 feet, and for several lengths down 
to panels of 65 feet long, all to be erected by a top chord traveller. 

The "K" truss design lent itself to the use of a top chord traveller, 
the normal stresses from live and dead load in the vertical posts requiring 
sufficient section to carry the erection stresses imposed by the traveller, 

39 



and the cost of the erection equipment for this arrangemeut would have 
been comparatively small. The long panels, however, required such 
heavy floor beams that there were difficulties in the way of manufacturing, 
transporting and erecting them. The stringers were also heavy, and 
the general arrangement did not realize the economy expected. More- 
over, the risk of using a top chord traveller with two panels reach was 
not in any way abated. 

Deformation diagrams were made for the sub-panel now used. It 
was found that this form of sub-trussing did not introduce objectionable 
secondary stresses, and did not interfere with fully splicing the joints 
of the compression members as the erection proceeded. Altogether 
this arrangement had everything to recommend it, and it was adopted. 

The trial of the sub-panel was initiated in the endeavour to use a 
through traveller which would avoid the risks attendant on the use of 
the top chord traveller. The through traveller is necessarily much heavier 
than the top traveller, and one to reach 84 feet or even 65 feet, was found 
quite impracticable. There was no doubt, however, that a traveller 
could easily be designed of moderate weight that would place members 
of the size to be handled at a distance of about 42 feet from its forward 
point of support. The early designs for this traveller were made with 
four swinging booms each capable of lifting the heaviest member to be 
handled and swung by guy-tackles from a horizontal frame-work attached 
to the top of the traveller. As the erection was worked out in detail, it 
was found that the use of the booms would not be entirely satisfactory, 
and we were led to adopt the travelling cranes, notwithstanding that 
this arrangement added much to the cost of the equipment. 

Touching on the outline of the whole structure, — the length and 
outline of the centre span had early been determined. The maximum 
height of the cantilever arms over the main piers was fixed by the specifica- 
tion. A lower height was undesirable, because even if it resulted in 
economy of total material, it would require heavier chord members near 
the piers and these large members, as already pointed out, fixed the size 
and capacity of the erection traveller and lifting devices as well as the 
cranes and many shop tools. 

Economy demanded that the end height of the cantilevers should 
be as low as practicable provided the end post from which the centre span 
is suspended should have an economical inclination approaching 45 degrees. 
The outline of the bottom chord of the bridge was also fixed by the specified 
height of the central portion to give clearance for navigation, and the 
requirement of straight chords between this central portion and the main 
shoes. 

For the sake of appearance, repetition of shop work and other reasons, 
it was desirable that the anchor and cantilever arms should have the same 
panel lengths and form of trussing, but it was found economical to shorten 

40 



the anchor arms by one full panel. There would probably have been 
economy of material in a still shorter anchor arm, but it would have 
increased the already heavy load on the main piers, 502 feet was the shortest 
anchor arm of the Official Designs, and a shorter design was not tried. 

Plates XIII and XIV. — Show the general arrangement of the 
traveller used for erecting the Quebec Bridge. It was equipped with two 
travelling cranes, each having two trolleys of 60 tons capacity, that 
could be run out far enough to lift the outside members of the trusses, 
and auxiliary hoists at the ends of each crane of 7 tons capacity. The 
four derrick booms each had a capacity of 15 tons. The total weight 
of the traveller and rigging was about 940 tons, and with both cranes 
run to the rear position for moving, the load was almost equally 
distributed on the front and rear trucks. When lifting the maximum 
loads the reaction at the front points of support was about 1300 tons. 

The top of the steel work of the traveller was about 210 feet above 
the rail on which it ran. The crane runways were about 140 feet and 
permitted a crane to work 49 feet in advance of the point of support. 

The above short description of the traveller is given here to convey 
an idea of its size, and the importance of this portion of the field equipment. 
It will be fully described in the forthcoming paper. 

After planning our erection for the "K" truss we found that this 
form of traveller could be advantageously applied to Official Design V 
and we estimated on using it for erecting that design. 

Figure 2. — Shows the method proposed for using this traveller in 
erecting Official Design V. It will be seen that as the traveller is moved 
out from floor beam to floor beam, it is necessary to support it, and a 
portion of the bridge, by means of temporary ties until it has reached a 
position where it stands on the fourth floor beam away from the shoe 
where the unit of the main truss system is completed, and the same 
condition obtains for the succeeding four floor beams. It will also be 
noted that until its rear portion has moved past the top of the inclined 
strut, the permanent sway bracing cannot be placed between the 
compression struts, and a large portion of the bridge is without 
transverse bracing above the floor system. 

To understand this we shall have to refer to the system of wind 
bracing adopted. It has been usual to place a system of lateral bracing 
in the planes of both the top and bottom chords to provide for wind loads 
and vibrations, and to keep the compression members in line. The 
stresses, or part of them, that go through the lateral bracing in the plane 
of the top chord are then considered as being carried to the piers by a 
system of sway bracing between the main posts. 

41 



This arrangement results in some ambiguity of stress, and the Board 
properly decided that the more correct and economical method would 
be by means of sway bracing between the compression members, to transfer 
all horizontal loads to the bottom chord, and thus carry all the horizontal 
loads through the lateral bracing of the bottom chord to the piers where 
they must eventually be resisted. 

This system of top lateral and sway bracing, excellent in itself, was 
not well adapted to a through traveller as in several positions of the 
traveller the sway bracing would have to be omitted from between the 
inclined struts until the traveller had moved out clear of them and there 
would thus be two panels of truss without permanent lateral bracing. 

Figures 3 and 4. — Show the traveller when erecting the "K" truss. 
It will be noted that in all positions of the traveller, each floor beam when 
placed is carried by the permanent construction of the bridge and no 
temporary work is required except for holding up the bottom chord, and 
that, only because it is convenient to place the lower members first. If the 
diagonal members were placed first they would interfore with the tackle 
necessary for placing the lower chord. 

Another advantage in the erection of the "K" truss is that the lateral 
and sway bracing can be placed between the vertical posts as soon as 
the traveller has moved past one of these posts and, moreover, can be 
erected by the traveller itself, so that the structure as it is erected is braced 
and securely held against all lateral forces by the permanent construction. 

Figure 5. — Is a deformation diagram for both anchor and cantilever 
arms showing the framed form of the trusses. It is shown here to illus- 
trate the absence of sharp bends and the generally uniform deformation 
resulting from the tc K" form of bracing, an advantage that had not 
altogether been anticipated when the designs were started. It was early 
demonstrated, however, as it was necessary to make deformation diagrams 
for each stage of the erection, to determine whether there would be 
bends or undue distortions during erection. As a further illustration of 
the uniform deformation of the "K" truss. — 

Figures 6, 7 and 8 — Show the anchor arms of the Phoenix truss, 
the Official Design and the "K" truss. The distortions of the Phoenix 
truss are very marked, particularly those of the bottom chord. The lower 
chord of the Official Design has a uniform deflection, the long diagonal 
members are kept straight through the adjustment of the upper chord of 
the sub-truss before referred to, but there are marked kinks in the upper 
chord where the long vertical redundant members push it out of line. 

Figure 9. — Shows a comparison of the bottom chord deflections 
under dead plus live load for the three designs. 

42 




0fO?A£f en #*? to 






Figure 2 



43 




Dotted tints show mattrtal tv be erected 
with the traveller in the position shown. 



Figure 3 



44 



' 7~r.,.'/.r 







:_. . /f . "yM ^J 



Figure 4 



45 




ci * 
i. 

5 



5 » 

3' 



'1-1.! '•' "' 



Nljtls.j's'^-TtS 



5i2SiS?3C* >:'*: 



1145.?.? Iisi. 



•-■■■ |JW 



T:X. 



- fst tfttf 






PUS. 



liip 




I 



46 

















/ 


A 






/ 


\/ 


















/ \ 











I 














Figure 6 




Figure 7 



47 







Figure 8 



ST LAWRENCE BR/DGE C? DES/GN 



OFF/C/AL DES/GN 




Figure 9 



RHOEN/X BR/DGE C9 DES/GN 



Figure 10. — Shows the connections of the web members of the bottom 
chord at the first main panel point of the Official Design, and the con- 
nections for a corresponding panel point of the "K" design. The 
Diagram also shows sections of the chords and compression diagonals 
of the Official Design and the "K" design. 

There are many objectionable features in the details of the connections 
of the Official Design. The main compression members must be forked 
for a long distance in order to connect on pins in the center of the chord. 
The clearances of the long forked ends were necessarily small and called 
for great care in manufacture to permit the field connections to be made. 

The pins on which these members abut carry heavy moments, and 
are very large. The plates to which the tension diagonals connect are 
very deep, and the pin holes a long distance from the bottom of the chord. 

The pins in the centre of the chord prevented the use of centre 
diaphragms, and to provide the necessary lateral stability for each pair 
of webs they were connected by cover plates on the bottom flange with 
a row of lattice bars at the centre of the section and another row on the 
top flange. The section of the cover plate on the bottom was compensated 
for on top by narrow plates riveted to the flange of each web girder. 
While the area of the metal was thus properly distributed about the centre 
line, the section was unsymmetrical and required for lateral stability 
an unduly large proportion of lattice bars not carrying stress. 

In the "K" design the stresses in the web members are practically 
halved, as there are two diagonals in each panel to carry the vertical 
shear. Owing to this and the more favorable angle of inclination of the 
diagonals, it was possible to make all connections on pins outside the 
chords, and to keep the webs of the compression diagonals in the same plane 
as the connection plates, thus relieving the pins of all moments and making 
it necessary to provide pins only of sufficient size to carry the direct 
bearing. 

There were no pin holes in the centre line of the chord, and centre 
diaphragms were carried through between the outside webs, making an 
exactly symmetrical section that tested well, was easy to manufacture and 
handle, and considerably more economical in the weight of details than the 
Official Design. 

49 




50 



There was another advantage in the chords of the "K" truss not appar- 
ent from the sketches — namely, in the more uniform increment of load 
received at each panel point from the cantilever end to the shoe, permitting 
a more gradual increase of section, and for this reason better splices and 
better details throughout. This will be better understood when it is consi- 
dered that in the Official Design there is no change in chord stress, due to 
vertical loads, from the shoe to the first main panel point, approximately 
170 feet from the shoe : and, similarly, there is no change in the chord 
stress from this panel to the next main panel point, 170 feet further on. 
In the "K" design the main panel points occur at half the distance of those 
in the Official Design, and the increments of stress are correspondingly 
less and more often applied, so that the increase in the chord section at 
any panel point of the "K" design is only about half the increase at the main 
panel points of the Official Design. 

Reference has been made to the large number of vertical members in 
the Official Design, and in our "M" design, introduced simply for carrying 
the weight of the top chord. It will be seen that in the "K" truss every 
member of the truss carries its proportion of live as well as dead load 
and there are no redundant members. As shown on the deflection diagrams 
of the Official and Phoenix designs, the redundant members cause consider- 
able distortions of the frame, and the omission of these members which 
have no part in carrying the live load of the structure results in important 
economy. 

The question of appearance must be largely one of individual prefer- 
ence, and it is perhaps difficult to compare designs shown only on a flat 
elevation. The redundant members in the Official Design and the varying 
inclination of the chords of the sub-trusses and stiffening struts, seemed 
to us to detract from its appearance. We think the symmetry and evident 
purpose of every member in the "K" design gives it a certain appearance 
of fitness and dignity. 

When referring to the webs of the compression diagonals being practi- 
cally in the same plane as the webs of the chords, attention was not called 
to the fact that this gave extra spacing between the outside webs, permitting 
each web to have double angles, thus making it symmetrical and of stronger 
section than the channel form of web, and moreover, giving a much larger 
and sufficient radius of gyration for the columns as they were shipped 
from the shop and erected, making it unnecessary to place much dependence 
on the lacing and tie-plates field riveted between these members after 
their erection. 

Another important consideration from the purchasers view point 
is that, with the construction adopted, access for inspection and 
painting can be obtained to all parts of the interior of the chords and large 

51 



compression members; whereas, with the long forked ends proposed for 
the Official Design there were of necessity many narrow spaces that could 
not be reached after the Bridge was erected. 

The other comparisons have principally to do with the cost of material 
and operations in detail and cannot easily be set forth. We, however, 
estimated there were many economies in the "K" design. 

Plate XV. — Shows a general elevation of the "M" design. It will 
be seen that the design is pleasing in appearance; the detailed estimates 
of weight were, as expected, considerably lower than for any of the other 
designs, being only about 85% of the next lightest, and it may be asked 
why we did not make a stronger effort to obtain the contract on this design. 

Plate XVI. — Is a portion of the "M" design to a larger scale, showing 
the general character of the details and connections proposed. 

Figure 11. — Shows the detail of the first main panel point in 
the bottom chord compared with the corresponding panel point of the 
"K" design. 

In order to carry out the design logically and. in such a manner that 
it could be erected safely, it was necessary to introduce pins at all panel 
points in the compression members and the use of pins introduced many 
objectionable features which Mr. Vautelet and the Board of Engineers 
had been at great pains to avoid in the Official Design, views in which 
we entirely concurred. When the design was started it was not known 
how these details would work out — indeed, as already pointed out, the 
details of any new design for a bridge of this magnitude could only be work- 
ed up by carrying the design through by the tedious method of working 
from the centre span and the cantilever ends back to the piers. Having 
carried the design to this stage of completion, we thought it worth while 
putting in a tender as a matter of record, although we did not in any way 
urge its claims. 

It has been shown in discussing the open erection joints of the Phoenix 
truss, that a truss of this form could not be safely erected without providing 
for angular movement at every panel point of the compression member; 
hence the reason for the pins. It was doubtful, however, if ordinary 
pins could perform the intended function of permitting this angular move- 
ment because the movement is gradual as the load comes on and on some 
pins, when the cantilever arm has been extended out some distance, must be 
made under heavy stress. If movement were to be assured, it is probable, 
that the unit pressure on the pins would have to be much less than the 
bearing pressure allowed for the transmission of axial stress. This is pro- 
vided for by large sleeves at the shoe-pins of the Official Design and of the 
Bridge as built. 

52 



It will be seen that to provide pin bearing even for the stresses permitted 
by the specification, it was necessary to build up the chord webs of the "M" 
design to great thickness, and that in some instances all the reinforcing 
was placed on the outside of the web in order to make room for the long 
fork ends of the members connecting on the pin. This, which may be 
termed eccentric reinforcing, is of doubtful value even if the inner pin 
plates are made sufficiently long to assure that the rivets all act in shear. 

The necessity of stopping the diaphragms towards the ends of the chord 
joints, thus making fork ends on the chords as well as on the web members, 
also seemed a very undesirable detail. 

Added to the above considerations, we concurred with the Board of 
Engineers that these very large compression members should be as nearly 
as practicable continuous from end to end, so that the distribution of stress 
throughout the member should not be disturbed, and that the stresses 
would continue without being deflected past the splices and past the panel 
points where the member would receive a fresh increment of stress to be 
provided for by additional material. With pin joints it takes a great deal 
of reinforcing material to concentrate upon the pins the compression 
stresses carried by the outer edges of the web and the flanges, and these 
stresses must be again distributed over the section of the member after 
the joint is passed. 

A question that may be asked is why the "K" bracing was not carried 
through the suspended truss, so many advantages having been found in its 
use in the cantilever arms and the suspended truss itself being of such 
very considerable length. "K" bracing was designed for the suspended 
span, and had we determined to erect this as a cantilever it would no doubt 
have been used, but it does not lend itself well to the parabolic form of 
top chord or to the low end heights which it seemed economical to use for 
this truss and, considering the method of its erection and the deformation 
diagrams, there ssemed no objection to the form of truss used. 

The discussion heretofore has referred to the designs submitted with 
the tenders and principally to Design "B" of the St. Lawrence Bridge 
Company recommended by the Advisory Board of Engineers for accept- 
ance by the Government. That Design was for the specified lengh of 
1758 feet centre to centre of piers, and its outline elevation necessarily 
differs from that of the present Bridge which has a span of 1800 feet centre 
to centre of piers. 

Plate XVII. — Shows the general elevation of the Bridge as built. 

The masonry of the old Phoenix Bridge was left in perfect condition 
after the accident, but the piers were too short to carry the projected span, 
it having been determined to increase the 67 ft. width of the Phoenix Bridge 
to 88 feet in the new Bridge. Mr. Vautelet had intended to make use of 
the old foundation of the pier on the South side, increasing its dimensions, 

54 



however, both in width and length by sinking new caissons alongside, 
and to build a new pier on the North side of the River clear of and to the 
South of the old pier — the plan working out to a distance of 1758 feet 
centre to centre of piers with the centre line about 15 feet down stream 
from the centre line of the old bridge. 

It was finally concluded that there might be great difficulty in sinking 
new caissons alongside the old foundations of the South Pier — indeed, 
it was thought by some that the proposed construction was quite impracti- 
cable. After Mr. Vautelet had resigned, the Board, in conjunction with 
the Advisory Engineers, recognizing the risk of at least serious delay from 
this cause, made a number of studies for using the foundations of the old 
piers. These studies were based on building up new masonry on the old 
foundations of the maximum length that the foundation caissons would 
permit, and reducing the weight upon the foundations by carrying a portion 
of the load on new piers to be built to the South of each of the present piers, 
thus giving four points of support for each cantilever instead of two as in 
the Official and the present design. Studies were also made for sinking 
three new pedestals on each side of the River and making use of the old 
foundations for the fourth pedestal. 

Figure 12. — Shows the proposed design of the superstructure for 
one of those plans. 

It was not found practicable to make the piers long enough for the 
width of bridge without too great intensity of pressure on the ends of the 
caissons, and the three new pedestals did not promise much economy. 
The plan to make use of any part of the old masonry or the old foundations 
was finally abandoned, and the Board recommended sinking new piers 
to the South and entirely clear of the old foundations but on the same centre 
line, thus restoring the span to 1800 feet. 

In adjusting the "B" design to the new length of span the Advisory 
Board considered that the Bridge would have a better appearance if the 
low panels towards the ends of the cantilevers and of the suspended span 
were made shorter, in order that the inclination of the web members might 
be kept at about the same angle. In conference with the Board, the length 
of the suspended span was fixed at 640 feet and the cantilever arms at 580 
feet. The span was divided into four more panels, one being added to 
each cantilever arm and two in the suspended span. 

Mr. Vautelet had specified that the height of the steel work above 
the centre pier should not exceed 290 feet and had refused to permit 
any greater height. The Advisory Board saw no reason for this limitation 
and the inclined upper chords of the anchor and cantilever arms were con- 
tinued until they intersected over the centre pier, the Board considering 
that this arrangement gave a better appearance, as it undoubtedly does on 
paper, than the two panels of flat chord over the piers. The intersection 

55 



II 



1 I 



'-A 



fa 



of the chords at this point made it necessary to extend the vertical posts 
to carry the shears from these chords and, having adopted this vertical 
post to omit the horizontal tie holding the heads of the first compression 
web members and replace it by inclined tie-bars, thus carrying the "K" 
system of bracing from the inclined posts at the end of both cantilever 
and anchor arms throughout all the panels to the main pier. 

In making our competitive designs we were necessarily governed by 
considerations of economy, and endeavoured to adopt an outline that would 
result in the lightest structure consistent with the specifications and the 
other requirements sought, including good appearance. 

The increase in span called for a heavier structure and the additional 
panels, as well as the changes in the pier panel, and also made considerable 
additions to the weight, but the Board considered that in a monumental 
structure of this character the extra weight was justified by the improved 
appearance. Personally, I am not at all convinced that the more economical 
arrangement of the "B" design would not have given an equally good 
appearance in the completed structure had it been adopted. 

Plate XVIII. — Shows a cross section of the floor of the Official 
Design together with the live and dead loads above the railway stringers. 

A cross section of the floor and the leads on which our Design "X" 
was based. 

And a cross section of the floor of the bridge as built. 

Throughout the consideration of the construction of the Official Design 
and of the alternative designs, we were faced with the problem of manufac- 
turing and placing material having weight and dimensions far beyond 
any precedent, and it was constantly being forced on our attention that 
because of the span and the construction approaching the practicable 
limits it was of great importance to avoid any unnecessary weight either 
in live or in dead load. 

A consideration of the heavy construction required to carry the high- 
ways specified, and the increase in the weight of the structure arising 
therefrom led us to make some approximate figures for a bridge to carry 
railway traffic only. These were so encouraging that they were followed 
up by complete designs omitting the highways, with the exception of a 
side-walk, which it was thought should be retained for the sake of inspection 
and such pedestrians as might wish to cross the bridge. 

These designs were otherwise in strict accordance with the 
specification and were similar to the designs carrying highways. This 
plan saved 3100 lbs. per lin. ft. of superimposed load and 4,480 lbs. per Lin. 
ft. in the weight of the roadway. The saving in the structure itself aver- 
aged 11,900 lbs of steel per lin. ft. and the finished structure was estimated 
to weigh only about 75% of that designed to carry the highways in accord- 
ance with the Board's specification. 

57 



I — 




/7/6 "' 

■B' DES/GN 
ST LA WRENCE 



/372°' 

"X" DES/GN 
BRIDGE CS 



203/°" 

^FFJCJAL DES/GN 
A/°/ 
Figure 13 

Figure 13. — Shows sections of the bottom chords next the main 
pier of the Official Design, the "B" design and the "X" design without 
the highways. 

In addition to the economy in weight to be realized, there were many 
good reasons for advocating this change of design, as it brought the size 
of the members within more practicable limits for manufacture and erection 
and the small interests to be served by the highways seemed out of all 
proportion to the additional cost, and the risk and difficulties in construction 
that the highways as designed entailed. After the Advisory Board had 
recommended the acceptance of our tender for Design "B", designed to 
carry the full highway and electric car loads as specified and had modified 
the span and outline as above noted, we pointed out to them the economy 
and advantages to be gained by omitting the highways and, when the final 
contract was drawn, the specification and plans accompanying it omitted 
the provision for these highways. 

In preparing our alternative design without the highways, it was not 
considered advisable to risk undue criticism by making any further change 
than necessary and the highways were simply left off, leaving the tracks 
as spaced in the Official Design. It was manifest, however, that advantages 
would accrue by separating the tracks as widely as the clearance of the sway 
bracing and the torsion of the structure would permit, these advantages 
being reduced stresses in the floor beams, a better staying of the top flanges 
of the floor beams and easier provision for traction stresses. The Board 
of Engineers readily accepted this change and the tracks were moved out 
to a width of 32'6" centres. 

During the discussion of this change Mr. Monsarrat advocated a fur- 
ther change to the form of floor system used under his direction by the 
C. P. Ry. on some of its high viaducts. This floor, which is shown on 
Plate No. XVII consists in placing each track in a through plate girder 
bridge with steel stringers and floors beams of its own, and with exception- 
ally well braced and reinforced top flanges, so that in the event of derail- 
ment, the derailed rolling stock would be kept in this trough and have no 
opportunity to plunge down and wreck other members of the bridge. 

58 



Another change from the accepted design was in the use of eye-bars 
for the top chord. We followed the Official Design in the use of carbon 
and nickel steel, and tendered on all of our designs in this way, in order 
that our tenders should not be set aside through not complying with the 
requirements. We also gave alternative tenders on what we considered 
more economical proportions of carbon steel, but we tried to avoid using 
carbon steel where its use could be criticised on account of adding unne- 
cessary weight to the bridge. While we tendered in some of our designs 
on using nickel steel eye-bars, experiments up to that time had not shown 
these bars to be entirely reliable, the quotations for them were exceedingly 
high compared to other material, and in our "B" design we estimated on 
plate nickel steel chords with riveted connections. Some of the members 
of the Board preferred eye-bar top chords, and after the contract was 
awarded we obtained quotations for carbon steel eye-bars, prepared esti- 
mates comparing the weights and cost of these with the material shown 
on the contract drawing, and under a supplementary contract substituted 
the eye-bar top chords in the final design of the bridge. 

When discussing the changes in length of span and modifications of 
outline, the Advisory Board of Engineers also discussed some changes in 
the specification to bring its form more in accord with existing standard 
specifications without sensibly changing its requirements except, in the 
substitution of E-60 for E-50 locomotives; and when it was finally decided 
to build the bridge for railway traffic only, a new specification covering 
this condition and the other changes was drawn up. 

The specifications finally issued by the Board of Engineers provided 
for a double railway track, but for no highways, except two sidewalks 
five feet wide. They will be given in full in the forthcoming paper. 

The greatest care was taken to comply with the requirements of 
the specification, and to submit with our tenders all the information 
called for therein. Had there been no change in the length of span, our 
plans were ready for the preparation of shop drawings, but the changes 
already enumerated necessitated an entirely new set of computations 
and, while the details and the make-up of the members remained of the 
same type as originally designed, modifications were necessary in nearly 
every instance by reason of these changes. 

Before tendering it was necessary, for purposes of estimating, to make 
approximate plans of shops and equipment, the erection travellers, false- 
work and other equipment required in the field, so that we might assure 
ourselves of the practicability of the work and make close estimates of the 
cost of the whole. These sketches were, however, not working drawings 
and as soon as the contract was let it was necessary to organize for carrying 
out the new design, the preparation of shop drawings, the manufacture and 
the erection of the structure. We felt that great perfection and a new 
standard of shop work would be required, that there was so much to be 

59 



kj 

| 
I 

1 

i 






c 


^ 


i 


"J; 


Hi 


% 


£ 


II 


5 

6 


i 

Hi 


1 

k 
k 

* 



v, 
ft. 



e! 









Vi X OQ 



5 ^ 



■ - M 5 

(*. <$ ft; ft; ^ 



vj s ^ 

? ^ i 

S ft: $ 
ft;<3Uj 



wi .. 5 * Is 



*l 



^ IS Vl 



* * 



Sll 









* S 2 2 8 SI? S 51 



5 B 



St 



in 









ft 
fc5 



? L ^ 

Ve V (j 



lo'OQ 



done, so many unusual problems to consider and so short a time in which 
to do it, that we must employ the best engineering and manufacturing 
advice that could be obtained. 

Figure 14. — Shows the working organization of the St. Lawrence 
Bridge Company for the construction of the bridge. 

Mr. Johnson as President kept an engineering as well as a business 
supervision of his charge. 

Mr. McMath was available for consultation when required, all 
drawings were submitted to him and were examined by him personally. 
Mr. Willard Pope, Chief Engineer of the Canadian Bridge Company, 
rendered much valuable assistance, personally checking all strain sheets. 

Mr. Porter was put in immediate charge of the Engineering force 
as Construction Engineer. To him belongs a very large share of the 
credit for the conduct of the work. In addition to his assistance in prepar- 
ing the details for our tender designs, he acted as Resident Engineer 
throughout, overseeing the preparation of shop drawings and the detailed 
design of erection equipment at Montreal during the Winter, and during 
the summer residing at the site in responsible charge of the field operations. 
From the outset, the work under his direction was carried on so satisfac- 
torily that much more responsibility than generally falls to the position 
was placed upon him with the utmost confidence. 

The Canadian Bridge Company released Mr. W. P. Ladd, then 
Superintendent of their shops at Walkerville, as Works Manager of the 
new shops at Rockfield and it became his duty to take charge of the 
layout, purchase of equipment, organization and, indeed, everything 
in connection with the manufacture of the bridge and the erection equip- 
ment. The excellence of the shop work and the wonderful precision in 
the lengths and fitting of the various members have been commented on 
by every engineer who has viewed the work. This excellence contributed 
much to the facility with which the bridge was put together in the field, 
and we all realize that the work of Mr. Ladd and his staff was one of the 
most important factors in the operations. 

We felt that the erection, involving as it did the lifting and placing 
of the heaviest pieces heretofore handled, and by far the largest tonnage 
in one span, required the best experience in heavy work and the best 
expert advice that could be obtained, and we were fortunate in securing 
the services of. Mr. S. P. Mitchell as Consulting Engineer on erection, 
he being considered best qualified to supply the experience and advice 
needed. Mr. Mitchell devoted a large portion of his time to the considera- 
tion of the erection equipment, secured for us the services of Mr. W. B. 
Fortune, our ^General Superintendent of Erection, and assisted in 
organizing thejield force. 

61 



Where so many experienced men have been employed and given 
their best thought to the work over a period of about six years, it is 
impossible to particularize further, but it will no doubt be appreciated 
that many helpful suggestions and much assistance were received from 
the Assistant Engineer, Mr. Harkness, the Chief Draftsman Mr. Kerrigan 
•and his staff, the Superintendents and the Inspectors both in the shop 
and in the field. The accurate work of Mr. Jost, Mr. Burpee and the 
field engineers should not be overlooked. 

Reference must be made to our harmonious relations with the Board 
of Engineers and the assistance received through the hearty .co-operation 
of its Members. The experience and judgment of Mr. Modjeski, 
Mr. Charles MacDonald and the Advisory Engineers when discussing the 
changes in plan was of great value, and it goes without saying that any- 
thing with which Mr. Schneider had to do must bear some evidence of 
his knowledge and ripe judgment in all matters pertaining to steel construc- 
tion. Notwithstanding Clause 5 of the Specification which threw the 
entire responsibility of the design, material and construction upon the 
Contractor, the Board of Engineers organized an experienced and highly 
efficient staff of engineers and inspectors, and our work was 
much facilitated by the co-operation of the Board and its staff. Every 
stress sheet and every calculation was investigated and checked, and 
every detail was checked to the last rivet. Material was inspected at 
the mills, in the shop, and in the field. Workmanship was most care- 
fully inspected both in the shop and in the field, and all field engineering 
(lines, levels and measurements) was carefully checked. It was very 
reassuring to have this supervision and to feel that it was practically 
impossible for an error to escape unnoticed. 



62 



THE DESIGN, MANUFACTURE AND ERECTION 

OF THE SUPERSTRUCTURE 

of 

THE QUEBEC BRIDGE 

by 

PHELPS JOHNSON, M.E.I.G., G. H. DUGGAN, M.E.I.C., 
and GEORGE F. PORTER, M.E.I.G. 

THE project of the long span required for the Quebec Bridge has been 
before the engineering world for a long time, it being over twenty years 
since the specifications and call for tenders issued by the Quebec Bridge 
and Railway Company, under date of 1st September 1898, brought the 
bridge prominently before the profession. The subject was at once taken 
up with great interest as it was realized that the magnitude of the work 
would call for the solution of new and untried problems, conditions in 
Canada not permitting the methods of construction used at the Forth 
Bridge — the only other bridge that at all approached the Quebec Bridge 
in magnitude. 

Much has been written on the subject but it is scattered through the 
engineering papers of a long period, and the engineers responsible for the 
design and construction of the bridge as built feel that a description of 
the work should be on record in readily available form within the Transac- 
tions of the Engineering Institute of Canada, all the Engineers taking 
part in the construction being members of the Institute. 

There may seem some plagiarism in what follows, some of the des- 
cription and many of the drawings having been already published; but 
as all of the data for these articles was taken from the designs, drawings 
and notes prepared by the St. Lawrence Bridge Company, the Authors 
feel entitled to present it a new as original matter. 

An introductory paper was read before the Society on January 10th, 
1918, entitled " Notes on the work of the St. Lawrence Bridge Company 
in preparing the accepted design for the construction of the Quebec 
Bridge." That paper refers to the important features of the official design 
on which tenders were called and the difficulties forseen in erecting it; 
it discusses the design of very long span bridges in general terms, the 
advantages of the " K " bracing and outline the considerations leading 
to the design of the St. Lawrence Bridge Company recommended by an 
Advisory Board of Engineers for acceptance, and on which the contract 
was awarded. These subjects are therefore either not discussed or only 
touched on in this paper. 

63 



The introductory paper reviews the history of the undertaking; 
the disagreement among and the changes in the personnel of the Board 
of Engineers, the setting aside of the official design and the adoption of 
the Contractor's design. 

These events may be briefly summarized chronologically as follows: 
29th Aug. 1907. Bridge under construction by Phosnix Bridge Company 
failed. 

31st Aug. 1907. Government appointed a Royal Commission to inves- 
tigate the accident. 

17th Aug. 1908. Government appointed a Board of Engineers, to 
" prepare plans and construct the bridge on these 
plans." 

Mr. (Now Sir) Maurice FitzMaurice, C.M.G., M. 
Inst. C.E. 

Mr. Ralph Modjeski,.M. Am. Soc.C.E. 

Mr. H. E. Vautelet, M.Can.Soc.C.E., Chairman and 
Chief Engineer. 

Mr. Vautelet completed plans and specifications about 
the 1st of January 1910, but the other members 
of the Board did not fully approve of his design, 
believing that a more practicable design could 
be made, and consented to tenders being called 
upon the official design only on condition that 
bidders might submit tenders on their own plans if 
they so desired. 

1st June 1910. Mr. FitzMaurice resigned. 

17th June 1910. Advertisement inviting tenders on Mr. Vautelet's 
official design and on Contractor's alternative 
designs. 

28th Sept. 1910. Mr. Chas MacDonald, Past Pres. Am.Soc.C.E., appointed 
to succeed Mr. FitzMaurice. 

1st Oct. 1910. Tenders received and opened by the Government. 

20th Jan. 1911. The Government appointed M. J. Butler, C.M.G., Past 
President, Can.Soc.C.E., and H. W. Hodge, M. Am. 
Soc. C.E., to assist the Board in selecting a design. 

8th Feb. 1911. Messes. Modjeski, MacDonald, Butler and Hodge 
reported in favor of the St. Lawrence Bridge 
Company's design. Mr. Vautelet dissenting. 
64 



28th Feb. 1911. Mr. Vautelet resigned from the Board. 

The remaining members of the Board then changed 
the span from 1758 feet to 1800 feet, and the masonry 
plans, modified the specifications and, in conjunc- 
tion with the Contractors, modified the design but 
retained all the essential features of the design 
submitted with the tender. 

4th April 1911. The Contract was signed, the design for the bridge was 
signed as part of the Contract, as were also the 
specifications. 

When the Contract was signed Mr. MacDonald desired 
to be relieved of further detail work and Messrs. 
Butler and Hodge ceased to be members of the 
Board by the terms of their appointment, thus 
leaving Mr. Modjeski the sole remaining member 
of the Board that had decided on the plans and 
specifications. 

It is pertinent to note that after this date when the contract was signed, 
no change could be made in the design or specifications except by agreement 
between the Contractors and the Board of Engineers. The calculations, 
all the working drawings and all the special methods employed in the 
construction were entirely developed by the staff of the St. Lawrence 
Bridge Company — the Board of Engineers thereafter acting in a super- 
vising capacity only. 

6th May 1911. Mr. C. X. Monsarrat, M.Can.Soc.C.E. was appointed 
to the position made vacant b} r Mr. Yautelet's 
resignation. 

loth May 1911. Mr. C. C. Schneider, Past Pres. Am.Soc.C.E. was 
appointed to complete the Board of three. 

8th Jan. 1916. Mr. C. C. Schneider died. 

Feb. 1916. Mr. H. P. Borden, who had been on the Engineering 
Staff of the Board from the beginning was appointed 
to fill the vacancy created by the death of Mr. 
Schneider. 

Reference was made in the earlier paper to the exceptional responsibil- 
ities for the DESIGN and ENGINEERING of the structure placed 
upon the Contractors in the clause of the Contract which required the 
Contractor. — 

" To GUARANTEE the satisfactory erection and completion of 
" the Bridge and to undertake the ENTIRE RESPONSIBILITY not 

65 



" only for the materials and construction of the bridge but also for 
"the DESIGN, CALCULATIONS, PLANS and SPECIFICA- 
" TIONS and for the sufficiency of the bridge for the loads therein 
"specified". 

The Contractors welcomed this responsibility, having every con- 
fidence in their design and plans for construction and being thus relieved 
by the terms of the contract from the possibility of being asked to 
adopt plans or methods of construction of which they did not approve. 

Reference was also made to the organization of the St. Lawrence 
Bridge Company. This Company was specially incorporated in the joint 
interests of the Dominion Bridge Company and The Canadian Bridge 
Company to combine their organizations and resources for the execution 
of the work. The working staff of the St. Lawrence Bridge Company 
is shown on Fig. 14. Further introduction is believed unnecessary. 

The present paper describes the bridge as built and the methods of 
construction. 

The paper is prefaced with a map of the vicinity, of the site, notes 
of the River conditions, tides, currents, etc., all of which may assist in 
a better appreciation of the conditions governing the delivery and storage 
of material and the erection as well as of the decision to make no provision 
for electric railways independent of the steam railway tracks or for vehicle 
traffic. 

It has been stated that the principal consideration leading to the 
adoption of the " K " form of truss bracing was the difficulty of erecting 
all other forms tried and the comparative safety and facility of erecting 
the " K " form. The problem of placing and connecting the important 
members of the structure was constantly before the Designers and, even 
after the general form of the truss and the make-up of the members were 
determined, the erection of each member and its field connections were 
always carefully considered in making the working drawings. It is 
thought, therefore, that it will facilitate an understanding of the design 
and construction to preface the detailed description of the work with a 
general outline of the methods of erection adopted, giving a short argument 
for their adoption and for the elimination of other methods considered. 
Reference is also made to the traveller and other appliances so far as this 
equipment affected the stresses and the details of construction. A more 
extended description of all important erection equipment is given in the 
chapter on erection. 

After the outline of erection methods there follow plans illustrating 
the completed structure with a description of the general design and 
details. 

66 



The manufacture and handling of the large members entering into 
the structure was beyond the capacity of any equipment existing in Canada 
at the time the work was undertaken, and it was necessary to build and 
equip special shops. The shops and their equipment are briefly described 
as well as some of the methods employed in the manufacture. 

The chapter on erection touches on the general accommodations for 
the field forces, the preparations for storage and equipment for handling 
material at the site. It describes the equipment and the method of erect ing 
the important members in the structure. While the latter description 
is tedious, it is felt that the record may be interesting to some — safe and 
easy erection having been the motif of the Designers in preparing the plans 
for the bridge. 

Summarizing the above outline, the following is a brief index to the 
several chapters: 

I General description of the site and conditions. 

II Outline of the methods of erection and transportation with reference 

to their effect upon the design of the superstructure. 

III Description of the structure as built, including the design. 

IV Description of the shops and equipment with reference to the unusual 

methods of manufacture. 

V Field operations and erection. 

(a) Camp and storage at site; 

(b) Erection Travellers; 

(c) Erection of approach spans; 

(d) Steel falsework and staging; 

(e) Erection of the north anchor arm; 

(f) Erection of main posts; 

(g) Erection of the cantilever arm; 

(h) Erection of the suspended span at Sillery; 

(i) Equipment for floating, mooring and hoisting the suspended span; 

(k) Floating and hoisting the suspended span in 1916; 

(1) Hoisting and coupling in place of the suspended span in 1917; 

(m) Final operations to complete; 

(n) Erection details; 

(o) Accuracy of work; 

Several subjects which may be found useful for reference which are 
not necessary to an understanding of the structure as constructed, are 
treated in appendices. 

APPENDIX "A" 

In discussing the broad considerations which led to the adoption of 
many features of the present design, reference was made to tests carried out 

67 



by the Royal Commission and by the Board of Engineers. Additional tests 
were called for in the original specification and the Engineers of the St. 
Lawrence Bridge Company made still further tests for their own inform- 
ation and guidance. The deductions from many of these tests find their 
expression in the specifications and in the design of the compression mem- 
bers; those relating to end details of tension members and riveted tension 
splices were carefully considered in the design of the details. A summary 
of these tests is given in Appendix " A ". 

APPENDIX " B ". 

The calculations for both primary and secondary stresses, being of 
a comparatively simple character, did not require the development of any 
new methods and naturally followed old and well established practice. 
It is thought, however, that the calculations were well arranged both for 
computation and reference and that it may be of interest to have a record 
of how these were carried out. 

Mr. A. L. Harkness has written a memorandum on this subject. 



Chapter I 

GENERAL DESCRIPTION OF THE SITE AND CONDITIONS 

Figure 15. — Is a photograph of the completed bridge. 

The Quebec Bridge is now a link in the Canadian National 
Railway, which saves about 219 miles between Moncton and Winnipeg; 
the distance from Levis to Winnipeg by the N. T. Ry. being 1355 miles, 
while the distance via Montreal and the Ottawa River is 1574 miles. 

Plate XIX. — Is a map of the vicinity of the Bridge site showing the 
connections of the railway lines crossing the bridge to the main lines of 
the C.N. Railway and the branches leading into the Cities of Quebec 
and Levis. It also shows the main highways in the vicinity of the bridge. 

Plate XCIX. — Is a chart of the River from Sillery Cove on the East 
to a short distance above the Bridge on the West. 

Plate XX. — Is a general elevation of the Bridge showing the profile 
of the River at the crossing. On the North side steep rocky cliffs extend 
for some distance on either side of the Bridge with only a very narrow 
beach at high water, and for about a third of a mile to the East of the 
Bridge high water comes to the base of the cliffs, cutting off the road connec- 
tion. The shore between high and low water is flat and covered with 
large boulders. At about the point of low water, or a short distance 
outside, the bottom falls away rapidly. On the South side the conditions 
are somewhat different, the beach being much wider. The South main 
pier is dry at all stages of the tide, while on the North side there is about 
25 feet on the outside of the pier at low water. 

The City of Quebec is situated at the East end of a high narrow spur 
which is almost an island, the deep valley of the Cap Rouge River at this 
point cutting it off from the high banks to the Westward. The St. Charles 
River flows into the St. Lawrence just below the City of Quebec after 
traversing a course almost parallel to the St. Lawrence and the valley of 
the St. Charles joins that of the Cap Rouge River at a comparatively low 
elevation. 

The main line of the Canadian National Railway crosses the Cap 
Rouge River on a high single track trestle 3,335 feet long, at an elevation 
of 154 ft. above mean high water. The railway then skirts the side hill 
of the Quebec Cape at almost a level grade until it reaches a point just to 
the East of the bridge where it begins to fall to Sillery Point, and from 
that point maintains its level to the projected terminal just below the 
Terrace at Quebec. 

69 




70 



It may be noted that at the time the contract was let it was intended 
to establish the Quebec Terminals at this point, bul subsequently arrange- 
ments were made witli the Canadian Pacific Railway for joint use of the 
C.P.R. terminals on the St. Charles River, the connection being made up 
the Valley of the Cap Rouge River around the West end of the Quebec 
promontory. 

A car ferry service during the construction of the Bridge was also 
established from a wharf about half a mile West of the end of the track 
opposite the Quebec Battlefields to the Intercolonial Railway wharf at 

Levis. 

The Quebec Bridge is nearly at right angles to the line of railway 
described above, and is connected to it by curves in both directions forming 
a " Y " in which the material storage yards were established. This " Y " 
and the yards were on a wide bench about rail level, but there was a high 
narrow rock ridge near the edge of the cliff which had to be cut through 
to give entrance to the bridge. The bench to the Westward of the Bridge 
was levelled down with the intention of establishing a large railway yard 
and shops. A roundhouse and much of the yard was put down but the 
repair shops were later established on the St. Charles River at St. Malo. 
On the South side the rail level at the abutment is considerably above 
the bank and the railway is carried for some distance on a high embankment 
through which passes the shore road leading to Levis. 

The Cities of Quebec and Levis are connected by an excellent ferry 
service, and ferries run occasionally from Sillery Point to New Liverpool. 
The shore road at the South end of the bridge was comparatively unimpor- 
tant, the greater part of the highway traffic into Levis being from the roads 
leading from a more Southerly direction. On the North side the St. Louis 
Road^the main highway opposite the bridge, has little traffic originating 
to the West of the Bridge, owing to the steep grade up the bank of the Cap 
Rouge River and practically all of the traffic is destined for Quebec. 
Connections for highways on the bridge to the road on the South side of 
the River could have been readily made, but the connections on the 
North side presented a good many serious difficulties; the deep rock cut 
at the entrance to the bridge, the crossing of the railway yards and the 
connection to the St. Louis Road, which at this point is at a considerably 
higher elevation than the bridge floor. 

The above considerations coupled with the greatly enhanced cost 
of the superstructure if highways were provided led to the decision to omit 
the highways. 

The tides at Quebec are heavy. The normal Neap tides have a range 
of about 12 feet and the normal Spring tides of about 18 feet. Occasional 
tides have a larger range up to 21 feet. In ordinary conditions of weather 
the heights of low and high tide as well as the periods can be very closely 
predicted from the Tide Tables. The average period of flood tide is about 

71 



? * * 




72 



five hours and of ebb tide about seven hours. The high tide begins to 
fall about an hour before the change in direction of current and the low 
tide begins to rise also about an hour before the current changes in direction. 
At the Bridge site serious cross currents often occur, and the River men 
state that these are most troublesome at the end of the flood tide. 

Figure 16. — Shows a series of tide curves, plotted from the Tide 
Tables, about the period selected for floating the suspended span in 1916. 
It well illustrates the general tide conditions. 

To the West of the Bridge site there is a straight reach of the River 
from one to three miles wide for about twenty miles; the high banks continue 
on either side of this reach. To the East of the Bridge site the River is 
narrower and not so straight and the site is comparatively sheltered from 
the East, but from both directions a heavy sea makes at the bridge crossing, 
particularly when the tidal current is against the wind. Owing to the 
geography of the River outlined above, there is no shelter from a wind 
up or down the River, but strong cross winds are very infrequent. 

The Board of Engineers established a wind gauge with an area of 
37.2 sq. ft., 54 feet above mean high water on the Contractor's wharf 
near the North Pier, but it was rather too low to give accurate records 
of the pressures that would occur at a higher elevation against the structure, 
and as the anchor arm was built it was to some extent sheltered. Its use 
was therefore practically abandoned at the end of the year 1914. The 
maximum pressure recorded on this gauge was 10.6 pounds per sq. ft. 
in December, 1914. As the substructure was erected anemometers were 
established at the highest and most exposed points. 

The foundations and all masonry work were constructed by Messrs. 
M. P. & J. T. Davis, Contractors. 

Plate XXI. — Shows the general dimensions and layout of this 
masonry. 



73 



Chapter II 

OUTLINE OF THE GENERAL SCHEME OF ERECTION IN SO FAR 
AS IT AFFECTED THE DESIGN OF THE SUPER- 
STRUCTURE: 

Plate XXII. — Indicates the order in which the different parts of 
the structure were erected. 

Following the order of designing, the erection of the suspended span 
was first considered. Although tenders had been submitted for erecting 
the official design by cantilevering, in order to fully comply with the 
specifications, there was no expectation that these tenders would be 
accepted. For the St. Lawrence Bridge Company's alternative designs, 
plans were made and estimates prepared for erecting and floating the 
suspended span at the high level which it would eventually have in the 
structure and also for erecting and floating it at a lower level and hoisting 
it to its place in the bridge after it reached the site. The high level scheme 
had many objections which became apparent after the drawings were 
prepared and the site investigated, but tenders were made on it on account 
of the prejudice known to exist against hoisting the suspended span into 
position, the method actually used. 

The span was erected on steel falsework having corner towers suffi- 
ciently strong to carry the span. After the intermediate staging bents 
were released and removed, barges were floated under it and the span 
blocked on top of the barges which rested on wooden sills bedded in concrete. 
Valves in the bottom of the barges permitted the tide to flow in and out 
until it was desired to move the span, when the valves were closed at 
extreme low tide. 

A favorable site for erecting the span was found at Sillery about 
314 miles to the Eastward of the bridge site. This site had a flat rock 
bottom, dry at most low tides, affording good foundations for the false- 
work and making it easy to prepare them. The Railway connecting the 
Material Yard at the Bridge with the Market Street Station at Quebec 
was about the elevation at which it was desired to erect the floor of the 
suspended span and swept around Sillery Cove in a curve, allowing an 
easy connection leading directly on to the falsework set for the erection of 
the span. 

Plate XXIII. — Is a plan of Sillery Cove showing the position in 
which the span was erected. 

74 



From the time it was decided to erect the centre span on falsework 
and float it into position it was foreseen that the greatest risk in placing 
would be when the suspended span was being anchored between the canti- 
lever arms in order that the suspension connection or the hoisting connec- 
tion as the case might be could be made. There were several factors in 
this risk. The range of tide of about 18 feet at the time of placing caused 
comparatively rapid vertical movements of the water surface; the strong 
current reaching about 7 miles per hour only a short time before the turn 
of the tide; the cross currents at or near slack water, and the large wind 
surface of the span and falsework, all making it difficult and hazardous 
to move and place the span in the position required without fouling and 
damaging some of the permanent structure either of the cantilever arms 
or of the span. 

Important considerations in favor of low floating were the high and 
very expensive falsework and the high centres of gravity and wind pressure 
for the elevated position of the suspended span. This raising of the centres 
of gravity and pressure required long pontoons to give stability, greater 
displacement in the pontoons to carry the weight of the falsework, and it 
would have been necessary to make the pontoons strong enough in them- 
selves to carry the weight of the falsework and span to the foundations, 
as there seemed no practical way of floating the pontoons under after the 
span had been erected on the falsework. The extra length and draft of these 
pontoons would have added considerably to the difficulty of controlling 
them when moving the span and to the risk in placing it in position. 
Moreover, there was no suitable site to make foundations for these longer 
and deeper pontoons. The site at Sillery while excellent for the pontoons 
used, could not well have been adapted to anything requiring greater draft. 
The actual method and arrangements for hoisting the suspended span 
will be given in Chapter V. 

The traveller which erected both the anchor and cantilever arms 
and set the falsework of the anchor arm, is shown on plates XIII and XIV. 
Its total weight was about 920 tons, almost equally distributed 
on front and rear trucks with both cranes placed at the rear 
end. When lifting its maximum load, the reaction under the forward 
posts was about 1,300,000 lbs. on each leg and for some conditions there 
was an up-lift at the rear end of 59,000 lbs. The traveller ran on rails 
laid on the top flanges of the permanent bridge stringers, and other 
stringers temporarily laid on the floor beams to suit the gauge of the 
traveller truck wheels, with temporary lateral bracing between the top 
flanges and sway bracing as required. The stresses resulting from the 
traveller load did not call for increases of sections in the permanent 
structure. 

Briefly, the plan followed in erecting the anchor arm was: The 
main shoes were set accurately on the pier; the bottom chords were laid 
from the shoes towards the abutment in a perfectly straight line and the 

75 



joints were completely riveted. After the bottom chords were fully 
riveted and made continuous from end to end they were jacked down to 
the position required to give the calculated camber in order that they 
might be straight when under load in the finished structure. The traveller 
set its own falsework as it proceeded towards the centre pier to erect the 
shoes ; set the chords as it retreated towards the abutment and was then 
taken back to the centre pier to erect the lower half of the web bracing and 
complete the triangles resting on the bottom chord. The upper tri- 
angles were then started from the abutment and as the traveller advanced 
toward the pier it completed the erection behind it. 

The above arrangement as carried out worked perfectly, but at first 
it seemed that it would be impracticable as the deformation diagrams 
showed that the upper triangles could not be connected without a consider- 
able bend in some of the members, necessitating very heavy forcing. 

The difficulty of connecting the upper triangles was overcome by 
slightly elongating the pin holes in some of the members. Tests were 
made, as referred to in Appendix A — to" ascertain what effect the elonga- 
tion of the hole might have upon the strength of the member, and it was 
found to be in no way detrimental. 

All field splices in the compression web members were fully riveted 
as the work was assembled, thus avoiding any possibility of open joints. 
This made it necessary to slightly spring some of the compression members 
in order to match the pin holes when connecting an " M " joint or when 
completing a bottom chord panel. The springing of the compression 
diagonals was accomplished by means of wire tackles, but for the bottom 
chords of the cantilever arm it was found desirable to use a temporary 
platform on which the chords could rest while being assembled and riveted 
and from which they could be jacked to the required position. Generally 
speaking, the cantilever arms required very little special provision for 
erection conditions. 



76 



Chapter III 



PLANS AND DESCRIPTION OF THE STEEL SUPERSTRUCTURE 

Plate XX. — Shows the outline of the structure with the governing 
dimensions. The clear head room of 150 feet above high water 
required by navigation interests and the rail level on either side made it 
economical to use a falling grade of one percent from each end of the sus- 
pended span. 

Specifications: 

The specifications were determined by the Advisory Board of Engineers 
at the time the contract was signed. 

The loads and unit stresses for which the bridge was designed are as 
follows : 

Loads — The loadings for which the bridge will be calculated are as 
follows : — 

A. Train Load — Two Class E60 Engines, followed or preceded, or 
followed and preceded by a train load of 5,000 lbs. per foot per 
track, on one or two tracks. Where empty cars weighing 900 lbs. 
per lineal foot of track in any part of a train produce in any member 
larger strains than the uniform load of 5,000 lbs., such empty 
cars shall be assumed. 

B. A Sidewalk Live Load of 500 lbs. per lineal foot for each of two 

sidewalks. 

C. Dead Load — The weight of all material remaining in the completed 
bridge, and a snow load of 500 lbs. per lineal foot of bridge. 

The weight of railway floor above stringers to be assumed at 
860 lbs. per lineal foot of each track. Timber to be assumed to 
weigh 5 lbs. per ft. B.M. 

D. Erection Loads — The weight of loaded travellers, erection plant 
and materials. 

E. A Wind Load normal to the bridge of 30 lbs. per square foot of 
the exposed surface of two trusses and one and one-half times 
the elevation of the floor and 120 lbs. per lineal foot of bridge on 
sidewalk fence (fixed load), and also 30 lbs. per square foot on 
travellers and falsework, etc., during erection. 

77 



F. A Wind Load on the exposed surface of the train of 300 lbs. per 
lineal foot applied nine feet above base of rail. (Moving load). 

G. A Wind Load parallel with the bridge of 30 lbs. per square foot 
acting on one-half the area assumed for normal wind pressure. 
(See paragraph E). 

H. Tractive Force assumed at 750 lbs. per lineal foot on one track. 

I. Temperature — A variation of 150 degrees Fahr. in the uniform 
temperature of the whole structure. 

J. Temperature — A difference of 50 degrees Fahr. between the 
temperature of steel and masonry. 

K. Temperature. A difference of 25 degrees Fahr. between the tem- 
perature of a shaded chord and the average temperature of a chord 
exposed to the sun. 

L. Temperature. Stresses due to a difference of temperature of 25 
degrees Fahr. between the outer web exposed to the sun and the 
other webs of compression members. 

M. Impact from railway load shall be assumed as 100% for stringers; 
75% for floorbeams and truss members carrying one panel load 
or less; 50% for hangers carrying two panel loads and 20% for all 
main truss members. For trusses of approach spans 

Impact = S — — — - , where L is the length of loaded track, 

and S = total stress from railway load. 

Loads Used to Determine Section of Members — All the co-existing loads 
and stresses and the deformation shall determine the section of the different 
members with the following restrictions: — 

Load "A" shall be used in all calculations where not otherwise provided ; 

Load " B " shall be used for floorbeams and sidewalk stringers, and 
members receiving their maximum strain from a length of moving load 
covering two panels or less; 

Stresses produced by " L" shall be considered as secondary stresses, 
and loads " K " and " L" shall be assumed to co-exist with one-half wind 
loads " E " and " F." 

UNIT STRESSES AND PROPORTIONING OF PARTS 

Unit Stresses (Carbon Steel). All parts of the structure shall be pro- 
portioned so that the sum of the maximum stresses produced by the 
loadings specified, including impact, shall not exceed the following amounts 
in pounds per square inch. 

78 



Tension: Lbs. per sq. in. 

Eye-bars 20,000 

Riveted members . . 18,000 

Including secondary stresses 24,000 

Compression: 

Short members with l/r 50 and under 14,000 

Long members with l/r over 50 17,500-70 l/r 

Including secondary stresses 18,000 

Bearing: 

Shop rivets 22,000 

Field rivets 20,000 

Rollers, per lineal inch 600d 

where d = diameter of roller in inches. 

Pins in eye-bars 22,000 

Pins in riveted members 20,000 

Bending: 

Pins 25,000 

Steel castings 16,000 

Shear: 

Shop rivets and pins 11,000 

Field rivets 10,000 

Webs of plate girders, gross section 10,000 

Steel castings 11,000 

Bearing on Masonry: 

Granite 800 

Concrete 250 

For nickel steel increase the unit stresses given for carbon steel by 
40%. 

Grip of Rivets — For rivets with a grip greater than four diameters 
reduce the units specified in Paragraph 20 by one per cent. (1%) for 
each one-sixteenth (l/16th) inch of additional grip, except in compression 
members having butt joints, but no rivet shall have a grip exceeding seven 
and one-half (7 l A) diameters. 

Laterals and Sway Bracing — Take both systems in calculation of 
strains, disregarding reversal of strains. 

For compression 16,000-70 l/r 

Anchorage Masonry — Anchor piers shall show a co-efficient of safety 
of two for all primary stresses including impact. 

Bending Stresses — All bending stresses in compression members 
produced by the weight of the member itself and by loads applied on the 
member shall be considered as primary stresses. 

79 



All such members shall be proportioned so that the greatest fibre 
stress due to this bending and axial strain together will not exceed the 
allowed units for the axial stress in that member. 

Secondary Stresses — All stresses produced owing to the deformation 
of the steel work under any and all loads, either by the absence of pins at 
the joints or by the friction on pins opposing the turning of members shall 
be considered as secondary stresses. 

Alternate Stresses— Members subject to alternate tension and com- 
pression shall be proportioned for either stresses and their section shall be 
made equal to the sum of the two sections. Rivets in connections and 
splices in all cases shall be proportioned for the sum of both stresses. 

Splices in Tension Members. Tension members shall be given full 
splice in material and rivets. 

Splices in compression Members — All splices in compression members 
shall be given full strength in material and a sufficient number of rivets for 
the axial stress. 

Net Section at Pins — Pin-connected riveted tension members shall 
have a net section through the end pin hole at least thirty-three per cent. 
(33%) in excess of the net section of the body of the member and the net 
section back of the pin hole parallel with the axis of the member, shall not 
be less than eighty per cent. (80%) of the net section of the body of the 
member. The net section through the intermediate pin holes shall be 
increased over that of the member by the section cut out by the pin hole. 

Latticing — The latticing of compression members to be proportioned 
for a cross shear per square inch of gross section of two per cent. (2%) of 
the unit stress for short struts of the same material; if the weight of the 
member produces additional shear this must also be provided for. 

Single flat lattice bars shall have a thickness of at least 1-40 and double 
lattice bars at least 1-60 the distance between nearest end rivets. Their 
unit stress in compression shall be 8,600 — 63j for carbon and 40% more 
for nickel steel. 

Lattice bars shall be so spaced that the portion of the flange included 
between their connections shall be as strong as the member as a whole. 

Radius of Gyration of Compression Members. Minimum radius 
of gyration shall be one one-hundredth (l-100th) of the length of member 
for trusses, and one one-hundred-and-twentieth (l-120th) for lateral and 
sway bracing struts. 

Materials to be Used — Approach spans, floorbeams, stringers, hand 
railings, stairways and all rivets shall be made of carbon steel. In case 
the main part of any member of the trusses is made of nickel steel, all the 

80 



details and connections of such members shall also be nickel steel. In case 
the main part of any other member of the bridge is made of nickel steel, 
the details and connections may be made of carbon steel. 

Minimum Thickness. No material shall have a thickness of less than 
one-half (%) inch, except floor bracing, lacing for sway and lateral bracing, 
stiffener angles, and sidewalk stringers which may be three-eight (finches. 

In no case shall any material be less than three-eight (%) inches, except 
fillers. 

Compression Members — The thickness of plates in compression mem- 
bers shall not be less than l-24th of the distance between the lines of rivets 
connecting them to the flanges. 

Forked Ends — When forked ends are used they shall be made of at 
least twice the sectional area of the member, and at least as strong as the 
body of the member. 

MATERIALS 

Rolled Carbon Steel 

All structural steel shall be made in an open-hearth furnace. 

Chemical Requirements — The ladle tests of steel as usually taken shall 
not contain more than the following proportions of the elements named. 

Acid. Basic 

Phosphorus 06 per cent. .04 per cent. 

Sulphur 04 per cent. .04 per cent. 

Manganese 70 per cent. .70 per cent. 

except rivet steel, .60 p.c. 

No chromium to be used. 
Silicon 10 per cent. .10 per cent. 

It is desired that the carbon contents be as small as possible to meet 
specifications. 

Rivet Steel — The ladle tests of the carbon rivet steel shall not contain 
more than .03 of one per cent, of phosphorus, and not more than .03 of one 
per cent, of sulphur. 



81 



Physical Requirements — Specimens cut from the finished material 
shall show the following physical properties: — 



MATERIAL 


Ult. Strength 

lbs. per square 

inch 


Minimum 
Yield Point 

lbs. per 
Square Inch 


Minimum 
Elongation 

per cent, 
in 8 inches 


Minimum 

Reduction 

per cent. 

of area 


Shapes and 










Plates up to 










and including 
1 in. thick. . . 


62,000 to 70,000 


35,000 


1,500,000 
ultimate 


44 per cent. 


Shapes and 




. 






Plates over 1 






22%; 20% for 




in. thick 


62,000 to 70,000 


33,000 


Sheared Plates 


40 per cent. 


Eye-bar Flats 
(unannealed) 


66,000 to 74,000 


35,000 


22% 


40 per cent. 


Rivets 


48,000 to 56,000 


28,000 


1,500,000 
ultimate 


50 per cent. 


Pins and Rollers 










(annealed) . . 


65,000 to 75,000 


35,000 


22% in 2" 


35 per cent. 



Yield to be determined by drop of the beam. 

Speed of machine for testing samples to be such that material under 
tension will not elongate more than one inch in two minutes. 

Bending Tests — Specimens cut from plates, bars and shapes two inches 
wide shall bend cold 180 degrees around a rod of a diameter equal to the 
thickness of the specimen; when at or above a red heat, 180 degrees flat. 

Specimens cut from rivet rods shall bend 180 degrees flat when cold, 
or when at or above red heat. A test piece two inches long when heated 
to a bright cherry red shall flatten longitudinally under the hammer to a 
thickness of one-quarter (}/i) inch without cracking on the edges. 

Full sized sections of eye-bar material as rolled without annealing shall 
bend cold about a rod of diameter equal to twice the thickness of the bar. 
Angles of all thicknesses shall open cold to an included angle of 150° 
and close to an angle of 30°, without a sign of fracture. 

All specimens in bending tests must show no signs of cracking on the 
outside of the bend. 

82 



Rolled Nickel Steel 
All nickel steel shall be made in an open-hearth furnace. 

Chemical Requirements— The ladle test shall contain not less than 3.25 
per cent, of pure nickel, and not more than the following proportions of the 
elements named. 

Acid. Basic. 

Phosphorus 06 per cent. .04 per cent. 

Sulphur 04 per cent. .04 per cent. 

Manganese 70 per cent. .70 per cent. 

No chromium to be used. 

Silicon 10 per cent .10 per cent. 

Carbon 45 per cent. .45 per cent. 

Physical Requirements— Nickel steel for plates, shapes and unan- 
nealed eye-bar flats must meet the following physical requirements in the 
finished material: — 



MATERIAL 


Ult. Strength 

lbs. per square 

inch 


Minimum 
Yield Point 

lbs. per 
Square Inch 


Minimum 
Elongation 

per cent, 
in 8 inches 


Minimum 

Reduction 

per cent. 

of area 


Plates and 
Shapes 

Eye-bar flats, 
unannealed % 


85,000 to 100,000 

95,000to 110,000 
90,000 to 105,000 


50,000 

55,000 
55,000 


1,600,000* 
ultimate 

15% ' 
1,800,000 

. nil 


40 per cent.f 

25 per cent. 
35 per cent. 


Pins (annealed) 


ultimate 



* For material thicker than one inch (1"), the required percentage of 
elongation shall be reduced by one for each increase in thickness of one 
quarter inch (J4") or fraction thereof above one inch (1"), but in no case 
shall the minimum elongation required be less than 14%. 

J Tests for information shall be made of annealed specimens cut from 
the rolled eye-bar flats. 

f For material thicker than three-quarter inch {%"), the required 
percentage of reduction of area shall be reduced by two for each increase 
in thickness of one-quarter inch (\i") or fraction thereof above three- 
quarter inch (%!'). 

83 



Bending Tests — Specimens of nickel steel not less than 2" wide and of 
the full thickness of the material as rolled shall bend cold 180° around rods 
of the diameters specified below for the various thicknesses, without fracture 
on the outside of the bend. 

For material up to Y 2 " incl 180° around D = IT. 

over Y 2 " and up to W 2 " 

incl " " D = 2T. 

" over \Y 2 " " " D = 3T. 

Angles of all thicknesses shall open cold to an included angle of 150° 
and close to an angle of 30°, without a sign of fracture. 

Steel Castings 

Steel for castings shall be made in an open-hearth furnace. 

Chemical Requirements — The ladle test of steel for castings shall not 
contain more than the following proportions of the elements named : — 

Phosphorus 04 of one per cent, for basic steel. 

Phosphorus 06 of one per cent, for acid steel. 

Sulphur 05 

Manganese 75 " 

Silicon 35 

Physical Tests — Test pieces taken from cupolas on the annealed 
castings shall show an ultimate strength of not less than 65,000 lbs. per 
square inch, an elastic limit of at least 35,000 pounds per square inch, and 
an elongation of not less than 20 per cent, in two inches. They shall bend 
without cracking 120 degrees around a rod twice the thickness of the test 
piece. 

The final stress sheets and the working plans were necessarily made 
concurrently and after many trials to determine closely approximate 
sections and resulting weights, the deadweight of this bridge being so 
largely in excess of the live loads that errors in assumed weights were not 
permissible — particularly in the suspended span or towards the ends of the 
cantilevers where the deadweight has a large influence on the stresses in 
the cantilever and anchor arms. It is interesting to note that the shipping 
weights agreed within less than one percent of the estimated weights used 
in the final calculation of the stresses. 

In making the working drawings the natural order of procedure, after 
the approximate designs had been prepared, was to take up first those self- 
contained units of the structure in which stresses did not result from the 
weight of other portions of the Bridge, but the deadweight of which had a 
marked effect upon the stresses and sections in the trusses of the structure. 
The description will generally follow the order of making the final design 
— namely, the floor system, the suspended span, the lateral systems, 

84 



the cantilever structure, the shoes, and the anchorage; but the subjects 
must overlap in places and the logical order cannot always be strictly 
maintained. 

Floor System: 

Plate XXIV. — Shows the general arrangement of the floor system over 
the whole structure. It also shows a cross section of the suspended span 
and at two points in the cantilever structure. Cross sections showing the 
floor over the main and anchor piers have already been shown on Plate XX. 

Plates XXV and XXVI. — Are stress and material sheets for the floor 
Bystem. 

Mention has already been made of the type of through floor adopted 
for the purpose of giving greater safety in case of derailment and for the 
sake of facilitating the erection. 

The tracks were designed to be laid with 100 lb. A.S.C.E. rails on 8" x 
10" wooden ties 12' long spaced 12" centres. Sixty-pound rails arc also laid 
as guard rails. The ties rest on 24" I stringers spaced 7' centres headed into 
the floor beams of the through plate girder spans which are spaced about 
14' apart. The plate girder spans have lateral bracing in each panel and 
the webs of the stringers are diagonally braced to the girders to provide for 
traction and lateral stresses. The top flanges of the through girders are 
well stayed from the floor beams by plate brackets and are reinforced with 
heavy 15-inch channels for protection in case of derailment. 

The sway bracing of the trusses above the floor is in the form of portal 
bracing with the inclined knee brace intersecting the plane of the truss well 
below the floor to avoid bending in the truss members. It was desirable to 
spread the tracks as far as practicable to reduce the moment and the 
deflection of the floor beams, but the spacing was limited to 32'6" centre to 
centre by the necessary clearance of the portal bracing. 

In spreading the tracks, consideration was given to the torque produced 
in the suspended span and the effect upon its sway bracing and lateral 
systems when diagonally opposite corners of the cantilevers were loaded by 
trains in positions to give the maximum difference in the deflection of the 
corners at the end of each cantilever. It was found that with the track 
spacing adopted the suspended span is sufficiently flexible to adapt itself 
to this twist without over strain in the cross bracing. 

Plate XXVII. — Is a shop drawing of the single girder floorbeams. 

The floorbeams were made as deep as practicable but owing to their 
great length the deflection is appreciable and special provision was necessary 
in the end connections to avoid bending the truss members to which they 
connect. With the exception of the floorbeams in the suspended span and 

85 



those at panel points 1 and 2 of the cantilever and anchor arms, the floor- 
beams are hung on pins placed in their neutral axis, the pin holes being 
bushed to give a low unit pressure and to permit deflection of the floor 
beams without cutting the bearings. 

The floorbeams at the main panel points of the cantilever structure 
are made of two girders, one placed on each side of the vertical tension 
members to which they are connected. The two girders are connected by 
diaphragms placed over the pins and at the centre of each track. The 
distribution of load upon the two girders is further provided for by con- 
necting the abutting ends of the longitudinal track girders. All other 
floorbeams, including those in the suspended span, have single webs with 
the bottom flanges stayed to the track girders to avoid undue vibration. 

Provision for Traction Stresses and Expansion Joints in the Floor System: 

The fixed points of the cantilever system are the main piers and 
provision for changes in length arising from temperature and deformation 
under loads was necessary at the anchor piers and at the junction between 
the ends of the cantilevers and the suspended span — the possible varia- 
tions, as shown on Plate XXVIII being 11 inches over the anchor piers 
and 14 inches at the junction of the suspended span. The floor is at varying 
distances from the neutral axis of the cantilever structure and when the 
whole structure deforms under load the floor does not exactly follow the 
longitudinal movements of the truss verticals to which it is connected. 
Moreover, being of lighter material it may be more rapidly affected by 
temperature changes than the heavier material of the trusses. Expansion 
joints were, therefore, provided at floorbeam 14 over the main piers and at 
the intermediate points, FB7, in the anchor arms and FB9 in the 
cantilever arms; also at the centre of the suspended span, thus dividing the 
floor of the cantilever structure between anchor piers into ten sections. 

Where the floorbeams of the sub-panels are carried on posts, the ends 
of the floorbeams and the posts are stayed longitudinally by braces to the 
floor system, but in the more usual case where the floorbeams are carried 
by tension members from the " M " joints there is no provision for fixing 
the ends of the floorbeams or to take up longitudinal stresses at these 
points, and it is necessary to fix each section of floor in such a manner that 
the traction stresses are carried through the truss structure to the main 
piers. The floor sections are anchored against longitudinal movement in 
panels 5-6 of the suspended span, and in both cantilever and anchor arms 
in panel 1-2 and in the 5th panel from the main pier. On the suspended 
span and at the ends of the cantilever structure where the floor system is 
near the main lateral system, the floor is attached to the laterals of the 
bridge. Lattice girders, spanning the panel length in which the traction is 
taken, are placed on the centre line of each track with their top flanges con- 
nected to the lateral system of the track girders and their bottom flanges 
connected to the lateral system of the bridge. The traction stresses are 

86 



thus transferred from the floor to the bottom chords of the trusses. The 
lattice girders have flanges of two 4x4x |^ angles connected by lattice 
angles in the form of a Warren truss. 

In panels 10-11 of the cantilever arm and 8-9 of the anchor arm, advan- 
tage was taken of the inclination of the sway bracing in the inclined com- 
pression members. Immediately below the floor girders deep lattice trusses 
are substituted for the regular sway bracing. The bottom flanges of the 
track girders are connected by brackets to these trusses and due to the 
inclination of the diagonal compression members, about 45°, the truss 
receives traction stresses from each floor girder and transmits them to the 
diagonal posts. The vertical component of the stress delivered to the 
traction truss is resisted by the track girder in bending, causing an increase 
or diminution of the bending moment from vertical loads depending upon 
the direction of traction. Provision for the additional stresses is made in 
the track girders and main floorbeams at these points. 

Plate XXVIII. — Shows the expansion joints in detail. 

Suspended Span: 

Plate XXIX. — Is a stress and material diagram for the suspended 
span which shows the form of the cross sections of all the members. 

Figure 17. — Giving a view of the span shows the general construction. 

Plate XXX. — Is a shop drawing of the lower half of the end post. 

Plate XXXI. — Is a shop drawing of Post UeLe, showing typical 
connections at top and bottom, the slot for the floorbeam connection 
and the bottom lateral connection. 

The chords and end posts are constructed of three webs 45" deep, 
connected on the top by cover plates 64 inches wide, also connected by 
diaphragms and the usual tie plates and lattice bars on the bottom flanges. 
The top chords were shipped and erected in half panel lengths, and spliced 
with material and rivets to take up the full stress after being erected. The 
connections at the main panel points are made on pins which also serve 
for the upper connection of the main web members. The end tension 
diagonals are eye-bars connected on pins, but all other web members consist 
of two built-up channels having their webs in planes parallel to the centre 
lines of the trusses in order that they could be connected at the ends by 
riveting to connection plates through which the chord pins passed. 

The lower chords are nickel steel eye-bars packed parallel in one tier. 
The floor beams are passed through and connected to the inside channel 
of the vertical members and also connected to the outside channel, thus 
distributing the load to both sides of the verticals. The end connections 
are given a bevel equivalent to the deflection under one-half live load, and 

87 




a 

CO 



■a 

ss 

a 

3 
CO 



o. 

OJ 

s-i 
•00 
O 
w 
O 

Cm 






88 



the verticals have a slight initial bend under dead load only but are straight 
and without braiding stresses when the floor beam is under usual conditions 
of live load. 

The lateral bracing ia of box section with the top and bottom flanges 
connected to the floor beams and the verticals at equal distances from the 
centre line of the bottom chord, thus eliminating bending stresses in the 
truss members. 

The suspended span is carried by eye-bars pin-connected at the 
intersection of the lower chords and the end post of the truss as shown on 
Fig. 17, but during erection it was carried by a bearing under the centre rib. 
The transfer of stresses from the outer ribs to the centre rib was accom- 
plished by inserting heavy diaphragms and reinforcing the tie plates 
adjacent to the shoe. 

The transverse horizontal loads acting on the suspended span are 
transferred to the lateral system of the cantilever by a heavy pin illustrated 
on Plate XXVIII. This pin is cantilevered out from a box girder forming 
an end strut for the suspended span and slides in a bronze bearing, spherical 
on the outer surface, so that it may be free to align itself with the pin under 
all conditions of vertical or horizontal flexure. 

The eye-bars suspending the span from the end of the cantilever were 
made but 10" wide to minimize the secondary stresses from bending, due 
to the expansion and contraction of the structure. Under certain com- 
binations of loading and temperature — the deformations may cause the 
point Lo of the suspended span to move 9" towards the centre of the 
bridge or 5" away from it in the other direction, giving a total range of 14". 

Experiments were made, as referred to in Appendix A to determine the 
resistance of pin friction (steel on steel dry) to angular movement, and the 
bending stresses caused by friction of 40% were provided for. The 
pin holes in the eye-bars are bushed with bronze to reduce the unit pressure 
and prevent cutting and it is probable that the actual bending stresses are 
very much less than those estimated. 

Cantilever Structure: 

Plates XXXII and XXXIII. — Show the axial stresses and the 
material of the cantilever and anchor arms. 

Plates XXXIV and CXIII. — Show the secondary stresses throughout 
the cantilever and Anchor arms. These did not affect the sections of 
the members but required some extension of the pin plates at CLie 
and ALu and special treatment of the floorbeam connections at CL.2 
Ala, as indicated on the plates. 

Plates XXXV and CXV. — Show the stresses and material in the 
sway bracing of the cantilever and Anchor arms. There is no top lateral 
bracing in the cantilever or anchor arms, the horizontal forces applied 

89 



at the top chord being transferred through the sway bracing between the 
vertical posts to the " M " joints, from which points they are carried to 
the bottom lateral system through sway bracing in the plane of the 
inclined compression members. The sway bracing between the vertical 
members is of the ordinary type consisting of deep top and bottom struts 
connected by a single set of diagonals in each direction. The bottom 
strut is stayed by a vertical from its centre to the intersection point 
of the diagonals. 

The wind shears at the " M " joints are carried by portal bracing to a 
point below the floorbeams where the diagonal bracing is resumed, as 
shown on the stress sheet. The knee braces of the portal are extended to 
an intersection with the diagonals below the floor system to avoid bending 
stresses in the compression members in the truss. Sway bracing of the 
" K " form is placed between the tension verticals of the truss below the 
floor to carry the transverse loads acting on the floor to the bottom lateral 
system. 

The omission of the top laterals eliminates ambiguity and makes the 
wind stresses for the assumed pressure determinate throughout the 
structure. While the vertical forces due to wind displacement con- 
siderably increase the stresses in the truss members, the arrangement 
worked out more economically than if the top chord wind loads had been 
carried to the centre and from thence transferred through the sway bracing 
to the main piers. It, moreover, simplified the erection as the cantilever 
arm was stayed against wind forces by the permanent bracing — without the 
use of temporary material — at all stages of its erection. 

The laterals were calculated as a double Warren system, but hori- 
zontal struts were introduced to assist in keeping the alignment during 
erection and to complete the " K " system of sway bracing in the vertical 
tension members. The lateral members and struts are all formed of 
double lattice girders the depth of the chord, latticed together on top 
and bottom flanges. The laterals are connected to the chords by heavy 
connection plates reinforced where necessary to transmit the longitudinal 
stresses from the connection to the centre of the truss. The pin plates 
connecting the truss members prevented the upper lateral connection 
plates from being carried across the upper flanges of the chords and the 
forces are distributed over the four ribs of the chord by wide tie plates 
at the ends of the lateral connections on both top and bottom flanges of the 
chords. 

Plate VII. — Shows the cross section and outside dimensions of all 
members in the cantilever arm. 

The main truss members of the cantilever structure are made of four 
ribs consisting of symmetrical I sections for compression members, and 
channel sections for tension members, built up of plates and angles. The 
compression ribs are connected in pairs by longitudinal diaphragms on the 

90 



cciitro lines and lattice and tie plates connect the flanges thus forming two 
II sections, which were again connected in the field by tie plates on the 
flanges of all members and by an additional line of tie plates on the centre 
line of the larger members. 

The tension webs were assembled in pairs with flanges turned in and 
connected by lattice and tie plates on the flanges only. 

Each member was completely assembled in the shop for the purpose 
of finishing the ends, boring pin hole- and drilling splicing material. Most 
of the pins connecting the web system were in two lengths. This construc- 
tion practically divided each main truss into two complete trusses placed 
side by side and field connected together, which facilitated the transporta- 
tion and erection of the large members. The butt joints of the compression 
members are in perfect contact throughout the structure, but when the 
work, was designed it was thought impracticable to attain this accuracy and 
all compression joints are fully spliced by material and rivets, no reliance 
being placed on the bearing of the faced ends. 

A description of the details would be too voluminous but a few typical 
drawings follow which show the general arrangement of the details and the 
assembly at the joints of the main members of the structure. 

Plate XXXVI. — Shows the size of lattice bars for all members 
of the cantilever arm, and plate XXXVII the method of their 
determination. 

The general make-up of the members is illustrated by the typical 
shop drawings shown on plates. 

Plate XXXVIII. — Shows the detail drawing of the bottom chord at 

AldS. 

Plate XXXI X. — Shows the upper half of a diagonal compression 
member with the detail at AMio. 

Plate XL. — Is a vertical compression member which rests on the 
diagonal shown on Plate XXXIX. 

Plate XLI. — Shows a typical joint of the bottom chord with the 
truss members and lateral bracing assembled. It will be noted that all 
the connections of the web members to the bottom chord were made on 
pins outside the chord permitting the webs of the chord and the com- 
pression web members to be in practically the same vertical planes 
which are parallel throughout. In elevation the chords are tapered from 
a depth of 82" at the shoe to a depth of 45" at the L2 joint. 

Plate XLII. — Shows the detail at Joint AL2. 

Plate XLIII. — Shows the assembled members at AM12. It will 

be noted that the detail is built as an integral part of the 

compression diagonal AMuMu. This arrangement enabled the 

91 




a 
o 

Ji 



J3 

a 

03 

i- 
«0 

o 

*•> 

o 

a 



92 




93 



material of the compression member to be carried through the joint 
and gave an independent connection for all the members assembling at this 
point. The combination of these heavy connection plates is quite 
complicated and cardboard models were made of all joints to facilitate 
the consideration of the distribution of stress amongst them. 

Plate XLIV. — Shows a typical top chord panel. Eye-bars could 
not be obtained in full panel lengths at reasonable cost and difficulty 
was foreseen in handling such long bars. There was doubt as to the 
proper allowance for bending stresses caused by the weight of the bars. 
The bars were, therefore, made in two lengths connected by pins in the 
centre of the panels. They were packed parallel to the plane of the 
trusses in two tiers, to insure an equal distribution of stress throughout all 
the bars by reducing the length of the pins. To prevent deflection of the 
bars they were joined at the center. The connecting pins were carried 
by trusses spanning the length of a panel. The trusses with their lateral 
bracing formed a box frame in which the bars were packed before erection. 
This arrangement greatly simplified the erection and held the bars in place 
until the pins could be driven. At the top of the main post the bars were 
connected by a special detail the shop drawing of which is shown on 
Plate XLV. 

Plate CXVII. — Shows the main posts and the bracing between 
them. 

Plate XLVI. — Shows the joint AU14, the top of the main post. 

Flate XLVII. — Shows the detail at joint AM14 about the middle of 
the main post and show the adjustment provided in the horizontal member 
to insure the proper distribution of stress and provide a means of tipping 
the main post out of the vertical position in order to be able to connect the 
members in the anchor arm. 

The form of the trusses, the arrangement of the joints and the methods 
of loading of the trusses assured the calculated distribution of stress through- 
out the structure. At the end of the cantilever the weight and loads carried 
by the suspended span are equally distributed over the four webs of each 
cantilever truss. 

Plate XLVIII. — Shows the joint at CUo the end of the cantilever 
arm. 

Deformation diagrams were studied concurrently with the preparation 
of the stress sheets, the design of the details and the erection, but the only 
special provision necessary for deformation or for secondary strains beyond 
the ordinary framing for camber and the elongated holes in the tension 
members of the anchor arm was in the suspension bars of the central span 
and in the pin joints at the main shoes. 

Plate XLIX. — Shows the anchor arm deformation diagram. For 
convenience in making connections the anchor arm was erected with 

94 



the point La 5 inches below its normal position. This and the framing 
lengths requi hi 1 that the vertical post over the main piers should be set 
with it.s top 14 inches shorewards, causing angular displacement in all 
members connecting at the shoe. As the cantilever was erected the post 
gradually approached a vertical position. 

To provide for this angular displacement without causing bending 
stresses in the members connecting at the shoe and to eliminate the danger 
of pin cutting, the bearing unit pressure on the pins was very much reduced 
by the use of sleeves and as a further precaution the moving surfaces were 
lubricated with paraffin. By introducing sleeves between the pins and the 
bearing surfaces of the members to give a bearing over the whole length of 
the pins, the unit pressure was reduced to about 8600 lbs. per sq. in. under 
total load and 4000 lbs. per sq. in. for dead load and the thickness of the 
members bearing on the larger diameter of the sleeve is kept within reason- 
able limits, about 9 inches. The experiments noted in Appendix A show 
that under the above pressure and with paraffin lubrication the friction is 
reduced to 2%, but to be entirely on the safe side it was calculated at 40%. 

Deformation Diagrams are shown on Figs. 5 and 8; also on Plate L. 
These show remarkable uniformity of curves and the absence of kinks in the 
members at all stages of the work. 

Plate LI. — Shows the general construction of the main shoe. 

Figure 18. — Shows the shoe being assembled in the shop. 

Figure 19. — Is a view of the completed shoe in the structure showing 
the connection of the lateral bracing. 

The maximum stresses calculated in the various members connecting 
at the shoe are as follows : — 

Main Vertical Post, 26,600,000 lbs. ; Cantilever Compression Diagonal, 
7,820,000 lbs.; Anchor Arm Compression Diagonal, 7,810,000 lbs.; Canti- 
lever Chord, 29,600,000 lbs.; Anchor Arm Chord, 24,100,000 lbs. 

The vertical components of the above stresses are 55,000,000 lbs. and 
the horizontal components at the centre of the shoe 32,000,000 lbs. 

The shoe proper is built up of four webs in the same plane as the webs 
of the connecting truss members. The vertical loads from the webs are 
distributed on the masonry by a grillage of cast steel box girders each about 
6'8" wide and 20'10" long, giving a bearing area on the masonry of 26' 4" x 
x 20' 10". Each grillage casting has 4 longitudinal ribs and 8 cross 
diaphragms one under each of the main ribs of the shoe. The material 
is no where less than 23^" thick, increased as necessary to carry the con- 
centrated loads under the ribs. The rough castings as they came from the 
foundry weighed about 40 tons and the shipping weight, after machining, 
of each casting averaged about 37 tons. 

95 



For convenience in obtaining material, handling in the shop and 
shipping, the main portion of the shoe was made in two storeys the lower 
one 9' 6" high and the upper one 5' 5", making the centre of the pin in 
the main post 15 feet from the base of the shoe and 19 feet from the masonry. 
The ribs of the lower portion were shipped in pairs and connected in the 
field, but the upper portion was shop riveted together and shipped as a 
unit. 

Plate LI. — Shows the diaphragms between the ribs and the brackets 
opposite these diaphragms extending from the top of the shoe to the outer 
edge of the base plate resting on the grillage castings. The brackets 
were added to give lateral stability to the shoe and to carry the transverse 
wind force amounting to 1,300,000 lbs. to the base. While the grillage 
castings were calculated to distribute the vertical loads of the ribs over 
the full area of the base, the stiff brackets must carry considerable loads 
and these were made as rigid as practicable, special 12 x 12 bent plate 
angles being used for connecting them to the ribs and all faces being planed 
to give a perfectly true bearing. Horizontal plates were introduced between 
the two parts and at the bottom of the shoe to make convenient connections 
between the parts and to distribute the lateral stresses throughout the 
member. The ends of the shoe ribs and of the brackets are covered by 
face plates to add to the rigidity, make a better appearance and keep out 
the weather. Manholes were provided to all parts of the shoe to make it 
accessible for inspection. After erection the grillage castings were filled 
solid with concrete and the base of the shoe was also filled with concrete 
up to the bottom of manholes in transverse diaphragms and sloped to the 
drainage holes shown on the end elevation. 

The grillage castings were machined with great care to exact dimen- 
sions on all faces, assembled on a carefully levelled table and the holes 
drilled for the bolts through the end flanges connecting the sections. Each 
rib of the lower story of the shoe was fabricated separately but the holes 
connecting it to the grillage casting and for the diaphragms and brackets 
were drilled after the ribs were assembled on the grillage casting. The 
connecting rivet holes in the upper portion of the shoe were also drilled 
after assembly. After assembly and match marking on the grillages, the 
shoes above the grillages were re-assembled on the bed of a large boring 
machine and all pin holes were finished at one setting of the shoe. 

The pressure on the granite masonry from the vertical loads is 690 lbs. 
per sq. in. which is only slightly increased by the displacement due to 
transverse wind force, but the longitudinal wind force is much larger, being 
6,200,000 lbs., and is transmitted to the masonry through the compara- 
tively narrow grillage girders. After making due allowance for these 
girders being riveted to the base of the shoe and for friction, it is calculated 
that the toe pressure under the leeward edge of the grillage girders may 
reach 915 lbs. per sq. in. for all loads. 

96 



Friction brakes were placed at the ends of the suspended span in the 
sections of bottom chord between Li of the cantilever and Lo of the 
suspended span to prevent any sudden movement of the span due to 
traction or from longitudinal wind force. The detail of these brakes is 
shown in plate LI I. They are calculated for a resistance of 250,000 lbs. 
and are made adjustable so t hat this resistance may be maintains I. 

There is no condition of loading which produces a positive reaction 
at the anchor piers and the anchorage bars are always in tension. The 
transverse horizontal forces of the lateral system are transferred to the 
anchor system through a pin cantilevered from a heavy steel frame built 
into the masonry shown on Plate LIU. 

Plate XXVIII. — Shows the detail of the sliding shoe engaging the 
top of this pin and the frame in which it slides. 

Plate LIU. — Shows the general arrangement of the grillage girders in 
the anchor pier and the anchor bars at the end of the anchor arm. Steel 
stairways for inspection purposes are provided over the main piers and at 
the ends of the anchor and cantilever arms. The tops of the girders 
carrying the upper chord eyebars provide a foot walk over the cantilevers 
which is guarded by a pipe handrail on each side. 



97 



Chapter IV 

DESCRIPTION OF THE SHOPS AND EQUIPMENT WITH 
REFERENCE TO THE USUAL METHODS OF MANUFACTURE 

At the time the construction estimates were prepared there was no 
Canadian shop equipped with cranes and tools of sufficient capacity to 
manufacture the large members called for by the design; the shops of 
both the parent Companies were fully occupied with ordinary construction, 
making it undesirable to re-model either plant for manufacturing the 
Bridge and it was decided that special shops must be built for the work. 

There was no crossing of the River below Montreal, and the high banks 
at the site, great range of tide, exposure to waves and ice conditions in the 
Winter made it impracticable to deliver by water. The site selected for 
he shops is at Lachine, near Montreal, on the main line of the Grand 
Trunk Railway, giving connection by the Grand Trunk Railway and the 
Intercolonial Railway with the South side of the River at the bridge site 
and a short spur connected the shop site with the Canadian Pacific Railway 
over which material could be shipped to the North side of the River, both 
connections providing excellent railway facilities both for receiving and 
shipping material. 

Plate LIV. — Shows the general arrangement of tracks and connec- 
tions to the trunk lines, the location of all buildings and the storage yards. 

Plate LV. — Shows in fuller detail the plan of the principal shops, 
the position of the machine tools, the track arrangement of the interior 
of the shops and of the storage yards. It also shows the three elevations 
of the main shop. 

Plate LVI. — Shows a transverse section through each end of the 
shop and the longitudinal section through the centre. The overhead crane 
arrangements for handling material are indicated in these sections. 

The main portion of the shop was 160 feet wide and 660 feet long, with 
an extension 60 ft. wide along the South side for a distance of 440 feet 
from the East end and 30 feet wide for the remaining 220 feet. 400 feet 
of the West or receiving end of the main shop was divided into four bays 
of 100 feet each between columns. The 260 ft. of the East or finishing 
end was made higher with a row of columns 20ft. apart, practically dividing 
the shop into two aisles of 75 and 85 feet, each covered by travelling 
cranes. 100 feet of the East end of the 60 ft. extension was utilized as 
part of the main shop for boring eye-bars. The remainder of the 60 ft. 

98 



extension contained the pipe shop, forge shop, machine shop, tool room, 
power plant, heat inn plant and electric supplies. The straightening 
rolls occupied the 30 ft. extension at the West end. 

In planning the construction of the shop it was decided to eliminate 
the risk of fire and possible consequent delay by making the shop entirely 
fire proof, and no wood whatever entered into its construction. The 
roof is of 3-inch reinforced concrete slab, resting on steel beams; metal 
sash is used for all windows and sky-lights; the walls are of brick below 
the windows and cement plaster on expanded metal above. 

The template shop, shown on plate LV, is a steel frame building 
60 x 176 ft. in plan. The curtain walls are of brick but the roof is of 
wood laid on steel frames, as it was considered useless to attempt to make 
this building fire proof, its contents being of so inflammable a nature and 
in such quantity that a fire in the contents might destroy a fire proof 
building. A sprinkler system was installed to make the fire risk to 
building and contents as small as possible. 

The main shop was laid out with a view to conducting all longitudinal 
movement of material on narrow gauge tracks which extend to the storage 
yard at the West end. Transverse movement at the receiving end of the 
shop where the operations of straightening, shearing, drilling, edge planing, 
punching were carried on, was accomplished by means of transverse 
overhead cranes. After the material was fabricated it was moved East 
on the narrow guage tracks and assembled under heavy longitudinal 
cranes which were of sufficient capacity to lift the heaviest members after 
assembling. The South bay where the large wall planers and boring 
machine were placed was equipped with two 70-ton overhead cranes and 
one 35-ton, and all the large compression members were assembled, 
riveted and finished on this side of the shop. The North bay was equipped 
with two 35-ton cranes and was used for assembling the long tension 
members, the lighter compression members, floorbeams and the smaller 
material that could be handled by the lighter cranes. The shipping shed 
was equipped with one 70-ton two trolley crane and one 30-ton crane. 
In addition to being used for loading material, it was also used for 
assembling large members and drilling the splice connections. A heavy 
10' x 10' Loudon Bros, planer was installed at the North end for planing 
the 40-ton grillage castings, the shoe ribs and other work of a similar 
character. This planer was later fitted with a boring bar to supplement 
the Newton borers in the main shop. 

The site chosen for the shop was under-laid by rock at a depth of 
about four feet, giving excellent foundations for the large machines and 
the skidways to carry the heavy members. 

99 



The shops were laid out after careful consideration of the length of 
members to be handled, their weight, the operations to be performed and 
the quantity to be manufactured each month. 

The Plate Straightening Rolls were designed to straighten plates 
120 inches wide and were of sufficient strength to straighten nickel steel 
plates 84" wide and 134" thick. 

Three edge planers were installed, two with a capacity of planing 
plates 2" thick and 45' long, taking a 5/32" cut at a speed of 40 ft. per 
minute and one with a capacity of planing plates 20 ft. long. The Plate 
shear had a capacity of shearing nickel steel plates 84" wide and 134" 
thick. The Angle Shears had a capacity to cut nickel steel angles 
8 x 8 x VA" at an angle of 45°. 

Most of the rivet holes were drilled in the solid and for this purpose 
40 radial drills were installed. All drills had a 6' reach. The columns of 
14 drills were placed along the centre of the second 100 ft. bay and were 
anchored to the masonry. The remaining 26 drills were mounted on 
heavy trucks which ran on special tracks to which the trucks could be 
anchored after the drill was set in the desired position. 

Punches, spacing tables, angle mills, lattice mills, and other machinery 
of a similar kind were standard for heavy bridge work. 

Compressed Air at 100 lbs. was used for operating the riveting 
machines all of which were of the Hanna or Murphy toggle type of yoke 
riveter and maintained a uniform pressure on the rivet for the last %" 
of the stroke. There was considerable variety in the capacity and reach 
of the machines used. For the long \}4," rivets the machines were of 
100 tons capacity some with a reach of 6'3", while for the short "j/%" rivets 
machines of 70 tons capacity some having only 48" reach were used. 
The heavy riveters were mounted on wall cranes which travelled along a 
track on each side of the central columns at the riveting bays of the shop. 

Three 60" rotary planers were installed all on turntables, two of them 
on the south side of the shop being mounted on a heavy bed permitting 
both ends of a member 90 ft. long to be faced at the same time. One was 
mounted at the northeast corner of the shop and equipped with a squaring 
table, for finishing the ends of the smaller members. 

For facing the ends of the heavy compression members a double head- 
ed vertical and horizontal planer was installed at the east end of the shop 
mounted on a bed permitting both ends of the heavy compression members 
to be finished at the same time. These machines had a vertical and 
horizontal stroke of 126" and could be set to face both ends of a member 
65' long. Care was taken to maintain a nearly uniform temperature while 
a member was being planed and the operation of facing the individual 
member continued day and night until finished. 

100 



Compression members of ordinary size and tension members were 
finished with the double headed rotary and single rotary as before 
mentioned. For boring the wry large pinholes in the shoe and heavy 
compression members a special boring machine was installed at the south- 
east corner of the main shop. The boring bar of this machine had a suffi- 
cient range of motion to bore all the holes in any piece without resetting 
the work after the piece was bolted to the bed, the boring bar having a 
horizontal movement of 22' and a maximum height from the top of the 
bed to the centre of the bar of 15'. The cutting head travelled through 
a distance of 14'. The large pinholes in the shoes and bottom chord 45" 
in diameter were bored by this machine. 

In the north bay of the Shop two Newton horizontal borers were 
installed for boring the holes in the tension members and those in the 
lighter compression members. These machines were used on holes up 
to 14" in diameter and 90' centres. 

A double-headed eyebar boring machine was installed in the east end 
of the 60 ft. lean-to, each head set on a heavy bed of sufficient length to 
bore holes 70 ft. apart centre to centre. The large pin holes in the chords 
and connecting plates were swept out 1" smaller than the finished hole 
in the individual plates before assembling. For this purpose there was 
provided two manhole borers each having a capacity to cut holes from 
10" to 45" in diameter and equipped to cut elliptical manholes. 

The best quality of workmanship was demanded. All sheared 
edges were planed; punching was not allowed in nickel steel and only in 
carbon steel not over 11 /16th in. thick. All drilled holes were drilled after 
assembling the members and punched holes were reamed \i of an inch 
larger than the diameter of the punch after assembling. Rivet holes were 
countersunk about 1/16 of an inch to remove the burrs. 

Some difficulty was found in drilling the members after the material 
was assembled together, in a manner that would avoid drill chips or an 
excess of compound working in between the several parts. This was 
successfully overcome in the following manner: in the thick material, 
of which most of the members was composed, stitch holes for bolting 
were selected which would give an area of unfastened plate not over about 
14 inches square. These holes were drilled Y % of an inch smaller than the 
finished hole and after the other holes were drilled and bolted they were 
reamed to the proper size. It was found impracticable to draw the 
plates close enough together by ordinary bolting and after the pieces were 
assembled the member was moved down to the riveting trestles where 
a pressure of about 80 tons was applied through a riveter alongside of 
each bolt as it was being tightened up. The ordinary carbon steel assem- 
bling bolts did not stand the stress of being properly set up and Mayari 
steel was early substituted for this purpose. 

101 




102 



All lattice liars were drilled and planed to jig. In the early part 
of the work the holes were reamed after assembling, but it was found 
that the heavy members were kept better in line by accurately drilling 
the member and the lattice bars separately and this practice was generally 
followed throughout the work. The ends of all built webs of both compres- 
sion and tension members were planed on the rotary planers after riveting 
but before assembling into the member, principally to relieve the large 
planers used for finishing the assembled member. 

The Contractors appreciated from the outset that very perfect work- 
manship would be necessary, it being of the utmost importance to the 
operations in the field with work of this magnitude to have all members 
finished so exactly that connections could be made without difficulty. 

Extreme care was exercised in setting members for machining to keep 
the member absolutely level both longitudinally and transversely. A 
centre line was marked on each outside web and an Engineer's level was 
set up off one corner of the member so that a sight could be taken along 
the longitudinal centre line and across the end. When that end was 
levelled exactly, the level was moved to the other diagonal corner to see 
that no wind existed. After the member was exactly level it was securely 
bolted to the machine bed. The transverse level was further checked 
by running a tool across the member at each end before starting to 
machine it. 

The truss web members are all connected by pins with the exception 
of the diagonals in the suspended span, but the lateral and sway bracing 
was made with riveted connections throughout. All field riveted 
connections were reamed to steel templates, heavily bushed to prevent 
wear. These templates as well as those for the pin holes were checked 
by pinning the templates together on the floor of the template shop and 
great care was taken in applying them to the members. 

The pin holes in the tension members were on the centre line of the 
member and required only the ordinary precautions in laying them off 
and boring them, but the pin holes at the " M " joints of the diagonal 
compression members and those of the bottom chords are in pin plates 
outside the members and great care was necessary in laying out these 
holes on the finished member. 

Figure 20. — Shows a typical "M" joint and the measurements used for 
laying out these holes on the finished member. Before assembling the webs 
of the member together the holes were marked off in the connection plates 
from templates and swept out in a manhole borer about 13^ inches smaller 
in diameter than the finished hole. In laying out the finished boring; 
after the member was truly levelled on the bed, the intersection point 
was measured from the faced end previously finished and the centre of 
each pin hole was then laid out by measuring the distance from the inter- 
section point to the vertical passing through the hole and laying off the 

103 



ordinate. After the pin holes were scribed, careful check measurements 
were made by measuring the distance from the back of each pin hole to the 
intersection point — • the back of the pin hole always being taken as that 
side on which the pin bore. That is to say, in compression members the 
check distance was from the intersection point to the inner side of the pin 
hole, while in tension members it was to the far side of the pin hole. All 
of the distances and angles for laying off the pin holes were calculated in 
the drawing office and shown on the drawings. 

The compression joints were spliced with material and rivets suf- 
ficient to carry the full calculated stress. The holes for the connecting 
rivets were drilled after the joint was assembled in the shop and the parts 
drawn into close contact. 

All measurements in shop and field were made with steel tapes which 
were frequently checked with the standard tape, kept for that purpose, and 
corrected if necessary. 



104 



Chapter V 



FIELD OPERATIONS AND ERECTION 

Transportation by Railway limited the size of some details, such as 
the end connection plates, shoe ribs, etc. The maximum weights and in 
many instances the maximum lengths that could be safely transported 
determined the limit of shop assembly. Sketches were prepared and 
submitted to the Railways which showed the weights of the heavier 
members, their dimensions including connection plates and all data necessary 
for the design of special rolling stock necessary to carry the exceptional 
members. The railways provided steel cars of 150,000 lbs. capacity with 
specially framed bodies to provide for the concentrated loadings and 
with openings in the floor to pass the projecting connection plates on the 
members. 

Plate LVII. — Show typical loading diagrams. 

For convenience in handling the large members in the shop, in loading, 
unloading and in erection, the centre of gravity of each member was 
marked in the shop. In many cases special connections for the hoists 
were bolted on and only removed after the member was in place in the 
bridge. These connections were of various forms to suit the details of 
the member to be handled; some of them are shown on Plate LVIII. 
The positions for the connections were determined in the Drawing Office 
by calculation and verified in the shop. 

(a) Camp and storage at site. 

Crane runways were erected at the bridge site storage yards, each 
having two trolley cranes of 70 tons capacity, 30 ft. lift and 83 ft. span. 
The distance between columns in the runways was fifty (50) feet and the 
total length of each runway 500 feet. The columns were staggered and 
members longer than the crane span were handled through the runway 
by moving the member transversely when passing a column. 

A convenient site for the North storage yard was found within the 
" Y " tracks close to the North end of the bridge, the yard thus connecting 
directly to the bridge and to Sillery Point where the suspended span was 
erected. At the yard site there were also a number of Railway sidings 
which were convenient for the storage of the lighter material that could 
be handled by locomotive cranes. On the South side the nearest point 
available for the crane runway was about a mile from the end of the 
bridge. The location of the storage yards is shown on Plate XIX. 

105 



The equipment for transferring material from the storage yards to 
the erectors consisted of: 

Two 4-wheel yard locomotives weighing 56,000 lbs. each; 

Four Bay City Locomotive Cranes each with a capacity of 30 tons 
at 13 feet radius; 

Eight steel frame flat cars each of 80,000 lbs. capacity; 

Ten trucks for transporting the heavy members to the erection 
traveller. 

At the North yard, well equipped repair shops were established for 
blacksmiths, machinists, pipe men and electricians; also store houses for 
small material, erection gear and paint. 

Bins for rivets and bolts, storage tanks for fuel oil required for heating 
rivets as well as facilities for watering and fuelling locomotives were 
established on both sides. 

An office was placed near the South end of the bridge with accom- 
modation for the South Shore force. 

The working forces and all buildings were connected by telephone to 
the Main Office. 

With the exception of the steam railway equipment, all of the power 
used in the erection of the bridge was electric. A.C. current was purchased 
at line voltage of 22,000 on the North side and 11,000 volts on the South 
side, and transformed to 2,200 volts for the synchronous motors driving 
the air compressors and motor generator sets which transformed to D.C. 
current for use in the traveller motors. 

Steel frame fire-proof power houses were built on both sides of the 
River. That on the North side contained four electrically driven air 
compressors each of 530 cu. ft. capacity, and two motor generator sets 
each of 250 K.W. capacity. The South side power house contained three 
530 cu. ft. air compressors and two motor generator sets. 

The offices and dwellings for the Construction Engineer, the Superin- 
tendent of Erection and their staffs were established at the North end of 
the bridge, as well as a substantially built camp with sleeping accommoda- 
tion for over two hundred men and boarding facilities for a much larger 
number. The camp included a hospital, police station and all the acces- 
sories of an isolated camp — it being about half a mile from the main road 
and three and one-half miles from the electric car line leading to Quebec. 
It was provided with water and sewerage systems, cold storage plant, 
electric light and fire protection. A road was graded to connect it with 
the main Quebec Road. 

106 



Plate LIX. — Shows the general layout of the camp in its relation to 
the North end of the bridge. The five bunk houses were each 24 x 53 ft. 
in plan; the dining hall 47 x 58 ft; the office building a two storey structure, 
28 x 53 ft. with a one storey wing 20 x 20. 

Two gasoline motor boats were employed for ferrying men and light 
materials to and from the South side. As soon as the work permitted, 
electric passenger elevators were installed on each side of the River con- 
necting the shore level with the floor of the bridge. 

(b) Erection Travellers: 

All of the material of the bridge proper between anchor piers, including 
the steel falsework, was erected by means of two inside tower travellers, 
one on each side of the River, having the main hoists suspended from 
trollies which in turn were mounted on electric travelling cranes, enabling 
the hoist to be placed any where within the range of these cranes. 

The general design and principal dimensions of the travellers are 
shown on the following drawings: 

Plate XIII Side elevation and top plan 

" XIV Front elevation 

" LX Stresses and material 

" LXI Deformation Diagram 

The considerations leading to the adoption of this new type of 
traveller were fully discussed in the introductory paper, Pages 29, 30, 
41 and 42. The original design was somewhat enlarged in scope, it 
being found desirable to add to the tender design the four swinging booms 
for handling the sway bracing, falsework and other light material; also 
the 5-ton auxiliary hoists at the ends of the travelling cranes for handling 
the working cages. These additions, with the machinery, added so 
considerably to the weight of the traveller as first designed that nickel 

I was adopted wherever an appreciable saving in weight could be made 
by its use as in the top cantilever trasses and the structural material of 
the travelling cranes. 

The Hoisting equipment of each traveller consisted of four 60-ton 
hoists hung from trollies, two on each side. Four pairs of 5-ton hoists 
also hung from trollies on the ends of the travelling cranes and four 
20-ton hoists on the ends of the 70-ft. derrick booms. 

The weight of each traveller in working order was about 920 tons 
approximately made up as follows: 

107 



Runway Trucks 65 tons 

Bottom supporting trusses and bracing 120 " 

Main tower frame floors and bracing. : 170 " 

Wood in floors, etc 30 " 

Top crane runway trusses and bracing 165 " 

Elevator Shaft and stairway 19 " 

Main travelling cranes 210 " 

Auxiliary Gantry Cranes 36 " 

Derrick Booms and Tackle 50 " 

Wire and manilla rope except for main cranes 8 " 

Electric wiring 3 " 

Derrick hoisting engines 32 " 

Miscellanous machinery on operating floor 3 12 " 



920 tons 



The bottom story of the tower was made 89 feet long to give a long 
base to reduce the reactions and lighten the rear anchorage against up- 
lift, but the length of the tower proper above the lower sway struts was 
limited to 37 feet to permit the sway bracing to be placed between the 
vertical posts last erected before the traveller was moved to a new position. 
This limitation in width caused heavy stresses. The resulting deflections 
were therefore large and as safety demanded there should be no grade 
in the crane runway which would tend to run the cranes off the end of the 
traveller, the trusses were so framed that under the worst conditions of 
loading the runway rail would be level. This, as will be seen, required 
an additional height of about 6-inches at the ends. See Plate LXI. 

The top trusses were designed to form a runway of sufficient length to 
permit the hoists to stand a little beyond the vertical of the furthest ahead 
members to be erected. The bottom chords and the web members were 
built of four angles and a plate in an " H " section. The top chord was 
a box section with a cover plate, — reinforced to carry the rail, — on the 
top flange and latticed on the bottom flanges. All connections were 
riveted. The trusses were connected to the tower by 10" diameter 
pins. 

The width of the tower was limited by the required clearance for 
erecting members between it and the truss members already in place and, 
further, by the economy to be found in using the outside girders carrying 
the permanent track for supporting the rail under the inside truck wheels 
the traveller being supported when moving on a double line of rails on 
each side of the bridge. The outer line of runway rails was supported on 
the main stringer girders provided for the panels of the permanent track 
but not yet erected and temporarily placed under the traveller 6'6" 
outside the permanent stringers. These were braced to the permanent 
girders with temporary laterals in the plane of the top flanges and with 
frames opposite the sub-floor beams in each panel. Special girders were 

108 



supplied for the outer rail support where the panel Length was shortened 
in the end panels of the anchor and cantilever arms, there being no regular 

main stringers available of the proper length. The rails were 85 lbs. 
fastened by dips bolted to the tops of the girders, field holes being left 
for this purpose. 

Plates XIII, XIV and LX. — Show the general construction 
of the trucks support inn the traveller. The centre of the forward 
trucks was placed 9'2" from the centre of the front column and 
the longitudinal distance between trucks was 66'11". The cast steel 
wheels were 33 inches in diameter, bronze bushed, running on 
ti" axles. When the traveller was being moved the two top travelling cranes 
were locked at the extreme rear of the upper runway and in this condition 
the load on one front truck was 486,000 lbs., while that on one rear truck 
was 450,000 lbs. In order to distribute the load equally on the six wheels 
of each truck, the reactions were taken by heavy spiral springs four to 
each axle. Each spring carried a load of about 21,000 lbs. while the 
traveller was in motion but was designed to take a load of 30,000 lbs. 
before it would compress to its solid height. The cross beams of the 
forward truck were connected by 6" pins to an equalizing beam placed 
between the webs of the truss chord. 

When lifting with the cranes in the forward position the reaction on 
a front post was for some conditions about 1,300,000 lbs. and as the 
trucks were only designed for about half of this load the posts were extended 
through the bottom chords, in order that they might be shimmed to a 
solid bearing when the traveller was in operation. Pedestals were provided 
to carry the post reaction directly to the floor beam and as a matter of 
safety the columns were bolted to the pedestals by 2" diameter bolts, 
the pedestals having previously been securely bolted and braced to the 
floor system. The traveller was anchored to the floor girders at the rear 
against uplift and shims were placed to avoid the possibility of over- 
loading the rear truck springs when the cranes were lifting towards that 
end. 

The rail on the anchor arm is on a 1% grade, but due to the framing 
for camber this grade was accentuated in the erection of the cantilever 
arm. The grade also varied from panel to panel as the erection of the 
cantilever arm proceeded and means were provided for keeping the upper 
crane track level by an adjustment in the rear trucks. The load at each 
of these trucks was suspended through a 5" diameter bolt, the length of 
which could be varied by nuts. When necessary to adjust the level of 
the traveller, both cranes on top were moved to the forward end, materially 
decreasing the rear reaction and permitting the adjustment to be easily 
made. 

For moving the traveller along the bridge floor 1}4" wire ropes 
were connected to the runway girders on each side of the traveller. These 

109 



ropes passed through sheaves between the'chords of the lower trusses and 
were connected to six part 1% manilla rope tackles which were hung on 
the rear face of the tower from the brackets carrying the swinging derrick 
booms. The two tackles were rove with a single line the ends of which 
were led to spools on the outside derrick hoists, thus equalizing the pull 
on both sides. This arrangement gave a speed of about 10 feet per minute 
and gave close and satisfactory control of the traveller movement. 

Plate LXII. — Shows the arrangement of the cranes on top. 

Each of the main cranes on top of the traveller was built of two single 
web plate girders 113' 4" long, 5'5" deep and spaced 8}4 feet on centres, 
connected by top and bottom laterals and brace frames for that portion 
between the runways, 54 feet. The working range of the 60-ton hoists 
was from 7 ft. to 21 ft. outside the crane runways. To lessen the move- 
ment of the cantilever extension the trolleys carrying the heavy hoisting 
machines were placed inside and as near the centre as the necessary travel 
would permit. 

The upper block of each main hoist was a trolley carrying six 24" 
sheaves connected to the main trolley carrying the hoisting machinery 
by a spacing strut to control its movement and take the pull of the fall 
lines leading to the drum. The sheave trolley was carried on an auxiliary 
track inside the main crane girders which was supported by girders three 
feet deep headed into a diaphragm between the main girders over the 
truck and into another diaphragm 24'6" further out. The auxiliary 
girders which were spaced 3'11" apart were connected to the main girders 
by brace frames and by quarter inch plates on the top flanges, thus provid- 
ing efficient lateral bracing for the cantilever extension of the main girders 
and for the auxiliary girders. 

The lower block carried five corresponding sheaves. With its con- 
necting shackle it weighed about three tons. A wire cable %" diameter 
was rove through these sheaves making a ten part tackle, both ends of 
the line being carried to the winding drums, greatly reducing the friction 
loss in the blocks and securing a more uniform distribution of the load on 
the different parts of the line. With the usual single hoisting line the 
friction load on the hoisting part when lifting or on the becket end when 
lowering would be doubled and a heavier line would have become 
necessary. 

The maximum lift of the hoist was about 330 feet, requiring 3500 feet 
of cable for each hoist. This was wound on two drums 36" diameter and 
31" wide requiring five layers of cable for the blocks in the upper position. 
The winding drums were mounted on the same shaft which thus carried 
the entire bending load, but in order to give greater strength in the driving 
gears and pinions each drum was separately driven by a heavy cast steel 

110 



gear next the frame One drum was keyed to the shaft to cause rotation 
of the shaft in its bearings, but the other was free on the shaft to prevent 
unequal loads on the driving pinions. 

The controller system was of the magnetic switch type arranged for 
dynamic braking with a maximum lowering speed of 12 ft. per minute. 
The armature shaft had an automatic disc brake capable of holding the 
maximum load. Special fuses were provided which would blow when the 
load reached its maximum, thus guarding against an attempt to lift loads 
greater than calculated for. As a further precaution a hand operated 
brake, independent of all other safety devices, of sufficient capacity to 
hold the maximum load was placed on the extension of the back gear 
shaft of the motor. To prevent over-winding a limit stop was installed 
which applied the brake by interrrupting the current when the block 
reached the limiting position. 

The hoist and sheave carriage had a traversing motion of 14 feet. A 
5-H.P. motor was geared to a rack rail to give a speed of 10 feet per minute 
under maximum loading. The traversing mechanism was made strong 
enough to resist a thrust of 30,000 lbs. due to fleeting loads from the 
centre of the traveller, but was not capable of moving against this pressure. 

In addition to the two main hoists carried by each travelling crane 
there were two light hoists of 5-tons capacity at each end. These hoists 
were supported by gantry frames which spanned the sheave carriages and 
were carried on rails over the centre of the main girder. Each gantry 
supported two trolley tracks having a travel of 20 feet normal to the axis 
of the main crane. The over-hang was not symmetrical, being two feet 
from one support and nine feet from the other. Locking wheels under- 
neath the. top flange of the eye-beams were provided to keep the gantry 
in place and take the uplift from the maximum load at the end of the 
9 ft. arm. The hoisting blocks were single sheave rove with Y% diameter 
wire rope wound on drums placed near the centre of the main cranes. 
The drums on these hoists were 16" diameter geared to give a rope speed 
under full load of 150 feet per minute. The trolleys were racked in and 
out by hand wheels. 

The runway rails carrying the cranes were 100 lbs. A.S.C.E. section 
planed flat on top. Each crane was carried on four two-wheel trucks 
with wheels 4' centres giving a total wheel base on the runway of 12'6". 
The wheels were steel, 30" diameter, 3" tread. The crane was driven 
by a 16 H.P. motor placed at the centre of and between the main girders 
operating a Z]/ 2 diameter squaring shaft leading to each side. The motor 
was geared to give a longitudinal travel of 25 ft. per minute. 

&l The two derrick booms at the rear of the tower were 70 feet Jong and 
hinged on brackets extended diagonally to permit material to be taken from 
the tracks at the rear and passed around the tower. The forward booms 

111 



were on brackets extending in front of the tower. A second bracket a 
storey lower down was provided at each front corner to carry the booms 
when extended to 90' for placing the anchor arm staging. The extension 
was made by inserting a 20' section. 

The booms were built of four 6 x 6 x % angles latticed, the outside 
dimensions at the centre being 3' x 2'6". At the upper end two sets of 
sheaves were provided, one for the four part % steel rope tackle of the 
lifting blocks and one set for the six part % inch diameter tackle used for 
the topping lift. At the lower end the booms were pin connected to the 
pedestal, which was cast solid with a double bull wheel 4 feet diameter 
rigged for revolving the booms through an angle of 230°. 

Electric hoists were placed on the second floor of the traveller about 
18 feet above the track for operating the derrick booms. The hoist drums 
were 15" diam. 31" long, designed for a rope pull of 10,000 lbs. with 
eight layers of % in. wire cable on a drum, power being supplied by a 
51 H.P. motor, 295 R.P.M., geared to give a lifting speed of 37 feet per 
minute under full load, controlled with a drum type controller. Each 
hoist had in addition four spools 12 in. diameter and 15 in. long for manilla 
rope. The spools were loose on the shafts operated by jaw clutches. The 
bull wheel was revolved by a 5 H.P. Motor, 700 R.P.M. placed on the 
third ficor, geared to give a rope speed of 25 feet per minute with a pull of 
6,000 lbs. on the rope. The motor was furnished with a reversible 
controller and was fitted with a mechanically operated brake on the 
armature shaft with which the machine could be locked when not in use. 

All hoisting ropes were of flexible plough steel, 6 strands with hemp 
centres, 19 wires per strand. All gears were of steel with cut teeth. 

Direct current at 220 volts was used in all motors. 

The main and auxiliary hoists on the cranes were equipped with full 
magnetic control and dynamic braking. Controllers for all the operations 
of the cranes were placed on the working platform immediately over the 
railroad tracks so that the operator could at all times see what was being 
done and get orders directly from the Foreman in charge. The controllers 
for the derrick operations, both hoisting and swinging, were placed near 
the hoisting machines, where the operators had a good view of the working 
space on the bridge floor. 

Iron contact bars were used on top of the towers to deliver current 
to the electric cranes, 144 being required of 13^" x jHs section. The 
connections from the contact shoes to the motors were made by copper 
wire as well as the connection from the contact bars to the controllers. 

Figure 21. — Illustrates the contact bars as well as the general 
appearance of the hoists. • 

Except for racking the auxiliary Gantry hoists, operators were not 
required on the upper level but the machinery was constantly under 

112 



* 


j 


— 


~i3| 


«^ 








113 



inspection and at least one man was kept on each machine to oil, inspect 
and apply the emergency brakes should occasion arise. The distance from 
the working platform to the crane girders was about 175 feet and for con- 
venience an electric elevator was installed, operated by a 15 H.P. motor, 
geared for a rope speed of 150 feet per minute. The controller was arranged 
for dynamic control and placed in the car. The elevator shaft was en- 
circled by wooden stairs. 

Field telephones were liberally distributed over the traveller and 
working parts and placed wherever communication could be facilitated 
thereby. 

The switchboards, magnetic control panels, resistance coils, etc., were 
installed in a fire proof building placed directly under the rear platform. 
The building was mounted on wheels running on the top flanges of the 
inside permanent track girders and was coupled to the traveller. The 
feed wires were wound on reels mounted on a platform attached to the 
switchboard house. The platform was also useful to carry shackles, rope, 
small tools and other extra equipment. 

In using the traveller for erection of material special cantilever track 
stringers were placed to extend the working tracks about 20 feet in front 
of the forward floorbeam. Material was taken from the storage 
yard to the work on two special trucks. The forward truck was run out 
on the cantilever track and that end of the member lifted by the outside 
crane until the member was carried by this crane and the rear truck. 
The member was then run forward until the rear crane could be connected 
and the member lifted clear of the rear trucks. 

Special fleeting tackles operated by the forward derricks were provided 
to keep the main crane blocks centred over the railway tracks. Truss 
members were moved forward in pairs, one on each track, and fleeted 
one against the other. When the members were lifted from the trucks 
the cranes were moved forward until the members could be moved trans- 
versely past the forward legs of the traveller. The fleeting tackles were 
then slacked until the falls hung vertically. When in this position the 
fleeting tackle was cast off and the members moved into the positions 
required by means of the crane and trolley motors under the control 
of the operator. If the truss members were too long to be fleeted past 
the forward end of the traveller in a horizontal position, the forward end^ 
was raised until the member cleared. 

Piute LXIII. — Shows the manner of rigging one of these tackles. 

Transverse members such as floor beams which could not be handled 
in pairs were loaded on trucks in such a manner that the centre could be 
carried beyond the forward legs and hoisted by the two 60-ton blocks 
from the forward crane, which were coupled together and connected to 
a special device. The lifting hook was provided with a ball-bearing 

114 




o 

o 

•00 

c 
c 



•00 

c 

o 

js 

09 

JS 

a 



11. j 



swivel to permit the floor beams to be easily turned when suspended 
(See Fig. 22). Laterals, sway bracing and other light transverse members 
were handled directly with the derrick, sling chains being used for connect- 
ing to them in the customary way. 

The lower supporting trusses of the traveller were erected with loco- 
motive cranes and the upper stories by an ordinary wooden guy derrick 
operated by a two drum steam hoist. The derrick mast was placed in the 
centre of the tower, special provision being made for suspending it as it 
was moved up. The crane girders on top were raised in two pieces and 
spliced at the centre with finished bolts when in place. 

The traveller on the North side was erected on shore during the winter 
of 1913-14. To insure a level foundation and that the traveller should 
be in the proper position to run on to the Bridge, four spans of plate 
girders were set on concrete foundations at the end of the approach spans. 
The girders were later used to carry the outside tracks for the traveller 
at the ends of the cantilevers as well as in other places during the progress 
of the work. 

(c) Erection of Approach Spans: 

The 107 and 151 ft. spans at the North end of the bridge were erected 
during the Summer of 1913 by locomotive cranes. Timber staging was 
used under the short end span and steel staging for the longer span reaching 
to the anchor pier. 

It was necessary to move the traveller from the position in which 
it had been erected on shore across these spans in order that it might reach 
the anchor pier — the position where it was first put in use. When the 
spans were completed, temporary brackets were attached to the outside 
trusses to support the girders carrying the outer rails of the traveller 
tracks. 

Plate LXIV. — Shows the method of supporting the temporary girders. 

The approach spans were two independent single track structures 
and the traveller loads were applied off the centre line of each span. To 
prevent the eccentric loading causing serious lateral stresses and to dis- 
tribute some of the traveller load on the inside trusses, the two spans 
were connected together by temporary struts opposite the brackets. 
Special clamps were also used to connect to the girder under panel K 3. 
This girder had been used as the top of the staging bent during the erection 
of the approach span. 

(d) Steel Falsework and Staging: 

The loads on the staging due to the weight of the structure and the 
erection equipment were so heavy that timber could not be used and 
steel construction on concrete foundations was necessary throughout. 

116 



The staging was arranged in two distincl parts. The first part called 
" [nside Staging " was designed to carry the floor and the traveller. The 

ad part called " Outside Staging " was placed on separate foundations 

directly under the main trusses and was designed to carry the entire 
weight of the anchor arm including the floor. 

It was planned to use the inside staging of the North shore for the same 
purpose on the South shore. It was necessary to leave the outside staging 
under the trusses of the North shore anchor arm until the cantilever arm 
had been built out so as to fully balance the weight of the anchor arm. 
Two sets of outside staging were therefore provided for the bridge but 
only one set of inside staging, to be used first on the North shore and 
then altered as necessary for use on the South shore. 

Plate LXV. — Shows the layout of foundations for both inside and 
outside staging with their relation to each other, also the arrangement of 
anchor bolts. 

Inside Staging: 

Plate LXVI. — Shows the general outline of the staging in elevation 

and section together with the stresses and material in all members. 

It will be noted that vertical posts were used in the upper storey of 
the bents, the posts being under the centres of the traveller legs, but the 
posts were battered below a plane which cleared the lateral bracing in the 
bottom chords of the trusses. This arrangement seemed to be the only 
way of providing room for the laterals and sway bracing with safe clearance 
while at the same time adopting a uniform construction for the bents and 
gaining a sufficient base for stability. 

The clearances between the staging and the bracing of the trusses 
wire in many places so close and at so many different elevations that to 
avoid any possibility of error it was considered prudent to construct a 
wooden model of the staging and of the anchor arm on a scale of }4 inch 
to the foot. A model wooden traveller was also built and the complete 
model used for instructing the foremen in the order of erection and the 
method of handling the members. 

Figure 23. — Shows this model. 

Bents 1 and 3, 7 and 9, 11 and 13 were connected by longitudinal 
bracing to form towers but bent 5 was braced longitudinally by struts 
connecting to the other towers. It will also be observed that temporary 
struts were used between 9 and 11 to stay bent 11 during erection. 

Bents 1 to 11 rested on concrete pedestals but bent 13 was located 
between the caissons of the old and new piers and it was necessary to 
carry it on heavy box girders which spanned from the old pier to a concrete 
seat on the new caisson, These girders were designed to be used later 
for supporting and lifting the ends of the suspended span. 

117 




■3 



x 

a 



8 



US 



The vertical columns of the bents were extended to a height that 
would permit t he columns of the traveller to be shimmed directly over 
them. Brackets were attached to the columns of the bents to support 
a longitudinal girder on each side. These girders were spaced 5' apart 
and were braced together in pairs. The permanent floor beams of the 
bridge rested on these girders but sufficiently removed from their final 
position to permit the vertical truss members to be erected without inter- 
ference, the connection of the floor beams to the trusses not being made 
until the truss was completed, as will be later explained. 

The floor stringers and floor system of the bridge were laid on the 
floor beams, special provision being made for the temporary shifting of 
the points of supports and where, as in the end panel, the stringers were 
notched out for the floor beams, short bents were used to carry the 
stringers. The outer temporary stringers supporting the traveller tracks 
were also placed as required and braced to the outer permanent stringers. 

Plate LXVII. — Shows the general arrangement for the first two 
panels which are typical of the construction throughout. 

Plato LXVIII. — Shows the location and displacement provided for. 

To allow the floor beams to be brought to a position for connecting 
before the truss had taken its final form, clearance was provided above 
the tops of the longitudinal girders and steel shims were inserted between 
the floor beams and these girders. Shims were also placed between the 
floor beams and track girders supporting the floor and traveller track to 
provide adjustment for the track grade and to give ample clearance above 
and below the floor beams when adjusting for attachment to the trusses. 

The inside staging was designed for use again on the South shore 
after the North anchor arm had been erected. The ground elevation was 
generally higher on the South side and short sections were used in the lower 
columns, the removal of which adjusted the bents to the heights of the 
South 6hore foundations. Sway brace connections were provided for both 
conditions on the same gusset plates. The sway bracing was fabricated 
the proper length for use on the North side and was cut off with a hack 
saw in the field to the required length for the South side, both sets of con- 
necting holes having been punched when the members were fabricated. 
The top and bottom sections were used a third time to form part of the 
ring for the suspended span during erection and were detailed with 
this purpose in view. 

Outside Staging: 

Plate LXIX. — Shows the general arrangement of the outside staging 
and 

Plate LX X - Is a stress and material sheet of this staging. Each bent 
was a self-supporting tower built of four columns 6 ft. centres well braced 

119 



on all four faces and having horizontal bracing. Each column was 
anchored to the masonry with two 23^ inch bolts. 

Temporary shoes were bolted to the bottom chord at each panel point, 
the ribs of the shoe bearing directly on the chord ribs. These shoes 
rested on pins carried in the tops of the outside staging towers — the 
connecting pins being in two lengths each only long enough to take the 
bearing of two ribs of the chord. 

Floor beams provided for the cantilever arm were temporarily used 
between bents 6 and 8, 8 and 10, and 10 and 12, to carry timber bents 
for supporting the bottom chords at the mid-panel splices. These floor 
beams were braced together in pairs and supported on brackets fastened 
to the columns of the staging. 

The columns were connected in pairs near the bottom by heavy box 
girders of sufficient capacity to transfer the load to the hydraulic jacks, 
used for adjusting the height of the bents. The bents were set to provide 
for camber as already noted, the point L-2 being 5 inches below its final 
position. The further adjustment required after the chords were riveted 
was provided by about 4 inches of steel shims placed under each leg. 
The shims were divided on the line of the anchor bolts, half holes being 
punched so that they could be easily taken out or put in place. The 
anchor bolts were made long enough for vertical adjustment and some- 
what heavier than required by the vertical up-lift to guard against any 
possible displacement during the operation of jacking. 

The North traveller was moved into position to begin erecting the 
inside staging on the 20th May 1914 and the staging and floor for the 
North anchor arm were completely erected by the 20th of July with the 
traveller in position to place the main pedestals. 

(e) Erection of North Anchor Arm: 

The scheme of erection starting from the base formed by the pedestals 
on the main piers and projecting both anchor and cantilever arms from 
this base required that the shoes should be set exactly as to elevation, 
distance apart and alignment : 

The bridge seats in the main piers had been dressed with extreme 
care to a perfectly plane surface at exact elevation. Copper plugs were 
leaded into the masonry on which were marked the intersection of the longi- 
tudinal axis of each truss and the transverse axis of the pier with the 
four edges of the cast steel base of the shoe. Each of the bases was 
20 feet wide, 26'4" long and 4 ft. high, made up of four separate castings. 
When assembling the shoes and bases at the shop great care was exercised 
to maintain the pins in their exact relation to the edges of the castings and 
marks were placed on the castings which corresponded to the marks on the 

120 




121 



plugs set in the masonry. The surfaces of the masonry and of the castings 
were so nearly perfect that it was considered the bedding could be best 
made with only a thin wash of cement. 

The pedestals were fastened to the masonry by twenty-eight (28) 
anchor bolts and sixteen (16) dowels, all 3" diameter. The holes for 
these were drilled in the top of the masonry with a 4" Calyx drill to a 
depth of 5 feet. The bolts and dowels were set to a steel template antl 
grouted in with neat cement before the castings were set in place. When 
the castings had been set and carefully checked for position, the space 
between the anchor bolt and the casting in each hole was filled with lead 
and all nuts turned down tight. The dowels projected into the castings 
through holes cored in the bottom. After the shoe was assembled the 
castings were filled solid with concrete. Having placed the base castings 
the main portions of the shoes were assembled and the connecting holes 
brought to match. Before proceeding with the erection of the truss suf- 
ficient rivets connecting shoes and bases were driven to insure against 
any displacement while handling the heavy pins and chords which con- 
nected to them. 

The pins and sleeves connecting the anchor arm bottom chord were 
placed in position in the manner illustrated on Fig. 24. 

Erection of Bottom Chords: 

Owing to their great weight these heavy members were shipped from 
the shop in quarters, being divided at mid length and mid width. Field 
splices in each chord had been arranged to keep the weights within the 
capacity of the traveller and of transportation. 

The total weight of one panel of chords 12-14, including the details 
was 392 tons, but by shipping it in four parts the maximum weight of a 
single piece was 92.5 tons. A temporary bracket was attached to the 
shoe under each rib to support the ends of the half panel, L 13-14, abutting 
against the pin and to hold them in close contact with the connecting 
sleeve. The other end was supported on a bracket extended from bent 
13 of the inner staging leg, a temporary wooden shoe being bolted to each 
rib of the chord section and adjusted to the correct height by shims. The 
plates for the centre splice were placed in position but only loosely bolted. 

The two sections 12-13 comprising the second half of the panel 
including the detail at 12 were then placed resting on column 12 and 
supported at the inner end by the splice plates. When making this 
connection the end of the chord of bent 13 was kept slightly below its 
normal position to insure the upper edges of the joint being in contact. 
These edges were drawn together by means of heavy bolts set up against 
buck angles bolted to the top flanges of the abutting pieces. The joint 
was then raised by jacks placed under each rib until the surfaces of the 
adjoining pieces were in close contact throughout when the splice plates 

122 



were thoroughly drifted and bolted for riveting, the shims being adjusted 
to hold the member in position. The bottom laterals connecting from 
panel point 12 to the shoes were connected and the alignment of each 
chord cheeked. 

For erecting the first section of chord L10L12 the traveller was placed 
at bent 11. The inner end of the section was supported by the splice plates 
while the outer end was carried on a timber bent on top of the temporary 
floor beams spanning between staging bents 10 and 12. 

Riveting was not started until this panel had been completely 
assembled and adjusted to insure that no displacement would take place 
at the first splice; thereafter riveting generally followed about a panel 
behind the assembling. 

Panels 6-8-10 were handled and supported in the same manner as 
10-12, the bottom laterals keeping the work in line. 

In panel 4-6 the chord being lighter, the centre splice was riveted 
in the shop before shipment and a middle support was unnecessary. 

It has been explained under the description of the design that each 
truss was laid out as to independent half trusses for the sake of ease in 
erection and to assure a proper distribution of stress in all parts of each 
member. Until the joint of panel 4 was reached the chords were handled 
in half sections as above and connected only by tie plates, care being taken, 
however, that the two ends were always in the same plane. The tie plates 
connecting the two half trusses were not riveted until the chord was 
completely assembled from end to end. 

The detail at L2 being very heavy and of complicated construction, 
was made in one unit without longitudinal division and was field spliced 
to the bottom chord 2-4. The section of 2-4 between field splices was 
thus without the heavy details required for panel connections and the 
area of the member being comparatively small it was possible to handle 
the whole section in one piece. After opposite chord sections were in 
position the splices were adjusted and the laterals connected. It was 
found in each instance that the abutting ends were in actual and close 
contact over their entire surfaces. When the bottom chords and lateral 
system were completed, the erection was stopped until the bottom chord 
splices were completely riveted and supporting towers for 4, 6, 8, 10 
were jacked down to the elevation shown in the Diagram, Fig. 2, of Plate 
XLIX which corresponded to the calculated deformations of the panel 
points produced by the framing for no load conditions. 

While the riveting and adjusting of the bottom chord was proceeding 
the traveller was being returned to bent 13 from which position was 
started the erection of the web members comprising the lower triangles 
of the truss. 

l.'.'i 



Erection of Lower Web Triangles: 

The erection was planned to avoid the use of temporary staging 
above the bottom chord with the exception of a short strut used to support 
M13L14 and a strut to stiffen the tension eye bars at M2L2. 

Plate LXXI. — Shows the position of the traveller and gives the 
order of procedure for erecting the first panel M.12-L.14. The vertical 
member M12L12 was first erected. It was handled full length but as its 
weight was about 116 tons it was divided longitudinally in two parts in 
the same manner as the chord sections. 

The inside half of the member was first erected in order that it might 
be braced, to stand alone, by timbers clamped on each side and connected 
to the projecting ends of the floor beams. The outside half was only 
placed after the sway bracing had been erected and connected to the inside 
half, and was then permanently connected by the tie plates and the 
connections to carry the pins supporting the floor beams. 

Plate LXXII. — Shows the temporary supports and brackets used in 
placing M13L14. The 42 ft. plate girder used as a strut was one of 
those built for a platform on which to erect the traveller on shore. It was 
set upon jacks and braced normal to the chord. At L14 a temporary 
tension connection was made with riveted steel brackets to provide 
against a possible up-lift from wind force. The member M12M13 was 
also erected in two halves, each weighing about 72 tons. The splices 
at M13 were only bolted until the pins at M12 were driven. After the 
splice was adjusted to a true bearing by means of the jacks and shims 
under the ends of the temporary struts it was fully riveted before 
the erection was proceeded with. 

The vertical M10L10 was erected and temporarily braced in a 
similar manner to M12L12. 

Plate LXXI 1 1. — Shows temporary wire rope supporting tackles 
which were first made use of in this panel. The drawing shows the method 
of attachment and the manner in which they were used in a number of 
panels. The tackles were provided with a special stopping-off device 
attached to the floor beams by which the load could be held in any 
position. A tackle was provided for each longitudinal half of a truss, 
making four in all. 

After the erection of M10L10, two of these tackles were attached 
to the ends of the pins at M10. The inner half of the member M11L12 
was stood on its pin at L12 and temporarily supported by one of the 
tackles. Possible uplift was provided against by making a temporary 
tension connection to the chord. After the sway bracing was connected 
for lateral stability the outer half of the member was placed and held 
up in a similar manner and the tie plates connecting the two halves bolted 
in place. 

124 



The sub-member MiiMu was then put in place, the pin holes being 

brought to register by the supporting tackles which were released as 
soon as the pins were driven. 

The upper half of the diagonal MioMu was then erected; first 
the inner ami then the outer half, and supported by the adjusting tackles 
attached near Mio. The alignment was adjusted by the tackles and the 
splice riveted up. After the splice was completely riveted the member 
was sprung by the tackles as necessary to adjust the distance between 
pins LlO and Mr: and the pins at the latter point were driven. The 
lower triangles of the remaining panels were erected in the same way. 

Plate LXXIV. — Shows the method of erecting M2L2. This member 
was made of eight eye-bars with temporary steel columns, to take 
compression during erection, placed between them in such a manner that 
they could be readily removed after having served their purpose. One 
column was provided for each half truss and the four bars clamped to it 
with the proper spacing, as shown on Figure 4. Plate LXXIV. They were 
then handled in a horizontal position to place and the pin L2 driven 
after which they were revolved to a vertical position and temporarily 
guyed in place. 

M I.; was placed and held in position with the rear crane while 
the pin at M3 was driven. Temporary wind connections were made to 
the bottom chord and the sway bracing placed. 

The diagonal M2M3 was then placed and held back by tackle while 
the splices were being riveted up. It was afterwards adjusted and 
the pin driven at M2 when the temporary guys were removed. 

The diagonal M1L2 was erected in one piece without splice and 
connected to the sub-hanger M1L1. 

The compression web members were temporarily connected at the 
bottom to the chords to resist wind forces. 

While the traveller was in this position the anchor bars were placed. 
The lower lift of bars below A2 were connected to the anchor girders 
and braced while the masonry was being built and a well was left for the 
bars above A2. These bars were made in four lengths and the lengths 
arranged to bring the Lo pin in a position where it might be supported 
on the masonry, thus carrying the weight of the string of bars while the 
upper connection at Uo was being made. 

The bars above the Lo were toggled to provide for the position 
of L2 being 5 inches below the normal, for facilitating the connection 
and to control the straightening of the bars as they took their 
load with the extension of the cantilever. 

125 



Plate LXXV. — Shows the arrangement of the toggle and the false- 
work for supporting it. 

An eyebeam grillage was placed across the well in the anchor pier to 
support the pins at Lo and a timber tower was erected to hold the bars 
in position above the pier. 

The bars below the pin Lo were kept parallel by means of plate 
links connecting the pins. There were 24 of these bars in each corner 
arranged in groups of six on one pin. Three lengths of each group were 
lowered to the ground and coupled. When assembled in this way they 
were hoisted by the rear crane and dropped into position through the 
grillage on which the point Lo was afterwards supported. The elevation 
of this point was adjusted to allow the two lengths to hang with a little 
slack to avoid the possibility of stress due to contraction in cold weather. 
The adjusting arrangement for the toggle was then laid on the timber 
tower and the upper portion of the tower completed. The upper lengths 
of bars were placed in groups of six and connected at Mo. 

The post M2U2 was erected and held in a vertical position by the 
forward crane until the sub-diagonal M1U2 was placed and connected 
at each end. The supporting tackles were then attached to M2U2 
as shown on Plate LXXV to support M2U0 which, with its U detail, was 
placed in one piece. The splice at Mi was adjusted by means of the 
supporting tackles to a true bearing throughout and completely riveted 
before further connections were made. 

TJ4M4 was next erected in one piece, and held in position by 
temporary brackets and bolts on the shore side and a wire rope guy adjusted 
with a steamboat ratchet on the other side. See Plate LXXIV. 

The tension diagonals M2U4 were erected and pins driven at each 
end; the pin hole at U4 being elongated to allow a small adjustment in 
the distance between U2 and U4. 

The top chord panel TJ2U4 was then erected. The bars were 
packed in two groups in their supporting trusses at the yard before being 
brought on to the bridge — one pair of trusses and a group of bars for 
each half of the main truss. 

Brackets were placed on each side of the vertical posts to support the 
trusses carrying the bars, and the trusses were bolted to these brackets, 
oblong holes were used to avoid any chance of stress being carried through 
the supporting trusses. Connections were provided in the frames for hold- 
ing the ends of the bars in a straight line at the proper spacing. 

Top chord panel U0TJ2 was placed in a similar manner and pins 
driven at Uo and TJ2, the adjustment of the distance between Uo and U2 
being made by the temporary holding tackle supporting M1U0. 

126 



The toggle in the anchor bars was then adjusted to register the pin 
holes in the upper ends of the bars with those in the diagonal. After this 
connection was made the timber tower was removed and the portal bracing 
place. 1. 

The erection of the succeeding upper triangles to the point Uio 
was similar except where the members were too heavy to erect in one 
piece, as in the case of the vertical post UsMs which was divided 
longitudinally, each half being held by wire rope guys and temporary 

brackets. 

The top chord panel U6Us was too heavy to be handled in two 
parts complete and only the bars inside the supporting trusses were packed 
in the yard, the bars on the outside being sent out singly and slipped on 
the ends of the pins. The oblong holes in the eye-bars, before alluded 
to, made this operation comparatively easy, the camber being so adjusted 
that the top chord panels were one inch short. 

The traveller was then moved to bent 11 and the sway bracing placed 
in U10M10 when the work of raising steel on the North shore was 
suspended for the season of 1914. 

The floor beams from F3 to Fs were then connected to the main 
trusses and the weight of floor transferred from the inside staging to the 
trusses, thus releasing a portion of the staging which was removed and 
shipped to the South side during the remainder of the season 1914. 

Soxith Approach Span: 

The South approach span was erected during the season of 1914 
on timber staging with an ordinary bridge erection car. 

The South traveller was erected over the anchor pier in position 
to start setting the staging and was ready to begin erection of the South 
anchor arm on May 20th, 1915. The erection of the South anchor arm 
was similar to that of the North and will not be referred to again. 

Continuance of North Anchor Arm Erection: 

The work of raising steel was resumed on the North side April 15th, 
1915. The length of the vertical post at 12 being about 121 feet, it could 
not readily be brought through the traveller and passed to the outside. 
Each half, moreover, weighed about 65 tons, exceeding the capacity of 
one main hoist on the traveller. A field splice was introduced 53'7" 
from the bottom end and the lower portion handled in one piece. The 
upper section was handled in two parts. 

The diagonal tension members M10U12 were 137'3" centres of 
pins, and for convenience in transportation and erection a field splice 
was made at about one-third the length. To erect these members the inside 
part was fleeted through the traveller in two pieces laid down on the floor 
beams outside the traveller and the splice riveted. The lower end of 

127 



M12U12 was then placed in position, guyed to M14 and connected 
by temporary brackets and bolts on the shore side as shown in Fig. 1, 
Plate LXXIV. 

The diagonal M10U12 was then raised, the pin driven at M10 and 
rested on the top of the lower portion of M12U12 already in position 
to which it was connected, thus giving a firm support for the vertical 
member. 

The upper outside half of the vertical member was then placed and 
spliced to the lower portion, when the outside tension member, which had 
been spliced on the floor, was raised complete and pins driven at M10 
and U12, after which the inside half of the vertical member was fleeted 
out and suspended by the forward crane. The rear crane was connected 
to the upper end of the diagonal which had been supported on the lower 
portion of the vertical member. This diagonal was then swung up until 
the upper portion of the inside vertical member could be placed and the 
splice connected, when the diagonal was revolved about pin at M10 until 
the upper holes matched those in the vertical, permitting the connecting 
pin at U12 to be driven. The vertical member was then lined up and 
the splice riveted. The top chord U10U12 was erected in the same manner 
as U8U10, the outside bars being placed after pin U12 was driven. 

(f ) Erection of Main Posts: 

After the completion of panel A10A12 the traveller was moved to 
bent 13 to erect the main vertical posts. The bottom sections of the posts 
were tapered to carry the loads to the lower pin, being 8 ll' wide at the 
top and 5'9%" at the bottom. They were handled in two parts of two 
webs each weighing 60 tons, or the full capacity of one main hoist. The 
pieces were placed on the sleeves and pins through which they bore on 
the shoe, carefully plumbed by a transit and held securely in position by 
temporary brackets shown on Plate LXXII. 

The next section above consisted of four main parts or corner posts 
and four minor parts made up of the lattice connecting the main sections. 
Each main section was 54 feet long and weighed 50 tons. The two 
sections next to the anchor arm were placed first and supported in a 
vertical position by temporary struts, one for each piece, connecting 
to the compression diagonal (See Plate LXXII). The minor parts contain- 
ing the lattice were then placed and bolted to the two sections erected, 
when the two remaining main sections were erected and connected to 
the lattice parts, the tie plates connecting the four sections being after- 
wards bolted on. 

The third section was similar to the second and was handled in a 
similar manner. After erecting it the lower sway bracing was placed 
between the columns, the floor beam erected and the main floor system 
laid to F14. 

128 



The middle sections Mn were 36'6" long and carried the details 
for connecting the Bub-members. After this section was completely 

mbled tlic sub-diagonal MisMu was erected and the pin driven at 
Mm, but the lower end at Mi.; was not permanently connected until 

later. 

The horizontal member M12M14 was provided wit h a screw conned inn 
and a heavy ratchet near Mi; so that the length could be adjusted as the 
work progressed and it was connected at both ends. 

Sections 5 and 6 were similar to Sections 2 and :! and were handled 
in the same manner but no longitudinal or transverse bracing was required 
other than the lattice and tie plates connecting the four main parts of each 
section, which were quite sufficient to give lateral support. 

The top section was similar to the bottom section but inverted and 
was Ik .red for a 30 inch pin without sleeves. 

The top link l'u was made of six ribs and divided for erection into 
two parts of three ribs each, each part weighing 70 tons. The side eleva- 
tion was 16 ft. square and the width of the cover plates connecting the 
three ribs was 5'5" each. Each half was carried through the traveller 
on the pit cars which were used for shipping it from the shop and raised 
by the two main hoists on one side of the traveller for which special con- 
nections were provided on the member. The 30-inch pins were placed in 
the links before lifting, making it only necessary to raise the member to 
its position, the connection being made at the top when the pin was 
dropped into the half pin hole in the top of the post. 

Plate LXXIV. — Shows arrangements for holding the top link in 
position. 

The three ribs were connected by transverse diaphragms which were 
directly over the outer edges of the posts. Angles were connected to the 
bottom of the diaphragms and to the top of the post with a space between 
for the adjusting shims required to keep the top link in place and prevent 
rocking on the pin. After the link was properly adjusted the connecting 
angles were bolted together. As a further measure of security, a tension 
connection was made between the outer edge of the top link and the 
cantilever side of the post to balance the. weight of the anchor arm top 
chord and tension diagonals. 

The allowance for deformation required that the top of the main post 
should be 14 inches shorewards from the vertical, as shown on Plate 
XLIX, but to maintain and check the alignment during erection it was 
erected in a vertical position and not tipped back until all splices and 
connections had been thoroughly bolted, when it was easily adjusted 
by means of the ratchet placed in M12M14. Before tipping the post 

129 




130 




131 



back the temporary brackets at the bottom and the struts connecting 
to the second section were necessarily removed. When in position the 
pin at M13 was driven. 

The main posts having been set to the calculated inclination, the 
tension diagonals M12U14 were erected and connected, the procedure 
being similar to that employed in placing M10U12. The top chord 
panel was then placed in the same manner as U10U12 and the panel 
completed. 

(g) The Erection of the Cantilever Arm: 

The design of the truss permitted each panel and sub-panel of the 
cantilever arm to be made self-supporting as the work progressed but, 
as already noted, equipment was provided of only sufficient capacity to 
lift one quarter of the weight of the heaviest chord panels. This neces- 
sitated some form of temporary support for the sections of bottom chords 
until the splices could be made and the full panel connected up. 

Plate LXXVL — Shows a movable platform provided for this 
purpose. 

Plate LXXVII. — Shows the links and the manner of their 
attachment for supporting this platform in the different panels and 
Plate LXXVIII shows the manner of handling the platform from 
panel to panel. 

Plate CXVI. — Shows the stresses in the platform and in the 
supporting links. 

The platform consisted of two longitudinal girders for each truss, 
spaced 15 ft. apart and connected with cross girders and lateral bracing; 
the platforms on each side being further connected by cross struts and 
lateral bracing, making a unit of the whole construction. The supporting 
chains were made of plate links, V2}i" by 1" and 1^ inch of various lengths 
and provided with a number of pin holes to enable the different lengths 
required to be readily made up. The links were packed in pairs and special 
bales were provided for handling them. They were connected at the 
upper end to heavy box girder saddles which were temporarily fixed in 
convenient places to erected parts of the bridge and moved forward as 
the work progressed. The longitudinal thrust due to the inclination of 
the links was taken by a hinge thrusting against a bracket connected 
to the shoe in the first panel and in the succeeding panels against brackets 
connected to the under side of the bottom chord. The space between 
the two girders on each side was planked to make a convenient working 
platform. 

A 60-ton jack was placed under each rib of the chord at the splice in 
the centre of the panel and a 100-ton jack under each rib near the end 

132 



of the panel. These were used for adjusting the alignment and after- 
muds for adjusting the lu-iirli t of the end to allow the pin to be 'Irivcn 
completing the panel. See Pigs. 25 and 26. 

The platform was erected by first assembling on the main pier that 
portion from the heel to the outside of the cross bracing, about 55 feet, 
swinging this into position, and after the traveller was moved to M15 
placing with the traveller the outer sections, ( \s, 28 ft. long. 

After placing the inner section of tb • platform, as shown on Fig. I, 

Plate I.XW 111. the pins and sleeves connecting the bottom chord to 
th- shoe were placed in the s me manner as the pins on the anchor side, 
and th ■ two half B ct ions, LlfiLu, of the bottom chord were lowered to 
position and supported at th ir outer ends on the cross girders of the 
platform. The half panel of bottom [at srals was th n placed, the outer 
ends being blocked up on the cross girders of the erection platform. 
The lower ends of th- compression diagonal, MisLie, were then placed, 
the inside half first, and supported by the wire rope supporting tackle 
used in similar cases on the anchor arm. 

Plate LXXIX. — Shows the manner of using these tackles in all the 

pan. 

The sub-member, MuMu, was placed with the derrick booms and 
connected, thus supporting the compression diagonal and releasing the 
holding up tackles. The sway bracing MisLie, the subpost M15L16, the 
floor beam F16 were then placed and the floor was completed to F15. 

To give lateral stability temporary sway bracing was placed below 
the floor beam F15 and the floor laterals were supplemented with extra 
bracing. 

Plate LXXX. — Shows the extra bracing, also the special pedestal 
used to carry the traveller reaction to the floor beam and the method of 
anchoring the traveller. 

The material erected having been thoroughly bolted the traveller was 
moved forward to Fl6, 

Plate LXXXI. — Shows the members erected with the traveller 
standing at F15. 

The erection platform was then completed by placing the outer 
28 feet and supporting this outer end by chains connected to saddle 
near Mr. on the compression diagonal. 

Chord section LuLui was then placed in two parts, carefully aligned 
and the middle splice riveted. 

The upper half of the compression diagonal MisMis was placed next 
•lid supported by holding tackles. The outer half was erected first 

133 



and thoroughly bolted to provide lateral support. When the inner half 
was erected the tie plates were connected and sway bracing placed below 
the floor level. The alignment with the lower half was carefully made with 
the holding tackle and the splice riveted up, when the sub-member 
M14M16 and the vertical M15F15 supporting this member in the centre 
were connected. The member M14M16 had a sliding connection at 
16 so designed that the member could take no stress and would adjust 
itself to the varying distance between M16 and M14 as the work 
progressed. 

The long diagonals Mh and U16 were fleeted through the traveller 
in two lengths for each half truss, assembled on the deck and riveted, 
then raised in one length to position and the pin at U16 driven. To 
make the M14 connection it was necessary to spring the compression 
diagonal 2.68 inches from a straight line. This was accomplished by 
means of the holding up tackles supporting it and required an additional 
pull of about 24,000 lbs. The hole in the tension member at M14 was 
oblong to simplify the adjustment of the distance between the con- 
necting holes. 

The vertical tension member M14L14 was handled next in (wo parts 
and full length. Each half was fleeted through the traveller and sup- 
ported from the upper pin hole by the outer crane in its extreme forward 
position. This allowed the member to be hung vertically about 7 ft. 
beyond its position in the truss. The lower end was then drawn in to 
connect with the bottom chord at L14 and the pin driven. A runner 
was placed at the outer end by which the member was revolved about the 
lower pin until the holes at the upper end registered. The upper holes 
were oblong to facilitate the adjustment of the distance between Mm and L14 
which was accomplished by means of the hydraulic jacks on the erection 
platform. The sway bracing, the floor beam F14 and the traveller 
track were then placed and the floor laid to Fh. The track girders 
only reached to a point half way between the two floor beams at F14 and 
a temporary extension bracket was bolted to the ends of the track girders 
to distribute the load on both beams. 

The vertical post U14M14 was placed next. The member was handled 
in three parts; first the lower part with four ribs complete was set on the 
pins M14 and secured against lateral movement by the temporary 
anchor bolts on the outside and on the inside by plates bolted to the detail 
of the diagonal. The outside half of the upper end was then placed, the 
splice bolted and guy lines attached. Afterwards the inside half was 
erected and guyed in a similar manner and the tie plates connecting the 
two halves bolted on. 

The top chord panel U14U16 was erected, the pins driven at U16 
and the supporting trusses bolted at both ends, thus holding the vertical 
post in position and allowing the guy lines to be released. The sway 

134 



bracing between the main posts, excepting the top strut, was then placed 
and the inside link supporting the outer end of the erection bridge was 
disconnected to allow the bottom laterals to be completed to Ln. 

The chord pins at In could not he driven until the chord U12I11 
was erected and as the support of portion of the succeeding panel caused 
stresses in I'mIim a temporary connection was necessary. 

Plate I. XX XII. — Shows links provided for this purpose. The links 
rested on the top chord supporting trusses and were attached by tem- 
porary connections bolted to the tops of the posts. The panel length 
was adjusted for driving the pins by means of the hydrauhc jacks. The 
links were so arranged that they could readily be moved forward from 
panel to panel as the work progressed. 

The erection of the succeeding panels to the panel point C4, was a 
repetition of the foregoing operations, but the shorter lengths and the 
lighter sections towards the end of the cantilever arm permitted a simpli- 
fication in some cases. 

Plate I. XXXIII. — Shows the members erected with the traveller 
standing at Fl4. 

Plate I. XXXIV. — Shows the members erected with the traveller 
standing at Fix. 

Panel Cud was erected with the traveller standing at CF4, the 
overhang being sufficient to cover the entire panel. 

Plate LXXXV. — Shows the members erected with the traveller 
standing at 2. 

Panel C0C2, 65 ft. long, was erected with the traveller at CF2 
in the following manner: The bottom chord L1L2 was connected by its 
pin Ls and supported at the outer end by the erection platform. See 
Plate LAX V 111. The sub-vertical Mild and floorbeam Fi were then 
placed and the floor, including the lateral and traction connections, 
completed to Fi — all being supported by the erection platform. 

The sub-member Mi 1*2 was then connected by its pin at U2 
and allowed to hang vertically alongside U2M2, M1L2 (weighing 
67 tons.) was erected and connected at L2 when the member M1U2 
was swung out and connected at Mi, thus supporting M1L2. 

I Mi (weighing 46 tons) was then carried out in a nearly vertical 
position by the forward crane and connected by a temporary hinge on 
the bottom of this member to a similar temporary connection on the 
top of M1L2, when the supporting hoist was slacked off and the member 
allowed to revolve on the hinge until in alignment with the lower part. 
It was then held in position by the holding up tackles until the splice 
was riveted. 

135 



The top chord panel U0U2 was then placed and the pins driven at 
U2, when the adjusting links were slacked, transferring their load to the 
top chord U2U4. 

The adjusting links were shifted to U0U2 to spring the member 
TJ0M1L2 until the pins at Uo could be driven. The distance necessary 
to spring the member to make this connection was 1.03 inches, requiring 
a pull on the adjusting links of 53,000 lbs. 

The pin at Ml was then driven, the jacks on the outer end of the 
platform under Li being used to make this adjustment. The erection of 
the cantilever arm was completed by placing the sway bracing U0L2 and 
hanging the eye-bars U0M0. 

The erection platform was then dismantled and shipped to the south 
side to be used in the erection of the south cantilever arm. 



The traveller was also dismantled and the lower portion sent 
Sillery to be used for erecting the suspended span as later described. 



to 



The North cantilever was completed November 15th, 1915, when the 
work of dismantling the traveller was commenced, but before the forward 
derricks were taken off they were used for placing the upper girders from 
which was later suspended the apparatus for raising the suspended span. 

The following table shows the progress of erection of the cantilever 
arm: — 



Traveller at 
Panel Point 


Date 


Days 

Required 

Each 

Move 


Days 
Required 
Full 
Panel 


Approx. 

tons of 

Steel 

Erected 


Main Post 14-16 


June 5 
17 

July 13 
19 

Aug. 11 
16 

Sept. 2 

8 

18 

20 

30 

Oct. 4 

9 

25 








15 


12 subpanel 

26 

6 subpanel 
23 

5 subpanel 
17 

6 subpanel 
10 

2 subpanel 
10 

4 subpanel 

5 

16 




950 


14 
13 


38 


1,993 

828 


12 
11 


29 


1,732 
452 


10 
9 


22 


1,450 
447 


8 

7 


16 


968 
493 


6 
5 


12 


747 
288 


4 

Last Move 2 




9 
11 


538 
610 
636 











136 



The total dead weight of the cantilever arm was approximately 

a .'100,000 lbs., and of the anchor arm 21,000,000 lbs., and when the 
cantilever arm was completed with the traveller in its last position there 
was an uplift on the anchor Pier of about 720,000 lbs. As the erection of 
the cantilever arm proceeded the elongated holes in the tension members of 
the anchor arm gradually came to a bearing and, as the weight of the 
cantilever arm balanced the anchor arm, the load on the outside staging 
was released. 

The first lifting movement was observed in panels 10, 8 and 6 when the 
traveller was standing at panel 11 of the cantilever arm just after the 
bottom chord section LioLn had been put in place. 

Table 11 shows the vertical movement of the north anchor arm, in 
inches, as the erection of the cantilever progressed. 





12 


10 


Panel 

8 


Points 
6 


4 


2 


Aug. 19 

" 24 

Sept. 1 

" 3 

" 10 

" 15 

Oct. 1 

" 8 




1/16 

1/16 

1/16 

3/16 

3/16 

5/16 

3/8 


1/8 
5/16 
5/16 
5/16 

1/2 

1/2 

3/4 

13/16 


3/16 

3/8 

3/8 

7/16 

11/16 

11/16 

1-1/16 

1-5/16 


3/16 
5/16 
7/16 
7/16 
3/4 
7/8 
1(1/8 
1-7/16 




1/16 
3/16 
3/16 
1/2 
9/16 
1 
1-5/16 





1/16 


1/4 

5/16 

7/16 

11/16 



When the cantilever arm was fully assembled and the traveller removed 
the cantilever and anchor arms practically balanced over the main pier, 
but the staging at AL2 was left in to prevent any possible negative 
'ion on the anchor bars due to wind or moving loads. The toggles on 
the anchor bars were slacked off and removed and the towers at L2 were 
jacked until the gauges placed on the jacks measured a reaction of about 
200,000 lbs. when shims were driven under the tower bases in this position. 

The vertical motion of the trusses due to the balancing of the cantilever 
arm was very small at AL12, and it was only after the suspended span was 
connected that the staging under this point could easily be released. As it 
was not required for use elsewhere it was left in place. 

The intermediate * taken out, shipped to the shops and cut 

into short columns to be used at Sillery this work being done during the 
winter of 1915-16, for use in the early spring. 

137 



The south anchor arm was erected during the season of 1915. The 
work was started May 20th and finished, including the erection of the main 
post, on November 8th. 

When the floor system of the south anchor arm had been connected to 
the trusses releasing the inside staging, part of it was shipped to Sillery to 
be used again in the erection of the suspended span. 

The erection of the south cantilever arm was started in the spring of 
1916, and completed on July 31st. The apparatus for lifting the suspended 
span was then installed before the traveller was removed on August 29th. 

( h ) — Erection of the Suspended Span at Sillery: 

Plate LXXXVI. — Shows a general plan of the Sillery site and the 
position of the falsework carrying the span in relation to the double track 
main line connecting the site with the north shore material yard about 
three miles away and with the car ferry to the east by which connection 
could also be obtained to the south shore material yard. Due to the direct 
connections with the storage yards and the small amount of traffic on the 
main line, only limited siding capacity was required. 

A trestle was built to connect the falsework for carrying the span 
with one of the sidings laid down. The trestle was constructed with the 
84 ft. plate girder spans used on the inside staging of the anchor arm and 
timber towers. It was floored with ordinary track ties, but having a 14 ft. 
tie at frequent intervals projecting to carry a foot-walk, air pipes and electric 
wires. 

Plate LXXXVII. — Shows the staging for carrying the suspended 
span during erection. The foundations for this and for the connecting 
trestle were prepared during the season of 1915. The bottom was a prac- 
tically level shale rock foundation, and most of the footings were made by 
simply chipping the rock to a little below the required level, the correct 
elevation being obtained by levelling up to grade with cement mortar. 
Anchor bolts were grouted into the shale which was moderately hard 
below the exposed surface. The site was only dry for a few hours at low 
tide, and, while the work was easily done, it extended over a considerable 
period on account of the short working time. 

An office and small power house was built during the season of 1915, 
in which were installed one of the 250 K.W. motor generator sets used at 
the north end, to supply direct current to the traveller motors, and one 
of the air compressors to supply compressed air for riveting. 

The towers at the four corners ( Plate LXXXVIII ) were designed 
as pedestals to carry the full weight of the span when swung and any 
lateral forces that might occur from wind or temperature. The heavy box 
girders forming the top of these towers and providing support for the span 
were those used for the purpose of supporting bent 13 under the anchor 
arm. 

138 




l.'I'.t 




.fi u 



ga 



u*0 

o ? 

J3 3 
+"• » 

GJS 

%" 

o 

si 

93 

a 



140 




141 



The timber blocking between the under side of the spun and the top 
of the columns was built up in a special unit for each position and prepared 
in advance to facilitate placing it rapidly as required. See Plate LXXXVII. 

Sand jacks for swinging the span were placed under all main panel 
points 2, 4, 6, 8, 10, 12, 14 and 16, those at 8 and 10, being of extra size. 

The depth of the suspended span at the centre being only 110 feet and 
the weight of the heaviest member to be handled 66 tons, it was possible 
to use the derrick booms (increased to 40 tons capacity) for erecting the 
span and the cranes and crane-runways were not required. 

The tackles on the derrick booms were increased in capacity by the 
addition of sheaves to the blocks but the booms themselves as originally 
made were of ample strength. The derrick hoists and swinging gear used 
when erecting the cantilevers were again used at this location. 

Figure 27. — Shows the traveller assembled and setting the first 
bents of staging. 

As it was necessary to remove the traveller from the span before 
floating it, the whole operation was facilitated and time saved by erecting 
the traveller on a special tower outside the span from which it could start 
to set the staging as it proceeded across the span and to which it could be 
returned when the erection was complete. 

From its position on the temporary tower the traveller placed the 
corner towers at the point Lo staging bents at 1 and 2 and the 
longitudinal bracing connecting them. Floorbeam FBi was then 
placed and stayed by the lateral bracing which was of a box form 
3' 9" deep, the verticals connecting to the floorbeam were erected 
and the temporary floor laid, so that the traveller could be moved to 
bent " 1." In this position it erected the bottom chords, FB2 and the 
lower half of the diagonal truss members. The succeeding panels were 
erected in the same manner as the traveller was moved across the span, the 
upper half of the diagonal truss members, the top chords, the top lateral 
bracing and the sway bracing being left for the traveller to erect as it 
was returned panel by panel to the west end of the span. 

Figures 28 and 29. — Show the span being erected. 

The first member of the span was placed on May 22, 1916, and the last 
pin was driven July 21st, 1916. The riveting was begun as soon as the 
traveller was sufficiently advanced and closely followed the erection. 

The top chords had a full bearing on the pins at U2, XJ4, U6 and Us and 
were without hinge plates. The camber blocking was set to make the top 
chord panels a half inch longer than the shop lengths and the pin holes in 
the bottom chord bars were elongated half an inch except the four end 
panels which were in compression when floating. In placing the floor 

142 



beams the distance between panel points was made l/8th of an inch short 
of the shop length of bars, which greatly facilitated the driving of pins and 
allowed for any change in the length of panels due to temperature. 

Plate LXXXIX. — Shows the deformation of the truss and the 
displacement of the panel points for the different conditions while being 
erected. 

In swinging the span the blocking at the sub-panel points 1, 3, 5, 7 and 
'.i was lir>t removed and afterwards the blocking on the inside columns of 
the staging at points 2, 4, 6 and 8 when the weight was carried on the sand 
jacks at these points. The sand was then removed from jacks at 2, 4, 6, 
leaving the entire weight supported at and 8. During this operation the 
top chord joints at l T 2, U4 and U6 closed and the opening at Us increased. 
The slack in the bottom chord between and 8 was taken up causing a 
horizontal displacement at 0, 2, 4 and 6. This displacement was provided 
for in the construction of the sand jacks and in the detail at Lo. When the 
sand was removed from the jacks at the points 8 there was both horizontal 
and an gular displacement at Lo. The movements are set forth on Plate 
IX XXIX, but the arrangements for meeting them will be referred to 
more fully when describing the loss of the suspended span in 1916, which 
was due to the failure of a portion of this joint. 

After the span was supported on the four corners the four points Ls 
were all brought to exactly the same level and the holes connecting Ls — 
Ms were reamed and riveted thus insuring an equal distribution of the 
live load stresses between the diagonal members in the centre panel. All 
other web members had been riveted while the span rested on the false- 
work, but the top chord and end post splices were completely riveted when 
the span rested only on the four corners. The laterals and sway bracing 
were thoroughly bolted but not riveted before floating. 

After swinging the span all but the centre bents of the falsework were 
removed to permit the scows to be fixed in position for floating it. 

( i ) — Equipment for Floating, Mooring and Hoisting the Suspended Span: 

Plate XC. — Shows the position of the scows and the method of 
loading them. The span in condition for floating weighed approximately 
5,300 tons. All the scows were blocked to the same elevation and were 
thus equally loaded under the panel points Li, L2, L3 at each end. The 
load was distributed over a length of 25 feet at each end of the scows 
through a grillage made of four longitudinal floor girders to be used later 
as part of the floor of the suspended span which were well braced together 
and rested through wooden blocking on steel cross bulkheads. This 
jrrillage was in turn loaded by an eye-beam grillage made of eight track 
stringers on which the floor beam was carried by tightly driven steel shims. 
Each scow as placed was sunk on a foundation prepared for it by opening 
the valves in the bottom and was securely blocked in position. The 

143 



scows rested while at Sillery on timbers bolted to concrete foundations 
placed vertically under the longitudinal trusses of the scows and accu- 
rately levelled. 

The three scows at each end were connected together by four girders 
made of the inside staging posts, the girders and connections being calculated 
to meet the wave conditions shown on the diagram (Plate XCI). The 
cross girders were connected to the scows by means of channels fastened to 
the cross frames and extended up above the top of the girders to carry 
saddles. The girders were tightly wedged in place between these saddles 
and the top walling. See Plate XC. In addition to this connection the 
scows were well lashed together around the side bollards and diagonal 
wire ropes were run from corner to corner to take up any possible longi- 
tudinal shear between the scows. 

Each scow was tested for tightness as it was placed by closing the 
valves and allowing the tide to exert the lifting effect of the full depth of the 
scow, a load about 50% in excess of that it would normally be called upon 
to carry when floating the span. The tests were made before the cross 
girders were connected to avoid any danger of straining these through 
compression of the blocking under test. 

Plate XCI. — Shows a general plan of the scows, the various 
conditions of loading assumed, the stresses and the material. The load 
of the span was carried by the stringer grillage directly to the steel bulk- 
heads which in turn transferred it to the three longitudinal trusses 
placed 10' 6" apart. These trusses were estimated to carry all the 
longitudinal stresses, no reliance being placed upon the planking which 
was regarded only as a skin to keep out the water. 

With a view to making the scows saleable for commercial work after 
serving their purpose in erecting the bridge the framing was calculated to 
meet the five conditions of loading shown on the bottom of the Plate. The 
width of the scows was fixed at 31' 11", it being desired to give them the 
full width of the panel with only sufficient clearance to insure placing them 
accurately. Six 8" valves were placed in the bottom of each scow and 
operated by gas pipe keys extending through the deck. These were left 
open while the scows were in place at Sillery to allow the tide to flow in and 
out without exerting any lifting effect until it was desired to float the span 
when the valves were closed. 

The draft of the loaded scow was limited by the requirement of having 
the valves in the bottom sufficiently above the level of low tide to be sure 
that the scows would be thoroughly drained before it was necessary to 
close the valves for floating the span. From the tide tables it will be seen 
that the elevation of 83 meets this condition for four or five days at each 
spring tide. The shoal outside Sillery Cove over which it was necessary 

144 



to float the span had a maximum elevation of 82 and from these two 
ooBsiderations the top of the foundations on which the scows rested was 

fixed at an elevation of 83. 

When moving the span it was desired to float the scows about two 
hours before high tide to give ample time to reach the bridge site and make 
the connections before the reversal of the current. A draft of 8' 2" or a 
surface elevation of 91.2 met this condition. The length of scow required 
to give the necessary displacement on this draft also gave sufficient end 

lity against wind forces or possible wave action from any seas likely 
to be encountered -a study of such records as were available indicating that 
the wave length was about 40 feet with a maximum depth of 4 feet from 
• to hollow. The scows were made 11' 0" deep, giving 3' 4" of free- 
board under probable conditions of floating. The freeboard was sufficient 
to provide for the possible contingency of one scow being injured and two 
scows at an end having to carry all the load. The trusses were also investi- 

I for this possible although unlikely contingency. 

The steelwork for the scows was manufactured at the Lachine shops, 
but they were assembled and planked at a Sorel shipyard. 

Before describing the floating and hoisting of the span to its final 
elevation the arrangements for anchoring it in position at the bridge site 
and the apparatus for hoisting it will be taken up. The hoisting arrange- 
ments used in 1017 only differed in a few details suggested by the experience 
of 1916 from those used in that year, and the description will generally be 
limited to the apparatus used in 1917. The anchoring arrangements were 
the same on both occasions. 

Much consideration was given to securing the floating span in position 
at the bridge site. The strong and varying current, together with the 
cross currents which existed at certain stages of the tide, particularly near 
the turn, made it necessary to have ample clearance in placing the span to 
avoid the danger of the steelwork of the span fouling the anchoring or 
hoisting gear. Having placed the span in position it was necessary to 
hold it vertically under its permanent place in the bridge against any prob- 
able wind under varying conditions of current ami elevation of surface 
while the hoisting chains were being attached and until the scows had 
floated out leaving the load suspended by the chains. 

Plans for anchoring to the ground were considered both for connecting 
the cables to the supporting scows and for connecting the cables to inde- 
pendent scows anchored on either side of the span which would act as 
docks between which the span might be moored. The water is, how 
about 200 ft. deep, the range of tide about 20 ft. and the wind surface about 
11,000 sq. ft., making it impracticable to place anchors of sufficient holding 
power up and down stream. Neither did it seem practicable to hold the 

145 



span in exact position by means of lines running to the shore on account of 
the great length required, the pull of the current upon the lines and the 
elasticity of any arrangement depending on submerged cables. 

Plate XCII. — Shows the plan finally adopted, a heavy cantilever 
frame hung from floorbeam 1 of each cantilever arm. The frame was 
calculated to resist a pull from the current of 65,000 lbs. and a wind of 18 
lbs. per sq. ft. acting on the span, giving a total horizontal force of 300,000. 
It was built of inside staging legs placed 54 feet apart centres, braced toge- 
ther and stiffened in the longitudinal direction by a truss. The frame was 
connected to the floorbeam at the upper end by hinges to allow it to be 
swung out of the way while the span was being guided into position. It 
was held in the longitudinal position desired by two heavy wire rope 
tackles leading to electric winches, all of which were obtained from the 
60-ton hoists of the erection traveller. 

At each lower corner of the frame there were independent sheaves for 
two 134" plough steel anchor cables hinged to revolve in a vertical plane 
through an angle of nearly ISO degrees. The anchor cables were attached 
to the span by loops slipped over steel bollards bolted to the end of the 
span and were led around the sheaves to eight part steel tackles, the falls 
of which were carried to electric winding engines previously used for the 
derricks on the traveller but later placed near the ends of the cantilevers. 
The lower blocks of the anchor tackle were fitted with guides running on a 
wire jack stay to prevent twisting or fouling. 

A working platform was provided at the lower end for carrying lines 
and for general convenience when anchoring. Connections were also 
fitted for tackles to hold the span against horizontal forces as it was being 
raised. 

Plate XCII. — Also shows the rigging and tackle for pulling the hoist- 
ing chains out of the way when placing the span — the dotted line 
indicating the curve of the chains when so drawn back. 

Plate XCIII. — Shows the general arrangement of the apparatus at 
one of the four corners for hoisting the suspended span. The upper side 
elevation shows the shoe of the suspended span in its lowest position and 
the lower elevation the shoe raised to a point where the permanent con- 
necting pins were driven. The permanent suspension eye-bars supporting 
the suspended span from the end of the cantilever arm were in two lengths, 
the top half being erected and hung from the end of the cantilever while 
the lower sections of the bars were pinned to the end post of the suspended 
span and raised with it, the final connection being made at the centre. 

The whole hoisting apparatus was suspended by means of pins from a 
heavy girder carried on a rocker bearing on the centre line of the truss, thus 
securing adjustment in all directions for the hoisting appliances. The 

146 



platform carrying the lifting jacks was built of two plate girders connected 
together by heavy diaphragms. The outer diaphragms supported the 

hydraulic jacks while the lifting chains passed between the next set of two 
which were made of sufficient Btrength to cany the load in the lifting 
chains and transfei ii to the side girders. The girders were riveted to the 
outside of the suspension frame hung from the rocker girder on top of the 
cantilever. These hangers were well braced together so that the jacking 
platform, the suspender and the rocker girder above could swing laterally 
only as a unit. 

The lifting girders were practically duplicates of the supporting girders 
with corresponding diaphragms for the hydraulic jacks to thrust against, 
but they were free to slide on and were guided by the suspenders carrying 
the jacking platform. All of the lifting was performed by the hydraulic 
jacks, but as a measure of safety and to permit the renewal of a packing or 
any portion of the hydraulic equipment which might fail in operation, four 
12" screw jacks shown on Plate XCIV were provided at each corner and 
during the operation of lifting, these jacks were always kept in contact with 
the lifting girder although exerting no pressure thereon. The jacks were 
supported in brackets built out opposite to the diaphragms carrying the 
suspension pins. They were geared together in pairs and operated by 
means of hand wheels from the level of the platform. The screws were 
counterweighted, shown on the drawing, and a counterweight was also 
attached to a drum on the gear shaft tending to keep the screws always in 
contact with the top bearing. 

Each lifting chain was made of seven links of 4 — 28" x 1}4" plates 
about 27 ft. long. The links were connected together by 12" pins, 24' centres, 
the end connecting pins being half way between the holes for the pins 
connecting to the jacking girders which were placed every 6 ft. 

The system was arranged for a lift of 2 ft. in each jacking operation 
and three holes were placed in the diaphragms at 2 ft. centres so that a hole 
might always be brought to register with the holes of the links which were 
6 ft. centres. The lower hole of the hoisting chain was elongated to 5 feet 
to facilitate the connection to the suspended span and the chains were made 
sufficiently long to connect to the span while afloat at the lowest stage of 
the tide. It was expected that the connection would be made on a falling 
tide and it was, of course, most important that no chain should take a load 
until all had been connected and, furthermore, that they should all be 
loaded simultaneously and equally. 

As the pins connecting the links came clear of the upper girder, the 
links were taken off and laid on the floor of the bridge. 

Plate XCV. — Shows the rigging of the tackles for taking off the links. 
It also shows the platforms from which the jacking pins were operated, a 
detail of the jacking pin in place and the method of supporting these pins 
from counterweights so that they might easily be shifted from hole to hole. 

147 



It will be noted that the suspension line was given a lead which would tend 
to release the pin when it was without load. A trigger was placed on the 
pin to prevent this action, also to automatically insure that the pin was in 
its proper position before transferring the load to it. 

Plate XCVI. — Shows the arrangement of the hydraulic jacks and 
the piping. The rams of the jacks were 22 inches in diameter and the 
working pressure about 3,700 lbs. per sq. in., giving each jack a capacity 
of about 700 tons. They were tested, however, for a load 33% in excess 
of working capacity by pinning the lifting and jacking girders together and 
working the hydraulic pumps to a gauge pressure of 5,000 lbs. Rocker 
bearings were provided at the top and bottom of each jack to prevent any 
possibility of eccentric loading. The four jacks at each end were operated 
by two direct acting double plunger pumps operated by compressed air 
connected up as shown on the diagram. There was a double controller 
valve opposite each corner and opposite this valve a telltale showing the 
comparative level at each jack so that the operator stationed at this valve 
could regulate the flow of water to each jack and easily keep the girder 
practically level. A similar valve was placed at the centre of the span 
and opposite to it a telltale showing whether the two corners of the span 
were level. The operations of jacking were generally controlled by these 
valves, those at the corners being used merely as auxiliary to keep the 
jacking girders themselves level. A signal system was arranged across the 
span so that all corners were kept practically level at all times. 

Plate XCVII. — Shows the arrangements for holding the span when 
afloat and for launching it from its berth to a position where the tugs could 
be attached to it. 

Part of the floating programme was to choose a tide that would float 
the span as near as practicable to daylight in the morning so that there 
might be ample time for making the connections and lifting the span to a 
position where it could be safely left through the night, it being considered 
impracticable to carry on the work by artificial light. This necessitated 
closing the valves in the bottom of the scows somewhere about midnight 
at a time when the morning weather is difficult to foretell and it was felt 
that if conditions at daylight proved unfavorable for moving there should 
be guides and anchors to force the scows to settle back with the falling tide 
on the foundations provided for them in exactly the old position. A 
timber framework was built around the outside corner towers to act as 
end guides and timber shims were fitted between the scows and the frame- 
work with only sufficient clearance to allow the scows to rise and fall with 
the tide. A steel framework was placed at the shore end of the centre 
scow of each end set which prevented the span from moving shorewards 
and the scows were anchored to the framework by a tackle led to the hoisting 
engine of sufficient power to hold or to pull the span against the assumed 

148 



wind force. All of this temporary framework was estimated for a wind 
force of 5 lbs. per sq.ft., it being considered thai any greater pressure would 

be foretold by tin \\< ■■< orological Office. 

sfosl of the regular Btringers which had been used as traveller tracks 
in erecting the Bpan were taken off to form grillages on the scows, but those 
in the L1I.2 Panels were left in place and a heavy timber platform was 
built between the two inside stringers to carry a steam hoisting engine, the 
engines being required for pulling t he span out from its berth, for adjusting 
the tackles and for taking up the three-quarter inch wire ropes with which 
.n was first connected to the hanging frames. 

Plate XCVI1I. —Shows the method of guiding the span from 
its berth to a position in the River where there was sufficient 
depth of water to permit the large tugs to be attached to it. 
The operation was undertaken at strong flood tide, requiring an 
anchorage down stream to guard against the possibility of the span 
drifting against the Billery Wharf, and a wire rope was run to a heavy 
crib anchor about 1240 feet to the eastward. An eddy often occurs at 
flood tide and it was thought prudent to have an additional anchor to the 
westward which was connected to 3" bolts grouted into the rock. Both of 
these anchor cables were connected by rope lashings to the span which 
were cut when the span was in a position where the cables ceased to be of 
service. Springing tackles, led to the engines, were rigged to the cables to 
permit end adjustment while the span was moving out. The span was 
connected to the anchor bents on the shore side by 2" rope tackles which 
were paid out as desired until the scows were clear of the supporting towers 
when the rope lashings connecting these tackles were also cut. 

To keep perfect control of the span and to pull it out from its berth, 
tackles were rigged from the inside corner of the end scows and connected 
to the outside towers, the falls of these tackles being led to the hoisting 
engine. Marks were placed at every 5 feet along the edge of the scows and 
every 5 feet of movement at each end was reported by telephone to the 
Superintendent. 

(k) Floating and Hoisting the Suspended Span, September, 1916: 

The preparations for moving and hoisting the span were practically 
completed about the 1st of September and the high tides near that date 
might have been utilized for the work, but the success of the operation 
seemed so dependent upon every man being familiar with his duties and 
upon team play that it was decided to postpone it until September the 
11th, and the intervening time was spent in practise drills, the operations 
being gone over as far as practicable with all men at their stations. The 
weather promising well on the night of the 10th the operation of closing 
the valves was completed by 12.45, about half an hour before the tide 

ted to rise (see Fig. 16), and an hour and a half before it reached the 

149 



bottom of the scows. At 3.30 a.m., the span was floating on the barges 
with a freeboard of about 3' 4"; it being then quite dark and there being 
ample time to move to the site, work was postponed until 4.40 when there 
was sufficient light, and the hoisting engines started to pull the span out 
from its berth. This operation took about 73^ minutes. 

It will be seen from Plate XCVIII there is a shoal at elevation 80 about 
600 feet outside the span over which the large tugs could not pass, but 
there was sufficient draft for two small harbour tugs of about 300 H.P. 
each, and these tugs were attached to the span as it was moved out of its 
berth. They then pulled it out swinging it around the eastern anchor 
crib until it was past the shoal, when four tugs each of about 500 H.P. 
were connected on the east side of the scows two at each end. A large tug 
of 1,000 H.P. was also connected to the centre of the span and the two 
smaller tugs took up their position on the west end of the barges for the 
purpose of guiding the span endwise or of swinging it. These connections 
were completed about 5.13, when the down stream anchor line was cast 
, loose. 

Plate XCIX. — Is a chart of the course followed by the span 
showing the ranges established for following this course and regu- 
lating the speed of the span up the river. Two large balls were 
hung in the centre of the span from wires connecting the outside 
of the cantilever ends which gave the exact course into the opening, 
but the span could not make this range until it had reached a 
point about one mile from the bridge. Course ranges were established on 
shore at measured distances from which the rate of progress up the River 
could be ascertained and a table was made of the time at which the span 
should pass each range. The motion of the span was checked after range 
No. 14 to see that it was under full control, was allowed to reach a speed of 
four miles per hour between ranges 13 and 12 and checked again almost dead 
at range No. 8. After passing range No. 9, the tugs brought the span to a 
standstill in three minutes. When the span was within about 150 feet of 
the bridge it was held motionless by the tugs and the %" lines wound on the 
drums of the hoisting engines were carried out by boats and connected to 
the frames. Then, with the assistance of the small tugs, the span was 
brought almost into position before the large \\i" lines were connected to 
their bollards. As these lines crossed in a rather confusing way the 
bollards were painted different colors and the loops on the lines were 
painted to correspond. 

The operation of mooring the span and connecting the hoisting links 
took considerably less than an hour and was all completed at 7.30, by which 
time the current had practically ceased. At 8.15 the hoisting chains began 

150 



to take a Load from the falling tide and at 8.51 jacking was commenced; 
at 9.21 the scows drifted out during the third lift leaving the span suspended 
on the hoisting chains. Everything had worked as planned up to this 
point, it was thought all risk in the operation had been successfully over- 
come and thai nothing remained but to carefully jack the span to its final 
position. The workmen had been on duty since midnighl , some all night, 
and, after the span was raised four lifts at the north end and five lifts at the 
south end, they wen- allowed a recess of an hour for breakfast and rest, the 
span at this time hanging about 30 feet above the water. Work was 
resinned again about 10.36, one lift was completed and the upper lifting 
girder was being lowered for the next lift when the span slipped off its sup- 
ports and fell into the river at 10.46. 

The evidence that could be immediately gathered as to the manner in 
which the span failed was most conflicting, but, as far as could be seen from 
the deck of the bridge, the supporting girders on which the span had rested 
were uninjured and the hoisting chains remained intact. The Engineers 
of the Company at once proceeded to examine the supporting girders, 
being lowered down in a cage by the locomotive hoist to each girder in 
turn. 

Figures 30, 31, 32 and 33. — Show the conditions of all four shoes. 
The photos will be better understood by a reference to Plate C showing 
the construction of the shoe supporting the span on the girders. 

On both of the north girders it was quite apparent, as will be seen from 
the photographs, that the castings had simply slid off the lower supporting 
pins, the tap bolts in the keeper plates being sheared off in a horizontal 
direction and all scores and abrasions showing the same movement, At the 
south east corner, the lower pin had disappeared altogether while the centre 
pin had been revolved through an angle of 90 degrees and fallen into the 
seat vacated by the lower pin. The western hanger chain was badly bent 
to the west, indicating that the span had fallen heavily against it in a 
westerly or transverse direction, but the manner of failure was not evident 
until the southwest corner was reached where the initial cause of failure 
became apparent. Here it could be plainly seen from the crushing of the 
keeper plates and the shearing lines on their bolts that the first fall had been 
vertical. The marks seemed to indicate that the span had struck first on 
the north side of the girder kicking it back from under the span thus allowing 

this corner to fall into the river. 

A photograph, Fig. 34, taken at the moment of failure and published a 
few days later, confirmed the opinions formed at the time of examination. 
It is quite apparent from the photograph that the southwest corner fell 
first and that the end floor beam revolved into almost a vertical position 
twisting the southeast corner off the supporting girder and that the two 
north corners were pulled into the river after the south end of the bridge 
had gone down. The torsion naturally destroyed the lateral and sway 
bracing which caused the collapse of the top chords. 

151 




y»*» 



00 

s 
o 

J3 

CO 

« 
hi 

o 

o 

si 

Oh 






152 




C 

% 

o 

00 

JS 

a 

« 






153 




Figure 32 Photograph showing the girder at the S.W. 

corner after the accident. 

154 




tn 






■on 

G 

% 

o 

09 

a 
« 

u 
-OB 
O 

o 

S3 
Oh 



S 



155 




s 



a 
a 
a, 

09 

8 



"OB 

C 

I 

o 
ee 

& 
« 



h 



150 



The impact of the span falling at the aoutb west corner, the partial 
transfer <>f the load at this ami the diagonally opposite corner by t he 
action of the lateral bracing to an axis from the south east, to the north- 
west corners; the aide fall at the southeast corner, the audden release of the 

extra load firet at the southeast ami afterwards at the north corners and the 

parting of the l ' [" anchor cables all combined to cause very heavy vibra- 
tions, both vertical ami horizontal, in the ends of the cantilevers, these 
being bo Bevere as to throw some of those on the cantilever ends off their 

feet. It was feared that the impacts and vibrations might have over- 
stressed some of the material or the joints, and a very rigid inspection was 
al once undertaken. Every joint wa.s inspected, rivets were carefully 
tested wherever excessive strains would be likely to occur and new lines 
ami levels were run over the span. With the exception of the diaphragms 
under the supporting girders on the top of the cantilevers on the south 
and northwest corners where a slight slip in the rivets was noticed, no 
indication of any injury could be found and the new levels and alignment 
corresponded exactly with those that had been recorded before the span 
was attached. All felt quite satisfied after this inspection that it was 
only necessary to build and place a new suspended span. 

The decision to duplicate the span and to use the same methods of 
erecting it was made on the day of the failure. It was thought at the time 

that all of the hoisting tigging would be available with the exception of a 
few pieces known to be injured, such as the lower links of the chain and the 
upper suspension frame, but it was decided to take it down and to carefully 
measure and inspect each individual piece; also to test the links while 
taking them off by pinning the hoisting and jacking girders together and 
putting the full pressure of the pumps upon them, this being about 50% 
in excess of the stress under the act ual load. When the links were measured 
in the yard it was found that a number had been stretched beyond the 
elastic limit and it was decided to test some of these links to destruction as 
well as some of those that had not received any apparent injury. While the 
- did not indicate any danger in using the links again, even those that 
had been stretched beyond the elastic limit, there were difficulties in re- 
matching them and there had been so much criticism of the Company for 
its decision to again adopt a method which, in the opinion of many, was the 
cause of the failure, that it was decided to order entirely new hoisting chains, 
to re-build all of the suspension gear and jacking girders and to make some 
experiments to determine the amount of yielding at the pin bearing in 
wide arid comparatively thin plates. 

Plate CI. — Shows the loads and stresses estimated. The experiments 
indicated somewhat greater strength for thicker and narrower material 
than had before been used and 28 x P. /'links were substituted for the 
30 X 1H" previously used. Sway bracing was introduced in the top 
pension frame and steel castings were omitted wherever practicable, partly 
as a matter of sentiment -a steel casting having been the cause of the first 

1.-.7 



failure. The steel cast shoes and pins unrler the upper rocking girders 
were replaced by a steel rocker bar bearing on heavy rolled steel plates 
at both top and bottom. The greatest change was, however, in the 
method of supporting the span on the lower girder. 

Plate LXXXVIII. — Shows the new bearing adopted. 

It will be seen that during the process of transferring the load from the 
staging to the supporting girders at Sillery, it was necessary to provide for 
a horizontal movement of two inches at both ends and for an angular dis- 
placement of about 5/16 of an inch in the length of the bearing. In the first 
shoe the angular displacement had been provided for by the transverse pin 
and the longitudinal movement had been secured by allowing the lower 
casting to slide on the bottom rocker pin. The expansion and contraction 
of the span while resting on the end supports at Sillery was also provided for 
in the same way, but when the bridge was being hoisted it was necessary 
to have the transverse pin over the centre of the girder in the same vertical 
plane as the pins connecting the chains. As the span floated it picked up 
the girders on which it had been supported through the temporary links 
shown on the drawing marked " C." The pins of these links were carefully 
placed in the vertical plane of the transverse pin and as the girder was lifted 
it automatically swung into its correct position. Keeper plates marked 
" D," which had been previously fitted in the shop were then tap bolted in 
place to keep the girders properly centered as they were being suspended 
by the hoisting chains and the load of the span was again being transferred 
to them. Several theories were advanced for the cause of failure, but the 
most generally accepted was that the cruciform casting "A" had split under 
the upper pin on the shore side throwing all the weight on the river side and 
causing the girder to tip. Experiments were made later with duplicate 
castings poured in the same manner as the casting which failed, which 
when sectioned showed gas pockets and spongy material in the centre and 
the authors have no doubt that the initial failure was in this casting. 

( 1 ) — Hoisting and Coupling in Place of the Suspended Span in 1917: 

Plate LXXXVIII. — Shows the method of supporting the span 
at Sillery and when being hoisted. The load of the span while 
being hoisted was carried by the nickel steel rocker bar shown 
in the sections which was planed to a top radius of 25 inches, 
but while the span rested on the end supports at Sillery, it was 
carried by shoes under the outer ribs of the end post. These shoes 
which were carefully planed and polished slid on bronze plates and allowed 
ample horizontal movement for temperature changes as well as for the 
operation of transferring the load from the staging to the corner supports. 
The supporting girders were suspended by links when the span floated and 
until it was carried by the hoisting chains at the bridge. As soon as the 
span floated and the load was relieved off the side shoes the bronze plates 
and shims were taken out. The keeper blocks holding these were reversed, 

158 



having been pre\ iously fitted in the reverse posit inn to take up the 3 inch 
clearance lie! wren the enda of the Bhoe and the ends of the shims. A key 
was also inserted at the end of the centre rocker bearing, as shown on 
section C-C. The supporting girder was thus firmly held centrally under 
the rocker bearing and perpendicular to the cent re line of the t russ. 

The angular movement which took place when the load was trans- 
ferred from the Btaging to the end supports was provided for by means of 
lead shims placed under the outside shoe bearing. Experiments made for 
the purpose showed that the lead would How under a pressure of about 
5,000 lbs. per sq. in. and the bearing surfaces were proportioned for this 
pressure. The bearing of i lie cent re rocker, which was nickel steel, was also 
determined by experiment on a short section and was made 50,000 lbs. per 
lineal inch. 

The new construction shown on Plato LXXXVIII required that the 
full weight of the centre span should be carried by the centre ribs when the 
span was being hoisted into position, while the outside ribs carried this 
weight when the span was resting at Sillery. It was therefore necessary to 
acid additional diaphragms and t ie plates in the shoe and end post to trans- 
fer these loads. It was also necessary to add a small amount of reinforce- 
ment to the centre rib at the bearing, but otherwise the new span was an 
exact du] lioate of the span lost. 

A change was made in the method of attaching the hoisting chains to 
the supporting girder to save lowering the links through the end strut as 
in the previous case, short links being introduced which brought the con- 
necting pin well above all interference and made it only necessary to swing 
the ends of the hoisting chain into position between these links. 

"ti of the New Spun: 

It was known that the cove at Sillery was subject to heavy ice shoves 
in the Spring and that during the Winter the rise and fall of tide in extremely 
cold weather would impose severe conditions on the falsework and connec- 
ting trestle remaining at the cove. All the bents of the trestle were cribbed 
in and filled with loose stone but, in spite of the precautions, the ice upset 
the heavy steel corner towers which had carried the span and very nearly 
wrecked the tower on which the traveller stood. Repairs were, however, 
completed in time to start erection on May 20th, and the erection proceeded 
as in the previous year, being completed on July 20th. Concurrently 
with the erection of the span the new hoisting gear was put in place and 
arrangements made to take advantage of the spring tides running from the 
13th to the 20th September. The towing fleet was assembled and on the 
night of the 14th the crews were called out and actually started to close 
the valves about 11.30, but bail weather and unfavorable reports from the 
Meteorological Office caused a postponement until Sunday night, the 16th. 
The rang' and time of the tide on Sunday night happened to correspond 

150 



very closely with those of 1916. The span was floated off its end bearings 
about 5.15, and was moved out from the berth at 5.47. At 6.15 the tugs 
were attached and the anchor cable cut, when the move up the River started. 
After the experience of 1916 the trip up the River was made with much 
more confidence, without any of the checking of the previous year, and the 
span was in position for connecting the anchor cables at 7.30. At 8.05 
the lifting chains were let down, but a little difficulty was experienced in 
entering some of these and it was nearly nine o'clock before everything was 
connected, inspected and ready to begin the operation of jacking, which 
started at 9.10. The tide was falling and after three lifts, or 6 feet of rise, 
the scows came free at 10.28 and floated out. As soon as the span came 
clear the men were allowed half an hour for breakfast, and during this time 
the Engineers inspected the bearings at all four corners and found that all 
connections had come to place as intended. 

To provide lateral anchorage in case of need during the hoisting of the 
span, two of the tackles from the top of the anchor lines were connected 
between diagonally opposite corners of the anchor frame and the scow. 
These tackles were shifted up the anchor frame as the span was raised. 

Nine lifts were made on the first day in cycles varying from 13 to 19 
minutes. Jacking operations stopped at 4.40, when the mooring lines were 
set tight and the span left for the night. On the morning of the 18th, the 
first operation was to take off and lower the upper set of links which had 
come through the top jacking girders. This required about three-quarters 
of an hour and the actual jacking only started at 8.16; it was continued 
steadily until 12.02, when the 26th lift had been taken. In the afternoon 
the first operation was to take off another set of links, in which a little 
difficulty was encountered, and it was not until 3.30 that jacking* was 
resumed. Seven lifts were made up to 5.31 and work stopped, when the 
span had been raised thirty four lifts or 6S feet. Oh the 19th twenty-six 
lifts were taken, giving a total rise at the end of the day of 120 feet and 
bringing the span within 30 feet of its final position. The weather had 
been perfect for the operations up to this time, but just as work closed at 
the end of the third day a thunder squall broke from the north and the 
weather was very unsettled throughout the night. On Thursday morning 
it was blowing from 30 to 35 miles per hour from the northeast, with strong 
gusts which probably exceeded the recorded velocity. There was some ner- 
vousness amongst the men and hesitation about casting the moorings free 
to allow jacking to be resumed, but after carefully easing off the tackles it 
was found that the span swung off centre from 5 to 7 inches without appre- 
ciable side sway although there was longitudinal oscillation of about one- 
half inch within a period of 10 to 12 seconds. Jacking was started at 
9.05. After the noon recess only eight feet of lift remained and soon the 
upstanding permanent hanging bars came through the jacking girders. 
After the first lift a short stop was made to inspect clearances and after the 
next, the 73rd lift, there was a delay of 30 minutes for taking down the 
free top links of the hoisting chains and some spacing angles which would 

160 



interfere with driving the pins. The 74th lift was taken very slowly as 
some of the wooden working platforms bad to be taken down, tbe clearances 
inspected and the lower eyebars guided into the spaces between the eyebars 

hanging from the cantilever. It was completed :it 3.10 and was followed 
immediately by the 75th lift, during which the pin holes in the top and 
bottom bars wen- registered. A*t 3.25 the firsl pin was driven. The clear- 
ances wen- perfect, the heavy pins requiring only a few taps of alighl rail 
to drive them home and the last pin was in place at four o'clock. After 
this the jacks were lowered to let the Load come on the permanent hangers 
and to let the Lifting girder bearings come free. The following day, Friday 
the 'J 1st, the permanent eon nee t ions were made between the lateral systems 
of the sus[>ended span and the cantilevers and all temporary connections 
released. 

Plate CII. — Gives a record of the lifting operation, 
(m) — Fitial Operations to Complete: 

When the span was thus made safe, after the continuous work and 
strain of the past ten days, the men generally laid off for about a week, 
but the floor system across the span was laid and a train was crossed on 
Oct. 17th. Locomotive derrick cars and shunting engines had, of cou 

aed before this time, but the regular train was not allowed over until 
the riveting of the floor system had been completed. Work proceeded 

iily with the balance of the steel and the bridge was handed over to 
the Government, opened for traffic on December 3rd, 1917. At this time 
there still remained the concreting of the sidewalk over the suspended span, 
some concrete to be placed in the anchor pits and a considerable amount of 
painting, but none of this work could be carried on during the winter months 
and it was not until August 19th that all was completed and ready for the 
final inspection of the Government Engineers. 

(n) — Erection Details: 

It may be interesting to call attention to Plate LVIII showing some of 
the special erection appliances and connections designed for the work. 

(o) — Accuracy of Work: 

Reference should be made to the extreme accuracy of workmanship 
which contributed so largely to the success of the erection and which inspires 
the confidence that every member of the Bridge is performing the work for 
which it was calculated. All compression joints throughout the bridge 
came to a true bearing over their entire surface before they were riveted 
up; there were no loose tension members and as far as could be judged by 
the ordinary methods of inspection all tension members were equally 
strcODcd. A< an instance of the precision of the shop work, of the field 
measurements and of the field operations, the starting points of the canti- 
levers were the large shoes on the piers, giving a base centre to centre of 

161 



88 feet. When the cantilevers were erected and extended out 580 feet 
the alignment of the cantilevers from both sides of the river was perfect, 
it being impossible to detect with a transit any variation whatever. The 
ends of the cantilevers on both sides of the river were exactly the same 
elevation and the deflection when the weight of the centre span was hung 
upon them was the same at all four corners. Another illustration of the 
accuracy of the shop work is to be found in the large vertical posts over the 
piers; each of these posts was composed of four separate columns riveted 
together in the field after erection. The total length of the post was 310 
feet, made up of seven sections 36 ft. to 52 ft. long. At each one of the 
posts when erected the tops of the columns were absolutely level and the 
tops of all the four posts were at the same elevation so far as this could be 
checked. 



162 



APPENDIX "A" 



SUMMARY OF TESTS — QUEBEC BRIDGE 



< toe of the importanl conclusions arrived at by the Royal Commission 
appointed to investigate the cause of the collapse on Aug. 29th, 1907, of 
the Quebec Bridge in course of construction by the Phoenix Bridge Com- 
pany, was thai the knowledge of the behavior of the very large compression 
members required in a Btructure of this magnitude was not sufficient to 
enable their design to be made with that degree of confidence essential 
to such an undertaking. 

There had been many theories advanced and supported by an im- 
posing array of mathematical calculations, but these, did not in themselves 
give sufficient data on which to design such large members. 

This fact was realized very soon after the failure of the Phoenix design 
and tests were first made on a model of the chord ''AoL," (the failure of 
which was believed to have caused the collapse of the structure) by the 
Phoenix Bridge Company under the supervision of Prof. Burr. This test 
showed a low ultimate unit stress and indicated that the lacing material 
used was not sufficient to develop the full strength of the web material. 

This test was followed by one made for the Royal Commission on a 
model built to represent two webs of the chord A9L, but having the strength 
of the lacing system increased about 50' , . As had been anticipated from 
the manner of failure of the first model tested this member carried a much 
higher unit load and failed in the web instead of in the lacing system as was 
the case in the other test. 

The Board of Engineers appointed to prepare a design for the new 
structure took full advantage of the knowledge gained from the collapse 
of the previous structure and of the subsequent tests when preparing their 
design, the chords of which had a cross sectional area nearly 2 ] 2 times 
greater than the previous design, due to the increased loading and to the 
lower unit stresses specified. After having tentatively decided on the 
details of the compression members they had several models of them made 
and tested to destruction by the Phoenix Bridge Company. These models 

loped a very high ultimate strength and proved that the proposed 
lacing system was adequate to develop the full strength of the section. 

After the contract to build the new bridge had been awarded to the 
St. Lawrence Bridge Company and the detailing of the members well 
advanced, it was decided to make further tests on some of the more impor- 
tant compression members as they differed in many important features from 

163 



the members previously tested. These tests all showed remarkably good 
results and gave a feeling of confidence in the ability of the members to 
carry the loads which would be imposed on them with a reasonable margin 
of safety. 

There were also several tests made on tension members. 

It was at first proposed to reduce the allowable unit stress in the 
horizontal and inclined tension members by an amount equal to the unit 
bending stress due to the weight of the member itself but in view of the 
results of several tests on built-up tension members and on eye-bars this 
was not done as the tests proved that the bending stresses did not apprecia- 
bly effect the ultimate strength of the members. 

Oblong pin holes were used in many of the eye-bars and built-up tension 
members and the ease with which the structure was erected and the pins 
driven is due very largely to their use. Before finally deciding to use these 
elongated holes some careful tests were made which showed that they had 
no important effect on the strength of the members. 

Several other tests were made of minor importance but all of value in 
solving the problems which presented themselves as the details of the 
design and of the erection were being worked out. 

A summary of the more important tests made, and of which a fuller 
description follows, is given below: — 

LIST OF TESTS DESCRIBED- 

1.— Test of Model of Chord AoL. 

2.— Test of Model of Two Webs of the Chord A9L Made for the Royal 

Commission. 
3. — Tests of 16 Nickel Steel Models of the Compression Members of the 

Board of Engineers' Design for the Quebec Bridge. 
4. — Tests on 6 Eye-bars, 28 Tension Members and 12 Compression 

Members. 
5. — Tests of 6 Carbon Steel Eye-bars with Oblong Pin holes. 
6. — -Pin Friction Tests. 
7. — Tests of 8 Tension and 10 Compression Models of Members Used in 

the Quebec Bridge as Built. 
8.— Tests on Lifting Hitch for Compression Members. 
9. — Test of Plates with Pin holes without Reinforcement. 
10.— Tests on 2—28" x 2" and 2—26" x iy 2 " Plates to Ascertain Their 

Ultimate Strength. 
11. — Tests on Links Removed from the Hangers used in Raising the 

Suspended Span in 1916. 
12. — Tests on Plates in Tension to Investigate the Effect of Inclining the 

Plate to the Axis of Load. 
13. — Tests on Lead in Bearing. 
14. — Test on Steel Rocker Bearing. 

164 



1.— TEST OF MODEL OF CHORD AoL 

After the collapse of the bridge the Phoenix Bridge Co., on its own 
initiative, built and tested a model chord having, as far as possible, the 
same relative dimensions as the No. 9 chords of the Quebec Bridge. 
The test was made on November 21st and 22nd, 1907, and was under the 
general direction of Professor W. II. Burr, who reported on the result of this 
in part, as follows: — 

A model chord section was built to a linear scale of one-third of the 
lower chord section '.• of the anchor arm truss of the Quebec Bridge and 

was tested to destruction at the shops of the Phoenix Bridge Company. 

The chord section No. ',» was built of four ribs 54" deep with 
4" x 3" x $ +," double angle latticing. Its area of cross section was 780 

Si), inches. 

All the linear dimensions of the model were exactly one-third of those 
of the full sized chord section, making the area of the cross section (86.526 
sq. inches) one-ninth of that of the full sized member. Each of the two 
interior ribs was composed of 1 - IS" x 5/16" plate, 1 - 18" x 3 i" plate, 
2 - 15-5/16" x 5/16" side plates and 2-2-11/16" x 1 \ [" x 5/16" angles; 
and the two exterior ribs were each composed of 1 - 18" x 5/16" plate, 
2 - 18" x ! [" plates, 1 - 12%" x 5/16" side plate and 2 - 11/16" x 2" x 
5/16" angles. The latticing was a double oblique system of 1 - 11/32" 
x 1 x ' -," angles, with 1 - 3/16" x 1" x Y%' crossing angles at the pane, 
points of the former at right angles to the axis of the member. All of 
these lattice angles had two - 7/24" rivets at the ends of each with the 
single rivet at each crossing of the interior flange angles. 

All the metal used for the main parts of the model column was medium 
steel, but soft steel was used for the rivets. Tensile tests were made on 
both plates and angles and the usual effect of rolling thin metal was 
apparent in the high elastic limit. 

The column was accurately placed in the machine with four fine wires 
stretched throughout its length in the general plane of the upper flanges 
and with two similarly placed in relation to the lower flanges. Longi- 
tudinal timber .scantlings on the two center lines of the exterior ribs, carry- 
ing steel scales at their ends, were used to measure the shortening of the 
column under loading for 16 ft. of its length to 1/128 of an inch. 

Progressive loading was applied in stages of 3000 lbs. per sq. inch of 
- section of column, beginning with an initial load of that value. 
lie end of every other stage of each loading, the column was relieved 
ress in order to make observations in that condition. This program 

165 



was adhered to up to a stress of 21,000 lbs. per sq. inch, when the next 
increment was made 1,500 lbs. per sq. inch after which the column was 
freed of load. 

After the application of each 3,000 lbs., or finally 1,500 lbs., increment 
of loading and upon each removal of loading an accurate series of measure- 
ments for shortening of the column and for horizontal and vertical deflec- 
tion were made. 

Under a stress of 12,000 lbs. per sq. inch one rivet in one of the lattice 
angle was found loose, but toward the end of the test it appeared to become 
less so. Up to the final loading all other rivets appeared to remain in 
good condition. 

After having attained a load of 25,000 lbs. per sq. inch, instructions 
were given to load the column to 25,500 lbs. per sq. inch, but inadvertence 
in signalling to attendants at the pump caused the load to reach 26,850 lbs. 
per sq. inch, at which stress the member suddenly failed. This failure 
was attended by a quick sharp report, and was so sudden that all failures 
of details appeared to be absolutely simultaneous. 

Aside from the raising of scale on the pin plates immediately in front 
of the 12" pins, the collapse of the column consisted in the failure by shear- 
ing of the majority of the lattice rivets at the central panel of latticing and 
of a considerable number of other rivets throughout the length of the 
column in both flanges. There were no permanent* strains or distortions 
of any kind discovered or apparently discoverable up to the loading 
producing failure. All the circumstances of the test indicated that no 
main part of the column was stressed up to its elastic limit; in other words, 
that the entire loading was insufficient to develop more than a part of the 
elastic resistance of the column as a whole, and that if the latticing details 
had been stronger the column would have carried a greater load before 
collapsing. 

It should be carefully observed that as the column lay in the testing 
machine the ratio of its length divided by the horizontal radius of gyration 
was 35, while the ratio of the same length over the vertical radius was 42. 
The column failed, therefore, in the plane of the greatest radius of gyration. 
Furthermore its failure was wholly in a horizontal plane, there being no 
sensible vertical deflection of the failed column. 



166 



PEST OF MODEL 0] TWO fl 1 BS "I THE 
CHORD \ 1. M MM FOB THE R01 \l. COMMISSION 

Extract from the report of the Commission. 

In December the Commission ordered the construction ol 
chord No. - for the purpose of determining the strength of the web of the 
design used in the Quebec Bridge. This chord had a section half thai of the 

chord No. l ; t he number of rivets in the hut ice connection was doubled, 
the section of the lattice bars was increased 50' , and the weak parts at 
their centres were strengthened by the use of a connecting plate. The 
webs were of the same section as the outer web of test chord No. 1. Material 

from the s:ime heals was used in the manufacture of the two test chords. 

This chord was tested at Phoenixville .on .Ian. 18th, 1908, and failed 
under a stress of 37,000 lbs. per sq. inch by buckling in the webs in the 
center bay, the latticing being sufficiently Btrong to fully develop the 
Strength of the webs. The nominal strength of the column was slightly 
less than tin- elastic limit of the metal in the webs. 



167 



3.— TESTS OF 16 NICKEL STEEL MODELS OF THE COM- 
PRESSION MEMBERS OF THE BOARD OF 
ENGINEERS' DESIGN FOR THE 
QUEBEC BRIDGE 

A number of tests to determine the ultimate strength of models of 
some of the typical nickel steel compression members of the cantilever 
arm of the 1758 ft. span Quebec Bridge as designed by the Board of En- 
gineers were made at the testing plant of the Phoenix Iron Co., at Phoenix- 
ville, Pa., during May, June and July, 1910. 

Eight pairs of models were made of the same material and of the exact 
shape and proportion of the bridge members as designed. The material 
was nickel steel containing as a minimum 3.5% of nickel, with the excep- 
tion of the rivets which were of soft carbon steel. 

All the members were tested in a horizontal position. With the 
exception of T6, all the columns had a flat bearing at one end and a pin 
bearing at the other. The pin bearing was so arranged as to practically 
eliminate bending in the pin. The flat end of the specimen abutted 
against the planed surface of a heavy casting-through which the pressure 
of the ram was applied. No attempt was made to counterbalance the 
effect of the weight of the column. To measure deflections fine wires were 
stretched in the general plane of the top flange. To measure the longi- 
tudinal compressions four compressometers of the Johnson Type were 
used, one attached to each side of the test piece. The loads applied were 
measured by a Shaw mercury gauge. 

The make-up of the members and the results of the tests are shown in 
Table 1 which is given here by courtesy of the Engineering News-Record, 
and is taken from an article which appeared in the Engineering Record of 
November 19th, 1910, describing this series of tests. The location in 
the bridge truss, of the members represented by the test model, is shown 
in Fig. No. 1. 

The Board of Engineers were represented by Professor H. M. Mackay, 
of McGill University, who reported on the results of these tests, in part, 
as follows : — 

TiA — The full capacity of the testing machine, 50,134 lbs. per square 
inch, was applied for ten minutes without failure. Then 2 - 2 h /%' holes 
were bored through the web of each of the four ribs, reducing the section 
to 46,335 sq. inches. It failed, with very little general distortion, by 
buckling of the webs at the section where the holes were bored at a load of 
59,894 lbs. per sq. inch of net section or 49,730 lbs. per sq. inch of gross 
section. 

168 



C ro Si 



L. € ng t h 



Actual Area 
Sq./n. 



Elastic Limit 
Lbs perSq/n 



■a Webs ig'x 'Pjs 

f. f/a ts e'* '/*' 
/oL? e'* i'/e'*'/s' 
8 Horz Web'. 4 



■w « K//V*' r* 
r*l..!LJl]..Ll,i 



n a 


„=! J l//£>>V 


\ 


/«&* J 


I 


--«/**** >j 



45,482 



42,309 



4 Webs s'x V'G 
i Cover 2S'i -J'ie 

i Cover - 

A Flats 2'x 'M'J io<g^ 
Section changes from 
56.66 So In toabt ii4.y5oh 



if* 



jIl JljIlj 



///I'- 



n^a 



66 66 
at rrtn 



o : y -is: 



Could not be 
determined as 
section changes 

continuously 



r* n -7'o'- >»• 



Sec tion 



56.66 

at m-n 



Ditto 



a Webs as%i'x"/3a' 
4p2 3 &'x2 3 4'x"/ K ' 
Afiats e 9trU "/se" 



HI 



.■■Pin ie%'* 2'S/ie" 



k- -S7'9 7 />6 X 



37/7/7 



4 Webs iS-X'z 9 /se' 

tLtiWA 1 **/ 
jFlotss'x 3 /*' 
4 Covers a'x %se 
i Cover 7>*'x 3 /ie,' 
slse'xa'- 3 /*' 
eP>ates m**46f 
6Se J x ifg"x 7a' 



)<a4«e *ke»l 
Tjr njif tit Titr 

r n r'-ilr'n 
■jIl Ji JljIl 



■ Pm a"* 



4-0.21 
At Br eo king 
Point 




4.0.2I 
At Breaking 
Point 



aWeboei'x 'W 
a Fiats e'x 'W 
Btee'xa'-sss* 
6Be'xiys"-'/a" 



ii 




im 


Ji 


r'=\ 


ILJI 



-35' 3'- H 



38,556 



\*a'^-a'4*a'r\ 



P.ns't 



2<&' 



Same as 
7\3 A. 



To break Ts Are holes 
a'/s'&were drilled 
in each Web 

To break Ts B-2 holes 
3*0 were drilled 
in each Web 



46.792 

Through 3'/?'? 
Poles 



4-4.371 
Through 3* 

holes 



■* Webs 2i"x 'U' 
A£s'xa , -w-3.a5* 
4 Flats e'x '/*' 
8tsa"xi'/e'-'/a' 



■Pm 9' 



IL 1 



■ adefe' H 



4-5,615 



sWebseo'&e'xZiz" 
■a Top/? e'xe'x 3 />6" 
a Top £a'xi%'x %x' 
4 Top Flats e'x Ve* 
4 Top Covers 4'x '/e" 
i Top Cover 6%'x '/a* 
aBottomLss'iS'x Vie' 
e Bottom Cov n'/e"x'/a" 
€ Ce" *er&e'si-%"x %sz' 






1%. \<- 
.T BE 



■19' 7& 



-~*\eok'. 



-Ii 


M 


ni 


M 


ir 




h" 


n 


M 




-> 




L_J 


L J 


i_ 



■-Pihro" f»>i ■■J'J'V , 
To break TiA-e holes 8%"*' 
were drilled m each Web 



46 335 

through <?V e 
hoes 



T i B 

57 497 



To break 
TiB 



First z holesa' 3 /'s" 

viere drilled m each Web 



Then thee holes in 
each Web 



4.7 069 

throughe% 

holts 



enlarged to J 3 /* 



r* 



43 7£2 
through holes 

' ■■-••-• 



BWebss^'x'/xs" 
4 Webs/o^x'/is' 
8Websio'-%*"x^i6" 

4Topl9-e , x2'x 3 /*,' 

4 Topt>2"xi 3 /B'x 3 /ia 
4 Flats 2'x '/a" 
4 Covers 4"x'/a* 



r' 

1 



■7ir--, 



,- Pm io"r 



2ofa' • 



552.49 
At Center 



39,716 



3I-' 



21'/*' L _ 



^r 



Actual Ar- 
Sq.ln, 



t !,>:. t>, I in„t 

Lbs per So In 



U'firnutr ' >irrtu}> 

1 1 . pet So vt, 



Proportion of Lmeor 
Jintcnohnt to full 



i 

around I*-, 
Axis 






1 



5 i 



■ »•**?■ 



42JOB 



tfTmtae'j *' 

Section changes from 
S6.6* S{ b '-V* upSgto 



^<S n 






r> n -7'o* 



Could not be 
determ'medas 
sec tion chongei 
continuously 



Highest load . 
a$3ao m per° 
\\ Buckling of Cover 



A 9 Sec tion changes 
often, ij lection of 
urie Rib will be con- 
sidered only, faken 
symmetrical about 
neutral- Axis 



-uv 



- 



StLong 



* Webs poU'x *3e' 
*f/afi r%*'s Sse' 



& 



"/se'of full 3/^e 
member or, 
l % /-'me s AVieo^ 
sizes of 7i Sfcor/ 



sBe'ifx*/*' 

'Flats s'xV*' 
4 Covers *'* $3/ 
f Cover r^'iy*' 
see'jie'-- 3 /*' 
sfates ie'i 3 /t&' 
eae'i if?"* 'la' 



ftti 

jIl Jg. jIljl 



■ Pine"" 



JI 



is ' i ''-' K *?'«# -H ., 

nfb'- 



*o.zi 

At Breaking 
Point 



sWebeei'jt'W 
e Flats e'x 'W 
e&e'xs'-sss* 
6/Se'xiye'-'^" 



■35's'--- ^ 



Applied bad of 
43, ieo lbs.per sqm 
'or eihrs. then 
i idded a bad of 
~ooo/6s. at center 



kS+j+.s'A 



efc'4 



Some 
rj A. 



break Ts A-e holes 

trilled 
in each Web 

b break Ts B-S holes 
'twere drilled 
in each Web 



ss -sis Ulf 

Unload at Center 



applied load of 

in. besides a had 
ofitooolosot 
center 



44.371 

Through 3'<r 
hofcs 



ss.eszUlt. 
Ho load at Center 



*iYebsai"+ 'U' 
AtSs'na'-'/J'-sas* 
4 Flats s'x '/*' 
elSe-iiVe'Va' 



r» ed?&-'- 



aWebseo'Sbz'xZis" 
4 Topl?e'xS'x 3 /i6' 
4 Top £?'jri%\ %?' 
" Top Flats n't '/a" 
* Top Covers-*'* '/&" 
i Top Cover 6%'j '/e' 
sBoitomlsa'ta'* */ 

?Bo ttom Cov II '/e"x 'le' 
6CenterlS S ' 1 i%\ii S > 






To break TiA-a holes £■%>"? 
were drilled in each Web 



4C335 

through?!* 

holes 



Highest load 
50,134 

Alot broken 



To break 
TlB 



Highest load c 

40.660 

Vet t'rten 



First Z holes. B ,3 /is" 
drilled m each Web 



TrVe . 
each Web i 

[ iraci I 



highest loodoi 
£3.939 

Not broker 



4 WebsioWn'/is' 
eWebsio"lt*"x 3 /i6" 
4 Toplia'is'* »/«" 

4 Topl>E"nlWK'/li' 

4 Flats a" a '/a" 
^Covers a'x'/b* 
> Cover e s /a"x '/a* 
a Splices e'n'/n " 
e Bottom B e'xe'x *e 
sCovers, ii '/a* x'/b' 
6Lfs'Ar'/a'x s /3s" 




■ Pin io"t 



35^43 

At Center 



Results of Tests of Nickel-Steel Models Made by the Phoenix Bridge Company for the Board of Engineers, Quebec B 

TEST 3 — Table No. 1 



ridge. 



TiB — The capacity of the testing machine, 48,660 lbs. per sq. inch, 
was applied for 5 minutes without failure. 2 - 2-13/16" holes were then 
bored in each web and a load of 59,989 lbs. per sq. inch was applied for 
20 minutes without failure. The holes in the web were then re-bored to 
3 - 34". The member failed by buckling of the webs at the reduced section 
under a load of 63,990 lbs. per sq. inch of net section, or 48,660 lbs. per 
sq. inch of gross section. 




T2A — The member failed by buckling of the ribs at the point of attach- 
ment of the transverse diaphragm near the pin end. 

T2B — One rib buckled at the point of attachment of the nearest 
transverse diaphragm to the pin end, the other three at the second set 
of diaphragms from that end. 

T3A — The full load of the testing machine, 49,120 lbs. per sq. inch, 
was applied for two and a hah hours without failure. Then a vertical load 
of 7000 lbs. was applied to the center of the column so as to be supported 
by the edges of the webs only, and the full load of the testing machine was 
applied without failure. Finally 2-2 l A" holes were bored in each web. 
The webs buckled at the reduced section. 

T.3B — A load of 49,000 lbs. per sq. inch was applied without failure, 
and then a vertical load of 17,000 lbs., resting on the webs, was imposed 
at the center and the full axial load applied without failure. Afterwards 
2 - 3" holes were bored in each web. The member failed at the reduced 
section under a load of 58,252 lbs. per sq. inch of net section. 

T4A — At failure the piece bent downwards and found a support on 
some blocking inadvertently left in place, thereafter carrying a load of 
52,650 lbs. per sq. inch. All four ribs were latticed on top and bottom 

170 



flanges, and the piece failed near the Center by buckling of the ribs near 
the center. The forked end showed ahghl signs of scaling near the pins 
at a load of 40,000 lbs. per sq. inch which gave a bearing stress on the pin 
of about -15,000 lbs. per sq. inch. 

T4B — The forked end showed slight signs of scaling at a load of 
36,800 lbs. per sq. inch, corresponding to a bearing stress upon the pin of 

about l:{,M(MI His. per Bq. inch. The mode of failure was almost identical 
with that of TiA, the inner ribs of both buckling outwards at the top 
flange. The bottom latticing at the point of failure showed very little 
distortion. 

TA Long] The member failed by buckling of the ribs close to one 
of the transverse diaphragms. 

ToB (Long) — The failure was similar to T&A (Long), and also took 
place at the point of attachment of a transverse diaphragm. 

ToA (Short) — Failure was by buckling of the ribs at the point of 
attachment of a transverse diaphragm. 

ToB (Short) — Failed by buckling of ribs at point of attachment of a 
transverse diaphragm nearest pin end. One lattice bar in the central 
plane failed in tension without shearing its connecting rivets, and one angle 
of the lower transverse diaphragm was torn from the rib throughout its 
depth. 

T6A — Two square ends. The sectional area changed almost con- 
tinuously from point to point, being 56-66 sq. inches at the weakest point. 
The highest load applied, 48,980 lbs. per sq. inch, did not cause complete 
failure, but a section of the 3/16" top cover plate bulged up at a point 
where the rivet pitch in the lini _'", that is to say. 10-2/3 times 

the thickness of the plate, and 9-1/7 times the diameter of the rivet. The 
upper flanges of the outer rib also bulged outward so as to take a permanent 
set. 

TeB — The results were almost identical with those of T< A. 

T:A Was similar to Ti, except that the webs were spliced longi- 
tudinally so as to avoid the use of wide plates. The distortion was greater 
from the beginning of the test than in the corresponding member with solid 
web. One rivet had failed at a load of 23,363 11 inch. Failure 

took place by bucklnig of the ribs at the transverse diaphragm nearest the 
central splice and a large number of loose rivets were found in the vicinity 
of the point of failure and in the reinforcement at the pin end. 

T:B— Failure was similar to T:A but took place at the diaphragm 
nearest the pin end. 

171 



As might be expected from the local character of the failures, there is 
but little relation between the ultimate strength and the ratio of length to 
the least radius of gyration. 

TiA, TiB, T3A and T3B would certainly have failed at lower unit 
stresses had sufficient power been available to test them to destruction 
without reduction of area. 

A noteworthy feature of the test is that in all cases where the points 
of failure were not predetermined by boring holes in the webs, with the 
exception of T4A and T4B, failure took place at the points of attachment 
of the transverse diaphragm. This might be expected in as much as, 
when the member as a whole is shortened, the action of the lattice bars 
tends to spread the ribs apart. The restraining action of the diaphragm, 
therefore, causes bending stresses of considerable magnitude in the rib. 

The design and construction of the members were such as to develop 
an unusually high and satisfactory percentage of the yield point value of 
the material employed, except in the case of T7A and T7B, which members 
were obviously weakened by the longitudinal splicing of the webs. 



172 



4.— TESTS OX 6 EYE-BARS, 28 TENSION MEMBERS 
AND 12COMPKKSSION .MEMBERS 

Made at the Phoenix Iron Company's Testing Plant under the super- 
vision of Professor II. M. Mackay of McGill University and Mr. James E. 
Howard, Engineer-Physicist of the Bureau of Standards, Washington, 
D.C., during January, February and March, 1912, for the St. Lawrence 
Bridge Co. 

Brief descriptions of these tests, taken generally from Professor 
Maekay's reports, follow: — 

Test on 6 Eye-bars: 

The six bars tested were each 10" x 1%" x 40' 0" c. to c. of pin holes. 
The end pins were 10" diameter and the pin holes approximately 10-1/32" 
diameter. Bars Nos. 1 and 2 were tested under their own weight. Bar 
No. 3 was counterpoised with three upward loads of 800 lbs. each applied 
at the centre and quarter points. Bar No. 4 was counterpoised with 
loads of 600 lbs. similarly applied; and bars Nos. 5 and 6 had loads of 50 
lbs. hung upon them by wire loops at intervals of 13-1/3 inches throughout 
their length. Extensometer readings were taken, (1) by Mr. J. E. 
Howard's micrometer gauges over 20" lengths at 9 points, (2) by Martens 
mirror extensometers over 4" gauge lengths at 4 points, and (3) 
by the Phoenix Bridge Co's Olsen extensometers over 200" gauge lengths 
on the top and bottom of the bar. 

A general inspection of the curves, plotted from the extensometer 
readings, shows that the elastic limit (limit of proportionality) and yield 
point occur under very different loads at different parts of the bars; and 
while the elastic limit can be pretty closely determined, some arbitrary 
rule is necessary to determine the yield point. Relatively large deviations 
from the stress-strain curve, and considerable "creeping" of the extenso- 
meters under a fixed load, occur at points considerably lower than would be 
indicated by the "drop of the beam" or "drop of the mercury column" in a 
rapid continuous test. 

Perhaps the load at which the stress-strain curve becomes sensibly a 
horizontal line may be as good a point as any to choose for the yield point. 
The following table is compiled on that basis. The figures given are the 
axial loads per square inch. 

173 







TEST 4 - 


- Table No. 1 — EYE-BARS 




Bar 


Elastic Limit 


Yield Point 


Ultimate 
Strength 




Ends 


General 


Ends 


General 


Remarks 


1 
2 


14,000 
12,000 


21,000 
20,000 


26,000 
24,000 


28,000 
27,000 


60,200 
59,470 


Under its own 
weight 
do do 


Mean 


13,000 


20,500 


25,000 


27,500 


59,835 




3 
4 


18,000 
16,000 


23,000 
26,000 


28,000 
28,000 


28,000 
30,000 


59,970 
59,620 


Counterpoised 
do 


Mean 


17,000 


24,500 


28,000 


29,000 


59,795 




5 
6 


16,000 
16,000 


24,000 
26,000 


28,000 
30,000 


28,000 
30,000+ 


60,100 
62,190 


Cross loaded, 
do do 


Mean 


16,000 


25,000 


29,000 


29,000+ 


61,145 





While the above values are tentative and might be selected differently 
according to the definitions adopted, they probably give as good a basis as 
any for comparing the bars with each other. 

From the above comparison and a study of the curves, it may be 
concluded, 

(1) That the elastic limit and yield point occur not only at much 
lower axial loads, but also at much lower measured fibre stresses at the 
ends than at the centres. The most reasonable explanation seems to be 
that the metal near the ends of the bars was affecting by the forging, so 
that the elastic limit and yield point there are, in fact, lower than in the 
central portion of the bar. 

(2) That the counterpoised and cross loaded bars, although represent- 
ing the extreme conditions tested, show very little difference as to their 
yield point or elastic limit. In the central portions of the bar the distri- 
bution of stress tends to equalize itself as the axial load increases. 

(3) While the elastic limit and yield point for bars No. 1 and No. 2 are 
smaller than for the others, an inspection of the measured stresses at which 
these points occur shows that the low values are in all probability due to 
the quality of the material. 

174 



(4) There is no evidence that the weights of the bars or cross loading 
had any influence in lowering the ultimate strength. 

(5) The bars all broke in the body between 5 ft. and 13 ft. from an 
end. No particular relation is traceable between the point of fracture and 

ptionally low load values of elastic limit and yield point. As is shown 
in the "measurements after testing," several bars exhibited marked 
elongations and reductions of area at points other than that at which 
fracture actually took place. 

Tests on J.' plate members. 

12 of these test pieces were made of single plates 36" x %" ', with pin 
plates of varying lengths while the remaining 10 were made of 3-36" x 5/16" 
plates stitch-riveted together and having pin plates all the same length. 

The make-up and general results of the tests are summarized in Table 
2. Column 8 gives the ultimate strength per square inch of gross nominal 
sectional area. The test pieces ran from 2.36% light to 5% heavy of the 
nominal calculated weights. The ultimate strengths corrected for the 
percentage by which the whole piece is light or heavy are given in Column 9. 
P"or the ultimate strength per square inch of net section given in Column 
11, three rivet holes one inch diameter are deducted. The values of the 
yield point and the elastic limit are not included in this table under these 
names as it is uncertain as to just what they refer to in a built up member, 
however, there is given in Column 12 the stress per square inch of gross 
section at which the stress-strain diagram over 200 inches of length on the 
scale plotted ceases to be a straight line. Column 13 gives the modulus of 
elasticity for the central 200 inches of length including the splice in case 
of spliced members. It will be noticed that the modulus for the spliced 
members runs a little lower on the whole than that of the unspliced ones 
notwithstanding the extra section. 

Several of the earlier test pieces failed by dishing back of the pin hole. 
Steps were taken to avoid this by driving strips of white pine between the 
pin plates and pulling shackles just back of the pin. The restraint thus 
introduced, although slight, was sufficient to prevent dishing. In nearly 
every instance where dishing occurred it was evident from the distortion 
of the web at the ends of the pin plates that failure was very near, and it i.s 
probable that the ultimate strengths obtained were but little reduced by 
the dishing. 

Extensometer readings were taken on 200 inch gauge lengths on two 
sides and two edges of each piece by means of Olsen Extensometers, and 
the elastic limit was determined by means of these readings. Numerous 
readings were also taken with a Howard gauge with a view to obtaining the 
distribution of the stress across the section, particularly at the ends of pin 
plates and splice plates. Numerous readings were also taken from the 
ends of pin or splice plates to the web in order to obtain the slip between the 
web plate and the pin (or splice) plates at the first line of rivets. 

175 



TEST 4— TABLE 2— 



Date of 
Test 
1912 



Webs 



Pin Plates 



Jan. 15 

Jan. 15 
Jan. 16 

Jan. 17 
Feb. 16 

Feb. 16 
Feb. 20 



1 PI. 36" x H" 

1 PI. 36" x H" 
1 PI. 36" x H r 

1 PI. 36" x H" 

1 PI. 36" x H" 

1 PI. 36" x H" 
1 PI. 36" x W' 



Feb. 21 1 PI. 36" x 



Feb. 19 

Feb. 20 

Feb. 17 

Feb. 19 

Feb. 21 

Feb. 22 

Feb. 22 

Feb. 23 

Feb. 23 

Feb. 24 

Feb. 14 

Feb. 15 

Feb. 15 
Feb. 15 



1 PI. 36" x H" 
1 PI. 36" x H" 

1 PI. 36" x H" 
1 PI. 36" x H" 

3 Pis. 36"x5/16" 
3 Pis. 36"x5/16" 
3 Pis. 36"x5/16" 

3 Pis. 36"x5/16" 
3 Pis. 36"x5/16" 

3 Pis. 36"x5/16" 

3 Pis. 36"xo/16" 

3 Pis. 36"x5/16" 

3 Pis. 36"x5/16" 
3 Pis. 36"x5/16" 



Splice Plates 



Material 

of Web 

Ultimate 

S Strength 



36" stf'x 4'-l" 

36" x H"x4'-1" 
36" x y 2 " x 4'-l" 

36" x H"x4'-1" 
36" x y 2 " x 4'-7" 

36" x y 2 " x 4'-7" 
36" x y 2 " -nb'-Y 

36" x y 2 " x 5'-l' 
36" x y 2 " x.4'-7' 
36" x y 2 " x 4'-7' 

36" x y 2 " x 4'-7" 
36" x H"x4'-7' 

36" x K"x4'-7" 
36" x K" x 4'-7" 
36" x K" x 4'-7" 

36" x K" x 4'-7" 
36" x K" x 4'-7" 

36 x H" x 4'-7' 

36" x J<" x 4'-7' 

36" x }{" x 4'-7' 

36" x K" x 4'-7' 
36" x K" x 4'-7" 



36"x5/16" x 3'-8 
36"x5/16" x 3'-8" 

36" x H" x 3'-2" 
36" x H" x 3'-2 



Member 

Light or 

Heavy 

% 



Ultimate 

Gross 
Nominal 
Section 



36"x5/16"x5'-6" 

36"x5/16" x 5'-6" 

36"x5/16" x 4'-6" 
36"x5/16" x 4'-6 



68700 

68700 
68700 

68700 
68700 

68700 
62900 

63440 
68700 
68700 

68700 
68700 

65620 
65620 
65620 

65620 
65620 

65620 

62910 

62910 

62910 
62910 



3.7H 

1.5H 
5.1H 

4. OH 
1.5H 

1.6H 
0.2L 



2.6L 
2.4L 

1.0L 
0.8L 

1.4L 
1.2L 
1.7H 

1.9H 

0.8L 

0.7L 
1.7H 

1.9H 

2.2H 
2.1H 



44660 

52000 
50980 

53330 
52270 

52000 
46930 

48670 
51330 
53330 

50670 
52270 

49510 
47110 
51730 

47550 
51380 

48710 

53330 

52000 

52440 
54930 



176 



PLATE TENSION MEMBERS 



9 


10 


11 


12 


13 


14 


15 


Ultimate 
< iroae 


M. in 


Ultimate 


Limit of 
Proport- 
ionality, 
200 ' 


]•: 

over 

200" 

Length 

(10") 


Mode of Failure 


Remarks 






40970 

55880 
52840 

5f)050 
51800 

53090 
57500 
59580 

55820 
57490 
54800 
52020 
55480 

50840 
56600 

53450 

57190 

55660 

56000 

58680 


28 

24000 
29300 

29500 

26700 

20000 

25000 
30700 

26700 

24000 
20000 
26700 
26700 
26700 

25300 
25500 

25500 

28000 

30700 

28000 
31000 


28.5 

28.5 
28.5 

28.5 

28.5 
28.5 
28.5 

26.6 
28.9 
25.0 

28.5 
26.6 
25.1 
26.0 
28.5 

27.0 
28.5 

27.0 

27 2 

26.9 

27.0 
27.0 


Rivet Line end of 
Plate. 

Dished Back of Pin. 

Rivet Line end of 
Plate. 

Rivet Line end of 
Plate. 

Dished Back of Pin. 

Dished Back of Pin. 

Rivet Line end of 
Plate 


Pin 

Pin 
Pin 

Pin 




51220 
18440 


47140 




51260 
51490 


49850 




51330 


51310 






47850 


Pin Ends Blocked, 4" x 


18670 


Rivet Line end of 
Plate 


Pin 


1 H" White Pine. 
Pin Ends Blocked as above. 


52710 


Rivet Line end of 
Plate 


Pin 






53665 


Pin Ends Blocked as above. 


54620 


Rivet Line end of 
Plate 


Pin 


Pin Ends Blocked as above. 


51170 


Dished Back of Pin. 

Rivet Line end of Splice 
PI 


Stretched to 30,000 lbs./sq. 
in. before Testing. 


52700 


51935 


Pin Ends Blocked. 


50230 


Rivet Line in Body. 

Dished Back of Pin. 

Rivet Line end of 
Plate 


Pin 




47690 
50860 


48960 






48730 


Pin Ends Blocked. 


46600 




Pin Ends Blocked. 


51880 


Rivet Line end of 
Plate 


Pin 






50490 


Pin Ends Blocked. 


10000 


Rivet Line end of 
Plate 


Pin 


Pin Ends Blocked. 


52420 


Rivet Line end of Splice 
Plate 






51720 


Load reduced to after 


51020 


Rivet Line end of Splice 
Plate 


reaching 24,000 lbs. /sq.in. 
Ensuing E= 30.1 x 10f. 

Load reduced to after 


51340 


Rivet Line end of Splice 
Plate. 

Rivet Line end of Splice 
Plate. 


reaching 24,000 lbs. /sq.in . 
Ensuing E = 29.2 x 106. 


53790 


52565 





177 



Large slips at the point might result from either comparatively loose 
rivets or high stresses on the rivets, and possibly on this account there 
seems to be no particular relation between the amount of slip and the 
strength of the test piece. 

The slip was usually found to b3 very small or zero up to a load of from 
4000 to 8000 lbs. per square inch on the test piece, but the relation between 
this slip and the load was usually not linear; it was often nearly so in the 
middle range of loading say from 8000 to 24,000 lbs. per square inch. 
It may be concluded that when the rivets are well driven friction 
between the plates prevents slip for the lower loads, and that the point 
where the slip begins marks the failure of friction to maintain its load. 
When the slip becomes proportional to the load the friction has probably 
become negligible and the slip likely denotes the elastic deformation of the 
rivet or the metal surrounding the rivet hole, which continues until the 
local elastic limit is reached. 

The stress-strain curves for the web plates indicate a very uniform dis- 
tribution of stress. It may be said of the web plates (1) that within the 
elastic limit the highest strain observed at any section seldom exceeds the 
mean for that section by more than 20%. (2) That the variation of the 
strain across the face of the plates from top to bottom, or of the mean 
strain of the two faces from top to bottom, are still less, the highest strain 
seldom exceeding by more than 6% or 7% the mean for the whole section. 
(3) That no relation is traceable between variation of stress across the 
section and the ultimate strength of the test piece. 

The results of the tests show that the stresses in the web and pin 
plates are probably equalized within less than two feet from the ends of the 
pin plates, and that the length of the pin plate apparently has no effect on 
the distribution. 

From a careful study of the data on the 22 tests it would appear that 
the variation in the arrangement of rivets, lengths of pin plates, and 
lengths of splice plates are of secondary importance as regards strength and 
rigidity. 

Tests on 6 two-web tension members. 

The make-up and general results of the tests of these members are 
given in Table No. 3. Numerous extensometer readings were taken in 
an endeavor to throw light on the action of built-up members. 

All the slip curves are quite similar in character to those of the plate 
members. 

Nearly all the curves show at the foot the effect of the weight of the 
member but the general impression given is that of a remarkably uniform 
distribution of stress at ' all points measured. No clear evidence was 
obtained to show that the stresses due to the weights of the members 
had any influence on the breaking strength. 

178 





E 

a 

— 

z 

o 

— 
z 


- 
5 

P 

— 

I 


cud of Bplice 
Ls one 
fnnii Bplice. 


3 

m 

o 

<- 

pa 


CD 

a 
— 

r. 

— 


end of splice. 

(>uicr Ls Rivet 
Line end of Bplice 
Inner La one space 
from splice 


i-H 

s 

CD 

s 

S3 


— 

n 

.s 

« 
o 


-r 
o 
PQ 

I 

o 

u 

pq 




c 

- 






CO 
CO 

uo 






/ 

CO 

■0 






o 

CO 

1- 




3 c • 


- 
•o 


o 

00 

•o 

U0 

'- 


O 
— 


o 
B5 

lO 


o 
qo 

00 

r 


o 
— 
— 


- 


5 

"f | -S 
.9 ^ « 

- - 7 


— 


o 

/ 
• - 


o 

00 
lO 


o 

01 
CO 

-V 


o 

00 

:: 
O 


to 

o 

i-O 


: 

- 
- 


D O >i 

— _ - 


= 

:: 

O 

9 


- 

oi 

3 


- 
f 

CM 

o 


i-3 

:: 

«o 
d 


P 
K 

O 

Ol 


p 


- 
- 

i 

n 

:: 


* 3 s 

- - = 

g .a a 

S p -z 


a o 

CC c 
M PC 

:: — 

EC EC 

J£ i 

1 


- - 

- - 


c o 

1- Tl 

— - 

M DC 

— CO 

s - 


- 
< 

- 

— 

- 

H 

p 


CO 

-_ 

•— 
— 
X 

■- 


Q . 
i - 

CO 

r- ! : 
2 - 
m in x 

- z > 

"' 7 as 

01 01 4 


1 ^ 

co 

02 

- : 

CO \00 
~ M\ 

i- X 
* ; 

; - M 

■r - 

— EC 

i i 
"* 00 


oo 

■ 

Ph 

" X 

X 

Ec 

co 
i 

M 




- 
z 

- 

1. 
a 
"8 


oi :: 

- _ 

^ E 


xjo 3s 

; * 
01 DC 

-1 — 
- 

> - 


to «b qd 

lO U5 W 

X X /• 
•- 5o ^ 

iH C* ^ 




— 
i 

5 
S 


o 

- 


01 
e 

H 


h 






1 



179 



A very noticeable feature was the reduction of section of the angles and 
their opening up from the webs at practically all points of lattice con- 
nection. Extensometer readings did not show, within the elastic limit, 
any stress of importance produced by the action of the lattice bars as the 
member elongates, and it does not seem that such stresses can amount to 
much at points remote from diaphragms or tie plates. 

Curves plotted from extensometer readings taken on the top and 
bottom splice plates between the two webs show a comparatively small 
stress in these plates, that the stresses get smaller away from the edges 
and that these plates do not do then- share of the work. 

In T-ll No. 2 the extensometer readings were limited to the top 
flanges, stay plates and lattice bars, the curves show considerable transverse 
compression at the ends of the end and center stay plates doubtless due to 
the action of the lattice bars in pulling the ribs together while measure- 
ments taken at the center stay (or splice) plate show very little distortion. 
Readings taken on the lattice bars gave very little information of value. 

On the whole it will be seen that the six members developed in their 
net sections about 90% of the ultimate strength of the material of which 
they are composed which seems a very satisfactory result. 

Twelve compression tests of built-up members. 

This series of tests included 6 short columns 10 ft c. to c. of pins marked 
TCi, TC2 and TC3; and 6 colunms 49 ft. c. to c. of pins marked TC4, 
TCs and TC6. Two pieces of each mark were tested. The make-up and 
general results of the tests are summarized in Table No. 4. 

The short columns differed only in the length of the pin plates. A 
prominent feature of these members is the great weight of the lattice bars. 
As might be expected, these members all failed by local buckling. The 
members with the longer pin plates sustained heavier loads, but the gain in 
strength is not so great as the increase in weight, so that the longer pin 
plates do not give higher economic efficiency. 

The mode of failure was nearly the same in all six members. In five 
members both ribs buckled outwards beginning at the end of the longest 
pin plate. In the sixth one rib buckled at that point, the other about a 
lattice panel nearer the centre. At failure scaling lines were developed 
all along the inside of the webs, some were also observable on the outside, 
particularly near the ends. Generally speaking no scaling lines developed 
in the angles. 

The long columns had values of 1/r from 73.7 to 80.8. As these 
members were tested horizontally without counterpoise the weights and 
resulting deflections had a very marked influence in inducing failure. They 
all failed by bending until they touched the cross beams at the bottom of 
the testing machine without any perceptible local buckling. No loose rivets 
or deformed lattice bars were noticed. 

180 



















i 

0) 


























T3 














it 


























g 












~- 


























u >- 












•/. 














- ? 




• 








ft* 


ja s 


r 


o 


o 


o 


o 


i- P 

> 


O 


o c 


o 


_o 




- 




TS 


■CO 


-3 




-c 


-c 


"C TT 




^ 






x : 




































c 














- 












— s 

_i- • - 

- ; 












z\- - 


co 




r^. 


o 


X 


1- 


■c 


CJ 


00 — 


t 


o 




3 •§ S g 

a = — 

- K a - 


r. 


t^ 


co 




>o 


re 


o 


co 


(35 «* 


- 


f 




?i 


Cl 


■- 


— 


CS 


<o 


ro 


o 


CD O 




ei 




r~ 


3 


/ 


OC 




o 


•j: 


- 


O 05 


CO 


»o 




•: 


r: 


-.-. 




— 


■^ 


CI 


CI 


CC CI 


Cl 


ei 




— — _2 / 


























_ /. - 


























u 


























-. . 


— 


^ 


to 


CS 


© 


CO 


/ 


ri 


1^ s 


o 


CO 




— ~L — O 


— 


f 


o 


5 


E 


1* 


cc 


•«ti 


co co 


c 






— - .- 


i 


Ci 


- 


»C 


co 




c< 


O OS 


c 


i~ 




.2g 


3 


■<*> 


-: 


cn 


•o 


W 


O 


o 


l~ 1 7 




Cl 




re 


CO 


re 


cc 


cc 


co 


0) 


(N 


CI CI 


Cl 


Cl 


V 


























— 




i 






















■^ 


(a 
























- 


B _ ■- 


c _ 












CD 


t- 


00 cs 


00 


o 


- 
















00 


1^ 


r- os 


"<* 


co 


i's 7 = 


Q S 


o 


z 


_c 


o 


c 


GO 


o 


Cl CN 


o. 


OS 




— 


— 




-a 


— 


X 


00 


o o 


00 


/ 


* 














pH 


1—1 


Cl CI 


' — ' 


^^ 


z 




£ 














































1 


























r 

- 


c -j 


?1 












/ 




1^ 


00 








o 


o 


r 


__£ 


_£ 




o 


c 




O 


— 
- 


zc~ 




— 


-r 


-c 


- 


~ 


c 


~G 


co ~ 


o 


D 


: 


























w 




















































CO 


























-5 •: i L 
S 7- 1 


N 












5 










— 
- 


■ - 


o 


Q 


Q 


C 


o 


iO 


O 


Z 


o 


o 


j 

OS 




" 


~ 


— 


73 


00 


~ 


X) ~ 


^ 


~ 


_ 
- 


>-, - s 
























< 


























- 


1 r- i 


- 












•x 




CO 


co 




1 


■* 


o 





c 


o 


O 


W 


o 


IO 


O! 


_c 


— 


= E" 


CD 

<0I 


-a 


TS 


— 


T3 


- 


i-O 


T> 


-o 

i-O 


i-C 


- 


- 

X 


55 x 
















































- 
- 




• 03 












y! co 




V. 7. — ; M 


go tg 




— - 


\po\oo 












Op 

- 9C 




sooNtc * X 
cc " " — 








y. 


_o 


O 


o 


O 


o 


y. 


O 


X X 


X 


~ 




•— — 


— 




-c 


~ 


"C 


T3 


TP 


T3 


M "* X M 


-r 


"Z 




- 


y X 












X X 




^x^x 


X X 






n 












N 




Cl -f 


CN 






r ,'T 












-,'7 




ei x — — 


r ,'-r 








n — 












CN Tf 




> ' • 


ri -r 






-^ 


— 


(N 


1—t 


ci 


I— 1 


CN 


*-< 


CN 


lH Cl 


»— i 


CN 




— 


_L 


J-, 


CN 


ci 


CO 


X 


— 


— 


LO IQ 


co 


'-T 




U 


O 


P 


O 


_ 


O 


P 





O 


O 






P 


H 


P 


E-i 


H 


H 


H 


E- 


H r- 


H 


r~ 




z 


i— i 


ON 


CO 


^ 


lO 


o 


t^ 


oo 


CS c 




01 






























~ 

























181 



Numerous readings were taken with Olsen, Howard and Marten's 
extensometers. The Howard gauge readings in the first member of each 
kind were devoted principally to investigate the distribution of stress in 
the ribs near the pin plates. In the second member particular attention 
was devoted to the relative motion of the two ribs. The behavior of each 
pair of members was remarkably uniform. 

Points of high local stress are obviously, in general, the points which 
limit the strength of the columns. Considering the 12 tests included above 
the highest stresses noted were just in front of the pin. These stresses are 
not of so much importance, as the high values occur only for a very short 
length, and in all the members of the present series transverse diaphragms 
made the webs secure against local buckling at that point. In the long 
columns the deflection due to weight is clearly the determining factor. 
Apart from these considerations the part of the column seeming to need the 
closest attention is that just at the end of the pin plates. High local 
stresses have been noted there. Among the causes are, (1) Inequality of 
stress between the two ribs due to lateral deflection or imperfect workman- 
ship. (2) Concentrations of stress in the webs from rivets of pin plates 
which do not extend the full width of the member. These loads are 
usually applied eccentrically as regards the web and must tend somewhat 
to cause dishing. (3) Pin pressure applied eccentrically as regards the 
centre of gravity of the rib. (4) The combined action of the tie plates and 
lattice bars in bending the ribs at the ends of the former. 

Slip between the pin plates and the webs was observed of the same 
character as the corresponding slip in the tension tests, indicating a large 
concentration of stress on the rivets near the end of the pin plates. 

Vertical readings over a 10" length at various points in the webs were 
taken, all showing the transverse expansion in the plates accompanying 
longitudinal compression. 

The transverse readings taken on the second member of each pair to 
determine the lateral deflection of the ribs due to the action of the lattice 
bars, are of considerable interest. The spread of the ribs in the three 
members measured was very nearly the same. 

Under the most favorable circumstances the curvature indicated must 
be accompanied by rather heavy bending stresses. 

The spreading of the ribs increases rapidly at high loads. 

Transverse measurements at the end of the tie plates denote a trans- 
verse tension stress as high as 7,500 lbs. per square inch. 



182 



5 —TESTS OF 6 CARBON STEEL EYE-BARS WITH OBLONG 
PIN HOLES 

Extracts from report by Professor H. M. Mackay 

This report includes the tests of six eye-bars, nominally 14" x W 2 " x 
12' 0" centre to centre of pins, made at Phoenixville, September 10th to 
13th, 1912. The pin holes at one end, called "East," wen- circular; those 
a1 t i u . ther end, called "West," were oblong so that the longitudinal 
diameter was '.. inch greater than the transverse. Details of the actual 
measurements of the bars are given in Table 1. The main object of the 
was to discover any differences in the strength or the distortion of 
the heads which mighl arise from the elongating of the pin holes. 

Method of Testing. 

The bars were measured and points were established on the axis of the 
bar, on both sides and at each end, at a distance of 3' 0" from the back 
of the pin holes. The bars were then placed in the testing machine and a 
load of 20,000 lbs. per sq. inch applied. Under this load the elongation 
from the front of the pin to the points established as above was measured 
by means of a scale graduated to one-hundredth of an inch. The bars 
were then removed from the machine and the permanent set in the three 
feet from the back of the pin holes was measured by means of a Howard 
gauge. The bars were then loaded to 28,000 lbs. per sq. inch and corres- 
ponding measurements made; after which the bars were again placed in the 
machine and pulled to destruction. 

The measurements from the front of the pin to the "three ft. point," 
under the applied load, are assumed to give elongation in the three ft. 
length from the back of the pin holes. The only error in this assumpt ion 
would arise from the flattening of the pin. It would be difficult to devise 
any scheme of measurement which would entirely eliminate this error, 
and it must be so small compared with the distortions observed as to be 
negligible, at all events for comparative purposes. 

Results of Tests. 

Table 1 gives the yield points and general results of the tests to destruc- 
tion. The yield points were determined by the drop or halting of the 
mercury gauge. Bars 3 and 4 gave unmistakable yield points at 24,350 and 
25,650 lbs. per sq. inch respectively; but when these bars were replaced in 
the testing machine after being loaded to 28,000 lbs. per sq. inch, new 
yield points of 29,220 and 29,120 lbs. persq. inch respectively were obtained. 
All bar, failed in the body, but the East head of bar No. 2 (which was 
actually tested first) dished back of the pin, the dishing being almost 
simultaneous with the fracture in the body. To prevent recurrence of the 

183 











-a 


T3 






















a 


3 












a 

3 




fH 






03 
O 
O 
O 


u 


u *> 








racter of Fra 

and 

Remarks 




j5 

a 

(-■ 03 

03 »h 

60 O 


•a 

M 





_oS t3 

-3 « « 


CO 
CN 

O ^ M 

v- § -2 M 

"S — "° G 

W _ ^ 03 

'co W ■< 


03 


03 ^ 

"» 2 

M S 
3 g 

<) rj 

l.s 

CO S 


T3 

Cj 




•a 

cu 

^1 

_ t> 

P-. 

^ CO 






13 3 


P3 

03 


03 3 

~ ° r^ 
CO CM W 


~ CQ 


a 




OS 

A 

O 




:3 O 
CO "* 


a 
as 

H 


g- S3 




CO Tjt 


■a 

w 


H3 cu 

co W 






CO 

3 


v 




^ 


^ 


^ 


N 




^ 


^ 


a S a 


cm 




cb 


05 


rt 


rt 




CN 
CO 


00 


ongatio 
i'-O" fro 
ck of Pi 
Hole 


CD 
CO 




10 

co 


CO 

•* 


00 

CO 


00 




■* 


43 


; 




; 


; 


; 


; 




; 


; 




O 






OS 


d 


Tt< 




O 




w .a « 


o3 


O 




00 




10 






lO 


OS 


lO 




Tji 


CO 


CO 


CO 




CO 


co 




-w 


^ 




N 


K 


^ 


^ 




^ 


^ 






















3 o> 


0J 


U5 




CN 


CN 


CM 


CO 




co 


CO 


.2o 

03 W 

s .5 


cm 






co 


CN 


IT) 

CO 




10 

co 


CM 

CO 




£ 




^ 


^ 


^ 


^ 




^ 


^ 






















.2 Ph 


03 






OS 


CN 


CD 


CO 




•* 


co 


c3 






CD 


TC 


co 







t>. 


OS 


M o 


H 


CO 




CO 


oi 


CM 


CM 




pi 


CN 


iduc- 
>n of 
rea 




d 




O 

10 . 



00 


CO 


co 




OS 
CN 






£%< 




•* 




■* 


CN 


co 


•* 




tH 










N 




^ 


^ 


v 


^ 




^ 


































co - 




lO - 


O ^ 


CM ^ 


^ 




O ^ 






3 t- 
O 3 
•« *j *» 




^D 




°d 


*.q 


^D 


n 1 ^, 




Id 








1-1 co 




l-H O 


>-l T}< 


^ 10 






11 as 








x ""*! 




K °° 


« n . 


M ^! 


Z q 




x *1 






" * § 




^ CM 






^ >o 


^ ■* 


^ CN 




v <M 






























«a £ 




© 11 




co .. 


CD ,, 


§ II 


CM ,. 




00 „ 








© II 




q II 


(33 II 


OS II 




q II 










rH 




■H 


6 


<N 







*H 










1-1 




rt 


1-1 


1-1 


rt 




1-1 






s 
























ti ^ 




■* 




I> 


10 


"5 


CO 




CM 






§ -2 7 tf 




■* 




co 


,_; 


o> 


CO 




t^ 








CN 




CM 


CN 




CM 




(N 






w 
























0) J3 




























O 




O 


O 


O 


O 




O 






S 60 




co 




CO 


co 


o> 


OS 




lO 






a a 

.S a 




CM 




00 


00 


c» 


co 




U5 















C3 










OS 






— - 




CO 







1T5 


CO 


CD 




O 






5 co 
























V « 




O 







O 


S 2 

"o SJ * 

CM &> 
" CN 


O 




O 








3 




CD 


S <*> » 

CN « 
"^ CN 


os 




CN 






.2 '3 











OS 




os 






»* P* 




CO 




CO 


CM 




CM 
















^^ 


*-* 
















^ 




^ K 


N ^ 






































en 




co *+ 




—I 


■* O 


O O 


O O 




iO O 










•* M> 




CD t^- 


CD 00 


CO CO 


OS CO 




00 <o 






Thick™ 
of 
Head 




t ^ 




■* Tjl 


■"*! 't 


TJH Tl< 


■*! "*! 




"* ■**! 


































-*-> — ■ 






-t> -p 




-»- ^ 








co CQ 




CO CO 


CO CO 


CO CO 


CO CQ 




CO co 








as . « 




03 ,CU 


03 CD 


03 <u 


03 « 




03 « 








W £ 




w £ 


« £ 


w £ 


w ^ 




H ^ 
















^ 


^ 


s 




^ 


































t^ 




10 


CO 


10 







m 










00 ; 




CN - 


CN ^ 


OS ^ 












rs a 




2^ 




"?p 


"?D 


^D 


"?□ 




"?q 






5 o 






—c OS 


rt CO 


•-< io 


rt 03 




rt CD 






3 •£ 




os 




M 2 


x 12 


m 2 


M 2 




X ^ 






3 s 

"*< CO 




„ O 




5»« 


; ^? 


lo? 


5^ CM 
CO ,. 




"°? 










l-H II 




l-H II 


l-H II 


q II 


q II 




■— II 










•* 




■<3< 


■* 


■* 


■* 




■* 
















rt 


1-1 


^ 




*"' 






03 








CN 


CO 


■<* 


10 




O 


a 

0! 


n 




















a 

s 





184 



dishing, the heads of all the other bars were rather lightly blocked with 
strips c>f white pine inserted be! ween the beads and the shackles, just back 
of the pins. 

Tin' elongations <>f the pin holes after testing arc in two cases greater 

in the Hast (round) holes, and in four cases greater in the West, (oblong) 
holes. The mean elongation of the round holes is 2.93 inches, and that of 
the oblong holes 3.26 inches. Similar relations exist in the elongations of 
the three foot length measured axially from the backs of the pin holes, the 
mean elongation for the Baal end being 3.91 inches and for the West end 
t. IS inches. 

Table '_' gives the permanent sets in the three foot length referred to, 
after loads of 20,000 and 28,000 lbs. per sq. inch. After the 20,000 lb. 
load these values are greater for the West end in bars 1 and 0; for the East 
end in bars 3 and 5; and the same for both ends in bars 2 and 4. The 
greatest permanent set .026 inch is at the West, end, while the mean for 
the east end is .0107, and for the West end .0130, a difference of .0023 inches. 

After a load of 28,000 lbs. per sq. inch, the permanent sets in the same 
distance are greater for the East end in bars 1 and 2 ; and greater for the 
West end in all the others. The mean for the East end is .0795 inches 
and for the West end .0906 inch, a difference of .0111 inch. 

Table 3 gives the elongations under loads of 20,000 lbs., and 28,000 
lbs. per sq. inch in the same three foot length from the back of the pin hole, 
the actual measurements having been made as explained above from the 
front of the pin. The elongations under the 20,000 lb. load are more 
concordant than the permanent sets. The mean elongation for the East 
end is .048 inch and for the West end .052, a difference of .004 inch. For 
four bars the elongation is greater in the West end, and for two bars the 
same in both ends. Under the 28,000 lb. load, the elongation is greater in 
the East end for two bars, greater in the West for three, and the same in 
both ends for one. The mean elongation for the East end is .134 inch and 
for the West .150 inch. 

From the above data it may be concluded that (1) There is no evidence 
that the ultimate strength of the bars is affected by elongating the pin 
holes, since all bars broke in the body. (2) At a load of 20,000 lbs. per 
sq. inch the elongation and permanent set in the oblong heads are on the 
average greater than in the others by about .004 inch and .002 inch respec- 
tively, but these differences are probably too small to be of much signi- 
ficance or of great importance. (3) At higher loads the distortions of the 
oblong heads continue to be greater on the average, but the differei 
are rather small compared with variations in the individual bars. In 
some instances indeed the omission of a single bar would make the mean 
permanent set or distortion, as the case may be, Less for the oblong end 
than for that with the circular pin hole. These variations are not aston- 
ishing in view of the fact that the elastic limit is undoubtedly reached 
locally in all heads at a lower load than 20,000 lbs. per sq. inch. 

is.", 



TEST 5— Table 2 



Permanent Set in 3'-0" Length. Measured prom Back op 
Pin Hole After Load of 20,000 lbs. per Sq. In. 





East 


= Round Pin 


West = Slotted Pin 








Hole. 




Hole. 






Bar 














Remarks 




North 


South 


Mean 


North 


South 


Mean 




1 


.007 


.008 


.0075 


.027 


.025 


.0260 




2 


.016 


.000 


.0080 


.016 


.000 


.0080 




3 


.022 


.011 


.0165 


.011 


.005 


.0080 




4 


.007 


.008 


.0075 


.011 


.004 


.0075 




5 


.011 


.011 


.0110 


.008 


.010 


.0090 




6 


.013 


.014 


.0135 


.019 


.020 


.0195 


Scaling behind Pin, 
West End, below 
20,000 lbs. / sq. in. 




Mean. 




.0107 


Mean. 


.0130 





After Load of 28,000 Lbs., per Sq. In. 



1 


.127 


.120 


.1235 


.072 


.035 


.0535 




2 


.090 


.078 


.0840 


.031 


.016 


.0235 


East Head scaling 
between 20,000 
and 28,000 lbs. 


3 


.081 


.068 


.0745 


.149 


.137 


.1430 


West Head heavily 
scaled at 28,000 
lbs. behind Pin. 


4 


.033 


.028 


.0305 


.045 


.031 


.0380 


Scaling general at 
26,000 lbs. 


5 


.053 


.055 


.0540 


.063 


.062 


.0625 




6 


.111 


.110 


.1105 


.219 


.227 


.2230 


Heavy Scaling Both 
Heads at 26,000 
lbs. 




Mean. 




.0795 


Me 


an. 


.0906 





186 



TEST 5— Table 3 

\\i\i. Elongation m :;'-(>" Length. Measured from Back of 
1'i\ Bole, Load 20,000 Lbs. pes Sq.In. 







Round I'in Hole. 


Weal =Slotted Pin Mole 


Bar 








North 


South 


Mean 


North 


South 


Mean 


1 


hi 


.04 


.040 


.05 


.04 


045 


2 


.06 


.04 


.045 


.05 


.01 


.045 


3 


.05 


.05 


.050 


.00 


III 


.060 


4 


06 


.04 


045 


.00 


.01 


.050 




.05 


06 


.055 


.07 


.05 


.060 


6 


.06 


.05 


.055 


.07 


05 


.060 




Mian. 




.048 


Mean. 


.052 



Load 28,000 Lbs. peb Sq. In. 



1 


.15 


.14 


.145 


.10 


.09 


.095 


2 


.1 1 


.15 


.145 


.09 


.08 


.085 


3 


.13 


.14 


.135 


.21 


.19 


.200 


4 


.09 


.10 


.095 


.11 


.08 


.095 


5 


.11 


.11 


.110 


.15 


.13 


.140 


6 


.17 


.18 


. 1 75 


31 


.26 


.285 




Mean. 




.134 


M. 


an. 


.I.",!) 



187 



6.— PIN FRICTION TESTS 

A test to investigate the pin friction in eye-bar heads and the efficiency 
of certain lubricants was made at McGill University during the early 
months of 1912. 

Extracts from the report by Professors Brown and Mackay follow: — 

An apparatus was designed whereby an eye-bar under tension could be 
subjected to a bending moment, the moment being applied to the pin 
through which the bar was loaded and being transmitted to the eye-bar 
by the friction between the surfaces of the pin and the head of the eye-bar. 
The principle of the apparatus is shown on Figure I. 

Levers Li L2 are keyed to the pins in the holes in the head of the e3 r e-bar. 
Links L3, L4 and Ls, the latter of which terminates in a platform P clear 
of the eye-bar, are connected as shown by pin joints at A, B, C, D, and a 
load Wi may be hung from P so as to be in the line of pull applied 
to the eye-bar. This gives tensions Ti and T2 in the vertical links L3 and 
L4. and thus moments Till and T2I2 of equal magnitude are applied to 
the pins at the ends of the eye-bar. These moments bend the bar so that 
its axis moves into a position shown by the dotted line. If with a given pull 
on the eye-bar, a series of tests be made to determine the load Wi which 
must be applied to the system of levers in order that the pins may slip in the 
holes, comparative values of the friction coefficient, using different pin 
clearances and different lubricants between the surfaces, can be determined. 

Further if r= radius of pin in inches 

T = tangential friction force between pin and eye-bar 
Tr = moment of pin friction 
= TihorT2l2 

™ Tih 

Then T = 

r 

T 

In the following report -has been called the friction coefficient for 

Pull 

the pin = u. 

The lengths of the levers and radius of pin (2-3^") are such that 

104.2 Wi 



Pull on the bar 
Pins and Eye-bars supplied for Test. 

Three eye-bars 5" x %" x 5' 0" c. to c. were used for the test. Two 
sets of pins of approximately the same diameter 5" were used, one set 
having a highly polished surface while the other set had the ordinary 

188 



machine turned surface. One eye-bar was bored so as to have a snug fit 
on the cud and the other two bad a clearance <>f aboujt 1/32" in one case 
and slightly more than this in the other. 

Method of test: A given pull, say 60,000 lbs. was applied to the 
eye-bar and the lever Le was then adjusted, by means of a device at D, 
to a definite position and the load Wi was then added in equal increments 
until .slip occurred. After each increment the lever Ls was adjusted to its 
initial position from which il bad descended when Wi was added. Another 
increment was then added and the process repeated until the slip load was 
reached. The apparatus itself and the weights used in balancing the lever 
system before adding Load Wi Kave a pull of 1,110 lbs. on the eye-bar. 
The addition of the weights Wi also affected the pull of the bar and these 
two items are included in the pull used to determine the friction coefficient 
amount given in the table is exclusive of these two items). 

The first eye-bar tested had a snug fit on the pin. It was tested without 
lubricants and the average value of u for several loads was .40. Tests with 
lubricants was not made on this bar. 

The results of the tests using the other two bars are given in Table 1. 

TEST 6 -Table 1— FRICTION COEFFICIENTS 





Pin 


Lubricant 


"u" under pull 


u 


Bar 


40,000 


60,000 


80,000 


Mean 


? 


Polished 


Dry 


1 25,000 1 
'■ 0.25^ 
0.11 


0.29 

0.14 
0.014 


0.27 

0.14 
0.018 


27 




Heavy cylinder oil. . 


0.13 
0.016 


2 


Turned 


Dry (Residuum of 
Paraffin) 

Dry (Cleaned by 
melting) 

Heavy cylinder oil. 

Tallow and white 
lead 


0.076 

0. 15 
0.114 


0.066 

0. 17 
0.112 

0.042 


0.134 
0.038 


0.071 

0. 16 
0.120 

0. 04 


3 


Turned 


Dry 


0.240 


it 236 


0.254 


0.243 


2 




Tallow and white 
lead 








013 




Graphite and Lin- 
seed oil 


007 


2 


Neeker's "XB 

Bridge lubricant 
Neeker's "XXXB" 

Bridge lubricant. 


0.025 
0.023 



189 



The paraffin was applied by heating the pins with a gas blow torch 
until the wax melted and ran freely in a thin layer over the surface. 
The clearance 1/32" was sufficient to allow the pins being slipped into 
position, but without damaging the wax layer. For the second series 
the same bar was used. The pin holes were thoroughly wiped out with dry 
cotton waste. The condition was assumed to be dry metal on dry metal. 
Unexpectedly low values of "u" were obtained (an average of .07). The 
previous tests showed an average value of u= .27. The difference was very 
striking and as it seemed possible that some of the paraffin wax had got 
thoroughly pressed in the uneven surface of the eye-bar heads and had not 
been removed by wiping with waste the bar was removed from the machine 
and heated with a gas blow torch so as to melt the residual wax. They 
were wiped off while hot and the test was then repeated. The value of 
u = .16 obtained was more than double that obtained before the hole was 
heated to remove the paraffin, thus demonstrating the marked effect of a 
small residuum of paraffin. The value is however much less than obtained 
in the first series of tests with the same eye-bar and the polished set of pins 
(u. =.27). This result may have been due partly to slight traces of 
paraffin which had not been removed by heating, to the fact that the eye-bar 
heads had been distorted in the previous tests, or it may be that the 
machined finished surface of the pin gave a bearing at a number of points 
more resembling a line of contact than an area of contact in the case of the 
smoother pins. 

The tallow and white lead lubricant was made of three parts of white 
lead to one part of tallow by weight. The white lead was well stirred into 
the melted tallow while it was hot and the mixture was then allowed to 
cool before being smeared over the pin. 




> a ~ 



— 41 

TEST G — Figure 1 



190 



7.— TESTS 01 8 TENSION AND 10 COMPRESSION M< >DELS OF 
MEMBERS I SED l\ THE Ql BBEC BRIDGE 
AS BUILT 

This scries of tests was made al the Phoenix Iron Company's Testing 
Plant during February and March, L913, for the St. Lawrence Bridge Co. 
and under the direction and persona] supervision of Mr. James E. Howard, 
engineer-physicist, Bureau of Standards, Washington, D.C. 

A description of th is published in the "Engineering Record" 

of March 21st, July _>."ith and November 14th, 1914, which is here reprinted 
by courtesy of the "Engineering Xews-Record." 

Various details, diagrams etc., which acccompanied these articles 
are grouped in proper order for reference from the text as noted. A 
Bummary of the results is given in Table 1. 

EMTEKBING RECORD, MARCH 21.ST, 1914 



ULTIMATE STRENGTH OF CARBOX AND XICKEL STEEL 
MODELS OF QUEBEC BRIDGE MEMBERS 

Destruction Tests of Precision on Large Riveted Compression M embers 
in the 640-Foot Suspended Center Span. 

• Ties of eighteen destruction tests of large-size models of tension 
and compression members for the trusses of the 1800-ft. cantilever span 
of the new Quebec Bridge has recently been made at the Phoenix Bridge 
Company's shop. The specimens were larger and stronger than similar 
members in many important railroad spans, and the results were carefully 
and accurately observed and recorded in great detail in order to furnish 
data for the analysis of the type of design, the character of detailing, 
fabrication and construction, the ultimate efficiency of the members, 
and especially for the comparative working value of duplicate compression 
members made of carbon or of nickel steel. A significant result is the 
demonstration of an elastic limit of a carbon-steel riveted compression 
member varying from 50 to 63 per cent of the elastic limit of a duplicate 
member made of nickel steel. The tests are also illuminating in their 
bearing on late developments of design for important tension and com- 
pression members. 

Scope of Tests 

The tests involved maximum loads of 2,800,000 lb. and included four 
tension tests on riveted plates, two tension tests on riveted members with 

.ngular cross-sections calculated to receive either tension or compres- 
sion, and twelve compression tests on members made with two or more 

191 



web plates and latticed flanges, models of compression members in the 
webs and chords of the cantilever trusses. The tests of two nickel steel 
and of two duplicate carbon-steel compression members showed the 
highest ultimate strength of the nickel steel members to be about 54 per 
cent greater than that of the weakest carbon-steel member and its elastic 
limit 101 per cent greater than that of the weakest carbon-steel member. 

The compression tests in this series were made on models of the vertical 
and inclined web members and of the inclined bottom chords of the canti- 
lever and anchor arms and suspended truss. The models have maximum 
cross-sectional areas of 70.65 sq. in. and maximum lengths of 45 ft. 10 in. 
The nickel steel models have a cross-sectional area of 35.2 sq. in. and a 
length of 34 ft. Some of the models represent members having shop 
weights up to 140 tons and lengths up to 90 ft. 

The design of the old bridge was fully described and illustrated in 
many articles published in the Engineering Record and in the report of the 
Royal Commission, which was published in full in the Engineering Record 
of March 14, 1908, page 309. In connection with the official design for 
the new bridge prepared by the board of engineers, a number of models of 
its principal truss members were made in carbon-steel to a scale of about 
one-fourth the natural size and were tested to destruction by the Phoenix 
Bridge Company in the same machine which was used for the tests des- 
cribed in this article, giving results interesting for comparison of strength 
and efficiency of material and design. The largest of the sixteen models 
had a cross-sectional area of 57.1 sq. in. and a length of 35 ft. 9 in., and the 
highest elastic limit obtained was 36,618 lb. per square inch. The details 
of these tests were published in the Engineering Record of Nov. 19, 1910. 

All of the models in both series were tested at Phoenix ville, Pa., in 
the Phoenix Iron Company's machine which was built by them from their 
own design and has recently been overhauled, calibrated and standardized 
and th6 frictional correction determined, which has been applied wherever 
necessary to the results recorded for these tests. The machine is of a 
horizontal hydraulic type, giving direct stress up to a nominal capacity of 
2,720,000 lb. tension or compression, and can receive members up to 50 
ft. long for compression and about 40 ft. long for tension, dependent upon 
elongation of metal. It has a plunger area of 3000 sq. in. and stroke of 
about 6 ft. The cylinder head is permanently attached to the plate girder 
bed frame and the movable head can be pin-connected to the frame at any 
point within the limit of length. Pressure in the hydraulic cylinder has 
been developed by a direct-acting steam pump, but an electrically operated 
hydraulic pump is now used. To provide for testing compression members 
a steel casting with a half hole bearing for a horizontal pin 10 in. in dia- 
meter was bolted to the plunger head and split steel bushings were provided 
for it and for the pin bearing on the moveable head to correspond with the 
sizes of pins in the compression members. (See Figure 3). 

192 



ink rebalancing 

Fulcrums consisting of Bteel pins in wooden half-hole bearings with 
clearance between them were Bel on the bed of the machine al three equi- 
distant points intermediate between the fixed and movable heads, and 
supported horizontal [-beams transverse to the axis of the machine al their 
center points. One end of each [-beam projected across the compression 
specimen teal and was attached toil by a 1 indie or sling inclosing the piece. 
Prom the other end of the beam was suspended a weigh! corresponding to 
the fraction of the weight of the truss member which it was intended to 
counterbalance, thus relieving the tesl specimen of secondary stresses due 
to its uu n weight. 

The elongations and compressions on short distances wen- measured 
by J. E. Howard for numerous locations near the pin bearings by the use of 
standard Brown A: Sharpe strain gages. Kadi gage had two conical steel 
points inserted in special holes of aboul 1-10 in. in diameter and about 1-16 

in. dee]), drilled and reamed with precision in the required positions, thus 
providing for very exact and unvarying setting of the instruments. 

Measurements of the elongations and compressions between points a 
long distance apart, at the center of the model and in its full length were 
made by C. Scheidl, Jr., of the Phoenix Bridge Company. He used 
Olsen standard compressonieters. Each had a graduated circular dial and 
an indicator needle having a horizontal axis with friction bearing interme- 
diate between its journals. The friction bearing consisted of a concentric 
cylinder with a circumference of ] ■? in., which engaged a long horizontal 
st.el rod fixed at one end to the test specimen at one of the points between 
which the elongation or compression was to be measured. The rod was 
supported intermediately on rollers and the other end rested on the friction 
cylinder and its weight developed sufficient pressure to produce friction 
revolving the cylinder and the axis of the needle to correspond with any 
longitudinal movement of the rod. 

The bearings of the needle being firmly attached to the specimen at an 
accurately measured distance from the attachment of the rod, any longi- 
tudinal distortion of the model produced a relative movement of the rod 
on the axis and caused the latter to revolve. The ratio of the diameter of 
the frict ion bearing and the diameter of the dial of the machine being 25 and 
the dial being graduated to five hundredths, gave a direct reading of 
1-1000 in., which could be reduced to 1-10,000 in. by the vernier scale of 
1.10. 

A I > plication of Instruments 

In order to correct eccentricity and irregularity and secure true values 

of the total distortion of the model a conipressometer was applied to each 

of the four sides of the cross-section and the mean of the results was taken. 

machine was bolted securely to a steel bracket, and the latter was 

193 



either bolted or riveted to the model. (See Figure 4). For readings 
taken on the vertical sides the brackets were riveted to the web plates, but 
for readings taken on the horizontal latticed faces the brackets could not 
be attached directly to the lattice bars, because the latter, not being 
directly affected by the compression in a member, are materially displaced 
by the compression in the web plates, and since they retain at first their 
original length while the longitudinal distances between their rivets are 
diminished, the web plates tend to move farther apart. To avoid errors 
due to this displacement, transverse expansion clips with compensating 
loops bent in them were riveted to the webs and the brackets attached 
to them, thus providing against displacement of the instrument by 
transverse movement of the webs. (See Figure 5). 

Brackets were also riveted to opposite extremities of the model, 
and to them were attached small sheaves over which a fine steel wire was 
stretched by a weight, thus providing a straight line from which lateral 
deflections were measured by scale. To secure greater accuracy and uni- 
formity of observation and avoid danger from possible fracture or movement 
at the time of destruction of the model, the compressometers were read 
through small telescopes mounted on standards and located at fixed points 
from 2 to 6 ft. distant from the specimen. 

Method of Testing 

The models, which were fabricated by the Pennsylvania Steel Com- 
pany, were carefully inspected and measured and the cross-sectional areas 
very carefully determined. Brackets were riveted to them to receive the 
instruments and the alignment wires, and the necessary holes were drilled 
to locate and receive the strain gages. The fixed head of the machine was 
locked in the required position, the bearings, castings and pins were adjusted 
and the model was delivered to the machine by an overhead crane and was 
centered and adjusted in exact alignment with the axis of the machine. 
In some cases the weight of the model was counterbalanced by the canti- 
lever beams already described, and in other cases its weight was not sup- 
ported between the ends. (See Figure 2.) 

Extreme care was exercised in making the pin bearings as accurate as 
possible; this was accomplished by a slight adjustment in the head of the 
machine. Usually one day was required for preparing the machine and 
making everything ready for the test, after which from one to three hours 
were necessary for putting the model in the machine, attaching the instru- 
ments to it, setting up the telescopes and making ready for the test. The 
test itself occupied from one to three hours and required four observers, 
one gageman, two recorders and about six laborers. Before making the 
test, sheets were prepared giving all the preliminary data, together with 
the different gage loads which were to be applied with the corresponding 
unit loads and the amounts of distortion which it was expected would 
result. 

194 



Deformation Diagrams 

The diagrams of deformation shown by the compressometers were 
plotted, giving Btraight linos from the origins to thcrelastic-limit points, 
beyond which they are, of course, broken lines, with the yield points noted 

at an arbitrary distance from the straight line, thus giving a graphic 
representation of the distortion ( ,f the specimen nude;' s! less. In every case 
the pressure at 26,000 lb. per sq. inch net for nickel steel 20,000 lb. 
gross for nickel steel, and 14,000 lb. gross for carbon-steel was released, 
and as a permanent set had then taken place, the deformation line for 
succeeding pressures commenced at a point offset from the origin and 
intersected the first line at the point of elastic limit, thus accounting for 
the two lines shown in the diagrams. 

Duplicate nickel and carbon-steel models, TXl6, 34 ft. long, made of 
two built channels latticed, were marked N and C respectively and corres- 
ponded to the upper sections of vertical posts in the center suspended span 
of the new Quebec Bridge. Each of them was spliced at the center, with 
cover plates on all sides, and had half-hole pin bearings at both ends. 
The weight was 7970 lb. and the area computed from it was 34.63 sq. in., 
being 1.61 per cent light of the nominal area of 35.2 sq. in. Each model 
was counterbalanced in the testing machine by a 1230-lb. weight applied 
at the center and at eacli quarter point to eliminate fiber stress due to the 
weight of the model. The models had a ratio L/R = 52. The specimen 
test results were exactly the same for both nickel and for both carbon-steel 
pieces. 

Strain gages were applied simultaneously to the latticed faces on both 
sides in nineteen 20-in. spaces on one model, and on both webs, and in 
eighteen spaces simultaneously on the other model. The compressometer 
was applied to lengths of 11 ft. 7 in., measured on the latticed faces at the 
center of the models, and to lengths of 34 ft. measured from center to center 
of end pins on the web-plate faces. The elastic limit of the body of the 
member was determined by observations on a short length near the center 
point, and that of the member as a whole, including its connection details 
at the ends, was determined by observations on the full length. In each 
case the initial gage reading was 30. (See Figures 1, 6 7, 8, 9, 10, 11, 
12, 29 and 30). 



195 



Engineering Record, July 25th, 1914 

ULTIMATE STRENGTH OF CARBON-STEEL MODELS 
OF QUEBEC BRIDGE MEMBERS 

Records and Diagrams of Destructive Tests; Anchor Arm Lower Chord 
Models Buckled under Unit Load of 50,886 Pounds. 

Among the eighteen tension and compression tests of models to a scale, 
about one-quarter the natural size of the members of the trusses in 
the 1800-ft. span of the new Quebec Bridge which have recently been 
made there are four sets on carbon-steel sections of the compression lower 
chords L10-12 in the anchor arm, located as shown by the diagram pub- 
lished in the Engineering Record of March 21, page 333. These members 
have the same position in the truss as those which failed in the Quebec 
Bridge collapse several years ago. In the new bridge structure they are 
proportioned for heavier total loads and for smaller unit stresses, and their 
design differs materially from that of the corresponding members in the 
former structure, which had four deep vertical webs connected only by 
end diaphragms and by top and bottom chord latticing. 

The models of the new bridge members have been made and tested 
in the interest of the accepted design which has been prepared after investi- 
gation of similar members in the other largest recent spans and may be 
considered as the latest development in compression members to resist 
excessively heavy stresses in long-span trusses. 

Models 

The models, about 30 in. wide, 18% hi. deep and 18 ft. 9 in. long, 
weighed from 7610 to 7650 lb. each and were marked TX13-A, B, 1 and 2. 
They had four deep vertical webs connected by full-length horizontal 
longitudinal latticing and tie plates on the center line and on the top and 
bottom flanges. They had ratio of L/R= 38 and a nominal cross-sectional 
area of 70.65 sq. in., which, however, owing to the inaccuracy of rolling 
the plates and angles, was reduced to an actual area of about 69 sq. in., 
making the sections a little more than 2 per cent light. Specimen tests 
made from the plates and angles showed an elastic limit varying from 39,780 
to 42,700 lb. They had half-hole bearings in both ends of the webs, which 
were reinforced to give 52 sq. in. of bearing for the 63^-in. pins. 

The accompanying illustrations are of special interest in that they 
include the first published details of the bottom chords of this great span. 
All of the models were symmetrical about their center transverse axes and 
those marked A were made with both ends exactly like the left-hand end of 

196 



the illustration, with the flanges of the center webs connected by lattice 
bars. The models marked B were exaol duplicates, except that they were 
made with both ends like the right-hand end of the illustration, with the 
flanges of the middle webs connected by tie plates. The models were sup- 
DOrted by end pins alone, without intermediate counterweights to carry 
part of the weight. (See Figure 13). 

General Resulls 

The elastic limit of the models, taken as a whole, varied from 16,866 
IK. to 22,441 ll>., and in all except the last case WBS much lower than the 
elastic limit determined on a long length measured in the center of the 
member. One model failed with a maximum unit load of 39,359 lb. The 
other three models endured without failure several repetitions of the maxi- 
mum load of the 2,S00,000-lb. capacity machine. Afterward they were 
weakened by holes drilled through the webs near the center of the piece and 
failed under loads of from 43,060 lb. to 50,886 lb. per square inch computed 
on the reduced sections. 

Former Similar Tests 

A very interesting comparison can be drawn between the results of 
the- i those published in the Engineering Record of Nov. 19, 

1910, page 564, on nickel steel models of the official design of the same 
bridge, which has since been superseded by the somewhat modified con- 
tract design now being fabricated. The nickel steel models were about 
t wice as long as these and had a considerably smaller cross-sectional area. 
They developed a maximum elastic limit of 43,947 lb. in the full-length 
measurements and had an ultimate strength beyond the capacity of the 
ing machine, so that it was necessary to weaken them also by boring 
holes, which reduced the cross-sectional area so that failure was finally 
produced by a maximum ultimate load of 63,990 lb. computed on the re- 
duced section. 

Manner of Failure 

Model TXisA-1 failed by lateral deflection in the planes of the webs 
and by buckling of the webs near the center of the piece. Diagram of 
locations of the strain gages and of the compressometers is here given. 
The locations of the compressometers were the same for all of the tests in 
this set, except that for the models marked B the distance measured on the 
latticed faces was only 127.." in. 

Model TXuA-2 was compressed to the full capacity of the machine, 
which produced a stress of 40,657 lb. without failure. Afterward two 1-in. 
holes were bored through each of the four webs near the center of the piece, 
reducing the section from 68.87 to 64.87 sq. in., and the same maximum 
total load was applied three times at high speed, producing a unit stress of 
43,164 lb. per sq. in., and on the third application caused the failure of 

197 



the model by a downward deflection and the buckling of the webs near 
the center. The strain gage locations were all made on one latticed face 
of the model, none of them being made on the webs. 

Model TX13B-I endured without failure the full load of the machine, 
producing a compression of 40,565 lb., after which two 1-in. holes were 
bored through all four of the webs near the center of the model, reducing 
the section area from 69.026 to 65.026 sq. in., and the same maximum 
load was applied again seven times at high speed, producing a stress of 
43,060 lb. and finally causing the failure of the piece by bending at the center 
in the plane of the webs and by the buckling of the webs near the center. 
The strain gages were applied at ten points on the webs on both sides 
of the model and at eight points at the upper latticed face. 

Model TX13B-2 was loaded to the full capacity of the machine, gage 
867.7, producing an ultimate compression of 40,565 lb. on the full section of 
the piece, which it endured without failure. The pressure was released and 
two 1-in. holes were bored through all four webs, reducing the section area 
4 in., and the same loading was repeated ten times. As this did not produce 
failure, two more holes were bored and the same loading repeated several 
times. Finally the section was reduced 14 in. by boring six holes through 
the two south webs and eight holes through the two north webs and the 
same load again slowly applied and maintained for 10 min., producing a 
compression of 50,886 lb. per square inch on its reduced cross-section of 
55.026 sq. in. After 10 min. application this load caused the failure 
of the model by bending transversely at the center in the plane of the 
webs, which were buckled somewhat at the location of the holes. 
(See Figures 1, 14, 15, 16, 17, 33 and 34). 



198 



Kmuneeiuxg Rbcoed, November 14th, 1914 

ULTIMATE STRENGTH OF CARBON AND NICKEL STEEL 
MODELS OF QUEBEC BRIDGE MEMBERS 

Tension Tests of Reinforced Steel Plates and Alternating-Stress Members, 
and Compression Tests of Four Struts. 

Tests recently made on eighteen large-scale models of members of long- 
span trusses of the new Quebec Bridge included comparative tests 
of nickel and carbon-steel models, described in the issue of March 21, 
page 333; compression tests of carbon-steel models of the lower-chord 
pieces with four webs, published in the issue of July 25, page 110, and 
tests of three other sets of models herein described. These latter were 
all marie on models which, like those previously discussed, were tested to 
destruction with an accurate and elaborate system of deformation measure- 
ments recorded for the study and analysis of the design. 

Tension tests were made on a set of four large nickel steel plates 
with riveted reinforcements for pin bearings and on two nickel steel riveted 
members of a t} - pe adapted to resist alternating stresses. A third group 
of tests were made of H-shaped carbon-steel struts in compression. The 
distortion of the metal was measured on the members considered as a whole, 
and locally on the webs and latticed faces and around the pin bearings, to 
investigate the construction and detailing. 

The tension plates developed a maximum elastic limit of 44,452 lb. 
and a maximum ultimate strength of 79,999 lb. in one of the two pieces 
broken. The other two pieces endured strains nearly as great without 
fracture. The alternating stress members under tension stress developed 
a maximum elastic limit of 41,526 lb., and an ultimate strength of 73,294 
lb., breaking in the body of the pieces. The H-shaped compression pieces 
developed a maximum elastic limit of 21,979 lb., which, as in the case of 
the previous compression tests, was considerably higher in most cases 
than the elastic limit determined by measurements on the full lengths of 
the pieces. The maximum ultimate strength was 35,146 lb. 

Tests of Alternating Stress Members 

Two nickel steel models adapted to resist either tension or compression 
were tested to destruction under tension, developing, like most of the 
compression models, a higher elastic limit for a measured length in the 
body of the piece than for the full length. The models were marked 
TioN-1, and TioN-2, and were duplicates, each being made of two built 
channels about 24 in. deep and 40 ft. long with their flanges turned in and 

199 



latticed and their webs bored for 8-in. pins about 37 ft. 9 in. apart. The 
models had actual gross and net sectional areas, figured from the weight, 
of 32.51 sq. in. and 26.83 sq. in respectively. The weight was 7110 lb. 
Test specimens of the angles and plates showed elastic limits of 58,500 
and 62,880 lb. respectively, and ultimate strengths of 93,000 and 93,360 
lb. The strain gages were applied in 10-in. lengths at twenty-five loca- 
tions on the flanges and webs and the compressometers were applied to 
measured lengths of 228 in. in the middles of the latticed faces, and to 
pin holes 37 ft. 9 in., center to center. 

Both models failed by breaking transversely through the body. 
(See Figures 1, 18, 19, 20 and 3l). 

Tests of Plates 

Models A2N-I, 2, 3 and 4 were nickel steel members made of single 
plates nominally 36 in. wide, Y% in. thick, and 30 ft. long, with reinforce- 
ment plates riveted to them for the bearings of two 12-in. pins 26 ft. 8 in. 
apart on centers. The specimen tests gave elastic limits of 62,880 lb. and 
66,520 lb., and ultimate strengths of 93,300 lb. and 89,360 lb. Three 
of the pieces weighed 3310 lb. each, and one of them weighed 3315 lb. The 
actual gross area computed from the weight was 22.38 sq. in. Strain 
gages were applied to one face of the model in twenty-two different loca- 
tions at the end and in the middle of the piece, to measure the elongation 
in 10-in. lengths, and four compressometers were applied on both edges and 
on the center line to measure the extensions in a length of 200 in. 

Models A2N-3 and A2N-I failed by fracture through transverse 
rows of rivets in end reinforcement plates. Models A2N-2 and A2N-4 
failed by dishing around the pin holes. (See Figures 1, 21, 22, 23 and 
32). 

Compression Tests of H-Shaped Models 

The four H-shaped carbon-steel models marked TXn-1 and 2 and 
TXis-l and 2, represented respectively vertical and inclined posts in the 
cantilever arms of the trusses. They were located as indicated in the gen- 
eral diagram published in the issue of March 21. Each was made of two 
built I-beams with latticed flanges and solid webs connected by an I-shaped 
horizontal latticed diaphragm riveted to them on the center longitudinal 
lines. At the center point they were spliced by web and flange cover 
plates, and at the ends the vertical webs were reinforced and had half-hole 
pin bearings. All of them were about 21 in. wide and 21 in. deep over all. 
The TXn models had a nominal cross-sectional area of 42.75 sq. in., a 
weight of 8950 lb., a length of 46 ft. 6 in. center to center of pin holes, and a 
ratio L/R of 78. The TX18 models had a nominal area of 56.65 sq. in., 
weighed 9550 lb. each, were 34 ft. 6 in. long center to center of pin holes 
and had a ratio L/R of 58 

200 



The TXn models were counterbalanced by 2000-lb. weights attached 

tn them at the quarter and center points, and the TXis models were 
similarly counterbalanced by S70-lb. weights. The distortions produced 
by the test loads were measured by strain gages applied to the lattice bars, 
flanges and webs of the models on short measured lengths, and by the com- 
preSBOmetera applied to measured lengths of about 19 ft. on the latticed 
faces and to the full length between pin centers on the webs of the models. 

All of the pieces failed by bending at the centers, accompanied in two 
cases by comparatively small buckling of the webs near the center. In all 
cases but one the elastic limit was higher for the measured length in the 
center of the model than for the full length between pin centers. The 
maximum elastic limit developed was 21,979 lb. and the maximum ultimate 
strength, 35,146 lb. (See Figures 1, 24, 25, 26, 27 and 28). 



201 




202 




Casting to Receive Models 
TEST 7.— Figure 3. 




V^ \ZS v3/ 

Attachment of Compressometer to Model. 

TEST 7.— Figure 4. 

203 




^2 x *\ 



f Tap Bolt 



Attachment of Brackets for Compressomefers 
\_[_9'I 



'4£-—i*"J'4£--!i 




^Platform 
under 
Vifeight 



Section A-A 
Arrangement of Compressometer Brackets 
and Counterbalance Levers. 



TEST 7.— Figure 5. 



204 



Weight of 
Test Piece *7970 



\> -II' 7'— ~*\ 4"x§' "Latticing 




\Z-l4WPIs. b VOXjePI. 
I \Z:22'd"Pk... . J4 'o'cfocof Pins- 



Considering Main Body of Member - Elastic Limit =36340 A« 
Considering Main Body of Member - Yield Point =40999V°" 
Considering Whole Member - Elastic Limit =30749V™ 

Considering Whole Member - Yield Point =38204*/ °« 

_Before Testing 



East 




South 

Models TX16 and Positions of Compressomcters 
TEST 7.— Figure 6. 



205 





20' Gage Lengths 


2i 


'.'_*U H 


3 


OoO OOOOO ooooo 


o o ooooo 


0) O OOOOOO o 

0:066000 do 


| 


7 6 5 4 


3 1 


o o 


j 


North and South Sides 


t 2 


o ' o*o ooooo o/* 
^...^._|e4jo oo of 
O _ of} ooooo oV^ 


J 




o o 


0000000 ° 
0000000000 


JO 


o'o oo o o.oo o o o o o 


o o ooooo 



East 



North 




20"Cage I South 
ejn 9 - s u-ind^io'-* H-taiL. 



Y-IO^IO'-* y-10^10'-^ *-K)'j*IO'-\ Y IO i l0 'l^op^ r 



East 
bothWebPls. 




Location of Strain Gage Measurements on Models TX16N—1 and 

TX16N-2 

TEST 7— Figure 7. 

[*- -22"--- -A L bent up /£ "in Top Fig. 

' bent down /j-'in Bott.Flg. After Testing 




0000 o o 0000 



Be rs bent Latt. Bar s bent_^ 
"in middle up 2?"/n 1 niddle /\ 



0000 00 0000 




South* Lbent down 3" in Top Fig. j_ — ^ n 

North Web, ^^t^^ f-^op^ 
3fi ] ^ 

L..-.-22"- ►! ^ I South Web had double bulge at 

"r 1 ; center, but only single bulge at top fig. 
SI 



South Web- 



a 



_j^ 1 .311 



h- -2^"— -"^15' ^"17"- A 

Model TX16G-1 at Point of Failure. 

TEST 7.— Figure 8. 
206 









_<#i. -«rf£_ 















,,.-.-_;.-.-.-= 










II 






West J 






E=- 


- 




; 


!=^---- - 



East 



L bent up §'in Top Flange 
L bent down 



North 1 After Testing 




bent down Ji in Top Flange\^y.. 21" '--^^t" ~l§ "Opening bet. Web&L 
Lbent up §■" in Bo ft Fiange~ ' 



0000 00 0000 




North Web? 



South Web- 



1 27'- 4"/4"H 




Model TX16C- 2 at Point of Failure, After Test 
TEST 7.— Figure 9. 




Member failed East of Center^ 
k— - 18'-— 

J . I H 



\ J OOOO OjO 0000 



Y---/8-—A 
Web bulged j at Middle ^L bent do wn Ji",2i k ppening bet. L & Web 



-J 

5 



^ |o O O O "o;Q o o o o <; 




_ bentdown2§^. 

W-'- 6 --.. 
Model TX16N-1 at Point of Failure. 

TEST 7.— Figure 10. 

207 



k- - 33'3$" North and South Sides *| 

hor. Leg of L bent down 3"in TopFlg. 

HorLegofLbentup3"inBottFlg.£ ^--22'--- +\ >JCz" Afte r Testing 
oooooooooo 




Member failed by buckling 



Pin Pl.^ 

h-s\ > > ,v,v, >»/ ■ / ■ /? ■ ■■ vw,w>; 

North Web' 



Hor Leg ofL bent down 2* in Top Fig. 
Hon Leg ofL bent up 3" in Bott. Fig. 
|*-— - 14* — -*j 

^2%&77 " — J3- ^4^^^P~ T ^'i/x-'-^-^ : \ 1 1 ; I / 1 / i 1 1 1 1 1 t 't t m 
\* 22" -A-— /<9" «J 



5W//7 Mfef> 



a 




E3SB <<< ■'-' -'-'j '- '■ ^ <■' ^' ^' 6 ' ' ' f f f ' 

PinP/S h 20"-- -+---/4*--- -| 

Model TX16N— 2 at Point of Failure, After Test 
TEST 7.— Figure 11. 



208 



35000 




5000 
2780 



.05 .10 .15 .20 .25 JO J5 .40 A5 .50 

Compressions in Inches 

Model TX16C-1 



35000 




.05 .10 .15 .20 .25 .30 .35 .40 .45 .50 

Compressions in Inches 

Model TX16C-2 

Stress-Deformation Diagrams of Phoenix Bridge Company's Readings. 

TEST 7.— Figure 12. 

209 



45000 



40000 



- 35000 
& 
<? 
^30000 

D_ 

c 25000 

w 
<u 

CO 



10000 



5000 
2790 



45000 















































Set 












fEL 














nr.p. 




















«S 


fe.L. 












41 






tl ( 




y^ 














A 








V^/ 
























oy 
























r* 









































































.05 .10 .15 .20 .25 .30 .35 .40 .45 .50 .55 .60 .65 
Compressions in Inches 

Model TX16N-1 




J .05 .10 .15 .20 .25 .30 .35 .40 .45 .50 .55 .60 .65 
Compressions in Inches 

Model TX16N-2 
Stress-Deformation Diagrams of Phoenix Bridge Company's Readings. 

TEST 7.— Figure 12a. 
210 




V <?&' ;-■ 

Section through Main 

Body of Both Types . 

of Chords 



2JiH'PL- 




5'xiH 



2'x2'xiL s 
5'xjk'Pls. 



Half Half 

Section Section 

b-b c-c 




Half Half 

Section Section 
■at Splice D-D 




O O p OO OJJ.O C 

-i 

O..P..°..P..P;P..°..P..P..P 



O"o"'6"O"o:S'(J"e0 6' 

.Q;.Q..ft.<?..ft.ff 



-- - - " e . -- -i 6 -- -- tf 



o"o""6"{Lr6' r 6"t5"o""o" 

■ 

0,30000. 



0"O"O O 6 
0"&!ol O..S..Q..9..9. 




. . - J 



6"o""6'"6'o"<5'e"«"B - ] 
3..P.P..Q.P..9.P-S.8.S 1 

u 



■9'4f- 



' — ~ ^~ ■ "-"'^ ■■^ — ! — -j" -_ "~": ~- . 







i=== 



\ ' — ^o J o o 

'..Q.p..P. P._P..P..P.. ,"?_p .;. ...Q. ft X. '-... [.ft. P..P.P..9..P.P..P.. .,P.,-P...°..9'-P..P..PP.-°— ..9..S..S..9 



J 

loo o o o o ooool 









■ . — -. 



OOOO 0:0 OOOOl 

p_o o o o!o o o o o| loEoi P.-Sl I? ? ( 

oT>"o"o"!"6i"o'o ol ooooooo o""6"p oooooooooooooocoo ~_o ~ 



' 



WW 



Outside- Elevation 

— ' ; .■'.": " 'C "~ ■ ■ IE 

loooo'po o o| 



iiZJ 




D* J 
Sectional Plan A-A 



B-' O' 
Detail of Carbon-Steel Model of Anchor Arm Lower-Chord Section. TEi T 7— Figure 13. 



211 





■cl 



S58J4S l\Uf) 




SS9J19 *! U D 



213 




SS3JJ5 ilUf| 



214 




































\*-T 


- K\ 


























-- v 






o 




: 






















V, 




by « 




3 


















^ 






"^ 






















~~ 


---" 


,; 


- 


^ 
























% 










- 








~"^- 
























53 




















<- 
1 


5 


y 
r 
i 




< 

c 




1 
c 

CO 


-> 
. 

*4S 


i 

1 '-T 


5 
-> 

1 


r 
<■ 




i 


i 
1 





iT 



or * 



# ^ 



a? 



03 
CO 

i— I 

H 

J 

w 

Q 
O 

S 

'- 
o 



i 


> 

Q. 


w 
OS 
P 

- 

<< 


i - 


A 


h 


u, 


H 






o 


i 






0J 


7 






z 


1 






£ 


r^ 






< 


H 






>i 


EG 






Q 


H 



u 



#>■ 



215 



4th rivet from 
endofspl.p, 



7 th rivet 
from end spl.pl. 




KZ£^4'3& i! » h "/ fig ' y 4'al*-— }• — I3'4j"—-j— 42f i! ~&Z / 



P 



~ffi 






1; I 



1 



er 



r 



39' if- — 

South Side 
Test HON 

DEFORMATION OF MODEL TlUN 1 
TEST 7.— Figure 18 



216 




217 











































































































... s 


















































» 


n\ 
























































-J 
J i 
























































































































# 






























































































































\. 


'k 




































































































x 


? 
















































'<5 


^_ 




















































fe. 1 * 




















































-- 


































" 


&< 


£2b 










< 


»> 






































*■ 


- 


> L< 





































































































































































































































SSSJjg IjUf) 























































































































































i 






























































































































































\ 






























# 
























1 


-J 
























5 - 
























































































































'< 


















































>"! 




















































<?'- 


















































*■& 




















































<y\ 


































Uj 
















.. 


















































S; 


v* 






































:A 


fe 










-^ 








































.-., 


^« 


^T 































































































































































































































OJ 




1 



TO 

•J ^ 

GJ O 

§ ^ 

I i 

fa E 

o | 






ssajjg +iuf) 



218 




Break - 60% silky, 40% granular 



& r -a 



.Hz k 



'2.'2y¥-j'/<>&'-*+ a# e -*i.--J'5"^^ //J ' 



Test A2N*l -After Testing 



_j^j£ 




1L033"-^- -3'6'~ ^5'7'^5'6'^5'7 I '^J'6,^--^I2.04I " 

**« 27'S.s --- — * 

Test A2N*2— Before Testing 




£-— 39'//" -- J —t—M 

Test A2N # 2- After Testing 




flrfoA- —50% silky, 

50%qranular _ 




r-"— -3'ioi'-^- — - 20'4%'~<\+--3'l0t'~*\ 

^ --2S'/,{" "* 

Test A2N*3 - After Testing 



fr" 



//•-^T- 




L- j'/o£»-.+U-.~30'4g!.—-4*S'/Of'»~. 

L, S pfl'/J." -»-l 



h 2Q<tf» 

Test A2N*4 — After Testing 

Details of Failures of Plates. 

TEST 7.— Figure 21. 

219 



50000 
































i 


































! 
1 


































i 

i 


































i. 


y 






































































El 
























































































































































50000 
























































































































h 


























































20000 


















































































































































































10000 




































I 


1 1 














fi 


^2N* 


1 














/ 















































































































































.1 .2 

Elongation in Inches 



ouuuu 








































































































































































50000 


























































































































































/ 
1 








































1 

I 














40000 




























If 














































































I 


:l 


























































<v 










































<^ 30000 


















































































z> 
















1 










































t 






































/ 


§ 




























20000 










1 








































~i 


7 




























































































































































10000 






















A2N 


*2 


> 












, 










































1 1 







































































































































































.2 .3 

Elongation in Inches 



STRESS-DEFORMATION DIAGRAMS OF MODELS A2N-1 AND A2N-2 
TEST 7— Figure 22 



220 




ipui 3Jont>3 jad spuncy ui poo - ) 



























^ 












































K) 






*i 




















CM 
















Sjjj 










< 






-\i 




































































































\^: 








I 


J 


r 


a 


i 


5 


' 


* 




> 


< 


- 



qou| ajDnbg J3d spunoj u; poo - ) 



221 




222 























































< 


■. 


























N 




a: 












<M 












\ 














X 


































































kl 












































*■.. 


^"^ 


a: 
























































-: 

kj 







































































Lpu| ajonbg jad spuncy u; poo-| 























































-^5 


\ 


























\ 


s \ 














i 












vl 




























— i 

Y 4 - 
































































































5^ 


X 


0: 
























































































































^ 





j. 


<M 


_c 


t- 


r 


l-H 




X 


o 


H 


v> 






Q 




4, 


F 


< 


3 


t-H 




t- 




»-( 




X! 




H 




W 




t-1 




W 




a 




o 




S 




ei 




o 




Ik 




OT 




s 




<: 




02 




a 




< 




— 




Q 




2 


o 


o 


* 


H 




«* 




S 


c 


as 




o 


<J> 


u. 




w 


a 


Q 


t 

o 


i 


1 ) 


Cfl 




w 




0J 




H 




W 



H 

DO 

- 



ipu| ajDnbg jad spunoj ui poo"i 






443"j<J%{" 
I PI. I5"x,{" 
2Pls.ZZt"xi* 





Compressometer Readings 
3i"north web^ y3^"sou(h web 




Test TX 18*1 —After Testing 

STRUCTURE, LOCATION OF COMPRESSOMETER 

READINGS AND DEFORMATION OF MODEL 

TX18-1 . 

TEST 7.— Figure 20 
















After Tesling 










• . 






— ' ] 
















* ' 


STT 


'"" 


UK8K 




J--N 


--.,--^. W -^>- 


.^"^^iii>o«»Ci>»OCS.*VVJ^&J-:-iyjC.5i* 


«««• 


i^^«Vt*>3G«A«£i^V>-;S=«"% 


Zar 




",-1 


(, 








J 


— ■ 


• 









Test TXi8*2 — CompressomeTer Readings 

DEFORMATION OF TX18-2 AND LOCATION OF COMPRESSOMETER READINGS 

TEST 7.— Figure 27 



224 




y 
































































N. 
























It 

s 












n 




ft: 

\ 




























V 






























* 


. N 
























^ 
































» 


^ 




kj 


'-, 




























- 


^ - 


^> 


































tV; 


^"^ 


^W^X 


























^^^ 



B3 



ssaj;g iiun 






^H 



H 
x 

H 
H 



fe 



H 
x 

H 



09 



H 



226 




a 




'- 






?i 


u 


-■-. 


<u 




— i 




■— 


3 




bl) 


-f 






— 


/. 


| 


^j 




< 


t - 


a> 


H 


t-i 


7 


r. 


- 


- 


r- 


■— 




o 




- 




= 




^: 





227 




228 




- 



I 

I - 

E- 
/. 

W 
- 



229 



TEST 7— 
Tests of Vertical- 





Mark 


Elastic 


Limit 


Ultimate 


Material 


At 
center, 
lb. per 
sq. in. 


In full 
length, 
lb. per 
sq. in. 


load, 
lb. per 
sq. in. 


Nickel steel 


TX16N-1 


36,340 


33,545 


47,522 


Nickel steel 


TX16N-2 


36,340 


30,749 


51,624 


Carbon steel 


TX16C-1 


19,457 


16,667 


36,597 


Carbon steel 


TX16C-2 


19,457 


19,457 


33,354 



Tension and Compression Tests of Carbon 



Mark 


Kind of 
steel 


Per cent 
elong- 
ation 

in 20 ft. 


Elastic Limit 




In portion 

at center 


In full 
length 


T10N-2. 


Nickel 

Nickel 

Nickel 

Nickel 

Nickel 

Nickel 

Carbon 

Carbon 

Carbon 

Carbon 




41,526 
41,526 


38,757 
35,989 
41,820 
36,424 


T10N-1 . 




A2X-1 


2.1 
1.875 
2.0 
1.95 


A2N-2. 




A2N-3. . 




36.422 


A2N-4 




44,452 


TX18-1. 


20,133 
20,183 
21,194 
21,979 


17,257 


TX18-2. 




17,257 


TX17-1. 




21,194 


TX17-2. 




20,409 









Compression Tests of Carron Steel 





Elastic Limit 


Ultimate 
strength 


Percentage 

model is 

light of 

nominal 

weight 


Mark 


In portion of 
member at 

center 


In full 
length 




Pounds per square inch 


TX13A1 

TX13A2 

TX13B1 

TX13B2 


19,677 
22,488 
22,441 
22,441 


16,866 
16,866 
22,441 
19,635 


39,359 
43,164 
43,060 
50,886 


2.51 
2.51 
2.29 
2.3 



230 



Table 1. 
Posi Modus 





Elevation of 




Time of 


Manner nf 


center point 


Da-e 


tests readings, 


failure 


at commence- 
ment of test 


of u 
1913 
















l'irst 


Last 


Web buckled 


1/64 in. high 


Mar. 3 


9.43 


1.15 


near center 






a.m. 


p.m. 


Web buckled 


3/32 in. high 


Mar. 10 


1 25 


5 33 


at one end 






p.m. 


j). m. 


Web buckled 


1/16 in. high 


Mar. 11 


1.13 


1. 16 


on oppoe 






I>. m. 


p.m. 


sides of 










center Bplice 










Both webs 


1/16 in. high 


Mar. 12 


1 09 


4.49 


buckled adjacent 






p.m. 


p.m. 


to center splice 











and Nickel Steel .Models 



Ultra 


Percentage 

under 
normal 
weight 


Maximum 

vertical deflection 

at center 


Time of 
test readings 


strength 


Before 

tests 


After 
tests 


First 


1. -■ 


70,983 

.'94 
79, 2". 7 








3.20 
2 13 
8.10 
1 15 
10.00 
4.05 
1.22 
1.23 
1.13 
1.23 


6 41 








3 34 








10 01 


7' '.556 








3 36 


257 








11 


77.956 








6 13 


766 

146 

32,576 

32,968 . 


ti 96 
0.96 
3.83 
3.83 


13/64 

—1/16 

—13/32 

3/16 


4 5/64 
+3 7/16 
—7 
6 15/32 


1 25 
4 14 
4.25 
4.59 



Lower-Chord Models 



Max. Vert. Deflection 
at Center, Inches 


Date of 
rts, 1913 


Time of 
test readings 


Before 
test 


After 


First 


Last 


:; 64 

1/32 

21/64 

1/32 


7 4/32 


February 2S 
March 4 

March 3 
March 5 


1 55 P.M. 

9.27 a.m. 
9.05 a.m. 
1.10 p.m. 


5 -''') P.M. 
1 14 P.M. 

1 .05 p.m. 


7 H 


1 15 P.M. 



231 



8.— TESTS ON LIFTING HITCH FOR COMPRESSION MEMBERS 

Tests made at McGill University under the supervision of 
Professor H. M. Mackay. 

The tests were made upon a specimen designed to represent the action 
of a lifting hitch, which it was proposed to use for handling heavy com- 
pression members during erection. (See plate LVII — Typical Hitches, 
Type A.) The specimen consisted essentially of a 40" x %" plate, 6' 2" 
long to which were riveted two 6 x 6 x Y" angles 6' 2" long, the com- 
bination representing one rib of the member to be lifted. To the flange 
angles were bolted two 12" channels 2' 11" long, constituting with the link 
and connecting pin a full sized model of the proposed hitch. 

When placing a member in the bridge, two of these hitches would 
always be required for each main hoist. (See plate LVIII.) Each hitch 
would, therefore, be required to carry, as a maximum, a load of 60,000 lbs. 

Under a load of 100,000 lbs. the maximum deflection observed at the 
outer margin of the flange angle was 0.047 inches and on removing the 
load a permanent set of .006 inches. The maximum deflection under a load 
of 140,000 lbs. was .079 inches and the permanent set .016 inches. At the 
end of the test it was observed that almost all the Y% inch bolts connecting 
the hitch channels to the flange angles were loose, so that in most cases the 
nuts could be readily turned with the fingers. It was evident, therefore, 
that they had been strained beyond the elastic limit and had acquired a 
slight permanent set. 

After these bolts had been tightened, a second test was made. Both 
deflections and permanent set were somewhat less. At the end of this 
test most of the bolts were again found to be loose. 

A test was then made with the bolts initially loose. The flange 
angles deflected rapidly at the earlier stages of loading but later the in- 
crease was about the same as when the bolts were initially tight. The 
permanent sets, however, were practically the same as for the case where 
the bolts were initially tight. 

Except as regards the loosening of the bolts, no visible evidences of 
weakness were observable throughout the tests, and at the conclusion no 
permanent distortion was noticeable to the eye. 



232 



9. -TESTS OF PLATES WITH PIN HOLES WITHOUT 
REINFORCEMENT 

Report by Professors II. M. MacKay and Ernest Brown. 

The following Report covers tests made for the St. Lawrence Bridge 
Co., during 1913-14, in the Testing Laboratory of McGill University upon 
tWO series of plates. 

FIRST SERIES 
The first scries consisted of four pairs of plates as follows: — 

2 Plates PI, 12^" x V 2 " x 6"^H" Net Section 3.625 sq. ins. 

2 " P2, 12M" x 5 A" x 6'-4H" " " 4.53 " " 

2 " P3, 15" x H" x 6'-4M" " " 6.09 " " 

2 " P4, 15" x %" x 6'-4H" " " 6.09 " " 

Plates Pi, P2 and P3 each had four pin holes bored to 534" diameter, the 
end holes being bored to two centres H" apart so as to make them5J4" 
long in the direction of the axis of the plate. Plate P4 had three of these holes 
replaced by a slot with semi-circular ends. The slots were 5 l A" wide and 
about 2'-ll" long. The location of these pin holes and slots are shown in the 
accompanying drawings. Slight variations were found in the dimensions of 
the plates and pin holes, but nothing to affect the general results of the tests. 
It was thought at first that the pin holes and slots might affect the strength 
of the plates so prejudicially that the maximum load of the Emery Testing 
Machine in the McGill Testing Laboratory (150,000 lbs.) might produce 
incipient failure. Such, however, was not the case. At very moderate 
loads, indeed, the elastic limit of the metal was exceeded locally; and at 
loads well within the capacity of the machine the yield point was also 
exceeded locally, as indicated by the scaling of the metal, as well as by 
extensometer readings. But no damage was done affecting the strength 
of the plates under a static load. The tests on the first series of plates, 
therefore, consisted largely in ascertaining the distribution of stress at 
various sections at loads within the elastic limit of the most highly strained 
fibres of the metal; the elastic limit and yield point of the most highly 
strained fibres; and the permanent deformations after subjecting the plates 
to the working load. The latter was specified as 21,600 lbs. per sq. in. of 
net section. 

The distribution of stress in Plates Pi to P4 at loads within the elastic 
limi t, is shown by the ordinates of shaded areas in Figures 1 and 2. 
These areas were plotted from extensometer readings taken at the points 
indicated. The gauge length was usually 2" and the extensometers near 
the edges of the plates and pin holes were centred about %" from such 
edges; so that the short portions of the curves between these points and the 

233 



edges of the metal were plotted to agree with the general form of the 
curves. As the pitch of the curves near the pin holes is very steep there is, 
in. such cases, some uncertainty as to the value of the maximum stress. 
Furthermore, the stress at the edge of a 5J4" pinhole may vary appreciably 
in a length of 2". To overcome this difficulty an extensometer was 
devised as the tests progressed, which read over a J^" gauge length, and 
which could be attached to the actual edge of an unloaded pin hole, thus 
giving the true maximum stress at such points. The readings of this 
extensometer indicated that the curves as plotted were nearly correct, 
although perhaps the maximum stresses shown are a trifle low on the 
average. A satisfactory check on the accuracy of the work is however 
provided by the consideration that, for any one plate, the stress areas at 
sections A, B, C, D, E and F should all be equal. The values of these 
areas for several plates of the series are shown in the following table. The 
value of E used in plotting the stresses was 29,800,000 lbs. per square inch. 

TEST 9— Table 1 



No. of 


Actual 
Load 


Stress Area Sq. inches at Sections 


Load 
estimated 


Plate 


A 


B 


C 


D 


E 


F 


from stress 
area 


PI 
P2 
P3 
P4 


40,000 
40,000 
50,000 
50,000 


2.02 
1.95 
1.98 
2.06 


1.97 
2.04 
1.95 


2.03 
2.02 
1.98 
1.95 


2.04 
2.00 
1.93 

1.98 


1.97 
1.94 
2.02 


1.93 
1.90 
2.02 


40,600 
39,560 
49,030 
49,925 



As these results include all errors arising from observation, variations 
in dimensions or properties of the plates themselves, plotting and measuring 
the areas, and all other sources, they constitute a very good check on the 
accuracy of the work. 

It will be noticed that the maximum fibre stress occurs, in every case 
at the edge of the loaded pin hole. As the diameter of the pin was 5", 
J4" less than the transverse diameter of the hole, there was ample oppor- 
tunity for the distortion of the metal. Doubtless the distribution of 
stress would be modified if the pin filled the hole completely. At the 
edges of the unloaded pin holes the maximum stress is somewhat less, 
running from 1.8 to 2.4 times the mean value for the section. In the body 
of the plates between the pin holes, where the distance between the centres 
of the holes is three diameters of the hole, the greatest stresses occur at the 
outer edges of the plate, and the variation of the stress from the edges to 
the centre is practically linear. On the other hand, when the centres of 
the holes are 6 diameters apart the stress at the section midway between 
them is nearly uniform across the width of the plate. In the case of the 

234 



Blotted plate, Pj, the maximum stress again occurs at the edge nearest the 
pm, and the distribution becomes more uniform at the succeeding sections, 
until at the end of the slot remote from the pin the stress at the outer edge 

of tlir plate is i:;'. ,' greater than the mean. 

These stress distributions are all at loads within the elastic limit of 
the most highly strained fibre. Whenever that limit is exceeded, the 

strains of course fail to give a measure of the stress, and the distribution of 
the latter cannot, therefore, be inferred from extensometer readings. 

Elastic Limit and Yield Point. 

The elastic limit and yield point were naturally reached first at tin 
edge of the loaded pin holes. The elastic limit occurred in all these plates 
at mean stresses from 11,000 to 13,000 lbs. per sq. inch of net section; and 
the yield points at mean stresses from 20,000 to 22,000 lbs. per sq. in. of 
net section, or in the neighbourhood of the proposed working stresses of 
21,600 lbs. per sq. in. These elastic limits and yield points were doubtless 
very local in character. 

The raising of the elastic limit and yield point due to repeated loading 
is tdearly shown in the case of Plate PS (Figure 1). Curves showing the 
relation between total load and elongation are plotted for points 1, 2 and 3, 
Section A A, and for four consecutive applications of the load. An interval 
of six days elapsed between the first and second applications of the load; 
and intervals of 18 hours and 24 hours respectively between the second and 
third, and third and fourth applications. The curves also show the total 
load expressed in lbs. per sq. inch of net section of the plate. Using this 
unit loading it will be seen that at point 3, the elastic limit was reached at a 
stress of about 11,000 lbs. per sq. inch, and was followed by a somewhat 
indefinite yield point at about 20,000 lbs. per sq. inch. The actual stress 
intensity at Point 3 can be inferred for this loading from the elongation 
which was about 0.0016" on a gauge length of 2 inches, when elongation 
ceased to be proportional to total load. This corresponds to a stress 

intensity of about 

0.0016 x 30 x 106=24,000 lbs. per sq. inch, 

2 
and as the stress intensity at the edge of the pin hole was probably some 
20% to 30% greater than this, the maximum stress was about 30,000 lbs. 
per sq. inch, a reasonable value for the elastic limit. The maximum elonga- 
tion at Point 3 for the first application of the load was 0.0048", three times 
the above amount, indicating considerable local over-straining. 

As already pointed out, the actual stresses in the plate cannot be 
infrrred from the extensometer readings, either under the first or subsequent 
loadings, once the elastic limit has been passed, even locally. Following the 
first application of the load, there was a set of 0.0017" at Point 3 and a set 

235 



of 0.00015" at Point 1. The latter does not necessarily mean that the metal 
at Point 1 had been overstrained, since the final configuration of the plate 
on removal of the load will depend on the strain of the most highly stressed 
fibres next the pin hole. An elastic strain at Point 1 will not entirely disap- 
pear when the load is removed from the plate, if the metal next the pin hole 
has been overstrained. Residual stresses will exist in the plate, and none 
of the observed elongations for the fibres at Point 1 suggest that the metal 
there was overstrained during any of the repeated loadings. 

The raising of the elastic limit by the overstraining of the"metal next 
the pin holes is shown by the continuously lengthening straight line portion 
of the curves for Point 3 for successive loads. Under the fourth application 
of the load the elastic limit and yield point became practically coincident at 
a load of 26,500 lbs. per sq. inch of net sectional area. Assuming the value 
oi the modulus of elasticity to remain unchanged (as is likely), the incre- 
ment of stress at Point 3 during the application of the fourth load up to the 
new and artificial elastic limit, as indicated by the extensometer, would be 
about 56,000 lbs. per sq. inch, and the maximum stress at the edge of the 
pin hole, being about 25% greater, would approach 70,000 lbs. per sq. inch. 
But as the residual stresses due to previous overloading cannot be estimated, 
it is impossible to infer the actual distribution of stress in this case, even at 
loads within the new elastic limit. 

Similar results are indicated by the load-elongation curves for certain 
points on Plates P3 and P4 (Figure 2), where the effects of two applica- 
tions of the load are shown. The slow yielding of the most highly strained 
fibre under a constant load is shown by the curve for Point 3, Section A A, 
Plate P4. 

Permanent Set. 

The largest permanent set observed after applying a load of 21,600 
lbs. per sq. inch of net section did not exceed 0.003" in a 10" gauge length. 
It may, therefore, be stated that the permanent sets were of no practical 
consequence. 

SECOND SERIES 

As a testing load of 150,000 lbs. was insufficient to cause failure in the 
plates described above, a second series of tests was made in July, 1914, upon 
three pairs of plates of smaller section. The dimensions were as follows : — 

2 Plates P5, 13" x %" x 5'-Sy 2 " Net Section, 3.0 sq. ins. 
2 " Pe, 11" x y 8 " x 5'-63^" " " 2.25 " " 
2 " P 7 , 9"xM"x5'-6" " " 1.5 " " 

These plates were bored to correspond with plates Pi, P2, P3 of the first 
series, except that all pin holes were 5-1/32" diameter. The corners at one 
end of each plate were clipped. The loads were applied through b" pins. 

236 



Specimens cut from the uninjured portions of the metal after testing 
gave the following results: — (See Table 2). 

rEST 9— Table 2 



Yield Point 

Ultimate Strength 
Elongation in 8". 
Reduction of area 



P5 



36S75 
65500 

56.695 



P6 



37950 
66600 
27.3% 
57.5% 



P7 



36950 
65600 
28.4% 
56.7% 



Loads were applied ininc rements of 5,000 and 10,000 lbs. until failure 
took place. In the case of Plates Ps the load was released after an appli- 
cation of 24,000 lbs. per sq. inch of net section, to observe the permanent 
deformations. Extensometer readings were made at one loaded and one 
unloaded pin hole. Extensometer swith 2" gauge length were placed as 
near as possible to the pin, and an extensometer with 3^" gauge length was 
attached to the edge of an unloaded pin hole. The elongations out to out 

of end pin holes were also observed by a scale reading to of an inch. 

1LHJU 

All these readings are plotted in Figures 3, 4, and 5. 

As indicated by the extensometer, the elastic limit at the loaded pin 
holes is reached in all the plates of the series at loads of about 13,400 lbs. 
per sq. inch of net section. As the ratio of the maximum to the mean stress 
in this case is estimated at 2.8, as explained in discussing the first series, 
the maximum fibre stress when the elastic limit is reached would be 13,400 
x 2.8 = 36,200 lbs. per sq. inch. At the edge of the unloaded pin holes the 
elastic limit is shown at about 18,000 lbs. per sq. inch of net section, as an 
average value for all the plates. As the ratio of the maximum stress at the 
edge of the pin hole to the mean stress across the section is, in this case, 
from 2.1 to 2.2 we have the value of the maximum fibre stress 18,000 x 2.15 
(say) =38,600 lbs. per sq. inch. These values for the maximum stress are 
so near the actual yield point of the steel (about 37,000 lbs.)as obtained 
from specimens cut from the plates, that it seems probable that the elastic 
limit in such cases means simply the point at which the most heavily 
stressed fibre reaches the yield point. 

The local character of the elastic limit as obtained by sensitive extenso- 
meters at the above points is shown by the elongations back to back of 
pin holes, which show little or no departure from proportionality to load, 
until loads approximating 26,000 to 27,000 lbs. per sq. inch of net section 
are reached. 

237 



The following table (Table 3) shows the values obtained for the ratio 
of the maximum to the mean stress for one loaded and one unloaded pin 
hole, in the case of all the comparable plates of series 1 and 2. 

TEST 9— Table 3 



Plate 


Section 


Loaded Pin hole 


Unloaded Pin hole 






max. /mean 


max. /mean 


PI 


12^" x y 2 " 


2.5 


1.9 


P2 


12H" x Vs" 


2.3 


2.1 


P3 


15" x %" 


3.1 


2.3 


P5 


13" x 3 S " 


3.0 




P6 


11" x y 8 " 


2.9 


2.2 


P7 


9" x %" 


2.6 


2.1 



These results show no consistent variation with the widths of the 
plates. 

The highest elongation noted back to back of pin holes at a load of 
24,000 lbs. per sq. inch of net section was 0.08" and the highest permanent 
set in the same distance after the above load was 0.02". The first two 
plates tested, Pz-l and P-6l, failed by dishing behind the pin at the end 
whose corners were clipped. The next three, Ps-1, P6-2 and P7-2, had the 
clipped end blocked behind the pin hole so as to prevent dishing there. 
Of these Ps-1 and P6-2 failed by dishing behind the pin at the square 
end. P7-2 (section 9" x %") failed in tension across an unloaded pin hole. 
The remaining plate Ps-2 had both ends blocked. It failed by splitting 
longitudinally behind the pin. The accompanying table (Table 4) gives 
the essential data as to the ultimate loads and modes of failure. 

The following points made evident by this table may be noted: — 

(1) Clipping the corners behind the pins increases the tendency to dish. 

(2) Only two plates were loaded to anything like their capacity to 
resist tension. Of these, one, P7-I, withstood without failure in tension, 
a tensile stress practically as great as the ultimate strength of a specimen 
cut from the same plate. The other, P7-2, failed in tension under a tensile 
stress per sq. inch of net section 1.04 times as great as the ultimate strength 
of a speciment cut from the same plate. 

(3) When the ends are unrestrained the ratio of total width or net 
width to thickness does not seem to have much effect upon the capacity 
to resist dishing behind the pins, that is to say, upon the strength of the 
plates. The wider plates, in fact, show a little higher strength than the 
narrower ones of the same thickness. From a comparison of the plates of 

238 







TJ 










T3 






c 






— 




a 






W 


50 








w 




J 


— 


c 




— 

£ 




13 
CO 
O, 

a 




- 

eS 


Lf" 


— 
o 




~ 




o 




S 


— 
gfl 

— 
- 


r. 
— 


1 

— 


u 

-. 
— 


-- 
E 
E 


••3 
99 






_* 




/ 




/ 


-^ 






o 


-t- 




-_i 


— 


w 






o 


o 
3 


- 


5 
S 


a 


- 




-*->.-. GO 

- - / 
















bio N 
idth t 
ickne 


co 


_z 


o 

1— I 


.£ 


© 


_z 




Tl 














- - - 
















Tensile 

Si rese mi 
Failure 
M- per 
Sq. in. 


g 
o 


= 


g 
g 
5 

— '■ 




— 
o 


o 
co 






ro 


CO 








CO 


— 


^ 


— 


-^ 


CO 




-a | g fe . 

^ - — — ,s 


§ 


g 
o 


- 

K 

zT 

>z 


g 
— 


o 


g 
g 


a 

j 
cs 
-- 
- 


I'm 11 
in- Si 
:it Fai 

Sq. 


- 
i- 

US 


— 


so 

■ -. 


M 

?r 


1Q 

UO 




TJ 




d 


-c 
p 


~ 






i 


— 
- 




@ 


H 


1 




cs 






~ 


- 


— 




- 


a 


D 

E 


.9 




— 


•_ 




02 








— 










•I 


GO 


g 


— 
— 


~ 


~ 






_i 


oj 


_2 


39 

— 


T3 


— 













- 


OJ 


L 






« 


~- 


— 






M 


1 




,— , 




— 


~i. 


on 








.2 


IX 


o 


- 


Q 


3J 




imate 

oad 

,bfl. 


o 


8 


o 


g 

o 


g 


3 




o 


CO 


o 
of 


71 

— r 


/ 


g 




5 


/ 


— 


- 


O 


:■ 


co" 




© 

IH 










o 

1— I 




_ z Q 
















r 




'- 










- z ~ 

* : r 


CO 




— 


z\ 


: 


1 13 


_z 




















X 


J 




5 




; 






a 


/ 




X 










.2 
















v. 


X 


-5 


X 


C 


K 


o 
— 






■" 














5 


*■* 




■"" ' 










■2 


■ 


z\ 


•—I 


Z\ 


,_, 


•M 




5? 


IQ 




■i 


■z 


1- 


l^. 




~ 




- 


- 




s 


Ph 






this series with the large plates tested at Ambridge in September, 1914, it 

would seem that the ratio of diameter of pin hole to thickness of the plate is 

an important factor. Thus two plates 28" x 2" with 12" pin holes 
j 

(— = 6) stood a bearing pressure of 79,000 lbs. per sq. inch without sign of 

failure by dishing and may be credited with a bearing capacity of at least 

80,000 lbs. per sq. inch. Two plates 26" x iy 2 " with a 12" pin hole 

d 
(——8) showed signs of dishing (although not enough to cause failure) 

at an average bearing stress of 71,700 lbs. per sq. inch. The average bearing 
stress at the point of failure by dishing for the three widths of plates included 

in series 2 is 56,000 lbs. per sq ; inch and — = 13.3. Plotting these results it is 
observed that a straight line is obtained whose equation is nearly repre- 
sented by p = 100,000 — 3,300 — , where "p" is the bearing pressure to produce 

dishing "d" the diameter of the pin hole and "t" the thickness of the plate. 
This suggestion is made with all reserve, on account of the inadequacy of the 
data on which it is based. But on the other hand the relation seems too 
striking to be altogether accidental. 

Table 5 gives the diameters of all pin holes before and after testing to 
failure. 

The broad conclusion to be derived from the tests of series 2, as well 
as from the full sized plates tested at Ambridge is, that when the ratio of 
thickness to the diameter of the pin holes is sufficient to prevent dishing, 
the plates will develop a strength in tension per unit of net sectional area 
very nearly or quite equal to the ultimate strength of the metal of which the 
plates are composed; and that this is true notwithstanding the very uneven 
distribution of stress at loads within the elastic limit. 



240 



TEST 9— Table 5 





D 


intensions of Pin boles before and after failure 


Plate 


Loaded hole 


Unloaded 

hole 


Unloaded 
hole 


Loaded hole 




Bor. 


Vert. 


Bor. 


Vert. 


Bor. 


Vert. 


Bor. 


Vert, 


Original 


















A 11 Plates 


5.03 


5.03 


5.03 


5.03 


5.03 


5.03 


5.03 


5.03 


l'-l 


5.02 


5.12 


5.02 


5.04 


5.03 


5.04 


5.00 


5.17 


l > ..-'_ , 


5.00 


5.90 


4.96 


5.20 


4.96 


5.20 


5.00 


5.36 


Pe-1 


5.00 


5.26 


5.00 


5.12 


5.00 


5.10 


5.01 


5.18 


1' ■-_' 


5.00 


5.24 


4.98 


5.15 


4.98 


5.15 


5.04 


5.32 


P:-l 


5.00 


5.60 


4.85 


5.48 


4.83 


5.48 


5.00 


5.48 


P7-2 


5.00 


5.74 


4.94 


6.14 

* 


4.82 


5.76 


5.00 


5.68 



*Tenaion failure al section through this pin hole. 



241 




Plate: Pi 



lEi > ± * bi-' lona Total load UO 000 Ids 

Stress diaarams I ~ 10.000 lbs iq m 

Cauqe lenath I 



a. 7 « 

■6 243/0 






Pom f / 








/°o 


n 


► 2 














































— 


'JO 

lit 
no 

=5/00 

~s 90 

■5 t' 

S 7" 

t, 
■" 

(5 '" 


























































L 














































































































jff 




























































1 "l-f 
c 70*0 

° "W 




































































































































" 












































' / 










J 


2 


















































' 


















































t 






























j 
















































/ 1/1 1 j 






































' / "/ 


































-* 


\pi: . 


I * f / 


















JS\ 






J 










I ! 










1 




















































j 





Elongations ottr 2' gauge length at Section An on P/ate P2 un </tr 




PLATE P2 I2i X j '* t >V ' lonq Total load SooOOlbr • IIOSO Iksfsg m on , Q 

Stress diaarams /' - 10 000 lbs I sg in Gauge 



I I S I • I gure 1 



_i. 



Plate Pi 



Hit i> 6V/imj Total load iO 000 lbs - il.ioolbs/sai 

Stress diaarams . / - 10.000 Ibslso m 
Cauae length i' 



net section - bkoo Ibslstj in on aross sectu 







Pomtl 










Po 




' 2 




















































flwif 


I 


























































A7M 


<s. 


._:':. : 




4-+ 








1 ' 1 












































^ 




























































_ =3/Oo 


lIUvTl 




TTT 










4= 
























































































































































































- 














































t ubo 


— 1 '° 

— S 7° 
i (i 






















































































































; .. iso 




; x*j ' 










































































































. u;*?" 


/■ : / ' _/ 
















































































































/ ■ ' / 


/ 






















































































































/// / 


1 ; /i L-* 


/ 
















































< 


































































a ff40 




y/f 7 


~ / V 




















































/ 






































































'/ / 










/ 












































































































//J / 












' 






































































































^ ?;.o 


ill J 


/ / 












































































































^ 


.If 7 ' ■ <i 


<r / 


























































1 




















































1 1 ! ! 






i 




















? 






















: 










| 










i 








































1 

























Eton 3 at,ons o*tr ? aouat length at S^han AA or, PUfe PZ under r.ptottd /o. 



I division - 0-0002" elonqation. 




© 



TEST '.)— Figure 1 242 




A B C D E 

iLATE P3 IS\fn6\'/ona. Total load So.000 lb: 

Stress diagrams 
Qauat Iznqth 2 e ° 




A 8 C D £~ 

PLATE P/f. IS\%*bX lon<] Total load So.000 lbs 

Stress diaarams I . 
C,au<jt lenath 2 < o 



I I ST 9 -Figure 2 



243 



Plate Pz Etted .1 r, pta uj f„j,-, 



. 7* 





rLATE l3 /5»|'x6V lon(j Total load So 000 lbs - 8,200 Ibslsa in on net section = SiiO Ibs/sq.in on aross sectn 

Stress Jiaqratns /'- 10.000 lbs I S(j in 
Qauae lenqth 2 except as noted 





Section U 


Pent 3 Section 


CC Pent «. 




am 


lit. ■ 


1 J III 


\\ . 


5 io Ha 


il' 


^ 




* ::r 


n /.o 








* tot 


,<T / 










^ z - 
















fl , 


' 7° ! rf 


/ 


1 ' i 






X Z ' // 




1 




J" 8:o« 


~so ^ZZ. 


/ 






fcuo 


■»« i // 


/ : 


I " 






3 * 1/ ' 










■s" // ' 








r : » A / 


//iii 






,A / 


, ,jfi , i i ,.i 



Elcnqati.ns Orel- /«.««! knqth / rfir.sion - .0001" 





rLATbZ P 1+ /5\\',bX lony Total load So.Ooolbs - 8200 lbs jsom on net section - 5 ilo lbs/so m on aross^section 

Stress Jiaqrams /' - lO.OOOlbsJsqin. 
Goua,e length 2' except as noted. 



PLATC P/j. ElUct si rtptattJ l.oJmq . and el tim inter™ s Jv.ni, Lading. F,rst loaj.nn *>0OO/is. 

S.c.nJ Ucdmq /SO.OOO/is Y.Mny ot S. -■ ,n H. k»t n.t at S«t.°n, CC «-* ££ 

Scct.nnCC Se, 





— -s'* 

= «» 

.- ?7< 
















i 








• E 


S«h 




» 




---• 


3 




















-.- --. 




















'; 






- 




































I 


























































. 




■ 










■: 
















































, 


1 i 


e v 
















































■ 














< 






i\ 






















































i 








r 






















































. 


1 






- 










































~r- 












\ 


I 


■j "'" 




t 










































/ 












... 




E 4tu 














































-b- 












. 


4 — 


i — — 


*" J 200 

S~trto 




-/ 












































- 












— 


m 


I — 


1 jjjj 


i 










































y 


7 










i 1 


\ 




f 


I 








! 




















s 












u 


*- 








j 








1 6- 1 .' j J : ; ; 



Elcnj.ti.ns .,,r 2" a.uat Ur.th . J-d.,is.~> - Coo 



© 







4'un io 



Plate Ps. 



Ib'x^*5'8i' lg < 



*tm 


//O 








A 


rffi 


p 


5 - 


/ 




























^ JO, IOS 
"v 26760 
c 23. uS 


/Oo 

— ,3 I" 

— 1 7° 




































































'. 


^> 






































J« 


y 




























M 
































c 2oojt 
~-.f6.7iT 

» 'o.ojr 


S < 












7 
































— 3 


^S Jo 












-J 
































c 












.7 






























_ J Jo 
— £ 2° 






























































































































"£ J, J** 


■ 
























































































& 01 oz 0* *■ OocS 00,0 oo.s 
Eloign btobofp'is Clonyn over \ andZ qau(jt lengths sho~n Id.vn - O 




Idt.n - O ./" 



3T 9 -Figure 3 



244 















■ 5 8i' 










D 










*■(.• 








— »■- 






' li 




-Ilk 




- Z'i' 






-7i" ■ 


'i 






/\ 




c 














; 


/ .j-i 

T 




£"^i" 9° m s* « n pJ>jp «f "°'* ^ 


,$M 


f J&\ 










i 


i 


k 


J V s " 5 * 












" 






^ 


3 








/i 


c 










« 


-/ITE /°5 li'xl{*5'8t 


h c <" j r 


/en^/i 2'erM . i 


at CC 

Plate P5 - 2 




P/Ve P5 - / 

4J.4H5 




/Jo 

Ita 

^? oo 

*o »0 

i 7° 

S to 
* 50 

- 40 
1 JO 

~- 20 
1-2 /o 






















































gu«si 

ij io. IOS 

"t zt.ybo 

' "^ /3. 380 
; lo.ojr 




/IO 

:£ *> 

"S Jo 

1 7° 

; .0 




























































































































































































































•/ 












i 


5^ 


















c 2*. 7=0 




































y 












i*i 






















































.:'/ 
































E ». M 


































a 


































~«r a. ]2s 


























A' 
§ Jo 

~5 so 

»S ,0 












^ 


































































































M /o 








































































$, «>'»• 




































































- 
































































































( 




Elony, btob °lf>,*S E.lon<jn O-tr ^ and, Z' aauoe /cnof/is 5>iO~n /J,™ - O OOl' 




Elonan tocA fe toeA of p.ns f<frvn - 0-/* 



TEST 9 — Figure 3 







/ li 




£ oauat on edor o( hole. 



-4'Ufz back to pa* 



Plate Pb. ti*i*s'u'/ $ . 



Plate Pb-I 




Elonnn back to bock o( pins Elena" ortr f Ond 2~ apuar Itngths sho 



I di.n • Ol 



I don - O OOI 



1 EST 9 Figure 4 



245 





C'ii' 
























•" 






4 


C 










1 


/ V 


p: 


3* 


,,«. 8 ^ - 




^\ /" 


"~\ 




1 








>«j 








f 


-*! f 


-n f 


■*} f 


^ 


! 




V 


[ 




V 


y v* 


P ^ 


J ^ 


y 












*o 




- 


1 

1 












1 




Lj 


/i !c 




















PLATE lb. IU%*5'bilq Gauqe lenqth 2'etAA. i'atCC. 




Plate Pb-I . Plate Pb - 2 






MtJB 




IOC 

1" 

.-£ 80 

** 7° 
1 bo 

1 5 ° 

40 
-i *> 








































"■.700 

^ J J. 760 

s,ol« "t 

C J'. Zoo 

° 24 820 

=S /7.«M 
i ».?»« 

> i.A-7° 


too 

— * ? ° 

=5 80 

- «?■ 
§ 60 

_ §«g 




































■ 1 | 






i 












































































J^ 


















, 






















«sJ 


























*2* 














7 
















i» 


















■-- 


01 


• 


























^•j 




























u 










r-N 


► ■ 


iJZ. 




























^ 














I 










7 














































^ 






*s 


."****~ 






















J 




















































































































































































"" JO 

h2 '» 


























































































































L 
































































1 














^ o» 02 oj o ^ os ^ o-oos o-o/o ^ 0-i o i oi o 4 os oi o 7 o 6 w aoos OOfO 




f/^ ?n bot-ktohock of f .l f/on," „»»/• i" W ,?',«„,€ /.nfltts 5* » n £/on ? n tot/, h. bod, ol prns £'•»)»' »! °"d * O.uO, /«„jf<.S Sho-n 




Id!,. . 01' ld„n -oool" /rf.vn - Cl" /*»» » Ooo.' __ 



1 EST I) — Figure 4 




4-Hjz back to back of 
iL A TE f J y\l,*5'b'/ono Caucje /en 



Piatt Pi - / 




I 02 03 o* o-S 04 07 o-s 
elongation back to back of pins 

/<//,* -- or 



Llongation ot*r X and 2 gauge lenqtht sho 
Id'." ' O oof 



l'9 — Figure 5 



246 













5'6' 








































4' 6" 










A" 






- /'/i 




l' li' - 








2'3" 




6' — ■ 


'I 






/1 




c 














1 £j! 




C^ qauJp on edat of holt 

SM / 




^\ 


.1 




7 




'Y. 


* 








v,^ 


/ v/ 


/ V 9" 




/ V 












j 










7 V 


.— ' 




•^ 








/I c 












•-a'-J 




- 






/LATE Ij a\^ x S'6'lonq Gautje len 


<jth 2" at A A ,j"afCC. 




Plate Pi - 1 


Rate Pi -2 

Elongation back to Lack of pms . t dir" - 0-/" 


to] 200 




100 










































(7,»h /oo 


»/' 02" OJ' oV ° 


s 1.0' is' 20 
















































































"i-B.Tk 




JJ «» 
















































*J}76o 


=3 80 
















prt 








































JB 










































^^"' 


































,3 litter 





1 fe 
3 so 
* 4* 






9^ 










































40, J 20 


io 
a So 

c 40 










































Y\ 
























$£ 


0°*' 






























r.c 


^i- 






































.> 


, t ^^ 




























^ 


i> 


t^i- 


* 




















j 1 


















^^ ^ 




































-"*! 




























•E 2».to 




1 * 

-2 20 








































„ ?o. >°o 


-« Jo 

^1 20 












J 


































1 


















I / 


























/ 


/ 


/ 


































* i.f' 




l-s 


* 


















1/ 




















^ 6,7,, 


hS '» 












b 


/ 






















































tf 
























P 






































^ 0/ 02 Oi o* 05 06 of OS w ooos' OOIO* 


^ ^ OOOS.' O-OIO 0-OiS' 




ClooQohon bock to bock of pms Llonoot/on over 4" ond Z" gauaz lenqtlis j/?owo. 


Elongation oyer ±* and z" oauge lengths sAown 




Id,,' ■ 0.(" /rf,^" - Ool" 


/ d<¥ n - OOOl' 

© 



I B8T 9 — Figure 5 



in— TESTS ON 2 28 a - WD 2 26 k 1) PLATES TO 

Hi i:r\i\ Till li: i LTIMATE STRENGTH 

The plates were tested on the large testing machine of the An* i 
Bridge Co., a" Ambridge, Pa., in September, L914, under the supervision 
of Professor B. M Mackay, who reported on the results ol the tests m 
pari, as follows: — 

The detail dimensions of the p shown in Fig. No. L. 

The object was tain the ultimate strength in tension and also 

to observe the behaviour of the plates under load, and more especially 
at loads of 18,000 lbs. and 25,000 lbs. per square inchofnel section. 

The fcnbridg machine is so arranged that, in the case of 

large test piece! very rapidly. It is, therefore, difficult 

to stop a. any predetermined load below .1.- yield point. It i, unpossible 
to hold the load stationary. Tin- above mentioned loads were, on tins 
account, considerabl: ■ '» some c 

A brief account of the individual tests follows:— 

Plate l. Nominal 

Actual thickness 2.02". 
Actual net section v > sq< ilis - 

At 19,500 lbs. persq. in. of net Bection scaling was distinctly noticeable 
back of and at the sides of the loaded pin hoi 

, lbs per sq. in. the sealing was more pronounced about the 
loaded pin holes. No scaling was uoticeable about the unloaded to 

\t 55 000 lbs. per sq. in. of net section (about 32, I lbs. per -1- inch 

of gross section) scaling been r the body ol the plate. 

\, 57,800 lbs. pei aq. inch the plal - ,llL ' 

loaded pinhol lead" end. The fractun <»"■<• 

with some tracefi of included slag. 

Plate '.. Nominal diment - •• -"• 
tal thickm 

\, iQQOOlbs. persq. in. mo ding had occurred back of the 

loaded pin holes. Tb """" 1 J he ^T 

pinholes, and the load n rmanenl d iW be 

detected in the unloaded holes with b micromel to one 

thousandth of an inch. 













7 







I" 
£ 1 



S r=i 



m 

m 



248 



At 29,400 lbs. per Bq. inch sealing was heavy about the loaded pin holes, 
and had extended to unloaded holes. Upon release of the load, the un- 
loaded pin holes showed permanent deformations of 0.016" transversely 
and 0.04S" longitudinally. 

At 46,900 lbs. per sq. inch, scaling was general over the body of the 
plate. 

At 58,600 lbs. per sq. inch the plate failed by splitting axially behind 
the loaded pin hole at the "dead" end. Fracture coarse crystalline. 

Plate 8. Nominal dimensions 26" x 1)4". 

Actual net section (least) 21.15 sq. inches. 

At 23,800 lbs. per sq. inch moderate scaling occurred back of loaded 
pin holes. 

At 29,800 lbs. per sq. inch heavy scaling occurred at loaded pin holes, 
moderate scaling at other pin holes. 

At 60,300 lbs. per sq. inch the plate fractured transversely across 
the unloaded pin hole nearest the "dead" end. Fracture coarse crystalline. 
A slight tendency to "dish" was noticed at the "live" end. 

Plate 4- Nominal dimensions 26" x 114". 

Actual net section (least) 21.09 sq. inches. 

At 18,000 lbs. per sq. inch slight scaling occurred back of loaded pin 
holes. Upon releasing the load no permanent set was perceptible in the 
unloaded pin holes. 

At 26,100 lbs. per sq. inch scaling was noticed at all pin holes. The 
permanent deformation of the unloaded pin holes was 0.006" transversely 
and 0.004" longitudinally. 

At 61,900 lbs. per sq. inch the plate fractured transversely through the 
loaded pin hole at the "dead" end. A tendency to dish was also evident, 
amounting to a deflection of about \% inch from a straight line drawn 
transversely across the plate behind the loaded pin hole. Fracture was 
coarsely crystalline. 

The ends of the plates behind the pins were not restrained by blocking 
or in any other way. While both l-}^" plates showed some tendency to 
"dish," no such tendency appeared in the 2" plates. The ratio of net 
width through the pin hole to thickness, in the former plates was 9.35, and 
in the latter 8.0. It may be noted that, in the case of the small plates 
tested at McGill, the only ones with unblocked ends which did not fail by 
dishing had the above ratio about 10.6 

249 



The following Table 1. gives the reduction of area at a right section 
across all pin holes. 

TEST 10— Table 1 



Plate 


Dimensions 

Nominal 

Inches 


Ultimate 
strength 

lbs./sq. in. 

Net section 


Reduction of area % 


No. 


P.H. 1 


P.H. 2 


P.H. 3 


P.H. 4 


1 
2 
2 
4 


28x2 

28x2 

26x1^ 

26x1^ 


57,800 
58,600 
60,300 
61,900 


11.2 
10.6 
14.6 
13.4 


13.1 
10.25 
14.0 
13.1 


14.0 
11.6 
17. 6x 
14.4 


19. 2x 
14. lx 

17.0 
17. 7x 



x — Point of Failure. 



250 



11.— TESTS ON LINKS REMOVED FROM THE BANGERS 
i BED IN RAISING THE SUSPENDED SPAN IN L916 

After the loss of the lit st span on September llth, 1916, careful 
measurements were made on all the links t<> ascertain if any had been 
seriously over-stressed. The measurements showed that all had received 
a permanent set, particularly the ones on the South-Easl and North-West 
corners, but none of an amount that would in itself condemn them for 
ose when raising the second span. As, however, there existed a prejudice 

againsl using these links again, it was decided to have several of them tested 
to destruction at the Ambiidge Plant of the American Bridge Co. 

Each link was built up of 2 — oO x lU plates stitch-riveted together, 
as shown in Figure No. 1. 

The ultimate strength developed was surprisingly low and as there 
seemed to be little doubt that this was largely influenced by the stitch- 
rivets, rather than by any injury which the links may have received in the 
accident, it was decided to order an entirely new set, and to avoid the use 
of stitch-rivets. The new links were single plates 28" x l}i" as shown on 
plate CI and were assembled as shown on plate XCIII. 

Professor H. M. Mackay was present when the tests were made on the 
links at Ambridge and reported as follows: 

The following Table 1. summarizes the results as regards the ultimate 
strength figured on the basis of a net sectional area of 39.375 square inches. 

TEST 11.— Table 1 







Ultimate 






Maximum 


Strength 




Member 


load 


lbs. per 
sq. inch 


Remarks 


ELHiE 


1 175,000 


37500 


Split behind slotted hole. 


ELHiF 


1,500,000 


38100 


" round " 


ELH3C 


/ 1,500,000 


38100 


One plate cracked. 




1,620,000 


41100 


Both plates split behind pin hole. 


ELH3D 


1,550,000 


39400 


Split behind pin hole. 


ELHaE 


1 ,."-75,000 


40000 


ii ii ii 


ELH3F 


2,090,000 


53100 


One plate split behind pin hole. 
" " dished about %". 


1 I.H4C 


1,635,000 


41500 


Split behind pin hole. 


ELH4D 


1,570,000 


40000 


i< 11 11 11 



251 





Sketch shonina arrangement of fit rets 
and ■ T/Pical Method ofFathre. 



TestW Fif.1. 



252 



The mode of failure was very similar in all plates, except as noted 
above. Usually B crack first developed suddenly and jvith a sharp report 
from the end of the plate either to the rivet on the axis behind the pin hole, 
or to one of the rivets nearest the end and adjacent to the axis of the plate. 
Upon continued application of the load the crack would then extend from 
the rivet in question to the pin hole. 

In the case of ELH2E, the first member tested, the loading pins were 
not parallel owing to the condition of the bearings. The divergence was 
about 5/1G inch in possibly about 6" of length, with an indeterminate 
divergence due to the shape of the pins. This member gave the lowest 
result. For subsequent tests the machine was corrected. In other tests 
some of the loaded holes were as nearly as possible at right angles to the 
BOOB, while in others variations of the bearing surface of the pin hole from 
1/64 to 3/64 from the perpendicular were noted. It is not possible, how- 
to trace any relation between this inclinat ion and the ultimate strength 
or mode of failure. 

In some cases, too, one of the two plates riveted together projected 
at the end beyond the other, the maximum amount of such projection 
being about two inches. So far as could be observed, or inferred from the 
fracture, the longer of such plates cracked first. 

Tests were made later at McGill University on fourteen model plates. 
These were, except as regards length, one-fourth the linear dimensions of 
the actual hangers and were designed to show the effect of the following 
variables: — 

(1) Sheared vs. Finished Edges. 

(2) Round vs. Oblong Pin holes. 

(3) Riveted vs. Plain Plal 

The results showed no appreciable difference as between sheared and 
finished edges or as between round and oblong holes. The plain plates 
developed on the average an ultimate strength equal to 99% of the test 
specimens, while the riveted plates developed only 79.5% of the speci- 
mens. Excepting one case of dishing all the riveted plates failed by split- 
ting behind the pin holes in the characteristic way, while none of the plain 
plates failed that way all breaking transversely. 



253 



12— TESTS ON PLATES IN TENSION TO INVESTIGATE THE 

EFFECT OF INCLINING THE PLATE TO THE 

AXIS OF LOAD 

Four Plates 28" x 1J^" were used for each hanger when lifting the 
suspended span In 1917. The stresses in these plates are shown on plate 
CI and the manner in which they were arranged on plate XCIIL 

The design was criticized by Mr. Joseph Mayer of the staff of the' 

Board of Engineers, who claimed, that under certain unfavorable conditions 

of loading, very heavy bending stresses would result at the ends of the 

plates, which combined with the direct stress would give a stress in excess 

of the elastic limit of the material used; that this stress might alternate 

from one side of the plate to the other during the lifting operation, and 

finally that the knowledge of the behaviour of plates under such condition of 

loading was not sufficient to warrant their use in the manner proposed. 

His argument was based on the assumption that the bending stress which 

1 dT 
might be expected was expressed by the formula S = — - where S = 

R aL-2 

the bending stress in lbs. per square inch due to the end moment, R = the 

net section modulus (across the loaded pin hole), d = the relative motion 

of the center of bearing of the two ends of the plate, T = the total load, a = 

v / — , I = the gross moment of inertia, E = Young's modulus = 29,000,000 
\/ EI 

L = the distance center to center of end pins. 

There was no disputing the accuracy of the calculations required to 
develop the formula for the bending stress but there did seem to exist 
reasonable grounds to question the accuracy of the assumptions on which 
they were made. A test was, therefore made at McGill University on four 
plates, one-fourth the linear dimensions, except as to length, of the 28 x 1H" 
plates, with a view to discovering if these bending stresses did exist and if 
so what effect they had on the ultimate strength of the plate. 

A copy of Professor H. M. Mackay's report, under whose supervision 
this test was made, follows: — ■ 

This Report describes a series of tests on four plates 7" x %" x 5' 334" 
over all, and bored with holes for 3 inch pins. The end pin holes were 4' 6" 
c. to c. of pins, and each hole was bored to two radii 1J£" an d 1-9/16" 
respectively, from centre 5/32" apart. The three intermediate holes were 
circular, and Q/%' in diameter. The object of the tests was to find how 
the strength of the plates was affected, when loaded through the end pins, 
by inclining the plates to the axis of the load, and also to find how the stress 
distribution on a section through the loaded pin hole was affected by such 
inclination. 

254 



Two specimens out from the same stock as the plates were tested with 
the follow ing results: — 

A Sfield Point 30,600 lbs. sq. in. 

Ultimate Strength, 63,200 lbs. Bq. in. 

B Yield Poinl 30,200 lbs. Bq. in. 

Ultimate Strength 63,400 lbs. sq. in. 

The first two tests (Plates Tl6-1 and 2) were made as follows: — 

The plate was assembled in the testing machine with the upper end 
placed centrally on the pin (in the axis of the machine). The lower end 
was deflected ' ■_, inch, say northerly, and a load of 36,000 lbs. was applied, 
or about 25,800 lbs. per Bq. inch of net section. This load was held for a 
short time, released and then the lower end of the plate was deflected 
1 _. inch off centre in the opposite direction, and the same load applied. 
The elongation hack to back of pinholes was noted in each case. After 
repeating the above loading thirty or more times, the plate was tested to 
destruction. 

Plate 1 failed 1)V fracture across a loaded pin hole at 92,500 lbs. total 
load, or 03,800 lbs. per sq. inch of net section, a little above the ultimate 
strength of the test specimen. Plate J failed by dishing behind the pin hole 
at '.10,00(1 lbs. total load or 02,070 lbs. per sq. inch of net section, about 
0.495 below the average ultimate strength of the test specimens. The 
ultimate bearing Stress was, however, 80,000 lbs. per sq. inch. From 
previous tests on plates an ultimate bearing strength for structural steel 

was deduced =100,000 — 3300 — , when d — diameter of pin and t = 

thickness of plate. This would give an ultimate bearing strength in the 
present case of 73,600 lbs. per Bq. inch, which was exceeded. It is apparent, 
therefore, that the ultimate strength of these plates whether as regards 
tension or bearing capacity was not lowered by the repeated applications 
of the inclined load. 

The third test was designed to measure the actual stresses on a section 
through a loaded pin hole with the plate inclined to the axis of the load. 
For this purpose mirror extensometers with J^ inch gauge lengths were 
used, one on each face of the plate. As the plate bends on application of 
load, all mirrors attached to it rotate. A third mirror, was, therefore, 
clamped to the plate as near to the extensometers as possible to measure the 
amount of such rotation and furnish a correction to the readings of the 
extensometers. Two complications were encountered in this measurement. 
First, in nearly all plates, owing to kinks, bends, accidental eccentricities, 
etc., the observed Btressee on the two faces of a plate are seldom the same 
no matter now carefully the load is centered. Second, the stress distribu- 
tion across a section through a pin hole is far from uniform, the stress at the 

of the hole often being as much as three times the mean stress. The 
first of these difficulties was overcome by loading the plate centrally to 
begin with, and measuring the deformations in that case, and subsequently 

265 



comparing these with the deformations when one end was offset. To 
obviate the second, the extensometers were set as nearly as possible at the 
point of mean stress on the transverse section through the pin hole. Care 
had to be taken, however, to keep the loads low so that the elastic limit at 
no point should be exceeded. For if the elastic limit were once exceeded 
at any point, subsequent extensometer readings might be incapable of 
interpretation. 

Table 1 gives the measured stresses under a central load on both 
faces of the plate on a section through the loaded pin hole. The extenso- 
meters were placed four-tenths of the distance from the edge of the hole 
to the edge of the plate, at the point where former experience showed the 
mean stress in the plate might be expected. As the extensometers bear, 
not on a point, but over a line of finite length, they can only be placed 
approximately. However, the agreement of the mean stress and the 
measured stress computed for E = 29,000,000 lbs. per sq. inch agrees very 
closely and constitutes a satisfactory check on the working of the instru- 
ments. It will be seen that although the plate was unusually straight 
there is a considerable difference between the stresses on the two faces of 
the plates under a central loading. 

Table 2* gives the corresponding results when the top of the plate 
is offset 5/16" to the south. The agreement of the mean measured stress 
with the actual mean is not quite so close but is fairly satisfactory, when it 
is considered that an error of one one-hundred-thousandth inch in measure- 
ment means about 600 lbs. per square inch in stress. 

The excess of stress on the north side gives the total fibre stress due 
to bending plus that indicated in table I as due to initial causes and is, 
except in the case of the 8,000 lbs. load, greater than that due to bending 
alone. The greatest value observed is 2,900 lbs. per sq. inch or 29 per cent 
of that theoretically due to bending. 

Table 3 corrects these stresses for the inequality due to initial 
causes as observed under a central load, as far as the data are available. 
In the case of the 14,000 and 16,000 lb. load the inequality for a central 
load is assumed the same as for 10,000 and 12,000 lb. load and the measured 
bending stresses may vary considerably from those indicated, but should 
undoubtedly he between the values of Table II and Table III. 

TEST 12— Table I— Plate Ti5-3, Centrally Loaded 



Total 
Load 


Mean Stress 
lbs. per 
sq. in. 


Measured 

Stress 
North Side 


Measured 

Stress 
South Side 


Mean 

Measured 

Stress 


Excess 
North Side 


8000 
10000 
12000 


5520 
6900 
8270 


5220 
7540 
8700 


5800 
6380 
7540 


5510 
6960 
8120 


-290 
+580 
+580 



256 



II ST 12 Table 2 Plate Ti5-3— Top Offset 5/16" 


to South 






Measured 


Measured 




• 


Theo- 


Mea- 


Total 
Load 


Moan 
Stress 


Stress 
North 


Stress 
South 


Mean 

Measured 


Excess 

North 


retical 
Bend- 


sured 

% 


per sq. in. 


Side 


Side 


Stress 


Side 


ing 


Theo- 














Stress 


retical 


8000 


5520 


6960 


4060 


5510 


1450 


8780 


16.5 


10000 


6900 


8700 


4060 


6380 


2320 


9260 


25 


12000 


s_>7D 


11020 


5800 


8410 


2610 


9670 


27 


14000 


9650 


12180 


6380 


9280 


2900 


10100 


29 


16000 


11020 


13340 


9280 


11310 


2030 


10500 


20 



TEST 12— Table 3— Mean Values 



Load 


Total 
Excess 

North Side 


Excess due 

Central 

Load 


Excess due 
to Bending 


Theoretical 
Bending 


Observed 

% 
Theoretical 


8000 


1450 


—290 


1740 


8780 


20 


10000 


2320 


580 


1740 


9260 


19 


12000 


2610 


580 


2030 


9670 


21 


14000 


2900 


(580) 


2320 


10100 


23 


16000 


2030 


(580) 


1450 


10510 


14 



It seems safe therefore, to conclude that up to the loads taken, the 
observed bending stresses do not amount to more than from 20 to 30 per 

T /l T 

cent of those calculated from the formula S= — . This formula 

R aL-2 

depends for its validity on the assumption that the axis of the plate at the 
point of bearing will remain perpendicular to the axis of the pin. The 
writer believes that this will not usually be the case. The bending mo- 
ments considered can be developed only by the eccentricity of the resultant 
of the pressure on the pin. In the case of the plates in question and for a 
5/16" offset up to total loads (T) about 16,000 lbs., the resultant of the pin 
pressure, in order to develop the bending stresses indicated would fall 
outside the middle third of the thickness of the plate, and the plate would 
tend to "tip" on the pin. At a load of 16,000 lbs., the above resultant 
would fall about the middle third. Very careful measurements show that 
at a load of 16,000 lbs., the thickness of the metal back of the pin hole 
increased 0.001", indicating some flow. Moreover, while the inclination of 
the plate to the axis of the pin at the initial load was 0°-20', the inclination 

257 



as shown by a mirror clamped on the centre line of the pin, was 0°-8' at a 
load of 16,000 lbs. Plates 1 and 2, as observed before, clearly indicate 
flow of metal under the conditions pertaining to them. The writer believes 
the tests indicate this as the most probable reason why the theoretical 
bending stresses are not realized, and that they are therefore no more 
likely to prove harmful to the ultimate strength, than the very high stresses 
obtaining within the elastic limit at the edge of the pin holes. 

The fourth Plate Tis-i was given a larger offset, viz., — J 2 inch, and 
the total load was carried to 29,000 lbs. This plate was not straight, and 
under an axial load of 12,000 lbs., the observed stress on the north side at the 
point of measurement (transverse axis of lower pin) was 3,860 lbs. per sq. 
inch greater than the mean observed stress. 

The upper end was then offset }o inch to the South, so that the 
fibre stress from bending on the north side at the point of measurement, 
would be added to bending stress due to initial curvature. The accom- 
panying table (Table 4) summarizes the results. Comparing columns 
2 and 3 it will be seen that the observed mean stress is anywhere up to 
10 r c less than the calculated mean stress. This is no doubt due, as 
previously mentioned, to the difficulty of predicting the point of mean 
stress on a transverse section through the pin hole, and making the cxtenso- 
meter bear at that point, if correctly predicted. Column 4 gives the 
maximum observed stress and column 5 the excess of the stress on the 
north side above the mean. This excess stress is certainly due in part to 
initial curvature, but it is impossible to say just how much is so caused. 
If however it be all attributed to bending due to the X A inch offset, it 
does not in any case exceed about 30^ of the theoretical stress caused 
in that way. The true percentage is certainly less, and probably much less. 
These results therefore seem to confirm strongly those obtained in the 
previous tests. 

TEST 12— Table 4— Plate Tis-4 





Mean 




Max. 


Excess 


Theo- 


Total 




St! 


Mean 


Observed 


Max. 


retical 


Excess % 


Load 


Calcu- 


> tress 


> tress 


Over 


Bending 


Theo- 




lated 


Observed 


North 


Mean 


Stress 


retical 




Per sq. in. 




Side 








16000 


11030 


10150 


1 1500 


4350 


16S40 


25.8 


18000 


12400 


11600 


15660 


4060 


17490 


23.2 


20000 


13800 


13050 


17400 


4350 


18090 


24.0 


22000 


15150 


11210 


19900 


5690 


1S700 


30.4 


241 11 10 


16500 


15950 


21460 


5510 


19230 


28.8 


26000 


17900 


16820 


22600 


5780 


19S40 


29.1 


28000 


19300 


18S50 


24400 


5550 


20370 


27.2 


29000 


20000 


20300 


26100 


5800 


20700 


28.0 



25S 



13. TESTS ON LEAD IX BEARING 

ire deciding upon the final dimensions <>f the lead bearing, used to 
carry the weight of the suspended span .-it Sillery (See page 138 and Plate 
I \ \ \\ 1 1 1 '. a test was made a i McGill Qniveraity to observe the beha- 
viour i)f lead under heavy loads. For the purpose of the test a small 
model, to represent the conditions of loading at Quebec, was made. Ii 
consisted of a '.•'_. \ :; ," plate, l' I' /' Ion- with four plates :i' ," x 

riveted to one fare and SO placed as to leave an opening 3" \ 10" \ : ' , " deep. 
POUT pieces of U" sheet lead 2 -15 16" wide of leli.nl lis Varying from ."/' to 
10 with the shorter piece on top were placed in this opening and co\ered 

with a plate 2 L5 I6"x%"x9 15 16" long. The load was then applied 
in increments, the fop or loose plate being removed after each incremenl to 
observe the behaviour of the lead. 

There was very little How up to a load of 75,000 lbs*, or 5,000 lbs. per 
sq. inch on the 3 x 5 piece of lead. The length of this piece was then 
5.1". At a load of 105,000 lbs. the length was 5.4". Measurements 
were not made beyond this point, the upper piece becoming imbedded into 
the lower pieces and the second layer of lead came into hearing. 

Another teat was made on a single piece of lead '.'>" x .">". At a load 
of 105,000 Ihs. its length had increased to 5.4" after holding the load for 
3 minutes. At a load of 160,000 lbs. its length after :! minutes was (i 1 ," 
thus giving a load of about 8,500 lhs. per sq. inch on the new bearing area. 
In both tests the lead was observed to extrude through the clearance at I he 

sides of the plate, but even under the heaviest loads applied equilibrium 
stablished. 

The lead bearing was used under the suspended span to avoid putting 

a bending stress in the end post as it rotated from its camber position to 
its position under full load. It is obvious that this condition would be 
fully attained if the pressure per square inch over the whole area provided 
for the lead was sufficient to c iuse it to flow, and from the results of the 
experiment it was thought that a pressure of 5,000 lbs. per square inch 
would be sufficient to give this condition. 

The lead was therefore confined to a sp-.. X '■',' 0", which, under 

the total dead load of 1,410,000 lbs. on each bearing 'See Plate CI), 

would give a pressure of 5,050 lbs. per square inch, and was arranged as 
shown on plate LXXXVIII, the sheet metal angles ELGi — B13 being 

provided to prevent it from extruding through the clearance between the 
casting and the bearing plate. 

When the bearing was removed after carrying the load for about six 
weeks it was found that the lead did not completely fill the space provided 
for it, from which it may be concluded that the object, of avoiding the 
bending stresses in the end post, was not completely attained and t hat while 
this bearing served the purpose for which it was designed better results 
would have been obtained if the area of the -p are provided for the lead had 
been reduced about 25 per cent. 

_'.VJ 



14.— TEST ON STEEL ROCKER BEARING 



A full sized test was made at McGill University on a rocker bearing 
representing a short section of the bearing used under each corner of the 
suspended span to carry its weight while being lifted from the scows to 
its final position in the bridge (see Page 138, and Plates LXXXVIII and 
CI.) The specimen was made of ordinary rolled carbon-steel instead of 
rolled nickel steel as used for the bearings themselves. Four test pieces 
cut from the same stock as used for the specimen showed an average 
yield point of 30,500 and an ultimate strength of 63,500 lbs. per sq. inch. 
Figure No. 1 shows the dimensions of the test specimen and the results 
of the tests are summarized in the following Table 1. 

TEST 14.— Table 1 



Load 
Lbs. 


Load 

Lbs. per 

Lineal Inch 


Deformation 
1/1000 inches 


Width of Contact 
Inches 


Convex 
Surface 


Flat 
Surface 


Convex 

Surface 


Flat 
Surface 


125,000 
175,000 
217,000 


50,000 
75,000 
86,800 


4.5 
4.5 
5.2 


3.0 

6.5 

10.3 


1.0 
1.4 
1.7 


1.0 
1.7 
2.2 



All measurements were made after the specimen had been removed 
from the machine and therefore represent the permanent set of the 
material. The width of contact given was clearly defined on the steel by 
reflected light. No measurements were made to obtain the amount of 
deformation or the width of contact under the load. 



260 




261 



APPENDIX "B" 
NOTES ON THE QUEBEC BRIDGE STRESSES 

by 
A. L. HARKNESS, B.A.Sc, A.M.E.I.C. 

LIVE LOAD STRESSES 

Specification: — The live load for which the bridge will be calculated is 
as follows, — 

Train Load. Two Class E60 Engines, followed or preceded, or followed 
and preceded by a train load of 5,000 lbs. per foot per track, on one or 
two tracks. Where empty cars weighing 900 lbs. per lineal foot of track 
in any part of a train produce in any member larger strains than the 
uniform load of 5,000 lbs. such empty cars shall be assumed. 

A Sidewalk Live Load of 500 lbs. per lineal foot for each of two sidewalks; 
to be used for members receiving their maximum stress from a length of 
moving load covering two panels or less. 



The maximum live load stresses in the different members of the bridge 
and the condition of loading giving these stresses are shown on plate 

No. cm. 

The direct stresses, in the Anchor and Cantilever Arm Trusses, are 
entirely determinate as will be at once apparent from the outline diagram 
and by consideration of the fact that the member CM14M16 is built with 
a sliding connection at CMio which makes it impossible for this member 
to carry direct stress. 

Formulae were developed expressing in terms of moments and elements 
of the truss, the stresses in all the members. Owing to the similarity of 
the different panels it was possible to arrange the calculations in tables 
thus greatly reducing the amount of labor involved. The formulae used 
and characteristic cases showing the development of the formulae and the 
method of tabulating the calculations for stresses will be given. 

Notation: The stresses in the members are represented by the capital 
letters T, D, H, P, U, L and Q, their lengths by the small letters t, d, h, p, 
u, 1 and q, while the perpendicular distances from the point of moments 
to the members are represented by the italic letters t, d, u, I and q. 

The loads are represented by the capital letters M, B, F, G, K, N and 
E. The small letter "m" represents the distance covered by the load 
.M. — b, f, g, k and n represent panel lengths and "e." represents the space 

262 



covered by the locomotives. The italic Letters m, b, f, g, k, n and e represent 
the distance from the points of moments to the centers of gravity of the 
different loads. The moments are represented thus-*Mm, B6, Yf, Gg, 
Kk, Nn, and Ee. 

R represents the end reaction from the Suspended Span. 
r " distance from panel point I to the end of the Canti- 

lever arm., 
W " weight of the uniform load = 5000 P.L.F., 

x " variable distance from the point of moments to the 

locomotive, 
a " " length of the Anchor Arm, 

c " " '• " " Cantilever Arm, 

s " " " " " Suspended Span, 

i the Angle between the top chord and the horizontal, 
i " " bottom chord and the horizontal, 

o the length of the end panel Ci o of the Cantilever Arm, 

P+h 

z represents the ratio 

Pi+hi 

y represents the ratio 



(p+h) (f+g) 

For purposes of reference each panel point was given a mark as shown 
on plates XXIX, XXXII and XXXIII. This mark preceded by "A" 
for the Anchor Arm, "C" for the Cantilever Arm and "X" for the Sus- 
pended Span, serves to locate any point in the trusses, as for example: 
AM12, CU4, XLs. A member joiniug any two points is referred to by 
combining the marks for the two points, as for example AM4U6, CM4M5, 
XU2L4. 

Moments: Nn, Kk, Gg, Ff and B6 represent the moments, about the 
panel point indicated, of all the loads to the left of that panel point, which 
produce stress in the member under consideration, e.g., when calculating 
the stress in the member " Di", Fig. No. 1. Nn is the moment about 
panel point V of all the loads between panel points V and I. It is obvious 
that loads to the left of panel point I do not effect the stress in the 
member Di. 

N, K, G, F and B represent the loads and n, k, g, f and b the distance 
to their centers of gravity respectively. 

Mm in Fig. I represents the moments about the panel point I of all 
the loads from that point to the end of the Cantilever Arm including the 
reactions at Ci and XLo from the uniform live load on the Suspended 
Span. As before " M " represents the total load and "m " the distance to 
its center of gravity. 

Mm in Figures 3 and 4 represents the moment about the far end of the 
Suspended Span, of all the loads from that point to the panel point in the 
Cantilever Arm beyond which they no longer effect the stress in the 
member under consideration. 

263 



M-m.\ 




FIG M>. I 



f£, 



'?-« 






1 



jT,4fe___ 



/7£ Afc. /? 



!«___* 



5 



264 




265 



CANTILEVER ARM 

Development of Formulae, — 
Stresses in Sub Members. 

Refer to Figure No. 1. 
Sub Hangers and Posts. 

Let. Rn be the stress in the sub vertical FiiMii 

Gg—Kk—Kg „ „ F/— B&— Bf 
Rii=— -+F— B— 



f 



f fg g 

Sub Diagonals. 

Q = Rn -? 

n 

Stress in Sub Member Mi Fi. 

The variation in the length of the channel span structure, due to change 
of temperature and to the live load on the bridge, was provided for at the 
junction of the cantilever arms and the suspended span, the floorbeam 
" CFi" moving with the Cantilever Arm and " XFi" with the Suspended 
Span. To allow for this expansion and contraction the stringers in the 
panel " C1X1" were fixed at Xi and were supported and free to slide on a 
bracket cantilevered out from the Floorbeam CFi (see plate XXVIII), 
the maximum distance from the center of bearing of the stringer to the 
center of the floorbeam being 2. 15 feet. 

In calculating the live load stresses in the sub members of the panel 
Co-2 this condition was considered but was neglected when calculating the 
stresses in the main members, the panel CiXi then being considered as of 
its normal length of 55.5 feet without regard to the eccentricity at " CFi." 

Hanger MiLi 

Refer to Figure No. 2 

k 
MiLi =R — 
k-v 

R = Go Kfc^f 1 + Nn _ v= 2.15 feet 

k kn n 

1 k+n ^ T k k = 44.15 " 

MiLi = G0. Kfc — +Nn 



MiUs = MiLi 
M2U2 = M1L1 



k— v n(k— v) n(k— v) n= 54.04 " 

Length M1U2 



" U2L2 

Length M1U1 

" U2L2 



266 



Main Truss Members: The development of the formulae for the top 
and bottom chords and for the members MiFi and MlFl will be given as 
an illustration of the. method followed. 

Bottom Chord 

Refer to Fig. No. 4 and to panel I-III of Fig. No. 1 

Nn — Kfc — Kn , . , _ . , . . „ 

U = Rr+ (r— o) + K (r— o— fc) + Qq 

n 

r — o r — o ^ 

= Rr-Kfc (1 + )+Nn + Qq 

n n 



1 f N Nn— Kfc— Kn \ 

R = -{Mm— Kfc— K s+o) (s+o)> 

Si n ) 



=IfMm-Kfc(l- S -±- )-Nn S -±- 
S\ n n 

q pf 

h q 

Rn = B&~ f/-±£ +Gg- 
f fg g 

L/ = M,» I_Kfei(l- S -±-°)-Nn L 8 +2-Kt(l+^)+N« 

s s n s n n n 

+B6 E_ Ff i f ±JL +G Pi 

h h g h g 

= Mm - +Bb f — F/ +G^ y Kfc Nn 

s h h g h g ns ns 

Loads to the left of I do not effect the stress in the member L. Equa- 
tion therefore becomes: — 

t, *, r wP f +g ^ P f rn (n— o) (r+s) ., o (r+s) 

LZ = Mm F/f- — =G? Kfc Nn 

s h g h g ns ns 

Top Chord: 

\ -Kfc Kn 

Uu = Rr+ — (r— o)+K (r— o— fc)— Rn f 

n 

Loads to the left of II do not effect the stress in the member U 

™ r n f m ("— o) (r+s ) XT o fr+s) 

Uu = Mm Ggf Kfc — Nn 

s g ns ns 

267 



Stress in Member M4 F4 (Refer to Figs. Nos. 1 and 3) (consider panel 
4-2 as panel I-III of Fig. No. 1). 

hi 
Hf = Di — +Ri 

d 

d hi 
= Di<2 HRi 

(p+h) (f+g) d 

= Didy+Ri 

N n XA Kn 

Did = Rr+ (r— o) +K (v—o—k) +B6— Lil+Qq 

n 

= Rr— Lil+Qq+Bb— Kk (1+— )+Nn — 

n n 

LiZ = Lih = Lihz 

pi+hi 

= Rz (k+o)+Rizk 

] r — o r — o 
Did=R { r— z (k+o) )— Rizk+Q?+B6— Kfc(l+ ) +Nn 

I J n n 

s s n s n 

Ri=Gg- Kfc— — +N« — 

k kn n 

n r/P ™ p f+g j_n„ p f 
Qq = Bb- F/- HG^— 

h h g h g 

Dirf = Mm-(r-z (k+o))+B& (!+-£-)— F/ f- — +G ? j ^ — z) 



s 



J h h g I h g 



_KJfc 1 1 (l-i±-°) (r-z (k+ o)j-z^t n + (1+— J 
I s n ( J n n 

— Nn <r— z(k+o) +z 

[ s n \\. J n n J 

Hr = M4F4 = Mm-|r-z(k+o)| — Bb |-g=- — y (!+-£) 1 

+Kfc^^- ) /z (s+o+k)-(s+r)l 
ns I. . 1 

+Nn - (z (s+o+k)— (s+r)} 
ns \ J 

268 



Stress is M1F1: Refer to Fig. No. 1. 

hi 
Hf = Di — +Ri 
d 

n , ' ,n J.U a (h+p) (f+g) 
= Did — — +Ri d= 

da d 

-Drfy+R, >-y (h+p !" (r+g) 

Did=Mm— Lii+Qg 

p+h 

LiJ = Lihz z= — 

pi +hi 

LiZi=Mm— M (f+g)+G?— Bb— B (f+g)+Qigi 

Qioi = Riy — k 
hi 

R iv = G<7 t — K fc -TT + Nn - 
k kn n 

pi „ pi k+n " pi k 

Qiqi = Gg f — Kfc l - -^- + Nn£- - 

hi hi n hi n 

LiZi=Mm— (M+B) (f+g)— B6+G? 1+ ^ ] — Kfc^- 1 — 

I hi J hi n 

», Pi k 

+Nti 5" - 

hi n 

Lrf=Mmz— <M+B) z (f+g)— Bfcz+G^z 1+ ^ 

pi k+n pi k 

— k/.z — + yn z — — 

hi n hi n 

^ „, P ^ P f +g , n P f 

Qg = B6- — F/- hGs- — 

h h g h g 

b+f 1 

Ri=-B6 -rr +F/- 

bf f 

Hf = v Mm-Mmz + (M+Bj z (f+ g )+B6z+ G0z'l+ ^ ! 
I hij 

pi k+n pi k p p f+g f 1 

+RAz • Nnz h B6 - — F/ - hGg-| 

hi n hi n h h fg gj 

H. =Mmy (i-z) + (M+B) yz (f+g)— B&j^ — y (z+ jjj 

/l p f+gl / _ pi P f 

+FK— — y- — f — G^y 'z (1+ -)— - — 

If h g J v hi h g 

.'•in 



pi k+n „ pi k 

+KA; y z — Nn yz 

hi kn hi n 

General Formulae: 

Formulae for all the members in the truss were developed as 

illustrated above and are as follows: — Refer to Fig. 1. 

f k+n k 

T( = Mm(i- z) + Mz (f+g)— Gg Kfcz + Nnz — 

g n n 

t 

D2d = Mm (i-z)+Mz (f+g)— Gg z (1+ ^)+-} +KAz ^ — 

I hi gj hi n 

Pi k 

— Nn z 

hi n 

p f+g pi p f 

Did = Mm (i-z) +Mz (f +g)— F/ ~ — — Gg z (1+ f-) —~ - 

h g I hi h g 

, pi k+n „ pi k 

+Kfc z,-— — Nn zf- - 

hi n hi n 

^ hl 

Hl = Di — 

d 

HF = Mmy (i-z) + (M+B) y z (f+g)— B6 l^-~ — y (z + ^) 

+F/(4-y^)- G3 yW£ , )-^| 

f h g I hi h gj 

pi k+n pi k 

+Kfc y z , — Nn y z , 

hi n hi n 

Rn = Bb 4 —W-T^ + Gg - 

f fg g 

Q = Rn v 
h 

Refer to Figure No. 1 and to panel I-III of Figure No. 1. 

Uu = Mm Gg KA; Nn 

s g ns ns 

t, ,* r tm P f +S , P f T , 7 (n— o) (r+s) o (r+s) 

LZ = Mm F/ |-G^ Kfc — Nn 

s h g h g ns ns 

Refer to Figure No. 3 : 

UoLo = Mm Nn — 

s ns 

TT TT TT T length U0U2 

UoUl = U0L0 x — — — 

" U2L2 

TT _, TT T length U0L2 

UoMi = UoLo x — 2. _ 

L2L2 

270 



MlLj = Ml-; 

T T AT Hu AT k+ ° T^ J D -°) (8 + 0+k) O (S + O + k) 

L2L4 = M2— > M2 = Mm K/c — Nn 

t I s ns ns 

1 



U2TJ4 = M: 
u 

s gk ^ k ns J 

. T yo (s+o+k) 
ns 

M 2 U4=T;andTt=Mm— r— z(k+o)\— Gg( — +z)+K ""^ Izfe+o+r) 
si J g )b ns I 

— (s+r)\ + Nn — (z (s+o+k)— (s+r) ) 

tt at at tt Lengt 11 U2M2 
U4M4 = M2TJ4 



M2M3 = M2l"l 



" M2U4 

Length M2L4 

~~ " M2U4 



M 3 L4 = Di; and Did = M m -jr— z (k+o) \—Ff~- — — Gg (z— — ■) 
m a{ J h g h g 

+KA; — z ( s+0 +k)-(s+r)> 
ns I ) 

• +Nn— z (s+o+k)— (s+r) 
nsl 

Length M2L2 

F4L4 = M3L4 — TTT~ 

" M2L4 

M4F4 = Mm — (r— z (k+o)l— B6 {-£— y (1+— ) 
s^ J { bf h 

y (n-o) f 1 

+Kfc ^ - ; { z (s+o+k)— (s+r) k 

ns 



v z (,s-|-o-t-K; — (.s-|-r; r 



yk f 

+ Nn — I z (s+o+k)— (s+r) 
ns I 



271 



- 
< 

- 

S 

o 

- 

-- 

K 
H 

O 

- 

H 

Z 

- 
— 

I— I 

- 
- 
- 
o 
u 

I 



£ 

<« 




•OOiOO^iN-i 
• CO CO t^ 00 © <N m 


m i-i o> -* co i-i o 

OtOBNNNS 


io 05 io r- oo i-i co 

CN f~ 00 CD CO CO O 


O 

►J 


• i-i <M t*- 00 >0 i-i t- 

•00OCNiO05COCO 
. _ ,_, rt f, CN CN 


i-i -h co f~ co oc >o 

CO O CN "O C5 CN CD 

— 1 ■-! rt i-l CN CN 


CO O ~* CO CO C CO 

t>- OS i-i co «* >o io 


fa 
O 

fa 

o 
o 

1 


a 




• r~ "O oo ■* oo cn co 

• O b- "O CN CO <-i "O 

■ iO 00 CN t~ CN 00 CO 

• -* r)< io "O co co t» 


r~- IO 00 Tt< OO CN CO 
CD MCI IN CO -i iO 
iO00CNt^.,CN00CO 
"rjl TtllQ io CO CO t~ 


00 O CN CO 00 iO Tf< 

Tfi io CO »o »0 CO »o 
co i-< r- co oo io co 

O Tj< CO "O i-H l-l i-H 






-H __,,-, ^ rt 






• -* O O 00 CO CO rt< 
' O OS CO 00 io cn © 

• ■* oo t)< o oo co co 

• CO CO t- 00 00 05 O 


— iOOOOCOCB'* 
ifJfflCCX >Q CN 3S 
■* 00 Tf O CO CO CO 
CD CD In. 00 00 05 © 


00 IO iO 00 CO O 00 
05 CO 0> CO i-H CO o 

rHtflrHCONHS 
i-< CO CO t» CO CO CN 


: r ~' 


rt 


CN CN CN CN CN IN 


Co 

o 


19.452 
44.444 
35.714 

31.250 
23.810 


• • CO CN O O © O 

• ■MNOOOO 

• -cooiooo© 

• -hhDIOOO 


■ 00 t-- "O CN ■* 00 
■NNOhOCS 
• 05 CO 05 I~~ I- CO 

•05 05000 00 00 


hMOOOOO 

t» 1> o o o p o 

000 lOOOOO 

CO— iCNOOOO 


. . rt 1-C I-l r-l FH rH 


— 




fe 


43 . 262 
83.659 
67.972 
56.250 
47.620 




■lOiOO^OOCO 
•00~COCN005 
■ CO ^ CD rn Tf CO 
■a»C»NNN 






I'-l'-l'-"-!'-"-' 




« 


23.810 
39.215 
32.258 
25 . 000 
23.810 




































































+ 

o 








•OOMNiOOO 

■ co io i-i co i-i co 






















• CO 05 CO 05 1^. IO 

• t» 00 O 05 05 C35 
























S 

3 


:::::::: 


ON 

i i- CO t^ CN CO 00 o 

S^cooo-hcncoio 

►_) — i— <CN WiflcON 


In. 

iO CD N CN CO 00 O 
-h CD 00 —i CN CO "O 
OSIOOOi^iiN 
m ■-« CN CO iO CO t>- 


IO iO O CN 00 "O ^ 
CN 00 CO 00 IO CO IO 

OS -^ ■* CN 00 iO CO 
-* CN CN CN -< i-i — i 

■1 1 1 1 1 1 


.•a ta 


i 

©: : : : ; : : 








------- 


s 1 

■go 

fa 


m : : : : : : : 


coco^: 


■*. . . . 
"=3 


com: : : : : 


0) 


MlLl 

M3F3 

M5FS 

M7F7 

M9F9 

MllFll 

M13F13 

M15F15 


UoU2 
2-4 
4-6 
6-8 
8-10 
10-12 
12-14 
14-16 


L2L4 
4-6 
6-8 
8-10 
10-12 
12-14 
14-16 


M2U4 

4-6 

6-8 

8-10 

10-12 

12-14 

14-16 




eavoixaa^ 
•sag 


dHOHQ 

1 


aaoHQ 
woxiog 


S1VNOOVIQ 
NOISN3X 



272 



z 

S3 
- 
- 



.-. .-. — _ — 



so o co c. 



re x — © — r» 

•: -i — r r r 



ENe-rNMrN-i 



re re re x re re 



r- — re - - 3 — S 
tD<HCOC — r- — — 
. ,—-- — -■- z 

-. re re — 



t- /-■-/• re ~ t~ EC 

S CC - ' - 



r ?i n m 

10 OB COOS'* COW ''• 

— — X X — EC — r~ 
vr; — re r i — r S 3 

-i -j -i m n — 



r — Z x r i- — 
///—.-> 

■- r>- re ei — —. — 



r* -* 

— - 2 •- — to — --'■ 

- - ^ - - r. i^ »* 

i- x - - - r. 



a i- © ~> x •-. — 
a ,7 oo co at ■ c '- 

— -i r ■-. -. 

;C»CINCHn«rt 



i- — 3 — {ON 

/ - ^ m ; r. - 
r — - — r co >o 

X BOINNNN 



/ — X ' - - ' 5 ' "- 



OfOC 



ifl 3 -r X © re — 
— CO -■ re re e>l ~* 

aCNSNNNf" 



re — x r. i- •" 



r- : 

■ - -i 

— - 

23 



1 x '-. ". 



X ■ X © - * o 



? i 00 cs 






re© O© t-S 



re X re re re 
<e to io *e us 



.. — j j. — — 

C — — re © — 

M M H M M M 

>* c:i- 

— EC X 

re III 

~ > e r > © — re 



— re 'e e> -r © m -r — 

SS2 ii i ~£ i ii', I 
.-_ ^ ^ 



BTVNOOVIQ K0ISS3HJK03 



;:i- — 



©re _»ei x ei — 
re - .- ei re - — 
10 EC t- »ONM 
re 10 i.e .e EC EC 0" 

M M M M M M M 

- - 2 -i- 
- ^ x 

— ei ■ ' III 

: / : :i 



S3' 



io >o © ei — ■* x x 
© '.e — — i- © ec x 



MHMNMKNK 

, _ ~ — re ie 

fe.fc.iflr>.©. 

~ce i I I I i I 
ie i- ~ — re io 



- Z f -x. — — - x 



wiaoxvfj 






«2 c , =r.-c 

_ 2 

— ~ '" '" 4. ' 4. J> 



gaao.>rvH kiboj 



BTVNOOVIQ 

aag 






./*- 






F/S. N* 5 




777A 



U * 



/V^v 




({((<f(SS/f/f(\ 






l///////. 



j;/s>;sfjs;ss/;ssjJ>J 



V e •!•--- 



__x 



f/S. A/*8 



*M»^ 



w»?> >>j»js;;;s/j,;jss/s;;sjjss;s. 



m 



t 



f 



U,-. 



- X -- 

F/6A 



274 



<A/7n 




/ 7/\ m 
k- M. _*-/?_ -J, 



fig. A/o. /2 






-.^ V 



„-/7__u. /I .. O/ -I 

J, O 




275 



S^ «- vi 

^i 2T *f> 
< •*) Oft 



& 2r $ 

fc- H) Oft 



§ <>- $ I 
* "> £ 3; 



&* ^ « 

N. ^5) \0 



S * * 
N» \© sj) 



* * R 

^ ^ ?i 



US 




25 £ Si 
> *o sr 



^ ^ >?5 



S °- St 
*> n s 



•o ^ R 

N» v© r^ 



03 ^> 5 

^ >o ^ 



>o ^ S 

^ NO iQ 



<* 5 $ 



* * 8 



5 cS 



The Moments may be expressed as follows: — 

1. For Figs, 3 and l. 
Refer to Figure No. 5, 

Let 'Mm" !><• the moment of all the loads between the points indicated, 

Let "E" be the weight o the locomotives, 

Let "e" be the distance to the c. of g. of the locomotives, 

Let "w" be the uniform Load per lineal foot of track. 

tasume locomotives headed towards the main pier, 

Mm = w^- + (Ee- w *- ) + (E— we) x+w ('— --) 32.5 

m 2 

= \v |-282x + 18470. 

2 

dMfii 

=282. 

dx 

2. For Fig. No. 1, 
Refer to Figure No. 6, 

R = Reaction at the end of the Cantilever Arm from the uniform load 

on the Suspended Span 
Ri = Reaction at CFi from the Uniform load in panel CiXi. 

w el , 

Mm=- (m2 — g2)— we m— x — +E (m — x — eJ+Rim+R (m+o) 

= - (m2— g2) + (E— we)(m— x)— (Ee— w -^-) + (R+Ri) m + Ro 

= - (m 2 — g 2 )+282 (m— x) + 1715m + 19652. 

dMm ^,„ 

— — =—282 
dx 

3. For Figure No. 1 

When g = o. 

Mm=- m 2+282 (m— x) + 1715 m + 19652. 

dMm 

=—282 

dx 

4. For Figure Xo. 1 
Refer to Figure Xo. 7, 

Let "Ei" be the weight of the locomotives to the left of panel point I. 

Let "to" be the distance to the c. of g. of Ei. 

\v 
Mm = — x2+wx (m— xi+E m— x— e)+Eiei+R (m+o)+Rim 

= wx (m ^)+E (m-x)— Ee4-Eiei + (R4-Ri) m+Ro 

= wx (m— -) 4-852 (m-x)+Eiei4-1715m— 12838 
277 



dMm 

— — =w (m— x — 852+Ei 
dx 

For Figure No. 1 

Refer to Fig. No. 8, 

m 2 e 2 

Mra = w |-Ee+Ex— w — +w e > 

2 2 

m 2 e 2 

= w — +Ee + (E — w e) x — w - 



= 


m 2 

w — +282 x- 


f 16602 


dMm 

dx 


= 282 




For Figure No. 1 




Refer to Figure No. 


9, 


Mm = 


m 2 x 2 

w w — 

2 2 


+Eiei 


dMm 


= Ei — w x 





dx 
7. Moment table for locomotives, page 276. 

A moment table of the wheel concentrations of the locomotives was 
used from which the load Ei and the moment Eiei of Figures 7 and 9 
could be easily obtained. 

Maximum Stress in any member. The position of the locomotive giving 
the maximum stress in any member was found by means of an equation 
obtained by differentiating the equation for the stress in that member. The 
first differential is equal to zero and contains only expressions representing 
the loads and the coefficients for the moments, thus giving an equation 
which was very easily solved. 

As an example refer to the calculations for the stress in the member 
M4F4, which follows, and to figures 5, 8 and 9. 

The first differential of the equation for the stress in this member is : — 

.4357 X 282—58.333 (Eb — wx) +22.688 (Ep-wx) 
—2.3090 (Eg— wx) + 1,0627 (Ek—wx) +.7484+282 = 0. 

where Eb, Ef, Eg, and Ek represent the weight of the locomotives to 
the left of panel points 4, 3, 2 and 1 respectively, x the distance from the 
point of moments to the uniform load preceding the locomotives and 
w = 5000 lbs. per lineal foot of track. 

To obtain the condition for the maximum stress in the member it is 
then only necessary to find a position of the locomotives such that the sum 

278 



cf the positive factors of the equation falls between the sums of the negative 
factors, obtained by first considering Eb as the weight of the locomotive 
to the left of the panel point 4 and second by considering that the weight 
Eb includes the weight of the wheel concentrated at that point. 

Stresses, — The full calculations for the stress in the members UoLo 
MjI/j, MiFi and the Hangers Hf (see Fig. 1) follow and will serve to illus- 
trate the method followed for calculating the stresses in all the members 
of the truss. 

The clause in the specification stating that the locomotives 
were to be followed or preceded or followed and preceded by the uniform 
load made it necessary to figure all the stresses twice, first with the loco- 
motives headed towards the main pier, as herein, and second with the 
locomotives headed towards the center of the span, the maximum of 
these two conditions being taken as the max'mum live load stress in the 
member. The difference in the stresses thus obtained was found to be 
very slight and would hardly justify the additional amount of labor 
involved. 

Stress in UoLo 

(See Figure No. 3.) 
UoLo = 1.5625 Mm -18.666 Nn 

sMm. 



37fe—* 

yzA 



VZZZZZZZZZZZL 






k. ss.'s 



x = sirs 



»c 



m -- 663'- 



C'hiterion: 

(See Figures Nos. 5 and 9.) 
1.5625X282= 441. 90 150 

90 90 

( 0, 60) X 19.06 = 0, 1140. 
Stress 

ra 2 
w— =1098920 
2 

282x= 144240 

18470 

Mm =1261630X1.5625 = 1971.2 

690 

6890 

Nn 7580 X 18.666 = 141.5 

1829.7 = UoLo 
279 



Stress in M2F2 

(Refer to Figure No. 3.) 
M 2 F 2 = 1.0258 Mm-58.155 G^ + 17.295 Kfc -4.611 Nn 



4.5 



■*M- 



sA/n- 



$t 






SM 



7*. 



L5 



SZ2. 



X* 609' 



3 % 



-2x:s^*> 



-42'- I 



-35. S- 



m = 7Z7S- 



Criterion: 








See Figures Nos. 5, 8 and 9. 






282X1.0258= 289 90 150 






426 


90 90 






300 


(0 60) X 58.155 = 


■ 0000 


3480 


126X17.295 = 2180 282 X 4.611 = 


1300 


1300 




2469 


1300 


4780 


m 2 
w — = 

2 


1323140 






282x = 


171740 
18470 






Mm 


1513350 X 1.0258 = 1552.2 
10488 
2982 
1400 






Kfc 


14870X 17.295 = 257.2 1809.4 
690 
456 






Gg 


1146X58.155= 66.6 






m2 
w — 

2 


36000 






282+1.5 423 








16602 






Nn 


53025X4.611 = 244.5 311.1 







1498.3 = M2F2 
280 



SlKKSS I\ \ III"! 

Si Figure No. 3.) 
MiFj = .t:r>7 Mw-58.333 B6+22.688 Fy-2.3090 Gg+1.0627Kft 



+ .74*4 X/i 








Criterion: 








282 X. 4357 =122 








387 




150 210 




242.5 




115 115 




144.5 X22.68S = 3280 




(35 95)X58.333 = 2040 


5550 


362 




516 




565 




355 




287X1.0627 =305 




161X2.309 = 372 


372 


282X.7484 =211 = 


o'.US 


2412 


5922 



£' 



i 



<fc 



-UE 






: & 



^Mttv. 



S. A 



4i 



5-4-S 
VZZZBZZZBZ 



X • 66Z 

' .V ?>>SS 777> 



^ 



-Z6- 



-255- 



•s'.s. 



-/7?r 781'— 



B6 



F/ 



Gg 



Kk 



Mm- 1730054 X .4357 = 753.8 

10022X22.688 = 227.4 

61127X1.0827= 54.3 

107227 X .7484= 80.3 

2077X58.333 = 12 L2 

20413X2.3090= 47.1 

281 



1115.8 



168.3 



Nn 



1440 




8553 


17544 


44832 


75255 


637 




194 


1032 


3408 


15370 


2077 




1275 
10022 


1837 
20413 


2887 
51127 


16602 




107227 


m 2 

w — 

2 


= 1524900 








282x 


= 186684 












18470 









947.5 = M4F4 



to 

X 

CO 
CO 

w 
oi 
H 

CO 

Q 
•< 
O 

i-J 

s 

< 

I 

6 

Z 

- 
- 
pa 




282 



00 

© 



-r Q 

<N l-H 



CM COO lOOCMO 



i- 1 t"» CM lOtDiHrt 

C5 CO ^H ! C3 l-H Tji OS 



5T 00 -r © n -i 

V * I OS oo cc ■ -r 35 

ol ™i ©co co i 




OOOO Ox ri 55 

CO CO »0 -<J< COCO 'f o 

OOiOO— J. / — X 

-HNMH CO CM i-i 

-H t^ CM ~H H O 

CM l.- / /_ 



CO CMi-i O 



3 



© »c 

CO lO 

M< ,-< 

CM >-H 



icoioo ir^t^oo-* 



001K3 NO«M 
i-i CM I - CO CO © CM 
<N CO O O i-4 CO CO 
CM i-t IfH 




Q O O O © 0E r i O 

© CO © Tfl t^ CO CO t~- 

oomii r^ co © os 

i^ r^ r^ i-i ro r j o © 

CO O >H o o 



CJSi-iCOO 

r. OS os — 



CO OS 

© "tf 
»0 I CM 

CM i-i 



lOtOOiON^N 

CO ■* Tfi I OS lO CO CI 
Tj< CM t^ i-i CO CO CM 

Tt< CO O CM i-i CO CO 





^ 



+ 



+ 



+ 



OOOO 


OCO 


ri 


CM CO ©*# 


OS CO 


■ t 


CO IC CM rj< 


t- oo 


OS 


ODN-tH 


X CM 


IC 


CO t^ i-i 


CM ^H 


— 


lO CO 


■* 


-r 




H 


iH 



CO i-i CMCM O 

t^ © CO CO ^h 
** COCM OS 



283 



ANCHOR ARM 

The live load stresses in the members of the Anchor Arm were cal- 
culated for the following conditions of loading. 

Case I The Locomotives and the uniform load of 5000 P.L.F. on 
the channel span with the locomotives placed to give the maximum uplift 
on the anchorage. No load on the Anchor Arm. 

Case II A uniform load of 5000 P.L.F. on the channel span. No load 
on the Anchor Arm. 

Case III A uniform load of 5000 P.L.F. from the outer end of the 
Anchor Arm to a point giving the maximum stress in the member under 
consideration and a uniform load of 900 P.L.F. (representing unloaded 
cars) on the balance of the Anchor Arm. No load on the channel span. 

Case IV The locomotives on the Anchor Arm in the position giving 
the maximum stress in the member under consideration, with a uniform 
load of 5000 P.L.F. extending from the outer end of the Anchor Arm to the 
locomotives and a uniform load of 900 P.L.F. from the locomotives to the 
main pier. 

Case V The live load on one track only. The tracks are spaced 
32' 6" c. to c. while the trusses are spaced 88 ft. c. to c. hence under a live 
load on one track only the one truss receives 68.47% of the load and the 
other truss the balance, i.e., 31.53%. Under this difference of loading the 
bottom chord of the one truss shortens more than the chord of the other 
truss and this tends to throw the outer end of the bottom chord lateral 
system off the center line. This tendency is restrained by the wind an- 
chorage pin (see plate XXVIII) and thus a shear in the bottom lateral 
system is produced resulting in stresses in the bottom laterals and a decrease 
in the stresses of the more heavily loaded chord and an increase in the other 
chord. This condition of loading gives the only live load stress in the 
bottom laterals but in no case gives a maximum stress in the chords. 

The maximum stress in the truss members is produced by either one 
of the conditions of loading described above or by a combination of them . 
Case I gives the maximum stress in the chords, the Anchorage eyebars and 
some of the members towards the outer end of the Anchor Arm. Com- 
bining cases 1 and III, gives the maximum stress in the Diagonals M3L4, 
M4M5 and M5M6, and in the hangers F4L4 and F6L6. Combining cases 
II and IV gives the maximum stress in the remainder of the web members. 

Formula?, for calculating the stresses, similar to those used for the 
Cantilever Arm were developed. The method used was, with but minor 
changes, the same and reference will therefore be made only to the points 
in which they differ. 

284 



Formulae: Refer to Figure N'n. 1 and consider it as a part of the 
Anchor Ann. Thr notation used for the Anchor Ann is the Bame with 
I hi exception of the momenta. Mm is used to represent the moment about 
the panel point Ao. of all the loads on the Anchor Ann. Nn, KJfc, G<7, 
Yf and Bb are used to represent the moments about the panel points 
indicated, of all the loads from that panel point to the vertical post over 
the mam pier, "r" is used to represent the distance from the panel point 
"Ao" to the panel point I of Figure I and "a" to represent the length of the 
Anchor Arm = 515 feet. 

As an illustration the formula for the member D. (Figure I) will be 
gi ven : — 

Dirf = M/H-Wz (a-r+f +g)— (a-r)\— Xn £i z — +Kk P z k — 
a { hi n hi n 

_G„ f zQ+ Ei )-P-4-F/ -f I+ g + B6<1 + JL ) 
hi h gj h g h 

Similar formulae were developed for all the members and were used for 
calculating the stresses for both Case III and Case IV. 

Calculations: The calculations for the stress in the member M 9 L 10 
for both Cases III and IV follow and will serve as an illustration of the 
method used throughout the Anchor Arm. 

Substituting the value of the coefficients the formula for this member 
becomes : — 

M9Lio = 0.2737Mm— 1.4480X«+2.4S20KA— 1.4006G^— 1.7188F/ 
+ 1.8594B6 

Case III Let w represent the uniform load of 5000 P.L.F. 
" wi " ' " 900 " 

" x " " distance from the panel point Ao to the 

end of the 5000 lb. load. 

The moments may then be expressed as follows, — 

wi w — wi w w— wi 

Mm= -r-a 2 + -r— x2+ — (50.5— 23) 23 = — — X 2 + 120930 

AAA It 

il.Mm 

= (w — wi x 

dx 

Wl . . , w-wi 

Nn = — (a— r+f+g+k+n) 2 + — - x— r+f+g+k+n)2 

dNn , 

— - =( w — wi) (x— r+f+g+k+n) 
dx 

K*«- y (a-r+f +g+k)*+ "'(x-r+f+R+k^ 

dKk 

— — = ( W _ wi) (x— r+f+g+k) 
dx 

285 



Wl W - Wl 

Gg = — (a-r+f+g)2 + — (x— r+f+g)« 



F/ 



dG£ 
dx : 
wi 



(w— wi) (x— r+f+g) 



W — Wl 

2 (a-r+f) 2 + — (x-r+f)2 



diy 

dx 



= (w — wi) (x — r+f) 



B6 



wi . . w — 1 

== — (a— r)H 

2 2 

dB6 



(x-r)2 



dx 



= (w — wi) (x — r) 



Maximum Stress: The condition of loading producing the maximum 
stress in the member may be found by placing the first differential of the 
equation equal to zero and solving for .x. 

Since the concentrations at all panel points AFi to AF9 inclusive, pro- 
duce compression in the member M9L10 and all concentrations AF10 to 
AF12 inclusive produce tension, it is obvious that the uniform load "w" 
ends somewhere between AF9 and AF10. 

The equation for the maximum stress in the member therefore is — 
x (0.2737— 1.4480+2,4820— 1.4006— 1.7188) +191 x 1.4480 
—221 x 2.4820+263 x 1.4006+305 x 1.7188 = 
or x = 342.7. 

Having obtained the value of "x" the stress in the member may then 
be obtained as follows : 



Mm: 



Nw 



Kfc 



= 2.05x342.7 = 240760 

1 20930 = 361690 x 0.2737 = 
.45 x 3242 = 47240 
2.05 x 151.72= 47180 = 94420 x 1.4480 = 



99000 



= .45x2942 = 38900 
2.05 x 121.72= 30360 = 



69260x2.4820= 171900 



Gg = .45 x 2522 = 28575 

2.05 x 79.72= 13025 = 41600 x 1.4006 = 
= .45x2102 = 19845 
2.05 x 37.72= 2915 = 22760x1.7188 = 



F/ 
B6 



136720 

58270 
39120 



.45 x 1682 = 12700 x 1.8594 = 23615 



294515 

234110 

Did = 60405 

d = 129.16 

M9L10 = 467. 7 

286 



234110 



Combining Case I and Case III the total stress in the member for this 
condition of loading becomes: 

Case I =711.0 
Case 111=467.7 



1178.7 
Case IV. The moments may be expressed as follows: — 

1 . 1 !efer to Figure 5 and 

let m = a = 51o feel 
" Xo represent Ao of the Anchor Arm 

XiCi be the stringer panel between the Anchor Arm and the 
Approach Span =50.5 feet, 

lot x be the length of the uniform load of w (=5000 P.L.F.) on the 
Anchor Ann. 

Then (m— x — e) = (a — x — e) will be the length of the uniform load wi 
(=900 P.L.F.) 

Then Mm =wi- + (w-wi) — wie (x+- ) +Ee+Ex + — (50.5-23)23 

m £ A 2i 

X 2 

= (w-wi) 749.4 x +164170 

dMm 

— — =4.1 x +749.4 
dx 

2. Refer to Figure 8 and 

let m = the distance from the point of moments to the post over the 
main pier. 

let x = the length of the uniform load w 
Then m— x— e " " " " " w, 

m 2 • x 2 e 

M//i=wi— +(w— wi)-- wie (x+— )+Ee+Ex 

, X 2 m 2 

= (w— wi) - +wi — +749.4 x +43244 

dMm 

— — =4.1x+749.4 
dx 

3. Refer to Figure 9 and 

let m = the distance from the point of moments to the post over the 

main pier, 
let x= the distance from the point of moments to the uniform load wi. 

wi 

Mm = - — (m 2 — x 2 ) +Eiei. 

dMm 

— — = Ei — wix. 
dx 



287 



Calculations for the member M9L10. 



^Nn-A 



X'Zzs' 



- ".": "■»■■■;■'■■ '■■.■•■:: "i 



m 



^Gfi 






50-01. 



■7 



-AZ'Ql- 



-42-01 



.AZ'-Ql 



Mm 



Nrc 



Kfc 



1663.7 880.6 757.6 

1663.7 x 2737= 455 

757.6x24820 = 1880 

2335 



Gg F/ 


B6 


509.4 




449.4 241.2 





880.6 x 1.4480 = 1270 


1270 


449.4 x 1.4006= 630 




509.4 x 1,0006 = 


715 


241.2 x 1,7188 = 415 


415 



2315 2400 



Moments : 










Mm 


N» 
2100 


Kk 
16 


Gg 


W 


101940 


47230 


38900 




2490 


167120 


23980 


1500 


20124 


1080 


164170 


43240 


43244 


26112 


19384 



433230 



116550 



83660 



46236 



22954 



Bb 



12700 



433230 x .2737 = 118570 
83660x2.4820 = 207650 
12700x1.8594= 23620 = 349840 



116550 x 1.4480=168780 
46236 x 1.4006= 64760 
22954 x 1.7188= 39450 = 272990 



76850 = D^ 
129.16 = d 

595 =M9Lu)CaseIV 
653 =MgLio " II 
1248 = Maximum live load stress 
in the member. 



288 



SUSPENDED SPAN 

The formulae used for calculating the stresses in the truss members 
of the Suspended Span and the value of the coefficients for the different 
members will be given. 

Figs 10, 11. 12, & 13. 

Refer to Kip. No. 10 note, g = k = n 

{ 1 s+o\lengthLoU2 

LoM.= Mm 7 -F/— ; „ ^ 

1 s ' fs / '■ U2L2 

MTT J' 1 ^ iVh> f+g] _ 1 1 l ength L0U2 

Mil 2 = ' M/h F/ 1 — Go — } 

I s y I fs 2fg J y 2gj " U2L2 

T , T r , f — g ~ P+k 1 

L 2L2 = F/ — — - — Go — — + N n — 

2fg 2gk T 2n 

Refer to Fig. No. 1 1 note, k = n 

T)id = M/n — — Xn 

s 2n 

D2i = Mm — hKft I ' f 2! =Nn|- - + 



I n J 



U -J 

V = Mm — ; — Xn — 

s (bi+r) 2n 

r + k + n 
U = llw — N n 



Lh = Mm — — 2Kfc + Nfl 

s 

Reversals, — Refer to Fig. Xo. 12. 

Dld = Mw L±^_ Xn j\±j: + 1 ] 

s I 2n J 

D2d=M w ^t^_K fc f b -+L +2 ] 

s n ) 



,. s + bi 1 bi+i 1 

= Mot . — Xn — I r— !— + li 

s (b/ +r) bi + r[ 2n J 

289 



CENTRE PANEL 

The rivets connecting the members M9L10 and M9L10 to the gusset 
plates at Ls and L10 were not driven until after the truss had been lowered 
from its camber position and was supported entirely on the staging bents 
under the four corners, Lo. Hence these members carry no dead load 
stresses; and the dead load stresses in the other members of this panel 
are determinate. With the members all connected into the truss system 
this panel has a redundant member and the live load stresses were there- 
fore calculated according to the elastic deformation of the members. 
The equations for the different members are given in the table of co- 
efficients. 



290 



TABLE No. 3 







S 


2 6 


COEFFICIENTS 
Unit = IO- 3 




Mm 


F/ 


Gg . 


Kfc 


Nn 


Bottom 

( hord 


L0L2 
2-4 
4-6 
6-8 
8-10 


10 
11 
11 
11 
13 


1.451 
1.451 
2 245 

2 991 
+4.452 
2.354 
3.048 
3.957 
+4.638 


17.699 










28.572 

22.100 

19.140 

—12.297 


14.286 

11.050 

9.570 

+ 1.696 












+ 1.696 


C <- 

o o 

^ U 


I'-jI'i 
4-6 

6-8 
8-10 


11 
11 
11 
13 








11.586 

9.759 

9.112 

—1.687 
















—1.687 


—5.905 


Main 

Diagonals 


LoMi 

M1U2 

M3L4 

1 -|M;. 

Mi La 

1V.M, 
M?L8 
1\M, 
M9L8 


10 

10 
11 
11 
11 
11 
11 
11 
13 
13 


2.132 
2.132 
1.166 
1.166 
1.220 
1.220 
1.625 
1.625 
—1 . 126 
+0.806 


26.011 
7.281 








20.986 








16.240 
37.232 
15.656 
33.733 
14.966 
30.721 
+2.868 
+2.861 






41.988 










36.155 










31.508 
+ 10.040 
—20.862 










Main 
Vert- 

ioala 


U4L4 
UeLe 
UsLs 


11 

11 
13 


0.855 

0.964 

+0.640 








14.286 

12.500 

—2.340 












—8.100 


< 

w 
> 


■r. 

T. 



o 
Q 


Ij.M 

M ::!.! 

1 ;M. 
M5L6 
I". M, 
M7L8 
I'sM'.. 
M.9L.8 


12 
12 
12 
12 
12 
12 
13 
13 


5.922 
5.922 
3.642 
3.642 
2.412 
2.412 
+0.806 
—1 . 126 








20.994 

18.078 

15.754 

—12.589 
+ 18.318 






41.988 










36.155 










31.508 












•- — 
^ .2 


U4L4 
UeLe 
UsLa 


12 
12 

13 
10 
10 
10 
11 
11 
12 


4.338 

2.886 

—1.002 








17.771 

14.414 

+ 10.802 














Sub- 
Bangers 


U2L2 
MiLj 

ML. 
MI.. 
MtI.t 
M + ■ 




6.376 
49.287 


30.77C 
30.77C 
30.77( 
28.57] 
25.00( 
25.00( 




> . ... 


15.385 


) 61.53S 

57 . 14: 

) 50.00( 

) 50. OCX 


> 30.770 
5 28.571 
) 25.000 
) 25.000 














■f. 

■r. 
A % 


L2M1 

LtMl 

L4M5 
LeM? 






MiLix .6823 






MsLsx .6823 






MsLsx .6321 





5 






M7L7 X 


.6297 





291 



DEAD LOAD STRESSES 

Specification: — 

The dead load for which the bridge will be calculated is as follows), — 

The weight of all material remaining in the completed bridge and a 
snow load of 100 lbs per lineal foot of each track, 150 lbs per lineal foot of 
each sidewalk and 150 lbs per lineal foot of each bottom chord of the canti- 
lever and anchor arms. 

The weight of railway floor above stringers to be assumed at 600 lbs 
per lineal foot of each track and the weight of sidewalk floor above stringers 
to be assumed at 200 lbs per lineal foot of each sidewalk. 



The dead load stresses were revised several times as the design 
developed. The final stresses, as used for determining the sections of the 
members and the weight of the material producing these stresses and its 
distribution to the different panel points, are given on plates Nos. CV1, 
CIX, CXIV. The weight of the steel given was estimated from preliminary 
shop drawings and sketches and to the weight thus obtained an addition of 
2% was made to allow for further increase in weight as the shop drawings 
were completed and to allow for overrun in the weight of the steel. It has 
developed however that the 2% addition was just about sufficient to cover 
the increase in the final estimated weight as taken from the approved 
shop drawings and that the weight of the steel in the completed structure is 
about .65% in excess of the weight for which the stresses were calculated. 

Table No. 3 gives the estimated weight, including the 2% addition, 
used for calculating the stresses and also the shipping weight. It will 
be noted that a large percentage of the increase occurs at points where it 
would not effect the stresses in the main structure. 

Table No. 3 also gives the percentage of nickel steel used in the struc- 
ture. The floor, swaybracing and top chord supporting trusses were built 
of carbon steel while the truss members and laterals were built of carbon 
or nickel steel as noted on plates XXIX, XXXII and XXXIII. 

The calculations for the stresses in the Cantilever Arm will be given, 
those for the Anchor Arm are similar and were obtained by the same for- 
mulae with the necessary changes in the coefficients. The only difference 
was that it was first necessary to obtain the reaction on the anchorages 
from the weight of the channel span, and then consider it as a concentrated 
load. The calculations for the suspended span were made in the usual 
way. The center panel was made determinate by leaving the connection 
at one end of the lower diagonals open until after the span was completely 
erected and resting on the end supports, thus insuring that this member 
would carry no dead load stress. 

292 




293 



eci>iMcoeo05t^cooo2t^t>'-iX'-it^c v Jt^'-ifo05>OGOfO 

H N i-l N H pj H CO 35 ."C i X -h O O O Q ^ to 05 <N >-H © 
"<# CO OS "# OS OS tO OS 00 i— i b- CO iO O CO <N i-( CO 

CO<M HNHCOHHHNrtH IH 



■NH 
■OSIO 



■t^O 



^(Mr-OSHr^CSINOJ© 

COCOOOXi-HtOOOOsOs 

Tfi^os'^cccsos-^i-t'ti 
co o ho m 



©*# 00 -*1^<M 
COt^iOOO -tf <N 



B 

c 

O 



to O iO ■>* ■* CO ■* 
<— i 00 t^. lO 
O h —( 




294 



Calculations for the Cantilever Ann dead load stresses — 

roiiMii.AK. Refer to Fig. No. 1 

Q-Un I 

Vu = M»i — Rnf 

l'f 
LJ-Mro+Qj 9= — 

= Mm-hltn V f 

h 

Tl -Mm— Rnf— 1 u«=t*u 

=Uu — Uimiz 

Did = Mm— Rnf— Lihz = Uw— LiZiz 

di 

Di = D 2 + Rn — 

h 

Hl =Di-7+Conc. Li 
d 

Hf =HL+Conc. F^ 
P = T— +Conc. Ul 

Panel 0—2. Refer to Fig. No. 3. 

1 
UoU»=» Ma- 
li 

1 

UoMi = M2— 

a 

lenth M1L2 
MiL2 = UoMi+Ri — 7l — =-j- 

Table No. 4. gives the elements of the truss and the coefficients of the 
formulae required for calculating the stresses. 

Table No. 5. gives the total concentrations at the different panel points 
and the moments. It will be noted that the concentrations differ slightly 
from those shown on plate No. CIX but the difference was not considered 
great enough to warrant another revision of the stresses with the exception 
of the sub members carrying one panel load only. 

Table No. 6 gives the calculations for the stresses. Concentrations 
under Ru are as given on plate No. CIX for the reason given in the last 
paragraph. 
Elements of the truss. 

(p+h) (f+g) 

cosi =0.9798 ' = " 

j (p+ h) (f+g) 
cosj =0.9732 d = 

P+h 
« =(p+h)cosi Z ~p+hi 

hi 

1 =( P +h)cosj y - (p+h) (f+g) 

295 





LO 


to 


I-H 


•<* 


o 


-" 


t^ 


•^ 


O 




01 


on 


t- 


LQ 


<N 


I— 1 


o 


o 


o 




CN 


-r 


IM 


<M 


i-~ 


CO 


co 




0-1 


(X) 


TtH 


i— 1 


CM 


— 


1- 




o 


CO 


00 




i-H 




CI 




I-H 


iH 





w 
z 
< 
&H 



O O O ■* 00 ■* CO 

O O O 00 LO o o 

O O O CO 00 CN 

LO IOO (OWN 

tj co h co oo o 

H H M I-H I— I 



M CO H HI O 00 O) 
"* 05 •<* 00 00 C5 C3 
00 i-H O NION 



o co co 


00 i-H 


O MCC 


co «fi 


r-H 1-H CM 


1-H 1-H 



O H H •* M CD CO 
00 00 CO M 00 •* N 
CO CO O O LO t>- 



00 CO i-H 
i-H CO 00 

115 1C O 



i-H -HH O 

tH OS O 

CO lO lO 



00N WON *N 
O J> 00 lO CN CM OS 
00 lO CO l> CM CM 



ijl 00 M 00 N CO (N 
if) «5 n i)( H Ol 00 
N CO "<*1 Tf Tfl •># 



NWO W 
•* CO 00 CO 
CS CO CM 


LO CM 
CM t- 

lo co 


00 ■* CO 
CO ■* 00 


lO CO 

cs *o 






-ssaax ao sxN3P«a7g[ 



"*CSCO00O0O0-*COCO©O0 
1>c0OO0000iOC0i-hi0O 

• io n co co co oo co 

MHaOWoOrtiONNif) 

O O lo ■* o 

CO CO 1-H 1-H 1-H O 



HOHr)t(«COiOCO^NM 

ONOHiOOltOffli- 1 O O 

• N (O lO N N ■* LO 

NLOO(OiOCOHiOht»iO 

co co its co o 

CM CM .-H 1-H 1-H O 



NHXO)CC*GONNNO 

<MI>COI>OOOI^Oi-HCO 

• O N 00 OS N O CO 

i-Hocoococoi-Hiocot^o 

CO CM ■* CO O 

CM CM i-H 1-H i-H O 



HJMNtOHNNHOOCON 
ONCOCONrtCCNMGON 
• • • • CO 1> CM CO O CO i-H 
lOCOCOCOCOOOCNCOCOCOlO 
O CS CO CN • • " • • • O 

1-H i-H 1-H 1-H 1-H © 



O-^iOCOCMrHOOOCMO'* 
00t>00COi-HCNCOCMCMC0CO 
••* N ■* N 00 ih Ol 
CONHCOCOOONMOOiO 
lO lO i-H O O 

I-H I-H I-H I— I 1-H © 



LO 

lOOlOlCOOOiOO'fHW 
NOONCOHi)(oOCOiOOCN 
• • • ION ■<)< wt^H N 
N(OOi*000(NtOiOOO 
CN CN CX CO • • • • • O 



OHLOCliOONiOMNfO 

WCNHiOtOMHfflM'* 

• LO N LQ O CO O 00 

CMi-HCO00CO00CMl>LOCO00 

O O N CO o 

i-H i-H i-H O 



• • 00 






• CO LO LO 


CM 


b- i-H 


• • CO O 00 




i-H O 


• oo t~ 


rt< 


t- r-i 


. co co oo 


N 


LO O 


• • LO ■ ■ 




o 



X A 



J3 ^T3 

~ ~ t3 Sua 



sxNiaioiaaaoQ 



296 



TABLE No. 6a— MOMENTS 



Panel 


Cone. 


Shear 


Lever 


Mom. 


Rnf 


Mm-Riif 


Mm 


Point 


[tax 10" 


Dbsz 10" 


ft. 


ft.- lbs 


ft.-lbs 


ft.-lbs 


ft.- lbs 










x 106 


X106 


X 106 


x 10« 





3391.5 


3391.5 


65 


220.5 








1 


356.3 




42 




15.0 






2 


575.3 


4323 . 1 


48 


207.5 




220.5 


235.5 


3 


161.2 




25.5 




4.1 






4 


661.5 


5145. S 


59 


303.6 




443.0 


447.1 


5 


202.6 




31 




6.3 






6 


899.4 


6247.8 


72 


449.8 




750.7 


757.0 


7 


249.3 




40 




10.0 






8 


L222 1 


7719.2 


84 


648.4 




1206.8 


1216.8 


9 


287.2 




42 




12.1 






10 


1576.5 


9582.9 


84 


805.0 




1865.2 


1877.3 


11 


347.4 




42 




14.6 






12 


1893.1 


11823.4 


84 


993.2 




2682.3 


2696.9 


13 


377.4 




42 




15.8 






14 


2361.8 


14562.6 


84 


1223.2 




3690.1 


3705.9 


15 


501. S 




42 




21.1 






16 


915.6 


15980.0 










4929.1 


4950.2 



297 



TABLE No. 6b— COEFFICIENTS AND MOMENTS 





CONCENTRATIONS 








MOMENTS 






_ 


lbs x 103 








ft. lbs X 


106 






o 



















o3 




























1 












Ph 


































1 RlTf: 














Ui Fi 


Li 


Rn 


Mffl 




h U Uw 


Ui 


wiz Li, 


1Z 1 


't D2d 




CO 


■* 


lO 


o 


i-O 






LO 












CN 


co 


CO 


co 


co 


id 






id 












d 


CN 


10} 


CN 


lO 


CO 






co 
















CN 


co 


CN 






CN 












x 


LO 


00 


CN 


o 


l-H 


cc 


r> 


O 


w 


CN 


OC 


00 


id 


<# 


r^ 


CM 


l^ 


co 


c 


co 


L0 


L0 


t^ 


I> 


CN 


o 


<o 


00 


co 


■* 




IC 


tP 




a 


Tt 


<* 








1—1 


T-H 


-* 




"tf 


-* 


CN 


<M 


1- 


T-H 


3 


CO 


r-i 


co 


cn 


o 


cc 


cr 


1^ 


C 


1> 


1> 


o 


00 


00 


o 


CN 


t^ 


i*: 


o 


d 


co 


e« 


t^ 


00 


t^ 


t^ 


o 


O 


i-O 




cc 


LO 


l« 


cr 


c- 


CO 




CN 


t— 1 


CO 


CN 


t>. 




t^ 


t^ 


iC 


L0 


1— 


i-H 


I 


T-l 


o 


co 


CO 


00 


r^ 


lo 


00 


o 


c 


CC 


00 


§8 


o 


d 

lO 


°2 


CO 

-H 


oc 


LO 

CN 


d 
5 


co 
co 


• g 


co 


00 
lO 


i 


CO 


CI 


T« 




CN 

i—i 




CN 

T-H 


CN 

l-H 


05 


q& 


CN 


CN 


o 

T-H 


OS 


00 


1—1 


lO 


co 


lO 


0C 


CN 


o 


CN 


(N 


q 


1—1 


co 


id 


,_! 


t^ 


c 


t^ 


id 


CM 


lO 


co 


d 




tM 


CN 


C5 


00 


t>. 


i-H 


| 


o 


00 


o 


oc 


CO 


00 


lO 


CN 


•o 


cn 


oo 




00 


■* 


LO 


co 


CO 












i-H 






T-H 


l-H 


T-H 






<N 


00 


o 


i— i 


MS 


OS 


t^ 


cc 


CO 


00 


cs 


U0 


i-H 


1—4 


























6 


id 


id 

CN 


9 


cn 


CO 

C5 


CN 

l-H 


8 


C9 


— * 

^H 


00 

CO 


o 


3 


t-h 


CO 


<N 


i> 




CO 




i^ 


CN 


es 


■<* 














CN 




<N 


CN 


(M 


CM 






«* 


00 


OS 


00 


CO 


Ci 


00 


t^ 


i—l 


o 


LO 


l-H 


CO 


i— i 


























i 


1— I 


co 


8 


^H 


id 


co 


03 


d 


CM 


co 


8 


d 


CN 


OS 


co 


1> 


o 


l-H 


i-H 


a> 


O 


CO 


lO 




t^ 


CN 


o 


co 


co 




co 


CO 

co 


CO 


CO 


U3 


U0 


CO 














CN 


CM 






l-H 


lO 


© 


co 


l-H 


i— ( 
1 














CO 


d 






d 


o 


CO 


05 


8 


■«* 














OS 


o 






(M 


o 


CN 


co 


i—i 














"* 


C5 






C5 


l-H 


CN 


J> 


t^ 



298 



TABLE No. 7— STRESSES 

lbs x 10 3 



U 



D> 



Di 



Hi. 



Hi 



3 



- 



= |g 

■* re 

>-i I <N 






1H l-H CJ 



tC r- 
r- X 



r. _ 
— r. 

^H 1-1 oc 



!9 



-' 



lO I i—i 
O '-Oi 



ION 
?l ?1 



o 


•r 


— 


i- 


o 


Q0 


•-C 


»o 



a 



X « 

— 



^O I 4 

mo © 
i-> I m 



53! 



a 



s ■ -. 
x - 
r. •-. 



3) 



— N 
re M 
r- to 



83 I 3 

M © GO 



XI I © 



3- N 
MOM 
M © — 
C<3 I rf 



■z ?i X 

"I© © 

CON © 

<n I n 



>~ 


re 




-- 


s 




















T. 


M 






<N 




















re 


I- 


s 


3 




















— 





























299 



OTHER STRESSES 

The specification for, the amount of, and notes on the calculations 
of the Impact, Wind, Temperature, Traction, Brake, Friction, Torsion 
and Secondary Stresses are given on the following plates, — 

Impact, Floor — Plates XXV and XXVI, 

Trusses — Plates XXIX, XXXII and XXXIII, 

Wind, Trusses and Laterals, Plates CV, CVII and CXI, 

Sway Bracing, Plates XXXV and CXV, 

Temperature, Plates CVIII, CXII, XXXIV and CXIII, 

Traction, " CV, CVIII and CXII, 

Brake, " CV, CVIII and CXII, 

Friction,- " CV, CVIII and CXII, 

Torsion, " CV, 

Secondary, " CX, XXXIV and CXIII. 

Plate CVI shows the lateral displacement of the structure under a 
wind of 30 lbs. per square foot, normal to the bridge, and the hori- 
zontal and vertical displacements of the trusses under the maximum 
variation in temperature. 

Plate XXVII shows the provision made in the floor and in the trusses 
to permit the free expansion and contraction of the structure, due to 
change in temperature and to the deformations under direct stresses. 

Torsion. 

As pointed out when discussing the live load stresses in the anchor 
arm, with one track only loaded the one truss carries 68.47% of the load 
while the other carries 31.53% resulting in unequal deflections in the 
two trusses. Owing to the absence of top chord laterals in the Cantilever 
and Anchor Arms no direct stresses due to this cause result in the truss 
members of this portion of the structure except as noted under the Anchor 
Arm live load stresses. Under such a loading the four corners of the 
suspended span will not all be at the same elevation and a condition may 
arise when two diagonally opposite corners will be at the same elevation 
with the other two corners at a higher or lower elevation. The maximum 
difference in elevation for the live load specified was found to be 1.1 inches. 
A solution for the determination of these stresses was developed by Mr. 
Cyril Batho, M.Sc, B. Eng., then Associate Professor of Applied Mechanics 
of McGill University, which has been fully described by him in " Engineer- 
ing " of London, Eng. of Oct. 15th, 1915, pages 392 and 393. 

300 



The maximum torsion b1 ist with the live load da shown for 

the maximum stress in the laterals on Plate CV.. These stresses are 
only a very small percentage of the total stresses in the truss members 
and were therefore calculated for only one condition of loading ('see | 
CV) and the stress thus obtained considered as the torsion stress in the 
members regardless of the condition of loading giving the maximum live 
load stress in the individual members. 

TOTAL STRESSES— SECTIONS OF MEMBERS 

Loads used to determine section of members. 

Specification: — 

All the co-existing loads and stresses and the deformations shall 
determine the sections of the different members with the following restric- 
tions: — 

The sidewalk live load shall be used for floorbeams and sidewalk 
stringers and members receiving their maximum strain from a length of 
moving load covering two panels or less. 

Stresses produced by a difference of temperature of 25° fahr. betw r een 
the outer webs exposed to the sun and the other webs of compression 
members shall be considered as secondary stresses. These stresses and 
the stresses produced by a difference of 25 degrees Fahr. betw-een the 
temperature of a shaded chord and the average temperature of a chord 
exposed to the sun shall be assumed to co-exist with one-half the wind 
load of 30 lbs. per sq. ft. normal to the bridge. 

^tresses. — 

Specification: All parts of the structure shall be proportioned so 
that the sum of the maximum stresses produced by the loadings specified, 
including impact, shall not exceed the following amounts in pounds per 
square inch. 

For Carbon Steel: 

Tension : lbs. per sq. in. 

Eyebars 20000 

Riveted Members 18000 

Including secondary stresses 24000 

Compression : 

I 

Short members with 50 and under 14000 

r 

Long members with — over 50 17"i00 — 70 — 

r r 

Including Secondary stresses ] ^000 

301 



Shear : 

Webs of plate girders, gross section 10,000 

Laterals and Sway Bracing: Take both systems in calculation 

of strains disregarding reversal of strains. For compression . .16,000 

r 

Bending Stresses: 

All bending stresses in compression member produced by the 
weight of the member itself and by loads applied on the member shall 
be considered as primary stresses. All such members shall be 
proportioned so that the greatest fibre stress due to this bending and 
axial strain together will not exceed the allowed units for the axial 
stress in that member. 

Alternate Stresses: 

Members subject to alternate tension and compression shall be 
proportioned for either stresses and their sections shall be made equal to 
the sum of the two sections. 

Erection Unit Stresses: 

Increase the unit stresses given above by 20% except in the case 
of shear in the webs of plate girders where a unit shear of 13,000 lbs. per 
sq. inch will be allowed. 

For Nickel steel: 

Increase the unit stresses given for carbon steel by 40%. 

Material Stress Sheets, 

Floor: The total stresses and the material used are shown on the 
following plates: 

Plate XXV Track Girders, Sub floorbeams and stringers. 

Plate XXVI Main Floorbeams. 

The material required to carry the loads specified for the permanent 
structure was not sufficient in the floorbeams of the cantilever arm to 
support the traveller during the erection of this portion of the structure, 
without exceeding the unit stresses specified for erection and it was neces- 
sary to increase the thickness of the web plates of the single web floorbeams 
and the lengths of the cover plates of all the floorbeams to provide for 
these stresses. 

Trusses: A summary of all the stresses, the total stress used for 
determining the sections of the members, the bending stress in the compres- 
sion members due to their own weight, the allowable unit stresses and 
the material used are shown on the following plates. 

Plate XXIX Suspended Span, Truss, Laterials and Sway Bracing, 
also shows typical sections of the different members. 

Plate XXXII Cantilever Arm, Truss and Laterals. 

302 



The sections of all sub-members supporting the front legs of the 
traveller during erection were determined by the erection stresses. These 
members and the floorbeams mentioned above were the only parts of 
the permanent structure requiring additional material to carry the erection 
loads. 

Plate XXXIII Anchor Arm, Truss and Laterals, 

Plate XXXV Cantilever Arm, Sway Bracing, 

Plate CXY Anchor Arm, Sway bracing. 

Plate VII shows the cross sections and dimensions of the members 
of the Cantilever Arm. The sections of the members of the Anchor 
Arm are similar. 



303 



BIBLIOGRAPHY 

FROM 
THE ENGINEERING INDEX ANNUAL 

BY COURTESY OF 

THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 



CANADIAN ENGINEER 

July 14, 1910. — Quebec Bridge Substructure. Illustrated article 
dealing with the new substructure, its difficulties and construction methods. 

Oct. 31, 1912.— The Construction of the South Main Pier of the 
Quebec Bridge. H. P. Borden. Illustrated description of the methods 
of construction and conditions. 

Feb. 13, 1913. — Progress in Connection with the Construction of 
the Quebec Bridge. Illustrates and describes the construction of the 
new piers. 

Jan. 22, 1914. — The St. Lawrence Bridge Company's Shops. 
Drawings and description of plant built and equipped for the fabrication 
of the Quebec Bridge and of the field equipment for its erection. 

Apr. 9, 1914. — Construction of the New Quebec Bridge. From an 
address by C. N. Monsarrat at Toronto, discussing the reconstruction 
of the new bridge. 

July 9, 1914. — Substructure of the Quebec Bridge. H. P. Borden. 
A resume of the construction of the piers and abutments, interesting 
caisson sinking and plant operations. 

Nov. 12, 1914. — Main Pedestals, Quebec Bridge. H. P. Borden. 
Notes on the design of the four 400-ton shoes to transfer the load from 
the cantilever and anchor arms to the main piers. The method of fabrica- 
tion and assemblage. 

Dec. 31, 1914. — Progress of the New Quebec Bridge. H. P. Borden. 
Reviews the superstructure erection to date. 

Sept. 23, 1915. — Progress of the New Quebec Bridge. H. P. Borden. 
Review of 1915 work. Illustrated. 

June 1, 1916. — The New Quebec Bridge. A. J. Meyers. Details 
of method for hoisting the suspended span in place. 

304 



17, 1916.— South Cantilever Arm., Quebec Bridge. A. J. 
Meyers. Details of progress during the present season. 

Sept. 14, 1916.— Quebec Bridge Centra] Span, with short editorial. 
Circular issued I plaining arrangements for placing span. 

Sept. 21, 1916.— Quebec Bridge Disaster. Hoisting of structure; 
incidents leading to collapse. 

Sept. 20, 1917.— No Rocker Bearings nor Si ags This Year. 

A. J. Meyers. Lower shoe of roller bearing riveted to girder while upper 
shoe is riveted to span and key carries the load during lifting. 

Sept. 20, 1917.— Erection and Floating of Central Span. Floating 
arrangements, influence of the tides and general plan. 

Sept. 20, 1917.— Canadian Engineering Has Triumphed at Quebec. 
Central span of Quebec Bridge successfully floated and hoisted. 

Sept. 20, 1917.— Mooring and Hoisting the Suspended Span. A. J. 
Meyers and M. B. Atkinson. General principles of last year's operations 
were again followed. 

Feb. 7, 1918.— Expansion Joints and Traction Trusses Quebec Bridge. 
A. J. Meyers. Sliding rail expansion joints allow motion of 17 H inches 
between suspended span and each cantilever arm. 

CONTRACTING 

July, 1916.— Erection of the Quebec Bridge. Serial 1st part, 
Methods and plant. 

October, 1916.— The Fall of the Quebec Bridge Suspended Span. 
Also two editorials. Description and cause of accident. 

CORNELL CIVIL ENGINEER 

May, 1916.— The Erection of the New Quebec Bridge. N. C. 
McMath. Methods used in this work. 

ELECTRICIAN 

Dec. 17, 115.— The Erection Plant for the Quebec Bridge. Serial 
1st part. Particularly crane equipment used. 

ENGIN EERING — London 

Sept. 2, 1910.— Caissons for the Main Pieis of the New Quebec 
Bridge. Illustrates and describes the work in progress and methods 
employed. 

Mar. 24, 1911.— Tests of Model Chords made for the Quebec Bridge. 
Describes various members tested and reports results of physical and 
chemical tests. 

305 



May 19, 1911. — Schemes of Erection Proposed for the Quebec Bridge. 
Gives line drawings and description of proposed schemes. Serial 1st 
part. 

Aug. 18, 1911. — Quebec Bridge Caissons. Illustrated description of 
the construction of the caissons. 

April 12, 1912.— The North Main Pier of the Quebec Bridge. Plan 
of the works and illustrated description of the construction, plant and 
methods. 

Dec. 27, 1912.— The South Main Pier of the Quebec Bridge. Plan, 
Illustrations and description of the construction work. 

Sept. 26, 1913. — The Construction of the Masonry for the New 
Quebec Bridge. A review of the construction of the piers. 

July 31, 1914. — Construction of the Quebec Bridge Members. 
Discusses the actual construction illustrating many of the operations. 

July 24, 1914. — The Superstructure of the New Quebec Bridge. 
Detailed description of the progress. 

Sept. 25, 1914. — The Erection Equipment of the New Quebec Bridge. 
Describes the plant required for the erection, and methods. 

Dec. 4. 1914. — Design of the Main Shoes of the New Quebec Bridge. 
Detailed description. Illustrated. 

Jan. 29, 1915. — Progress of the New Quebec Bridge. Reviews the 
work accomplished since July 15, 1914. Illustrated. 

Oct. 29, 1915. — Progress of Erection of the New Quebec Bridge. 
Erection since April. 

July 16, 1916. — The Erection of Quebec Bridge. Programme for 
season 1916. Work to be carried during the present season. 

Oct. 13, 1916.— The Accident to the Quebec Bridge. Also Editorial 
on the supposed cause. 

Oct. 27, 1916. — Erection of the South Cantilever Arm of the New 
Quebec Bridge. Serial, 1st part. Detailed account. 

Nov. 2, 1917. — Erection of the Suspended Span of the Quebec Bridge. 
Detailed account of this important engineering feat. 

Feb. 15, 1918.— The Quebec Bridge. Review of the undertaking 

and itd accomplishment. 

Mar. 24, 1911.— The Reconstruction of the Quebec Bridge. Frank 
W. Skinner. Brief description of the location and conditions, and of the 
design and construction of the new structure, with information relating 
to methods of removing the old bridge and of testing models of the new 
design. 

306 



Dec. S, 1911. — The Quebec Bridge. Editorial review of the article 
bj Guatav Lindenthal, published in the Engineering News. 

Nov. 3, 1916.— The Accidenl to the Quebec Bridge. Frank W. 

Skinner. Serial 1st part. Methods adopted to ensure, safety, with 

details of accident. 

<>ct. 28, 1917. -The Quebec Bridge. Reviews the history of the 
bridge and its successful completion. 

Sept. 13, 1918. — The New Quebec Bridge. Account of its erection, 

ENGINEERING \\I> CONTRACTING 

Sept. 27, 1916. — The Erection of the Suspended Span, New Quebec 
Bridge. ('.. V. Davies and \". ('. MeMath. Details of plans. 

Sept. 27, 1916. — What Lesson Does the Second Quebec Bridge 
Disaster Teach? Editorial on danger of applying test data beyond 
limits of actual tests. 

Sept. 26, 1917.— The 640 Ft. Suspended Span of Quebec Bridge 
Hoisted to Permanent Position. From the " Montreal Daily Star." 
Various stages of the undertaking. 

lAGINEERING MAGAZINE 

December, 1913. — Lifting the One Hundred and Thirty Million 
Pound Quebec Bridge. H. F. Stratton. Describes the part of electric 
equipment applied to erection problems. 

ENGINEERING NEWS 

May 19, 1910.— An 1800 Ft. Steel Arch as a Quebec Bridge project. 
Drawings and general description of a design for a voussoir arch to cross 
the St. Lawrence River at the site of the collapsed bridge. 

Sept. 8, 1910. — Caissons for the Main Piers of the New Quebec 
Bridge; Launch of the North Pier Caisson. Illustrates and describes 
the substructure work for the new Quebec Bridge. 

Apr. 20, 1911. — Designs for the new Quebec Bridge and the Accepted 
Design. Gives outline diagrams of the accepted design, with six compel ing 
designs, with notes. 

Aug. 14, 1911. — Schemes of Erection Proposed for the Quebec Bridge. 
Gives a summary of the erection project of the more important bids for 
this structure to be built over the St. Lawrence River. 

Nov. 16, 1911. — Notes on Quebec Bridge Competition — Gusta\ 
Lindenthal. A critical discussion of submitted designs, the accepted 
design and matters related. 

307 



May 30, 1912. — Remarks on the Quebec Bridge and a Proposed 
Cantilever Design. C. A. P. Turner. Gives an outline sketch and 
general description of the Author's design. 

Nov. 7, 1912. — The Construction of the South Main Pier of the Quebec 
Bridge. Describes the methods and the Contractor's plant. 

Apr. 30, 1914. — Design of the Superstructure of the Quebec Bridge. 
H. P. Borden. Brief description of the completed substructure and the 
progress on the superstructure. Inset sheets showing design and other 
illustrations. 

May 14, 1914. — Special Shop Work on the Heavy Members of the 
Quebec Bridge. H. P. Borden. Large detail photographs with descrip- 
tion of equipment and shop methods. 

Jan. 7, 1915. — Progress of Work on the New Quebec Bridge during 
the First Erection Season. H. P. Borden. About 80 per cent of North 
Anchor Arm erected during 1914. General procedure. 

Mar. 4, 1915.— The Erection Traveller. New Quebec Bridge. H. 
P. Borden. Describes the type adopted and the method of erection. 
Illustrated. 

Sept. 2, 1915. — Quebec Bridge Work in 1915. Statement of progress. 

Jan. 6, 1916. — Quebec Bridge Erection Progress in 1915. H. P. 
Borden. Completion of north cantilever and entire south anchor arm. 

Aug. 17, 1916. — South Cantilever Arm of Quebec Bridge Completed. 
A. J. Meyers. Also editorial, 13,000 tons of steel erected in 92 days. 
Erection methods. 

Aug. 31, 1916. — Quebec Suspended Span Hoisting Details Completed. 
A. J. Meyers. Outline of plan for placing this span. 

Sept. 14, 1916. — Erection of Quebec Bridge Suspended Span. A. J. 
Meyers. Details of method followed and the failure. 

Sept, 21, 1916.— The Full Evidence of the Fall of the Quebec Bridge 
Span. Also editorial. Investigation of the structural conditions preced- 
ing the fall. 

Oct. 5, 1916. — Computing the Stresses in the Quebec Rocker Casting. 
Also editorial. An analysis of the detail that failed. 

Mar. 29, 1917. — Spring Friction Will Hold Quebec Span Against 
Drag of Braked Trains. A. J. Meyers. Traction brake described. 

NEWS-RECORD, Sept. 27, 1917.— Quebec Suspended Span Success- 
fully Hung from Cantilevers. Harry Barker. Details of the work, 
describing new features in hoisting arrangements. 

308 



ENGINEERING RECORD 

Sept. id, 1910. — The Reconstruction of the QuebecBridge. Explain 
the general conditions, describing the original structure, its erection ana 
collapse, and the general design of the raw structure. 

Sept. 24. 1910.— The Plant for Constructing the Quebec Bridge Sub- 
structures. Information concerning the reconstruction, illustrating and 
describing the costly contractor.-' plant for carrying on the work. 

Oct. 1. 1910. The North Caisson of the Quebec Bridge. Illustrates 

and describes details of design and construction. 

Oct. 15, 1910. Building and Sinking the Quebec Bridge North 
Caisson. Illustrated detailed description of this feature of the work 
and of the temporary house protecting the caisson during construction. 

Nov. 19, 1910. — Tests of Nickel Steel Models of Compression Mem- 
bers in the Official Design of the New Quebec Bridge. Illustrated detailed 
description of tests with editorial comment. 

Max 27, 1911. — The Accepted Design of the Quebec Bridge. Line 
drawings and discussion of details. 

Aug. 12, 1911. — Quebec Bridge Ciassons. Diagram and illustrated 
description of methods of construction. 

Nov. 30, 1912.— Progress on Quebec Bridge Substructure. Illustrates 
and describes recent work and methods used in the construction. 

March 14, 1914. — Fabrications of the Quebec Bridge Members — 
the Longest and Heaviest ever built. Illustrates and describes how- 
massive steel plates and shapes for 60,000 ton superstructure are assembled 
in a specially equipped §1,000,000. shop. 

March 21, 1914.— Utimate Strength of Carbon and Nickel Steel 
Models of Quebec Bridge Members. Illustrated account of the destruc- 
tion tests of precision on large members in the 640 foot suspended centre 

span. 

March 28, 1914. — Progress at Site of Quebec Bridge. Reports the 
substructure and approach spans completed; storage yard and contractor's 
i-ainp established, steel received and traveller tower erected. 

Apr. IS, 1914. — Quebec Bridge Anchor-Arm Spans. Illustrates 
and describes important developments in main trusses, bracing, twin 
tension and compression members, multiple-pin connections and field- 
spliced shipping units. 

June 13, 1914. — Main Pedestals for New Quebec Bridge. Four 
400-ton built-up riveted and cast-steel members supporting cantilever 
and Anchor Arms on the main piers. 

309 



July 25, 1914.— Ultimate Strength of Carbon Steel Models of the 
Quebec Bridge Members. Records and diagrams of destructive tests; 
anchor arm lower chord models buckled under a unit load of 50.886 pounds. 

Aug. 8, 1914. — Quebec Bridge Girders and Wind Anchorage. 
Describes features of the floor system and the anchorage. 

Sept. 26, 1914. — Quebec Bridge Anchor Arm Bottom Chords. 
Design of 43 ft. compression members. 

Oct. 10, 1914. — Quebec Bridge Anchor Arm Diagonal and Posts. 
Describes the longest web members of 515 foot trusses which are erected 
in several units. * 

Nov. 14, 1914. — Ultimate Strength of Carbon and Nickel-Steel Models 
of Quebec Bridge Members. Tension tests of reinforced steel plates and 
of alternating stress members and compression tests of four struts. 

Jan. 16, 1915. — Erection of New Quebec Bridge. H. P. Borden. 
The north anchor arm has been completely placed by 1000 ton traveller 
since Aug. 1st, 1914 with exception of two upper panels. 

Apr. 17, 1915. — Provision for Traction Stresses in Quebec Bridge. 
C. A. Norton. Types of trusses used. Illustrated. 

July 24, 1915. — New Methods Involved in Building World's Largest 
Bridge. Details of Quebec Erection. 

Sept. 16, 1916. — Suspended Span of New Quebec Bridge Falls into 
River While Being Hoisted Last Monday Morning. Also editorial. 
Account of failure and suspected cause. 

Sept. 23, 1916. — I — Breakage of Casting of Rocker-Joint Bearing 
Responsible for Quebec Bridge Disaster. II — Revolutionary Methods 
Used to Float and Hoist Center Span of Quebec Bridge. 

Oct. 7, 1917. — What Was the Cause of the Initial Failure at the 
Quebec Bridge ? Computations and specifications for steel rocker casting, 
also editorial. 

ENGINEERS' CLUB OF PHILADELPHIA 

October, 1916. — Pertinent Remarks on the Quebec Bridge Accident. 
William P. Parker. A study of the cause of the failure. 

IRON AGE 

Sept. 21, 1916. — Quebec Bridge Disaster Charged to Casting. One 
of the four cast steel bearings believed to have collapsed. 

310 



JOURNAL OF FRANKLIN INSTITUTE 

Sept, 1913.— Design of Large Bridges with Spocial Reference to the 
Quebec Bridge. Ralph Modjeeki. Discusses the more important ques- 
tions which should be considered by designers of such structures; referring 
to structures designed and constructed by the Author, particularly the 
Quebec Bridge. 

SCIENTIFIC AMERICAN 

Feb. 12, 1910.— The Design of the New Quebec Bridge. Line 
drawings and criticism of the proposed plans. 

Supplement. May 10, 1913.— The Reconstruction of the Quebec 
Bridge. G. Kuwaschein. Illustrates and describes a design proposed 
by a Russian engineer, giving calculations, showing a large saving 
and advantages. 

RAILWAY AGE GAZETTE 

Sept. 26, 1913.— Adopted Design of the Quebec Bridge. Ralph 
Modjeeki. Discussion of elements considered in designing the longest 
span in the world. 

Sept. 11, 1914.— The Erection Equipment of the Quebec Bridge 
H. P. Borden. The method adopted and the traveller built for this 
1800 ft. span. 

Dec. 24, 1915.— Progress on the Erection of the New Quebec Bridge. 
H. P. Borden. 

May 26, 1916.— The Season's Work on the Quebec Bridge. A. J. 
Meyers. Programme for completion of structure during 1916. 

Sept. 22, 1916.— The Cause of the Quebec Bridge Disaster. Failure 
of bearing. 

Sept. 28, 1917.— Quebec Bridge Central Span Successfully Hoisted. 
A. J. Meyers. Roller or key bearings used as supports during raising 
instead of rocker bearings and steel castings. 

RAILWAY ENGINEER —LONDON 

Dec, 1916. — The Quebec Bridge. Critical discussion of the cause 
of failure of lifting arrangements. 

RAILWAY AND LOCOMOTIVE ENGINEERING 

Nov., 1916.— Jacking the Quebec Bridge. The lifting jacks; their 
number and position. 

311 



Vol. XXXII. Transactions Part II. 

of 

The Engineering Institute 
of Canada 



January to June 
1918 



MONTREAL 
1918 



The Right of Publication and Translation is Reserved. 








/UUx. 






CONTENTS 

Page 

Recent Advance? in Canadian Metallurgy. By Alfred Stansfield, 

D.Sc. M.E.I.C. 319 

I s of the Chain Fenders in the Locks of the Panama Canal. 

By Henry Goldmark I M.E.I.C) 329 

Discussion on abo ve Paper 348 

Nickel-Copper Steel. By Lieut.-Col. R. W. Leonard (M.KI.C). 361 
Discussion on above Paper 384 

Kettle Rapids Bridge. By W. Chase Thomson (M.E.I.C.) 391 

Discussion on above Paper 403 

The Champlain Dry Dock for the Quebec Harbour. By U. 

Vahquet M.E.I.C 415 






INSTRUCTIONS FOR PREPARING PAPERS, ETC. 



In writing papers, or discussions on papers, the use of the first person 
should be avoided. 

They should be legibly written on foolscap paper, on one side 
only, with a margin on the left side. 

Illustrations, when necessary, should be drawn on the dull side of 
tracing linen to as small a scale as is consistent with distinctness. Black 
ink only should be used. They should be drawn so that all details and 
lettering will show distinctly when reduced to a height of 7 inches. 

When necessary to illustrate a paper for reading, diagrams or lantern 
projections may be furnished. Diagrams must be bold, distinct, and 
clearly visible in detail for a distance of thirty feet. 

Papers which have been read before other Societies, or have been 
published, cannot be read at meetings of this Institute. 

All communications must be forwarded to the Secretary of The 
Institute, from whom any further information may be obtained. 



316 



TRANSACTIONS — VOL. XXXII — Part II. 
PREFATORY NOTE. 



This volume (Vol. XXXII, Part II) bridges the activities of the 
Canadian Society of Civil Engineers and The Engineering Institute of 
Canada. The papers presented are those for which arrangements had 
been made under the Canadian Society of Civil Engineers. At the 
Annual Meeting held January 21st-22-23rd, 1918, the Report of the 
Committee on Society Affairs was adopted which changed the name to 
The Engineering Institute of Canada and brought into operation a decid- 
edly different method of procedure in respect of the presentation of 
papers. 

Under the new ruling, all papers presented at any branch have equal 
status dependent upon their intrinsic value. Although the new name 
was adopted at the Annual Meeting it was not until April 15th, 1918, 
that the Bill changing the name was passed by the House of Commons. 
Between the time of the Annual Meeting and the legalizing of the new 
name, the Montreal Branch was started as part of the newly adopted 
regulations. Under the Montreal Branch the programme of papers 
previously arranged was carried out and these are included in this volume. 

The authors' names are given with their title in The Institute although 
at that time the title was not official. In the text the former appellation 
has not been changed. 



317 



RECENT ADVANCES IN CANADIAN METALLURGY 

By ALFRED STANSFIELD, D. Be, F.R.S.C, M.E.I.C. 

February 14, 1918 

In writing of "recent" advances it becomes necessary to decide on 
some definite period of time on which to report, and this surprising war 
has given such an impetus to Canadian metallurgy that it will be appro- 
priate to regard the commencement of the war as the epoch from which 
to measure these advances. 

Metal Markets in War Time.— The first effect of the war was to 
restrict the regular operation of metallurgical plants and to stop all 
new developments. This followed naturally from uncertainty in the 
financial situation and an immediate lack of money. Looking back at 
the situation it seems strange that markets should fall, and metal 
production decrease, when it must have been certain that nearly all 
the metals would be needed in increasing quantities for warlike 
operations. After about six months this demand became apparent, 
the prices of metals began to rise, and their production was consequently 
stimulated. An early development was caused by theShell 
Committee's need of brass for cartridge cases. Copper for this 
purpose could be obtained without great difficulty, but zinc of a 
suitable degree of purity was very difficult to obtain even at a greatly 
enhanced price. Zinc ores had been mined in British Columbia for a 
number of years, but it had not been found practicable to smelt 
them in this country, and the zinc ores, or concentrates, were shipped to 
smelters in the American zinc districts. After the outbreak of the war, 
the American smelters refused to accept ore shipments from Canada; a 
refusal which was considered by many as being caused by German control 
or interest in the smelters. Whatever was the cause the situation was clear: 
the Shell Committee needed zinc, and the British Columbia miners 
had the ore, while the American smelters were unwilling or unable 
to convert the one into the other. Zinc differs from many metals in the 
fact that the smelting process yields a metal which is marketable 
without the need of any refining. The purity of the "spelter" depends 
mostly upon the character of the ore from which it is produced, thus 
"high-grade" spelter from specially pure ores will contain less than 
0.1% of total impurity, while "Prime Western" spelter, obtained from 

319 



leady ores, may contain as much as 1.5% of lead. This difference in 
the purity of different brands of spelter is not objectionable, since the 
uses to which they are put vary greatly in requirements; "Prime- 
Western" spelter is amply good enough for galvanizing, while "high- 
grade" spelter is none too good for making brass for cartridge cases. 
The British Columbia zinc ores are in general mixed with ores of lead, 
and it would be impossible by the usual smelting process to obtain a 
spelter that would meet the requirements of the Shell Committee. 

A process was under investigation, at that time, for obtaining zinc 
from these complex ores by roasting, extracting the zinc as sulphate by 
means of sulphuric acid and water, and separating metallic zinc from this 
solution by electrolysis, using insoluble anodes. This process was not 
new, but it had failed hitherto owing to technical difficulties in the 
leaching of the ore, the purification of the solution and the production 
by electrolysis of "compact zinc that could be remelted. The cost, 
moreover, was too high to make it profitable in view of the low price of 
zinc before the war. A Commission was appointed by the Minister of 
Militia to investigate the situation with respect to the supplies of copper 
and zinc. This Commission concluded that the electrolytic process 
offered the greatest probability of filling the need, and made arrange- 
ments with the Consolidated Mining and Smelting Company which led 
to the establishment at Trail of a plant for producing electrolytic zinc 
from the British Columbia ores. At the present time this plant has a 
capacity of about 300 tons of zinc per week. 

Before the war the writer carried out an elaborate investigation for 
the Department of Mines on the electric smelting of zinc ores, but prac- 
tical success was not reached. During the war this research has been 
taken up again for a mining company with more success, both in the 
laboratory and in a small plant at Shawinigan Falls. Commercial 
success has not yet been obtained, but it is expected that the work will 
be resumed during the present year. The writer has also been interested 
in the production of zinc oxide for paint by the Wetberill process; he 
helped to design and operate an experimental plant in Montreal, and 
following on from this a full-sized plant has been erected at Notre Dame 
des Anges. 

The metallurgy of zinc, as well as that of lead, copper and silver, 
has been materially improved during recent years by the application of 
the flotation process by means of which the zinc minerals can be separated 
more perfectly from the gangue, and in some cases from other metallic 
minerals. 

Copper Situation. — The copper situation, investigated by the Com- 
mission, was entirely different: copper ores are mined and smelted in 
British Columbia on an enormous scale for the production of copper 
matte, and in some cases the matte is bessemerized for the production 

320 



of metallic copper. Crude copper, however, is of no use until it has 
been refined, and all the copper derived from British Columbia was sent 
to the United States, as ore, matte, or blister copper, in order to be 
refined in the American refineries. The problem before the Com- 
mission waa the establishment in Canada of an electrolytic refinery, 
so that Canadian copper could be made available for Canadian 
needs of that metal. The problem was complicated by American 
interests in the Canadian copper smelters, by contracts with 
American smelters and refiners, by heavy freight charges between 
British Columbia and the points at which the copper would be used, and 
by the lack in Canada of rolling mills and other works for turning the 
refined copper into sheets, rods, wires and other forms required by the 
Canadian market. Although the price of copper rose to a very high 
figure, the Shell Committee was able to obtain supplies of pure copper 
from the States, and there was not sufficient urgency to justify the 
Governmenl in establishing a copper refinery in British Columbia in face 
of the natural obstacles already mentioned. In the meantime the Con- 
solidated Mining and Smelting Company established at Trail a small 
refinery, which is now in operation, and has an output of about 10 tons 
per day. and the copper and zinc from Trail have both been used in the 
Dominion Copper Products plant in Montreal for making brass for the 
manufacture of cartridge cases. It must not be supposed that the 
erection of this small refinery solves the general problem outlined 
above, although it may serve by way of example to help in settling it at 
some later date. The copper furnaces at Trail smelt ores of gold and 
silver as well as ores of copper, and the resulting crude copper is so rich 
in the precious metals that the Company finds it more satisfactory to 
refine the metal at Trail, thus securing the gold and silver contents, 
instead of shipping the crude metal to distant refineries which would 
entail very serious difficulties in ascertaining and obtaining credit for 
the gold and silver that is alloyed with the copper. For economical 
operation, a copper refinery should have a very large output, say 10 ) tons 
a day. The whole output of copper from British Columbia amounted 
in 1913 to about 23,000 tons, or about 60 tons per day, so that a practically 
unanimous co-operation of all the producing companies would be needed 
to support a single refinery. We may bope that in course of time a large 
refinery will be operated in Brit ish ( 'olumbia for the treatment of Western 
ores, but this will depend on the erection in that Province of mills for 
working up the refined copper and of the growth of a market in Western 
Canada that will absorb a large part of the output of such a refinery. 
The production of copper in British Columbia has increased steadily 
during recent years as the following figures show, this increase being 
largely due to the development of the Granby Company's smelter at 
Anyox. Recently the copper smelter at Ladysmith has changed hands 
and has again been put into operation. This increase of copper produc- 
tion will tend towards the establishment of a large refinery on the coast. 
The introduction of the flotation process has modified copper metallurgy 

321 



in recent years. Low grade ores, which were at one time smelted in 
blast furnaces, or even treated by wet (chemical) methods, are now 
crushed and ground to a fine powder and concentrated by flotation. The 
flotation concentrate, on account of its powdery condition, can be treated 
most easily by roasting in mechanical furnaces and smelting in rever- 
beratory furnaces. 

PRODUCTION OF COPPER IN BRITISH COLUMBIA 
IN TONS 

1913 1914 1915 1916 

23,000 21,000 28,000 32,000 

Outside of British Columbia the largest production of Canadian 
copper is from the Sudbury district, where it occurs in association with 
nickel. The production in Ontario, nearly all being from the Sudbury 
district, is about half the production in British Columbia, but its separa- 
tion and refining is dependent on the separation of the nickel and copper, 
and can only be considered as a by-product from the extraction of nickel. 

Nickel Production and Refining. — At the beginning of the war 
two companies, — the International Nickel Company and the Mond 
Nickel Company, — were operating in the Sudbury district. These 
companies smelted the ore to a matte containing from 20% to 30% of 
nickel and copper, and then bessemerized this to obtain a matte of 
about 80% of nickel and copper, containing little else than sulphide of 
nickel and sulphide of copper in proportions depending on the relation 
between these metals in the original ore. The Mond Company shipped 
their matte to England for treatment by the Mond process at their 
works at Clydach, in South Wales, while the International Company 
treated their matte by the "salt cake" process at their refinery at 
Constable Hook in New Jersey. It has for a long time been regarded as 
an economic injustice to Canada that the copper-nickel matte should be 
sent to the States for refining, but it was maintained by the Company 
officials that the cost of refining it in Canada would be very much 
higher than in New Jersey. After the outbreak of war, attention was 
repeatedly called to the situation, because it was feared that the nickel 
passed into German hands, and as a result of this outcry the Interna- 
tional Company agreed to erect in Canada a refinery sufficiently large 
for the British requirements of nickel. This refinery :s now being 
built at Port Colborne in Ontario and should be in operation early in 
the present year; it will have an initial output of 7,500 tons of nickel 
per annum. In the meantime a new company has entered the field. 
The British America Nickel Corporation, a strong British-Canadian 
company which is controlled by the Imperial Government, has acquired 
about 17,000 acres of mineral land in the Sudbury district, and early 

322 



in 1917 began the construction of a large smelting and refining works 
at the Murray mine. This company will smelt the ore to matte, and 
bessemerize the matte, on substantially the same. lines as the older 
companies; but the production of refined nickel and copper from the 
matte will be carried out by the Hybinette electrolytic process which 
is now in operation in Norway. The Dew plant is expected to be in 
operation in 1910, and will have an output of 6000 tons of refined nickel 
per annum. In 1913, the last whole year before the war, nickel matte 
containing 25,000 tons of nickel was sent abroad to be refined; during 
1917 the output was aboul -1 '2.000 tons, and by 1919 we may expect a 
production of some 13,000 tons of nickel in Canada in addition to perhaps 
three times this amount in matte refined abroad. 

During the year 1914 one million tons of nickel-copper ore was mined 
and smelted in the Sudbury district, and the resulting matte contained 
22,700 tons of nickel and 14,400 tons of copper. Some 400,000 tons of 
iron, equal to half the Canadian production of pig iron, was slagged in 
the furnaces and thrown over the dumps, and 300,000 tons of sulphur 
was discharged into the atmosphere. This sulphur would produce one 
million tons of sulphuric acid, which is equal to one-fourth of the consump- 
tion of the United States, but it is not considered worth while to save 
this at the present time on account of the cost of transportation of the 
acid to points at which it could be used. Projects have been considered, 
however, for saving the sulphur in the elemental state so that it could 
be shipped easily, or for collecting the SO._. gas in the liquid state and 
using it for the manufacture of wood pulp. With regard to the waste 
of iron, it may be remembered that more than half the nickel produced 
is used in the manufacture of nickel steel and with respect to this part 
of the output, there is therefore no need to separate from the nickel the 
iron which was originally associated with it. Attempts to make a nickel 
pig-iron which could be converted directly into nickel steel were met 
in the past by the difficulty that the copper, which is almost always 
present in the ore, was supposed to be harmful in the steel. During 
recent years, however, it has been found that if the copper is present 
in moderate amounts, not more than one third of the nickel content, a 
nickel-copper steel is entirely satisfactory and may be expected to replace 
the usual nickel steel for many purposes. Many tons of this steel have 
recently been made from Sudbury ore, by the process devised by Mr. 
Colvocoresses; and its mechanical properties and suitability for many 
purposes will soon be ascertained. It should be possible to employ 
this process on a large scale, although of course it is limited to ores 
with a small content of copper. Care will be taken, by suitable admixture 
of the ores and otherwise, to maintain a uniform product ;and it may be 
added that the small amount of precious metals contained in the ore 
cannot be saved if this process is used. 

In connection with the advances in the nickel industry mention 
must be made of the investigation of the Ontario Nickel Commission, 

323 



which was appointed in 1915 and brought out its report in 1917. This 
Commission has investigated the ores, smelting processes, refining 
methods, properties and uses of nickel, not only in Canada, but in all 
parts of the world where nickel ores are mined. They have considered 
the subject from many points of view, and have produced a very valuable 
report. Geo. T. Holloway, the metallurgist of the Commission, died 
soon after the completion of his work. 

The production of iron and steel in Canada fell off considerably at 
the beginning of the war, but recovered later, and at the present time, 
owing to the large demands for the manufacture of munitions, is larger 
than ever. This will be seen from the following table: — 



PRODUCTION OF IRON AND STEEL IN SHORT TONS* 

1912 1913 1914 1915 1916 1917 

Canadianl 2 16,000 307,000 245,000 398,000 340,000 190,000 
Iron Ore J ' 

Pig Iron 1,014,000 1,129,000 783,000 914,000 1,169,000 1,187,000 
Steel ) 

^j° tS > 957,000 1,169,000 828,000 1,020,000 1,454,000 1,735,000 



ex 



stingsj 



Electric ] 

furnace [ 61 5,600 43,000 50,000 

steel 



*The figures for production in this paper have been taken from the 
reports by John McLeish, who kindly furnished the writer with approx- 
imate data for 1917. The amounts, in most cases, have been stated 
in even thousands, as in this form they are far more easily read and 
compared. 

Electric Iron Smelting. — The production of steel has advanced 
faster than the production of pig-iron, the latter being restricted by 
the need of importing the ore, and in some cases even the fuel, for 
smelting in Canadian furnaces. The amount of ore mined in Canada 
is but a small proportion of the amount smelted, but if we include 
Newfoundland ore, as we reasonably may, we would find that rather 
more than half the iron is derived from Canadian ore. Just before the 
war the writer visited Sweden in order to study the electric iron 

324 



smelting industry in that country, and made a report to the Mines 
Branch in the full of 1914. At that time it did not appear that 
electric iron smelting could be undertaken profitably in this country, 
but in the year 1916 he made a fresh study of the situation in connection 
with a report by Mr. John Dresser on "Part of the District of Lake 
St. John, Quebec", and came to the conclusion that there is hope for 
electric pig-iron in Canada. In considering the possibility of smelting 
electrically the titaniferoua magnetites of that district it appeared 
that although pig-iron produced in that way could never compete, as 
regards price, with the product of the blast-furnace using coke as fuel, 
still it should be possible to make a limited production of high class 
white pig-iron which would command a high price and thus sell at a 
profit. Besides producing a very desirable quality of pig-iron, and inci- 
dentally helping the steel makers, this would have the advantage of 
utilizing ores which are at present worthless, and of employing water- 
power which now runs to waste. 

Electric Furnace Steel. — Although electric smelting of iron ores is 
still in the future, the electric furnace has been employed very largely 
for making steel, as is shown in the table. The electric furnace has 
not been used in this country to any extent for making new steel, but 
mostly for remelting steel scrap into shell billets. For this purpose 
several furnaces are in operation in Montreal and district and at 
Toronto a large plant containing ten 6-ton furnaces has been built by 
the Munitions Board for remelting shell turnings and other scrap. 
During the period under consideration the Armstrong Whit worth 
Company have built up an electric melting plant for the production 
of all kinds of tool steel in their works at Longueuil. The electric steel 
plant at Belleville has made a quantity of very good steel from titani- 
feroua iron ore, but at that time it was found cheaper to employ steel 
scrap than to smelt ore. The plant has recently been fully occupied 
with the production of ferro molybdenum for the British Government. 

Ferro Alloys. — There has been, for a number of years, a moderate 
production of ferro-alloys in Canada amounting in 1913 to 8000 tons, in 
1914 to 7,500 tons and in 1915 to 10,800 tons. This production consist- 
ed mostly of ferro-silicon made in electric furnaces at Welland by the 
Electro-Metals Limited, and in 1915 a little ferro-phosphorus made at 
Buckingham, P.Q. by the Electric Reduction Company. In 1916 the 
production had risen to 28,600 tons, valued at SI, 777. 000 and included 
besides ferro-silicon and ferro-phosphorus, a considerable production of 
the valuable product ferro-molybdenum, which is now made in 
electric furnaces at Belleville and Orillia. Ferro-silicon is now made at 
Shawinigan as well as at Welland. For the manufacture of steel there 
is a considerable demand for ferro-manganese and spiegel-eisen, and 
experiments are now in progress for making these alloys from 
Canadian manganese ores. The production ■ of ferro-chrome is also 
under consideration. 

325 



Gold and Silver. — More than 80% of Canadian silver comes from 
the Cobalt region, and during the period of the war the production has 
been falling slowly as the mines became exhausted, and on account of 
the scarcity of labour and supplies. The recent rise in price of silver 
has however stimulated the production. The metallurgical treatment 
of the ores had been worked out prior to the war, but the recent 
development of the flotation process has improved the concentration 
processes at Cobalt. The production of silver in 1916 was 25,600,000 
ounces, valued at $16,800,000 and in 1917 was about 23,500,000 ounces. 
The production of gold in Canada was varied from about 800,000 ounces 
to 1,000,000 ounces per annum during this period; gold is obtained from 
Yukon, British Columbia and Ontario. Before the war the two 
Western provinces had the largest outputs, but since the growth of 
the Porcupine district the production of Ontario is nearly equal to 
the sum of the other two, and the total productionhas somewhat 
increased. 

Lead. — The need of lead for bullets caused an increase in the 
mining and smelting of lead ores, with the result that in 1917, when 
this demand was reduced, there arose a difficulty in disposing of the 
lead already on hand and undergoing treatment. The management 
of the Trail Smelter informed the miners that they would only take 
ores that were relatively free from zinc, so as to reduce the ore supply 
and at the same time obtain a better grade of ore. This ruling gave rise 
to a great outcry, and arrangements were made with the Munitions 
Board to take a larger amount of the metal. The success of the 
Victory Loan should enable the production to be taken up by the 
Board, but it may be pointed out that the normal Canadian market 
for lead and lead products is amply sufficient for the Canadian 
production, and it is only necessary to arrange for the lead being 
converted into marketable forms. Some advance has been made in 
this direction and there are two "corroding" plants in Montreal for the 
production of white lead, but red lead and litharge should also be 
manufactured. The usual production of lead is about 20,000 tons per 
annum, but in 1917 it amounted to 28,000 tons. The Cottrell electrical 
precipitation process for the collection of smoke and fume has proved 
of great service in the treatment of ores of lead and zinc, and has been 
introduced at the Trail Smelter. 

Antimony. — The metal antimony was selling at about 8 cents per 
pound, but early in the war its price rose to 16 cents, and later to 
more than 40 cents a pound. Antimony is used to a large extent 
alloyed with lead for making shrapnel bullets, and the supply was for 
a time insufficient for the need, so that bullets have been made of lead 
encased in a steel shell, the whole being arranged to have the same 
size and weight as the bullets of alloy previously in use. Antimony 
is not produced to any extent in Canada, although a small amount is 

326 



derived from the electrolytic refining of lead at Trail. In view of the 
high price and unusual demand for the metal a mine of low grade ore 
at Lake George, New Brunswick was reopened and .after much patient 
work an efficient process WBt devised for extracting the metal from 
the ore by volatilization. An account of this process was recently 
presented to the Society by Mr. J. A. DeCew, M. Can. Soc. C. E. 
By the time that the process had been brought to a satisfactory cond- 
ition, other supplies of antimony became available, the price fell, 
and the mine and smelter had to be closed. 

The metal aluminium is made at Shawinigan Falls from bauxite 
which is imported from France and elsewhere as no ores of commercial 
value have been found in Canada. Figures for the production of this 
metal are not published as only one firm is engaged in the industry, but 
there is no doubt that the production has increased materially during 
the war, aluminium being needed in large amounts for the construction 
of aeroplanes and other military equipment. The value of metal exported 
from Canada in 1913 was $1,700,000; in 1914, $2,300,000; in 1915, $3,300,000, 
and in 1916, $5,200,000. 

Aluminium aud Magnesium. — The process of extracting aluminium 
from alumina (purified bauxite) consists, as is well known, in electro- 
lysing fused salts to which alumina is added. Before the war this was 
the only operation in use in Canada, that depended on the electrolysis 
of fused salts, but a fresh industry of this kind has now started — the 
production of the metal magnesium. Before the war this metal was 
made in Germany from the natural deposits of carnallite, but the war 
stopped this source of supply. The writer succeeded in producing 
magnesium in the laboratory, and the process was developed into 
commercial operation at Shawinigan Falls. At present the crude 
materials are imported, but ultimately it is intended to use Canadian 
magnesite as the ore of this metal. Magnesium is now made in consid- 
erable quantities at this plant; it is used in the form of powder for 
star shells and al oyed with aluminium for the construction of aero- 
planes. It is also used as a deoxidizing addition in melting certain metals 
and alloys. 

Cobalt. — The metal cobalt, which gave its name to the largest silver 
producing district in Canada, has not as yet found very many uses. 
It can be employed like nickel for plating, but being more costly 
its use is limited. It has also been employed in steel making. 
Valuable researches in regard to its use have been made by Professor 
Kalmus at Queen's University. A recent development is its use in the 
production of "Stellite", an alloy of cobalt, chromium and tungsten 
which is used for cutting tools in place of high speed steel. 

Brass Melting and Rolling. — In connection with the production of 
zinc and the refining of copper in Canada the need arises for converting 
these and other metals into marketable forms. The Dominion Copper 

327 



Products plant at Lachine has been built during the war for melting 
and rolling brass and copper for use in shell and cartridge cases. As the 
need for such products decreases, it is expected that this plant and 
Brown's Copper and Brass Rolling Mills in Toronto will be employed 
on a large scale for making brass and other non-ferrous alloys and for 
rolling these and the metals copper and zinc into forms suitable for 
use by Canadian manufacturing industries, so that a market will be 
found in Canada for Canadian copper, zinc, nickel, lead and other 
metals. 



328 



TESTS OF THE CHAIN FENDERS IN THE LOCKS 
OF THE PANAMA CANAL 

By HENRY GOLDMARK, M.K.I.C. 

Read before meeting of the Montreal Branch, February 28, 1918.) 

GENERAL 

When the alternative plans fur the Panama Canal were under 
discussion, advocates of the Bea level type laid great stress on the dangers 
to navigation inherent in a lock canal. Such dangers undoubtedly exist, 
although experience lias shown that the risk of serious accident is very 

small m locks that arc properly designed and carefully operated. Even 
at the Soo where the traffic, for many years, has been extremely heavy, 
only one serious accident is on record since the first lock was opened in 
1855. 

In comparing the two types in the case of Panama, it should be 
borne in mind, that the broad and deep channels provided by Lake 
Gatun, possess elements of safety which would have been absent in the 
smaller cross-sections of a sea-level canal. 

On the other hand, the accidental destruction of certain of the 
lockgates, would not only involve the risk of injury to vessels but might 
also set free the water impounded in Lake Gatun, and lower its level so 
much as to stop navigation for a long period of time. 

In working out the detailed plans of the locks, it was thought wise 
to take all possible precautions against injury to the gates and to provide 
special safeguards against further damage in case, after all, one or more 
were accidentally destroyed. 

The safeguards adopted with these ends in view are the following:— 

(1) Electric Locomotives for towing all vessels through the locks. 

These travel on a rack railroad close to the edge of the lock 
walls and have, so far. proved entirely satisfactory in controlling 
vessels and keeping them centered in the lock chambers. 

(2) Chain Fenders for protecting the most important gates. 

(3) Duplicate Gates in certain parts of the locks. There are the usual 

"guard gates" at both ends of each lock flight and besides these 
a second pair of lower operating gates is provided in Pedro 
Miguel lock and the upper chamber at Gatun and Miraflores. 
329 



(4) Emergency Dams of the drawbridge type at the upper end of each 
lock for shutting off the flow of water in case of serious injury 
to the gates. 

The first of these devices forms a part of the machinery used in 
normal locking. As long as it functions properly no further safety 
mechanism comes into operation. 

The second device, the chain fender, protects the gates, when, for 
any reason, a vessel is not under the control of the towing system. 

The third safeguard, the duplicate gate, in its turn, does not come 
into play, until the fender protecting it has failed to fulfill its proper 
function and finally: 

The emergency dam is needed only after all the preceding safety 
appliances have failed so that it becomes necessary to check the current 
of water flowing through the locks. 

A full description of the various safety appliances is given in a series 
of papers on the Panama Canal written by the members of the engineering 
staff responsible for the different parts of the work and presented to the 
International Engineering Congress at San Francisco in 1915. * 

They are also described in a concise, but readable and comprehensive 
article, published in "Engineering and Contracting" Jan. 7, 1914, which 
is the best general account of the Panama Canal known to the writer. 
Reference should also be made to an excellent paper **, which records 
the experiences obtained in the actual operation of the locks, since the 
opening of the canal. 

The chain fenders, the second of the safeguards mentioned above, 
were adopted at the suggestion of the writer who was in immediate 
charge of their design and construction. While similar fenders have been 
used in English locks for a number of years, they are believed to be 
inferior in strength and reliability to the Panama design. 

The present paper is intended to supplement previous articles on the 
Chain Fenders and more especially to put on record, in a systematic 
form, some unique tests which were made during the construction of the 
fenders, as well as since the canal has been opened for use. 

In view of the many novel features and the fact, that, in actual 
service the fenders proved entirely satisfactory in bringing vessels to 
rest, it was thought that an account of these tests might be of interest 
to the Society. 

* Transactions of the International Engineering Congress 1915. The 
Panama Canal Vol II. Also published separately by the McGraw 
Hill Publishing Co., N.Y. 1916. 

** First Year's Operation of the Locks of the Panama Canal; F. C. 
Clark and R. H. Whitehead. Journal of the Western Society of Engineers 
Vol. XXI No. 4. April 1916. 

330 



The fenders were placed in the upper and lower approaches to the 
lock flights, thus protecting the upper and lower guard gates, and also 
just above the intermediate and lower gates in the Pedro Miguel lock 
and the upper chamber at Gatun and Miraflores. 



DESCRIPTION OF FENDER MACHINERY. 

It is deemed unnecessary to give a detailed description of the fender 
machinery, although a brief account is requisite for a proper comprehen- 
s i, )U q| t i [he fenders consist of heavy chains, which normally 

span the lock chambers, near the top, being lowered to the lock floor, 
when a vessel is about to pass. Each gate and its protecting fender are 
interlocked electrically, so that the chain cannot be lowered, until the 
gate is opened, and hence is no longer in danger from collision. 

The chain is arranged to pay out under stress, when it is struck by a 
sel, so that the energy of the vessel is absorbed and it is brought to 
rest without damage. The machinery must, therefore, not only make 
provision for lowering and raising the chain, in daily operation, but must 
also include some reliable means of putting the chain under stress when 
it is stopping a vessel. Evidently the success of the entire fender depends 
upon the mechanism for producing a suitable resistance to the travel of 
the chain in its emergency action. 

In the English fenders, mentioned above, the friction of the chain 
about a horizontal cast iron cylinder placed on one of the lock walls, is 
depended upon, to give the necessary resistance. A small hoisting 
engine on the other wall raises and lowers the chain. 

The writer examined one of these fenders at Avonmouth near Bristol 
in 190S, and discussed their details with the designers and builders, 
Messrs. Brown, Lenox & Co. of Pontyprid, Wales. They are simple in 
construction, but the frictional resistance is likely to be variable in 
amount. It is also believed that lowering the chain from one end only 
is undesirable, as it often forms a loop at the bottom, which may foul 
vessels in the lock. As far as could be learned, no tests in actually stop- 
ping vessels have ever been made with these fenders. 

It is proper to add that the Panama designs were well in hand before 
the writer had heard of the English fenders, although their inspection 
proved of much interest. He would also like to record here, his indeb- 
tedness to his friend, Mr. E. B. McHenry. M. Can. Soc. C. E., for most 
valuable suggestions in connection with the first inception of the Panama 
chains. 

The adopted design was the result of an extended investigation. 
Frictional resistances of different kinds were studied, also the use of heavy 
weights, for stopping the vessels, but the hydraulic apparatus finally 
selected was considered to have advantages over all otherforms. 

331 



With some minor variations, the design shown on Plate (1) was used 
in all the twenty-four fenders built at Panama. There are three cylinders 
of the plunger type, the upper of which is suspended from beams spanning 
the machine pit while the bottom plunger rests directly on the concrete. 
The intermediate cylinder is movable, and slides on the inner surface of 
the upper and the outer surface of the lower cylinder. The chain passes 
through a hawse-pipe casting of steel, secured to a heavy anchorage, and 
is connected to the moving cylinder by a system of grooved sheaves. 
The pull of the chain when stopping a vessel is transferred to the anchors 
embedded in the concrete. 

The lowering and raising of the chain is brought about by pumping 
water under pressure into the bottom and top cylinders respectively. 
The maximum stroke is*21'3"and the multiplication given by the sheaves 
is four fold, so that the chain pays out 85 ft. from each wall, a length 
which is sufficient for the deepest lock and also provides ample stopping 
power in emergency operation. 

The chains were made from wrought iron bars 3" in diameter and 
have links 10" wide and 17" long. The sections spanning the lock 
chamber have standard Navy stud links, while open links are used for 
the part that passes around the sheaves. Considerable difficulty was 
met with in obtaining chains of proper strength, especially the open 
links which have rarely been made of so large a size. The specified 
breaking strength was 500,000 lbs. for the studded and 450,000 lbs. for 
the open links, but all shots of chain were subjected to proof tests of 
300,000 lbs. and 250,000 lbs. respectively. 

The operation of lowering and raising of the chain will be readily 
understood from the plans, especially the small diagram on Plate 1 
which shows the arrangement of the piping and valves. 

In order to start the cylinder, on either the upward or downward 
stroke, it is necessary to start the centrifugal pump and also to reverse 
the position of the operating valve which controls the direction of the 
flow. The latter is of the double piston type and operated by a small 
electric motor. Both the pump and valve motors are normally started 
from the central control house, from which all the gate and valve machines 
in the lock flight are controlled, but local control is also provided for. 

The cylinder is brought to rest at each end of the stroke by a limit 
switch which stops the pump automatically, and it also starts the same 
whenever leakage has caused the cylinder to move up or down a 
predetermined distance from its end position. 

The maximum pressure in the cylinders is from 100 to 150 lbs. per sq. 
in., the higher pressure being required in lowering the chain, as the heavy 
intermediate cylinder has to be lifted in this case. 

332 



Typical indicator diagrams taken on the first fender erected (at 
Gatun locks) are given in Figs. 1 and 2. The high pressure prevails in 
the upper cylinder when raising and the lower cylinder when lowering 
the chain. 

The pump has two stages, the first being of the volute, the second 
of the turbine type, a somewhat novel arrangement, which lias proved 
entirely satisfactory. The pump has a 6" suction and 5" discharge pipe 
and is operated at 460 r.p.m. by a To II. P. 250 volt 25 cycle induction 
motor. The lowering <>r raising of the chain is done in about one minute 
in a perfectly satisfactory manner, the chain dropping into a pit in the 
floor so as to offer no obstruction to the passage of vessels. 

EMERGENCY OPERATION. 

Lb the sole function of the fender is the checking of vessels, the 
device for maintaining a heavy tension on the chain, as it pays out, 
after being struck, is the most vital part of the entire apparatus. 

It consists of a pair of resistance or relief valves placed as shown in the 
Diagram of Operating Machinery on Plate 1. When the chain is struck 
by a vessel, there is a tendency for the moving cylinder to rise, so that 
the water pressure in the piping increases rapidly. The resistance valves 
must permit the water to escape, as soon as the pressure reaches a point 
corresponding to a suitable working tension in the chain links and then 
keep the pressure as nearly constant as possible. As a rule it will be 
necessary, in order to accomplish this result, for the opening in the valves 
to vary slightly as the chain pays out. Their movement must, of course, 
be reliable and they must close promptly when the strain on the chain 
is entirely relieved. 

It should be noted that the travel of the chain is resisted not only by 
the hydraulic resistance to the motion of the cylinder but also by the 
weight of the cylinder and other moving parts, by the friction of the 
chain at the hawse-pipe casting, as well as by frictional resistance in the 
machinery itself. As will be seen later, it proved entirely feasible to 
measure these supplemental forces accurately. They proved about 
equal in amount to the internal hydraulic resistance, making it nee 
to set the valves which control the pressure in the cylinder, for a much 
lower pressure than originally contemplated. 

ft may be seen by reference to the piping diagram, that, with the 
chain in its normal operating condition across the top of the lock, all 
gate and check valves are closed, so that the resistance valves provide 
the only means by which the pressure can be relic veil. 

In view of the importance of the subject, various types of valves 
that seemed suitable for the purpose were carefully studied, and three 
different designs were finally selected for detailed tests. The first of 
these was a differential piston valve of special design which originated 

333 



in the writer's office and is referred to in the sequel as the I. C. C. Valve. 
It did not prove entirely satisfactory in the tests. The second valve 
was made by the H. Mueller Manufacturing Co. of Decatur, 111., and 
differed from their standard design mainly in the use of bronze for the 
valve body in order to withstand the high pressures. This valve is quite 
simple. There is a disk \\" in diameter with a conical seat. The stem 
is directly controlled by a heavy helical spring, which keeps it from 
rising until the pressure reaches the predetermined amount for which 
the valve has been set. The third valve was built by the Ross Valve 
Manufacturing Co. of Troy, N.Y. from their own designs. It is practi- 
cally identical with the valve used with much success in maintaining a 
uniform pressure in the high pressure fire mains of New York City. The 
main valve has a movable stem with two pistons 6" in diameter, arranged 
so as to be very nearly balanced. There is an auxiliary valve of safety 
spring diaphragm type, which opens at a definite pressure, for which it 
may be set, and also a small needle valve permitting the escape of water. 
The pressure at which the main valve opens depends upon the setting 
of the auxiliary valve and the needle valve. The pressure actually 
maintained at the main valve is usually somewhat higher than that for 
which the auxiliary valve is set. 

The tests were very carefully made, with delicate apparatus, so that 
they may be called laboratory experiments on a large scale, and proved 
of sufficient interest to warrant their publication in some detail. There 
were three series of tests: — 

(1) Preliminary tests on the three valves at a large pumping plant in the 

United States, which provided water under high pressure. 

(2) Tests made on the first fender machine erected in Gatun Lock, the 

chain being put into tension by a large winding engine. 

(3) Actual working tests of one of the Gatun fenders in stopping large 

vessels. 



FIRST SERIES OF TESTS (AT NEWARK, N.J.) 



(1) The first set of tests was made in May 1912 in the power plant of 
the Prudential Life Insurance Co's. building at Newark, N.J. The public 
spirit of the Company in permitting the use of its plant, was much appre- 
ciated, as the tests proved of great value in giving greater confidence in 
this type of fender. 

The arrangement of the apparatus is shown in Fig. 3. Water under 
pressure was supplied by three high-pressure pumps and regulated by 
three accumulators on the discharge lines, the pressure at the accumul- 

334 



ators being about 750 lbs. per sq. in. while the discharge was as much as 
3,400 gal. per minute. A pipe 8" in diameter conveyed the water from the 
accumulators to the resistance valve. Beyond the valve an increaser 
was placed for connection to the 12" Venturi tube. On the high pressure 
side of the resistance valve was placed an 8" diameter gate valve, with 
by-pass, which was generally kept entirely open during the tests. Next 
to this was placed an 8" diameter quick-opening valve also with a by- 
pass. The flow of water was regulated by this last valve. It is believed 
that the throttling of the water in passing through the latter valve was, 
to a great extent, the cause of the reduction of pressure during heavy 
flow, shown in the tests. Beyond the resistance valve the discharge 
pipe lead to a 12" Venturi Meter for measuring the rate of flow. 

By means of three gas-engine indicators, the drums of which were 
arranged to be revolved uniformly by a small electric motor, the pres- 
sures on the high pressure side of the resistance valve, at a point just 
above the Venturi meter, and at its throat, were continuously recorded. 
From the indicator cards thus obtained, the pressures at the resistance 
valve, and the rate of flow at any given instant of time, can be readily 
found. The Venturi Meter had been calibrated very accurately by its 
makers, the Builders Iron Foundry of Providence, R.I. The use of a 
Venturi where the discharge varies so rapidly as in this case, was some- 
what novel. It is believed, however, that the results obtained were 
sufficiently accurate for the purpose. 

I. C. C. VALVE:— 

Seventeen tests were made in all with this valve, a summary of which 
is given in Table I. It will be seen that the valve worked satisfactorily 
for discharges as high as 750 gal. per minute. For greater rates of flow, 
as shown in Tests Nos. 2, 7 and 17, the results were unsatisfactory. The 
plunger moved up and down violently, causing severe vibrations in the 
piping system. 

The unsatisfactory results, under heavy flow, were ascribed to the 
large momentum of the plunger at the time of opening, resulting in 
oscillations and causing water-hammer. 

This type of valve was abandoned as the result of the above tests. 

MUELLER VALVE:— 

A total of twenty tests was made with this valve, the results of which 
are given in Table II. Tests Nos. 1 to 11 gave a rate of flow of 2,000 
gal. per minute, or less. Tests Nos. 12 to 20 gave higher rates of flow. 
The accumulators ran down during test No. 14, showing that the capacity 
of the plant had been reached. In these tests the valve gave results 
that were satisfactory in every respect, proving itself capable of reducing 
pressures from 550 lbs. to about zero under a flow of 3,000 gal. per minute. 

335 





W 








> 








K-l 








<J 








> 


c4 






w 


1— 1 






u 


CT> 






5? 


„ 






•< 


o 






H 

02 


>> 






i— i 
CO 


£ 


co 






d 

o 


bU 




d 




.3 






o 


~ 




o 


C5 


0) 








CJ 














co 


Tl 


1— 1 


w 

H 


CD 

EH 
-a 

a 


CI 

— 

co 

d 


(-5 


co 
H 


CM 


CD 


< 


co 


OS 


CD 


W 


T— 1 


bl 


H 


H 


oT 


S 












U 

i— ( 


>» 

rt 


S-l 




3 


§ 


S-l 




fc 


T3 






^ 


o3 


0) 




Q 


CD 


o 




O 


t- 






Cm 


£ 






« 


oo 






>*■ 


o 






W 








ta 








o 


CO 








CD 






hJ 








P 








CO 








W 








rt 











CO 

M 

■~ 

S 
CD 




CO 

d 

+3 
- 












f CD 

•J § 

S"8 


73 
t- 
O 
o 

CD 

o 


- 


d 
US 

r— 1 
4< 


d 

3 


CO 

CO 

co 
4 


^b 




d 

• S cc 

+J CD 

P ° 






cj 

CD 
co 

00 
CM 




co 
co 


CM 

CO 




Flow 
Gallons 
per min. 


o 
o 


3 

'fcH 

> 


O 

o 


o 


o 

o 

CD 


© 

CD 




V. 

— 

- 

■1. 

/. 

■~ 
S-. 
Oh 

-4-3 

a 

CD 

> 


Difference 
lbs. pr sq. in. 
or (lbs. pr D ") 


co 

d 


CD 

3 

> 


CM 

d 


CM 

d 


CO 

d 


"5 

co 

d 




CO ! — ' 

a s 


d 


Variable 

o +5.5 
o -3 


Mi 

d 


d 


o 


d 




d;T 

*^> a 1 <- 

CD 33 Oi 
CO -— ' 

±: o 


d 


Variable 

+4.5 t 
-3.5 t 


>o 
co 

d 


CO 

d 


d 


d 




Water Pressure at 

Indicator A 

Lbs. per sq. in. 


o 
o 
r- 

+j 

O 


© 

CD 
OS 

oj. a 

'(H ^ 

Oj .. 


o 

o 


o 

00 
CD 


CD 


o 
oo 

CD 






to O 
CD >jr 


- 


CM 


co 


!«* 


"O 


CD 



336 



00 

c 
_c 

+~ 

83 

u 
> 


















3 

>: 

_c 


cc 
c 
_g 

h 

> 


E~ 

is 

S 1 * 


o 


CO 

CC 

z 


cc 

Z 


ib 


"5 


Tb 


To 


iO 


«o 


© 


u 

00 


CS 


© 

T— 1 


© 


3 

© 


© 


3 

o 


»o 


a 


cc 

0> 


3 

-- 

CI 


— 

- z 


"3 
E 

00 

>> 

u 

> 


« 


"5 

E 
co 

>> 

u 
o 

> 


3 


O 

3 . 


s 


8 




© 

25 








c 


CO 

>> 

«- 

> 


n 

E 

>> 

t- 
o 

> 


© 


U3 

d 


•o 

d 


33 

E 

CO 

>> 
o 

> 


d 


43 

2 

.2 

> 


— 

99 

r. 
> 




© 


o 
d 


o 
d 


o 
d 






© 


W3 

d 


•O 

— 

d 


iO 
iO 

d 


£ 

■a j 

a a 


§ 


13 
t- 

o 

u 
Q 

o 


8 


3 


o 
i~ 

CO 


o 

CO 


o 
1 


© 
•o 


© 
1 


© 

o i 

■S E 
■g'E 


t^ 


00 


© 


o 


i— i 


c^ 


cc 


-r 


iO 


© 


t^. 



337 



- 

e 
< 



Eh 
> 

- 
- 
- 
- 
- 



m 
H 

Eh 
O 



P 

o 
- 
- 

= 
o 

»2 
H 
1-5 

P 

- 
- 



2 bfi 

2 S 



3 <3 



o 























'0 


















*^ d 


















?3 






^ 


^ 










— -+^ 


















. GO 






CM 


vr 










^S 


O 

3 


3 


co 






oc 


00 


00 


s 


O 




t— 1 












G 
















., , 






















































c 


















— 




u 














C^H 






02 


a> 




1-0 


-* 


•HH 


<M 


3 




CM 














P 




















bfi 

■SI 














ci 




£ 


3r1 














V 




O 


P H 














to 




5 














00 










































a 




















s3 


a.a 














O 




gj 


a 














O 

00 


"3 

s 

02 




. , , 




















3 ; 




















u — a 


































"5 






D - / 

— 3- - 














© 
























CO 
03 


J 3? 5 


















3 

CO 

-x. 






3 




n 











, 


s- 


r. •— 


5c 




3 
CO 






~a 




"3 


Ph 




* 


3 


« 


E 




£ 




- S W 

5 — £ 
a — ' 


CO 

55 











f. 




02 




= 5 






> 


3 


■* 









£h 




















3 


= ; 


















-*J 


.□ 


















3 

> 


- 5; — 
= '- s 

— 1 3..Q 














































































"2 . 




















32 3 






































42 


















" 


















Shjj-9 








O 


O 











i fc- cr 










<m 


1^ 




s 


s 


r. C x 








3 





O 











Pre 

licat 
per 


1^. 


re 


ro 


-1-2 


-u 


-1-3 


-1-3 


-1-3 








O 

O 


8 




LO 


O 




CO 


>- 3 • 








t^ 


t^ 


O 





CO 


X — ■ 


















- — -1 


















o9 — 


















£ 


















0D ^ 


1— 1 


CM 


rt 


•* 


10 


O 


t^ 


00 




Eh£ 



















338 



+ 




| 


1 










1 


+ 


+ 


+ 


- 


J 


J 


+ 




01 


- 




Oir 


ei 


Ol 


CO 


tN 


CN 


CO 






co 


CO 


CO 


CO 


co 


co 


^H 


CO 


co 




qo 


CN 


CO 


^ 


^ 


:t 


co 


CO 


eo 


t^ 


C5 


iO 


co 


™ 




l^ 


t>- 








CJ 


3 


3 








3 












00 
























tN 


CN 


O 


o 


- 


OJ 


CO 


o 


oo 


CO 


CO 


00 


*™ ' 


~ 








~H 














6 
















































oa 










«o 














«o 


kO 


in 


Ifl 


■- 


lOCO^H 


CO CN 


•* CO 


■^ co 


■^ co 


re ?i 


lOrf CO 


O 


c 


© 


o 


o 


s 


C: a 


3C = 


a g 


Q Q 


g 3 




- 




§ 


o 


5 


c — x 





3 ~ 


g 5 


5 5 


C: — O 


/ 






CO 


BQQCN 


I- O 


re — 


?l DO 


H CN 


— ei 


i^ — ei 


i-H 




CN 


CN 


ri 


ri re re 


•M CO 


COCO 


coco 


co co 


co co 


ci co co 




iC 


«o 








lO 


»o 


«o 




iO 


lOlO 


co 


:: 


ec 


•* 


■~ 


N00O 


CO 00 




~ 35 


coo 


0OO3 


CO 00 CS 




iO 


w 




■ -. 


i-O lO o 




o 


•O 




lO 


»0 LO 


CN 
1 


1 


CN 

1 


. 1 


~ 


IG EON 

1 1 1 


IC EC 

1 1 


EC I- 

1 1 


-o 1- 

1 1 


•- 1- 
1 1 


TT 


I 1 1 










kC 


tQ IC lO 


■- 




'- IC 




















— N 


coco 


ri ei 


CN 71 


Tl Tl 


0) CN CN 


O 


- 


- 




= 


CJ 


= 












35 


• - 


•o 




— 


t- 


■- 












o 


■- 


CO 




CO 


■ - 


lO 


















S 


~ 






c 


r 


- 


o 


O 


*J 


o 

5 


o 

s 


3 


— 

- 

re 




o 
o 
as 


■3 


V3 


to 


>o 


C5 


— 


B 


^_ 


ei 


re 


— 


■ 


-o 


r- 


CO 


- 


ei 












» 













o 

— 



339 



rr en 

^ >> 

CO ,5 

CO § 

k ^ S3 

Eh o . 

H ^ M 



CO 

EH 

co 

W 

So 
w 

Eh 



(* 
P 

O 

p 
W 

Eh 

o 

co 
H 
.J 

P 
co 
W 



».s 



cd r 

^1 
73-0 
C oo 
o3 co 
~ o3 

S « 

C 



<M OJ 

.03 oj 



0J<! 
73_ 
o3 o3 

C *- 
S CD 



is 







co 
































^ 
































Ih 
































o3 
































s 
































0J 
































£ 
































c , 
































.S1S 
































•+J <D 
































g^ 
































P*ta 
































P° 






. 




























60 -, 


































C OJ 

•i a 




























CD 


a 


o 


P H 




























CO 
CM 


co 


£ 


































o 




-H 






























03 


3.9 


*o3 

a 






o 


a 








o 


O 


-3 


© 


o 


o 


- 


M 


CO 






»o 


e3 








lO 


o 


o3 


lO 


iO 


o 


5 








fr- 


a 

CO 








fr- 


CI 


a 

CO 


o 


o 


oo 


i^ 




c3 s- 


OJ 


















1—1 




*"" ' 


*""' 






> 


































































(B Cv 


































0- _ D 


































oj C7 q_ 








lO 










lO 


fr- 




o 


o 


o 


iO 


GO 


53 G.JD 








© 










© 


© 




»— 1 


.— 1 


CO* 


-]' 


































00 


P 22 X 

£ o 


































.sr 
































Ph 


*- .D 

c3 o* fa 

H £ 

CO V* 

5S 








JH 
























't- 








3 

03 


































3 


g; 
















3 
















-*^» 










tn 
























a 


' □ 








03 
























OJ 

> 


oj co ►* 

3 ^ 5£ 

co v" 

a 5 








>■ 


































o 


o 


o 


o 


o 


o 


o 




o 






c3-< . 










§8 


oo 


co 


oo 


56 


oo 


o 




o 






.. . C7 










CD 


CD 


CD 


CD 


CD 


fr- 


t>. 


fr- 






Pressure 
ndicatoi 
bs. per s 


o 

CD 
O 




O 
fr. 


O 
CD 


o 

-*3 


O 


O 


O 


O 


O 


o 


o 

+5 


o 


§ 

CO 


- 
:0 










o 


O 


O 


o 


O 


o 


o 


o 


o 














CD 


fr- 

CD 


fr- 
ee 


fr- 
ee 


CD 


fr- 
ee 


s 


co 

CD 


s 






M hJ 
































co o 


iH 


(M 


CO 


-C* 


>o 


CO 


fr- 


CO 


OS 


o 


1—1 


<M 


CO 


TJ< 


Ifl 




E 


Sfc 

































340 



i 



cico 



W w 
weo 



3 3 



§8 

- 1 n 



o 

1 - — 



o W 



S! 



o w 



55 : 



— — ■ 

t> W<=5 



CO <M -< CO !M MM 



ri 



wow 



W W 
W I- 






3 3 

CO CI 



© w 

X 90 



WW 



©© 



o w 

CJ -H 



o wo 

•M -h ^h 



WW © w 

i-i r-i 



£ 



S 



S 



:ui 



Typical indicator diagrams from Test No. 16 have been reproduced 
in Figs. 4, 5 and 6. 

ROSS VALVE: 

A total of twenty-seven tests was made with this valve, the results 
of which are given in Table III. Tests Nos. 1 to 18 gave rates of flow 
of 2,000 gal. per minute, or less. Tests Nos. 19 to 27 gave higher rates 
of flow. 

Indicator diagrams from Test No. 23, are reproduced in Fig. 7, 
8 and 9 as typical of these tests. 

The action of this valve was also satisfactory in every respect. 

It will be noted that in this, as in the other tests, there was a decided 
drop in the pressures maintained, in the larger as compared with the 
smaller discharges. 

As the results of these tests made at Newark, four valves, two, of 
the Ross and two of the Mueller type were purchased for further trial 
on the Isthmus. 

SECOND SERIES OF TESTS (MADE IN GATUN LOCK) 

These tests were made in the winter of 1913 on the first fender erected 
on the Isthmus, which was in the upper approach to the Gatun locks. 
One Ross and one Mueller valve had been attached to each machine. 
Only one valve was used at a time, the other being entirely shut off 
from the piping. At this time there was no water in the locks. It was 
intended that these tests should be made, as nearly as possible, under 
the conditions that would prevail with the fenders in actual operation. 
The chain had been connected to the machines on both the walls and 
was stretched across the lock at the top. It was proposed to fasten 
a cable to the chain at a point half way across the chamber wall, and 
attach the other end to a large winding engine on one of the walls. In 
this way, a pull along the axis of the lock would have been exerted, the 
effect being similar to the thrust of a vessel striking the chain. It 
proved necessary to modify this programme slightly. The middle 
section of the chain was detached and the cable connected directly to 
the end of the chain where it emerged from the hawse-pipe on one of the 
walls — . The arrangement is shown in Figs. 10, 12 and 14. In 
the first series of tests (Fig. 10) the cable made an angle of 12 l A° with 
the wall so that there was friction between the chain and the hawse- 
pipe casting. In the second Series (Fig. 12) a snatch block was fastened 
to the opposite wall so that the cable made an angle of nearly 90° with 
the wall and did not touch the hawse-pipe. In the third arrangement 
(Fig. 14), a second block was added in order to increase the pull on the 
chain. 

342 



For these tests a Lidgerwood Unloader, consisting essentially of 
a 60 ton winding engine mounted on a flal car, and supplied with steam 
from a locomotive, was placed on the wall opposite the fender machine 
to be tested, and about 600 ft. from the same; the tension in the chain 
was produced by winding in on the unloader, thus causing the moving 
cylinder of the fender machine to tend to rise, producing a pressure in 
the upper cylinder. The amount of pressure depended on the setting of 
the resistance valves. The cylinder pressures were shown by gauges 
and continuously recorded by the indicators which were used in the tests 
made in the United States. The valves were set, by trial, for gradually 
increasing pressures. 

A number of runs were made with the arrangement shown in Fig. 
10, the pressure varying from 170 to 350 lbs. per sq. in. When the pressure 
exceeded 350 lbs. the unloader was apparently unable to overcome 
the hydraulic resistance and the hawse-pipe friction and internal resis- 
tances in the machinery. 

A typical run is shown in Fig. 11. 

After the valves had been adjusted, the pressure curves were very 
uniform for both types of valve, with practically constant pressures 
throughout the stroke, except for the small oscillations due to chain 
friction on the hawse-pipe. The plunger speed varied from 6 to 25 ft. 
per minute (equivalent, to flows of 350 to 1470 gal. per minute), being 
limited by the capacity of the unloader. The low speeds correspond 
to the highest pressures. 

Four runs were made with the second arrangement, when the cable 
parted owing to imperfections. The pressures ranged from 310 to 370 
lbs. per sq. in., as shown on the typical card, Fig. 13. The pressures 
were perfectly constant and steady, without any of the small variations 
due to hawse-pipe friction. 

For the third set of tests there was a three fold multiplication of the 
rope (Fig. 14), the pull being again nearly normal to the walls. A series 
of runs was made, using the two valves alternately, with pressures running 
up to 550 lbs. per sq. in. at plunger speeds as high as 8.4 ft. per minute. 
In the final test a maximum pressure of 630 lbs. was reached when the 
chain broke in a flaw. Fig. 15 shows a typical card. 

The pressure curves obtained with both valves, in the last series 
of tests, were also entirely uniform and satisfactory. In these tests, 
as in those previously made in the United States, both types of valve 
gave equally good results. The choice became a difficult one. The 
simplicity of the Mueller Valve was in its favor. The Ross valve appear- 
ed, however, to have two advantages. In the first place, it is more 
readily set for the desired pressure, as the auxiliary valve is controlled 
by very small springs. In the second place, it was possible to connect 
the auxiliary valve by a small pipe directly with the large machine 
cylinder, and thus control directly the pressure in the cylinder itself, 

343 



and not the pressure in the piping close to the valve. In this way, the 
variable drop in pressure between the cylinder and the point in the 
piping at which the valve is attached, was entirely eliminated. For 
these reasons, the Ross Valve was finally selected for use in all the 
fenders. 



THIRD SERIES OF TESTS (.MADE IX GATUX LOCK) 
TO DETERMIXE FRICTIOXAL RESISTAXCES 



In order to determine the most suitable pressure for setting the 
valves, a further set of experiments was made in February 1914. Their 
purpose was to measure the friction of the chain on the hawse-pipe and 
the frictional and other resistances in the machinery. The arrangement 
of the apparatus is shown in Fig. 16. It differed from that previously 
used by the addition of an hydraulic dynamometer, for measuring the 
strain in the wire rope, close to the winding engine. This dynamometer 
was compared with a standard spring dynamometer and found to be 
practically frictionless and to give correct results. Indicator cards 
were taken simultaneously at this dynamometer and at the upper cylin- 
der of the fender machine. From these the effect of friction, etc. was 
readily determined. 

Two series of tests were made. In Series I, the chain pull was at 
90° with the lockwall, so that there was no friction at the hawse-pipe. 
In Series II, the pull was at 20-5 with the wall so that the total resistance 
included hawse-pipe friction. 

The observations are plotted in Figs. 17 and 18 in which the abscissas 
and ordinates are respectively the pressure (P) in the machine cylinder 
and (p) at the dynamometer, both in lbs. per sq. in. 

There is a four-fold multiplication between the movement of the 
machinery cylinder and the travel of the chain, and a further three- 
fold multiplication in the wire cable, while the cross sections of the 
machine and dynamometer cylinders are respectively 1134 and 380 sq. in. 
Hence, if friction and the weight of the moving cylinder be entirely 
disregarded, the corresponding values of p and P whould be given by 
the equation: 



1134 

p= xP = 0.25P 

1 3 x 4 x 380 



In Fig. 17 the simultaneous readings in the two cylinders, when 
there was no hawse-pipe friction, were plotted. It was found that a 
straight line could readily be drawn, which fairly represented the actual 

344 



readings, and was at the same time parallel to the line of no friction 
p = 0.25 P. This showed that the sum of the internal machine friction, 
and the weight of the moving parts is constant for all pressures. It is 
represented by the difference between the ordinates of the two lines and 
found to be equal to 30 lbs. per sq. in. in the dynamometer, corresponding 
to a pull of 3 x 380 x 30=34,200 lbs. in the chain. 

The second series of measurements is plotted in Fig. 18. The 
observations cannot be as closely represented by a linear equation as 
in the previous case, though the line drawn in the figure gives a fair 
average. The difference between the ordinates of this line and the 
upper line in Fig. 17 represents the effect of the hawse-pipe friction. 
Its effect on the pressure at the dynamometer is p = 0.20 P. 

Taking hawse-pipe friction, and all other resistances, into account, 
we have: 

p = 0.25 P+30+0.20 P = 0.45 P+30. 

If now T x and T., represent the pull on the chain inside and outside 
the hawse-pipe respectively, we readily obtain from the previous 
equations: 

1 191 

T l = -^P4-34,200 = 283.5 P+34,200. 
4 

and 

= 3 x 380 p= 1140 p = 513 P+34,200 

while their difference — 

T..—T 1 =229.5 P represents the effect of hawse-pipe friction. 

As noted above, series II gives rather irregular results for the value 
of p. This was doubtless due in part to the wear of the chain and hawse- 
pipe during the tests. Flat places were worn on the side of the chain 
links as wide as 2\" and as long as 7". and on the hawse-pipe, 7" 
wide and 40" long, the maximum depth in both cases being about \". 
This wear of the chain and hawse-pipe evidently increased the friction 
as is shown on plotting the values of p, which were larger for the later 
observations. 

It becomes still more apparent on computing the coefficient of friction. 
This can be done by the formula: 

I T fu ^ T 

f ='° g (^j*1ir X2 - 73 - 1O8 Tr 489 

in which f = coeff. of friction. 

345 



64.5 = arc of contact of chain around hawse-pipe, while the log. 
is a common or Briggsian logarithm (See Unwin Machine Design, New 
Edition, Part I p. 447). The coefficients are plotted in Fig. 19. The 
higher values of f in the later experiments become quite clear on inspection. 

The principal purpose of the foregoing tests was to determine the 
proper pressure for the setting of the resistance valves. 

For a given cylinder pressure, as shown above, the maximum pull 
on the chain; i.e., the pull outside of the hawse-pipe, is 

T.=513P + 34,200 

The proper value to be used for T 2 depends on the strength of the 
chain; as shown by test. 

The minimum breaking strength, permitted for any link, was 400,000 
lbs. and all chains had also to stand a proof test of 245,000 lbs. without 
sign of failure. 

As there will also be some secondary stresses due to bending of 
the chain around the sheaves and to other causes, it seemed best to 
limit the working stress to 220,000 lbs. For this value of T ; we have — 

„ m 34200 220,000—34200 

P = T.. = — = 362 

" 513 513 

It was therefore decided to fix the cylinder pressure at 360 lbs. per sq. in. 

At this pressure, of the total chain pull of 220,000 lbs., I l x 360) or 

104,120 lbs. is due to the cylinder pressure i.e., only about one-half the 
total stress, the rest being the result of the weight of the moving parts, 
the internal resistances of the machinery and the friction at the hawse- 
pipe. 



FOURTH SERIES OF TESTS (MADE IN GATUN LOCK) 
TESTS IN STOPPING VESSELS 

The foregoing tests gave a reasonable assurance that the fenders 
would function properly in stopping vessels. Twenty-two of the fenders 
were therefore built, practically identical in plan, while two others 
(in the lower approach to Miraflores Lock) differ only in having two 
chains stretched across the lock at different levels, to provide for the 
great difference (22 ft.) between high and low tides in the Pacific. Their 
machinery is absolutely identical with that in the other fenders, the high 
and low level chains being alternately connected and detached, as the 
tide changes. 

346 



It was, of course, desirable, to make an actual test of the fenders 
in checking a vessel in the lock. 

In October and November 1915, after the writer had left the Isthmus, 
a number of such testa were therefore made by a Board appointed 
by the Governor of the Canal. They proved of great interest and value 
especially as the vessels were of considerable size. Two ships were 
used, the "Allianca" having at the time a displacement of 4,221 tons, 
and moving at speeds varying from 1.23 to 3.38 miles per hour, and 
the "Cristobal" with a displacement of 18,000 tons and speeds as high 
as 2.45 miles per hour. 

The resistance valves were set to open at 360 lbs. per sq. in. in most 
of the tests, and the propellers of the vessels were stopped in every 
case before the chain was struck. Indicators were connected to the 
piping system in both the machinery rooms (Nos. 810 and 811 — Gatun 
lock), and the pressures and also the travel of the moving cylinder of 
the machines were automatically recorded. 

A rope mat was woven around the central portion of the chain and 
a similar protection given to the stem of the ship. The Photograph PI. 
3 shows the "Cristobal"' approaching the fender. No damage to the 
ships occurred as a result of the tests, and the chain was marred only 
very slightly. Twelve runs were made with the "Allianca" and ten 
with the "Cristobal", and the vessel was brought to a stop in every 
case before the chain had paid out to its extreme limit. 

The tests with the "Allianca" were not entirely satisfactory, as 
the resistance valves had not been cleaned for a long time, and there 
was a slight sticking of the valves which prevented them from closing 
promptly when the pressure was reduced. All the valves were, there- 
fore, thoroughly cleaned, a new leather placed in one valve, and the 
other leathers softened up. 

The tests with the "Cristobal" were made after these changes were 
made and proved entirely satisfactory. The indicator cards taken 
with this vessel are shown in Figs. 20, 21, 22 and 23, which also give other 
information in connection with these tests. 

The pressure curves have a decided peak at the beginning, which 
is in every case decidedly above the setting of the resistance valve. 
Beyond this point, and throughout the greater part of the stroke the 
pressure remains remarkably uniform, with very few oscillations. The 
vessel was brought to rest from 51.5 to 62.0 ft. beyond the center of the 
fender, its speed when striking the chain being from 2.06 
to 2.45 miles per hour. 

There was little difference between the pressures in the machines 
on the two lockwalls or in the length of chain paid out from each side. 
The travel of the cylinders was hardly over 6 ft. out of a total possible 
stroke of 21.5 ft. 

347 



The distance travelled by the ship before being stopped was less 
than the shortest distance from any of the fenders to the gate it is intend- 
ed to safeguard, so that there seems to be every assurance that the 
fenders, if ever called upon, would fulfill their purpose, even in the case 
of a ship as large as the "Cristobal" and moving at a speed of over 2 
miles per hour. As this vessel is about 500 ft. long, and 58 ft. wide, 
few larger ships are likely to use the Canal, nor is the speed of two miles 
likely to be exceeded in the approaches or the locks. 

It is a matter of considerable interest to note that the distance 
in which the "Cristobal" was stopped agrees very closely with the 
theoretical curves which were computed before the designs were complet- 
ed, but after the working stress of 220,000 lbs. had been adopted for 
the chain in stopping vessels. These curves are shown on PL 2, which 
is copied from the Annual Report of the Isthmian Canal Commission 
for 1911. This close agreement with theory is, of course, very satis- 
factory. 

Accounts of the various tests, though in less detail, are given in 
the Annual Reports for 1913, 1914 and 1916. 

DISCUSSION 

Mr. R. DeL. French, Assoc. Mem., Can. Soc. C.E. — The most inter- 
esting part of this paper, to the writer, is the account of the tests on the 
resistance valves carried out by Mr. Goldmark at Newark. 

As he remarks, the use of a Venturi meter for measuring such rapidly 
fluctuating flow is somewhat novel. The Builders Iron Foundry, in their 
bulletins on the use of the Venturi, state that the meter should be preceded 
by a straight pipe at least six times as long as the diameter of the meter 
tube. This condition does not seem to have been complied with in Mr. 
Goldmark's experiments. However, the makers are prepared to issue 
special instructions for cases where this requirement cannot be met, and 
it may be that it was taken into account in the "accurate calibration" 
referred to. 

From the description of the "resistance valves," they seem to be 
practically identical with the reducing valves used in many water works 
systems. Little data is available regarding the loss of head through 
these valves under different conditions of flow, and it would be of great 
interest to water works engineers in general if Mr. Goldmark were able to 
furnish some information on this point. The apparatus shown in Fig. 3 
is well fitted for loss-of-head experiments, requiring only the addition of two 
pressure gauges, one on each side, close to the valve under test. Possibly 
Mr. Goldmark made some observations of this sort, and has purposely 
omitted them from his paper as not relevent to the subject. 

Mr. Goldmark is to be congratulated on the close agreement between 
the expected and the actual results, as shown by the tests with the 
"Allianca" and "Cristobal," particularly as there was no precedent to 
help in the solution of the problem. 

348 



W* Criiuoma. . 




10WEHINO CHPIN. 



1 


o~. 


«< 


c 


rilMO£A. 

MM 


1 








































































•- 


' 




































: 




















































N 




} 




' 


















































— 


100 ' 


























































i So • 

V 


f #r***Jp***r 


■ ^MJJVAA iait 










































1 *• 




X 










































1 


- 












N 



Lowm/Nc ch/j/n. 



•so sa 



r/Q.i 



76JTS Of lOWEmNQ £~ A.ff/J/M$ CttfilN BT faTM . 
/.owe*. Cri'Moe*.. 




up?*,*. Cn./MO*A. ■ 



+* J* j: 

Aff/J/NO Ctfff/H 

A \- 



~3£CO"£>S . 




ZO z* ^ Z _ȣ 

A0/J/MG CHff/H . 



so se 



F/G.Z 



349 



1 

I 



5 










ffat 


W£ > 


Q£Jt 


15 7-/9 


/VC£ 


„ -ISO , 


























-1 


■ — - 1 












-«*~ 


















/ 








"S^ 


















/ 














^1 






































ffrpmjf***/c 






















V 








/*ASJt*l*jU*M 




I 
















^_ 
































'1*1 

































a n it /» f 7 £ ■£ ■* -* z ' 



Tusrs or AfuELLE.fi. VffLVE ffr N£.wag.K.H 'J . 



F/G 4 



VLNTUn.1 INLET. 




351 



T£JTJ OF &0JS V/7LV£ /7T N£.Wfffi./L J M.J . 
J?AdY£ /t!£J/Jr/7//C£ V/7/.VE . 



f#£SSVA£ HUB. 



z 



y izs 



600 * 



-*»o 



t 



ZM 5 



/* 9 



&G-7- 



6 7 6 <S 4 3 Z / 

S£CONJ>S. 



VENTOA/ /NLET. 



0r/*fosprtE.*/c 
f/ixjiitac Line. 



sm^ 



/o 9 8 7 s & 4 




F/Q. 8. 



VEMTl/R/ THAOffT. 



ArTAtOSfWf/lIC 




352 



CO To*, ia^lAWQOD l/*i.oaO£JL 



TESTS ffT QffTUN.JffNFt5.HIS . 



cuatM rtNDt a Mm J 



-f^ce. Of tot* u/#u. , 



/V'« 



-5*<* 



F/0 JO. 



ChPIH f INDIA Hh 





-+- 






















































! 






























































i< M ^ 


U| 




K^ 


■'A' 


M 


w 


\ 


fl 


M 


Wl| 


HI 


*w 


ih 


tfli 


l| 


id 


'r* 


>w 


tm 


nil 


irii 


k - 












































• 


,, 












































































_ 




£"• 




44 


K • 




>» 


. 














































\T~ 


~ 


" 



















































•s« -Jo ** j« j2 r< zo /^ g z 

F/{f- JJ. -J* COM2 J . 



Vj 



- \ 



GQ TQM I O^SAh/OOO tSHLO*QL/L 



TESTS ffT (fffTUhf . JffN Fid. /f/3 . 



Chq/h renom Am 



£ or itiMt. 




r/Q./z. 



li 



'W, 



fAAUU* I L'** 






I 
46 Jo 41 Je JZ. 2* 20 



r/Q/3. 



C»*tu f*HO£A Am. 



_- S»o 

X 
Jco ■? 



. . _ mo 



353 



*° n " *"*" wo °° 4 "™* TEST 3 >r CRT UN. JM Kb. If 13 • 



£ af loc hs- 



r/Q.14. 






C*JV,AT f4+>0*. A *A*\ 




Cm mm ftvoeA Am 



->*■«* 




/oo "^ 



T£JTS J7T ,^r *-•*««« QA 'TUN. F£3./fA£ . 




/"Vf""~- vwsifipi 

LMr~ t4 ' 30 ' 

MIL tlOUMD HffWJtflPt 



fVtL Lmt Junes Ifi/iL fo'r*Licx w*t.L. 






J7QJ7* 




354 






1 

i //4<. 



% 



J SJooo 



.&*. tfC't 1MB**. ,J8#*u - HJ*jt.M 



T* iicK VmifM '»iirtf t OCTIHIj) 







* Ff$J8. 



■se ?s /eo /Z5 /so ips- 200 zzs iro z/sr 

-I i * * I 

3 , C * & I 



355 



TE JZJ AT CfffTCW. FE5. tfjf . 



«j .4c 
\ 



X .5o 



< 



.3>< 



i; .Z0 



« ./5 



X 






























V 












/ 


o 


X 












/ 




\ 












/ 




\ 












/ 




« 


k 




^r.»7v /< 


1RD£ fSB 


Z*" -' 


' 






\ 




Ixcepr. ff.e.*//fff£ ft, 


s./^ V 








\ 




0V£*/tf£ 


/-* 


/ 

/ 








\ 






















































/" 










I 






/ 


O 




o 




1 






( < 


1 


( 


. 




1 






\ 
















\ 




f , 




s < 


I 






* 


V 
X 


lie 


> 










































































































1 


' < 


i 




\ 








1 
1 




o 


o 


\ 

\ T 


TSrj MR Ok 


i FEB./S 




1 


» 







S 4 


<£A#G£.y 


= .Z? 




















\ 




\ 

\ < 


1 


/ 
/ 












N 


















































"*" 











SO 100 JSO ZOO ZSO JOO JSO jC&S./**3+t**. 

P~ Crz./"0£A Paesjoas 

Pac of CoNracr ffr Htrwjtpipc. «= €4°-30' 

C0£ff. Of fA,CT,0«f- /Of.?? +*$£ 2-73 ~/o f .r±~. f8f 

73. «» c/f/rtu fojLL ours toe. qp 4f#t*s*M/H&— //4op . 

T/ m it „ /HS/0£. Of * - » Z&3.5/+34 / zoo 

P =r J>yAf#»TO»TEr/;A, />A£JSl/A£ VJ A£<f/Jr£*.£D . 

P — Cyz/Afoe/L. « « * 



f/q.fZ 



356 




F/Q. ZO. 



T£Srj ffTGBTUN , N0VJ4I5 . 





•SSC O AT D3. 
.5" /o is- 2o 2C Jo 


\ 




i 

s 


! 


1 

-- 

s 

1 


\ 
I 


8; 


I 


« : 
}'! 
*-. 

n 

•> 

s 

1 

? 


1 

: 




Sio 














i 

<* 

J/U *// 


| 

//0 J// 


* 


4— \\ 


f 


P*l3t 


VAtS 




_ 


"t' 


•> 


J.. 




IV^" 






^<^~ 






ft 


v 




M 


x*f>" 








,~) 






&> 


































N 




"'" 




























zs 


H 


a**. 


zzc 


&6 


J 


#/ 


s& 


sw 


$a 


KM 


«2 






J*Q 








*y ' 






























\ A *"* 


uaU 


y' 










V \ - 


y 








•■ 


r» / 


1 ^ 








" \ 


• 

V 


c 




*>' 








\ 




^' 


























Z4 


"■♦. 


QHfl 


Zll 


SCJ 


5 


Sr 


M 


fit 




a 



/7<?. «?/. 



TE3TJ ffr Qffruu. NOV, /f/S. 



357 





<S~ /J .<& lc X€T -JO 


% 

<* 


14 

K 

3 




1 


es 

n 

1 

K 
-J 

"5 


8 


K 
•v. 


I 


* 

a 

i 


g 

1 
















s.is' 


s 

g/0 X/l 


f 

810 Si/ 


* 


*. 1 


f 




/ve«/ 


IC/A&S 




,"' 




ij 


J °° 1 


r 




^A/\j 








*: 


Z06 1 














\ 

SI 


-£zJ 




^ 










\ 


s 






















a 


,.-- 


- 
























etz 


zs 


* 


tot 


*# 


sw 


^ 


as 


442&2BSzsxan 


IZ 






^■ssa 








s' 




\ 


\ 






















* 

$ 


4c e 


A A 


puts 


■SVKtS 


•' 






300 I 


f V 












3 


Zoo 1 




y< 


o- 








8 


i^ 




°y 










>s 


a 


„-" 












■M 
















U 


H 




Zfss/s 


j 


Ml 


S.SS 


530 


5.45 HD.ll 


It 



F/Q.Z2. 



TEST5 ffTGffTC/N. NOV.W5 . 





^r /0 /S ZO 2S 30 




s 


5 


•a 

1 


S4 

>M 

R 


I 






Si 

<*> 
S 

>< 


I 














_,. 


_JJA_\_ 


v4 

a 

R 

Aft V? 


f 

stio si/ 


i 




flo 






y 


' 






3<x> j 




PUIS Si 


nil 








J 








1 






,1* 


?"' 








« 






&' 












\ 




















<t 




,- 
































V 


"/% 




Zjc 


SIS 


3 


Hi 


6/9 


iSB 


4so 


ma 


IZ 














^ . 


£*!■'— 






























t «to 






, 


"' 




J 




\Kr 


/>A£* 


rvxts. 


,'' 






? 


Zoo 


IV 


1** 


gs' 










^n 




■&■''' 










•> 


Zz 






























zt s 


"/% 


St0« 


ZJO 


HO 


J 


£a 


HI 


410 


SID 


MM 


IZ 



F/G. 23. 



T£JTJ ffT Cr (7 TON. NOV l<?/5 - 



358 




/'LATE 3. 






7 



<; ^/^*f^ 







^ W^"* ' ^ 







PUAN. 



Typical Pit tor Fenders at Approaches 




X-Xr^'.-LWKjvi*- 




mr&pxs 



Elevation 

360 




ton 1>iru Cylinder . 




Section thus Cylinder . 




^POO 



to.ooo 



1 1,000 



I2POO 



Note - Th» ealculotic 
the- fo'. !■■■ 

The chain is stretched acn 

lock ond , when struck by o vessel, pays 
Out uncter a constont tension of IOO 
qro&S. tons. I*s resistnn-e is the only 
force tendmq to chonge the speod of 

the - esse! .i f '^ H strikes the chain 



The abscissas of the points 20", 25' 
ergy absorbed by 
the chom , poyln3 out eaual \y from each 
wall, offer the v«.»»*| hat, Iravelod 20', 25' 

The ordinate* to the 
points 20', 25' etc give the speeds fr< 
which the vestels will be stopp*- 
traveling 20, 25 etc 

Thus, a v*ss*l of 5,000 tons »■ 
stopped in TO' if Us fmttnl »pe*d 
5-5 k"ots 




Energy of Vessel 
Foot Tons (Gross) 



NICKEL-COPPER STEEL 

By R. W. LEONARD, M.E.I.C. 

(Rea.l at a Meeting of the Institute on March 28th, 1918) 
Revised to date — August 31st, 1919. 

In the early eighties, during the construction of the Canadian Pacific 
Railway through what is now known as the Sudbury District, some copper 
ores were discovered, and subsequently a Company was formed to develop 
the ore bodies. This Company — The Canadian Copper Company — sent 
its ore or matte to the Orford Copper Company's refinery at Constable 
Hook, N.Y., for treatment, which plant was established for the purpose 
of treating the copper ores mined at Orford Mountain, P.Q. 

When it was realized that these copper ores contained substantial 
quantities of nickel — for which metal there was very little demand at that 
time except for the purpose of making German Silver and for nickel- 
plating — the Canadian Copper Company and the Orford Copper Company 
were merged into the International Nickel Company. This Company 
developed very large properties at Sudbury and greatly stimulated the 
demand for nickel, especially for the purpose of alloying with steel to be 
used for the many purposes so well known to all engineers of the present 
day. 

During the past three years the International Nickel Company has 
constructed a refinery at Port Colborne, Ont., for the purpose of com- 
pleting the process of separation of the nickel from the copper and of refining 
these products in Canada for the Canadian and foreign trade. 

The Mond Nickel Company, of England, also acquired properties in 
the Sudbury District, and ships its partially manufactured product to 
England in the form of matte. 

Latterly, the British-America Nickel Corporation has acquired large 
mineral claims which it is developing, and is erecting in the Sudbury 
District metallurgical works to treat its ores, and at Deschesnes, P.Q., 
an extensive refining plant. 

Although the different Companies are pursuing somewhat different 
processes, in general the operation consists in mining and sorting the ore, 
then roasting it (generally in open heaps or in stalls) to largely eliminate 
the sulphur, and thereafter smelting in the ordinary type of copper-smelting 

361 



furnaces to a matte, consisting of sulphide of iron, nickel and copper, which 
matte is further bessemerized for the purpose of eliminating as much of 
the iron as possible and producing a matte much richer in nickel and copper 
content, and containing therein some small proportion of the precious 
metals. 

Until the Port Colborne plant started operations a few months ago 
these mattes were exported either to England or to the United States 
for refining or separation of the nickel from the copper, and in one or two 
cases for the recovery of the accompanying precious metals. 

The above-outlined process results in the waste of all the sulphur 
content of the ore, amounting to many hundreds of tons of elemental 
sulphur per day, to the serious damage of all plant life in the immediate 
neighborhood. It also results in the waste of some thousands of' tons per 
day of iron in the slags, which until recently was considered a necessary 
waste preparatory to separating the nickel from the copper. 

The Sudbury nickel deposits, as developed by the mining companies 
and as worked out by geologists, consist of an oval saucershaped basin 
about 36 miles in length by 18 miles in width, around the South rim of 
which are located most of the properties of the Canadian Copper Company 
and the Mond Nickel Company, and around the northerly and easterly 
rim of which are principally located the properties of the British-America 
Nickel Corporation. The Towns of Sudbury and Copper Cliff are on 
the south rim. The Canadian Northern Ontario Railway passes in a 
Northerly direction through the Eastern portion of the basin, and the 
Canadian Pacific Railway passes almost through the centre in a north- 
westerly direction. 

An excellent geological map of the deposits accompanies the 
Monograph on the Sudbury Nickel Region, by Dr. A. P. Coleman, 1913. 
This basin is generally conceded to be one of the greatest mineral deposits 
of the world, containing nickel-copper pyrrhotite of unequalled quantity 
and richness, which can be mined and the nickel and copper extracted and 
refined at a cost defying competition. 

It is now the principal source of the world's nickel, and is also the 
source of a considerable amount of platinum and palladium, which are 
recovered as by-products. 

The magnificent Report of the Royal Ontario Nickel Commission, 
1917, which bears on the nickel production of the world, is probably 
one of the most complete and valuable reports on any mineral industry 
extant. 

A number of men have experimented with alloys of nickel, iron and 
copper, commencing with Alexander Parks in England in 1844, who 
patented a "useful alloy of nickel, iron and copper." Hybinette and 
Shulcr, of Sudbury, made experiments about 1902 in the manufacture 

362 



of nickel pig from Sudbury ores, and El A. Sjostedl carried out some 
experiments in the direct manufacture of nickel pig and nickel steel at 
Sault Ste. Marie, Out., for the Lake Superior Corporation about the 
same time. In 1005, Dr. E. ECaanel, Director of the Mines Branch, 
Geological Survey, Ottawa, made, experimentally, some nickel-copper-iron 
pig from roasted pyrrhotite in an electric furnace at Sault Ste. Marie, 
under Canadian Government auspices. W. S. Honey experimented in 
the smelting of nickel-copper-iron ores at Sault Ste. Marie in 1808. 
"Metallurgical and Chemical Engineering" of February, 1013, gives a 
description of the manufacture of nickel steel from nickel pig in an electric 
furnace at Trondjeti, Norway. 

In all these experiments an endeavor was made to select ores in 
which the copper bore the smallest possible proportion to the nickel 
content, it being believed or feared that the copper was an injurious con- 
stituent, except in the case of Shuler, who claimed that the presence of the 
copper in certain proportions was not objectionable. 

Mr. G. II. ('lamer, of Philadelphia, has used Monel Metal (a natural 
alloy of nickel and copper as obtained from the Sudbury ores after elimina- 
tion of the iron and sulphur) in the manufacture of a nickel-copper steel 
which is in commercial use and is said to have been successfully employed 
even in the manufacture of armour piercing shells for the United States 
Government, and the results are reported to have been very satisfactory. 
This nickel-copper steel is also being manufactured into commercial sheets. 

Mr. George M. Colvocoresses, at one time in the employ of the Orford 
Copper Company and who has had a valuable experience in the mining 
and metallurgy of nickel and copper in Canada and Xew Caledonia, 
made laboratory experiments in the production of nickel-copper steel 
direct from the Sudbury ores, and has taken out patents on his process. 
In these patents he claims the direct smelting of Sudbury nickel-copper 
ores or the slags wasted in the present process, either by electricity or 
with fuel, into a nickel-copper pig, and ultimately refining resultant pig 
to Xicu steel, claiming that in this direct smelting process the copper, up 
to a certain proportion, may be considered as taking the place of an equal 
amount of nickel, and that not only is the presence of copper not detri- 
mental, but that, on the contrary, it may be advantageous in that it pro- 
duces, — owing to certain qualities of the copper, — a superior product which 
can be manufactured at much less cost than nickel steel made by the 
ordinary practice of alloying refined nickel with steel in certain definite 
proportions. 

During the past year Mr. Colvocoresses, with some associates, has 
experimented in the manufacture of nickel-copper steel direct from the 
Sudbury ores, for which purpose about 200 tons of ore and 40 tons of 
slag were obtained from the Sudbury District and experiments were 
carried on at the plant of the Canada Cement Company, at East Montreal. 

363 



This ore was roasted in a hand-rabbled furnace and smelted to pig 
in an electric furnace of the Heroult type, and some of it was afterwards 
refined into steel in the same type of furnace, and the balance in an open- 
hearth furnace using producer gas. 

The experiments at Montreal were under the direct supervision 
of Mr. H. A. Morin, who had previously been associated with 
Mr. Colvocoresses in the smelting of Sudbury ores, and I think I cannot 
do better than quote substantially and at some length from Mr. Morin's 
report on the result of these experiments. 

In his report, dated December 7th, 1917, Mr. Morin explains that 
the experiments consisted in desulphurizing iron-nickel-copper sulphide 
ores mined in the Sudbury-Ontario District for the purpose of smelting 
and reducing these ores, with suitable fluxes, and producing an iron-nickel- 
copper pig of homogenous composition which, preferably in its molten 
state, could be refined to a nickel-copper steel, with or without foreign 
ferrous addition, according to the grade of Nicu Steel desired. 

Another experiment was made in the smelting of blast furnace slag, 
which slags are produced in large quantities in the smelting of the Sudbury 
ores (partially roasted) in a blast matting furnace. While these slags will 
produce a pig low in nickel and copper, it is a simple matter to increase the 
nickel-copper content by the proper addition of roasted nickel-copper ore. 

The following is an average analysis of such ores and slags: 

Nickel-Copper Ore Blast Furnace Slag 



Iron 40-50% 

Nickel 3-4 % 

Copper 1-13^% 

Sulphur 25-30% 

Silica 12-20% 

Alumina 3-4 % 

Lime 2-3 % 

Magnesia 1-2 % 



40-45 % 

35-5 % 

25-35 % 

20-30 % 

6-7 % 

2-3 % 

1-2 % 



Balance Oxygen. 

These experiments were carried on in accordance with the description 
given in the Patent papers issued and granted, both in Canada and the 
United States, to Mr. G. M. Colvocoresses. 

ORE SUPPLY 



According to the Royal Ontario Nickel Commission Report, published 
in April 1917, a total of 75,000,000 tons of ore had been developed by the 
Operating Companies up to that time, and the Report further states that 

364 



out of 110 miles of nickel-bearing formation, only about ten miles have 
been developed, partially, by diamond drill, and that consequently it is 
fair to assume that this ten mils of partly developed tarnation is capable 
of further extending the ore bodies within this area. 

The ore secured for these experiments, amounting to about 200 tons, 
was obtained from the Algoma Steel Corporation and was of rather low 
made. It theoretically should have produced a 3 to 3J^% nickel-copper 
steel, but in actual operation the nickel-copper pig was considerably diluted, 
by reason of the fact that the electric furnaces used had built-up banks 
and bottoms of iron which had formed during the previous operation of 
the furnaces in the production of pig from scrap. 

The ore obtained for treatment was mined from one of the properties 
of the Algoma Steel Corporation about 14 years ago and, having been 
exposed to the air and weather during all that time, was decomposed and 
the nickel-copper-sulphur content was considerably leached out. The 
following is a close approximation of the composition of the ore when it 
was first mined and as it is to-day: 

P'reshly Mined Ore Ore Received 

Iron 45% 46% 

Nickel 2.9% 1.35% 

Copper 75% 25% 

Silica 17% 19% 

Sulphur 30% 8% 

SMELTING COPPER CLIFF SLAG TO PIG 

The slag was smelted in the same manner as the roasted ore, having 
a very similar composition and therefore requiring but a slight variation 
in the proportion of the fluxes. It is particularly interesting to note 
that the slags used in this experiment contained 2.2% sulphur, and 
after smelting this slag in an electric furnace the resultant pig contained 
.065% sulphur. 

Mr. Morin reports that the conversion of the pig to steel in the open- 
hearth furnace proved entirely satisfactory, the operation being identical 
with the production of steel from ordinary pig iron. About 70 tons of 
2H% Nicu steel were made. 

Since the experiments at Montreal the Nicu Steel Corporation has 
produced an additional 60 to 70 tons of Nicu pig and steel containing 
approximately 3 l A% combined nickel and copper, at the plant 
of Electric Steel & Engineering Ltd., Welland, Ont., considerable of which 
has been brought into actual use in various industries, the 'desired changes 
in the composition of the steel having been made for the various purposes 
for which it was required. 

The ore was smelted in electric furnaces, after a preliminary roasting, 
and the physical properties in every case have proven very satisfactory 
when compared with the properties of nickel steel, as shewn by the fol- 
lowing report furnished by Mr. Morin : — 

365 



Engineering Standard Committee 
E.S.C. 3% Nickel-Chrome Steel. 

Carbon 0.30% 

Silicon 0.30% 

Manganese 0.60% 

Sulphur 0.04% 

Phosphorus 0.04% 

Nickel 2.75 to 3.50% 

Chromium 0.45to0.75% 



NICU STEEL 
Independent test by Dept. of Mines, 
Ottawa. 

0.28% 
0.014% 
0.58% 
0.038% 
0.005% 
2.16 to 2.62% 
Copper 0.46% 



Physical properties of above steel when 

oil hardened from 820°C and 

tempered at 600 °C. 

Tensile strength. . 100,800 lbs. 

Yield ratio 75,600 " 

Elongation 15% 

Reduction 50% 



Physical properties of above steel when 

heated to 800 °C, quenched in water and 

drawn back at 400 °C. 

Tensile strength. . 101,200 lbs. 

Yield ratio 78,900 " 

Elongation 21.5% 

Reduction 58.8% 



Chemical and Physical properties of Nicu Steel as compared with 
Nic kel Steel 



American Society for Testing Material, 
O.H. Nickel Steel Specification. 



Carbon 0.46% 

Silicon 0.066 

Manganese . 70% 

Phosphorus 0.021% 

Sulphur 0.034 

Nickel 3.36 



Nicu Steel, Independent test by Royal 

Ontario Nickel Comm., Page 415 of 

Report. 

0.43% 

0.30 

0.47% 

0.05 

0.05 

Nickel 2.101 

Copper 1.20J - 3 - 30% 



Physical properties of above steel, rolled 
natural. 

Tensile strength. . 122,000 

Elastic Limit 74,625 

Elongation 16% 

Reduction 34% 



Physical properties of above steel, rolled 
natural. 

Tensile strength.. . 110,400 

Elastic limit 82,600 

Elongation 22% 

Reduction 48% 



" Nicu Steel users comprise the automobile trade, makers of different 
kinds of mining machinery in different forms, such as pistons for pneumatic 
tools, rock drills, etc. A considerable tonnage will, no doubt, enter the 
market as castings, crusher parts, ball mill liners, balls, and many other 
uses of high-grade steel, where strength, toughness and durability are 
required, and in general competition with nickel steel and other simila 
alloys." 

366 



Analysis and Tests of 'Xicu" Stool Specimens: 

'Xicu' Steel A f Nicu' Steel B 

Carbon 0.32 0.31 

Silieon 0.0- 0.02 

Sulphur 0.04.V, 0.0:, 

Phosphorus 0.07 0.003 

Manganese 0.83% 0. < 

Nickel 1 • 1-73% 

Copper 0.27 0.31', 

Soft test of above steel after forging: 

tic Limit, sq. in 66,800 1b. 68,320 1b. 

Tensile Strength, sq. in 90,720 lb. 91,840 lb. 

Elongation on 2 inches 

Reduction of Area 0. 190 0. 190 

Description of Fracture Cupped Silky Cupped Silky 

Hard test after the following heat treatment: 

Heated to 1800 deg. F. Quenched in Oil. Re-heated to 700 deg. F., 
and cooled slowly: 

Elastic Limit, sq. in 92,9601b. 89,600 1b. 

Tensile Strength 113,120 1b. 110,880 11.. 

Elongation on 2 in 

Reduction of Area 0. 150 0. 155 

Description of Fracture H Cupped Silky J 2 Cupped Silky 

Cold Bending Test Showing no crack Showing no crack 

The above Cold Bending Tests were made on Specimen 4.5 in. 
0.75 in. x 0.375 in. Bent cold, through 180 deg. with a Presser of 1.5 in. 
diameter. 

Physical Tests and .Analyses of Xicu Steel Heat Xo. 6, made by 
Dr. Alfred Stansfiold, M.E.I. ('.. McGill University; and comparison with 
Xickel Steel as per Specification for Plates and Shapes, Ontario Xickel 
Commission Report, Page 365. 

Natural Natural 

Xicu Steel Xickel Steel 

Carbon 0.37', 0.45 

Manganese 0.88% 0.70', 

Phosphorus 0.(h 0.04', 

Sulphur 0.047% 0.04', 

Xickel 1 -89% 1 M , . Xickel 3 21 



Copper 0.37%/ 2 26 ' c 

Tensile Stress lbs. per sq. in 96,300 85,000 to 100,000 

Yield Point lbs. per sq. in 56,350 50,000 

Elongation on 9 in. ( j 18.7% 16.2 

Reduction of Area% 36 . 3% -'"><>', 

367 



Njckel Steel Specification in connection with the fabrication of 
the large bridge to span the Mississippi River at Memphis: 

Tensile Strength 85,000 to 100,000 

Elastic limit not less than 50,000 

Elongation in 8 in. not less than. . . . 1,600,000 

T.S. Average 17% 

Reduction of Area not less than 30% 

D 

Nicu Steel produced commercially at the Canada Cement Company's 
Steel plant, East Montreal, and tested by Dr. Alfred Stansfield.M.E.I.C, 
— in comparison with Nickel Steel of similar composition, as given in tabu- 
lated form on Page 387 Marked (C) and Page 416 marked (1) Royal 
Ontario Nickel Commission Report. 

Nicu Steel Nickel Steel Nickel Steel 

Heat No. 6 (C) Page 387 (1) Page 416 

% % % 

Carbon 0.37 0.47 0.47 

Manganese 0.88 0.86 

Nickel 1.891 2.15 2.92 

Copper 0.37J ° .... 

lbs. lbs. lbs. 

Yield Point 52,800 52,000 56,000 

Tensile Stress 96,500 93,000 95,400 

Elongation on 2 in 24 . 3% 24 . 5% 22% 

Reduction of area 50.8% 51.8% 44.6% 

Bending Tests 180° Shewing no Shewing no 

crack crack .... 

E 

Results of the Royal Ontario Nickel Commission, Page 415, Table 3, 
obtained with Nicu Steel and Nickel Steel produced under exactly the 
same conditions during their investigation of the Colvocoresses process. 

Nicu Steel Nickel Steel Nicu Steel 

Heat No. 2 Heat No. 4 Heat No. 6 

Carbon 0.43% 0.53% 0.53% 

Nickel 2.10%) 3.43% 2.45%| 

Copper 1.20%/' 3dO% 0.80%/ <i -^ /o 

lbs. lbs. lbs. 

Elastic Limit 82,600 77,400 80,000 

Tensile Strength 110,400 115,400 111,600 

Elongation % on 2 in ... . 22% 20% 19 . 1% 

Reduction of area% 48% 38.3% 38.3% 

These steels were produced under the direct supervision of Geo. A. 
Guess, Professor of Metallurgy, University of Toronto, in conjunction with 
the Royal Ontario Nickel Commission. 

368 



Extracts from Prof. Guess's report: 

"It is evident from the results shown in Table 3 that these laboratory 
"made steels are of good quality." 

"The value of this process of producing Nickel Copper steel is based 
"on the belief that Copper may replace a very considerable amount 
"of the Nickel in a •'!.">' ,' Nickel steel without producing an inferior article, 
"which belief is, I think, well founded." 

F 

Royal Ontario Nickel Commission Report, Page 421, Comparison 
of Nicu Steel with Nickel Steel. 

Nicu Steel Nickel Steel 

Carbon 0.44% 0.46% 

Silicon 0.034% 0.066% 

Manganese 0.50% 0.70% 

Phosphorus 0.013% 0.021% 

Sulphur 0.013% 0.034% 

K::::::::::::::::::::::*^SK 

The Physical Tests of the rolled natural steels showed: 

Elastic Limit 72,400 74,626 

Ultimate strength 115,000 122,000 

Elongation in 2 in 22% 16% 

Reduction of Area 51% 34% 

In the annealed condition the results were: 

Elastic Limit . . : 63,750 64,750 

Ultimate strength 107,300 119,000 

Elongation in 2 in 25% 17% 

Reduction of Area 48% 37.5% 

Mr. Colvocoresses, from his experiments, has arrived at the conclusion 
that the best results are obtained when the ratio of copper to nickel is 
as 1 to 3 or 1 to 4 and that the total copper content should not exceed 
1% if the copper is to be considered as replacing an equal percentage of 
nickel and the steel produced is to be put to the ordinary uses for which 
nickel steel is employed. The greater part of the nickel steel produced 
to-day contains about 3% of nickel, and it has been found that nickel- 
copper steel containing 2^ of nickel and .75% of copper is similar and 
equal to a straight 3% nickel steel. 

Nickel-copper steel possesses qualities, however, which give it the 
advantage over straight nickel steel, namely, a greater uniformity in 
composition and a decreased liability to corrosion owing to the presence 
of the copper. This latter quality, when fully demonstrated, should 

369 



give Nicu steel the preference over ordinary nickel steel for ship plates and 
machinery parts where such are subject to the action of acids, salt water 
and other corrosive agencies. Experiments made by Clamer to discover 
the effects of these destructive agencies on nickel-copper steel were highly 
favorable, and it has been known for a long time that a small addition of 
copper to ordinary steel is used by steel manufacturers in making special 
kinds of non-corrosive steels for use in locations where there is danger from 
corrosion. 

The following extract is quoted from an article on "Corrosion-Proof 
Steel," by F. Rowlinson, which appeared in the "Scientific American," 
of August 16th, 1919:— 

NICKEL-COPPER STEELS 

Since it was found that both nickel and copper gave immunity from 
corrosion when added to steel, it was expected that a steel containing 
both would also be comparatively free from corrosion. This was found to 
be the case. But an additional feature was found to be present in the more 
complex steel so made. Providing the proportion of nickel to copper was 
never less than five to two, the copper does not segregate as it tends to do 
in carbon steel. In this way the copper content, and, therefore, the 
powers of resistance to corrosion may be very considerably increased. 
Moreover, as regards mechanical tests, the copper seems to replace perfectly 
an equal percentage of nickel, so that a nickel-copper steel with, say 2J^ 
per cent nickel and 1 per cent copper has the mechanical properties of a 
3\4 P er cen t nickel steel, with increased immunity from corrosion. 

Of course, it is too soon to say yet how Nicu steel will stand up under 
the variety of uses to which nickel steel is put, such as in the manufacture 
of armour plate, steel rails, etc., but the tests and experiments which have 
already been carried out give every promise of a highly satisfactory product 
for these purposes, and it is confidently expected that when the many 
advantages of nickel-copper steel are fully realized, and the existing pre- 
judice against the presence of a small percentage of copper in steel has been 
removed, a large demand will develop for this product. 

The question of costs is an important one, and as Nicu steel has not, as 
yet been manufactured on a commercial scale, estimates of cost must be 
approximate; but Mr. Colvocoresses estimates — and his figures would 
appear to make due allowance for the several operations from the mining 
of the ore to the production of the finished steel — that Nicu Steel can be 
produced from ore at a cost of about $20.00 per ton over the cost of ordinary 
carbon steel. 

Some fears have been expressed that segregation of the nickel or copper 
might take place, thus destroying the uniformity of the product; but the 
accompanying sketch shewing one half of an 18" octagonal ingot, which was 
supplied for the purpose of testing and thereafter drilled and analysed, 
indicates clearly that these fears are groundless. 

370 



Ia/qot Top 

$HOW/NGi PfP/NG 




■TtX 



H" * The following partial report of T. W. Hardy and John Blizard, with 
diagrams and photomicrographs, is the result of an investigation by the 
Mines Branch of the Dep't of Mines, Ottawa, at the request of the Im- 
perial Munitions Resources Commission, of the properties of certain sam- 
ples of Nicu Steel made by the Nicu Steel Corporation : 

"The advantages obtained by the addition of nickel to steel have 
been well known for many j-ears and nickel steel is one of the most common 
of the alloy steels in use to-day. Nickel dissolves in iron in all proportions, 
but the steels of the greatest commercial importance contain less than 
5% nickel and not more than 1% carbon. Steels coming within these 
limits are, like plain carbon steel, pearlitic on slow cooling. Notwithstand- 
ing this structural similarity, however, the physical properties of pearlitic 
nickel steels are considerably superior to those of the corresponding carbon 
steels. The addition of nickel to plain carbon steel in the production of 
pearlitic nickel steel results in a considerable increase in strength, parti- 
cularly in elastic limit, while the ductility is decreased but little, if at all. 
To develop fully the high physical properties which they are capable of 
possessing, nickel steels must be heat-treated, and in this condition their 
superiority over similarly treated carbon steels is more apparent than when 
neither the nickel nor the carbon steels are heat-treated. Heat-treated 
nickel steel is not only stronger, but is tougher and has greater resistance to 
wear and shocks than a similarly treated carbon steel." 

"In the majority of the nickel steels manufactured to-day, the nickel 
content does not exceed 3*/£% and the carbon content is generally less 
than 0.50%. Steels of this class find a wide application in structures 
where a high combination of strength and ductility combined with a 
saving in weight are essential factors. Among the more common appli- 
cations of these nickel steels may be mentioned rails and bridge members, 
for which they are used without heat treatment; other perhaps more 
common applications are automobile and engine parts, machine parts and 
gun forgings, for which purposes the steel is practically always heat- 
treated." 

" THE INFLUENCE OF COPPER ON THE PROPERTIES OF 

STEEL " 

"The effect of copper on the properties of steel has been the subject of 
a great deal of discussion. For a long time copper, even in very small 
amounts, was supposed to cause red-shortness, and this view is still held 
by some. Of later years, the researches of several able investigators have 
proved that copper, within certain limits, is not only harmless, but really 
improves the properties of the steel. As a result of these researches it 
may be conservatively stated that in proportions not exceeding 0.75% 
copper does not make steel red-short provided the sulphur content is 
not high. There seems to be no doubt that an amount of sulphur, say 
0.08% to 0.1%, that would not cause serious trouble in steel free from 
copper, would, in conjuction with a few tenths of one per cent of copper, 
produce serious red-shortness." 

372 



"The results of these investigations also show that the presence of 
"small amounts of copper makes steel more resistant to corrosion. They 
"also show that the general effect of copper, in small amounts, on the tensile 
properties of the steel, is to increase the tensile strength and yield point, 
"and to decrease the elongation and reduction of area." 

It seems probable that the red-shortness and segregation that some- 
times accompany copper when present in steel in an amount exceeding 1%, 
have their ultimate explanation in the fact that copper is but Blightly 
soluble in iron. Since nickel unites readily with both copper and iron, it is 
reasonable to assume that the presence of nickel will permit larger 
amounts of copper to be used without introducing the above defects. 
This effect of nickel in increasing the solubility of copper in steel, together 
with the fact that copper resembles nickel in its effect on the cold pro- 
perties of the steel, lends weight to the probability that copper may 
replace a considerable proportion of the nickel in a nickel steel without 
materially altering its hot or cold properties." 

Acknowledgments are due to Mr. H. S. Foote, who prepared and 
photographed the microsections; to Messrs H. A. Leverin and F. W. 
Baridon, who determined the chemical composition of the steel; and to 
Mr. E. S. Malloch, who assisted in the tensile tests. 
aao-c HEATING COOLING 



eoo'c- 




6oo*c 



TEMPERATURE DIFFERENCE - DIFFERENTIAL CURVES 
HEAT NO. 4 




TEMPERATURE DirrEPENCE - DIFFERENTIAL CURVES 
HEAT NO 10 
373 







NICU STEEL TESTS 
Heat No. 4 
CHEMICAL ANALYSES- 










Analysis by 


Bar 


Car- 
bon 


Phos- 
phorus 


Man- 
ganese 


Sulphur 


Silicon 


Nickel 


Copper 


Cobalt 


Nicu Steel Co.. . 

Mines Branch . . 

do 


A 
B 


0.28% 

0.262 

0.267 


0.005% 
0.012% 
0.006% 


0.58 % 

0.522% 
0.532% 


0.038% 
0.061% 
0.036% 


0.014% 
0.009% 
0.006% 


2.16% 
1.98% 
1.96% 


0.41 % 
0.470% 
0.450% 


0.16% 








S 


ERIES . 


\ 











PIECES from BAR A, Quenched in Water from S00°C, and Drawn to 
Various Temperatures 

HEAT TREATMENT and PHYSICAL PROPERTIES 





Heat Treatment 


Physical Properties 






Test 














No. 


Heated 


Quenched 


Draw- Yield Point 


Ultimate 


Elonga- 


Reduc- 




to 


in 


ing 
Temp. 


Strength 


tion (in 
2") 


tion in 
Area 


1 


800 °C 


Water 


350°C 77,400 lbs. /sq. in. 


104,400 lbs./sq.in. 


23.5% 


64.0% 


2 




" 


400 °C 78.900 


101,200 


21.5% 


58.8% 


3 




" . 


4.50°C ,70,200 


99,400 


25.0% 


64.0% 


4 




" 


500°C 74,400 


95,800 


27.0% 


66.1% 


5 




" 


550°C 72,100 


91,000 


29.0% 


66.2% 


6 




" 


600 °C 67,500 


86,400 


29.5% 


70.8% 


7 




Cooled F 


reely in p 












Air. 




56,000 


79,000 


30.0% 


59.0% 


8 


" 


Cooled S 


lowly in 














Furnac 


e. ,51,000 


72,000 


31.0% 


53.0% 


9 


As Rolle 


d 


47,400 


76.700 


33.0% 


59.0% 



374 



- 1 i : r i a b 

PIECES from BAB B, Quenched In Oil from 800°C, and Drawn to 
Various Temperaturea 

BEAT TREATMENT and PHYSICAL PROPERTIES 





1 [i it Treatment 


Physical Prop 




Test 














No. 




Quenched 


Draw- yield Point 


Ultimate 


Elonga- 


Reduc- 




to 


in 


ing 
Temp. 


Strength 


tion (in 
2") 


tion in 
Area 


1 


BOO C 


ou 


350°C 08,000 lbs. >., in. 


91,000 lbs./8q. in 


20.0% 


66.3% 








400°C 67,700 


90,300 


29.5% 


66.3% 


3 






450°C 66.200 


88,700 


31.5% 


68.5% 


4 






500 °C 65,000 


87,100 


31.5% 


66.3% 


■ i 






2,600 


85,900 


32.0% 


iw; 2- ; 


6 






600°C 61,900 


83,000 


35.0% 


70.5% 


' 




Colled Fr 


eely ir. 












Air. 


57,500 


77,900 


34.0% 


61.5% 


8 




Cooled SI 


owly in 












Furnac 


e. 54,000 


73,100 


31.0% 


59.0% 


9 


As Rolle 


d 


47,800 


73,300 


34.0% 


58.8% 



NoTOi: 

1. All heat treatments made on the full section (1" Round), and testpieces (AS T.M. 
Standard 2 inch with threaded ends) machined from the heat treated stock. 

2. This heat was made from Sudbury ore. 







NICU .STEEL TES1 - 

Heat No. 10 

CHEMICAL ANALYSES 








Analysis by 


Bar 


Carbon 


Phos- 
phorus 


Man- 
ganese 


Sulphur 


Silicon 


Nickel 


Copper 


Cobalt 


Nicu Steel Co . 
Mines Branch . 
do 


A 
B 


0.34 % 


0.017% 
0.025% 
0.024% 


0.63 % 
0.554% 


0.041% 
0.043% 
0.050% 


0.019% 
0.017% 
0.023% 


1.33% 
1.31% 
1.30% 


0.46 % 
0.570% 
0.572% 


0.13% 



.17.-, 



SERIES A 

PIECES from BAR A, Quenched in Water from 800 °C, and Drawn to 

Various Temperatures 

HEAT TREATMENT and PHYSICAL PROPERTIES 





Heat Treatment 




Physical Properties 




Piece 














No. 


Heated 


Quenched 


Draw- 




Elonga- 


Reduc- 




to 


in 


ing ! Yield Point 
Temp. 


Strength 


tion (in 
2") 


tion in 
Area 


1 


800 °C 


Water 


350°C 87,600 lbs. /sq. in. 


123,000 lbs./sq.in. 


15.5% 


51.0% 


2 


" 


" 


400 °C 83,400 


118,000 


18.5% 


53.8% 


3 


" 


" 


460°C 83,600 


115,000 


19.5% 


56.5% 


4 


" 


" 


500 °C 


84,000 


112,500 


22.0% 


61.6% 


5 


" 


" 


550 °C 


79,000 


106,000 


23.0% 


61.6% 


6 


" 


" 


600 °C 


74,400 


102,800 


25.0% 


64.0% 


7 


" 


Cooled F 

Air. 
Cooled SI 


reely in 


52,000 


81,000 


30.5% 


59.1% 


8 


•• 


owly in 














Furnac 


e. 


48,400 


76,400 


32.0% 


53.8% 


9 


As Roll 


ed 




53,500 


84,600 


32.0% 


59.1% 



SERIES B 
PIECES from BAR B, Quenched in Oil from 800°C, and Drawn to 
Various Temperatures 
HEAT TREATMENT AND PHYSICAL PROPERTIES 





Heat Treatment Physical Properties 






Piece 














No. 


Heated 


Quenched 


Draw- 




Elonga- 


Reduc- 




to 


in 


ing Yield Point 
Temp. 


Strength 


tion (in 
2") 


tion in 
Area 


1 


800 °C 


Oil 


350°C 62,200 lbs./sq.in. 


98,300 lbs. /sq. in. 


25.5% 


61.6% 


2 


" 


" 


400°C 65,200 


99,400 


24.0% 


61.6% 


3 


" 


" 


460°C 64,700 


99,400 


26.0% 


61.6% 


4 


" 


*' 


500 °C 72,900 


99,800 


25.5% 


61.6% 


5 


" 


" 


550°C 65,700 


94,200 


27.0% 


64.0% 


6 


" 


" 


600°C 66,200 


94,200 


29.0% 


64.0% 


7 


" 


Cooled F 


reely in 












Air. 


53,000 


81,500 


29.0% 


59.1% 


8 


" 


Cooled si 


owly in 












Furnac 


e 40,700 


76,400 


31.5% 


59.1% 


9 


As Roll 


ed. 


53,500 


82,000 


29.5% 


56.5% 



Notes: — 

1. All heat treatments made on the full section (1" Rounds). Testpieces (A.S.T.M. 
Standard with threaded ends) , were then machined from tho heat treated stock. 

2. This heat was made from Sudbury slag. 

376 




NICU STEEL 
HEAT H?4 

TENSILE TESTS 



377 



dMm 

— — = w (m— x — 852+Ei 
dx 

5. For Figure No. 1 
Refer to Fig. No. 8, 

m 2 e 2 

Mm = w — +Ee+Ex — w — +w c > 
2 2 

m 2 e 2 

= w — +Ee + (E — w e) x — w - 

m 2 
= w— +282x + 16602 

dMm 

——=282 
dx 

6. For Figure No. 1 
Refer to Figure No. 9, 

m2 x2 ^ 

Mm = w — — w — +Eie! 
2 2 

dMm 

— - — =Ei — w x 
dx 

7. Moment table for locomotives, page 276. 

A moment table of the wheel concentrations of the locomotives was 
used from which the load El and the moment Eiei of Figures 7 and 9 
could be easily obtained. 

Maximum Stress in any member. The position of the locomotive giving 
the maximum stress in any member was found by means of an equation 
obtained by differentiating the equation for the stress in that member. The 
first differential is equal to zero and contains only expressions representing 
the loads and the coefficients for the moments, thus giving an equation 
which was very easily solved. 

As an example refer to the calculations for the stress in the member 
M4F4, which follows, and to figures 5, 8 and 9. 

The first differential of the equation for the stress in this member is : — 

.4357X282—58.333 (Eb — wx)+ 22.688 (Ef— wx) 
—2.3090 (Eg— wx) 4-1,0627 (Ek—wx) 4-. 7484 +282 = 0. 

where Eb, Ef, Eg, and Ek represent the weight of the locomotives to 
the left of panel points 4, 3,- 2 and 1 respectively, x the distance from the 
point of moments to the uniform load preceding the locomotives and 
w = 5000 lbs. per lineal foot of track. 

To obtain the condition for the maximum stress in the member it is 
then only necessary to find a position of the locomotives such that the sum 

278 



of the positive factors of the equation falls between the sums of the negative 
factors, obtained by first considering Eb as the weight of the locomotive 
to the left of the panel point 4 and second by considering that the weight 
Eb includes the weight of the wheel concentrated at that point. 

Stresses, — The full calculations for the stress in the members UoLo 
MjLj, MiFi and the Hangers Hf (see Fig. 1) follow and will serve to illus- 
trate the method followed for calculating the stresses in all the members 
of the truss. 

The clause in the specification stating that the locomotives 
to be followed or preceded or followed and preceded by the uniform 
load made it necessary to figure all the stresses twice, first with the loco- 
motives headed towards the main pier, as herein, and second with the 
locomotives headed towards the center of the span, the maximum of 
these two conditions being taken as the max'mum live load stress in the 
member. The difference in the stresses thus obtained was found to be 
very slight and would hardly justify the additional amount of labor 
involved. 

Stress in UoLo 

(See Figure No. 3.) 
UoLo = 1.5625 Mm -18.666 Nn 




-M\ 



£5.£_ 



Tn. 



X = SI/ s 



* m -- 663'- 



Criterion: 






(See Figures Nos. 5 and 9.) 






1.5625X282= 441. 


90 


150 




90 

( o, 


90 

60) X 19.06 = 0, 1140. 


Stress 






m 2 
w— =1098920 
2 






282x = 144240 






18 170 






Mm =1261630X1.5625 


= 1971.2 




690 






6890 






Xn 7580X18.666 


= 141.5 






1829.7 


= UoLo 




279 






Quenched in water from 800°C. Drawn to 600°C. 




As Rolled. 
380 




Annealed at 800°C. 

PHOTOMICROGRAPHS. HEAT No. 10. 

MAGNIFICATION 75 DIAMETERS 




Quenched in water from 800°C. Drawn to 400°C. 
381 




Quenched in water from 800°C. Drawn to 500°C. 




Quenched in water from 800°C. Drawn to 600°C. 

382 




As I lolled 




Annealed at S00°C. 
3S3 



DISCUSSION 

lapt. Jamieson Capt. E. A. J amies on, A.M.E.I.C— This paper is one that should 
interest every engineer at this time, and if this process be carried out on 
a large commercial scale, it means not only a rejuvenation of our steel 
industry, but means the greatest boost it has yet received. 

This process means not only the reclaiming of the 8,000,000 tons of 
iron said to be held in the slag dumps in the Sudbury District, but means the 
production of a steel of excellent qualities. The presence of copper in the 
Nicu Steel is not too high. The copper in steel used to be considered to be 
very harmful, but now it is known that when it is present in small quanti- 
ties, it has no serious influence on the physical properties of steel. 

According to M. A. L. Colby (Stahl and Eisen, 1900, pp. 20.55), 
neither a propeller shaft with .565 per cent copper nor a gun tube with .553 
per cent showed any defects after forging and hardening, and armour plates 
with .575 per cent flanged well and answered all the requirements of the 
American Navy. Various samples of Bessemer steel with carbon ranging 
from .11 to .65 per cent and with .292 to .486 per cent copper showed no 
red shortness on rolling. 

A steel containing a high percentage of copper, although not red short 
in the ordinary sense, yet will not bear the same amount of heat as the same 
metal without copper, thus showing that the effect of the latter resembles 
that of carbon, rather than that of sulphur. 

After making various physical tests on cuperous samples, the investi- 
gators concluded that copper does not possess the property of making steel 
red short. 

C. F. Burgess and J. Ashton (Electrochemical and Metallurgical 
Industry, 1909, vol. VII, pp. 527-529), have also examined the properties 
of iron-copper alloys, and found that the alloys up to two per cent copper, 
forge well at red heats. Those from 2 to 7 per cent copper will not forge 
at a low temperature, and rather poorly at a white heat, the case of work- 
ability varying inversely as the percentage of copper. The same authors 
(Ibid, 1910, vol. VIII, pp. 23-26), on examining the magnetic and elec- 
trical properties of these alloys found that above 2 per cent copper the per- 
meability falls off rather rapidly as the copper content is increased, but only 
bars of 4 per cent and more are conspicuously poor. Annealing greatly 
improves the quality of the bars. 

From Col. Leonard's paper, on page 9, the chemical analyses show the 
copper constituent to be well below the figures given by Mr. Colby. As 
regards the nickel constituent, its percentage is somewhat lower than that 
required in the manufacture of gun steel but as this system appears not to 
take into consideration the addition of any elements other than those 
already contained in the ore or slags used, the increase in the per cent of 
nickel can be artificially carried out. The physical tests appear very 
satisfactory and must be very gratifying to the makers of the Nicu steel. 

384 



Last year, it was my privilege to visit a number of the large steel Capt. Jamieson 
producing corporations of Croat Britain, and everywhere t lie same ques- 
tion was asked, "What are you doing in Canada to improve the steel 
stiuation?" Steel is required in larger quantities and of a better quality 
than the world has ever demanded before, and if we as Canadian engineers, 
know how to increase this production of steel, it is our duty to see that it is 
carried out. Why should Canada not compete in the meat call for arma- 
ment? Previous to the War, Great Britain and Germany supplied the 
largest percentage of the World's armament, and what has it done for 
Great Britain? Huge industries can be built here in Canada, as well 
as over there. 

Do the engineers or general public realize what it would mean to our 
Empire should the Huns succeed in their efforts to bomb the large steel 
plants in England. 

I was present during raids of the Huns and saw them make efforts to 
reach the steel plants in the North of England. An extremely high wind 
II that saved them one night, and judging from the damage caused by 
one Zeppelin bomb, a steel plant would be wiped out in no time and past 
all hopes of repair, as several tons of molten steel let loose are not as snow, 
regards removal. 

It would be the greatest national catastrophe that could befall us and 
would we in Canada, as a part of our Great Empire, not be sadly lacking in 
our duty as such, if we were not ready to step in to help fill the breech. The 
production of high grade steel in large quantities, in Canada is a national 
necessity at present and in the future post war days, it would enable us to 
compete with the world. 

From the viewpoint of armament, which at this time is before us all, 
why should Canada not be independent and manufacture her own arma- 
ment ? Col. Leonard shows us that we have the raw material and that this 
system can produce high grade steel. If the Government of our country 
does not see the necessity of encouraging the increased production of steel, 
then we as members of the Canadian Society of Civil Engineers are not 
doing our duty if we do not appeal to them to see as we do and urge upon 
them the necessity, if necessary, of subsidizing corporations to enable 
Canada to take her place in the front line with the great steel producing 
countries together with their allied industries. 

Mr. G. C. MACKENZIE, 'Sec, Munitions Resources Commission, Mr. Mackenzie 
Ottawa.) — I have been very much interested in Col. Leonard's paper 
as it reminds me of a discussion I had some years ago with Mr. 
R, W. Beelye, late manager of the Mines Department, of the Algoma Steel 
Corporation. Mr. Seelye had an idea that some use could be made of the 
large slag dumps at Copper Cliff and Victoria Mines, and suggested that 

385 



Mr. Mackenzie the iron content might be separated by means of electro magnets. This, 
of course, was out of the question because the iron exists in these slags in 
the form of a silicate, chemically combined. Subsequently, I proposed the 
smelting of these slags to the blast furnace superintendent who was then in 
charge of the Soo furnaces, but he was not at all enthusiastic over making 
use of this material which he claimed, quite rightly, was very similar 
to mill-cinder, a slag product of puddling furnaces, which is sometimes 
used in the manufacture of pig iron. 

Mill-cinder, as every iron blast furnace man knows, is a most objec- 
tionable material to smelt because its fusion point being low it melts high 
up in the blast furnace stack and runs ahead of the charge in the form of 
unreduced iron silicate with ill affect upon the furnace lining and also upon 
the quality of the pig iron produced. On the other hand, oxides, such as 
hematite, limonite and magnetite are gradually reduced and carbonized 
in their descent to the tuyeres where the iron is finally brought to the fusion 
point and falls to the hearth. 

I have myself smelted a considerable tonnage of mill-cinder and have 
a lively recollection of its behavior; and while I have never heard of a 
furnace burdened entirely with this material I would anticipate consider- 
able difficulty in smelting in a blast furnace the Sudbury slags which are 
similar in physical condition to a mill-cinder. 

A mixture of roasted copper-nickel ore and slag might be easier to 
work than the slag alone; but I would anticipate that fuel consumption 
would be relatively high, possibly over 3,000 pounds of coke per ton of 
copper-nickel pig on the supposition that a furnace operating with this 
material would require to be run very hot and with a very basic slag. 
It might be possible to cany out experiments in one of the idle blast furnace 
stacks at Parry Sound or Midland, Ontario. Such experiments would, of 
course, be costly but would I think demonstrate the practicability or other- 
wise of manufacturing Nicu pig from a mixture of roasted Sudbury ores 
and Sudbury slags. 

In so far as the electric reduction of these materials is concerned I 
would think that electric power should be secured for not more than 
$10.00 per horse-power year, if commercial success is to be attained. 
From a purely metallurgical standpoint the production of Nicu steel in 
electric furnaces may be said to have been a success; but it is quite another 
matter to produce this steel to compete in the open market with nickel 
steels made by standard practise. It may be said, therefore, that Nicu 
steel might eventually become a commercial success, by first manufacturing 
Nicu pig in coke iron blast furnaces, providing the smelting of iron silicate is 
proved a success and then converting this pig to Nicu steel either in electric 
or open hearth furnaces. 

386 



('<>i .!>.< ' \i;\i ..n , [mperial Munitions Board). Col. Leonard is to be < 
congratulated on the work he has done in connection with the manufacture 
of copper-nickel steel from the Sudbury ores. This workis of the bighesi 
importance to Canada bul 1 feel there is much yel to be done in an 
experimental way before it is proved really a commercial proposition. 

The [mperial Munitions Board are dow conducting experiments with 
regard to the quality and uses of the Bteel produced by this company. 

Mb. W. J. Dick, M.E.I.C., (Conservation Commission). TheMi 

successful and economic manufacture Of NlCU steel is OIIC of 

lerable importance both from the commercial and conservation 
stan.lp.iini. By the present practice all of the iron and sulphur 
content of the ore is lost , the former representing a loss of about three- 
quarters of a million tons of iron per year while the loss of sulphur 
amounted to about 100 tons per day. It istruethal the markets could 
not consume this quantity of sulphur or sulphuric acid, but owing 
to the nature of the process now in use it is not possible to recover the 
sulphur as elemental sulphur or as sulphuric acid. 

Canada has always realized the importance of having an iron and steel 
industry. 

Sine- 1896 a total of $16,785,827 has been paid by the Government 
of Canada in bounties for the production of iron and steel, the annual 

pay nts on pig-iron, puddled iron bars, steel, and manufactures of steel, 

being shown in the following table. — 



387 



Total Bounties on Iron and Steel Paid by the Government of Canada 

since 1896 



Year Ended 



Pig-iron 



Puddled 
Iron Bars 



Steel 



Manu- 
factures 
of Steel 



June 30, 1896 

" 1897 

" 1898 

" 1899 

" 1900 :.. 

1901 

1902 

" 1903 

1904 

" 1905 

1906 

Mar. 31, 1907 (9 months) 

" 1908 

" 1909 

" 1910 

" 1911 

1912 

" 1913 



Total. 



104,105 
66,509 
165,654 
187,954 
238,296 
351,259 
693,108 
666,001 
533,982 
624,667 
687,632 
385,231 
863,817 
693,423 
573,969 
261,434 



5,611 

3,019 

7,706 

17,511 

10,121 

16,703 

20,550 

6,702 

11,669 

7,895 

5,875 

312 



59,499 

17,366 

67,454 

74,644 

64,360 

100,058 

77,431 

729,102 

347,990 

676,318 

941,000 

575,259 

1,092,201 

838,100 

695,752 

350,456 



7,097,041 



113,674 



6,706,990 



15,321 
231,324 
369,832 
338,999 
347,135 
333,091 
538,812 
526,858 
166,750 



2,868,122 



These bounties have been the means of greatly stimulating the 
industry and, therefore, of great importance in connection with the manu- 
facture of munitions in Canada. At the same time, the iron and steel in- 
dustry is largely dependent on Newfoundland and the United States for 
supplies of iron ore. There are only two iron mines in Canada — the 
Magpie and the Helen. 

The importance of Nicu steel to Canada is evident when it is con- 
sidered that the Sudbury deposits consitute the greatest known reserves 
of iron in Canada. 



Mr. J. B. Ch allies, M.E.I. C. (Supt. Water Powers Branch, 
Dept. of the Interior). — We have at last heard of a real constructive 
conservation scheme of national moment. If the process for 
nickel-copper steel production so instructively described by Colonel 
Leonard and his colleague, Mr. Colvocoresses, proves to be the economic and 

388 



commercial success which we arc all so reasonably assured it will, it will Mr. Challies 
mean the salvation of two of our great industries and the rejuvenation of a 
third. The assured product ion of high grade Xicu-stee! will save our great 
steel industry and apparently withoul the necessity of Government bonuses. 
The production of large quantities of sulphur — the milk of chemical manu- 
facture — will prove a boon to a great many important electro-chemical 
processes now largely dependent upon imported sulphur, and probably of 
equal importance will be the rejuvenation of the great industry of hydro 
power development in Central < inada. It is not clear yet that electric 
energy is an essential feature of the process necessary to produce Nicu-steel. 
It is probable, however, that, when all the factors involved are carefully 
weighed, cheap electric power will prove to be an integral part of the pro- 
ject. Fortunately Canada has in her vast water power resources potential 
possibilities which can, if necessary, be called upon. 

Probably the most interesting feature of the whole matter, however, 
is that raised by Capt. Jamieson, namely, the relationship of this process 
to the world production of steel. We sincerely hope that Capt. Jamieson's 
prediction that Canada, as a result of this process will be at least able to 
take the position occupied by Germany in the world production of steel 
previous to the war. If so, the future of our country is indeed a rosy one 
for the engineer. 



Mil. AoamShortt. — Colonel Leonard's paper and much of the discus- Mr. Shortt 
sion which followed it leads to the practical conclusion, so often verified in 
other cases, that the real line of progress in the development of national 
resources lies in a close study of the peculiar nature and possibilities of these 
resources. Yet, lacking original and independent initiative, an addition 
to the careful stud}- of the world's past experience, the popular tendency is 
towards the mechanical reproduction of the methods and processes which 
have been justified elsewhere. As this mechanical reproduction of foreign 
methods is commonly accompanied in its early stages with considerable 
loss, it is sought to compensate for this by drawing on the national resources, 
either through direct bounties or indirect taxation via customs duties. A 
closer examination commonly reveals the fact, however, that the successes 
achieved elsewhere, and which it is frequently so difficult to reproduce under 
new conditions, have themselves been due to just that enterprising study 
of the special f< the peculiar combination of factors which every 

new country or area presents. 

The possibilities of turning out from the iron, nickel and copper ores 
of the Sudbury region, under the most favourable market conditions, 
indefinite quantities of nicu-steel, without first separating and then recom- 
bining these metals, suggests a revolution in the production of this highly 
valuable alloy, comparable only to that accomplished through perfecting 
the methods of producing carbonized steel directly from the ores as com- 
pared with the older and enormously more costly methods of carbonizing 

389 



wrought iron. It would appear that what chiefly remains to be done is to 
work out the details of the most efficient and economic method of pro- 
ducing from the ores and slags of the Sudbury region the most effective 
alloy or series of alloys of iron, nickel and- copper suitable for the uses to 
which the finished product may be applied. Here we have a series of 
problems of the highest national interest which at once calls for and justi- 
fies liberal assistance on the part of the government. , It is just at this stage 
in the development of national resources that government assistance is 
certain to be most effective and most completely justified. Such a policy 
contributes to the national interest as a whole, since the results of the 
demonstrations which may be made through government assistance will be 
available for any corporate or private interests which may be able and 
willing to undertake as a commercial venture what has been demonstrated to 
be a practicable use of a great national resource. The government would 
be in an excellent position to attach whatever regulations may be necessary 
in the public interest to the private development of these resources. In this 
way, and at a more fraction of the outlay incurred in the line of iron and 
steel bounties, the government may place an incalculably important in- 
dustry on a sound financial basis attractive to either private or foreign 
capital. 

Mr. J. A. DeCew, A.M.E.I.C. — Mr. President and Gentlemen, lam 
pleased to be able to express my interest and appreciation of this paper 
as there are several aspects of it which are of particular interest to the 
chemical engineer, one being the remarkable chemical inactivity of this 
new product. These copper steels which contain such a small quantity of 
copper in solution in the iron, become so resistant to corrosion that they 
compete directly with the almost chemically pure iron, which is now 
widely used for those purposes where corrosion is a great disadvantage. 

The economic importance of this "Nicu Steel" to the industries of 
Canada is perhaps not yet fully appreciated and I sincerely trust that 
nothing will occur to retard its development. 



390 



KETTLE RAPIDS BRIDGE 

By w. Chase Thomson, M.E.! < , 
(Read at a meeting of the Society on April 11th, 1918 . 

GENERAL 

The Budson Bay Railway extends from The Pas, the northern 
terminus of the Canadian Northern Railway in Manitoba, to Porl Nelson 
on Hudson's Hay, a distance of 424 miles. Although primarily intended 

i short route for the export of grain to Europe, the railway opens up 
a valuable territory, rich in minerals, fish and pulp-wood and of great 
agricultural possibilities. The grading has been completed throughout, 

and the rail.- have been laid to mile :;:;_'. 

There are three important bridges on the line: The first crosses 
the Saskatchewan River at The Pas. ami comprises four fixed-spans 

of about 150 feet i gether with a swing-span of about l'.'.ii feet; 

the second crosses the Nelson River at Manitou Rapids, mile 242, ami 
is a handsome structure of conventional deck cantilever type, with 
a channel-span of 304 feet 6 inches ami anchor-spans of Pis feet 'I inches. 
supplemented by an 85 foot plate-girder span at the north end: the third, 
which is the subject of this paper, is at the second crossing of the Nelson 
Paver, or Kettle Rapids, mile 332, the present end of steel. The latitude 
of the bridge-Site N., and the longitude, 514 34'-49" \Y. 

The Nelson is one of the great rivers of ( ianada, its drainage including 

the prairies of Alberta, Saskatchewan and Manitoba on die west, the 
Ked River Valley <>n the south and part of Ontario on tin- east; but. 
owing to the intervention of Lake Winnipeg which servo as a huge 
reservoir, the How of water in this river throughout the year is remark- 
ably uniform. At Kettle Rapids, the Lowesl water-level recorded to 
date is 316.0, as observed December 31, I'd?: and the highesl water- 
level, with the river unobstructed by ice, 319.0, as observed November -i, 
1916; a difference of only 3 feet. But. with the freezing of the river 
ami the consequent jamming of huge quantities of drift-ice, the channel 
is greatly obstructed and the water rises suddenly. In the winter of 
1916-17. this phenomenal rise began at Kettle Rapids on January 19th, 
when the water was at elevation 3l7.il, and the greatest height. 338.5 
hed on February 3rd. Within two days, however, the water 

391 



dropped 7 feet; after which, it gradually subsided to 325.0 on May 23rd; 
it then went down suddenly and reached a normal elevation of 317.0 
on June 6th, when the river was clear of ice. Owing to the exceptionally 
low temperature throughout December 1917, the ice-jam occured on the 
last day of that month, which was much earlier than usual. The water 
was then at elevation 316.0, but rose 9 feet during the next 24 hours; 
and, on January 6th when the last observation was made, it had reached 
elevation 335.0, or within 3.5 feet of the recorded maximum. 

The highest ice-peaks have always been found on the islands, where 
piers 2 and 3 are located. In the winter of 1916-17, when the water was 
at its maximum height of 338.5, there were ice-peaks as high as 344.5, 
the same as had been observed during the winter of 1913-14; but, with 
the fall of the water, they settled to elevation 342.0, and remained there 
until melted. 

The main channel at the bridge-site is only 350 feet wide, and 
estimated to be about 200 feet deep at the centre; the current is very 
swift, and there is always a certain amount of open water. Directly 
above and below the bridge-site, however, the river freezes all the way 
across, but only after the jamming of the ice and the consequent rising 
of the water. It is evident that there can never be any danger from 
ice, either to the superstructure or to the piers; for the steelwork is 
15 feet clear of the highest fixed ice-peaks, and there is running-ice only 
when the water-level is much below its maximum elevation. 

A diagram of water-levels, from June 1, 1916, to January 6, 1918, 
is shewn on plate I. 

In locating the line, advantage was taken of two very conveniently- 
placed islands, allowing a central span of 400 feet, with piers and abut- 
ments on the solid rock. This rock is of pre-cambrian origin and is a 
tough granitoid gneiss. 

The general design and dimensions of the bridge, including the 
substructure, are shewn on plate II. It will be noted that it is a con- 
tinuous structure 1000 feet long, having a channel-span of 400 feet and 
two side-spans of 300 feet each. The trusses, or main girders, are of 
the sub-divided Warren type, 50 feet deep throughout, 24 feet apart 
centre to centre, having 25-foot panels. There are two lines of stringers, 
8 feet apart centre to centre; and the base-of-rail is 17 feet 6 inches above 
the centre-line of the bottom-chords. The structure is ri vetted through- 
out, and all bracing is rigid; it is fixed at pier 3, and provided with 
expansion-rollers at all other bearings. The ties are 8 x 12 inches, 14 feet 
long, spaced 12 inches centre to centre; they are notched K inch on the 
stringers, and every fourth tie is fastened thereto by a %-inch hook- 
bolt. The outer guard-timbers are 8 x 9 inches, spaced 10 feet 10 inches 
in the clear; they are notched one inch and secured to every fourth tie 

392 



by a ? 4 -inch bolt. Steel guard-rails, weighing GO lbs. per yard, are 
provided inside of the running rails, with S inches clearance between 
heads; they are brought together in a frog beyond the ends of the bridge. 
The main (or running) rails are of the American Society of Civil Engi- 
neers' standard section, weighing 80 lbs. per yard; at abutment 1, where 
the total expansion and contraction of the bridge will be about 8 inches, 
they are provided with specially-designed expansion-joints of the split- 
switch form, with points of manganese-steel. Refuge-bays, for pedes- 
trians, are provided at intervals of 200 feet. 

Three types of bridges were practicable for this location: 1st, simple 
spans, with temporary members over the piers for cantilever-ereci ion 
of the channel-span; 2nd, the conventional cantilever bridge, with a 
central freely-suspended span; 3rd, a true continuous-girder bridge. 
The first would have been satisfactory, but un-econoraical, owing to the 
great weight of extra metal required for erection stresses only. The 
second was rejected partly on account of the objectionable articulated 
joints at the ends of the suspended span, but principally because of the 
expensive shop and erection work in connection therewith; for an econom- 
ically-designed cantilever structure would have required a much 
greater depth over the piers, with considerably less depth at the abut- 
ments and for the suspended span, resulting in sloping chords and 
irregular webbing; besides, in order to obtain such economical propor- 
tions, it would have been necessary to locate the bottom-chords as 
close to the base-of-rail as possible, thus largely increasing the quantity 
of concrete in the substructure. 

The third type, as designed and built, is the most rigid of all, and 
the most economical; for it required no extra metal for erection-stresses, 
except in the bottom-chords of the channel-span adjacent to the piers, 
where the increase of section was slight; the simplicity and uniformity 
of the framing reduced the cost of fabrication to a minimum; and the 
continuous horizontal chords, without adjustable joints, greatly facili- 
tated the work of erection. It is admitted that continuous-girder 
bridges have been regarded somewhat unfavourably in the past; for 
it has been claimed that the usual theory for computing the stresses 
therein, which assumes a constant moment of inertia, is inexact; that 
the computation of the stresses is too difficult and tedious; finally, that 
the least settlement of any support would radically alter the stresses 
and thus endanger the structure. No doubt, in the old days of pin- 
connected bridges, continuous girders were undesirable in many respects - 
although a notable example of such a structure, which has received 
much praise and which gave excellent service for many years, was the 
old Canadian Pacific Railway bridge at Lachine. Now that pin- 
connections have been almost entirely superseded by rivetted joints, 
continuous girders are growing in favour; and the most remarkable 
example of such construction is to be found in the recently-constructed 
Sciotoville Bridge over the Ohio River, comprising two continuous spans 
of 775 feet each. 

393 



Regarding the objection, that the computed stresses are inexact, 
it may be stated that, in the present instance, the reactions were first 
computed for panel-concentrations by formulae as given in Merriman 
and Jacoby's "Roofs and Bridges", part IV, pp. 12 and 13, and afterwards 
checked by the elastic method. The difference in the results obtained 
by these two methods was less than ^2 of 1%, which should satisfy the 
most exacting. This close agreement was undoubtedly due to the 
parallel chords and nearly constant moment of inertia; but, in the most 
extreme case, the error due to the use of the formulae would probably 
not exceed 6%. 

The objection, that the labour of computing the stresses for con- 
tinuous girders is too difficult and tedious, is unworthy of notice, espe- 
cially where a large and important structure is concerned. 

Finally, the claim with reference to results that would be produced 
by a possible settlement of one or more of the supports, is more rational, 
but much exaggerated; for continuous girders are very elastic structures, 
and can accommodate themselves to small settlements of supports 
without developing serious alterations of stress in their members. In 
this case, the ends of the trusses, if unsupported, would deflect 25 inches 
below the horizontal line from dead-load; and the alteration in the dead- 
load reactions at the abutments due to raising or lowering the end 
supports a whole inch would only be 9500 lbs., or 4%, whilst the reactions 
and stresses for the live-load would not be affected at all. Moreover, 
it is inconceivable that any settlement of the foundations can take 
place, as they are all on the solid rock; otherwise, this design would 
not have been adopted. Furthermore, in order to provide for possible 
small inaccuracies in fabrication or erection, the ends were made adjust- 
able by allowing 1% inches for shims between the shoe-castings and 
the bottom-chords; and hydraulic-jacks, with gauges attached, were 
used to set the ends at the height necessary to obtain the computed 
dead-load reaction. In this connection, and referring again to the 
Sciotoville Bridge, the following quotation from an article by Clyde 
B. Pyle, published in Engineering News-Record, January 31, 1918, 
will be of interest: 

"One of the most striking features of the entire erection was the 
curve for the last few inches of the jacking, after the steel towers were 
both free. The computed increment for each inch of lift was 7.5 tons; 
and the actual increase in load was too small be read on the gauges. 

"It is quite evident from this condition that, for long-span continuous 
trusses, it is not as vital a point as was formerly thought to be the case 
to have the supports at exactly correct elevations. In this case an error 
in setting one of the end supports, say as much as 3 inches, would have 
changed the end reaction 22.5 tons, or the stress in the end-post 32 tons, 

394 



which would l>e less than the probable error in computing the actual 
Btreaa in thai member. The worst condition of ahop-work, erection 
.•111*1 setting of shoes could nol possibly total more than our inch; so 
that the certainty of stresses and therefore the safety of such a bridge 
is left without question. 

"The fact thai complications enter into the design and erection 

cannol bar the use of such bridges as long as they are economical. The 

ions usually given, thai the stresses are nol statically determinate 

and that uncertainties of stresses result from slight errors in elevation 

of the supports, are no longer valid." 

DETAILS OF DESIGN 

The structure has been designed in accordance with the General 

Specification for Steel Bridges, issued by the Department of Railways 

and ( 'anals in 1908, excepl for a slight modifical ion in the impact formula. 

affecting alternating stresses only, and a change in the allowable unit- 

• compression-members. 

In the matter of impact, when dealing with alternate live-load 

the Department's specification requires the impact to be 
computed by squaring the arithmetical sum of the tension and com- 
pression stresses due. to the live-load (or the range of stress >. and dividing 
by the maximum algebraic sum of CO-existing dead-load and live-load 

jes, or /= ' Now, taking an extreme case in which a member 

max. 

's subject to alternating live-load stresses of equal amounts, but no 

(/, _i. i) 2 
dead-load siress. the impact would be = 4L, or four times the 

live-load stress of either kind, which is certainly excessive. If. however, 
we take for the range the live-load stress of one kind and add to it t/lO 

of that of the other kind, we have =2L, approximately, or an 

impact equal to twice the live-load stress of either kind, which would 
seem to be ample. In conformity to the above argument, impact has 

ranqe ^ 
been computed bv the formula. 1 = ■ — . with the arbitrary stipu- 

lation that the range shall be taken as the arithmetical sum of the 

live-load stress of the greater kind and 1 Id of that of the Lesser. When 

/ : 
the live-load stress i> of one kind onlj . 1 he formula reduces to 7= — - 



L + D, 
in which L = live-load stress and D = dead-load stress. 

Concerning the unit-stresses for compression-members, the Depart- 
ment's specification calls for 16000 lbs. per square-inch reduced by 
Gordon's formula, using in the denominator the factor 18000 for square- 
ends, 12000 for one Bquare and one pin-end. and 9000 for pin-ends. It is 

395 



now quite generally recognised, however, that 16000 lbs. per square- 
inch is entirely too high for short columns; and the Joint Committee 
on Columns and Struts in the United States, which has recently sub- 
mitted its final report, recommends a maximum working unit-stress 
of 12000 lbs. per square-inch, which is therein shewn to provide a factor- 
of-safety of 2, or the same as for tension-members when designed for a 
unit-stress of 16000 lbs. per square-inch on the net section. In the 
General Specification for Steel Highway Bridges, recently adopted by 
the Society, the formula for compression-members is 12000 — 0.3 (Ijr ) 2 , 
which becomes zero when l/r = 200. In this bridge, the compression- 
members have been designed in accordance with the formula, 

(I r) 2 
12000/1 + qfir>n n> which agrees closely with that adopted by the Society 

for values of l/r up to 70, but gives somewhat higher unit-stresses 
for greater working ratios. 

The stresses, with data for same and the make-up of the members 
of the structure, are shewn on plate III. The live-load indicated is 
"Class Heavy" of the Department's specification, above noted; and the 
dead-load concentrations at panel-points represent the actual weights 
of steelwork and floor, carefully computed from the detail drawings. 
The bottom laterals have been proportioned on the assumption that the 
whole of the specified wind-loads, both during erection and afterwards, 
would be resisted thereby; and the wind-load stresses in the bottom- 
chords include the vertical effect of the wind-loads, equal to their 
moment about the bottom-chords divided by the horizontal distance 
centre to centre of chords. The design includes provision for cantilever- 
erection from piers 2 and 3 to the centre of the channel-span. Wind 
and erection-stresses are shewn only where they affect the sectional 
area of members. 

Provision for traction and braking forces has been made by hori- 
zontal trussing attached to the top-flange of the stringers and to the 
floorbeams at points M0, M4, M8, M12, etc., or 100 feet apart, as shewn 
on plates II and III; which forces are transmitted to the main girders 
through the inclined struts MO-Afl, M3-M4, M4r-M5, M7-MS, etc. 

The end floorbeams are provided with stiffeners and bearing-plates 
at points 16 feet apart, for jacking-up the bridge; and the floorbeams 
at M12 have been specially designed for lifting the bridge, with unit- 
stresses increased by 50% and having stiffeners and bearing-plates at 
points 14 feet apart. 

Latticing of main members has been avoided as far as practicable; 
but the open sides of compression chord-members are double-latticed 
with 5 x 5^-inch flats, having two rivets at ends and at intersections; 
tension chord-members are similarly latticed with 5 x J-^-inch flats. 

396 



All of the principal w « I >-i n« ■ 1 1 1 1 n ■ rs are provided with substantia] longi- 
tudinal diaphragms, which arc considered as pan of the effective section 
thereof; and the heavy compression diagonals, U&-LS, VlO-LU, L12-C714 
and L16-U1S, are further stiffened by tie-plates on (heir out-standing 
flanges. All joints and splices are fully rivetted throughout. 

Rocker-hearings are provided throughout, having 8-inch bearing- 
pins at the piers and 6-inch bearing-pins it the abutments; and the shoes 
are steel castings. The bridge is fixed at pier 3 and movable at pier 2 
as well as at the abutments. At pier 2, the expansion-rollers are 8 inches 
in diameter, and each set is provided with four 12-tooth cut pinions to 
prevent skewing. Substantial curtain-plates are supplied for the pro- 
tection of the gears and to keep out the dust; but they are removable 
for inspection and cleaning of the bearings. At the abutments, the 
expansion-rollers are 6 inches in diameter and similarly provided with 
alignment gears and curtain-plates. These expansion-bearings are 
shewn on plate IV. The fixed-bearings at pier 3 are similar to the 
expansion-bearings at pier 2, except for rollers and bed-plates. The 
bridge-seats are tool-dressed perfectly level and to the exact elevations 
called for on the drawings; and sheet-lead, y % inch thick, is provided 
to equalise any minor irregularities of the surfaces. 

Owing to the small deflection of this bridge, which is only 3 inches 
at the centre of the channel-span, for dead-load combined with the 
maximum effect of the live-load, it was considered unnecessary to 
provide for a perfectly straight bottom-chord under any particular 
condition of loading; so the trusses have been cambered, in accordance 
with the more usual method employed for simple spans of moderate 
length, by increasing the length of the top-chord panels, as shewn in 
the upper diagram on plate V. Members U10-U12, and C712-I714 have 
been correspondingly shortened; and % inch has been added to the 
verticals U12-L12, to obviate a slight kink at panel-points C/12. At panel- 
points LY2, the ends of the abutting chord-members have been bevelled 
to accommodate the form of the trusses when fully loaded. This method 
of cambering has greatly simplified the shop-work; and the results are 
entirely satisfactory. 

The total estimated weight of steel in the structure (including 
floor-bolts), computed from the writer's detail drawings before the 
contract had been awarded, was 4,424,000 lbs.; and the actual shipping- 
weight, as determined by the scales, was 4,415,000 lbs. 

ERECTinx 

Erection was started on June 6th, the earliest date possible; for, 
before the falsework for the southern anchor-span could be placed, it 
was necessary to blow up with dynamite huge masses of ice. This 
anchor-span was then erected in the usual manner by a 75-ton derrick- 
car, having trucks 35 feet centre to centre and a single 50-foot boom. 

397 



At LO, bottom-chords were set 10 inches low, by omitting the upper 
shoe-castings and using flatted pins for bearings. This was to provide 
for the deflection of the channel-span during erection, and to insure 
that the connections at L20 could be effected before the chords at U20 
would meet. After the anchor-span had been fully ri vetted, the southern 
half of the channel-span was then erected as a cantilever by the same 
derrick-car, the ri vetting following closely behind the work of erection. 
Panel-point L14 was supported temporarily by wire cables from panel- 
point {712, until the connection had been made at t/14; likewise, panel- 
point LIS was supported from panel-point [716, until the connection 
had been made at 1/18. By August 18th, or eleven weeks from the date 
of beginning, the first half of the bridge was fully erected; and the rivet- 
ting on this portion of the structure was completed one week later. 

The next, and perhaps the most difficult piece of work in connection 
with the entire erection, was the construction of a double cableway 
for transporting to the opposite side of the river the materials for the 
northern half of the bridge. The cables were supported on a rocker- 
bent 40 feet high, standing on the top-chords of the southern cantilever 
at panel-point 1718, and on a timber tower 120 feet high from the ground- 
surface, located behind abutment 4, with centre-line 60 feet from panel- 
point LO. The span of the cableway was 611 feet; the sag, under maxi- 
mum load, 36 feet; and the horizontal distance of the anchorages, at 
both ends, from adjacent supports, 400 feet. A triangular equalising- 
girder, suspended at its ends from the cables and having a lifting-hook 
at the centre, was provided for loading the cables equally. The cable- 
way was designed for a live-load of 14 tons, the weight of the heaviest 
piece to be transported. Its general construction and method of opera- 
tion are clearly indicated in the reproductions of photographs, herewith. 
In addition to its principal function of transporting materials, in which 
service it gave entire satisfaction, the cableway was of great assistance 
in the erection of steelwork. 

The falsework for the northern anchor-span was then constructed, 
with extension-bents reaching to the floor-level, for the accommodation 
of the special traveller provided for the erection of the northern half 
of the bridge. As this traveller was to be used on the top-chords as 
well as at the floor-level, its four trucks (of two 24-inch double-flanged 
wheels each) were spaced 24 feet centre to centre transversely, the same 
as the trusses, and 50 feet centre to centre longitudinally, to coincide 
with the panel-points; it was fitted with two 62-foot booms and a hoisting- 
engine; and its entire weight, including counterweight, was 60 tons, 
equally distributed on the four trucks. To provide for the weight of 
the traveller only, when moving, the top-chord members were supported 
at their middle-point by temporary timber posts, resting on special 
seats at panel-points MB, M5, Ml, M9, etc. 

398 



After the delay incident to the construction of the cableway, and 
the falsework for the northern anchor-span, erection of Bteel for the 
northern half of the bridge commenced on September* 17th. Beginning 
at pier :> (with the traveller al the floor-level and working towards 
abutment 4) the floor-system, bottom-laterals and Lower members of 
the trusses were placed, and the falsework-extensions removed. The 
traveller was then blocked up to the beighl of the top-chords, requiring 

a week for this operation; and the upper part of the steelwork for this 

anchor-span was erected, working from abutment 1 towards pier :;. 
When the traveller had passed panel-point f/6, an additional rocker- 
bent, 40 feet nigh, was set up there as an intermediate support for th< 
cables, thereby reducing the span to Hit Meet, the maximum Bag to 22 feet . 
and providing ample working clearance above the top of the steelwork. 
The cantilever-erection of the northern half of the channel-span was 
omplished in the same manner as for the southern half, except that 
rhe members were placed principally by the special traveller, which 
required temporary timber supports at the centre of the top-chord 

panels, already mentioned. 

From the time of placing the traveller on the top-chords to within 
two days of the end of November, when the weather suddenly turned 

rely cold, rapid progress had been made; and it was confident lv 
expected that the bridge would be entirely completed before the end 
of the first week in December. Accordingly, the writer was instructed 
to proceed to Kettle Rapids for the purpose of making a final inspection 
of the bridgl and to supers ise t he adjustment of the end react ions. ( )n 

his arrival at the bridge-site, December Mh. the work was still held 
up; but it was found that the centre connections at L_'n had been made. 
and without the least difficulty; for, on meeting, the trusses had been in 
perfect alignment and the deflections of the cantilevered arms, exactly 
equal; thus it had only been necessary to jack forward on its rollers 
the southern half of the structure, which had purposely been set back 
.") inches to facilitate the erection of the closing members. On the 
southern half of the bridge, the timber floor was practically complete; 
and the ties bad been roughly distributed over the northern half, except 
at one panel adjacent to the centre of the channel-span, where the stringers 
had not at that date been placed; thus the structure was practically 
supporting its full dead-load. 

Under these conditions, careful levels were taken to ascertain tic- 
exact deflections of the trusses or main girders, with results as shewn 
m the middle diagram on plate V. It will be noted thai the curves 
of th<- bottom-chords are remarkably uniform : and it wa< a great source 
of gratification to the writer to find that the centre ordinate was exactly 
the same as had been computed. At panel-point U20, the distance 
- of adjacent chord-members was one inch. 



The weather having moderated slightly, although still very cold, 
work on the bridge was resumed December 16th; for it was hoped that it 
might yet be possible to complete it this winter. On the 22nd, the ends 
were lifted V/2 inches, which was just sufficient to bring the ends of the 
top-chords at U20 to a firm bearing. Owing to frequent stoppages due 
to weather conditions, it was not until the last day of the month that 
the ri vetting of the main members was completed, and jacking of the 
ends was resumed. This operation was again interrupted by New 
Year's day (which was too cold to work, in any case); but the ends had 
been raised sufficiently by January 2nd to permit of placing the upper 
shoe-castings, without shims. Although the ends were thus 1% inches 
below their normal position, the load at each of the four corners, as 
indicated by the gauges on the hydraulic jacks, was exactly 118J/£ tons, 
the amount of the computed, dead-load reaction. The bridge at the 
time, however, was covered with many tons of ice and snow; thus it was 
impossible to determine very accurately the reactions for the normal 
dead-load. 

It had by this time been decided to give up the attempt to complete 
the bridge during the winter of 1917-18; for the men could not work to 
advantage; a satisfactory job could not be made of the track-work; 
the painting could not be done until the advent of mild weather; and the 
bridge was perfectly safe. A final adjustment will therefore be made 
under more favourable conditions, when it is expected that the ends 
will require to be raised about another inch. 

With the ends \y% inches low, levels were again taken on the bridge, 
with the satisfactory results indicated in the lower diagram on plate V; 
for the camber at the centre of the channel-span was found to be 1% 
inches, whereas the maximum computed deflection due to the specified 
live-load is 1% inches. 

The closing panel of stringers was placed January 4th, which ended 
the work for the season. The remaining work comprises a small amount 
of rivetting for secondary parts; some minor adjustments; the completion 
of the timber deck, including the laying of the rails; and painting. 

On January 6th, the Superintendent of Erection pulled out, with 
his men and such of his equipment as will not be needed to finish the 
bridge, leaving a watchman in charge. 

SUBSTRUCTURE 

The substructure is of concrete throughout, composed of pit-gravel 
and cement, in such porportions as were found by trial to give the best 
results. It had been intended to construct at least the abutment and 
pier on the southern side of the river during the autumn of 1916; but 
the track did not reach the bridge-site until the end of October; cold 
weather set in shortly after, and there was barely time to construct 

400 



the foundation for abutment 1. Excavation for this foundation was 

carried to a depth of over 10 feet, through frozen (day and silt, to the 
solid rock. The concrete was placed during the second wick In Decem- 
ber, and in very cold weather; but the materials had been heated, the 
mass was large and the result was entirely satisfactory, as found from 
a careful inspection the following spring. The abutment was completed 
during the month of April, 1917. 

I >pcrat ions at pier 2 were begun on April Kit h. and under very adverse 
circumstances; for the ri\er was t hen at elevat ion 328.0, or 10 feet above 
the average rock-surface at this point; and the rock was covered with 
a solid mass of ice. _'."> feet thick. However, it was necessary to get 
ahead with the work as rapidly as possible; so the ice was excavated, 
and the rock was bared by May 5th, at which date the water had fallen 
to elevation 325.0. Uthough the ice-walls of the exacavated shaft 
appeared to he perfectly solid throughout, the water percolated through 
and stood at the same elevation as that in the open river-channel; but 
it was perfectly still, without current or surge. A timber caisson, 
conforming on tin bottom to the irregularities of the rock-surface, was 
then constructed; and all small openings therein were sealed by sheet- 
piling, carefully scribed and driven so as to broom the ends thereof. 

ry inch of the rock-surface inside of the caisson was then picked 
with needle-bars, to insure that it was entirely clear of ice; and heated 
concrete was deposited by deep-sea buckets. The rock-surface at this 
pier had previously been carefully examined during low water, and 
found to be absolutely sound; thus every confidence may be placed in 
the foundation. The footing for this pier was completed on May 9th; 
the construction of the main shaft thereof offered no difficulties, and 
was effected without incident. 

At pier 3, no difficulties incident to water or ice were encountered; 
or work at this point was not started until June 29th; but there was a 
horizontal fissure in the rock at about elevation 322.0, which necessitated 
blowing up by dynamite the overlying mass. This resulted in giving 
an entirely satisfactory though very irregular foundation, to which 
the footing for the pier was made to conform. 

Excavation for abutment 4 was commenced on June 14th, and was 
carried through about 10 feet of frozen clay, silt and boulders to the 
solid rock. The footing, up to elevation 341.5, was completed July 21st. 

The pit-gravel, used throughout on this work, was invariably frozen 
and required to be thawed by steam; thus all of the concrete was placed 
warm, and with most gratifying results; for, on removal of the forms, 
not a single bad spot was discovered. 

401 



The butterfly wing-walls of the abutments were reinforced by 
twisted steel rods, one inch square, placed 3 inches from the rear surface. 
There were horizontals, 6 inches apart, wired to verticals, 3 feet apart. 
In addition, two such rods were placed along the upper edge of the wings. 

The total quantity of concrete in the work is about 3000 cubic- 
yards; and, of reinforcing steel in the wing-walls, 2300 lbs. 



The laying out of the work was difficult and tedious, owing to the 
irregularity of the ground and to the necessity of locating pier 3 by 
triangulation; but the instrument-work was done with such care and 
precision that all important dimensions and distances were afterwards 
found to be practically exact. In locating the centre-line of bed-plates 
on pier 2, and that of the shoe-castings on pier 3, where great accuracy 
was desired, the piano-wire method of measurement was used, taking 
into accoimt the pull on the ends of the wire and the corresponding sag, 
as determined on a level surface, and making the proper correction for 
temperature. The distance between centres of bearings on piers 2 and 3 
was afterwards found to agree with the steel structure, as built, within 
5/16 inch. It had been specified that the centre-line of the expansion- 
shoes should coincide with that of the corresponding bed-plates at a 
temperature of 30 degrees, Fahrenheit; and, on inspection, with the 
thermometer at zero, the centre-line of the roller-shoes at pier 2 was 
found to be exactly 1J4 inches northerly of the centre-line of the bed- 
plates, instead of 15/16 inch, the amount of contraction in the steelwork 
in 400 feet for a fall in temperature of 30 degrees. Thus the distance 
between the bearings was too great by 5/16 inch. If there had been any 
appreciable error in the setting of the bearings on these piers, it could 
have been rectified, as provision had been made for jacking-up the struc- 
ture, if necessary; but, as the dead-load reactions here are about 500 
tons each, and the shoes very awkward to move, any such adjustment 
after erection would have been difficult and expensive. 

The entire work has been under the general supervision of Mr. W. A. 
Bowden, M. Can. Soc. C. E., Chief Engineer, Department of Railways 
and Canals, Ottawa; and of Mr. J. W. Porter, M. Can. Soc. C. E., 
Chief Engineer, Hudson Bay Railway, The Pas. It was designed in 
full detail by the writer, who has been retained throughout for consult- 
ation in connection therewith. The superstructure was fabricated and 
erected by The Canadian Bridge Company, Limited, Walkerville, 
Ont. Mr. T. B. Campbell, M. Can. Soc. C. E., Bridge Engineer, Hud- 
son Bay Railway, was in charge at the bridge-site, under whose 
immediate supervision the concrete-work was constructed, and by 
whom all lines and elevations for the erection of the steelwork were 

402 



given; Mr. 1. E. Malum was the Superintendent <>f Erection for the 
bridge company; and Mr. .lames Carr, Representative of the Canadia n 
Inspection and Testing Laboratories, Limited, attended to the 
field-inspection. The entire work has been carried oul without lose 

Of life and without a serious accident. 

Great credit is due to Messrs. McDonald Brothers, for the excel- 
lence of the concrete-work; to Mr. Campbell, for the accuracy of his 
lines and elevations, and for his efficient supervision; to the engineerf 
of the bridge company, for their splendidly-conceived and carefully- 
prepared scheme of erection; to the shops, for the neatness and accuracy 
of workmanship; finally, to Mr Malion, for his skill and care in handling 
under difficulties this important and somewhat unusual piece of erection- 
work. 



discussion- 
mi-. 1'. P. Skbabvtood, M.E.I.C. — The continuousspan design, adopted Mr. si 

by Mr. Thomson, must have been very economical compared with either 
simple span or cantilever construction, but looking at the profile of the 
crossing, the reason for the long anchor spans is not evident, and if these 
spans could have been reduced by about 100 feet a considerable saving in 
the weight of steel and total cost of the bridge would have been 
effected. 

An unusual and good feature of the design is the half through type. 
Placing the floor near the neutral axis of the truss instead of at the 
upper and lower chords, has among other advantages, that of keeping the 
longitudinal stringers in the position where the changes in length due to 
stresses in the chords, are zero. It also facilitates the floor beam and truss 
connections by placing them clear of one another. 

The adoption of the 11 and I shaped sections for the web members 
simplifies the connections, provides the best form of diaphragms to equalize 
the loading of the inner and outer material of the members as well as 
cheapening the manufacture. It is mentioned that plates are used to 
stiffen the outstanding flanges and it would bo interesting to know how far 
apart these tie plates are placed, so as to compare it with the spacing adopted 
on tin' tie plated columns which were tested and condemned by the A.K.C.A. 
recently. The former have the tie plates merely as secondary stiffening, 
whereas the latter relied on them entirely to make the channels act together. 
The contrast in spacing may show that the columns tested by the A U.K. 
should have had the tie plates placed about one third the distance apart to 
jepresent usual practice and obtain a fair result. 

403 



!r. Shearwood The paper mentions that all the horizontal loads are earned by the 

lower lateral bracing. Sway bracing is provided at every vertical and the 
verticals are very stiff in the direction required to resist horizontal forces. 
It would be interesting to know whether the author considered the question 
of omitting them and really, taking all upper horizontal forces directly 
to the lower laterals, instead of merely assuming this effect. Top laterals 
often appear to be superfluous members, in fact it is even possible that they 
may be harmful. 

The pier members have an ingenious method of controlling the rollers 
which should certainly be effective in keeping them from skewing or creep- 
ing. Past experience has shown that this is very necessary. The arrange- 
ment has the advantage of having the mechanism exposed for inspection. 

The design for the cast shoes and beds is unusual in having a box 
section which does not seem to have the metal so advantageously distri- 
buted to resist the leading as the usual open ribbed types. It has also the 
disadvantage of more difficult moulding. 

The erection scheme is interesting and proved very successful, but why 
was it necessary to carry the traveller on extended falsework to the main 
pier so as to lay the lower chord, and then elevate it to the upper chord 
height, instead of at once starting off at the upper chord height to assemble 
the whole truss and falsework as it travelled out from the abutment towards 
the main pier ? 



Mr. M. J. Butler ; C. M. G., M. E. I. C— Mr. Thompson has 
contributed an interesting and valuable paper on an important bridge. 

The railway in question is, however, one that, in my opinion, ought 
not to have been built by the Government of Canada. It was entered 
upon at a time, when the resources of the country were under severe 
strain due to the construction of the Grand Trunk Pacific and the liabilities 
incurred by the guarantees given to the Canadian Northern System. 

Aside from the financial reason for not going on — no information of a 
reliable character existed as to the navigation possibilities of the Hudson 
Bay route. All the data available seemed to point to the fact, that a 
possible IYl months might be had on the average — but, until a systematic 
study over a period of several years it would seem to have been wise and 
prudent to wait. Whether the actual period of reasonably safe navigation 
is two months — as I understand the information available— or four 
months, as I have heard estimated — the question of securing ships for 
such a route has not, I feel sure, been dealt with. 

The information given in the paper, that there is along the line, 
"a valuable territory, rich in minerals, fish and pulp wood and of great 
agricultural possibilities," is certainly very important, if true. 

404 



[ understand thai the route is over a barren, sub-arotic tundra -with Mr- Butler 

DO land fit for sett lei unit that any spruce found along the line is scrub of 

no commercial value The minerals, it would be \fery interesting to 
learn what they are and where located. 

I am, of course, well aware, that the North West belie V68 in the Hudson 
Bay route and that i he undertaking owes its existence to pressure on the 
Government from that part of the country. 1 only hope that their san- 
guine expectations may tie realized. 

I have read with interest the reasons which actuated Mr. Thompson 
in selecting a continuous truss and agree with him that the labour of 
computing the stresses is not an important matter. The design and details, 
both of manufacture and erection seem to me very good and reflect great 
credit on all concerned. 

I — nine that there is a general agreement that the assumptions 
upon which the elastic theory are based, viz. a constant modulus of elasti- 
city and a constant moment of inertia, do not in fact exist. That certain 
approximate methods which have been developed check sufficiently with 
the "elastic theory" to re-assure the designer — that he is about as near 
right as is necessary to secure a safe structure. Too much refinement in 
bridge calculation, seems out of place. The increasing wheel loads of 
locomotive and rolling stock, demand that bridges be built adequate for 
all reasonable increase in weight. 



Mr. Geo. F Porter, M.E.I.C — I wish to congratulate Mr. Thomson Mr. Porter 
on the capable piece of work he has presented to the Institute this 
evening. His remarks in regard to the use of continuous girders where 
such conditions exist as in the present instance, are very well taken. 

There has been a very strong prejudice in the minds of American 
engineers against their use for fixed spans, while the usual type of draw 
bridge has of necessity been of this type. The fact of continuity has 
seldom been the cause of trouble except in cases where the foundations 
were poor and settlement took place. 

1 would like to ask Mr. Thomson, who has, no doubt. Looked into the 
matter carefully, whether the saving of masonry due to the placing of the 
floor well above the bottom chord, thus making necessary the introduction 
of bottom struts and sway bracing beside special provision for traction 
Btrusses, really proved to be economical. 

Mr. W.Chase Thomson, M.E.I.C -Replying to Mr. Shearwood: — A Mr. ThomsoD 
preliminary design for the bridge had side spans of 200 feet, but was 
subject to uplift at the ends, with consequent hammering. Counter- 
weighting, to overcome the uplift induced by the live-load, was con- 
sidered, and would, no doubt, have been a satisfactory solution of the 

405 



Ir. Thomson problem, as this scheme was later adopted in connection with the new 
continuous-span bridge over the Allegheny River, described in Engineering 
News-Record, May 2, 1918. Side spans of 300 feet were finally adopted 
principally because it was considered advisable to provide the wider 
water-way, which latter will be restricted by a rock-fill embankment at 
the north end of the structure. 

The supplementary tie-plates on the principal compression diagonals 
are spaced about 7 feet centres: they are about 21 inches long, with five 
rivets on each side. 

Top laterals may be superfluous in the completed structure, although 
it is difficult to see how they can be harmful. In this instance, their chief 
function is to hold the chords in alignment, which might otherwise have a 
tendency to kink at the joints. In the Quebec Bridge, which has no top 
laterals, the upper chords are in tension throughout the anchor spans and 
cantilever arms. 

The idea in adopting the box section for the cast shoes was to secure the 
best possible distribution of the load over the masonry. With the open- 
ribbed type, the edges of the bottom plate cannot be so well supported. 

Replying to Mr. Butler : — Although the writer has no brief to defend 
the Hudson Bay Railway, he is of the opinion that the undertaking is less 
foolish than that in connection with many other lines which have been 
constructed during recent years; moreover, there are many who believe 
it will, when completed, fulfil its original function: — to help in the shipment 
of grain to Europe. It has been pretty well demonstrated that the Straits 
can be navigated not less than four months during the j'ear, and it is 
hoped that this period may be extended to six months. The muskeags, 
which are so numerous, are generally only a few feet deep, and overlie a 
clay formation; when drained, they should be exceedingly fertile. Along 
the embankments, the writer has seen splendid specimens of oats and 
timothy, which had sprung up from seed accidently spilt; and several 
of the section-men have raised excellent vegetables. The lowest tempera- 
lure thus far officially recorded, between The Pas and Port Nelson, is 49 
below zero, whereas 60 below is quite common in districts growing the 
best wheat. 

Great accuracy in the determination of stresses may not be necessary; 
but it is important to note that the usual formulae developed from the 
clastic theory are entirely reliable, especially when applied to girders of 
uniform depth. As stated in the text of the paper, the reactions derived 
from formula? applying to beams of constant moment-of-inertia had been 
checked by the more exact method of deflections, with practically identical 
results; moreover, the measured deflections of the structure and the scaled 
reactions compare very closely with the calculated amounts. As is well 
known, the modulus-of-elasticity for structural steel is one of its most 
reliable and constant properties. Considering the observed facts, Mr. 

406 



Butler's remark aboul ••certain approximate methods which check Buf- Mr. Thomsoi 

ficiently w i< b I he elasl ic t heory to reassure t he designer thai he is aboul as 

near righl rv to secure a safe structure" is Bomew hat misleading; 

for there can be no doubt thai the calculated r emit inuous-girder 

bridges are quite as reliable as those for so-called simple-span bridges. In 

both eases, there is far greater uncertainty regarding the actual live-load 

ami impact. 

[n compliance with the expressed wish of the meeting before which the 
paper was read, plates VI and VII, covering typical structural details, 
have been added. These details have beei I with a view to 

econon y c< n i atible with strength. All joints are fully rivetted for the 
full value i f the connected parts. The use of outside splice-ph tes where 
the main diagonals are attached to the gusset-plat< s, 1 1 upper pai el-points 
-. <'■. Ki, 1 I and is, and at lower panel-points 4, 8, 12, 16 and 20, has the 
double advantage in permitting the use of comparatively small gusset- 
plates, with short rivetted connections. At all middle panel-points, and at 
lower panel-points 2, 6, 10, 14 and IS, the gusset-plates are only '■'■ g inch 
thick, but of ample Btrength to resist the stresses transmitted thereto by the 
hangers and sub-diagonals; and the remaining splice materials are just of 
sufficient width to conform to the dimensions (in elevation) of the main 
members. ( )\\ ing to the large areatif these gusset-plates, this arrangement 
results in a very considerable saving of metal. At upper panel-points 10 
and 14, the gusset-plates are of sufficient size and thickness to resist the 
vertical and horizontal shears acting thereon. Owing to the slight bevel 
in the chords at lower panel-points 12 (1/64 inch in one foot), the gusset- 
plates extend below the chords Y% inch, with bottom edges planed straight 
to bear upon the shoe casting; furthermore, additional bearing area is 
provided by using tapered fillers under the bottom flange-angles of the 
chords, as shewn on the detail drawing of the large shoe, plate IV, the load 
being transmitted to these bottom flange-angles by the outside splice- 
plates which are ground to bear thereon, and in which additional rivets are 
provided, over and above the number required for their proportion of the 
bottom-chord -' 

The bridge was completed and the end bearings were finally adjusted 
on June 10, 1918. Without any shims, the deadload reaction at all four 
corners exceeded the computed amount (118.5 tons) by exactly 5 tons, or 
about 4%, which was considered entirely satisfactory. Thus the ends 
remain one and five-eights inches below the normal; and the lower diagram 
on plate V represents the actual form of the completed structure, when 
unloaded. This is as might have been expected; for the cantilever erection 
of the channel-span naturally tended to increase the length of the top 
chords throughout, resulting in lowering the ends, — though but slightly. 

Although the winter of 1917-18 was the most severe of any, along the 
line of the Hudson Bay Railway, according to official records, the maximum 
height of water and ice-peaks at Kettle Rapids, due to the jam, only 

407 



Mr. Thomson exceeded previous records by five-tenths of a foot. Thus there can be no 
doubt that the clear distance of 15 feet between the bottom chords and the 
previously-recorded maximum elevation of ice-peaks is more than ample to 
provide for all contingencies. 

Replying to Mr. Porter: — The additional steelwork necessitated by 
the placing of the floor above the bottom chords is principally included in 
the bottom struts; and the entire additional weight of steel due to this 
arrangement is only about 75,000 lbs. which, at the contract price of 7.15c, 
amounts to $5362.50. This does not take into account the lower sway- 
bracing, which would otherwise be offset by deeper overhead sway-bracing; 
nor the inclined struts, in the plane of the trusses, which are necessary for 
stiffening the vertical members at panel-points 4, 8, 12, etc. The saving 
of concrete effected is about 350 cu. yds. which, at the cost of $15.00 per 
cu. yd., in place, is equivalent to $5250.00. Thus the cost of the extra 
steel is almost exactly balanced by the saving of masonry. But the plan 
adopted has the advantage of keeping thefloorbeam connections clear of all 
panel-points, thus greatly simplifying the details; the application of the 
floorbeam concentrations to the verticals at a considerable distance 
from the end connections of the latter insures a much better 
distribution of stress among the component parts of the main truss- 
members; finally, the location of the track-stringers near the neutral-axis 
of the main girders obviates, largely, the usual racking on the stringer 
connections, due to the change in length of the chords under the action of 
the liveload. In addition to these advantages, it was desired to provide 
girders over the piers, strong enough for jacking up the structure in case of 
necessity; and the depth of the ordinary floorbeams was insufficient for 
this purpose. Moreover, the arrangement permitted of placing the end 
floorbeams underneath the stringers, thus obviating the necessity of brac- 
kets for carrying the floor-ties adjacent to the ballast walls. 

In conclusion, the writer desires to express his thanks to the gentle- 
men who have taken part in the discussion of this paper for their general 
approval, so kindly expressed. 



408 




1 i'_ r . 1: -Abutmenl 1 under construction. 




Fig. 2: — Pier 2 under construction. 
409 



u/2 | Mlll M -^Tfr^' M *"^y^^ gs 


i 




IMHHHn 


mV'^m^Ar A w^yiVA%%^'^M^m^ r 4m% 


- 4 fmbHTIW H^Ti IL<l^^ii^^^^^T 

- v • - - 



Fig. 3: — Southern anchor-span erected; and beginning of cantilever- 
erection, shewing temporary supports for panel-points L14. 




Fig. 4 : — Southern half of bridge erected. 
410 




Fig. 5:— Falsework under construction for northern anchor-span. 




pig. 8:— View of cableway; and beginning of erection of northern anchor- 
span, with traveller at floor level. 
411 




Fig. 7: — View from top of cableway-tower. 




Fig. S: — Lower steelwork for northern anchor-span erected; traveller 
raised on blocking, for working on top chords. 

412 




Fig, 9: — View shewing cableway with its equalizing-girder; derrick-car 

on southern cantilever; traveller on top-chords of northern anchor- 
span. 




Fit'. 10:— Northern snchor-Bpaa erected, also 100 feet of adjacent 

lever. Note additional rocker-benl at rti for supporting cables. 

413 




Fig. 11: — View of bridge from south shore before cableway and traveller 
had been dismounted. Note that the lower members of the 
portal-struts and sway-bracing are missing: they were omitted 
temporarily for the accommodation of the derrick-car. 




Fig. 12: — General view of bridge, taken from the south shore and looking 
up-stream. 

414 



CHAMPLAIN DRY DOCK FOR QUEBEC HARBOUR 
By U. VALIQUET, M.E.I.C. 

Superintending Engineer Department of Public Works. 
(Read at Montreal and Ottawa Branches April 25th, 1918). 

For a number of years the River St. Lawrence has been frequented 
by ocean steamers of such dimensions that they could not be accom- 
modated in the Lome Dry Dock, completed in 1886, at Lauzon, in the 
Harbour of Quebec. 

In 1906 the Canadian Pacific Railway Steamship Company brought 
out their steamers "Empress of Britain" and "Empress of Ireland", 
of 65-foot beam; the Allan Line steamers "Virginian" and "Victorian" 
of 60-foot beam were also placed on the St. Lawrence route in that year. 
The "Bavarian" of somewhat narrower beam, 59 feet 3 inches, came to 
Quebec in 1905; thereafter the number of large ships placed on the 
St. Lawrence traffic increased rapidly until in 1912, there were 25 vessels 
that could not have been repaired in the long stretch of the St. Lawrence 
navigation for want of sufficient dock accommodation, the width of 
entrance of the present dry dock being only 62 feet. Any of these 
vessels that required docking had to be repaired temporarily as well as 
possible, while afloat, and taken either to Halifax or New York, which, in 
some cases, was a risky undertaking. 

The case of the S. S. "Bavarian" was an unfortunate experience in 
this respect. On the 5th November 1905, this steamer ran aground 
with a full cargo from Montreal and Quebec, about 40 miles below Quebec, 
opposite Grosse Isle; although late in the fall she could have been 
raised and brought to Quebec had there been dock accommodation for 
her. Her beam was 59-3", but through the accident her sides had 
bulged out beyond the width of the dry dock entrance. She was raised in 
the following spring, although further damaged by ice during the winter, 
and brought on the beach a short distance below the dry dock, where she 
was sold as scrap. This is the worst case on record in the history of the 
St. Lawrence navigation. This vessel was only six years old and of a 
registered tonnage of 10,387 tons. 

415 



In the summer of 1898 the writer was instructed to prepare a report 
on the practicability of widening the entrance of the Lome Dry Dock, 
which had been completed in 1886. A plan was submitted, showing the 
possibility of obtaining an entrance 70 feet wide, by removing part of 
the timber slides at the outer end of the dock; increasing the length was 
also suggested. The first was reported to be inadvisable as it would 
greatly disfigure the dock and do away with the convenience of the 
timber slides; the only feasible way would be to remove and rebuild in 
another position the eastern side wall, thus depriving the harbour of 
all dock accommodation for probably two seasons. A new caisson 
wold necessarily have to be provided; the cost would have been 
considerable. Further, it was considered that a new dry dock would 
be required in Quebec before many years. 

The suggestion of lengthening the dock was adopted; the length 
was increased from 484 feet to 600 feet; this consisted merely in moving 
the circular head, stairways and timber slides 116 feet further, after 
excavating the rock to proper width and depth. The work was performed 
under contract awarded in the year 1900, for the sum of $100,000.00, 
and completed in 1901 without interfering with the use of the dock. 
The details of construction of this dry dock have already been described 
in a paper read before the Canadian Society of Civil Engineers some 
years ago, by Mr. St. George Boswell, Chief Engineer of the Quebec 
Harbour Commission, who was assistant engineer during the con- 
struction. This dry dock was built by the Quebec Harbour Commis- 
sioners under an Act, 38 Vict. Cap. 56-1875, by which the issue of bonds 
was allowed to obtain the necessary amount. The work was started in 
1878 and completed in 1886 at a total cost of $921,130. 

In 1888 the Canadian Government relieved the Harbour Commission 
of all obligations to refund the principal sum or interest expended on the 
dry dock and in 1890 it was placed under the control of the Department 
of Public Works, the writer was then placed in charge. 

NEW DRY DOCK 

In 1906 the Quebec Board of Commissioners urged upon the Govern- 
ment the necessity for a large dry dock for the harbour of Quebec. In 
the fall of the same year the writer was instructed to make a survey of 
the locality surrounding the old dry dock and report on the best location. 
Two sites were examined, but the position to the east of the present 
dock was considered the most advantageous for three principal reasons: 

A larger area of land could be acquired. 

A better foundation could be obtained. 

The repairing plant of Messrs. G. T. Davie & Sons could have 
better access to both the new and old docks. 

416 



/ Q I £ 


/y lb 

- June \ July -■* August * September-* October ^November* Dece. 


~r x : 
















- 












- 




J 










! 










K — 


i ' 1 u "~~ 










, . 




1 



Plate I 




































































































- June \ July -*-Aug 


/y lb 
us/ -+September± October ^NovemberiL 


zccmber ^January -^February -f 

MS 


March 


-^April — 


^May - 


^—June -4— July — 


^August -September *0dober -\ 


^Noyember ^December ^-January -> 

3A$ 










Ice j.^.Peaks 




































































































t X<<"-, " 






































































































__/ n w : 
























































































































































































































































































































































































































i""- 






































































































.. L. 










































































































































































































! ' '< 


Hi 






















































































,1« 




























































































































































































































-J-L 
























































































.. 














4 3 




































































































. 


JJO / ^"X^N 






















































































IV) 


J 














/ ! 




































































































i 


r 

1 






































































































































































































































































































































































































v.* 




























































































































































































































3 






































































































.. 






























































































HUDSON BAY RAILWAY 


































































































1 1 






,_,_ 






























































































Bridge over the Nelson River 






-- -L 


— N 






















v 7 


































































at Kettle Rapids 






















































































""N 








«. 


































































































-<;--- 
















































































































W Chase Thomson, Nor Birh Ibik/trq 








































































































Consulting Engineer MonlteolMbetlX 



















- 
































































































3/0 








































































J IS 








Water Levels 




Base of fo'U 376-5 



kissississi^^ississssi^ 



IE panels of 25-0"= 30O-6 c to c 




FLATE II 




HUDSON BAY RAILWAY 

Bridge over the Nelson River 

at Kettle Rapids 

WChexThomscn 

CcnSL.- yE-J -.'f- ".---.•••'. 



I LATE II 




.' 209000 

\-Oor+7SOOO 
I - 264 OOP 
-827 000$ \€X0-SI6 



?-33>s/Spls-4l25- 5500(1}) 
4t>*4-s)*l!' 23.44 ' IS 44 It) 
64-60 S3 4->o 



L5 
0- 245.000 

I - SOI 000 or 1 233 000 
/ - 415.000 
■l,22l.OOoe\gOOO>7eb 



2-33-s/spls'4i:S' 5500(10) 
2 21 > s/aplt- If 2S- 22 50 (6) 

4 6. 4 >ii/iSV - 25 1?4 - 20 14 iS) 
33 14-77 64o- 



• ISO ooo 

-332 OOO or 

-J6S00O 

• 3S7 000£I2000=78I 





Horizontal Bracin g at Top of Track Stringers 



12 panels of ES : 0"-3OO'-6 



Stringers 
Shear, d 6.250 

L 89.500 
I 84.250 
ISO.OOolbiSlOOOO' moi 

Web pi. S4'j/s- iazss 
Moments, o 39000 

L 48O.250 
1 445.000 , 
364.2SO Hlbs 

<3425'*22e.0OOlbs.8l6O00-l4-lo 

Flangn i of- 5 * • J /* "** * «" si 

14 2gaH. 



F-J Ffaar beams 

Shear, 10.500 
l 98.000 

I 890OO 



Web pi 72 
Moments, 



270- 



D S4000 
L 784.00O 
I 7IOOOO , ,, 

IS73000 ftlbs 

9575'- 274000 lbs dieooo- /7/o" 

Flanges. & of ' 72*3/8 wei-- _'3' 
2 o .0. 1/2 L' ■ 

/ li.i/Apl ' 4 81 12) 
(5 SSti H 



' ■ • - .-■'-.'■-•■• 

Shear is. 500 

HOB 40O 
I 95I0O 
219.000 

Webpl 72*3/3.2700 
Moments 124000 

L Si' 

I 'SCOOP 

1.7 5 i.OOO ft lis 
S 575'* 304.50C .-. 






.v.-.v- • - 

*;.-.-. ----- .- ..«.-..•. -: 
Ktntrel I9t-: 

Moment s .. 



' V ■ 
I-I3*9ll€p 












PLATE III 




To p L oteral Brocir. y 



2^270 






TJlS'lOSSIt' 



m 



. : iO.000 

- - 191000 

)00'74.lo 



I 33- 

!-33'iftpla ■ ->i o- 

4.6-4-i/lU • 19 oo 

80.87a 



l-33'«/cpl ■ 22 70 

<4 368.000 l-33-S/gn/j. • 4125 

' 166.000 or -100.000 2-21-5/epls • 26.25 

. S 34.000 4 6-4-Vtl! . 17/76 

• I. 128.000 $ 11200- 11873a 1 1 7 $6'" 







5SO0 

1108400 
I 95 !0O 

Help/ 7t'VS'27°<>~ 

-000 

- 'OOO 

I 760000 

ijsi.oooff lbs 

:~.soo lbs s leooo .I9oo- 

t of 7?.i/atrel>- 3 37 
2-C-C-'/2U •95on> 
I ii-9/xpl. • 6 ib (?) 
19 OS on 



Jackin q-u p Floorbeams at M/2 

(Jacking points 14' c/c) 
Shear d 912000 

• '5f e 48.000 (for overhft) 
1 020.OOoltu (315000 • 68O0 

Web pis /end:) 2-96-VS' 72.oo 
(Centre) I ■ 96 ■ 3/8 • 3 6 6 

Moment 5 100 000 fl lbs. . 

9 775'- 6S800o/bs <a 24000- 27 4 d 

Elonges g f 96. if g web • 4 so 

2 6- €. 9/fU - 10 61 111 

2 13' 9/lCpls ■ - 12 36 It) 

27 41 'on 



,»1 




u 


M 


.:. 


Tola 'i 


1 







12 SOO 


17 100 


:'?7.oo 


rV 


/ 




20 000 


30 tOO 


'0400 




.- 


26.000 




36.900 


0.7 OOO 


'<, 


3 




t 700 


29,900 


iH 000 


$ 


■> 


27.000 


:j 5,10 


19 100 


70 0..'. 


1 




8.200 


37 400 


40 600 


': 


/■ 


14300 




41.000 


7: 300 


,". 


7 

9 


2S.S0O' 


. 300 


32 300 


44 600 




;j 000 


71400 


• : 10: 1 


; : 


V 




11.700 


. ; ■> 


44 100 


10 


44200 




40 100 


84 300 


•■ 


11 




2-0,800 


31.4 00 


■ 4 : 0. 


12 


ii 600 


30.S00 


ls.200 


•1/ i,A» 




13 




IS 000 


\3iaoo 


56.400 


■. 


11 

IS- 

If. 

•7 


4 4J00 
22900 


14200 

24 000 
13 800 


40100 

31 000 


84.H.:- 
45$00 


^ 


26 600 
33700 


73,' 00 
4 7. 500 




IH 

70 


14.200 
2 7 300 


e2oo 

24 000 


42,700 
33.U0V 
27 000 


77400 
47ono 
73300" 



16 panels of ZS-0 -400'-0* 

Dead load, as per fable of conceal raiions, which includes floor weighing 600 Its. per I in ft of bridge. 



«S **l *""> **) 



OOOO 



«s ■*> •*> **> 



m«« ^M$®, to&MM®, ftkCfe 



47Solk.perlinft 

1 . . . 



Impact, r ^Jo'. . '" which ronge* greater live-hod stress* jj of lesser 

Wind-Loads, 400 Ibs.per lineal foot, applied at base of 'rail, considered as fixed load. 

400 lbs. per lineal fool, applied 8 ft. above base of rail, considered as moving load 

(1)2 

Tension, I6.ooolbs.persq.in; Compression, iZ.ooo '■»■ ■/ '■> 36 r 000 Ibs.per sq. in. 

Erection Conditions-. Dead-Load as above; Hem'ck -Car iso.ooolh trucks is ft c/c, with 79.000 lbs. on 

front truck. placed a! panel point 19. ond 71.000 lbs. on rear truck; or Top Chord 
Traveller rYeighmg \20000 lbs.,equal loads on trucks Soft c/c- Wind, Soo/bs. 
per tin. ft, applied at base of rail Unit Stresses increased 5o<fe. 



Bottom Latere! Bracing 



Rivet 

Ma'r- 
Specification f t. v . V.v — 
and Canah f: 



HUDSON BAY RAILWAY 

Bridge over the Nelson River 

at Kettle Rapids 

Stress Sheet 

W ChaseThomson. NewBir*s8i+khaf 




guide-pmicns, 8'p.d, IZt.,3£f.,4+% cut-teeth 



- 7-0" 



End Elevation, with End Cur tain -Plate removed Ex pansion- Bearings at Pier 2. 



Side Elevation, >nth Side Curtain- Plate removed 



<* 



PLATE IV 





v hS guic/e-pimons.6p.d.,IEt,3i,4--jek,cutteeth. 
3-7/" [&-\ 

\ 

End Elevation, with End Curtain-Plate removed Ex pansion-Bearings at Abutments I &4 Side Elevation, with Side Curtain- Plate remored 



Castings filled with Concrete 



r- mmm" 



HUDSON BAY RAILWAY 

Bridge over the Nelson Riyer 

at Kettle Rapids 

Exponsion-Bearings 

W. Chase Thomsoi.. New Birks Buildina. 

Consulting Engineer. Montreal, Mar ,'3t91& 



I plan and report were submitted in the early pari "f 1907; the dock 
then proposed was 1000 feel long with an entrance width of LOO feet. 

The proposition was not immediately acted upon; the question 
as to whether the Governmenl ahould build the dock or induce aome 
shipbuilding firm to build it under a subsidy from the Government was 
unsettled. The result of the discussion was the passing <>f an Act of 
Parliament at the session of the year 1911 . assented to on the 1th May, 
called an Act to Encourage the Construction of Drj Docks. 

Under tins Art the dry docks were divided into three classes. The 
first class included dry docks estimated to cost not more than four million 
dollars, and capable of receiving and repairing the largest -hips of the 

British Navy and of the following dimensions: 

(a) Clear length on bottom 900 feet; clear width of entrance 100 
feet, with depth on sill at high water ordinary spring tides of 35 Oct. 

(b) Floating dry docks of a lifting capacity of 25,000 tons. 

The second class included dry docks' estimated to cost two and one 
half million dollars of the following dimensions: 

(a) Clear length on bottom 650 feet; clear width of entrance s."> 
feet; depth of water on sill at ordinary high water spring tides 30 feet, 

if in tidal water-: or 25 fi et on sill, if constructed in non-tidal waters. 

(b) Float inn dry docks of a lift ing capacity of 15,000 tons. 

The third clat ed of dry docks estimated to cost not more 

than one and one-half million dollars, of the following dimensions: 

(a) Clear length on bottom 400 feet; (dear width of entrance 65 
feet; depth of water on sill at ordinary high water spring tides 22 feet, 
if in tidal waters; and IS feet, if in non-tidal water. 

(b) Floating dry docks of a lifting capacity of :;..",nii tons. 

The estimated cost in all en 2s includes the totally equipped 
repairing plant capable of effecting all - »rts of repairs, including machine 
-hops mid tool-, foundry, administration buildings, etc., together with 
the dock itself, but does not include marine Blips or other installation 
used in the construction of ships. 

the \ ' the subsidy on dry docks of the first class 

is 3J per cent per annum on the estimated cost for a period of 35 years 

from the time it has been reported thai the dry dock is entirely 

completed. The subsidy on the second clac i cent per annum for 

from the I ime of complet ion. ' >n the third ■ idy is 

3 per cent for a period not exceeding 20 years from the time of completion. 

In all cases the Company making the application must furnish plans 

with a detailed list of the plant and a complete estimate of the cost. 

|"1,, -cl and corrected, if found advisable: and, upon a report 

from the Chief Engineer of the Department of Public Works that, the 

117 



works intended to be built are in the public interest, the application is 
granted upon certain conditions of management and maintenance. The 
works are to be executed under the superintendence of an officer of the 
Department. 

The above Act was amended in April 1912, by making the length 
of the first class dry docks 1150 feet, the entrance 110 feet and the 
estimated cost five and a half millions. 

Another amendment was made in May 1914, by which the subsidy 
of four per cent on the estimated cost is allowed for first class dry 
docks. 

The Act was further amended in 1917, by which the dimensions of 
the first class dry docks shall be: length on bottom 1150 feet; width 
of entrance 125 feet; depth on sill at high water spring tides 38 feet. A 
subsidy of 4£ per cent on the estimated cost of five and a half million 
dollars is allowed, payable half-yearly for a period of 35 years from the 
time of completion. By this amendment no bonds or debentures are to 
be issued until one million dollars shall have been expended on the 
construction of the dry dock. 

After the passing of the Act of 1910, shipbuilding firms were invited 
to build a dry dock at Lauzon, in the Harbour of Quebec, under the 
subsidy Act of that year. Two companies submitted plans and offered 
to build under contract without reference to the subsidy Act. In 1912 
another company submitted plans for a dry dock to be built on the 
Quebec side of the harbour, just below the mouth of the St. Charles 
River, according to the subsidy Act, as amended in 1912. Some 
objection having been made to this location and with no prospect in view 
for any other applicant, the Department of Public Works decided that 
a dry dock would be built by the Government. 



THE NEW DOCK 

In the early part of 1913 the writer was instructed to prepare plans 
and specifications on which tenders could be called as soon as possible 
for the construction of the new dry dock, the location being to the east- 
ward of the Davie Shipbuilding yard, so that both the old and new dry 
docks would be easily accessible from the shops. 

Tenders for the construction of this work were advertised on the 
12th May 1913, to be received on the 30th June following. The contract 
was awarded to the lowest tenderers, Messrs. M. P. & J. T. Davis. 
The contract was signed on the 7th October, 1913. 

418 



The new .lock was at first intended to be buill on i line parallel 
to the old dry ilnck. hut this was objected t" from the poinl of vvevt of 
navigation. A commission was appointed in the fall of 1018 to 
investigate and find out which direction would best suit the entrance 

facilit its, and it \\ as decided that the centre line of the dock should form 
an angle of ti'.t wnh the direct inn of the old dry dock, or approximately 

1" N. E. and it was so laid out . 

Owing to the limited time available before the calling of tenders, 
general plans only were prepared, together with in est iii i ate of t In' i 
The requirements as to details for the machinery and caissons were 
stated in the spec iticat ion ; the contractors were requested to furnish 
during construction all detail plans, to he submitted for approval by 
the Department. 

The dry dock has the following general dimensions. Total length 
from outer caisson to head wall 1 150 feel , divided into two compartments. 
Outer part 500 feet : Inner pari 650 feet. 

Width of entrance 120 f( 

Width at coping 144 " 

Width on floor 105 ■ 

Depth on sill at high water S.T .* 40 " 

Depth on sill at low water. S.T 22 ■ 

Spring tides rise 18 " 

Coping of side wall above high water S.T 7 " 

Floor at outer end below outer sill 4J " 

Slope of floor transversely. . 1 in 100. 

Western guide pier 400 feet. 

Eastern guide pier 500 

Depth in entrance channel at low tide 30 " 

The land expropriated in connection with the construction of the dry 
dock has a superficial area of 2o| acres, of which 11$ acres are reclaimed 
beach land. 

The outer entrance of the dock is closed with a rolling caisson, the 
top of which is provided with an automatic folding bridge; a floating 
caisson closes the inner entrance. This caisson can also be placed to 
close the outer entrance in cases when repairs are required to be made 
to the rolling caisson. 

Three main centrifugal pumps each of 63,000 gallons per minute 
capacity are used to empty the dock; two pumps of 6,000 gallons per 
minute each are used t o keep the dock dry. All pumps are run by electric 
power. Eight boilers of a total capacity of 3,600-horse-power furnish 
the steam at 200 lbs. pressure to run the three direct current turbo 
generators of 1,500, 750 and 300 kilowatts respectively, which furnish 
the current at 550 volts to run the pump and other motors. 

U9 



A direct current generator of 100 kilowatts at 220 volts, driven by 
a steam engine, will furnish the current for the lamps around the dock 
and in the buildings. There are 24. lamps of 500 watts, hung from poles 
around the dock. The poles are made of gas pipe, with the lower end 
set into sockets fitted with electric connections, and made removable 
in case of necessity. All electric wiring for lamps and motors outside 
of the buildings is placed underground. 

The approximate quantities of the materials in the principal items 
entering into the construction are: 

Rock excavation above and below coping 342,000 c. yds. 

Submarine rock excavation in channel '. . 65,000 " 

Dredging entrance channel 530,000 " 

Concrete 100,000 " 

Granite steps, altars and quoins 140,000 c. ft. 

Steel beams, reinforcing bars and manhole covers.. . . 150,000 lbs. 

Cast iron for roller casings and sluice valves 125 tons. 

Cast steel for caisson rollers 65 " 

Gun metal for caisson roller and valves 4,500 lbs. 

Cast iron in keel blocks and bollards 990 tons. 

Forged steel spindles for rollers 11,000 lbs. 

Bricks for chimney and flues 345,000 

Fire bricks ! 125,000 

Crib work in approach piers 63,300 c. yds. 

Concrete in approach piers 13,300 " 

Steel in rolling caisson 930 tons. 

Total weight in rolling caisson and machinery 1,125 " 

Steel in floating caisson 960 " 

The work was started in May 1914. The concrete retaming walls 
on each side of the dock, specified to be built from the natural rock 
surface to elevation +24 and intended to prevent seepage through the 
filling, were completed during the season's work, as well as the cofferdam 
between the outer ends of these walls. Rock drilling in the prism of the 
dock was also carried on in the part not affected by tides. The largest 
part of the drilling was done by two well drillers, the holes being sunk 
down to grade and plugged for future blasting. The average depth of 
perforation for each drill was about 80 feet per day, although as much 
as 130 feet was done occasionally. Ten or twelve ordinary steam 
drills were also used on the work. 

The rock consisted of hard shale, irregularly stratified, at an angle of 
about 45°. Considerable rock slides occurred on the west side of the cut, 
which necessitated a much larger quantity of concrete for the dock wall 
on that side, also the use of rock bolts, to prevent the sliding tendency 
of this wall. 

420 



km shovels and dump can urere used to remove the blasted rock, 
which was used for filling, wherever required, <>n t Ji<- Government 
property. 

The cofferdam was buill of ihuI.it cribwork, 20 feel wide, sunk in 
an average depth of one fool of water, at low tide, and built to the 

elevat imi of three feel above high t Lde; a layer of concrete was deposited 
along the bottom of the outer face ami this face w&b sheathed with 
plank. 

The floor and walls of the dock :ire built of concrete, the mixture 
being 1-3-5. All exposed faces are finished with a fine concrete of 1-2-4 
mixture for a thickness <>f six inches. The concrete for the walls and the 

floor was cast in alternate sections of approximately 30 feet, with 
expansion joints. 

All the cement used was subjected to a Laboratory test; apart from 
other requirements the tensile strength was required to he GOO lbs. 
per square inch after _'7 days immersion, for neat briquettes, and 
275 lbs. per square inch for 1-3 mixture. 

The steps at the top of the walls are built of granite, with treads 
and risers of 12 inches; the altars are 2 feet 6 inches wide and consist 
of granite 12 inches thick tailing 9 inches into the concrete. The caisson 
stops of both entrances and all culvert openings are built of granite. 

The floor is 5 feet thick and finished level from end to end; the 
sides slope down inches to the side gutters. The floor is provided 
with three strips of granite slabs, IS inches thick, intended to receive 
the cast iron keel and bilge blocks. The middle strip is 10 feet wide 
and level; the side strips are feet wide. 

In order to prevent the possibility of hydrostatic pressure under 
the floor and behind the side wallf m of drains is provided, that 

will take the seepage water tot he sumps. 

There are twelve stairways from the top of the walls to the floor 
of the dock, two at each end of the two compartments and two half-way 
between the ends <>f each compartment. Four timber slides, built of 
granite sh.l.s. is inches thick, are provided alongside the last set of 
stairways. There are also eight ladders, four on each side of the dock, 
that may be used to reach the floor. These are built with galvanized 
iron gas pipe, ami set in n he walls. 

The coping of the walls stands at elevation f 25, or 7 feet above 
high tide. They are provided with the ordinary cast iron bollards, 
set in concrete blocks, 60 feet apart. There are nine electrically driven 
capstans with 15-horse-power motors, four on each side of the dock and 
one at the head. 

121 



The keel blocks are each built of three pieces of castings; the middle 
piece being wedge shaped so that it may be knocked out and the block 
removed from under a ship, when in the way of repair work; the upper 
part of the top piece of casting is provided with a piece of white oak 
tenoned into the casting. All rubbing faces are planed true and smooth. 
The keel blocks are 4 feet 4 inches long and 2 feet 3 inches 
high. On top of these are placed temporary hard wood timber blocks 
to obtain the required height above the floor. It had been intended to 
build bilge blocks, so arranged as to slide under the bilge of vessels. 
However, this was objected to by the British Admiralty, who insist on 
having all blocks made of the same pattern, so as to enable building a 
bed that will conform to the bottom of the vessel. 



CAISSONS 

The outer entrance is closed by a rolling caisson built of steel and 
operated by an electric motor of 125-horse-power; the bottom is provided 
with two heavy scantlings of steel, resting on flanged rollers, 3 feet 
in diameter, placed at 8 feet centres. These are made of cast steel 
and bored to receive bronze bushings. The forged steel spindles, 4 
inches in diameter, are also provided with bronze sleeves. The cast 
iron casings, containing the rollers, are set in the concrete alters, on 
each side of the caisson berth and chamber. At an elevation of 15 feet 
9 inches above the sill of the dock the rolling caisson is provided with 
6 culverts, 42 inches in diameter, closed by sluice valves that are 
operated from the upper deck by a 15-horse-power electric motor, driving 
a longitudinal shaft provided with the necessary gearing; and, by means 
of clutches, any one or all of the valves may be worked. The culverts 
are used for flooding the dock. The caisson is divided horizontally 
by a water-tight deck at the elevation of 23 feet 6 inches above the 
bottom, forming the ballast and tidal chambers. As the tide rises 
the sea water comes on this deck through valves in the outer face of the 
caisson, which are kept constantly open during the summer to prevent 
the caisson from floating. A sufficient quantity of ballast is provided, 
so that the total weight of the structure resting on the rollers is 
approximately 150 tons. During the winter, when the dock is not in 
operation, the lower or ballast chamber of the caisson is filled with water, 
which is kept from freezing by a constant jet of steam. The 
tidal chamber is then kept dry by closing the valves. The caisson is 
closed and opened with heavy chains, supported on alters on each side 
of the caisson recess, and passing over pulleys worked by worm gears 
connected with the motor. The top of the caisson is provided with a 
folding bridge for light traffic across the dock; as soon as the caisson 
starts to open, the apron and railings of the bridge are automatically 
lowered to allow them to pass under the flooring over the caisson recess. 

422 



The middle entrance of the 'luck is closed by ;m ordinary floating 
or ship caisson. When in place, the deck is used M a bridge across the 

dork. This caisson may also be used to (dose the outer cut ranee by 

placing it immediately outside the rolling caisson, where the necessary 
stop is provided for it. This, however, will !»■ necessary only in cases 
of repairs being required to the submerged parts of the rolling caisson. 

These eaissons were built by the Dominion Bridge Company, under 

a subcontract. The mode of construction and other particulars were 
fully described in a paper read before the Canadian Society of Civil 
Engineers by Mr. L. R. Thomson, A.M.Can.Soc.C.E. Volume 30, 
Part 1, 1916. 

BOILERS AND ELECTRIC POWER 

Six water tube boilers of 500-horse-power and two of 300-horse-power 
furnish steam at 200 lbs. pressure to produce electric current. The boilers 
are provided with automatic stokers, ash and coal conveyors. The 
coal is unloaded from cars into a coal crusher run by an electric motor, 
and elevated to a hopper of 500 tons capacity, over the front of the 
boilers. Water heaters are provided, but the steam is not superheated; 
one of the small boilers will be constantly under steam pressure to run 
the drainage pumps and the lighting dynanio. 

The electric power consists of three direct current turbo-generators 
of 550 volts, one of 1,500 kilowatts, one of 750 and one of 300 kilowatts. 
The steam turbines are of the Curtis condensing type, built by the General 
Electric Company. In the large unit the turbine runs at 3, GOO r.p.m. 
It is geared down to 360 revolutions for the generator; the second is 
geared from 5,000 to 750; the third is geared from 5,000 to 900 r.p.m. 
A 100-kilowatt generator driven by a high speed direct connected steam 
engine, furnishes the current for lighting purposes. 

This power installation is more than ample for all the machinery 
connected with running of the dock proper. It is, however, anticipated 
that the whole of it will be used when large repairing and shipbuilding 
shops are in operation together with the pumping of the dock. 

This electric installation has been criticised on the ground that 
the large expenditure is not justified when electric current is available 
from private companies in the vicinity of Quebec. When the electric 
installation was proposed by the writer the idea in view was that no 
company would be interested or willing to furnish over 3,000-h. p. at 
any time of the day or night for the short period of about 60 hours in 
the year, without interfering seriously with their general service. It 
had also been ascertained by personal visits to five of the principal 
Navy Yards of the 1*. S. Government that each of them has provided 
its own electric power for pumping their dry docks. Out of five, only 



one had installed alternating current machinery. It has developed since 
that the only electric company that could furnish the power current is 
not willing to entertain the proposition unless at a much greater cost 
to the Government than the private installation can be run, including 
the interest on the outlay, which is approximately $240,000.00. 



PUMPS 

The dock is emptied by three main pumps of the horizontal 
centrifugal type, each having a capacity of 63,000 gallons per minute. 
The bronze shafts are connected to the armature shafts of 800-horse-power 
motors, running at 750 revolutions per minute. The motors are built 
to stand an overload of 25 per cent for two hours; the total lift will very 
rarely be more than 33 feet. The suction and discharge pipes are 48 
inches; the water is discharged into a chamber provided with non return 
valves, and to a culvert through the entrance wall outside of the caisson. 
The main pumps are guaranteed by the builders to deliver 63,000 gals. 
per minute against a total head of 25 feet. At the time of writing these 
pumps have not been tested as to efficiency. 

Two auxiliary pumps each of 6,000 gallons per minute capacity, 
driven by electric motors of 125-horse-power will take care of leakages 
and seepage; these pumps will also help while the dock is being pumped. 
The pumps were manufactured by the Allis-Chalmers Company. 

The time occupied in emptying the dock will vary according to the 
height of tide when the pumps are started and the size of the vessel 
being docked. At high water of spring tides the dock contains over 
38,000,000 gallons of water. This quantity of water, however, will 
very rarely, if ever, exist, when pumping is started. It is estimated 
that the average time for pumping out the dock will be about two and a 
half hours. 

Underground culverts 9 x 10 feet convey the water from the sumps 
in each compartment of the dock to the pumps; these culverts 
are provided with sluice gates, so as to permit of operating each compart- 
ment separately. The gates are operated from coping-level by 15-horse- 
power electric motors. The pressure against the gates may at times be 
due to a head of 50 feet of water. 

From the non-return valve chamber the discharge culvert is 7 x 12 
feet; it is also provided with a sluice gate. The capacity of discharge 
of this culvert was obtained from Chezy's formula V=« ^ TS , being 
obtained from Kutter's formula. Under a head of 4 inches the capacity 
will be ample to take care of the output of the pumps when discharging 
in open air. 

424 



The dock is idled through the six culverts in tl uter caisson, 

each having a sectional area of nine square feet, also two culverts, one 

eh Bide wall of a sectional :it-«-a of .'to feet, the vafves of which are 
operated by electric power. These culverts are made exceptionally 
large due to the faol thai each may only be partially opened until the 
water in the dock has reached the centre of the culvert opening, to prevent 
the heavy current that would result from a large opening from disturbing 
the beds prepared to receive a vessel; further, as the head between tin- 
outer and inner levels of water di the valves are fully opened, 
thus obtaining a large flow. The time required to fill the '!<>ek may al 
times be as much ae four hours. The middle entrance is similarly 
provided with filling culverts as the outer entrai 

In order to obtain sea water by gravity for the purpose of washing 
the floor of the dock, Bix-inch pipes were laid in the concrete Bide walls 
of the dock, at an elevation df two feel above low tide; each pipe has 
six hose-connections and valves al the face of the walls, where 50-fool 
lengths of 2$-inch hose may be attached for the purpose. The water 
is available within one hour of extreme low tide. Washing the floor is 
necessary owing to the sediment accumulated while the dock is flooded. 

GUIDE PIERS 

The western guide pier is 400 feet long and 7~> feet wide; the one 
on the eastern side is 500 feet long, 75 feet wide at the outer and 200 feet 
wide at the inner end. Each is built of two lines of 12 x 12 timber 
cribwork substructure up to six feet above low water, Bpring tides; the 
outer face of each line of cribwork is built close, and Bheathed vertically 
with 10-inch hardwood planks. The cribs facing on the channel were 
sunk in a depth of 30 feet at low water, spring tides; those on 
the eastern side of the east pier were sunk on the natural surface of the 
rock. Those on the western side of the west pier, as well as those for 
the landing pier, were sunk in a depth of 24 feet at low tide. From the 
elevation of six feet above low tide the superstructure consists of mass 
concrete walls, stepped at the back and filled 1 ated 

material. The railway spur track from the I.C.R. will be extended 
to the end of the western pier. These piers are intended to be uBed, 
when necessary, for unloading pai rom vessels to be docked. 

The entrance channel lias a depth of 30 feet at low water. 
springtides. The landing pier on the west side of the entrance is 
intended for unloading the dock supply of coal, when delivered by water. 

BUILDI1 

The power-house is 120 i LOO feet, divided by a brick wall into two 
r0OI being the boiler room and the other 

the generator room; the walls I brick, built on concrete 

foundation; the roof is buiH of reinforced ''one supported by 



steel I-beams, which were procured from the unused steel of the first 
Quebec Bridge. The building is provided with extra large windows with 
steel frames. Skylights and ventilators are also provided. The floor 
is concrete, overlaid with red tiles; and the lower part of the interior 
walls for the generator room is finished with a white tile wainscoting, 
6 feet high. Each room is furnished with water closets and wash basins; 
the water is obtained from the Lauzon village aqueduct. 

A special pump in case of fire and the necessary hose are provided. 
The generator room has an overhead travelling crane of 15 tens capacity. 
The lifting is done by motor; the travelling gear is worked by hand. 

The pump-house is 70 x 47 feet, with foundation walls of concrete, 
over which solid brick walls are built. The floor is at elevation of 16 
feet below low water, spring tides, or 41 feet below coping. It is finished 
with red tiles. The interior walls up to coping level are finished with 
white tiles. The pump-house is also provided with an overhead 
travelling crane of 10 tons capacity. The chimney is 180 feet high, 
built of brick, with an inner shell of fire-brick 100 feet high. There is 
an air space of 6 inches between the inner and outer shells; the inside 
diameter is 11 feet; the top consists of a cast-iron cap; four lightning- 
rods, well grounded, are provided to protect the chimney. 

It may be stated that the length of the dock was decided on not 
merely in anticipation of vessels of, say, 900 feet or over being employed 
on the St. Lawrence trade, which may not happen for a great number 
of years, but owing to the great number of applications received every 
fall from owners of moderate sized vessels for accommodation during 
the winter, so that repairs may be done at cheaper rates, and the boats 
be ready for traffic as soon as navigation opens. 

The dock is not yet quite completed: small portions of the floor 
and walls at the head remain to be finished; the boilers, machinery and 
pumps, although in working condition, require some final adjustment 
before they are tested and accepted; — the rolling caisson was operated 
in November last, — the contractors' floating plant was docked and the 
dock was pumped out. It is fully expected that everything will 
be entirely completed during the month of July next. 

The several classes of works in connection with the construction 
of the dock have been accomplished in a thorough manner both in regard 
to materials furnished and workmanship; several minor changes which 
were found to be advantageous were made during construction. The 
contractors, in all cases, have shown their willingness to give satisfaction 
in every way irrespective of cost. It must be noted that the works were 
started shortly before the war and continued without interruption, except 
in winter, in spite of increased cost of materials and labour. The time 

426 



required for the construction «>f the dock is somewhat over four years. 
It must, however, be remembered thai the working season is only six 
months in each year, — concrete works have to be suspended during the 
first days of November and cannot be resumed until the beginning of 
May. 

The total cost of the works under contract will be approximately 
$3,365,000.00. 

The works have been carried on by the Department of Public Works 
with Mr. Eugene D. Lafleur, M. Can. Soc. C.E., as Chief Engineer,— 
the writer as Superintending Engineer, — Mr. J. K. Laflamme, 
A. M. Can. Soc. C.E., as Resident Engineer, — Mr. S. Fort in, 
M. Can. Soc. C.E., Steel Structural Engineer, has had the approval of 
plans submitted for the steel structures. The Contractors are Messrs. 
M. P. A J. T. Davis, and Mr. 8. Wbodard is their Superintending 
Engineer. 



427 




•Q3g 3 ne) Noznvyyiooq AMQ NIVIdWVHO^/Wd 



428 




129 




-•aaaanfc 



wo *n\n--)| oQ A MQ NIVId^VHO/OWVd - 



430 




WlWiIi4ji4J4J 



iW^THTfKH: 




-: 



bl 

U 

z 
< 

H 

z 

UJ 

u 
z 

o 

r 

0) 

z 
o 

h 
o 
u 

V) 

-I 

< 

z 

5 

3 
h 

5 

z 

s 

H 
C 

£ 



- D393n6 'MOinvi-'iOOa AMO NIVldWVHDiO'M'd- 



431 




432 



*?oo 




m 

z 
o 

u 

z 

I 

o 





Z 

o 

X 

Id 



z 
< 
a 

H 
Z 

UJ 

$ 

Z 

h 

u 

in 

</> 

VI 

o 

(J 



-D3Q3n6 4 N02nvT->IOOa A«a NitMdKNVHD-DTWd- 



433 



jc?o<7 



- r 




a 
m 
10 
2 
< 
i 
u 

z 



< 
u 

o 

z 





a 

x 





a 
I* 

z 



h 
u 

UJ 

to 

Ifl 

t/) 



a 
u 



-o3aan6 *Noznvi-->iooa AUQ NiVldWVHDO'M'd- 



434 



• 






*';!•':••••'>» 


• 




n \ ' *t^*-< t ^ 


- 




II' '■■♦"'»■■« ^^^fes* 


m 


L 1 n rt F"j n ) "i r? -n rrh. 


■'. <& —I 




i 






^V^ ';-» X 














f.W ° 












■ 




:■ Ucy £ 
















• . fT*^ 


5 
















^'fl 


a 
z 


t 














•'" '« 


.J 





















'6 















tit-' 


o 

QC 

a 

z 






o 


o 






B 


4 * 


1 






«i 









m 


.oB 


o 






f 


6 






ft 
i 




z 
in 






si 


$ ! 




« 


9 


It 


z 










1 - ^ 






H 






k 


k 


"1 ^ 


o 


k 


■ 



z 






m 


* 




*J 


c 




'■»!■ 


a 






6 









■ 
•> 


' 'L 


z 








3: 

4 






3 


;4- = 


2 

















?i 


3 


"o 




o 

O 




IT 




I 


!| 


z 




yO 




S3 

* 
5; 










few 


u 

M 






TSr 
















C' 










' K ^ (/» 






Q. 




1 T~ 


Ufc 5 ^ 


J 




<S 




*o . / 








:^^ U 








»F'-' ■'« ' *■' "'" ' ' •• •'•''• • -^ 






T *" *" * * . ■• r fl-^U*'*/ \ "^ 


* 




■?;.'■;?• " ■- * ', i^C^ ?BW ^'^ 


'■» 

_J 
















.'338306 


N02nvi->iooa Ada NlVldWVHOiO'AA'd- 



435 



"i 




a. 

z 

u 

a 

s 

3 
M 

Z 
< 

a 

Q 

i/) 
to 
o 
a. 
O 

o 

z 





I 
</) 

z 


h 


UJ 

«n 
o 
a 
U 



-03S3n6 4 NOznw>J30QA«a WewieJWVHDO'Mcl 



436 




- 




-oasanfc 'Noznvi-**OOQ AdQ NIVIdWVHD-O/Wd- 







428 




J ^i_ 



*=s 



=> 



^=^ ; fc^J 



z 



>- 
u 

H 

A 

] 

a 

U 



£ 






o 



^ 



z 



h 

1 

u 




y 



I 



i 



L_L 





z 
o 

in 

< 

U 

o 

z 



o 



< 

< 

s 



J 
< 

Z 


p 
u 

ui 
01 



-oaaanb Noznvi-MOoaAaa NlVldkMVHO-DWVd 

* # 

439 




440 




- oaeanG Nozfivi-'^DOQ AfclQ NIVIdWVHO-O'AA'd 



441 




m 

3 
O 

z 
it 

I 

a 

x 
o 

D 

o 
o 

I 

o 

z 
o 

h 
U 
ul 

</) 



-33ean6 'Noznvi-'vooa AMQ NIVldWVHO^DVWd- 



442 













k=fc 






: I 




HOOJ.SN3<J 



f i\ i i ' i V v 




ID 
tfl 

D 


r 

i 

z 

D 
Q. 

X 




a 



Z 


p 
u 
u 



- 



- oaeanb Noznvi-^ooa A«Q NIVIdWVHD-OVVV'd 



443 



INDEX 

Page 

Abbott, Harry Braithwaite, (Obit.- HAM.).. . 32 

Annual Meeting 1 

Alberta Division, Report of, (R.A.M.) 73 

Alexander, Henry Ayhvard (Obit. — R \ M ;{4 

Anderson, George S. f (Obit — R.A.M. ). 34 

Anderson, Goldie Eraser, (Obit. Roll of Honour, R.A.M.) 46 

Anderson, Win. P., Li.ut.-( !oL Discussion on Fuelfl of < ianada L09 

Appendix A — Summary of Tests, Quebec Bridge L63 

Appendix B — Notes on Quebec Bridge Stresses - A. L. Darkness. . 262 

Armstrong, John Douglas, (Obit. — Roll of Honour) 46 

Barbeau, Fidele, (Obit.— R.A.M.) 34 

Belanger, Ernest, (Obit.— R.A.M.) 35 

Bertram, Brig. -Gen., Sir Alexander. Discussion on Fuels of Canada. 

(R-A.M.) ; 110 

Bertram, Brig. -Gen., Sir Alexander. Discussion on Report of 

Council. (R.A.M.) 114 

Bibliography (Quebec Bridge) 304 

Board of Examiners and Education. (R.A.M.) 79 

Branches, Membership. (R.A.M.) 8 

Branch, Societies. (R.A.M.) 9 

British Columbia Division, Report of. (R.A.M.) 72 

Butler, M. J., C.M.G. Discussion on Kettle Rapids Bridge 404 

Calgary Branch, Report of. (R.A.M.) 66 

Campbell, John James, (Obit.— Roll of Honour) 47 

Carnegie, D., Col. Discussion on Nickel-Copper Steel 387 

Challies, J. B. Discussion on Fuels of Canada. (R.A.M.) 107 

Challies, J. B. Discussion on Nickel-Copper Steel 388 

Champlain Dry Dock for Quebec Harbour 41S 

Chapman, Charles Herbert, (Obit.— R.A.M.) 35 

Committees, Report of. (R.A.M.) 10 

Conservation, Report of Committee on. (R.A.M.) 74 

DeCew, J. A. Discussion on Nickel-Copper Steel 390 

Dennis, J. S., Col. Letter of President. (R.A.M.) 3 

De Lancey, James Arnold (Roll of Honour) 47 

Dick, W. J. Discussion on Fuels of Canada. (R.A.M.) Ill 

Dick, W. J. Discussion on Nickel-Copper Steel 387 

Dickson, Adam Scott (Roll of Honour) 47 

Discussion on Committees 74, 84 

Discussion on Report of Council 5 

Discussion : On Fuels of Canada 104 

On Tests of the Chain Fenders in the Locks of the 

Panama Canal 348 

On Kettle Rapids Bridge 40i 

On Nickel-Copper Steel 384 



R.A.M.— Report of Annual Mci-tin*. 1918. 

445 



Page 

Drummond, Lindsay (Roll of Honour. R.A.M.) 47 

Drury, Edmund Hazen (Obit. R.A.M.) 35 

Duggan, Geo. H. Notes on the Work of the St. Lawrence Bridge 

Co., in Preparing the Accepted Design of the Superstructure 

of the Quebec Bridge. (Vol. XXXII, Pt. I) 17 

Duggan, Geo. H. The Design, Manufacture and Erection of the 

Superstructure of the Quebec Bridge. (Vol. XXXII, Pt. I) . . 63 
Duggan, Kenneth L., B.Sc, Major. (Obit. — Roll of Honour. 

R.A.M.) 47 

Edmonton Branch, Report of. (R.A.M.) 70 

Elections and Transfers. (Report of Council. R.A.M.) 14 

Electro-Technical Commission, Report of. (R.A.M.) 77 

Emblem of Society. (R.A.M.) 121 

Finance Committee, Report of. (R.A.M.) 21 

Financial Statement. (R.A.M.) 23 

Frame, William Layton, B.Sc, Lieut. (Obit. — Roll of Honour. 

R.A.M.) .' 48 

French, R. de L. Discussion on Tests of the Chain Fenders in the 

Locks of the Panama Canal 348 

Frost, George Henry. (Obit.— R.A.M.) 37 

Fuels of Canada, The 89 

Galway, John Campbell, Lieut. (Obit.— Roll of Honour. R.A.M.) 48 

General Clauses for Specifications. (R.A.M.) 80 

Goldmark, Henry. Tests of the Chain Fenders in the Locks of the 

Panama Canal 329 

Grey, The Right Honourable Henry George, Earl of, G.C.M.G. 

(Obit. R.A.M.) 37 

Gzowski Medal, Award of. (R.A.M.) 88 

Haanel, B.F.C. On Fuels of Canada, (R.A.M.) 89 

Hamilton, Henry Edward Raymond, Lieut. (Obit. — Roll of Honour) 49 

Harkom, John W., Col. Discussion on Fuels of Canada. (R.A.M.) 104 

Harvey, John Brown. (Obit.— R.A.M.) 37 

Hay, Thomas Alexander Stewart. (Obit.— R.A.M.) 37 

Henderson, Henry Edward Raymond, Lieut. (Obit. — Roll of 

Honour) 49 

Hill, Lieut.-Col. On Unveiling of Roll of Honour 113 

Hobson, Joseph. (Obit.— R.A.M.) 39 

Honour Roll — see Roll of Honour. 

House Committee, Report of Library and. (R.A.M.) 28 

Inauguration of President. (R.A.M.) 118 

Irving, Thomas Craik, Jr., D.S.O., Lieut.-Col. (Obit.— Roll of 

Honour. R.A.M.) 49 

Jamieson, E. A. Discussion on Nickel-Copper Steel 384 

Johnson, Phelps. On The Design, Manufacture and Erection of 

the Superstructure of the Quebec Bridge. (Vol. XXXII, Pt. I) 63 

446 



Page 
Kerry, J. G. G. Discussion on Fuels of Canada 112 

Kettle Rapids Bridge 39] 

arnell, Arthur Joseph, B.A.Sc Lieiri Col (Obit.— Roll of 

lloiHMir. I; \M !■, 

Leonard, R. u k-l-( loppi 361 

Leonard, R. W., Lieul -Col. Discussion on Fuels of Canada 109 

Library & Bouse Coi ■■•• Repoii of. R.A.M.) 

List of Members on Roll of I tonour 21 

Lowe, Edward Ja( Obit. Roll of Honour. R \M. 40 

Lucas, Frederick Travers, Q.l Obit. Roll of Honour. 
R.A.M 

Mackenzie, G. C Discussion on Nickel-Copper Steel 386 

Macpherson, Duncan, Lieut. 1 m on Fuels of Canada. 109 

Maodonald, Sir William < ' I (bit. R. \.M lo 
MacLennan, George Gordon, B >bit. K<>11 of Honour. 

B.A.M. so 

Macktin, John G. (Obit. R. \ \! id 

Manitoba Branch, Report of . R \M 55 

Maritime Branch. I R. \.M. 119 

McKenzie, William Lyon. (Obit.— R.A.M 11 

McLean, Edward Byron, Lieut. (Obit.— Roll of Honour. R.A.M. 51 

McLeod, Clement Henry, Professor. (Obit. R.A.M.) 41 

Memorial to C. B. McLeod. R \.M 5 

Meetings 8 

Moncton-Case, John, Lieut. (Obit. -Roll of Honour. R \ M 51 
Monsarrat, C. N., Lieut.-Col. On the Substructure of The Quebec 

Bridge (Vol. XXXII, Pari I I 
Morris, Basil Menzies, B.A.Sc, Lieul Obit.— Roll <>f Honour. 

R.A.M.) 51 

MungalL Andrew Norman. (Obit. — R.A M. 42 

Murdock, William. (Ob I 12 
Murray, Harcourl Armory, B.Sc, Major. (Obit. Roll of Bonour. 

R.A.M.) 

Needs, Charles Richard, Flt.-Lieut. (Obil Roll of Hoi 

Nickel-Copper Steel 361 
Obituaries. (R.A.M 

■ >tt;iw:i Branch, Repoii of 59 

Offer of Laboratory Facilities. R.A.M 121 

Papers, Report of Committee on Papers & Meetings. R AM .... 82 

entente, Committee on Roads and. R.A.M.) 76 
Plates to accompany Text of The Quebec Bridge.(Vol. XXXII.I'i. I A 

Benry Skeffingl il R. \ M 43 
r<>r'' 1 ( )n The Design, Manufacture and Erection of 

the Sup' The <,|u' \ '.■!. X XXII 

Parti) 

Portei . I leorge F. I Bridg 105 
Portland Cement Specifications I.'.A.M 

147 



Page 

Powter, Arthur Lawrence, B.Sc. (Obit. — Roll of Honour) 52 

Prize for Student's Paper, Award of. (R.A.M.) 88 

Quebec Branch, Report of. (R.A.M.) 59 

Quebec Bridge The (Vol. XXXII, Part I and Part IA) 1 

Raney, Paul Hartley, B.A.Sc, Fit. -Lieut. (Obit.— Roll of Honour) 52 

Recent Advances in Canadian Metallurgy 319 

Reinforced Concrete. (R.A.M.) 80 

Reports of Branches: (R.A.M.) 54-70 

Calgary, Quebec, 

Edmonton, Saskatchewan, 

Manitoba, Toronto, 

Ottawa, Vancouver, 

Victoria. 

Reports of Committees. (R.A.M.) 74-86 

Report of Council for year 1917. (R.A.M.) 5 

Report of Finance Committee. (R.A.M.) 21 

Report of Library & House Committee. (R.A.M.) 28 

Report of Scrutineers. (R.A.M.) 117 

Roads and Pavements. (R.A.M.) 76 

Roll of Honour, List of Members of. (R.A.M.) 21 

Roll of Honour, (Obituaries). (R.A.M.) 46 

Ross, R. A. Discussion on Fuels of Canada. (R.A.M.) 104 

Roy, Joseph Rouer. (Obit.— R.A.M.) 43 

Saskatchewan Branch, Report of. (R.A.M.) 68 

Schneider, Charles Conrad. (Obit. — R.A.M.) 44 

Sewage Disposal (Report of Toronto Branch Committee. — R.A.M.). 86 

Shearwood, F. P. Discussion on Kettle Rapids Bridge 403 

Shortt, Adam. Discussion on Nickel-Copper Steel 389 

Society Affairs. (R.A.M.) 81 

Speakman, Richard Edward. (Obit. — R.A.M.) 44 

Spelman, James. (Obit. — R.A.M.) 44 

Stansfield, Alfred, D.Sc. On Recent Advances in Canadian Metal- 
lurgy 319 

Steam Boiler Specifications. (R.A.M.) 80 

Steel Highway Bridges Specifications 78 

Tests of the Chain Fenders in the Locks of the Panama Canal 329 

Thomson, W. Chase. On Kettle Rapids Bridge 405 

Tooker, Noel Longfield. (Obit.— Roll of Honour. R.A.M.) 53 

Toronto Branch, Report of. (R.A.M.) 58 

Unveiling of Roll of Honour. (R.A.M.) 112 

Unwin, Wilfrid Peyto, Lieut. (Obit.— Roll of Honour. R.A.M.)... 53 

Valiquet, U. On Champlain Dry Dock for Quebec Harbour 415 

Vancouver Branch, Report of. (R.A.M.) 54 

Victoria Branch, Report of. (R.A.M.) 64 

Whiteside, James LeRoy, B.A.Sc, Lieut. (Obit. — Roll of Honour. 

R.A.M.) 53 

Wright, Percy Andrew, 2nd Flt.-Lieut. (Obit. — Roll of Honour. 

R.A.M.) 54 

448 



1929 




l 

E5 






ring Institute of 
Cer nsactions 



" A«u>m*u u" 



T-^ 



Engineering 



PLEASE DO NOT REMOVE 
CARDS OR SLIPS FROM THIS POCKET 

UNIVERSITY OF TORONTO LIBRARY 



ENGIN STORA9?