(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)
See other formats

Full text of "Modern Mechanical Engineering Vol-I"





M.Inst.C.E., M.I.Mech.E.



C.B.E., M.Inst.C.E., M.I.Mech.E., M.I.E.E.


66 Chandos Street, Covent Garden, London

This work contains a general account of the art of mechanical
engineering, which has for its object the harnessing of all natural
powers for the service of man. So vast a subject cannot be described
exhaustively in any single work of reasonable size. Selection must
therefore be very carefully made. The principle of selection has
been that of keeping the practical requirements of engineers as much
as possible to the fore, leaving aside current speculation as t©
possible new sources of energy, new prime-movers, and so on.
Theoretical matters are discussed in their proper places, and an
effort is made to present these subjects as concisely and clearly as
possible to readers who are likely to be interested in them.
Volume I deals with the organization of a modern works, be-
ginning with the DRAWING OFFICE and going through the PATTERN
the finished machine.
Volume II begins with a section on the TRANSPORT OF PLANT,
a branch of the subject which deserves, but seldom receives, ade-
quate treatment in books on Mechanical Engineering. This is
followed by a section on PIPE-WORK, a subject of vital importance
to operating engineers, as anyone who has had to operate a plant
with a faulty pipe system must know. Then follow three sections
on theoretical subjects: APPLIED MECHANICS, ELASTICITY OF
In Volume IV will be found sections on MECHANISM, MACHINE
Volume V is devoted mainly to STEAM ENGINEERING PLANT,

vi                                        PREFACE

OPERATION OF LAND POWER PLANTS. In this last article the needs
of power-station engineers have been kept more especially in view.
This volume also contains a section on ENGINEERING CHEMISTRY.
Volume VI is devoted to Internal-combustion Engines, and




CHAP.                                                                                                                                         Page
INTRODUCTION.........       3
I.   THE ESTIMATING AND COSTING OFFICES        -       -       -       -       4
II.   THE DESIGN OFFICE   -                ......     13
III.    THE DETAIL OFFICE   -       -......     17
IV.    TRACING OFFICE........     35
V.   THE PHOTOGRAPHIC ROOM -......     37
I.   THE ELEMENTS.......--52
1.  Methods of Moulding......     52
2.  Pattern Construction......     58
3.  Core Prints and Core Boxes                                       61
II.   EXAMPLES OF WORK........     66
1.  Cylindrical Work......     67
2.  Sheaves, Pulleys, and Flywheels                                 77
3.  Gear-wheel Patterns   ------     80
4.  Machine-made Wheels.....     87
5.  Beds and Allied Forms        -----     89
6.  Screws      -        -        -        -        --       •       -92
7.  Plated Patterns.......     96
8.  Sectional and Skeleton-like Patterns                            98










II.   MOULDING IN DRY SAND   -       ...


IV.    CORE-MAKING......








-  in

-  H3

-  125

-  126

-  131

-  137

-  139

-  H5

-  14?

-  *49

-  J59

-  172




II.    THE TOOLS......



V.   THE SHOPS        ------

-    177

-    180

-    198

-    211

-    230




JOINTS      ------


-    243
-    246
-    262
-    264



Editor of u The Draughtsman "

Vol. I.
3     LIBRARY

V.   **

Drawing-office   Organization

The draughtsman and his work form an essential link in the chain of any
engineering organization. He not only indicates the finished article desired,
but largely the routine and progress of the work, and his drawing is, in
effect, a series of directions as to how the work is to b6 done. When the
drawings or order sheets leave his hands, the method of distribution is so
arranged that they shall go to the various shops and yards in a certain definite
order and time, so that there shall be no hitch, neither forgetfulness nor
overlapping, and if possible some system of checks is introduced, so that,
where forgetfulness may on occasion take place, the omission can, be made
good with the minimum loss of time and cost.
There is a tendency on the part of some to regard drawing-office
work, and, in a lesser degree, pattern-shop work, as unproductive labour.
This argument, if true, would require to be logically extended until we
came to the fact that, as all machinery was only a means to an end,
the making of machines themselves was unproductive. But employers
would be very unwise if such crude materialism were allowed to interfere
with good and even elaborate staff work, as it is in the initial stages of
design that the largest ultimate economies can be effected. The best
firms realize this, and neither stint staff nor equipment, and make large
allowances for experiment, research, design, and administration.
It is necessary, therefore, to consider very carefully the detailed organiza-
tion of the drawing office, and the routine to be observed from the time the
work enters as an estimate or a contract until its final dispatch, for the work
of the drawing office is not completed with the issue of the drawings to the
shops. An endeavour will be made in the following pages to portray a
form of up-to-date organization which, if carried out and possibly extended
to suit particular circumstances, will enable a firm to compete successfully
against less efficiently organized rivals.

The Estimating and Costing Offices
Costing System.—Large modern businesses keep the estimating
department distinct from both the design and working drawing offices, but,
of course, with absolutely free access to the working drawings, sketches,
orders, and data-books of the drawing office, and with access to the books
of the costing department. The costing department is sometimes run as
a branch of the estimating department. A good costing system is the founda-
tion stone of successful estimating. To neglect this important section—
and it is shamefully neglected in all but comparatively few firms—is to run
very serious and quite unnecessary risks. " The working of this section
would demand a monograph on itself to describe it: it is taken for granted
Scanty Material for First Estimates*—In estimating for new
contracts, one is usually given only a few general outlines in the first case;
indeed, there may be three or four tentative offers before the definite con-
tract is fixed. This is very noticeable in large ship contracts, where the
shipowners have to study carefully a great number of problems before
coming to a definite decision, such as the quantity and kind of trade anti-
cipated, dock accommodation, fuelling facilities on the proposed route,
repairing facilities, tides, and consequently most economical speeds, &c.
Many tentative schemes must be submitted, and it may take months before
a satisfactory conclusion is reached.
This sort of thing makes it essential that the estimator should enter
against each estimate the date on which it is made, so that, by the aid of
his " material " charts, he may be able to make the necessary corrections when
the final order comes along. For instance, there may be one or two pounds
of an increase per ton of steel between his first estimate and the time when
the contract is fixed. On a ship, say, in which there may be several thousand
tons of steel this makes a considerable difference in price.
Tendency of Market Prices.—It is usual, unless the firm's accountant
or buyer is convinced of an impending fall in the market, to fix the sub-
contracts as early as possible, both for reasons of cheapness and to secure
priority of supply. It may happen that it is impossible to be sure of an
accurate estimate, particularly if the work be absolutely fresh, and if no
particulars be available for a similar class of contract. This is not an unusual
circumstance in bridge-building, say, or in the design of a very large and
speedy liner. In this case it is usual to leave the estimate a little " lucky ",
but discrimination must be used in this matter, and attention paid to the
" tendency " of the market, whether moving up or down.
Register of Weights of Previous Jobs.—A very important item, if
correct estimates are to be forthcoming, is that the register of weights of

previous jobs, complete and in detail, should be faithfully kept and tabulated.
Part of this tabulation is usually done in the working drawing office, and part
of it either in the counting-house or, better still, the separate costing de-

A book of finished weights should be kept, showing the weight of each
part or unit which it is practicable to consider separately. For instance,
suppose it is a reciprocating engine installation and perhaps boilers, the
record of this installation may be set down thus:

Estimated I.H.P.,
Size of cylinders,
Number and size of boilers,

Cubic feet of cylinders' capacity,
as a heading, which could be amplified considerably.
Then would follow headings of the principal parts, such asc
Main engines.
Fittings on main engines.
Auxiliaries in engine-room and in boiler-room.
Fittings apart from main engines.
Shafting, &c.
Fittings in boiler-room.
Water in boilers.
Condensing plant.
Water in condensing plant.
Spare gear.
Refrigerating plant.
Electric plant.
Each of these should be divided into as many sections as is found desirable.
As an example again, take the heading " Main Engines "; this would be
split up into sub-headings something like the following:—.
Air-pumps, &c.
This "finished weight" book should be carefully indexed.
Style of Job.—Again, the cost of a job depends a great deal upon the
characteristics of the intending purchaser and his firm's practice.
This may be illustrated by considering three firms known to the present
writer—all well-known shipowners.

Taking 100 as the index figure for a normal, plain, and straightforward
job, it was found that:

Firm " A " demanded a high finish, the use of brass where the normal
practice was cast iron, and many refinements and specialities. It was
found in this case that almost 50 per cent had to be added to the esti-
mated cost for the normal job.

Firm " B ", on the other hand, insisted not so much on specialities and
finish as on weight. Shafts had to be 20 per cent over Board of Trade
requirements; in approving drawings, thicknesses were generally increased,
resulting usually in a net increase of weight of almost 10 per cent.

Firm " C " got on the whole a very plain job, but the job he finally got
was seldom the job he originally ordered. It was usual for this firm to have
several ships building in different yards, and changes in the job were con-
stantly being made, and were a source of vexation and annoyance and extra
expense. Further, the changes involved extra cost, but it was very difficult
indeed to get these extra expenses paid. As the purchaser was a good
customer it was not considered politic to insist always on payment for these
extras, so that an extra 3 per cent was generally added to the estimated
price to cover these anticipated vagaries. This may not seem to have much
to do with scientific estimating, but it led to the erection of charts based
on pounds per indicated horse-power which looked something like the











3,000      4pOO

6,000     7,000      SjOOO      9,000     10,000     4000    8,000     0,000    MOOO

Fig. i. — Chart of Costs per I.H.P.

This chart shows a method of checking the detailed estimates.   It is
largely based on estimates received from sub-contractors, i.e. prices based

on overall weights, on indicated horse-power, on volumetric capacity, or on
some other recognized unit of measurement.

The estimator should keep an estimate book, and it is good practice to
keep a separate book for each job, if it is of any size.

Each item should be shortly but clearly specified. The estimated
weight should be put against it, the cost of the raw material, the estimated
time of workmanship and the rate of pay, and a column should be written
up for the total cost, thus:

	Price per Ton.
	Material Price.

	Bloom D

	steel, B.S.S.
	i c. 2q. 21 Ib.
	2C. i q. 7lb.
	£10 10
	£n 10

An extra cost would be added, because the total cost of machining is not
merely the cost of the actual ingot plus the labour charge—there is an " over-
head charge " to be added.
Specialities of other Makers.—In the case of specialities, such as
auxiliaries and furnaces, the prices of different makers would be put down,
although only that of the successful tenderer would be carried forward to
the finished column. This would enable the work to be referred back to
expeditiously, in case of a change later on. The customer, for instance,
might prefer to pay more to get some particular make of boiler feed-pump.
Final Estimate Book.—From this book the finished estimate would
be made up, and oncost charges and profit would be added. The finished
book containing these two items is confidential, in many places the manager
reserving the care of this book to himself.
Should a contract be concluded on the basis of this estimate, the details
as they are actually finished should be entered up in an abstract book, in which
double columns should show the estimated weights, costs, &c., alongside the
actual weights and costs. Only thus can the work of the estimating office be
properly supervised and checked.
Scales of Wages, Rates of Mechanical Operations, &c.—A scale
of wages for different classes of work must be kept, also rates of speeds at
which the work can be turned out; say, in machining, the table should give
the surface which can be rough turned in a given time, also the rate for finish-
ing cuts, &c., where this process obtains. In other classes of work a piece-
work rate per 100 rivets may hold, or the rate may be so much per foot for
smithwork on angle-iron, or, in the forge, a price per hundredweight for
" light" and a price per hundredweight for " heavy " forgings. This all
implies that the estimator shall make himself thoroughly familiar with his
own shop practice. It may be necessary for him to get estimates from the
foremen, or from rate-fixers, but so far as possible this information should
be tabulated inside the estimating office, and as little reliance as possible
should be placed on shop estimates, because shop conditions are peculiarly
unfavourable to the accurate making of estimates.


Material Charts.—Material charts on squared paper should be kept
in the estimating office, and the day-to-day fluctuations marked thereon.
Each chart should extend for a quarter- or half-year, but when taken down
it should be carefully filed, as, over and above the market fluctuations, it will
be found that there are general seasonal fluctuations which it may be advis-
able to take account of in fresh work.

The material charts shown are completed for a full year. It will be
understood that this curve will be gradually traced either by noting daily

TIN ( English   Ingots)

£per ton
 120 115 110 105 100 95 90 85 80 75 70






	' (S






£per ton
 380 360 340 320 300 280 260 240 220 200














	A i

	\ A
	\ t
	* _

	\ /








or weekly fluctuations. As a rule,
the curve is what is commonly
called a step curve, as shown for
iron, and is generally marked by a
clerk, who watches the markets
from day to day and observes the
current quotations. Notes of special
circumstances may be shown in
the margin. It will be known from
the data-books how much percen-
tage of these metals enters into total
weights, the amount that must be
credited to workmanship per ton of
metal bought, &c., so that the intelli-
gent use of the charts may bring an estimate out reasonably close to actual

On Charges and Profits.—The last two items in the estimator's
calculations do not call for any special remark, that is the addition of over-
head or running charges and profit. These are generally fixed percentages
which are under the direct control of the management. These percentages
may be whittled down considerably to gain contracts in a strong competitive
market, or even done without, if it should seem advisable, in a period of
depressed trade. It may be better to keep the machinery running, even

Per -ton



240/ 230/-22O.A




2IO/-200/ I9ty-i«r>/.

	— f







Fig. 2.—Cost Charts of Material

at some loss, rather than dispense with an organized staff ofxmanagement
and operatives.                                                                   *         """"— ~"~

Buying. — In large engineering and shipbuilding worfeg 4 is aH
common practice to keep the buying of material in the estimating depart-
ment; but, where this is done, often the general instructions to buy particular
classes of goods are issued by a financial manager, whose special province it
is to look after the whole commercial side of the firm's activities, leaving
the technical side to the technical manager. He keeps his eye on the market,
and only sends up such a general instruction to the buyer as: " Buy all
copper tubes required for next six months ".

The buyer's abstract book should show him, at any time, what material
is to be ordered, when it will be required by the shops, when ordered, when
to be delivered, terms, &c., and it should always be kept thoroughly up-
to-date and properly indexed day by day.

Filing System. — Needless to say, a good system of filing for corre-
spondence, returned estimates, and prints should be adopted. This will be
discussed more fully under the heading of correspondence.

The date of sending out for estimates should be carefully noted, also
replies tabulated. The date of fixing sub-contracts must be entered as
well as the terms. Moreover, a duplicate copy of all prints sent out should
be kept, and a fresh copy of the working drawing, initialled by the chief
draughtsman, should be sent with formal acceptance.

It may sound very much like an advocacy of the use of red tape, but a
very rigid and formal procedure should always be adopted in receiving and
dispatching prints. Everything should go through the head of the depart-
ment concerned, and it is a good plan to keep little order-books for anything
required and have the prints initialled. The form of the book, which should
be carbon duplicated, might be:

Job No.
	Copies Required.
	Drawing No.

	581/1068     { ^
	Main boiler furnaces

(Signed) E. Forster, estimator. (Countersigned) E. Harper, chief draughtsman.
This slip would be handed in to the chief of the detail department, who
would give it to the proper section leader, who would say whether it was
up-to-date. When this was verified, the slip would be given to the photo-
grapher, who would take the requisite prints and give them back to the chief
draughtsman for initialling before being sent into the purchasing
Commercial Information and Contract Law.—The buyer should
make himself familiar with the requirements of the departments and their
peculiarities, and should see that a copy of his purchases and contracts are
'given to the estimator and the ordering clerk. It may even be advisable
to send those also to the drawing office, and the department to which the
goods are to be consigned, for checking purposes. It is desirable that he
should make himself familiar with terms commonly used in shipment and
carriage of goods; also with the Law relating to Contract and Sale of Goods.
The importance of the 1893 Law cannot be overlooked, and a copy can always
be obtained from H.M. Stationery Office.
At all times the estimating and buying departments should be self-
contained, with their own clerks and typists, as most of the work is of an
essentially confidential nature.
Specifications.—The work of this office will generally cover the
preparation of detailed specifications for the owners as to what is being
contracted for and supplied. In specifications to sub-contractors the most
detailed and clearest specification possible should be aimed at, as it is unsafe
to leave out any detail which can be mentioned at all.
In calling for estimates from sub-contractors, it is usual not only to
specify the requirements very fully, but in every case to clearly state the
time, method, and date of delivery required, also the standard conditions
of the firm regarding invoicing, receipt, and payment, and their practice
regarding delays due to accidents, such as fire, industrial disputes, &c.
These clauses, of course, will naturally be added to the main contract in
order to cover the firm with regard to the purchaser. The departments to
which deliveries are to be made must be clearly stated, and a copy of order,
with price deleted, sent to department.
It is a good plan to insist that, on every invoice and receipt-note sent in
with the delivery of material, each item shall be stated, and its actual weight.
These weights should be at once transferred in the invoice department to
a book specially kept for this purpose, and to books or specially prepared
sheets in the different departments, which books or sheets will be sent in
regularly to the estimating office or detail office, whichever has charge of
the data-book, so that the data-book may be gradually filled up as the job
proceeds. If such entering-up is left to the end of the job, much hurry,
confusion, and delay may ensue, particularly if the contract be a lengthy one.
Inquiries for estimates should be sent out on standard inquiry forms, say
on white paper. When it is decided to accept a particular tender, the accep-
tance should be again fully detailed and kept on different-coloured sheets,
say yellow tissues. These acceptances should be kept separately and filed
by themselves for each particular job.
Apart from clerks and typists, the estimating office should be staffed
with men who have had good general drawing-office experience, able to
understand drawings quickly. All prints sent out for prices should be
returned with same, and a note put on the inquiry form to this effect. This
procedure is adopted for two reasons: firstly, so that the firm's drawings
may not get into the hands of people for whom they are not intended; and
secondly, so that, when the contract is fixed and being executed, there shall
be no possibility of work being finished to drawings which may have been
intended for prices only, and which it may be desirable to amend as working
drawings. No sub-contract should be begun until working drawings are
received, and a note to this effect put on the acceptance.
A good estimator must watch carefully and try to arrange to do with
the minimum extras, or, where the use of these cannot be avoided, he must
allow for same. The difficulty may be exemplified in boiler plates, for
instance, which are not simply so much per ton. There is a basic price
per ton, say .£20. An extra of 2s. 6d. per ton per 3 in. over 8 ft. broad may
be payable, another extra per 5 cwt. on any plate over 4 tons weight,
another for certain surveys, one for different thicknesses of plates, and
one for tensile strengths, &c. It will be clear from this how important it
is to keep in close touch with the design office; for instance, in a large boiler
installation, the shells may be specified in one or two strakes, and the differ-
ence in prices so caused may be hundreds or even thousands of pounds.
Tables of these extras can be got from steel-makers, and up-to-date lists
should be kept in the office. A good designer, by skill and knowledge of
these extras, may save almost incredible sums to the firm.
Concurrently with the preparation of the estimate should go the drafting
of the specification and necessary schedules. This will be submitted with
tender drawings to the purchasers, and the schedules will be completed
when the contract is entered into, and the whole document will become
the guide and general working instruction for the detail office. These
specifications are usually printed, and a copy is given to each of the main
departments. Copies are carefully executed by the contracting parties
with full legal formality. The preparation of specifications is a very re-
sponsible job. Many firms have standard specifications, which are used as
a basis for the preparation of those finally approved.
Standard Specifications.—During the progress of the late War, an
increasing need was felt for the standardization, not only of parts within
a firm, but for standardization in relationship to materials, tests, &c. As
an instance, it may be stated that the manufucturers of steel plates for
boilers and ships sent out lists of standard plates they were prepared to
roll, and which, tested at the works by the surveyors of the classification
societies, could be had as if from stock, and with the minimum delay.
The previous multiplicity of requirements and tests made the work of
both designer and detailer very onerous, and not infrequently led to very
costly construction.
These differences have been realized by practical draughtsmen for years,
but it was the urgency of the War that forced on reform. Recently the
Board of Trade, Lloyds, and British Corporation, have combined to make
their rules for boiler construction identical, a vast reform when it is
remembered that previously every rule was different, and most work for
first-class jobs was made under two surveys at least, scantlings to suit the
Board of Trade often meaning an increase of weight of about 5 per cent
with steel at £20 per ton.
12                   DRAWING-OFFICE   ORGANIZATION
Another side on which much progress has been made is in the pre-
paration by the British Engineering Standards Association (incorporated in
1918) of standard specifications.
The Association has issued, up to the time of writing, 136 Standard
Specifications and five interim reports on Ball Journals, Screw Threads,
Tyres, &c. All classes of work are covered, and a catalogue can be
received by sending a postcard to the Association at 28 Victoria Street,
London, S.W. i. The Specifications themselves cost is. Specifications
are in course of preparation for aircraft materials and components, and
a number have already been issued.
The Costing Department.—This department comes more properly
under the purview of the clerical staff, but it very closely affects the
efficiency and accuracy of the drawing-office work, and particularly affects
all future estimates. It is fairly common to find it in charge of a man who
has had a good technical training in the drawing office.
How the department is actually organized depends, to a very considerable
extent, on the methods of accountancy and stock-keeping adopted in the firm.
In the first place, by keeping a proper set of books, it should be possible
at any moment to get the actual cost of all the items delivered against a
job. Any material taken from stock should be shown on these books, as,
if this be not done, a particular contract might appear to work out cheaper
than it really is. Moreover, it is necessary to show what has been taken
out of stock, if the amount of stock carried is to be maintained at a level
which will enable emergencies to be met easily.
So far as the cost of labour is concerned, separate columns must be kept
for the different classes of work involved, also a statement must be made of
whether time, piece-work, or premium bonus is the method of payment.
If a premium bonus system prevails in the shops, the rate-fixing depart-
ment is usually attached to the costing department, and this department
may attain to very considerable dimensions with an extremely elaborate
system of books.
It would take us too far afield to discuss all the items involved, and
indeed it is unnecessary, as the system does not prevail to any large extent
in this country, and at present there is no indication of any considerable
The costing department has been mentioned here only because, when
a costing system is introduced into any firm, its main outlines are generally
worked out in the drawing office, although the details are left to the clerical
staff, and because the opportunity seems favourable to emphasize the neces-
sity and desirability of having this department put on a well-organized and
scientific basis, with a view to the facilitation of drawing-office and esti-
mating work. The costing system and costing department in most firms
are of the most rudimentary description. Temporary expedients and make-
shifts have been adopted as the business grew, even where it would have
paid over and over again to call in the assistance of a good accountant to
make a thorough overhaul and examination of what would require to be
THE  DESIGN  OFFICE                               13

done to put the system on a sound basis. One well-known firm of engineers,
of no great eminence fifteen or twenty years ago, created at the time quite
a large costing and estimating department, thoroughly examined all the
processes of production, tested the capacity and efficiency of machines, and
installed new ones where necessary, so that now these departments work
like clockwork, and costs, the firm claims, can be known almost to a penny,
with the result that even in very bad periods the firm is practically never idle.
It has reduced costs to a minimum, and is able to meet successfully the com-
petitive market prices.

The Design Office
Closely related to the estimating office is the design office, or, as some
large firms prefer to call it, the scientific department. This is a department
that only exists as a separate entity in large progressive, up-to-date establish-
ments whose line of business is largely influenced by fresh theory and inven-
tions and is dependent on experimental research. It comes to mature growth
where large prime movers are manufactured, or very variable structures such
as ships are built. The same need does not arise in structural steelwork,
for instance, though many large structural steel firms do maintain such an
office. In such cases they are used largely as estimating offices, and it is
probably in this direction, rather than in that of research, that their main
importance lies.
Technical Questions dealt with.—The design office is usually much
smaller than the detail office, and has its own departmental head. It is the
function of this office to prepare the original drafts and sketches for new
work; to estimate the quantities of material required; to ascertain how far
specified requirements can be met. When the main lines of design have
been sketched out, the quantities are estimated, and technical questions,
such as stability, if the job is the building of a ship, are looked into.
This office, working in close conjunction with the estimating office, will
prepare the tender drawings and sketches, and will generally feed the esti-
mating office with fairly detailed technical information. If the contract be
placed, the design office will lay down the main outlines of the job, and will
generally fix the principal dimensions and scantlings, calculate stresses on
parts, and tabulate them in an easily accessible form.
Stress-book.—The stress-book is a highly desirable and valuable
record. It should be kept as part of the office work and completely up-
to-date by the checker or section leader. If this book be not kept regularly
and carefully, the very valuable comparative data kept by the individual
draughtsman may be utterly lost to the firm, if such an individual should
cease to be employed with them.

';. i

Breaking up the Job.—The job can mostly be broken up into a
number of well-defined and easily recognizable units, each of which can
have a separate section in the data record. These units will be further sub-
divided, and under each subdivision will be indicated the job identification
number and the important dimensions. The material to be used should
be entered, the thickness, the working pressure for which the part is designed,
the test pressure, the stress per square inch at working load, &c. It is a
good plan to mark in the data-book, at the head of each section, the charac-
teristic formulae applied, and also the safe working load.

Data-books.—It is usual to keep a list of significant dimensions and
arrangements, because frequently the thing wanted most rapidly from the
design office is the arrangement of an installation, and the accommodation
which will be necessary to house it. At the head of such a section should
be a list of normal clearances. A specimen page of such a section is given
herewith for a marine engine and boiler installation, which will fully explain



JJ.        L.RH


A =  distance between H.P. and L.P. centres.

B =  distance between M.P.j and M.P.2 centres.

C =  distance between M.P.2 and L.P. centres.

D =  distance to outside of H.P. casing + 3 ft.

E =  distance to outside of M.P.j casing + i ft. 9 in.

F =  distance from C.L. of cylinders to C.L. of shaft.

G =  distance from C.L. of shaft to tank top + 2| in.

H =  A + distance to H.P. casing + distance to L.P. joint + i ft. 9 in.

J =  2 distance from C.L. to column foot + 2 ft.

Job No.
	Size of Cylrs.


Fig. 3.—Specimen Page of Data-book (i)
N.B.—Sketch, table, and column all form page of Data-book in each case.


A =   10 ft. 6 in.

B =  length of boiler.

C =  2 ft.

D =  length of boiler.

E =  10 ft. 6 in.

F =-.  length of cross bunker.

G =  mean diameter -f- i ft. 9 in.

H =  mean radius + 2 ft.

J =  overall length of boiler room.

K =  2 ft. 6 in.


Job No.
	Cub. Cap.

Fig. 4.—Specimen Page of Data-book (2)
Special circumstances will always call for special arrangements, but a
few normal figures, worked to approximately in every case, very considerably
lightens the designer's task, and avoids, as far as possible, chances of large
and serious errors. With a system of book-keeping highly elaborated, such
as this, the main features of an arrangement design could be sketched on
the back of an envelope, and the shipbuilder enabled to make his arrange-
ments accordingly. Even where such a complete record is kept, it is still
advisable to go very closely into new designs and check results. Nothing
should be left to chance. The designer should know almost instinctively
what clearances to test and what scantlings are suitable.
Leading Particulars of Job.—An important work of the design
office is the preparation of a fully detailed sectional drawing, showing the
important dimensions and scantlings necessary to obtain the certificates of
some of the classification societies, which are frequently necessary before
the installation can be insured. Typical examples of these are the midship
section of a ship, main steam-pipe installations, and marine boiler design.
In the two latter cases, however, the working out of these arrangements
and scantlings is left to the detail office, the demarcation of what shall be
done in each office being a matter of internal arrangement. The classifica-
tion societies generally considered are the Board of Trade, Lloyd's Registry,
the British Corporation, and the Bureau Veritas. In structural work on land,
the span of bays for overhead cranes, the distance between columns, the
scantlings of columns, crane rails, the size of foundations, the thickness of
16                    DRAWING-OFFICE   ORGANIZATION
retaining walls, and block plans have to be prepared for the local Building
Authority. These rules are generally very rigid, and must be closely adhered
to. If the contractor fail to comply with them, he is liable to be asked to
pull down much of what he has erected and to build afresh. It is therefore
important to get the plans approved by the competent authority as early
as possible.
Tests.—Another important aspect of the work of the design office is
to attend at all tests, and to collect and collate the results of them for future
guidance.   It is very essential to note very carefully the conditions under
f|                               which any test is carried out.    These conditions should be all carefully put
n                              down on the standardized data-sheet, in which the results themselves are
!                                shown.    The usual method of keeping these results is to have white prints
I                               made from a tracing, which show all the various items to be noted.    The
I,                              figures and remarks are marked on these sheets in pencil, and the whole sent
|                               into the tracing office and traced.    Photographs, either in white or blue,
I                               can then be filed for reference.    In addition, it is usual to enter the more
I                               significant items in a book, as, if dependence is placed absolutely on loose
I                               sketches, there is always the possibility of some of them being misplaced or
I                               lost.    It should be the duty of someone in the office to see that all data-
|                               books and sheets are carefully put away at night in the fire-proof safe gener-
'£,                              ally provided for this purpose.
1                                     It is usual to plot test results in a graphical form, and to find how much
;[,'                              any particular job may vary from normal practice, and if necessary to bring
F|I                              that normal practice up to date.    It may be found, for instance, that for
ftp                             some special reason higher working-stresses than usual have been used.
^                              If this special practice be repeated on several occasions, and the results are
');                             found to be satisfactory, it may be possible to bring the normal practice
into line with this special practice, and to alter the basic formulae accordingly.
Functions of Design Office.—The main detailed drawings and cal-
culations should be submitted to the design office, in order to ensure that
the general principles of the design have been carried out.
It will be seen that the design office has a double function in the pre-
paration of designs. In the first place, it has to prepare designs for esti-
mating purposes. The design for this phase must be accurate, but gener-
ally need not be given in so much detail as when prepared for the detail
office. Indeed, when prepared for the latter, it may be found highly desir-
able to considerably modify it so as to suit existing patterns, standard gauges,
templets, and conditions, which could not have been foreseen when the
original draft designs were prepared. The " estimate design " itself may meet
with considerable alteration at the hands of the purchaser.
Necessity for Full Information.—It is essential in the design office
that the fullest possible information should be put before the draughtsmen,
both in the shape of correspondence, similar designs from the firm's own prac-
tice and from elsewhere, and the latest scientific and technical information,
either in technical publications or the proceedings of learned societies.
Staff and Discipline.—From what has been said regarding the arrange-
THE   DETAIL OFFICE                            17

nient of the estimating and design offices, it will be observed that much lati-
tude must be given to the highly skilled men employed in them, in regard to
freedom of movement, opportunity for observation in shop or on site, and
time taken to particular portions of work. This does not mean that discipline
need be more lax in these offices; only that it must be of a different kind.
It is very probable that the administrative head, whether he be a chief over
the whole of the offices or a manager, will spend a considerable proportion
of his time in these offices. It pays to staff, and even slightly overstaff,
these departments and give them the maximum facilities for carrying on
their duties. Heating and lighting are by no means negligible factors, as
also satisfactory arrangement of the offices, lavatory accommodation, record-
ing and special instruments, &c. These remarks apply also, in degree, to
the detail offices, which it is now proposed to describe. Much of what will
be described in the next section applies to these offices, and has simply been
omitted so that what is common to them may be treated all at one time.

The Detail Office
Organization.—Whether or not the firm be large enough to support
separate estimating and design offices, it is-certain that the detail office must
always exist. In size, it is generally reckoned as the main office, and it is
always responsible for the issue of directions to the various shops in the
shape of drawings, order-sheets, standards, &c.
The detail office itself is generally split up, in certain classes of work,
into two or three more or less well-defined departments. In an electrical
establishment, it may possibly be that one portion will deal with the mechani-
cal design of the motors, dynamos, commutators, transformers, &c., whilst
another portion will.deal with the general installation, placing of switch-
boards, wiring, &c. In a ship office, we may have a section devoted to the
steelwork, another to piping arrangements, and yet a third dealing with
accommodation, including shipwright work, upholstery, &c. In a land or
marine engineering establishment, the sections will probably be a turbine
department, reciprocating-engine department, pipe and machinery arrange-
ment department, and a boiler department. The usual procedure is to
have a chief over the whole office, with an internal office with clear windows
looking out on to the main office. Under him, and working near him, will
be the assistant chief, who will generally look after the discipline of the
office, give out work to the section leaders, and correlate their work and gener-
ally approve of the finished drawing, discussing points of peculiar importance
or difficulty with the chief. All the correspondence will come through
him to the section leaders. It will perhaps help to make our meaning clearer
VOL. I.                                                                                                                 2
18                   DRAWING-OFFICE  ORGANIZATION
if we describe in some detail the working of a large marine engineering
establishment. We shall describe the routine of the work, including the
preparation of drawings, the circulation of correspondence, the methods
of ordering material, the issue of drawings to the shop, &c. This section
will be treated in much fuller detail than the previous sections, as, by doing
so, we shall be describing at the same time much of the routine work of
the estimating and design offices. The routine of the office having been
fixed, it must be strictly adhered to, and only departed from for very special
reasons and with the full knowledge and concurrence of the responsible
head. This is absolutely necessary if overlapping, omissions, and friction
are to be avoided.
The Chief Draughtsman.—This official has the ultimate and sole
responsibility for office discipline, and both the qualitative and quantitative
production of the work. He must see, as far as he can, that the proper time-
table is worked to, and should constantly check the progress made in the
various sections. He is responsible for the taking on of new men, and for dis-
missals, and for the taking on of men or youths from the shops. It will be
his duty to grant, or at least advise, what changes shall be made in the staff
organization at the proper time, and to investigate any grievances which may
be brought to his notice. From his room he can usually overlook the whole
office, and yet be easily accessible to callers and to men who may require
his attention for important decisions. It is usual for him to have various
forms or tables showing a time-table, progress rates, &c.
Programme Table.—A characteristic time-table for a marine-engine
office will show, against each job number, the name of the purchaser, the
size and capacity of the installation, the contract date, the date of launch
and delivery date proposed, and it will be for him to take such steps as he
thinks necessary to ensure that these dates are made possible, by regulating
the drawings and orders sent out through the order office.
Table of Drawings.—A very desirable table to be kept is one giving
the drawings and their characteristic numbers, the material orders, which
should be issued in connection with it, the date when the drawing was
finished, when traced, and when and to whom issued. Typical forms of
these are shown on opposite page.
Each table would be filled up by one of the juniors on each job, who
would also do a duplicate for his own section leader.
In addition, an abstract would be kept in the chief draughtsman's room
to show the parties with whom the sub-contracts were fixed, with their date;
also the date of promised delivery, and of actual delivery.
The time-books kept by each draughtsman usually pass through the
chiefs hands on their way to the time-clerks, so that he can scrutinize them.
Personal Control by Chief Draughtsman.—It is highly desirable
and very usual for the chief draughtsman to go round the office as frequently
as his other duties will permit, and so to keep himself thoroughly conversant
with the work in the office. It is usual for him also to work in close touch
with the heads of the shop departments in order to appreciate properly any

Job No.
	Drawing No.
	Pattern Shop.
	Finishing Shop.
	Fitting Shop.

	S I 12/20



Job No.
	Dg. No.
	Order Sheet.
	Sent to Copy.
	Pattern Shop.
	Finishing Shop.
	Fitting Shop.

	1 1 12/20
	J. Brown









difficulties that may arise in the execution of the work. Most large firms
have a daily council of heads of departments, where ideas are interchanged
and matters concerning two or more departments can be discussed.
Office Correspondence.—The chief draughtsman will be responsible
for the office correspondence. It is usual for him either to consult the section
leader or to get him to draft a suitable reply to a letter in which a number
of details are considered. He is also responsible for the issue of all draw-
ings and order-sheets. His own order clerk looks after this work. He
keeps a registry of drawings and orders issued, and sees them through into
their respective shops or to the dispatch clerk. The order clerk usually
keeps the chief draughtsman's books and carries any instructions or messages
he may have to give. The chief draughtsman in a large office keeps a typist
for his own particular correspondence. This typist generally does the
filing of correspondence and keeps the indexes up to date.
General Procedure.—The drawing office must keep in touch with the
foremen and erection engineers. The normal procedure is for foremen,
inspectors, or sub-contractors who wish to elucidate any point or who desire
an alteration to suit shop practice or purchaser's usual requirements, to go
in the first instance to the chief draughtsman, who will probably send them
on to the man in charge of the job. It is very desirable that this procedure
be followed out, so that the chief draughtsman may be thoroughly conversant
with any change made during the progress of the work in the office. This
procedure eliminates as far as possible controversy at the conclusion of the
contract, and mistakes which may arise from departments not knowing
of changes made which may affect them.
Assistant Chief Draughtsman.—The work of the assistant chief
draughtsman is to be reasonably familiar with the correspondence and the
general duties of the chief. He is expected to concentrate his attention on
the technical aspect of the work, and it is for him to interpret carefully the
intentions of the design office. Drawings going out of the office should be
scrutinized by him generally to see that the terms of the specification are
complied with and that they correspond with the original designs.
Section Leaders.—The section leaders have charge of one or perhaps
several jobs, and work with several juniors under them. The section leader,
who is generally his own checker, gives out the work to the juniors, and
generally superintends the drawings on the boards, and does a considerable
amount of the drawing himself. Having given a drawing to a particular
man, he guides him generally, and when the drawing is finished it is taken
off the board and carefully checked. The process of checking is one of the
most difficult and harassing parts of an experienced draughtsman's duties,
as he has to watch the specification very carefully, to see that the general
dimensions correspond to the design or guidance drawings, and to assure
himself that the various detailed sizes correspond to those on other detailed
drawings. Every pipe and valve-flange on drawings and on order-sheets
must be individually checked. This is no small matter when it is stated
that in the machinery pipe arrangement alone of an intermediate liner of
THE  DETAIL OFFICE                           21
7000 i.h.p. there may be as many as 1000 pipes and possibly 300 valve
Having checked the drawing, the details and preparation of which will
be more fully dealt with in a later paragraph, it is initialled by the section
leader and given to the assistant chief. The latter, after inspecting it, will
have it sent into the tracing office, whence drawing and tracing are returned
to the section leader. It is usual then to check the tracing with the drawing,
the draughtsman's initials being put in the corner with those of the checker,
when the drawing can be photographed and sent out for circulation.
Size and Style of Drawings; Instruments; Handbooks.—The
size and style of drawings should be standardized as much as possible. It
must always be recollected that the tracing-papers and cloths, also photo-
printing papers, are made in rolls of 30 in. and 40 in. broad, and drawings
should be made accordingly.
A very good size of drawing-paper is the ordinary double elephant size,
40 in. X 27 in., although a smaller sheet, the imperial, 30 in. X 22 in. is fre-
quently adopted. The former is not only a very convenient size when on the
board, but is a satisfactory size for handling in the shops, and is economical in
tracing-cloth and photo-paper. For large arrangement drawings, paper from
the web roll is generally used. This may be the well-known sand-grained
paper, or it may be some form of mounted hand-made paper. These can
generally be procured in long rolls of 30 in., 40 in., 54 in., and 60 in. width,
and the amount required cut off. Where a drawing will be on the boards
for a long time, instead of attaching it to the board, as is usual, with small
brass-headed drawing-pins, the paper is stretched by soaking, and, whilst
wet, glued to the edges of the board. When the paper dries, it of course
contracts and gives a very tightly-stretched surface to work on, and which
will remain stretched without any ruffling up, as long as the job lasts. It
is essential to have the edges of the board planed perfectly true, and also
to have a very true T-square, also good set-squares, one of 45° and the other
a 60° one. Scales may be of paper, but are more generally of wood, and are
much more satisfactory when edged with white celluloid. Ivory scales are
frequently used, but they are very costly, and after a time the marking gets
rubbed off, and they require to be recut. Where English measures are
adopted, the usual scales are J in., J in., f in., i in., f in., f in., i J in., and 3
in. to the foot. In modern drawing-office practice, the slide rule is constantly
used for multiplying, dividing, squaring, cubing, extracting roots, &c. Each
different branch of engineering and shipbuilding possesses its own favourite
pocket and handbooks with tables, See., but these tables are frequently
standardized on sheets hung round the office.
Beginning the Drawing.—On beginning the drawing, the draughts-
man plans in his own mind how he will space it out for easy reading in the
shops, which are seldom so well lit as the office. Centre lines are used as
datum lines, and all dimensions should be calculated from them and checking
done with reference to them. Generally, two views at least are necessary,
and half a dozen may be needed, including, perhaps, an outer elevation,


a sectional elevation, an outer end, and a sectional end elevation, and simi-
larly with the plan. Half-sections are very common. It is common practice
to draw the plan, looking down on the article, immediately below the side
elevation. The end elevations are usually drawn on the left and right hands
respectively of the plan and side elevation. The end elevation on the left-
hand side is the end view when looking from right to left; and the end
elevation on the right-hand side is the end view looking from the left to
the right.

The aim of a drawing should be to portray the article drawn simply,
exactly, and completely. All the necessary instructions for manufacture
should be given on the drawing. Sometimes this information is given in
the form of notes, but it is better to give it in tabular form. It is usual to
give overall dimensions to assist the shop foreman to understand at once

Looking Oft B

Lookiruj ot\ A

Looking On, C

Fig. 5 — Disposition of Views on Drawing

the size of piece he will be called upon to handle, and to make his arrange-
ments accordingly.
Dimensions.—Dimensions should be written in large bold figures,
and where, as in an arrangement drawing, there are a number of similar
parts, each should have a distinguishing mark, such as capital letters of the
alphabet. Thus the ground plan for a large works would have the columns
marked A, B, C, D, &c., and the different piping systems might be lettered
H!, H2, H3 for hydraulic pipes, Sj, S2, S3, &c., for steam pipes, &c. These
distinguishing marks are of very great assistance in the identification of
pieces in the shop, where they will be painted on if wrought iron or steel,
and probably cast on if cast iron or gun-metal.
The practice in many offices now is to give all dimensions in a drawing
up to 2 ft. in inches, i.e. 23 in., but after 2 ft. in feet and inches, as 2 ft. 7 in.
At least one standard, whether it be this or another, ought always to be
adhered to. It is usual in large firms to give each part a cost number, so
that the actual cost of every detail as it passes through the various shops may
be known. This cost number should be shown on the piece, and an arrow
should indicate precisely its location. This cost number will also be given
in the table at the foot of the drawing, with the location, material, " number
off ", &c. In some cases a refinement is made on the cost system, so that

„        ,

Mb* !«CM
	H.P. kL.P





	!&sueo Tb:-


[Facing p. 22, Vol. I

there is a different number for the material and for the classes of workman-
ship, but this leads to an enormous notation, which would not seem to give
commensurate results. To keep the cost number from being confused with

dimensions or " numbers off " it is usually ringed thus:

In the drawing proper, each part should have printed under it its dis-
tinctive name and " number off ", also the scale to which it is drawn, if
different scales are used in different parts of the same drawing.

All spare parts required should be marked on the working drawings, so
that they can be made at the same time as the working parts.

Finally, at the foot of the drawing, its well-known title, such as " piston-
rods " or " cylinders ", should be given, and the number of the job, draw-
ing number, date of drawing, and scales.

A characteristic title would be


SCALE: i J in. = i ft.

4 Sheets: Sheet No. i

Drawing No. 783/1

The table of particulars set out at foot would be something like the

Cost No.
	Particulars of Part.
	No. off.
	Order Sheet.

	Junk ring studs

	Junk rings
	Drg. 783/11

Colour Work.—In some offices a considerable amount of colour work
is done, chiefly by juniors, generally to distinguish the classes of material
used. Very faint washes only should be used for these, and these should
be applied to black-and-white prints rather than to the tracing. There is
a fairly well-known convention for materials, such as neutral tint for cast-
iron, blue for steel, brown for copper, yellow for gunmetal, and light pink
for lead.
Sections are frequently hatched to show up more clearly and to indicate
unmistakably that they are sections.
Diiferent offices may have different conventions for colour schemes and
for hatchings, but those shown in diagram are those in most general use.






(Light Grey)

(Light Blue)

(Liyhl- Brown)

(Ye/tow )


In a coloured drawing, a key diagram of colour scheme should always be
given on the drawing itself. This saves waste of time and any possible

In making many of the arrangement drawings, care should be taken to
simplify them to the greatest possible extent. The necessity for this becomes
very evident in any piping arrangement, where pipes may cross one another
or be hung above others in a great multiplicity of arrangements, and which
it is extremely difficult to show clearly. It is usual to make one arrange-
ment drawing showing all the different systems of pipes, for the sake of
checking clearances in the drawing office, but such an arrangement is of
very little use in the shops, yards, ship, or to men on the site. It is therefore
necessary either to colour pipes by systems or, better still, to prepare separate

drawings for each system,
such as a steam and ex-
haust arrangement, an oil-
pipe arrangement, a lubri-
cating arrangement, a
hydraulic arrangement,
a sanitary arrangement,
bilge and ballast arrange-
ment, &c., as different
squads of men will be fit-
ting different systems in
all probability. In pass-
ing, it may be said the
old practice was to take

sets and make a great number of such pipes to place. This has largely
given place, particularly where pipes of large bore are concerned, to a
system of detailing these in the office, and only using closing lengths which
are made to place. This considerably adds to the drawing-office work,
but saves much time and delay in the shops, particularly as there is a
growing tendency to use steel and iron pipes where copper and lead were
once very common. The steel and iron pipes can be procured easily in
standard lengths, also bends and junction-pieces, at prices very much less
than they can be made for on site.

Where wrought iron or steel is used it is generally necessary to send out
block sketches of the material required, as it may come in " rough forged ".
Details are not shown on these sketches, but they give the outline and outside
dimensions with the usual extras for working and machining. When draw-
ings or order-sheets are sent out, a copy should on every occasion be filed
for drawing-office use. This filing should always be done by one person,
say the safe-attendant or drawing-office clerk, and each item should be
entered up in a register, giving date and characteristic number. Notes to
photographer, authorizing the taking of prints for the shops or for prices,
should be initialled by section leaders, and the recall of all drawings from
shops for alteration should be done by a note in the office duplication book.


Key to Colour Scheme
THE  DETAIL OFFICE                               25
Order-books.—One or more large order-books should be kept for
each job, and a duplicate copy of each order sent out should be inserted.
The drawing-office order-book is generally quarto size, and made with thin
sheets of white paper upon which the orders are pasted. A standard index
should be at the front of each book, and the order-sheets for different jobs
entered always in the same numbering of pages. The book should, more-
over, be split up into convenient sections, and a number of spare sheets left
between sections, so that a space may be provided for unusual orders, which,
of course, must be specially and appropriately indexed up.
Standard Drawings and Data.—In any large office a considerable
number of standard drawings are kept, whereby a very considerable saving
of time and labour is effected. It is obvious that it is necessary to have
a uniform standard of bolting throughout the work, and, indeed, if the firm
can see its way to the adoption of the British Engineering Standards for
pipe-flanges, &c., so much the better, and the nearer we shall be to a standard
practice and the simplification of design and avoidance of difficulty in repairs.
Standards are generally constructed for pipe-flanges of different pressures
and material, standard dimensions of bolts, glands, riveting, and other parts
of the work that lend themselves to this process. In addition to this, books
containing all the dimensions and sketches of the different classes of small
fittings used, such as valves, cocks, &c., are kept, and it is only necessary to
indicate position of flanges and pieces they join and to add the standard
number, to completely specify the piece it is desired to have made. Portions
of the work which can be easily standardized in design, although perhaps
not in dimensions and scantlings, should be so treated that the addition of
the one or two variable dimensions should complete the sketch or order.
Miscellaneous Drawings and Sketches.—In any drawing office
there is always a mass of sketches and drawings received from outside,
which must be indexed and kept in an orderly fashion. These are either
kept in appropriate drawers or in individual pockets or dockets. When
drawers or dockets are not available, large square envelopes with tongued
flaps are a good temporary substitute. When the drawings are finished
with, they can be bundled together and put away in the storage safe, where
old records are kept.
These square envelopes should be marked on the; outside with the job
number and packet distinction, say, A, B, C, D, &c. Each print or tracing
kept in them will be Alf A2, A3, A4, &c., or B^ B2, B3, B4, &c., as the case
may be.
In folding prints it is a good, neat, and satisfactory plan to fold them in
Admiralty style, with title, number, and date received, and the origin of the
drawing marked clearly OB the outer portion. In putting prints back in the
dockets, they should always be put back strictly in order so as to minimize
loss of time in future searches. It is attention to these small details which
tells favourably on the efficiency of an office—saving time, worry, and
misund ers tanding.
All alterations to prints should be made in red " blue-print " corrector,
26                   DRAWING-OFFICE   ORGANIZATION
and if new prints are sent out in place of any recalled, a red chalk-mark
should be added to draw the attention of the shop foreman to the modification.
Time-book.—Each draughtsman keeps a time-book in which he
should enter up the time he spends per day on each job. Generally the
complete job number will be held sufficient, as few offices do more than
make an overall calculation of time spent on any individual contract.
Record of Alterations.—An experienced section leader will keep a
list of the alterations made during the course of the job, with a note of the
authority from whom he received instructions to make the changes. This
practice not only keeps him right in the case of disputes later on, but is
necessary, as a set of correct tracings of the job as finished may be required.
In this set of finished tracings alterations made outside the office, sometimes
without office sanction and at the request of a surveyor or inspector, are
expected to be incorporated. This will mean that a few journeys to the
ship or site may be necessary and a note of these alterations taken. The
working drawings should be altered accordingly in red.
Key Diagram.—In large arrangement work it is usual for the section
leader or his most experienced man to make outline key arrangements, and
sometimes what may be called a connection diagram. This diagram would
show in end vertical column the sources of power, and on top horizontal
column the auxiliaries to be driven. In the square common to each put
a circle with the bore of the connection. By this means a complete tabular
statement of all connections can be given. Such a diagram is the key by
which the section leader will check subsidiary drawings and orders.
Such a key diagram is shown herewith, dealing with auxiliary connections
for a large marine-engine installation. It is quite possible that the same
item shall appear on upper horizontal as well as on left-hand vertical column
heading. To show how it is worked, take the item " Steering Engine " in
vertical column. We find a circle with the size 2\ in. under the heading
" Reduced Steam ". This means that steam to steering engine is not taken
direct from boiler, but from a reducing valve. Of course the lead to the
reducing valve itself will probably be taken direct from the boilers, but this
does not affect the key diagram. Similarly, it will be evident that exhaust
steam from steering engines must be arranged by a suitable arrangement of
valves and connecting pieces to go either to main or auxiliary condensers,
feed heater, atmosphere, or L. P. turbine. This connection diagram is very
often translated with a key sketch actually showing the place of main engines,
boilers, and auxiliaries, and lines drawn connecting them. This is not really
necessary, as the connection diagram shown should give all that is necessary,
but the diagrammatic sketch makes it clear to juniors. It is understood that,
in column marked " Makers ", the name of the makers of any auxiliary
machinery should be inserted, simply as a convenience to the section leader.
It is usual in large contracts to send such key drawings and diagrams,
as well as the principal drawings, to the owners for approval. It is highly
desirable to get this approval at an early date, so that material can be ordered
early The usual plan is to send two thin prints of each, one of which will


be returned signed or stamped with the owner's approval, and the other
retained by the owner for his records. Great discretion is necessary to
know which plans it is important to have approved in the first place, as also
which parts of the material should be, and which parts can be, ordered first.
Orders for material of which only a long delivery can be given should be
pushed out immediately; also those portions which, from their position,
must be made and fitted first of all.
In the course of his work it is necessary for the draughtsman to make
himself thoroughly familiar with the ordinary shop practice, machines, and
facilities, such as maximum sizes the machines will take, facilities for handling,
crane-lifts and heights, jigs and gauges, patterns in store, dies, and all the
implements of manufacture generally.
There should be books in the office containing these items of information,
including a list of taps in stock, &c.
Catalogues of Special Parts.—As, in large-scale production, a con-
siderable amount of specialities are bought in finished from outside firms,
it is very desirable to have the figured catalogues easily available where such
specialities can be seen, and their duties and sizes found. This information
facilitates the ordering of the same, and makes for much greater accuracy in
the finished drawings. These catalogues should be kept in one place and
indexed, and a register of them kept by the office clerk or safe-man, or other
person deputed for the task.
In a well-organized office the boards and benches are cleared every
evening, loose drawings and tracings are put away in the fireproof safe, order-
books put in their correct place, as well as catalogues, &c. Not only are
these things saved if a fire does break out, but a great deal of trouble and
time is saved, should a particular item be required at an unusual time or
when someone may be off work.
Library.—Another very useful adjunct to the office is a library where
the larger works, other than handbooks, which deal with engineering matters
pertaining to the particular branch of industry in which the firm specializes,
are kept. The technical press should also be available in the library for
reference. In one or two cases, books are lent out from the library to juniors
who are keen to learn anything about the industry they are engaged in. In
such cases, the eldest apprentice may be responsible for their issue and safe
return. In one office in the writer's experience this plan worked very satis-
factorily. This same apprentice frequently has charge of pencils, rubbers,
inks, drawing-pins, &c., which are usually supplied to the draughtsmen by the
firm. These items constitute a fairly heavy expense in the office, and so far
as is commensurate with efficiency should be used as economically as possible.
Use of Tracing-paper.—It is becoming more and more common
practice in the larger offices to dispense as much as possible with drawing-
paper, using tracing-paper instead. There is much to be said for the prac-
tice both on the score of expense and convenience. The economic side of
the question need not be laboured, but a considerable amount of the work
in an office is of the nature of repetition work, with a few alterations to suit



particular cases, so that the use of tracing-paper may save considerable time.
Consider for a moment or two the case of, say, the lubricating pipe arrange-
ment round a large engine or the arrangement of platforms. If it be done
on drawing-paper the engines have to be drawn down to scale afresh each
time, and it is likely some apparently small details will be overlooked which
may seriously affect the arrangement. In a case like this a sheet of tracing-
paper can be laid over, say, the carefully made up drawing of the engines,
and the leads of lubricating pipes or platforms drawn in on tracing-paper,
without a single line of the engines proper. Of course, when finished, it
will be advisable to show the main outlines of the engine to convey a quick
picture of what is required to the men on the job. Against the use of tracing-
paper may be urged the fact that it tears more readily, gets dirty, does not
lend itself to erasure, and consequently leads to a good deal of annoyance
and irritation, and consequent inefficiency; but if properly and discreetly
used, great economy, financial and otherwise, ensues.
Fluctuating Nature of Work.—It is inevitable that there will be
periods of extreme pressure. Certain shops will demand work greedily to
keep them going, and to get the contract forward at the greatest possible
rate. Never may it be hoped that a whole contract can be designed and
detailed and modified to the draughtsman's satisfaction before he is called
upon to pass drawings and orders into the shop. Skill, judgment, and
experience are very necessary to know what things must have precedence.
For instance, it is obvious that where large castings form part of the product,
time will be needed in the production of the patterns, and a further period
must elapse before the foundry can deliver the castings.' In the machine
shop, moreover, many different operations may have to be performed on
one piece alone, and all of them at different times, so that the preparation
of such a drawing is generally a first call. But there are obvious risks. The
facings have to be fixed definitely when one would gladly do it tentatively,
in view of what may crop up at a later stage of the work. But this may not
be, and it is seldom indeed that a whole job is finished, and can be looked
back upon without a wish that it had been possible to alter many things.
It follows that often, in a squad, one or two of the men are trying the most
likely small arrangements to ensure that a reasonably suitable arrangement
of the major pieces can be made thus early. It is here that experience and
judgment are so necessary. But, whilst this is often the case, there are
periods of red slackness, when not much current work is on hand. The
staff is generally kept up, as it is usually bad policy to deplete a staff which
knows the run of the work and the office. During such periods the men
are generally turned on to the task of working up the data which may be
neglected during times of pressure; of preparing and altering, where found
necessary, standard drawings and sketches. This work, whilst it has no
visible return at the moment, proves useful in the long run.
Drawings of Standard Parts.—In making standard drawings it is
important not only to make the standards for different sizes of the same
group of articles show differences which shall be definite and progressive,


such as thickness, &c., but to make the drawings show similar views in
similar parts of the drawings, and to make all the drawings of exactly the
same size and style. For many classes of work it will be found very good
practice to take a sheet of double elephant drawing-paper and divide it in
four equal parts by drawing both vertical and horizontal centre lines. In
each quarter thus made one standard may be put, complete in all its details.
This sheet after being traced will be photographed, so that four standards
will be on one sheet. After photographing, each standard may be cut up
into a sheet by itself, and perhaps a dozen or twenty of them bound in one
book, the covers of which may be formed of stiff drawing-paper, stitched
with strong twine.

An example will make the above procedure clear.    Most drawing offices
have to use steam or water valves in some part of the work.    These
either globe pattern or L pattern, high pressure or low pressure, and
of cast iron or gun-metal.   It is obvious here that six standard books will
made up, showing valves, say, from i|-in. bore to ic-in. bore or thefeabojut,
rising in the smaller sizes by J in. each time, and from 3 in. to 7 in. bj^ An
and thereafter by i in.   The outer cover would be marked, for instange


220 lb. per square inch



For standards such as the above, which are in everyday use by a number ^N «*'
of men in the office, it is usual to have several sets of each standard, which
will be kept in the safe and given out on the requisition of a section leader.
On no account should the tracings leave the safe, nor indeed any other
tracings, unless for photographic work or for modification, for the loss or
misplacement of a tracing is a serious matter entailing considerable work
and annoyance in having to be remade, as such may perhaps cause very
serious inconvenience at a very busy time, say when a telegraphic request
is received for a photograph of some important part which has to be repaired
or replaced immediately.

Circulation of Drawings. — It is important to have a fixed routine
for the circulation of drawings. Each department may have only a small
part to do on any one drawing, but it is general and even advisable to issue
the complete finished drawing in each case. It is general practice to do so,
because it saves the preparation of several drawings — an important point
both as regards time saved and the reduction of error, for the preparation
of each fresh sketch or order-form involves risk of error, particularly as
such subsidiary sketches would be left for juniors to make. Not only so,
butjt is a false notion of efficiency to show a craftsman only the little por-
tion of a job he must do. A better job is done because of the knowledge
he has of the whole and its general purpose, and he may be in the position
to save some part of the process by his practical knowledge when he knows



what transformations it must yet undergo. Incidentally, when this practice
is adopted, one tracing instead of half a dozen has to be altered, if an altera-
tion be found necessary.

In designating where prints are to be sent, it is desirable to mark depart-
ments rather than the initials of particular foremen, which is a common
practice in many places. For instance, in a large engineering work the
initial letters of the different shops may be used, as P. S. for pattern shop,
F. S. for finishing shop, similarly for erecting shop, smithy, boiler shop,
machine shop, dock engineer, works manager, &c. It is usual to send a
copy of all drawings to the works manager as well as to retain one for drawing-
office use. Moreover, it may be necessary to send one to the purchaser's
superintendent and possibly his inspector on the premises. In ordering


lirculation of Drawings

photographs it should be clearly stated whether thin, thick, or mounted
paper is desired, and also whether black-and-white paper or the more com-
mon ferro-prussiate blue print is required.
Up-to-date practice is to take only the tracings as authoritative text.
The drawings after being traced are used no further, nor are they altered
if alterations become necessary. They are stowed away systematically, so
that, should the tracings be destroyed, the work could be gone over afresh
from the drawings. Drawing-office copies are printed off, and are marked
for drawing-office use only. These are the drawings in use for reference
in the office. The tracings remain in the safe. Alterations are made to the
tracings and photographs only.
In order-sheets the designations of the orders may be indicated on the
top, either in a printed form or put on fresh on each sheet. The printed
form is preferable, with the initials of those departments which are not to
receive the drawing crossed out. Thus a typical order heading would be:
Copy sent to: P.S., S.S., F.S., E.S., W.M., D.E., Inspector.
A glance will show where everything has gone to, and its recall be easily
arranged if necessary

Many order-sheets are ruled and printed in considerable detail, and only
want a few sizes filled in by the draughtsman and completed by the tracing
or copying office. Indeed, these offices can do a considerable amount of
work which will save the draughtsmen very considerable time. For instance,
in the case of a large piping-arrangement plan, it is usually necessary to give
a list of all the pipes in the job with their distinguishing number, material,
thickness, and standard pressure; also a list of fittings, and a list of auxiliaries,
such as the following:


Pipe No.
	Order Sheet Page.

• {
	Main steam from) boiler A . . /
	Si in.

• (
	Main steam from) boilers A and B J
	| in.




	Order Sheet Page.

R.i3 {
	Main stop valve \ on boiler C /
	Si in.
	5 in.

	Aux. stop valve \ on boiler C J
	4   in.




	Size of Cylinder.
	Size of Pump.
 6 in.

	6 in. — 6 in.
		12 in.
	Blank, Blank & Co.
	Sanitary pump
	7 tons per hour





Now these lists will have been made up early in the job on foolscap sheets
for the sake of ordering material, and long before the drawings are completed.
In such cases it is only necessary to hand these in to the tracing room and
have them incorporated in the drawing, indicating only the place and list
it is desired to have traced in.

Correspondence.—There is an amazing amount of correspondence
passing through the drawing office, much of which does not materially
affect the drawing-office work, but which the draughtsmen should see for
purposes of information. Unless the correspondence is kept in a very
orderly method, hopeless confusion is likely to arise. It is necessary to be
able to lay hands on particular letters at a moment's notice, as these letters
may contain the records of decisions arrived at in a very early stage of the
contract, and which it may be important to know and appreciate at a much
later period. Only copies of letters should be retained in the drawing-
office files; the originals of all incoming letters should be retained in the
typists' room or in a general reference room. Copies of incoming letters
should be kept on differently coloured paper from that of copies of outgoing
letters. A good practice is to have the former copied on thin white sheets,
and the latter on thin yellow sheets. All these sheets should be of the same
size, and a stamp in each case put at the top, giving necessary information
of the process of circulation or designation.

The white incoming letters have a stamp at the top,' such as:

Copy to: W.M., D.E., D.O., B.S.
Referred to: D.O.
Answered by: F.K.

and the outgoing letters:

Copy to: W.M., CH., D.O.

The letters sent out the previous evening are generally available for cir-
culation in the morning. Of course the originals, probably signed or
initialled by the chief draughtsman, would be checked by him before being
sent off, so that there is no need for him to peruse these letters, but he should
read letters which are sent out by other departments. Having finished
looking over the daily file, he will pas^ it out to the assistant chief, who will
assort them under their respective job numbers and subject-matter, and
give them to the appropriate section leaders. The case of incoming letters
demands a little closer scrutiny for any new points of importance which will


emerge. The originals of those referred to the drawing office for answering
will be retained by the chief draughtsman, whilst at the same time copies
of all incoming letters will be circulated similarly to the outgoing ones. The
section leader must peruse them carefully, and make a precis of the more

Fig. 10.—Circulation of Letters in Firm

important ones in a book he should keep for that purpose. The book will
have a column for the name of the firm from whom the letter came, and its
date, also short statement of contents written in pr&is form. It is desirable
to have a further column giving the date on which anything of importance
in the letter was given effect to. This means a little labour at the time,

Fig. ix.—Circulation of Correspondence inside Drawing Office

but it is well repaid at a later period. After the section leader or one of
his juniors has finished with the correspondence, it will be put in the office
filing-basket, from which the clerk or typist will take letters once or twice
a day and file them in-the proper filing cabinet. There are a great number
of filing cabinets on the market, but for drawing-office work a loose folder
system is the most suitable, as the letters have to be turned up so frequently.
VOL. I.                                                                                                                 3


I i





One drawer of a cabinet will be used for one contract, which should be
clearly indicated on the label. A number of stout manila sheets, alpha-
betically indexed, will be in the drawer, and in each lettered division a folder
for each firm under that letter will be inserted. Each folder will contain
the correspondence with one firm on one subject.

Perhaps at this point it may be well to indicate diagrammatically how
the incoming correspondence in a large firm circulates, and the place the
drawing office occupies in the general scheme (fig. 10).

The circulation in the drawing office itself is shown in fig. n.

Orders for Material.—Orders for materials, either on typewritten
sheets or on printed order-sheets, are generally sent out through the order-
clerk in the counting-house. There may be very good reason for delaying
to issue these orders, but in such delay there is a distinct chance of the order
being altogether overlooked. At least, if the sending out of the order-sheet
is all the drawing office knows about it, there is no chance of a forgetful
order clerk being reminded that the material will be required in a measurable
period of time, and if an order-sheet gets lost, serious disputes may arise
as to which department was at fault. In many cases now the original order-
sheet sent to the order department is not sent out, but is split up, if necessary,
for buying purposes, and a fresh order made, say on a differently coloured
paper. When this is received by the drawing office, it is a guarantee that
the order has been passed through by the order-clerk. The section leader
in his notebook for order-sheets will have several columns which will clearly
mark its progress and destination, such as:


	Prepd. by.
	Sent to Copier.
				From Order

	Order Clerk.


All drawings and order-sheets issued should be stamped with the name
of the firm, the department, and date of issue. They should also be initialled
by the chief draughtsman or his assistant.


Tracing Office
In the larger offices, women are now employed as tracers. Many
of the best technical men are by no means the neatest draughtsmen.
Whilst it is important to have drawings made carefully, neatly, and
to scale, it is of only secondary importance that the actual drawing
should be of a high finish. Neatness of line finish is only an incidental
accomplishment to the expert designer. But all the same, it is most desir-
able that the prints sent down to the shops should be neat and clear. For
neatness girl tracers cannot be surpassed, and it is wonderful how even the
most complicated and elaborate arrangement, say of general piping for a
large battleship or liner, can be made clear by people who do not know the
mechanical details or understand what each line signifies.
The tracing office is kept apart from the drawing office, but for obvious
reasons should be contiguous to it. It is under the charge of a head tracer,
who takes her instructions from the chief draughtsman and apportions the
work amongst her own staff.
Linen Tracings.—Most tracings nowadays are made on tough linen,
made clear with a highly starched glazed surface. The drawing is pinned
down, and a piece of tracing-cloth is stretched over it. It is necessary to
tear off a strip along the borders, as at this portion the rolls are generally
wrinkled, and if this selvedge were allowed to remain on, it would be very
difficult indeed to get the cloth properly stretched. This stretching does
prove rather troublesome, as the tracing-cloth is apt to stretch very con-
siderably. It is usual to stretch it tightly over the drawing for an hour or
two, or, in the case of a very big plan, overnight, before beginning work on
it, otherwise it would be found that if the tracing were right over the drawing
in one place it would not be so in another.
Making the Tracing.—The surface worked on is a highly glazed
surface. Water takes out the glaze by destroying the starched surface,
and makes the cloth opaque and useless for photographic purposes. It is
therefore essential to take care that no water gets on to the tracing. More-
over, a crack in the tracing will show clearly on the photograph, so the tracing
should never be folded, but should be either kept flat or carefully rolled up.
To prevent the ink running on the surface too freely, ground French
chalk is rubbed over it to enable the ink to grip.
The purpose of the tracing is to obtain the sharpest line photograph
possible. No half-tones are required. For this reason the tracing-cloth
should be as transparent as possible, and the ink as opaque as possible.
Many of the opaque papers have a strong yellow tinge, and if an ink without
much body in it be used, we get either a faint blue background, where ferro-
prussiate paper is used, or indistinct white lines, making a more difficult
photo print to read than need be.

fc-                                                       °

1                                     Ordinary blue-black or red writing ink is absolutely useless to get a clear

^                                line.    The main portion of the drawing, certainly, as well as the printing is

|                                done in black Indian ink.   The ink should be mixed freshly every morning,

I                                and ground down in a white-enamelled china palette to a consistency which

l                                will at once run freely and at the same time give a perfectly black line.   A

|                               little gamboge mixed with the ink helps to make it more opaque.   For

Hi                                    centre lines and dimension lines a less prominent line will do.   It used to

;i                                    be common to mix up crimson lake with water to a very thin syrup for this

I                                    purpose, but it was not generally dense enough, and has largely given place

|                                    to the use of burnt sienna.    Several firms, indeed, use nothing else but

|                                   black, chain-dotting centre lines to distinguish them from outlines.   Where

I                                        it is desired to show things faintly, such as ladders and platforms about an

II                                       engine installation, Prussian blue is employed.   As a general rule, tracings
i                                   should be made with firm, slightly heavy lines, if the most satisfactory work-
|                                    ing drawings are to be obtained.   Very thin lines do not come up well in
*                                    the photographic process.   Some of the inks can be washed out easily with
|                                   water, and it has become fairly common of late to use bottled waterproof
jl                                   inks in the tracing office.   These are much more difficult to erase. If altera-
!'                                   tions are desired in a tracing, it is better to have a small sketch of the altera-
*'                                   tion sent into the tracing office, and the erasure and alteration made there.

j'                                         When the tracing is finished it should be checked, size for size, with the

I                                   original drawing, before being allowed to leave the office.   The printing

|                                   and figuring should be as clear as possible, and, in fact, it is becoming general

practice to tolerate straight up-and-down lettering only.

Copying Order-sheets. — The copying of order-sheets is generally
done in the tracing room. The old method, was to press the sketch and
lettering through from the original sheet, by means of carbon papers, on to
the half-dozen copies required. Modern appliances have got rid of this
laborious and rather barbarous practice. The order-sheet sent in from the
drawing office is only drawn in pencil. The tracers go over it with a special
ink capable of taking a considerable number of copies. This is put on top
of special gelatine sheets and a roller run over it, so that an impression is
taken on the gelatine. This gelatine impress is now used as the original
to take the required number of copies, generally on thin tissues. Two or
more coloured inks can be used in the process, which leaves the order-
sheets very clear and satisfactory.

Several odd jobs find their way to the tracing office, although it is not
strictly tracers' work. These are the correction of a number of specifications
from an original copy, the writing up of the data-book in ink, which has
been filled in in the drawing office in pencil. In short, the tracing office
does any job arising in the drawing office that calls for neatness.


THE   PHOTOGRAPHIC  ROOM                       37

The Photographic Room

Sun Prints.—In the olden days the method of printing was for one of
the apprentices to run up to the roof with the tracing and put it in a flat
printing-frame, similar to that used by the amateur photographer but of
course very much larger, and to leave the sun to do the rest. This method
is only tolerable where the number of prints required in a day is small. In
large offices, not only is someone required to take off prints all day, but a
much more rapid method is necessary. The photographic room has become
a well-equipped and indispensable portion of the drawing office. In an office
employing about twenty-five men, the writer has known 120 large photo-
graphs being taken, dried, and dispatched in one day, including all the neces-
sary indexing, &c. This keeps one man busy the whole day, as quite a number
of the prints are black-and-white ones, which take about six times longer
to print than the usual blue prints, and require considerably more washing.

Equipment of Photograph Room.—Not only must the photographic
room be well equipped with an up-to-date electric printing-frame, but it
should have several large baths for washing, a plentiful supply of water, a
permanent squeegee, a good drying oven, and also a considerable number
of laundry rods for natural drying. A table with a hard wooden top should
be provided, a large steel straight-edge, or better still an automatic grip,
and a deep-cutting knife. Ventilation is highly important, as the drying
oven is generally a gas one. The fumes of the oven are apt to lie about the
room, and the process of natural drying is very slow if a good current of air
is not available.

Procedure for Obtaining Prints.—When prints are wanted, the
section leader wishing them enters up a requisition form, which is printed
in a small manifold book. Alternate leaves of this book are printed perhaps
on pink paper, and have perforations for detaching them. The copy leaves
are unperforated and may be white paper. The ordinary carbon sheet is
used to get the duplicate. A sample page is given herewith.

	JOB No. 531.   Date 11/11/20.

	TITLE: Foundation Plans.

	DRG. No.   531/87-
	No. off.
	Sent to
	Kind of Print.


	Blue mounted

	„   thin

	/    Black-and-white \       mounted

	Blue, thick

	»       *»

	„    thin

Kinds of Prints.—The blue print is the cheapest photo print and the
one most suited to general shop purposes. It costs less than a shilling
per yard.
Blue prints for mailing purposes are taken on a very thin paper. For
machine shops, where a print will be required a great deal, and where it will
probably be pasted up in frames to keep it flat and to prevent it from going
amissing, a mounted blue print is preferable. This is a paper blue print
mounted on a tough linen backing, and it forms a very firm print indeed,
being like a thin cardboard when washed. Sometimes a linen cloth is used.
This is durable, very soft, and is suitable for folding, but the parts of the
fold are apt to rub off.
Prints which have to be coloured, say for the approval of owners, are
generally taken on black-and-white paper, i.e. black lines on a white ground,
and if occasion demands it, on black-and-white mounted paper. This
paper is much dearer. It costs probably four times as much as the blue
paper, and it takes longer to print. The white ground if slightly under-
exposed is apt to look dirty, and if slightly over-exposed the lines may, if
traced thinly, come out rather faintly.
Printing Machines.—A fairly common type of printing machine is
one formed of two semicircular cylindrical pieces of plate-glass, which together
form the curved wall of a cylinder which is open at the ends. Two prints
are usually put in at the same time, one on one semicircular portion and one
on the other portion. An arc lamp is hung from the roof over the centre of
the cylinder. The cylinder swivels, so that it may lie horizontally when
putting in the tracing, the face of which lies against the glass. Over this
is placed the photo paper, and then a felt backing is strapped on to keep
it in position. When these adjustments are satisfactorily made, the cylinder
is tilted on end so that the electric lamp may travel down its axis. The
speed of travel of this lamp is adjusted by a clockwork arrangement. As
the lamp falls gradually to the bottom the light is reflected on to the glass
and the tracing, which it penetrates. The light affects and fixes in some
degree the chemical surface of the paper, The black lines of the ink pre-
vent penetration, and the unfixed chemicals are dissolved away in developing,
leaving a white line on a blue ground, or a black line on a white ground,
and there are papers with white lines on brown grounds, Sec. The defect
of this type of frame is that the length as well as the breadth of the print
is limited, at least without folding and to some extent damaging the tracing.
•Not only so, but unless the lamp has been carefully wound clear of the
frame, it may be broken in swinging the frame to the horizontal position.
The semicircular cylindrical glass, moreover, is very costly to replace, and
awkward to handle in such a contingency, and the portion of the print at
the bottom is liable to have longer exposure than the top portion. The
latter defect betrays itself in a slight unevenness of ground-tone. Of late
years a flat plate-glass horizontal frame, which works on an endless roller
system, has been introduced. This frame takes prints of any length, say
those common in shipyards. The arc lamp travels horizontally at a fairly


quick speed, and on either side of it is the flat plate-glass of the width of
the machine. A print can be taken on either side, and each side can be
geared to different speeds of feed, so that they may be kept geared one for
blue prints and one for black-and-white prints. The lamp travels back-
wards and forwards like a shuttle, and is operated by an electric switch.
The frame is never removed, so there is no danger to the lamp, and if a

Fig. 12.—Haldcn's Duplex Radial Electric Photo Copying Frame.   Prints being inserted.
sheet of glass does get broken, there are generally a number of spares kept
so that it can be replaced very quickly and readily.
Prints of both circular and flat glass types of machines are shown (figs. 12 to
14). The former is technically known as Messrs. Halden's Duplex Radial
Electric Photo Copying Frame, and the latter is the same firm's Double Pattern
Single-lamp Type Continuous Electric Photo Copying Machine, and both
are shown by the courtesy of Messrs. Halden, who very kindly supplied
the blocks for these illustrations. In the case of this latter type, the single-
lamp type has now almost entirely superseded the original machine, which

had two or even three stationary lamps instead of one moving arc lamp.
The current consumed is less, and less expense is entailed in replacing chim-
neys and carbons; moreover, the glasses can be placed nearer the lamp as
a less amount of heat is generated, thus avoiding to some extent the danger
to the glass by overheating. Prints can therefore be taken as rapidly with
one lamp as with several, and a more even exposure is obtained.

Fig. 13.—Halden's Duplex Radial Electric Photo Copying Frame.   In position for photographing.
A recently improved form of machine (fig. 15), also supplied by Messrs.
Halden, has been put on the market, called the Rowsley Super-continuous
Electric Photo Copying Machine. It is claimed for the machine that it is
more economical than previous patterns in the use of electric current and
that it enables the operator immediately to increase the output.
The tracing and photographic paper are fed from a table, and are taken
close up against the glass by slow-moving rollers. When the end of the
tracing comes out, the photographer draws his knife sharply along the photo
paper at the top of the table, which is a glass slab, and lets it work its way

Fig. 14—Halden's Double-pattern. Single-lamp Type Continuous Electric Photo Copying Machine
down in a few moments to the receiving trough underneath. A considerable
number of photographs may be taken off before the washing process is begun.
Often the photographic room is placed high up in the building, a relic of
the time when sun printing was the common practice. This is often respon-
sible for an insufficient supply of water. The bath should be kept perfectly
clean, as considerable sediment comes off some of the photo papers. Indeed,
Fig. 15.—Rowsley Super-continuous Electric Photo Copying Machine


the ideal system is to have a continually running supply, so that the bath is
kept constantly fresh. The photograph is immersed, and a small hose
made to play on it to drive off every particle of surface chemical. Separate
baths should be kept for black-and-white prints and the ordinary blue prints,
if satisfactory results are to be obtained. When thoroughly washed, the
print should be drawn through a squeegee, permanently attached to the
side of the bath, to take off as much of the surplus water as is possible. If
the groundwork of black-and-white prints comes up slightly muddied, it
may be chemically treated to bring it up white, but great care must be exer-
cised in this treatment lest the black lines of the drawing should get obliterated.
It may be of some interest to the operator, or draughtsmen with some
knowledge of chemistry or photography, to briefly indicate the chemical
reactions with either the ferro-prussiate blue paper with white lines or the
ferro-gallic white paper with black lines.
In the former the paper is coated with potassium ferricyanide and
ferric compound of iron. When exposed to the influence of actinic light,
either from the sun or the electric arc, part of the iron in the sensitive
compounds is changed from the ferric to the ferrous condition, which with
potassium ferricyanide gives an insoluble blue compound which is pre-
cipitated on the paper. Side by side with this reaction a portion of the
potassium ferricyanide is reduced to potassium ferrocyanide, which, with
the unchanged ferric iron, also deposits a blue compound on the paper.
The net result is that a complex mixture of blue compounds is laid down
on that portion of the paper, i.e. the background, which has been sub-
mitted to the ultra-violet rays. The portions unaffected because of the
protection afforded by the ink on the tracing are washed away in the bath,
leaving the white lines on a blue ground.
Similarly with ferro-gallic photo paper, i.e. paper which gives black
lines on a white ground. In this case the paper is coated originally with
a solution of iron salts, the ferric compound being reduced by the action
of light to the ferrous state. The paper is now treated with a solution of
gallic acid, which changes the ferric iron on the parts shielded from light
into a bluish black compound. The exposed portion, where the trans-
formation from the ferric to the ferrous state has taken place, is unaffected.
In water bath, or one bath ferro-gallic paper, the coated material
carries its own developer in the form of a powder on the surface. With
this paper, on immersion in water, after printing, the ferrous salt, with
the gallic acid in contact with it, is washed away, leaving fixed on the
paper the black compound of ferric iron and gallic acid.
Drying of Prints.—The prints are caught by spring clips at the edges,
and hung up to dry. It is better to let the prints dry naturally, as artificial
drying is apt to distort them badly, especially where they dry last.
Register of Prints.—A photo register is kept by the photographer,
showing when each print was sent out, and what was its destination.
The Safe
The protection of drawings, tracings, and books, from loss by fire,
theft, or careless destruction, is important. The safe is usually a strong-
room, a room built of brick and iron, asbestos-lined, and provided with
steel-shuttered windows and steel doors. The safe may be of con-
siderable dimensions, and is generally staffed by a man in charge and
one or two office-boys. The safe should be well fitted up with drawers,
pigeon-holes, serving-table, voice-tube connections to chiefs room and other
departments, and should be well lit and ventilated.
When a new tracing is made and checked, it should be at once given
into the custody of the safe-man, who will enter it up in his book, also the
date of receipt. This entry will be transferred to his permanent tracing
register, where it will be entered under the proper job and drawing number.
The original drawing, which will be passed in at the same time, will be filed
away, probably in another part of the building, as it will not usually be
required again. In any particular job the drawings may number anything
from 20 to 200, and probably in certain cases many more. It will usually
be found undesirable to roll up more than ten or twenty tracings together.
Drawings i to 10 will be in one roll, n to 20 in another, and so on. These
rolls are best kept in japanned tins to keep them from dust and damp, or,
failing that, in canvas covers. They should always be very carefully rolled
up and handled, the surface never being cracked nor the corners allowed
to be folded back. Every crack in the tracing means a line not intended in
the print.
Tracings are only given out for photographic purposes, or when it is
intended to alter the tracing; and when a tracing or a print is given out,
the date of issue and the draughtsman's name should be jotted down in a
day-book kept for that purpose.
When a new tracing is given in, the safe-man should see to it at once
that the proper office copies arc taken off, as prints are now almost universally
recognized as the standard form of drawing-office copy.
The same procedure will be adopted with the prints and sketches received
from outside, or copies of sketches sent out, except that, in the latter case,
it would be the only available copies which would be given out when
All the order-books are kept in the safe, and occasionally the data-books,
although these latter are more generally kept in an ordinary small iron safe
in the chief draughtsman's room.
Estimate drawings are, of course, entered up in an estimate-book and
filed away appropriately.
In the best firms no one but the safe-man, and whatever assistants he
has, enters the safe, all transactions taking place over a counter. The drawing
required is called in, and it is usually left to one of the office-boys, attached
to the safe, to bring it down to the draughtsman who requires it. These


boys generally do any clearing away of benches which may be necessary at
A well-kept safe not only ensures the safe custody and well-being of
records contained therein, but facilitates the usual routine work of the office
in providing what is required with the minimum delay and vexation.
Evolution, not Revolution, desirable in Drawing Office.—Such, in
outline, is the usual office organization, which we have discussed less in a
systematic theoretical manner than as good common practice in many
offices throughout the country. There are many items of organization to
which radical alterations may be made with advantage, but we have to con-
sider that, in most cases, even a relatively small alteration to modes in common
practice may produce very considerable dislocation for a time. An alteration
in the size of order-books, for instance, or the sequence of their pages, causes
a certain amount of confusion in an office, because all the old records are
done in another fashion. This point can only be appreciated by those
who actually have worked in an office at the time of such changes.
The Human Element.—To the ordinary draughtsman each individual
job is a job by itself which he must seek to do as satisfactorily as possible.
He follows the instructions of a senior, and if he interprets them intelligently
he is not likely to go far wrong. To the section leader each job is merely
a small thing in a very large contract; he has to look a good deal before
and after, and may, out of his long experience and knowledge of what may
be expected at a later period, make many decisions and give many instruc-
tions to a junior which do not at the time appear very convincing. The
contracts in the engineering industry are so large that large sums of money
are generally involved in the smallest decisions, and mistakes are likely to
be very costly ones. It is therefore necessary to let men think out the tough
problems that fall to their lot, and it is a false economy that keeps the section
leader's nose to the grindstone when he can perform a much more valuable
service in supervisory and advisory work.
The work of the chief draughtsman, whilst it includes that of the section
leader, calls especially for personal qualities. He must see to it that no
serious friction arises in the office, and that information is freely given.
Occasionally serious errors result from feelings of jealousy and bad feeling
which prevent one man giving another the fullest information. It depends
very much on the character and tact of the chief, whether this spirit or one
of good fellowship shall obtain between the members of his staff.
There is a tendency with the larger firms to achieve efficiency by means
of stringent discipline. Discipline, of the Prussian type, can be carried too
far—better results can often be achieved by giving conscientious men some
freedom of action. It has become very frequent of late to introduce time-
clocks into the office. No one, of course, denies the value and desirability
of punctuality, but it must be remembered clocks measure time, not work
Grievances should never be allowed to grow, but should be attended
to at a very early stage. Frank and free discussion will frequently remove



the most serious misunderstanding, and it should be realized that serious
grievances are often due to, and are kept alive by, a lively sense of some
real or supposed injustice. Such grievances are seldom confined to indi-
viduals, but quickly spread to large bodies of men. Fortunately with tact
they can usually be met and rectified.

One or two other points of a general nature have to be considered, and,



Message        Massage

Fig. 16.—Relationship of Officials
although they do not enter into the daily routine of drawing-office work,
they can be quite justly considered under the heading of organization.
One is the question of apprentices' entrance to the drawing office. From
the nature of things, no universal system of recruiting drawing offices exists,
nor is any standard of efficiency and ability demanded for full membership
of the profession, although it is obvious events are moving in that direction.
The day is possibly much nearer than many people suspect when draughts-
manship will be a profession like medicine, the law, accountancy, &c., in
which qualifying examinations are necessary.
A practice, in some of the leading firms at present, is to staff the drawing


office with men selected as a result of an examination held by the firm.
Where this course is adopted, the examination is confined to the firm's own
apprentices who have had at least two and a half years' shop experience.
Those who receive the highest marks come up to the drawing office if their
works record is satisfactory, and this policy ensures a leaven of practical
experience in the drawing-office staff.

Design of the Drawing Office.—Another point of organization which

Fig. 17.—Arrangement of Offices
is worth consideration is the disposition of the various drawing-office depart-
ments. It is advisable to have them all on one flat, if possible. Certainly
all should be in close proximity, and the chief draughtsman's room should
be in a central position and easily accessible to all departments. A very
satisfactory arrangement is shown in fig. 17.
The method of arranging the boards, the question of how they shall
face, and how the light shall come in, the convenience of lockers, the places
to lay drawings, the height and position of them to entail minimum fatigue,
&c., must all be carefully considered.



VOL. 1.




The work of a competent pattern-maker is both exacting and com-
prehensive. He must be skilled in woodwork, an accomplishment which
he shares with the carpenter, joiner, and turner, who, however, may not
understand how to construct patterns that will deliver from the sand, how
to economize in material and avoid the employment of complete patterns by
the substitution of skeleton-like structures, how and when to use sweeping
boards, sectional pieces, or cores.
He has to know the best methods of countering the effects of the damp
sand on porous timber, by the judicious employment of open joints, of
segmental pieces, and of framed structures of many kinds.
He has to be fully conversant with the different systems of moulding
—green and dry sand, and loam—and core-making in all their branches, and
with the handling of light and heavy work. It is necessary to be familiar
with the evils that result from shrinkages in unequally proportioned castings.
He is primarily responsible for the methods of moulding (since he
has to determine how patterns shall be constructed for the mould joints),
for ramming and delivery, and for the determination of upper and lower
faces for pouring.
An intimate acquaintance with the operations of the machine-shop is
necessary, as machining allowances vary considerably in different classes of
castings, while the variations that occur in similar pieces are often large,
due to the presence of hard cores, the straining of top-boxes, the absence of
risers, and the differences between the results that are associated with the
practice of hand-rapping and delivery and machine-moulding.
Elementary knowledge of arithmetic and geometry are required for the
estimation of weights and the laying out of work. In all shops some men
have to specialize in toothed gears, or in motor-work, or marine-castings, in
plating metal patterns, in odd-side work, and so on. In truth, the craft of
the pattern-maker is a many-sided one.
5^                               PATTERN-MAKING
The Elements
Pattern-work includes two very broad aspects, that of the method of
moulding to be adopted, and that of the actual construction. It is necessary
to determine the first before the second can be proceeded with.
Castings may be made from (a) complete patterns, (b) incomplete or
skeleton patterns, (r) loam patterns, (d) moulds swept directly in loam.
(a) Complete patterns are those whose shapes, except for cored portions,
are identical with their castings. With the employment of these, many
side-issues are involved: the directions of their jointing; the amount of
shrinkage allowance and taper; the adoption of middle parts, loose pieces,
and drawbacks or false cores; and the formation of internal portions by
self-delivery or with independent cores.
(ft) Incomplete patterns are made of strips or frames that have the out-
and-out dimensions and the main contours the same as for their castings,
but which leave interior spaces to be completed with sand cores, or with
strickles. The object here is to economize timber, and incidentally to
lessen weight.
(c) and (d) Loam patterns, and moulds swept out in loam are only used
for circular bodies, so that flanges, bosses, and brackets must be prepared
in wood as complete pattern elements and attached to the main pattern
or set in the loam.
On the pattern-maker falls the responsibility of deciding by which of
methods the pattern-work and the moulds are to be made.   In many
the most suitable method is self-evident to a man with experience.
Full patterns are always made for work of small and medium dimensions.
Skeleton patterns, those of loam and loam moulds, have preference for very
articles, but subject in a measure to the number of castings required.
A            casting, though of medium dimensions, would seldom have a full
pattern, provided  its  shape were  suitable for skeleton  construction  or
sweeping; a large one, if repeated in considerable numbers, would.   The
problem always is just one of the relative costs in the pattern- and moulding-
A large pattern is expensive, but so is a large loam mould, for'which
numerous attachments may have to be prepared.   It will often happen
that a quantity of castings of large dimensions can be more cheaply
from a skeleton pattern, perhaps from a complete one, than from loam
or loam, patterns.   In such a case the moulder has a grievance if the
him with unnecessary expense in order to lessen the
of its own. department.


Fig. i.—Pattern of Sheave Wheel built up

(a) Complete Patterns.—In these, the first question that arises is
that of the direction or directions of jointing the mould, with the usual
though not necessary concomitant, that of jointing the pattern similarly.
This very often admits of alternative solutions. In a fair number of
instances, only one is practicable, though others may be possible if the
cost of moulding is overlooked. The best way to approach the subject
is to consider the simple elementary geometrical forms which are constantly

Jointing.—All moulds, except the
relatively very small number which are
" open", comprise bottom and top
portions, included in bottom and top
box-parts (" drag " and " cope "). An
exception occurs in bedded-in moulds,
for which the bottom box is not re-
quired. The jointing between top and

bottom is determined by the facility afforded for delivery of the pattern,
with the least risk of damage to the mould, and bearing in mind too the
extent of subsequent details of finishing, coring, and of pouring, and the
disposition of upper and lower faces. This latter consideration is most
important when tooling enters into the case, since machined portions must
be free from specks and blowholes. In general, if one portion is of greater
depth than another, the deeper section goes in the bottom. The reason is
that it is much better to withdraw a pattern from a bottom mould than to
lift the top sand off the pattern. This is not always necessary, because when
a top box-part is turned
over, and the pattern
parts along the joint, the
upper portion can be
left loose from that
below, to come up with
the top sand, and be
withdrawn after turning

Over.                                                                     Fig. 2.—Pattern of Trolley Wheel built up

Elementary   sections

that deliver well are illustrated in many subsequent diagrams. The
patterns may or may not be jointed along the same planes. Very often
they are not; seldom in those of small dimensions used by brass moulders,
because dowels work loose with usage, and the edges of the pattern parts
overlap. When unjointed, the moulder makes the joint face, guided by
the eye alone, or, in repetitive work, some form of joint-board, odd-side,
or plate is provided.

In many instances, the pattern joint cannot coincide with that of the mould
(fig. i and fig. 2 are typical examples). The patterns must have divisions
to permit of withdrawing them from the moulds, but the joints of the latter
do not coincide with those of their patterns. In the examples (figs, i and 2)



Fig. 3.—Rib and Boss dowelled to come with Top Box

the mould joints are made along the centres of the convex edges, but the
pattern joints are elsewhere.

Taper.—The deeper the mould, the more difficult is the pattern to with-
draw and the greater the care that must be exercised to avoid disturbing the
sand. The first inch or two is the stage at which fracture of the sand is

most likely to occur. After the
pattern has been loosened by
rapping, and drawn slightly out
of the sand, the principal care
necessary for the remainder of
the lift is to keep the pattern
level, the difficulty of which in-
creases directly with area. Slight
rapping is continued until the
pattern, by reason of its taper,
has cleared the encircling walls
of sand. Taper or draught there-
fore assumes much importance.
Its amount varies widely. Some
patterns have a large amount, in

localities where it does not interfere with the fitting of parts, as on the
outsides of deep bedplates, of sewer boxes, of machine frames, or of stiffening-
brackets. No rule can be stated to meet all cases, but common practice is
to give J in. to J in. per foot of depth.

Loose pieces.—These are a particular provision, made to ensure free
delivery of portions on pattern sides, without making down-joints. In some

cases they afford alternatives to
coring and to drawbacks or false
cores. In others, as in copes, pieces
are often left loosely on the main
pattern (figs. 3 and 4) in order to
permit of their being taken out of
the mould after turning the cope
over in preference to lifting the
mould off them. In the latter case,
the alternative is to impart as much
taper as is permissible to the por-
tions that come in the top, and to
avoid keen edges and angles there.
In a sheave wheel the convexity of

the internal rim (fig. i) provides for a free delivery. In a trolley wheel,
internal taper and a well-rounded edge are necessary. Top bosses are
almost invariably left loose, unless they are very shallow. Thin facing
pieces and core prints are fast, but these have well-tapered edges.

Generally, when loose pieces are visualized, the most frequent case of
those attached to the vertical sides of patterns occurs to the mind. In the

Fig. 4.—Rib dowelled to come with Top Box


typical example (fig. 5) the top flange is fast to the pattern, but the rib
below and the middle strip must not be (apart from the employment of a
drawback plate). Loose pieces are located during ramming, either with
skewers or with dovetails, the first-named being removed during ramming,
previous to the withdrawal of the pattern.

Obviously, before a loose piece can be drawn inwards, there must be an
open space large enough to receive it. If, in fig. 5, the pieces have to be drawn
into the narrow space left on the withdrawal of the main rib, they must be
thinner than the space thickness, as is the case in A. But the bottom strip
in c is thicker, hence it must be divided into two or three thicknesses, one to
follow the other. Since the pricker has to be inserted diagonally (B), getting
the pieces out of so deep a
space is troublesome, and                              Q

no mending-up or cleaning
can be done if the sand
breaks down. But the
conditions are altered if
the interior has to be taken
out with cores, or if,
though rammed wholly in
green sand, the ramming
is done on a grid that
permits of the removal of
the interior mould. Ample
space is then left, into
which strips of greater
width than those shown
can be withdrawn. But
even then there are limi-
tations to the widths that can be dealt with in this manner, where fracture
of the sand and convenience of cleaning and blackening have to be con-

Drawbacks or false cores.—These avoid this awkward method of with-
drawal, but they have a vastly wider scope. They are either grids or plates
on which outer portions of moulds are rammed, to be lifted bodily away from
the pattern, to be replaced and reset accurately by some form of joint between
the plate, or its sand, and the sand in the main body of the mould. This is
capable of very extensive applications, since there is no limit to the width of
the encircling portions that can be carried thus.

Internal portions, Cores.—The conditions that control the delivery of
internal parts differ from those of external. Thus, it will be obvious that
depths and diameters are related. A shallow hole will deliver satisfactorily,
though it has but a slight amount of taper. A deep hole of the same diameter
will not, and therefore it must be taken out with an independent core. Frames
of large dimensions may be regarded as patterns having large holes, relatively
shallow. They deliver freely as well within as without, and they are tapered

Fig, 5.—Loose Pieces
56                                   PATTERN-MAKING

by the same amount. The question of coring scarcely arises here, but it does
in all cases where interiors will not deliver.

The plain interior of a casting, made by ramming sand within a pattern
which is the exact replica of its casting, may be termed a core. But the
accepted meaning of the term is, a body of sand, generally dried, rammed in
a box distinct from its pattern, and inserted, and located in the mould by the
impression of a core print attached to the pattern.

The alternative of coring an interior to making the pattern like its casting
arises. In the majority of instances no doubt exists. No intricate shapes
can be delivered. These must be rammed in a separate box or boxes, and
inserted in the mould. In some instances it is more convenient to make
cores than to self-deliver. In a fair number of cases a core is preferred,
because, using prints, a stronger pattern can be constructed than if the timber

pig, 6.—Skeleton Frame united with Halvings, with Interior Strickled
were cut away to allow the interior to deliver. Lastly, projecting portions
are frequently cored over in preference to making a joint in the pattern, or
" down-jointing ", or to employing loose pieces or drawbacks.
(£) Skeleton Patterns.—Patterns of large dimensions, and those of
moderate sizes, when they are of shapes that would require considerable
quantities of timber and much tedious cutting, are not made of solid, con-
tinuous stuff, but are of more or less open construction. The numbers of
moulds required count in this scheme, so that while a skeleton pattern might
be used for a few castings, a complete pattern would be more economical for
large numbers.
The open frame (fig. 6) is the simplest example of the skeleton pattern.
Narrow strips jointed at the corners provide the outside dimensions. The
interior is strickled, or it is occupied with loose removable strips (fig. 7) for
repetitive mouldings, the latter being better to ram the cope on than sand is.
The same method is employed for plated portions that are curved in outlines
(fig* 8), but with increased economy, because more timber and labour are
required for working these than for plane frames. Here the strickle may be
used, or strips be fitted at intervals to form a discontinuous guide, the spaces
between the strips being filled with sand, to be rammed on. This method


is extensively adopted for large cylindrical bodies, which would be most
costly to shape in solid timber, and be very heavy to handle.

Fig. 7.—Skeleton Frame with Dovetailed Corners, and Interior occupied with Loose Boards to Ram upon

(c) Loam Patterns.—These (fig. 9), of cylindrical form, are swept on
bars when they are too small to be swept in loam moulds on bricks, and

Fie. 8.—A Curved Tank Plate with Interior Strickled

too large, having regard to the number of castings required, to bear the cost
of timber and labour.     The field for their employment is thus rather re-

Fig. 9.—A Pattern Column swept in Loam, with Flanges of Wood
stricted.   They are awkward to handle, being heavy, and do not deliver
cleanly from the sand, especially in the cope, because their surfaces are rough,

58                                    PATTERN-MAKING
and they are unjointed. All non-symmetrical fittings, as brackets, feet,
bosses, and flanges, if large, are prepared separately in wood, and attached to
or laid on the loam pattern. A large number of moulds can be taken from it,
since it is hard, and its surface is protected with a coat of hot tar. Among
the articles commonly made thus are cylinders, e.g. for hydraulic machines,
for Corliss engines, for large gas-engines, and for pumps. Allied to this work
is that of strickling patterns of bend pipes in halves. The loam is swept on
grids; and flanges, feet, or other attachments are prepared in wood and fitted.
(rf) Swept Moulds.—Moulds are swept in green sand and in loam
to avoid the expense of complete or of skeleton patterns. The scope of the
first is limited because the fragile character of green sand does not permit
of deep sweeping. That of the second is very extensive, and is practically
the only method available for very large cylindrical moulds. In some details
where contours are irregular and unsymmetrical, loam is laid and worked
against pattern parts of wood embedded in it. But the main moulds are
swept to their symmetrical profiles with boards attached to a central revolving
bar, set concentrically in a step bearing, and their main cores are swept on
the same or a duplicate bar.
From these elements we turn to the consideration of the most suitable
methods of pattern construction.
The pattern-maker's craft differs in many ways from that of the carpenter,
joiner, and wood-turner. The first and chief contrast lies in the necessary-
provisions that have to be made for delivery, taper, and the other matters
instanced in the preceding division. In addition, measures have to be taken
to minimize the effects of the severe and destructive treatment to which
patterns are subjected. It is that of insertion in wet sand, of rapping, and
delivery, alternating with storage, and what is as injurious, the alterations
that have to be made in many patterns from time to time. And, in all but
highly standardized work, wood alone is used, yellow pine mostly, soft and
porous, and mahogany to a limited extent for small articles.
Very broadly, pattern construction falls within three great groups: plane
areas, cylindrical articles, and circular work.
Plane Areas.—In dealing with these the aim is to lessen the widths
of individual pieces to relatively narrow strips in order to localize the expan-
sion due to moisture and the shrinkage consequent on storage, which regularly
alternate. The solid glued-up table tops and side-boards of the cabinet-
maker have no analogues in pattern-work. Instead, wide pieces are always
made with " open joints " (fig. 10), that is, a space of about J in. or less is
left between strips, to the extent of which they are free to expand when
moisture is absorbed, so that the out-and-out dimensions of a broad width
are not affected.
Since the edges of the open joints are not united, there is no cohesion
between the pieces as there is when edges are glued, nor can the pieces lie


flat. The method adopted therefore is to drive into the adjacent edges
tightly-fitting dowels, which prevent the faces from getting out of level.
The strips are maintained in one plane by the attachment of flanges, ribs, or
other pieces, as in fig. 10, or, if
these do not happen to be avail-
able, then temporary battens are
screwed across, the impressions
of which are " stopped-off " in
the mould. An alternative to
this, which can be adopted when
the plated portion lies horizon-
tally, is to make an open frame,
jointed at the corners, and to fill
up the interior with loose strips

(%• 7)-

Boxing-tip.—This, except in
narrow widths, is combined with
open jointing. It is adopted in
all rectangular sections that are
too large to be cut from solid
plank, and is the only way in
which swelling and shrinkage
can be avoided in these. Longi-

Fig. 10.—Portion of Boxed-up Bed, showing Open Joints
and Cross-bars

tudinal strips are screwed to
cross-bars, a single strip for
narrow sections, several strips
with open joints for the larger

In these constructions the vertical pieces should occupy the entire depth
of the pattern, as in fig. n A, B, and should never lie between the top and
bottom plates, as in fig. n, c. The reason is, that the delivery of the
first is clean, that in the second is not, because a very slight shrinkage of


Fig. ii.—Right and Wrong Ways of Boxing-up

the strips produces lapping edges that tear up the sand.   Also it is better
to fit the pieces with a rebated shoulder (A) than to make abutting joints
only (B).
Cylindrical Articles.—These include engine cylinders, pump barrels,


pipes and columns, and work of which these are typical. Only when these
are of small dimensions, say of 6 in. diameter and under, are they made in
solid stuff, usually jointed along the central plane and dowelled. When of
over this size the patterns are built up with narrow strips, glued on cross-
pieces, located at short intervals—" lagging ". This method is adopted up
to the largest diameters for which entire patterns of wood are constructed.

As diameters increase, the num-
ber of lags is multiplied. In all
cases they are very narrow to
localize shrinkage, ranging with-
in small limits, say from 2 in.
in small bodies to not more than
5 in. in the largest. It is neces-
sary to secure rigidity as well as
freedom from changes of form.
Strips must not be too thin,
nor may the cross-bars be
spaced very far apart. Propor-

M2.—Ring built with Segments                   tions .are governed by diameter

and length. Subsequent draw-
ings (figs. 34 and 35) indicate suitable proportions. Good close joints
must be made between adjacent edges, and be united with glue. If the
work is done carefully and well, and seasoned stuff used, the patterns will
retain their accuracy for an indefinite period. There are details in the
methods of construction that are dealt with in later sections, where
examples of work are illustrated.

Circular Work.—This is built up with sectors of circles—" segmental

work ". Obviously, if rings
were cut from solid material,
they would shrink into ellip-
tical forms, and fracture
along the short grain. Built
with sector pieces overlap-
ping, " breaking joint ", they
mutually reinforce each other,
shrinkage is minimized, and
the circular shape is main-
tained. (See Chapter II,
Section 3.) To secure this result perfectly, it is necessary to limit the
length and the thickness of the individual pieces. Those too long would
shrink in width, and those too thick would shrink and lack something of
reinforcement by other pieces. The maintenance of a judicious relationship
between these proportions is necessary to secure permanence of form. This
method of building-up is suitable alike for rings that are shallow or deep.
The shallower the work, the thinner the sectors are. A thin flange of large
diameter should not be built in less than three or four courses. Rings of

Fig. 13.—Semi-spherical Pattern built up
THE  ELEMENTS                                61
all sections are made in this way, parallel (fig. 12), bevelled, and semi-spherical
(fig. 13). The pieces are glued singly, with carefully planed joints, checked
with chalk. In general it is not necessary to reinforce the joints, but as a
precaution wire nails are frequently driven in. When the section of a
pattern is that of a cone, as in the rim of a bevel wheel, or has any curved
outline, wooden pegs are preferable, because, if they should happen to come
to the exterior, they will not damage the turning-tools.
Methods of Union.—The union of elements is in some respects
peculiar to pattern-work, being due to the necessity of making alterations
from time to time. The tenon and mortise joint is seldom used. Other
forms of joints are frequently not glued, but screwed only. Unions of a
more or less temporary character made with battens, the impressions of
which are filled up in the sand, and do not appear in the casting, occupy a
useful place in alterations. Very many alterations are made only partly in
the pattern, being completed in the mould by the method of " stopping-off ".
One of the commonest joints is the half-lap, used for uniting flat strips. It
is either a plain half-lap fastened with screws only, when there is a proba-
bility of future alterations being called for (fig. 6), or, if permanent, the
dovetail form is cut (fig. 7), and the joint is both glued and screwed.
Screws occupy a larger place in pattern-work than in the more permanent
methods of the joiner. They take the place of tenons and mortises and
dovetails in the attachment of parts.
Dovetails.—These are employed chiefly for the corners of deep open
frames that deliver their interiors, of which sewer boxes are typical, and for
loose pieces, as an alternative to the skewers. Their use is generally restricted
to standardized work. They are safer than the skewers, since these afford
but a doubtful indication, by their small holes, of the position of a loose
piece if mislaid.
Dowels.—These play a large part in pattern-work. They include the
tightly fitting dowels used in open joints, and those loosely fitting in one
piece, in the joints of patterns that are divided for delivery between bottom
and top parts (figs. 3 and 4). These are of wood or, for permanent work,
of brass or malleable-cast iron.
Angles, Fillets, or Hollows.—These are peculiar to pattern-work,
being employed to fill up re-entrant angles that would, without them, invite
fracture in the castings. They should never be omitted. They are made
of wood, leather, or soft metal, to be bent round curved portions. Illustra-
tions of all the elements here noted will occur in the subsequent sections.
Core Prints.—The function of a core print is to locate, by the im-
pression which it leaves in a mould, the exact place for the insertion of a
core. There are exceptions to this general statement, since some large cores
are set without print impressions, as when moulds are made from sectional
and skeleton, patterns. Also, when portions of metal are cored over, in

02                                     PATTERN-MAKING

order to avoid the employment of loose pieces or of drawbacks. Core
prints fall under two fundamental types, the " round " and the " drop " or
" pocket " forms.

Round Prints.—Tne^e are set either vertically or horizontally. The
first-named have taper or draught, the second are, as a rule, parallel, invariably
so when a pattern and its prints are jointed longitudinally through the centre
to be withdrawn from cope and drag. They are tapered when they are
attached to bosses or pieces that have to be left loose and be drawn hori-
zontally back into the mould.

Vertical Prints.—There is no recognized rule for the length or the
taper of these. Both call for the exercise of judgment. As the diameters of
prints are increased, their length or thickness a (fig. 14, A) is lessened re-





Fig. 14.—Vertical Core Prints, and Cores
latively, because the larger diameters afford better support to the core at
the bottom than at the sides. A print for example i ft. in diameter need not
be more than J in. thick, while one of i in. diameter will be i in. long. Up
to about 3 in., lengths and diameters are about equal, beyond that the pro-
portionate thickness lessens.
Prints are thinner at the top than at the bottom (fig. 14, A). Usually
they need not be more than half the thickness, since they have not to support
the core, but only to steady it against risk of lateral displacement during
pouring (fig. 14, B).
Bottom prints may have from J in. to J in. taper on the diameter. Top
prints should have more, because the sand in the cope has to be lowered on
the upstanding core, with risk of a crush if the taper is not ample. Often,
to avoid this risk, the portion of the core that enters the print impression
has an excess of taper, with the result that close contact does not occur until
the cope is down to its bedding. (Compare c and D, fig. 14.) It is desirable
in work that is standardized to put the taper in the core boxes. Ordinarily
the moulder rubs the taper on the cores to match the print impressions, and
his is a frequent cause of inaccurate setting.

Fig. 15.—Drop or Pocket Prints

Horizontal Prints.—Round core prints disposed horizontally, as in
pipes, columns, and work of which these are typical, are not tapered. The
lengths of the prints are
about equal to their
diameter in the smaller
dimensions. As sizes in-
crease, the lengths are rela-
tively less. But they may
never be very short, because
in that case the weight of a
heavy core would cause the
sand to crush. In most
cases the core is bridged
between two horizontal
prints. When it has to be
supported from a single
print impression the length

must be sufficient to counterbalance the weight of the overhang of the
core. But this is only necessary in those cases where no assistance can
be obtained from chaplet nails.

Drop or Pocket Prints.—These
(fig. 15) are employed for horizontal
cores when the joint of the mould
does not coincide with the centre of
the core print, as it does in the pipe
and column types of patterns. Even
then in some cases round parallel
prints are attached, and a sloping
" down-joint'1 is made to the centre.

Or round, tapered prints are skewered on loosely. But these are excep-
tions to the usual practice.

The drop print only indicates a portion of the outline of the core to be
inserted — the lower part,
semicircular for round cores,
other shapes for other forms.
The portion of the print above
the centre is tapered to deliver,
but its impression is filled up,
following the insertion of the
core—" stopping-over ". This

is   done   by   the   moulder,   Or,                  F*S- i?.—Boss Facing covered with Drop Print

in standard work, the core is

made in a box (fig. 16), which includes the stopping-over portion in
addition to the actual core. The thicknesses of these prints are similar to
those of the plain horizontal kind. Thin prints will not provide sufficient
support in the sand to sustain the weight of a core without risk of crushing.

Fig. 16.—Core Box for Stopping-over Drop Prints



The fitting of drop prints is often associated with the presence of boss
facings which have to be left loose (fig. 17). These are cut to fit round the
print, or the print is notched to fit over them. The portion of the boss that
is covered by the print has to be made good in the mould during the stopping-
over (fig. 18), or, for permanent work, it is put in the core box.

Fig. 18.—Stopping-over a Core with Boss Facing in Mould

Fig. 19.—Iron
Box for Round

Core Boxes.—When determining the forms of these, similar methods
and precautions have to be observed as in the construction of patterns, with
regard to freedom of delivery, taper, loose pieces, and so on. In addition
there is the excess length necessary for the location of the cores in the print
impressions. Often a main core will contain prints, the im-
pressions of which will serve for the location of other cores.
Box portions must generally be taken apart to permit of
the removal of the core, so that they are only held tem-
porarily with dowels, clamps, or screws. The subject of
core-box work is therefore nearly as extensive as that of
pattern construction.

Standard  boxes of iron have  their  halves fitted  with
tongues and grooves (fig. 19);  those of wood are very similar (fig. 20).
The  ends of rectangular boxes may be retained  in place with blocks
screwed against the sides, and the sides may be screwed to the ends
(fig. 21).    This entails loss of time in removing the screws as often as
the sides  have to  be taken away from the core.
Clamps of wood (fig. 23) or of iron are to be pre-
ferred.  The sides of long boxes will become rammed
outwards, with consequent enlargement of the core,
unless they are retained about the centre with a bolt
(fig. 22) or with a clamp.   The fitting of ends into
shallow grooves (fig. 22) is to be preferred to their
abutment against end blocks.  Frequently the interior
of a rectangular  frame  is occupied  with contour
fittings.    In one example shown (fig. 23) for making one of the numerous
cores for a turbine ring the blocking of yellow pine is lined with mahogany
to favour durability in service.   Here the actual width of the turbine ring
* that of the curved strip which represents the metal separating three tiers

Fig. 20.—Wooden Box for
Round Cores

Fig. 21.—Core Box with Joggles

Fig. 22.—Wooden Core Box, with Ends Recessed, and Bolts

Fig. 23 —Core Box for Turbine Rings, with Recessed Ends, Clamps, Internal Blockings, and

Mahogany Facings





of buckets. The supplementary open spaces are equivalent to the extra
length allowed on cores to enter print impressions. But in this case the
cores are simply set to lines described on a levelled bed of sand, and,
mutually abutting by their supplementary portions, they complete the ring.
Plain, rectangular boxes are rammed on a core bench. A bottom board
is necessary when bosses and other fittings have to be located correctly.
The sides fit this with dowels or with strips on the board. Often one or

Fig. 24.—Core Box with Strickle for Curved Portions

two faces of a core are curved. Then, when practicable, strickling is resorted
to (fig. 24), as it is also for the upper plane faces of cores. This economy in
curved portions is that due to the saving of timber and of the time occupied
in shaping it to the curves.

Examples of Work
Pattern-making includes many departments. The work done on patterns
for a brass foundry is wholly different from that done on patterns for the
heavier castings of the marine engine, the locomotive, and the larger types of
pumps, while the making of core boxes for gas and automobile cylinders calls
for special ingenuity and skill. The construction of patterns for cranes,
gear wheels, pipes, and columns, each enlists the services of men who have
developed into specialists. In every large shop certain groups of patterns
go to men who seldom handle anything else. But a trained, intelligent man
is, or should be, able to take up any branch of his trade when required to do
so. The principles that underlie the practice are unchangeable. It is from
this chiefly, the general standpoint, that the subject will be regarded in this



There are certain groups of patterns which possess one feature in common,
that of being jointed through the longitudinal centre. Pipes, columns, and
cylinders of all kinds are typical of a very large number of patterns jointed

Turning Patterns in
Halves. — Patterns, being di-
vided for convenience of mould-
ing, are jointed and dowelled
before they are turned, since it
would be inaccurate to saw
through solid patterns. Being

, they have tO be Secured             Fig. 25-— Clamping Pattern Halves with Dogs

during turning with dogs (fig. 25) ,

screws (fig. 26), or centre plates (fig. 27). Dogs are driven into the ends in
small and large work alike, in the latter, as an additional reinforcement
to the centre plates. For very light articles the dogs alone may suffice, the

Fig. 26.—Securing Pattern Halves with Screws

centring being done directly in the wood instead of on plates. Screws
are generally used for very light pieces. They are inserted near the ends,
in supplementary portions to be cut off after the pattern has been turned.
If one or two must come also in the body of the pattern, as when it is of
considerable length, the heads
must go in countersunk recesses,
to clear the turning-tools, and
the holes are filled up subse-
quently. Centre plates, smaller
and larger, are used very gene-
rally, not only to secure jointed
patterns, but also to receive the

Fig. 27.—Securing Pattern Halves with Centre Plates

lathe centres in those that are
solid, as these wear the soft
woods when turning is being
done, causing the pattern to run eccentrically.   The plates are made of
iron or brass, and are formed in the smallest sizes like dogs, to be driven
in, but in the larger sizes they are attached with screws or with nails.
' In some cases it happens that jointed pattern portions are less than semi-


i     i

circles, as when boss sections have to be fitted on plates or webs. Then a
piece of the web thickness is interposed before the boss is turned, and is
afterwards thrown away.

Flanges and similar attachments are turned in halves, usually dowelled,
and then attached to their bodies. They are held on the face plate with
screws inserted through the plate from the back. The hole is bored entirely
through, or it is recessed, leaving a portion to be removed with the band-saw.

Pipe Patterns.—Pipes and columns have several cardinal aspects in
common. Both are jointed longitudinally, dowelled, and moulded by turning
over. Both are lagged when the smallest diameters are exceeded. Each
has flanges and other attachments fitted. Loam patterns are used for those
of large dimensions. Patterns are plated for quantities. For very large
numbers, metal patterns, unjointed, are employed.

Pipe Patterns for General Service.—It is necessary to make use
of this phrase, because, outside of the general shops, pipes are made by highly

Fig. 28.—Pipe Pattern with Body Flange for Alterations
specialized methods. They are cast vertically. Metal patterns, collapsing
core bars, and a number of special appliances, associated with the moulding,
coring, and casting, are used. In America large numbers of pipes are made
in permanent iron moulds. The methods of making bend and tee-pipes are
similarly specialized.
In the general shops the outstanding feature is that pipe patterns have
to be utilized not only for standard lengths, with flanges, sockets, and spigots
of standard dimensions, but with slight alterations may have to be used for
all kinds of odd jobs and make-up lengths. In these shops therefore it is
customary to keep one set of patterns strictly for standard sizes, and a lot of
odd lengths and nondescript pieces for occasional orders. The cutting and
scheming necessary exercise the judgment of the pattern-maker, and very
much has to be done with stopping-off pieces, which increase the work of
the moulder. In the last case the pattern is not wholly like the casting pro-
duced, the shape of which is revealed by the stopping-off pieces supplied
and the corresponding sectional parts put on the pattern.
Pipe flanges are fitted into recesses turned between the termination of
the body and the core prints (fig. 28). A flange being retained correctly in
its recess need not be screwed in place. For casting shorter lengths, a body
flange A is screwed on, and this indicates the length at which the mould has
to be stopped off. The stopping-off piece supplied carries the half-core
print. Socketed pipes are stopped-off by providing an iron socket piece


that can be moved along the body of the pipe and screwed in any required
position.   As this carries the print, a stopping-off piece is not wanted.

When turning long pipe patterns, the correct diameter is set in at each
end. A very light cut is taken about the centre,
not quite down to the finished size, because of the
spring and vibration present. The reduced section
is then embraced by a " steady " mounted on the
shears of the lathe bed, and a rough cut is taken
with the gouge from the centre to the ends. A flat
is then planed from end to end, checked with a
straight-edge, and rubbed with chalk or red lead.
This serves as a guide to turning down intermediate
sections, without the need of having frequent re-
course to the calipers and straight-edge to check
the progress of the work.

Bend Pipes.—These may be long pieces of
straight pipe with a bend at one end, or they may
be entirely curved. On the degree of curvature
depends the method of their preparation. " Quick
bends ", those of small radius (figs. 29 and 30), are
turned in halves on the face plate. Four quadrants being screwed to the
plate by their joint faces, and turned, provide two complete bends. Bends
of large radii are worked in halves by hand methods. From a rectangular
cross-section a polygonal shape is cut, leaving only minute angularities to

Fig. 29—Pattern for Bend Pipe
of Small Radius

Fig. 30

Illustrations of Pipe Bends


be removed, to produce the semicircular shape, which is checked with a
When bends are attached to straight lengths of pipe, abutting joints are
used (figs. 30 and 31), secured with dovetailed pieces let into the joint faces
and screwed. The same method is employed for uniting branch pipes at
right (fig. 32) or other angles, as for tee-pieces. Sometimes a plate of iron

Fig. 32.
Pipe Joints at Right Angles

is used instead of a dovetail (fig. 33), or to reinforce dovetails (fig. 31).
Abutting joints are reinforced with dowel-pins fitting tightly, and with
screws put in diagonally. Flanges, sockets, and spigots are fitted as in straight
pipes, and stopping-off is practised.

The larger pipes and bends, and those of
awkward shapes, for which the demand is
limited, are frequently moulded from loam
patterns, for which the pattern-maker supplies
strickles and fittings. As the principal work
is thrown on the core-maker and moulder,
the subject is reserved for treatment in the
article on Foundry Work.

Column Patterns.—These are only made
solidly when of small diameters, say not ex-
ceeding 4 in. or 5 in. Beyond these, and
apart from quantity methods of moulding,
they are always parted longitudinally along
the centre and dowelled. Solid timber is
rarely used when the diameter exceeds 6 in.
or 7 in. The reason is that the stuff is liable
to become convex or concave in the joint

faces, and the pattern to lose its circular section. It is also liable to warp
and curve lengthwise, an evil that results from the incessant wettings of
the joint faces with the swab.

Patterns from about 6 in. in diameter upwards are, like the larger pipes,
built with " lags " or strips of timber screwed on cross-bars (fig. 34) and

glued to each other with longitudinal joint
edges. No rules can be stated for the cross-
sectional dimensions of lagging strips, nor
for the spacing of the cross-bars. These
are proportional to the diameter and the
length of the column, but are never very
thick nor very wide, since, as in segmental
work, the object sought is to localize shrink-
age as much as possible. The cross-bars
must be set at distances sufficiently close to
one another to sustain the lags bridged over
them against the pressure of ramming.
Thus the stiffness of the pattern must be
secured without unduly increasing the timber
sections. A little experience teaches the pattern-maker how to proportion
these details, the relations of which are correctly proportioned in the
accompanying drawings.

When building up divided columns, the cross-bars for one half are laid
down on a true joint board, and the lags are fitted to that first. They are
planed on faces and abutting edges, the latter being chalked to show contact,

Fig. 33.—Iron Plate for Pipe Joints

and corrected with the trying plane.   Each is glued to its fellow, a man
stationed at each end imparting a reciprocating movement to the lag about

Fig. 35.—Alternative Methods of Fitting Mouldings

Fig. 34.—Construction of a Lagged Pattern

half a dozen times to work out the surplus glue. Iron dogs driven in keep
the joint in contact until the glue has dried, and one screw is put in through
each lag into each cross-bar. The heads of these are sunk in to permit of
turning. When one half has
been prepared thus, it is
turned over, the other halves
of the cross-bars are set in
position by their dowels,
and the lags for that half are
fitted, glued, and screwed.
The halves are united with
centre plates, and the turn-
ing is done with hand-tools
or from a sliding rest. Done
by hand, the same method
is pursued, and time saved,

as in the turning of pipe patterns.   A steady is used to prevent sag about
the central portions.

Column Fittings.—All columns have flanges, with or without mould-
ings. These are nearly always prepared separately from the shaft, which is
necessary, both to keep the
thickness of the lags within
reasonable limits, and to
avoid short grain. Gene-
rally flanges, and frequently
mouldings, are fitted into
shallow grooves turned in
the shaft (fig. 35, top), and
with the grain running
transversely. They are
either glued-up in segments,
the better way, or cut solidly
and not recessed (fig. 35,
bottom). In some cases it is better to glue blocks on the lags, and to

Fig. 36.—Shows Print Continuous with Lags

turn the mouldings from these,
of these supplementary parts.

The choice depends on the proportions
72                                  PATTERN-MAKING

The fitting of the end core prints depends on the relative diameters of
prints and shaft. If the difference is only that due to the thickness of metal
in the shaft, prints are turned on an extension of the lagging (figs. 34 and
36). But if they have to core out a large moulding, then they are better
fitted separately (fig. 35), the lags terminating with the moulding. When
large square bases are fitted to columns, these are prepared separately and
attached. The square prints being large, are boxed up and screwed
against the end of the column and its flange (fig. 37).

Fluted Columns.—The problem in these is that of providing for
delivery of the undercut flutes. The pattern shaft is built with lags, having
flats to receive loose strips in which the flutes are planed. The divisions

Fig. 37.—Method of Fitting a Square Base and its Print

between the strips are determined by the amount and direction of undercut.
In the withdrawal the shaft is taken out first, then the loose pieces adjacent
to the mould joint are removed, and finally those in the bottom. This will
be clear from the section (fig. 38).

In the construction of these patterns, the internal shaft, the body which
forms a backing for the fluted strips, is prepared; the strips are attached to
it with screws put in from within the body, to be taken out in the mould;
and the strips are turned. The edges of the flutes are divided round and
marked along, the strips removed, and the flutes planed. To permit of
planing through, the end pieces where the flutes terminate are screwed
temporarily. Fig. 39 shows a column section where the body is of cast iron
made for permanent service. Metal screws hold the lags, being tapped into
plates sunk in the flute strips.

The cores for columns are usually swept against the edges of boards,
unless large numbers are required, when they are rammed in a half-box,
each half-core being united to its fellow. In plain and moulded columns
the whole of the core can be swept, including the enlarged portions for the
mouldings. But if there is a large square base, as in figs. 37 and 38, or a



heavily foliated capital, cores for these sections must be rammed in boxes
having a central print of the same diameter as the core for the shaft.   This
leaves a hole to  fit over the

Cylinder Patterns. —
Most patterns of this class are
divided longitudinally through
the centre, notwithstanding the
fact that they are in the majority
of instances set vertically for
pouring. An exception occurs
in the largest cylinders, which
are moulded vertically, fre-
quently from skeleton patterns,
or swept in loam. A fair num-
ber of moulds of medium
dimensions are taken from loam
patterns, which also are un-
jointed. Patterns of metal are
used for highly repetitive cast-
ings in the smaller bores.
With these exceptions, cylinder
patterns are built with lags simi-
larly to the pipes and columns
just noticed. They have parallel
prints for the main core, usually
head metal, and flanges pre-
pared separately from the body.
All this is simple, plain work.
The difficulties that occur in
cylinder patterns and moulds
are those associated with the
preparation and the setting of
cores, which increase with their
number and tenuity, and are
the most frequent cause of
the rather high proportion of
" wasters " that are produced
in some foundries.

Any cylinder, whether
simple or complex, must be
drawn to actual size on a shop
board, with the machining
allowances and the positions

and dimensions of core prints included, in all aspects and sections.   On and
from this the pattern parts and core boxes are tried and checked as the

Fig. 38.—A Fluted Column with Square Base



work proceeds. Some pattern parts have to be left loose, fitting with
dowels or skewers. The locations of these are determined by the method
of jointing and moulding adopted, this being settled by the pattern-maker*
In some cases, alternatives present themselves, in others only one method

Fig. 39.—Example of Fluted Strips attached to an Iron Backing

is practicable. Usually the most convenient and the safest method of
setting and securing the cores determines the choice. A very slight degree
of inaccuracy in setting, or due to shifting from position in casting, will
produce spoilt work. Moulds are divided horizontally, because it is easier
to set cores thus than in a vertical mould of small diameter. But it is set

vertically to be poured,
in order to float all
sullage up into the head
metal, which if present
on portions to be tooled
would spoil the casting.
Typical Patterns.
—In the plainest cy-
linders the steam-chest
is distinct from the
main body, and a flange
on the latter is provided
to receive it. When

practicable, the flange is always moulded downwards, because it is con-
venient to insert the cores for the passages in the bottom of the mould,
instead of in the joint face. Prints are attached to the flange, and this is
necessarily dowelled loosely to the passage block. The cylinder foot, when
at right angles, is made fast to the body. But it may happen that other

Fig. 40.—Cylinder Pattern with Steam-chest


dispositions of the foot may entail coring over it or jointing the pattern at
right angles with the steam-chest face.

In many cases the steam-chest is cast
in one piece with the cylinder body (fig. 40).
Then the interior is produced with a core
for which a print is attached, wide enough
to afford adequate support to the core, and
prints are inserted in the box for the steam
and exhaust passage cores (fig. 41). The
pattern portion for the steam-chest is pre-
pared by boxing-up in order to reduce
shrinkage and to economize timber.

Many cylinders are jacketed. The an-
nular core is made in a box, complete in
all details. All jacketed castings require
especial care in both pattern-shop and
foundry, because the metal is thin and the
risks of displacement of the cores and
obstruction of the vents are very great.
Steam, gas, and petrol cylinders are made
with jackets, and the last named are the
most difficult of all, because of the large
number of cores and their interdependence, Fig> 4I._core BOX for steam-chest Cast
and the very thin walls of metal between                    with Cylinder

them, ranging from about J in. to f in.
When two, three, or four cylinders are cast en bloc the separate cores may

Fig. 42.—Cylinder Pattern for Motor-cycle Engine




number from twenty to thirty, depending on the design. These are almost
invariably rammed in iron boxes to ensure permanence of form, and their
positions in the mould are tested carefully by means of metal gauges.

Fig. 42 shows a plain cylinder pattern
for a motor-cycle, by Messrs. Ernest M.
Brown & Co. of Huddersfield. One-
half the core box is seen at the right.
The relation of the core to the pattern
and its prints can be observed in the
half pattern open in the joint face to
the left. There the thickness of metal
is painted black, a practice which is
commonly adopted in cored work, since
it is of assistance to the moulder when
inserting the cores. A cover is seen at
the left of the figure.

Fig. 43 shows the method of lagging,
with other details for a small compound
engine, in which the high- and low-
pressure cylinders were cast together with

T« 43-—Pattern for Compound Cylinder

Fig. 44.—Pattern of Diesel Engine Cylinder
and Base


their connecting passages. The upper view is a cross-section, the lower
is one-half the pattern open in the joint face. Head metal is provided.
The lags, the flat strips, the battens, bosses, and prints are obvious.

Fig. 44 gives complete views of the pattern for a Diesel engine cylinder
cast in one with its A-legs. The whole of the interior is formed with cores.
The upper portion A, the cylinder, is lagged and turned as a separate section,
and is united to the pattern frame B with two dovetails. B is boxed-up on
three cross-bars, being formed with strips having open joints. The feet are
attached to this, as shown in the lowest view, which is a plan taken from the
top of the cylinder, and the print is fastened to the bottom of the pattern.
Two diagonal brackets are fitted loosely with skewers, and four hold-down
bosses that lie below the joint of the pattern are left loose, with drop
prints. A bracket c at the top is prepared separately, with the print for
its lightening core, and attached to the cylinder.


The features which sheaves, pulleys, and fly-wheels possess in common
are: their outlines are circular, their depths relatively shallow, and they have
central arms (or discs), through the centres of which the moulds and often
the patterns have to be parted. The shrinkage stresses in arms and rims
cause fracture in these castings unless the pattern-maker exercises care in
proportioning of parts.

Sheaves.—Patterns for these are made in wood (fig. i) for moderate
numbers of castings, in metal
(fig. 45) for quantities. When
wood is used the rims are
built up with thin segments,
the centres being made in a
similar manner when they

are    SOlid-plated.      Arms    are                   Fig. 45 .—Iron Pattern for Sheave Wheel

locked at the centre and let

into the rim, these methods being identical with those employed in the con-
struction of toothed wheels. Patterns are divided through the centre of
the arms, or these are left of the full thickness, and the upper portion of the
rim is registered to the lower (figs, i and 45). Moulding is generally done in
a three-part box. the joints being made in the planes a, a (fig. 45). But a
circular grid can be used, as shown at the right hand, to carry the sand in the
recess, and then a two-part box with its joint at b can be used. Alternatively,
an annular core print can be fitted around the rim, and a series of short
lengths of core laid in. This is seldom done when complete patterns are
made, except in the case of cupped wheels which are provided with recesses
to receive the links of chain that lie flat and edgewise alternately. But the
method is of value when very large pit wheels are made with wrought-iron
arms, in which case the entire rim, interior as well as exterior, is frequently
formed with cores. Another form of sheave is that with a wavy gorge to



i I

g> 4.5.—Alternative Methods of Jointing Bottom Flange of
Trolley Wheel

prevent slipping of a rope, which also is provided for in a core box rather
than in the pattern.

Trolley or truck wheels resemble sheaves in the fact that the presence

of double flanges (fig. 2)
entails the employment
of a three-part box. Only
the lower flange is left
loose. This may be done
in either of the ways
shown (fig. 46). When
vertical arms are fitted,

either to trolley or sheave wheels, they are screwed fast in the bottom, but

left loose in the top, to come up with the cope and be withdrawn therefrom.

Pulleys.—Patterns of wood are useless for pulleys.    They must be

of iron. And, except
for repetitive work, they
are not made with rim,
arms, and boss in one
solid piece. Each is a
separate element, rim
and arms in iron, and
bosses in wood, from
which pulleys having
different widths of face,
and bosses for any
bores required can be
made up.

The system adopted
is to have a large stock
of pattern rims, turned
inside and out, with a
very slight taper and no crowning, of maximum depths likely to be
required, say of 12 in. width of face in the smaller sizes and 16 in. to
18 in. in the larger. Widths narrower are produced by stopping-off in the

mould. Diameters may advance
by i in. in the first, and by 3 in.
in the second. When the volume
of trade is large, one series of light
pattern rims and one of heavy is
stocked. The arms are made of
cast iron to fit easily within the
Fig. 48.-Core BOX for Fly-wheel Arms              rims< These also are made light,

having only the elliptical section,

and heavy with shallow vertical ribbings.   The bosses of wood fit to any of

the arms with a standard size of stud in a centre hole, say if in. diameter.

From these elements the moulder produces pulleys of any widths of

* 47-—Pattern of a Fly-wheel



Fig. 49.—Segmental Piece for
Rim of Fly-wheel

Fig. 50.—Core Box for Fly-
wheel, half boss

face by stopping-off, and centring the arms with a gauge, and, if required,

pulleys of wider faces than the rims by " drawing". Double-armed
castings are made from the same sets. From the
same pattern parts, castings are " split" in halves by
the insertion of lugs and prints to receive the split-
ting plates.

Fly-wheels.—The rims of these (fig. 47) are
built up with segments, and the arms, locked about
the centre, are sunk into the
rim during the course of build-
ing-up. Bosses are studded
at the centre. Patterns of wood
are suitable, except for highly
repetitive orders, for which
metal is substituted, in which
case the work is machine-
moulded. All fly-wheels, ex-
cept those of small diameter,
have arms. These may be
straight, but are preferably
curved to accommodate
shrinkage movements. The

smaller wheels have single curves, the larger generally double.     But solid-
cast arms are not safe for the larger wheels, which are either provided

with those of wrought iron, or the arms and

rim are cast in separate pieces and bolted or

cottered  together.     When large wheels are

cast with arms intact, these are made in cores,

for which the pattern-maker provides a box,

and also sweeping boards to form the rim.
When a wheel has cast-iron arms, the form

of core box used is shown in fig. 48.   The

arm piece is one-half the thickness of the arm

section, so that two cores are jointed to include

the mould for the complete arm.   The box is

shown as for a six-armed wheel, the jointing

angle at the centre which contains the boss

section being therefore 60°.   The outer radius

is that of the interior of the rim.   The notches,

cut in the edges of the   box   frame, receive

the grid which sustains the core.   Rims of any

section can be produced with sweeping boards

and sectional ramming blocks.

When  wheels   have wrought-iron arms,

these are cast into bosses in rim and central boss.   Obviously down-jointing

cannot be done, and therefore the upper halves of the rim bosses and their

Fig. 51.—Pattern Boss for Fly-wheel


prints must be cored over, and the mould and cores be coveted with a plain
top. Fig. 49 shows the provision made for the rim. A short length of
sweep has a half-arm boss with its half print, covered with a block print.
Into the impression made by this the core, rammed in the box (fig. 50),
is set. Taper is given, as shown, to the sides of print and core. The
central boss that corresponds with this type of wheel is shown in fig. 51.
This is rammed in a parted box distinct from the rim, having the joints
separated for the insertion of the arms, and is dried. The boss mould is
centred relatively to the rim, and levelled before the arms are inserted.
These are then covered with the top half of the boss mould.

Although this department of work has been deeply invaded by the insistent
demand for cut gears, a very large volume remains. Wheels with cut teeth
are expensive, and they are not usually found in common machines, such as
ordinary cranes, contractors' machinery, and the like. Another important
fact which favours the retention of cast gears is that the patterns now
made are far superior to those of some years ago. A high grade of work-
manship has been demanded and met, partly due to the employment of
machines for cutting pattern teeth, and partly to the fact that firms make
these for the trade, the pioneers being Messrs. Ernest M. Brown & Co.
of Huddersfield. And Wadkin & Co. of Leicester have revolutionized
the methods of some shops by the introduction of the " Mechanical
Wood-worker " in core-box work, and in the teeth of gear-wheel patterns.
In the general shops these patterns are the speciality of one or two only of
the hands.
Tooth Forms.—It is essential that the teeth of all wheels of the same
pitch shall be made to a correct contour, so as to secure a rolling contact as
far as may be and a uniform velocity-ratio. In cycloidal or double-curved
teeth this is secured by making the diameter of the rolling circle, to be
rolled on the pitch circle, equal to the radius of the smallest wheel of
the series. This gives radial flanks for the smallest or basic pinion, and
undercut flanks for those below that size. This is embodied in an odontograph
For involute or single-curved teeth, which have been largely substituted
for cycloidal, the basis is the rack, having teeth with straight, sloping flanks.
The point of contact of the teeth lies on the line passing through the point
of contact of the pitch circles and tangential to the base circles. In the
cycloids, curves are generated from the pitch circle; in the involutes, the
pitch circles have but an arbitrary relation to the base circles. This explains
why correct tooth contact occurs whether the ideal pitch cylinders are or are
not in contact, and why, by increasing the addendum in small pinions, under-
cut of the teeth can be avoided. The circular pitch is most generally used
for pattern gears, but the diametral is commonly associated with the involutes.


Tooth lengths are proportioned to pitches, but teeth are always made shorter
now than they were formerly. Proportions are given in many textbooks,
and they are standardized in the shops.

Spur Wheels, Pattern Construction.—For these, two materials are
used chiefly: yellow pine and Honduras mahogany, or Baywood, the first
for the bodies, and the second for the teeth. Yellow pine is suitable for
teeth when only moderate numbers of moulds have to be taken.

Only very small pinion patterns are made solid, that is, with the teeth
in one with the centre body, and the grain running longitudinally. Pinions
of over 6 in. or 7 in. diameter must have their centres built up. In the
smaller sizes, courses of sectors are glued up, the grain running radially.
In those above say 8 in. or 9 in., seg-
ments are used, the grain running
tangentially. Thicknesses will range
from \ in. to I in. in small and large
patterns respectively. Gluing is done
carefully, and nails or wooden pegs
reinforce the joints against the rough
usage of the foundry. The rims are
turned and finished before the teeth
are taken in hand, these being always
made distinct from the rim to get
longitudinal grain.

Methods of Constructing
Wheels.—The larger pinions, and
smaller wheels, have solid-plated
centres, built into the rims. All large
wheels have arms made separately
from the rims, which are built up.
„ Plated centres are built up with sectors
having the grain radiating, in not less
than two thicknesses. Or narrow strips with open joints are prepared, and
the courses of rim segments are glued up on the discoid centres. Rims for
armed wheels are built up and turned as separate elements into which the
finished arms, usually of T-section, are fitted. When they have the section
of a + they are built into the rim at the half-way stage of the courses of
segments. During the fitting, care must be taken not to drive them into
their recesses so tightly as to distort the rim. Only light hand pressure is
employed, with glue and fine screws. Though the locking together of the
arms at the centre is rather flimsy, the screwing on of the central boss and
the fitting of the vertical arms provide additional strength. The latter abut
against the boss, or fit in shallow grooves cut in it. They also abut against
the rim. Fillets or " hollows " glued in all angles further stiffen the struc-
ture. With arms of -(--section, the ribs that come in the top should be
do welled loosely with that boss portion, for reasons previously stated
Fig. 52 shows a wheel pattern with split lugs.

VOL. I.                                                                                                                         6

Fig. 52.—Portion of Pattern Wheel with
Splitting "Lugs


Tooth   Formation   and Fitting. — Teeth  are  either  shaped first
and attached to the rim afterwards, or they are worked in their places.

r—~,  r™!

Fig. 53.—Method of Fitting Teeth with Dovetails

The latter is to be preferred when a dividing-machine and a fly-cutter can
be used. But if not, the best way is to fit each tooth with a dovetail (fig. 53).
Turn, pitch, and mark the teeth out in place, remove them to be shaped

with planes, return and glue them permanently.
The best pattern wheels, apart from those
machine-made, are constructed in this way.

Other methods are, to plane each tooth sepa-
rately, fit and glue it to the rim, or to glue rough
blocks on the rim, turn, pitch, and mark out
(fig. 54), and cut the shapes through with chisel
and gouge. These methods are however not
entirely satisfactory.

Using a fly-cutter, the rim is turned J in.
or A in. below the roots, and rough blocks, each
wide enough to include three or four teeth, are
glued on in contact. Material is thus left below
the roots on which the radius or fillet is cut
without leaving a " feather edge ". The blocks
being turned, the tooth spaces are shaped in a
machine, which also divides for pitch. No taper
is given, and since the teeth are accurate, they
can be drawn through a stripping plate, in
hand-made moulds, or on a machine.

The chief advantage of planing  the teeth
previously to attaching them to the rim is that
they can be shaped accurately.  They are planed
in a box made of hard-wood, having the cross-
section of a tooth.     The difficulty lies in gluing the teeth to the rim.
Setting is done by centre lines, or by the edges of the flanks, to lines
pitched and scribed around the rim.   Errors in pitching and in getting

Fig. 54.—Method of Fitting Teeth
to be Shaped in Position

out of square arise. The first is checked with calipers as the work
proceeds before the glue has hardened, the second with a set-square
tried along a flank and working from a straight-edge.

To glue rough blocks on and work them in place with gouge and chisel
requires great care to get straight flanks. Using a small straight-edge, and
carefully glass-papering, the method is reasonably accurate, though tedious.

Since rough blocks, glued or dovetailed, have interspaces, these may be
filled in with wedge-shaped bits (fig. 54) to afford a continuous surface. It
is convenient for the turning, but not essential, since if light cuts are taken
with a sharp gouge, the teeth will not be knocked off nor the grain split.
And although a continuous surface is useful for locating tooth curves on,
the centres may equally well be set on a zinc templet piece as shown in the
upper part of fig. 54, worked round the periphery.

Tooth centres are pitched round on one side, and squared over to the
side opposite. The tooth thicknesses are set to right and left of the pitch
points, and the curves starting from these are described.

Bevel Wheels —These are based on precisely the same principles
and elements as the spurs, in regard to the shapes and proportions of the
teeth. But the pitch and related dimensions are always taken on the
major diameter, those on the smaller diameter being controlled by the
width of face of the teeth. The teeth are not developed on the real
diameters, but on conical surfaces at right angles with the pitch cone

(%• 55)-

Bevel gears are marked out as shown by fig. 55. The pitch cones ab
are the primitive rolling surfaces. The diameters A, B are the real diameters
for the actual pitches. Through #, the point of intersection of these, a
line is drawn perpendicular to ba, meeting the axes of the primitive cones
in c and d. Circles described with radii ca and da are the pitch circles
on which the teeth are drawn. In other words, they correspond with the
curves of spur gears of radii ca, da. As the teeth taper from the major
diameter to the apices of the cones, the tooth curves on the minoi
diameters are obtained on the developed surfaces having radii fe, ge.
The tooth forms for both are shown to the left, and those for the minor
radii are repeated at the right.

Pattern rims are built up with courses of segments that overlap sufficiently
(fig. 56) to include the cone section. Two chuckings are essential; whether
the back or the front is done first does not matter, since a straight-edge is
laid across the rim and each is turned with the aid of templets. Nails cannot
be used so conveniently to reinforce the glued joints as wooden pegs, though
the risk of segments starting after the teeth are attached is nearly negligible.

Teeth are fitted and worked in either of the ways described in connection
with spur gears. If fly-cutters are used, they are not selected for either
diameter, but for a location at about a third of the tooth length from the
major diameter, and two settings of a flank are needed.

Worm Wheels.—These are made much less frequently by the pattern-
maker than formerly, since the practice of generating has grown in favour,

; I;


with the employment of double- and treble-threaded worms. For ordinary
service, cast gears with single-threaded worms are still used, and are less
costly than those produced by hobbing.

Worm gears have the helix for their basis, though this is somewhat dis-
guised in the case of the wheel.    The worm is a continuous thread of ex-

Fig- 55-—Development of Bevel Gears
tremely short axial pitch or lead. The wheel which it drives has a number
of short helical teeth of extremely long axial pitch. The axial pitch of the
worm, if single-threaded, measures the same as the circular pitch of the
wheel teeth. The wheel may contain any number of teeth. The worm
diameter is usually from twice to three times the pitch. The curvature
therefore being small, the teeth of the wheel should form envelopes of the
worm thread to ensure durability and smooth movement.
The section of the worm thread is that of the involute rack. ' A worm

pattern can be constructed in wood, divided
longitudinally, but this is not a very satisfactory
method. If a cast worm is used, the pattern
should be cut in metal and moulded vertically,
screwing it out of the mould through a stripping
plate, and relieving its weight with a counter-
balance suspended from a rope passing over a
pulley. But it is better to cut worms in the
machine-shop, in which case the pattern-maker
can employ the actual worm as a perfect guide
by which to shape the teeth of the wheel.

Pattern wheels must be jointed along the
middle plane, either through the central plate,
or leaving this in one piece, undivided, the half
depth of rim is registered to it, as in the case of
sheave wheels. Segments are built up, overlap-
ping, and the concavity for the tooth blocks is
turned with a templet, the interior of the rim
being similarly dealt with. Blocks for the teeth
are fitted and glued to each half-rim, and the
abutting ends that come in the joint face are
turned at separate chuckings of each half-pattern,

and at the same time the curves are imparted      Fig. 56—Rim for Bevel wheel
to the points with the aid of a templet working t
from the joint face.    The outer ends of the teeth are then finished.

The teeth are pitched and their thicknesses and shapes marked in the

Fig. 57.—Worm-wheel Pattern mounted for Cutting the Teeth


central joint plane of the pattern precisely as for an involute spur wheel.

The tooth sections change constantly from the centre to the outer ends.

The larger the angle or
slope of the teeth, as in
worms of small diameters
and those with multiple
threads, the more marked
are the changes in section.
Here the advantage of em-
ploying the worm as a
templet guide for cutting
the wheel teeth is apparent.
The worm is set between
lathe-centres, and the wheel
is mounted on a stem in
the T-rest. The wheel
and the worm are moved
into contact (fig. 57). The
application of chalk or of
red lead to the worm
indicates, by its trans-
ference to the wheel
teeth, the high parts from
which material must be

If the wheel contains
a large number of teeth,
the work of cutting may
be hastened by shaping,
say, half a dozen correctly
from worm contact, and
then marking the shapes of
the other ends so obtained
on the remaining teeth.
These can then be roughed
out rapidly with gouge and
chisel, leaving the finish to
be imparted by the assist-
ance of worm contact.
A massive spur pinion

Fig. 58.—RoUing-mill Pinion, 15 teeth, 7-in. pitch, 26-in. face         pattern, which Stands higher

than a tall man, is shown

in fig. 58. The teeth are shrouded to the pitch circles. There is a joint
in the pattern along the face of each shroud. These are built up with,
segments, as are also the bosses. Fig. 59 shows a segmental pattern, from
which large toothed rings are built up, being bolted together by the end


Kg- 59-—A Segmental Pattern for building up a Toothed King

flanges, for use on revolving cranes, turntables, and swing bridges,
are  by Messrs.  Ernest                  .. ..

M.   Brown   &   Co.   of


The economies of
this kind of work are
associated chiefly with
the larger gears, and for
those casual orders when
only two or three cast-
ings are required, for
which the cost of com-
plete patterns would be

The employment of
a few teeth from which
to mould an entire wheel
rim, and of a core box
to include the arms, had
been practised long be-
fore the wheel-moulding
machines were invented.
Teeth attached to a seg-
ment are worked round,
and rammed in succes-
sive stages, at the end of
a radius bar centred on a
pin (fig. 60). Another


Fig. 60.-—Segmental Block moulding a Spur Wheel




method is that of ramming a few teeth in a core box, and laying the cores
round a circle. The wheel machine includes a dividing apparatus with
change gears for all pitches, and mechanical slides for withdrawing the
segmer.tal blocks that carry two, three, or more teeth.
Two designs of machines are made, one having a
moulding table on which the smaller gears are moulded
in top and bottom boxes, the other having the me-
chanism carried on a column sunk in the floor, in the
sand of which the teeth are moulded, to be covered
with a plain top box.

Fig. 61.—Tooth Block in

Fig. 62.—Sweeping Board for a Spur Wheel

The Pattern Parts.—The essentials, varied in details with the class
of gear and the sections and outlines of arms, &c., are the tooth block, the
sweeping boards, and the core box.

Tooth Blocks.—A tooth block is like part of a wheel rim having a few
teeth cut on it, attached to a backing of suitable shape and dimensions, and

I screwed to the tooth carrier of a
machine. The simplest blocks are
those for spur (fig. 61) and bevel
gears, which are withdrawn vertically.
Those for helical, double helical, and
worm wheels are withdrawn in the
horizontal direction, except in those
machines which do not include this
provision. For use in these, the
pattern-maker divides the block,
separating the main backing from the actual teeth, which are carried on
a thin backing, and dovetailed loosely to the portion that is attached to
the carrier. The latter is first lifted vertically by the machine, followed
by the teeth, taken away horizontally with the fingers.

Sweeping Boards.—These (fig. 62) are necessary to form a bed to
receive the cores, to make the joint faces at their correct heights between
drag and cope, to indicate the radii of the teeth, and, in the case of bevel

Fig. 63.—Sweeping Board for a Bevel Wheel


wheels, to correspond with the tooth points. These vary in details with the
shapes of wheels. A plane top is general for spurs, but the top for a bevel
wheel is curved to follow the edges of the vertical arms (fig. 63). The edges
for sweeping the bottom and top moulds are usually cut on the same board
which is reversed on the bar. The latter is of a standard size (fig. 62), so that
all boards are shorter than the real radius of the wheel by the radius of the
bar, to which they are attached with
an iron strap. The radius of the
tooth block, though set by the bed
swept, is checked, and, if neces-
sary, finely corrected with a strip
that gives the exact distance from
the central bar to the point or the
root of a tooth.

Core Boxes. — Arms of all
shapes can be made with cores,
but the most convenient are those
of H-section, and these therefore

are mOSt  COmmon (fig.  64).     Bevel         Fig. 64.—Core Box for Wheel Arms of H-section

wheels   have  arms   of  T-section.

Cores are rammed, dried, blackened, and set in the mould on the swept
bed without aid from prints. The spaces between the cores corresponding
with thicknesses of metal are set with wooden gauges. Their own weight
and the pressure of the cope when the mould is closed prevents them from
shifting. Central bosses and prints are swept, or bedded-in.


These being the bases for engines, pumps, machine-tools, cranes, &c.,
occur in an immense variety of outlines and dimensions. Only broad
principles can be stated here.


Fig. 65.—An Engine Bed suitable for Self-delivery—the top face being lowermost in the mould
Method of Moulding.—This is the first thing to be determined.
Usually the top face of the bedplate pattern goes to the bottom of the mould.
This ensures that sound metal shall be present in those surfaces which

have to be machined later.   The question of moulding by bedding-in
or turning-over is settled by the numbers of castings required and the

Fig. 66.—Section through Crank-shaft Bearings of Engine Bed.   Pattern shows boxing-up,
printsjfor cores, and loose pieces

boxes available. The latter method is preferable, except for beds of the
largest dimensions. The choice of self-delivery, or of coring the interior,
depends chiefly on the bed section, and on the
relative proportions of width to depth of interiors.
As there is no objection to giving plenty of taper,
a good slope is always given to the outside, and as
the thickness of metal is equal throughout, the
internal taper favours delivery (fig. 65). Many
deep beds therefore with wide internal spaces
deliver themselves, the interior " green sand core "
being carried on a grid suspended from the stays
of the top box, or, if special boxes are made, the
stays are brought down inside to a distance of about
| in. from the pattern all round. But this method is
not practicable when beds are narrow and deep. In
these cases, the interior is taken out with cores in-
serted in print impressions (fig. 66). This is very
convenient when loose pieces have to be attached to
the outsides, as these can be withdrawn laterally
through the open interior.

There are few patterns which do not carry some
loose pieces, and core prints for the insertion of
small cores (figs. 66 and 67), for bearings, and recessed

portions in varied forms.   In some cases it is convenient to carry all the
outer mould on an encircling plate for the purpose of getting at recessed

y.-shows Loose BOSS
*on end °f Engine

portions for cleaning and coring.   Many beds of large dimensions are made
without full patterns from motives of economy.     The exterior mould is

Fig. 68.—Formation of Curved

Fig. 69.—Method of making a Curved Corner

made from a skeleton frame, aided with strickles or sweeping boards, and
cores impart the shapes to the interior portions. In all long and narrow
beds, solid-plated on one face only, the effect of unequal shrinkage is to

Fig. 72
Methods of Blocking adopted for Curves

cause curving or camber, the solid-plated portion when cold becoming
concave lengthwise to the extent of from J in. to | in., depending on the
length. The pattern must be curved in the opposite direction to neutralize
this effect.


92                                   PATTERN-MAKING

Methods of Union.—The union of corners and the formation of
curved ends are shown in figs. 68 to 72. In fig. 68 sides and ends abut at
right angles, and a square block is glued in, which when set is worked to
interior and outside curves. In fig. 69 a similar method is employed for
interior and exterior radii, all being connected with a plated covering, after
the inner as well as the outer curves have been cut, the interior being for
self-delivery, as is the previous figure.

The next group represents portions of patterns, the interiors of which
have to be cored. This permits of making stiffer constructions. The
thicknesses of stuff are greater, and screws can be used if thought desirable
to assist the glue. Figs. 70 and 71 are alternative methods for the outside
curves. Both are strong, and are reinforced with the screws that secure the
plated portions, halved at the corners, to the verticals. Fig. 72 is a semi-
circular end, made in the strongest way. Risk of shrinkage is reduced by
making the segmental blocks short, and they are reinforced with the strips
glued in the angles. The inside curve is made of two pieces having the
grain running perpendicularly, in order to avoid short grain at the ends.
The plate is made with strips having half-lap joints, and is screwed to the
sides. This method is suitable for the semicircular ends of beds and of
other patterns of that type.


This work includes the production of helices in pile screws, conveyors
for elevators, worms, and propeller blades, cut in wood or swept up. The

patterns for these contain one or more than
one revolution, or a fractional portion of a
revolution. Although made in different
ways, the principle involved in all is the
same, viz. the development of a helix is an
inclined plane, or conversely a helix may
be imagined to be an inclined plane wound
round a cylinder. This is translated into
actual practice in many small patterns by
cutting an inclined plane in paper and
wrapping it round a cylinder as a guide for
working by. One templet of this kind may
be used for the base of the screw, the other
for its tip. The pitch is alike in each, but the
lengths of the envelopes and the angles of the
helices differ. The pitch is the distance be-
tween the centres of a helix or blade when
FIR. 73.—Marking the Tip of a Pile Screw it has made one revolution. The diameter

is measured across the tips of the blade.

Pattern Construction.—In pile and conveyor screws (figs. 73-75),
and in worms, which are members of the same family, the blades or



threads would suffer from very short grain if they were cut in one solid
with the cylindrical body. This therefore is prepared first, being jointed
along the centre and dowelled. The blade is fitted in short segmental
divisions and glued permanently into shallow grooves cut around the body,
but if the blades are deep they must be screwed temporarily, to be with-
drawn from the mould after the delivery of the cylindrical body.

It is generally convenient to mark the width of the grooves on the templet

Fig. 74-—A Pile Screw

sheet of paper, through which they are cut on the turned body. The length
of the pitch and of the circumference at radius r form a rectangle, whose
diagonal is the development of the helix at radius r. Another line is drawn
parallel with this to represent the width of the groove to be cut in the
cylinder. If there is more than one turn of the helix, the construction is
repeated. The paper when glued round is an accurate guide to the work-
man who cuts the spiral groove with saw, chisel, and router. The segmental
blades are fitted into this, prevented from overlapping by the insertion of a
flat dowell next the outer edges, and secured with screws put in from the
joint faces, for which the pattern has to be removed from the lathe, and
taken apart.




Although this method is convenient when a continuous cylinder affords
a good basis for the paper, it is not practicable for the tips of the blades,
since the helix is not cut in solid stuff. Then the method of intersecting
lines is adopted. Here the circumference and the pitch are divided into
the same number of equal parts; the larger the number the more nearly
accurate will be the results. A diagonal drawn through successive inter-
sections will delineate the screw thread (fig. 73). A line drawn parallel with
this is required for the thickness of the blade at the tip. As there is a gap
between the threads, the divisions are marked on a slip of wood. The

i  1


1 <•



Fig. 75.—A Group of Conveyor Screws
divisions are scribed off on the pattern revolved in the lathe. The blades
have to be removed to be worked. This is done with a narrow plane,
slightly convex on the face. Radially, every portion of the surface from
centre to circumference must be straight.
Propeller blades are short sections of multiple screws, two, three, or four
in number. When pattern blades are made for these, in the smaller sizes,
the boss is included, and the blade is glued up with strips that overlap at
the edges to embrace the screw formation, worked through with planes.
The method of intersecting lines is adopted.
Screws produced with Templets.—Large propellers are swept up
in loam by the aid of sheet-iron templets having the upper edge cut to the
inclined plane that corresponds with the slope of the face of the blades. As
many templets are cut as there are blades. These are set round in a circle,


their upper edges guide the movements of a sweeping board which pro-
the shape of the loam beds on which the strips, that correspond with
^hanging sectional shapes of the blades, are laid.    This work is repeated
^any times as there are blades, the templets being set equidistant round

Screw Drums.—These, grooved spirally to receive the wire ropes
"the chains used on large cranes, are seldom cut in wood, because the
is too great. They are sometimes cored, but a cheaper and more

Fig. 76.—The Use of a Templet Screw to control the Striking Board for the Loam Pattern
of a Spiral Crane Drum
ourate method is to sweep them in loam.   Drums up to about 3 ft. 6 in.
diameter are swept as loam patterns to be moulded. Those over that
ce are made as loam moulds with the axis of rotation set vertically. In
.cti case the pitch of the screw-thread is reproduced from a templet which
urtrols the longitudinal (fig. 76) or the vertical movement of the sweeping
>a.rd. The templet is cut by the guidance afforded by inclined planes
ajrked on paper and glued within and without. The grooved sections are
it on the edge of the sweeping board. Since this is moved through a dis-
rtce of one pitch during one revolution, the result is a true screw in loam.
space equal to the thickness of the board has to be filled up and made
>od by hand, because the board has to be moved back to its starting position
times before a smooth loam surface can be completed.



The practice of attaching patterns to plates has grown enormously in
consequence of the immense developments of machine-moulding. But it
ante-dated this, and is in extensive use apart from the aids afforded by
machines. It is derived from and is an extension of the employment of
joint or bottom boards.

Bottom Boards.—The bottom or joint boards, which are stocked in

many sizes in foundries,
are made of thick, narrow
strips with open joints
united with battens. They
have holes bored to receive
the pins of the bottom
parts of moulding-boxes,
and are of general utility,
since any patterns that
will go in a box can be
rammed on the bottom
board. Two results are
achieved, one being that
the ramming of a dummy
mould, for the sole pur-
pose of getting a joint
face, is avoided; the other
that the board affords a
level bed for the pattern,
so avoiding risk of its
winding during ramming.
Permanent Plates.
—At an early stage, when
work becomes repetitive,
an obvious economy is
secured by attaching pat-
terns to boards, and
making these and the

fitted boxes a permanent working unit. But then only the bottom box
can be rammed on the board; the cope must still be rammed on the joint
face of the bottom box, turned over to receive it. The next stage there-
fore is to attach the two portions of a pattern to a single board (fig. 77),
without battens, and to fasten both box parts together with the pins passing
through holes in the board. Here, though turning over is necessary, the
advantage remains that both joint faces are provided by the board, and that
both halves of the pattern are prevented from bending or winding. A more
advanced stage is that in which each half or portion of a pattern is attached

- 77.—Wooden Cock Pattern mounted on a Joint Board


to a separate plate. This enables two men or sets "of men to be working on
the same mould, one on copes, the other on drags, a very great economy,
which is necessary when a large output is required.

Metal Plates and Patterns.
—These are necessary for the
highest production, not only for
machine-moulding, with which
they are chiefly associated, but
also in the hand-moulding of
the smaller articles required in
large numbers. Patterns are
mounted on opposite sides of
an iron plate (fig. 78), or on one
side of separate plates. Weight
is kept down by making the
plates thin, say from f in. to
% in., and by lightening the in-
teriors of the patterns, though
these provisions are of less
moment when work is moulded
by machine than when done by
hand. Great care is necessary
when fitting the pattern parts to
their plates. Holes are drilled
through both in place, and these
receive dowel-pins or screws.

In   SOme   Cases    a   portion   Of   a        Fig. 78.—Iron Cock Pattern mounted on an Iron Plate

pattern may go right through a

plate.   All this work is rather of a special character, since patterns have

to be finished in the lathe, the grinder, and with files and scrapes.

In a good many instances patterns are cast integral with their plates, or
are made so by the method of their attachment (fig. 79).   This is most

Fig. 79.—Iron Pattern mounted on the Turn-over Table of a Machine
desirable when the jointing faces are irregular, having depressions on one
face and corresponding elevations on the other. These are readily cast,
after which the parts must be smoothed with file and scrape.
One great advantage of plating is that several small patterns can be put
VOL. I.                                                                   *                                                   7


on a plate, which in ordinary moulds would be arranged by hand, and that
all ingate patterns can be included, instead of cutting the channels laboriously
in every mould (fig. 80). The economies of these last developments are
such that the moulds for twenty or more small castings may often be made



Fig. 80.—Four Patterns mounted on a Plate

in the time that would be occupied for one in hand-work by the ordinary
method of turning over.

Both these are employed extensively for the largest castings when required
in small numbers, in order to economize timber and labour. Extra work is
always thrown on the moulder, but the question is one of relative cost. It
is rather remarkable how much can be done with strickles, sweeping blocks,
and skeleton frames, with the assistance of cores for dealing with the interiors.
Often the pattern-maker has to spend a considerable time in the foundry
assisting in the setting of parts and the checking of measurements.
Sectional Patterns.—This term includes a large variety of work,
:haracteristic of which is that the provision for making the moulds
_sts of strickles, strips, sweeps, boards, bosses, facing pieces, prints, and


-----rr~^-fo{~- - ~:==^--rj===^?*=£r

boxes, elements which in most instances bear only a remote resemblance
le casting, and which have to be supplemented with drawings, sketches,
-erbal instructions. Nearly all this work is moulded in. the floor and
:red with a plain top, in which pattern parts may be set by measurement,
ome cases the moulds are " open ". Most foundry tools, such as loam
2S and rings, gaggers, back plates, core plates, and the larger moulding
2S are made thus.

The first stage in making moulds of these kinds is the preparation of a
I   bed,  for  which   the

Jlel strips and the spirit-         ____LL

1 are requisitioned, or a
2ping board is worked
id a central bar. The
;r method is usually em-
red when a central boss,
alar facings, or shoul-
?d sections are wanted.
;se are formed by the
3 of the board, suitably
iled, the top edge of the
rd, parallel with the
:om, being set horizon-
7 with the spirit-level.
5 bed is vented to a
ler bed below, and the
ild is made on it from
sectional parts which the
:ern-maier supplies.
Broadly, moulds may be
aped as rectangular or
ular in plan. The first
be produced by the aid

hallow Strips Or Of deeper    Fig. 81.—Skeleton Pattern from which aU the Outer Mould is taken

rds set on the bed by

isurement, and retained in position with weights, against which the
d is rammed. Any extraneous portions, as strips, lugs, bosses, or
its, are set in their positions and rammed. A complete frame may be
ie (fig. 81) instead of separate strips, the outer mould being rammed
and it and supplementary parts attached, thus relieving the moulder of
responsibility of setting parts by measurement. If the interior is that of
bbed casting, that is formed wholly with cores.

Moulds which are circular in plan are produced by the aid of " sweeped "
ions. Any shape required can be readily Imparted to these (fig. 8z and
. As the length of the section is only 10 in. or 12 in., it has to he moved
nd and reset for successive rarnrnings. It is convenient to attach the
cxk to a radius bar worked round a centre pin. But this is not necessary,



if I

since having a level bed swept, the block can be set and reset on a circle
struck round with a trammel. It is held securely during ramming with a
weight. The interior is formed with cores. The two methods just de-
scribed are in common use for crane beds and centres which are of fairly
large dimensions, but which are seldom ordered in considerable numbers.

With the exception of open moulds, only used in making foundry ap-
pliances and the roughest castings, a top box-part is necessary. When a
sectional mould is made, the top cannot be rammed in its place on it, as is
done over a complete pattern. Then it is either swept with a strickle and
turned over on the mould, or it is rammed on a hard levelled bed of sand
away from the mould, transferred to the latter, and set on it by measurement.

Fig. 82. — A Sweeped
Pattern Segment, from
which a ring is moulded

Fig. 83.—A Sweeped Pattern Segment set for Ramming an
External Mould

The second method possesses this advantage over the first, that supple-
mentary pieces, as facings, bosses, brackets, Sec., can be laid on the prepared
bed in their correct positions, and the top box be rammed on them. This
is rather better than cutting away the sand in a strickled top and bedding
them in.                                                                                              .
Skeleton-like Patterns.—These differ from those just described in
the fact that they include the correct outlines, the complete contours, and
cardinal dimensions, but that the timber construction is not continuous.
The outlines are represented by a series of ribs, which leave open spaces to
be filled with sand. A large quantity of timber is saved, and labour is
economized, with no disadvantages to set-off. The method is employed for
large pipe-bends, large cylinders, condensers, and the casings of steam
turbines. It is used also in making alterations to some patterns. Enlarge-
ments of portions of patterns and reductions in diameters of core boxes are


effected by fitting strips of the required thickness round the curves, leaving
spaces between the strips about equal to their width to be occupied with

Fig. 84 is a group of various pipe patterns, two of which on the left
are skeleton structures.    The cylindrical portions are represented by discs,

Fig. 84.—Group of Pipe Patterns, including Skeleton Structures

leaving spaces between, which are filled with sand at the time of mould-
ing. The core prints are treated in the same way. The method is only
used for work of fair dimensions, and the larger the patterns are the
greater is the economy. These examples are by Messrs. Ernest M. Brown

Essential Machines
It is necessary to use the qualifying adjective, because some machines that
are absolutely essential in some shops would be like white elephants in others,
where they would be only partially employed. The larger the number of
hands, and the more varied the kinds of work done, the more extensive is the
selection of machines. Small shops handling specialities cannot afford to
neglect the facilities that special machines offer. There is a wealth of labour-
saving machinery now available, much of which is of comparatively recent
growth, notwithstanding that pattern-work is still mainly that of the handi-
102                                PATTERN-MAKING
The Lathes.—These take the first place in all shops, since all the
turning is done by the pattern-maker, who alone is competent to estimate
matters relating to taper, jointing, loose pieces, and moulder's requirements
generally. Lathes of from 6 in. to 8 in. centres are for common use. It is
desirable to have one with a set-over headstock for taper-turning. The lathe
has the ordinary tee-rest. The heads and rest are usually mounted on a
wooden bed, but iron beds are common. Lathes of higher centres, say iz
in., and having long beds are necessary in shops where pipe and column work
is done, and these frequently have a sliding rest. Large face work, as that
of fly-wheels, gear wheels, &c., is done on one of the long bed lathes, fitted
with a headstock spindle extended at the rear to carry a large face plate, and
having a floor rest there. It is better to have a face lathe with a deep head-
stock bolted to a floor plate, which may also carry a loose poppet, and having
a sliding rest on a stand mounted on a floor plate at the front. The chucks
used are simple and few, comprising the fork, the bell, and the face plates
of various diameters. The sheet anchor of the pattern-turner is the large
assortment of wooden chucks, made and used for a variety of patterns,
attached directly or through the medium of blocks, screwed on and recessed
to receive patterns for rechucking instead of cutting into the solid plates.
The Saws.—The circular and the band saws should form part of the
equipment of every shop. Suitable sizes of circular saws are from 14 in. to
18 in. diameter when new. The table must have a fence for cutting strips
to uniform widths, and a canting movement to the table is desirable for
sawing lags to a bevel without waste of material. A rising and falling table
is of value for rebating and shouldering. The band-sawing machine is
indispensable for cutting curves, and a tilting table permits of cutting bevelled
The Planing Machines.—Though a number of small shops do not
include these in their equipment, they are great time-savers. There are
three chief designs, the first machines one surface only, the second machines
parallel surfaces, and the third, by adjustments of the lower table, imparts
taper. So much of this kind of work has to be done in the pattern-shop that
the fully equipped machine soon recoups its outlay. The procedure is to
plane one face of the stuff over the top table, taking care not to exercise too
much pressure on the board, especially when it is thin and liable to spring
and produce a winding surface. The trued face is then placed on the lower
table, and carried along by the feed rollers, while the upper face is planed
with the revolving cutters. A fence is fitted to the top table for use when
the edges of boards are being planed.
The Wood Trimmer or Mitre Cutter.—This machine is hand-operated
through a lever, and saves a good deal of time otherwise spent in planing ends
of shorter pieces, held in the vice or laid on the shooting board. The
fences, two in number, for right- and left-hand cutting can be set to anj*
angle. Some of these go on a bench, others on floor-stands. The knives
in machines of different dimensions will take a good range of work, from.
7 in. long by 4 in. thick in the smallest, to about 18 in. by 5 in. in the largest.


The Mechanical Wood-worker.—No single machine has effected so
great economies in certain departments of pattern-shop work as the Me-
chanical Wood-worker, developed by Messrs. Wadkin & Co. of Leicester.
Previous to its advent, the statement that a single machine would tackle the
cutting of the teeth of gear wheels, the shaping of sweeps, of bend pipes, and
of the most intricate core boxes, would have been received with incredulity.
Yet this machine performs these functions, in addition to others of a more
general character.

The machine (fig. 85) is supported on a main frame, curved deeply in-

Fig. 85.—Mechanical Wood-worker operating on small Spur Wheel.   The whole of the teeth
cut in eight to nine minutes

wards to receive articles of considerable width. On this the overhanging
arm carrying the spindle-head floats up or down on sensitive bearings, with
a range of movement that will permit of its being raised above the hori-
zontal position, or lowered until the spindle is below the level of the work-
table. It can be set exactly horizontally or in any intermediate position.
The spindle head, at the outer end of the arm, swivels between the vertical
and horizontal, and can be locked in each or any intermediate position. It
carries a spindle and a chuck solid with it, ground to a No. 4 Morse taper.
It runs on two double rows of Hoffman ball bearings in dust-proof housings.
It can be rotated in either direction by means of a lever, a feature which is of
much value because it enables cutting to be done with, instead of against, the
grain. The spindle is fed to the work quickly by a hand-lever, and slowly
with a fine screw adjustment by a hand-wheel. The lever motion is controlled





by a spring plunger taper pin working in holes in a quadrant and having an
index, by which the depth of cut may be predetermined and the cutter
gradually fed into the work.

The work-table is massive, and is provided with tapped holes to secure
holding-down clamps. It has two motions at right angles, one operated by
rack and pinion, the other by screw and hand-wheel. It is mounted on a
pillar that travels along a runway which is bolted to the main frame. The
base of the pillar runs on anti-friction rollers, and is moved by rack and pinion.
The table can be turned through a complete circle on the pillar and locked,

Fig. 86.—Mechanical Wood-worker operating on Sectional Work of large Radius.   Any length
may be operated upon

and can be raised and lowered. An auxiliary table turning about a centre
pin is provided for small work.

The Range of Work Done.—Fig. 85 illustrates the cutting of the teeth
of a spur pinion, done in about nine minutes, a fair day's work if done by
hand. It is held in the universal head that is used for spiral and helical
gears. A gear-cutting fixture is inserted in the spindle, carrying a fly-
cutter having the same section as the tooth spaces.

Another large group of work is that which concerns the cutting of sweeps,
done at the bench with gouges, spokeshaves, and planes. They are cut with
an adze block (fig. 86), the spindle being set vertically, or canted slightly
if taper is required. The table carrying the sweep is moved around the curve,
round the centre of the top table if of moderate radius, or attached to a light
former of wood as in the figure for larger radii, the table being moved along



the runway past the front of the main frame. Sweeps 14 in. deep can be
cut. Pipe bends are treated similarly. Fig. 87 shows a half-pattern of small
radius carried directly on the table. The fly-cutter has nearly the same
sectional curve as the bend, and operates on both sides in succession. At
the same setting a regular curve can be combined with a straight length.
When core boxes for bends are being cut, the same method is adopted, the
cutter having a semicircular contour, and the half-box being carried round
a radius or along a straight line as required. But boxes for branch pipes

Fig. 87.—Mechanical Wood-worker operating on small Section as Pattern Bend of small Radius
Any section up to twenty-four inches can be operated upon, and any radius.   The whole segment of a circle
may be operated, and straight parts to any extent may be left at either or both ends of bend.
and those with recessed portions are cut with the spindle set horizontally
(fig. 88), with cutters of the sectional shapes required.
Miscellaneous Machines.—The circular and band saws, the various
machine cutters, and the bench edge tools have to be sharpened and kept
in good working order. Though often done without mechanical aids, these
become necessary in the large shops. On the regular setting of circular
and band saws their efficiency mainly depends. Circular-saw teeth can be
set in one machine and evenly sharpened in another. Band-saw teeth are
set and sharpened in one machine. The cutters for planing machines and
other kinds are ground while held in a fixture traversed past the face of an
emery wheel of cup shape. For the small machine cutters and for hand-


tools, grinding wheels and circular oil-stones are obtainable. Included in
the equipment of some shops is the conical oil-stone for sharpening the con-
cave bevels of paring gouges.

All the machines in a shop, the lathes excepted, should be in the charge
of a man or men who alone operate them and are responsible for their

Fig. 88. — Mechanical Wood- worker operating on Three-way Valve Core Box with numerous Internal Chambers
The core box shown was completed in forty-five minutes.  Approximate time by hand forty-five hours.

efficiency. This is both economical and safe, since circular saws and planing-
machines are fruitful of accidents to inexperienced hands. The lathes are
used by all the pattern-makers, who also grind their own tools.

The Shop and the Stores
The lay-out of the pattern-shop does not reveal those aspects of interest
which are associated with the machine-shop and with some of the later
foundries. The real attraction centres in the work under construction.
The lay-out is similar to that of the carpenters and joiners. Rows of benches
disposed across the shop accommodate two men each, working at opposite
sides. One long, wide bench with several vices is reserved for the larger
patterns when the work is of such a character as to require them. Machines
are arranged along one side or one end of the shop, in close proximity for
convenience of driving and of operation. The circular saw and the planing-
machine must have unobstructed spaces at front and rear for the movement
of boards. The most suitable drive is a gas-engine or an electric motor,
either of them driving a length of shaft from which the machines are driven.
If metal pattern-work is done, this engages a separate department, or is
relegated to another building. When large patterns are constructed, large
doors are required at one end of the shop. Whether the shop shall occupy
a ground floor or an upper story is a matter of no importance. A ridge roof
with north light is desirable, or, having a ceiling, side windows must be of
sufficient area. The shop should be heated with hot water, unless a regular
hot air and ventilating system is installed in the works, which may include
the pattern-shop.
The timber should be stored adjacent to the shop. It is stripped during
seasoning, but may be laid edgewise when ready for use. As timber is
expensive, economy can be practised by storing all odds and ends, which are
numerous in pattern-work, on racks at one end of the shop. The selection
of suitable fragments will often save the expense of cutting into a board.
Core prints are turned in quantities for stock by the apprentices. Some are
nailed on patterns, but a fair proportion are turned with studs of some
standard diameter to go on bosses for gear wheels and pulleys. Wooden
dowels may be stocked, but the metal kinds are more durable. Wooden
fillets, hollows, or angles are required for all patterns except the roughest,
tut those of leather are supplied to the trade. Pattern letters of various sizes
and shapes are made in lead, tin, and brass, but these are better bought.                          !
Rapping plates to suit all patterns are purchased.    All these are kept in the                          • {
shop stores.
Method of Working.—The organization of the machine-shop is not
represented in the pattern-shop.. Methods have been modified by the
introduction of the machines just now described. The result is that much
laborious hand-work formerly done on the bench is performed much more                          {
expeditiously on machines. All the hands are trained craftsmen, who have
served a lengthy apprenticeship, and who work under the direction of an
experienced foreman. And although the practice in the large shops is to
keep certain men or groups occupied with definite tasks, these are men with
a general training, who have drifted into specialization.
Men are paid by time in most shops. The variable character of the work
done, the fact that the greater portion of it is handicraft, that alterations are
sometimes seen to be desirable during its progress, and that one foreman is
easily able to keep the entire shop under observation, are causes that favour
payment by time rather than by the piece.
The method of constructing a pattern is settled by the foreman. When
uncertainty exists as to the selection of the best among alternative methods
of moulding, it is well to discuss the matter with the foreman of the foundry.



During the progress of the work he keeps it under observation, both with
a view to save labour and to detect error on the part of the workman, but
without obtrusive interference with the idiosyncrasies of the craftsman,
who often has his own peculiar ways of doing things. When the pattern is
complete the foreman measures it carefully before sending it to the foundry.
To deal with the many thousands of patterns that accumulate, some form
of registration is essential. A pattern register is kept by the foreman, in
which is entered the actual name of every piece and the order for which it
was made, but opposite the name are letters and numbers, and these are
stamped on the patterns. The letters are those of the alphabet, the numbers,
commencing with i, run up to a predetermined limit, 1000 or higher. These
are stamped on the main pattern, on every loose piece, and every core box
belonging to it. If any portions stray in the foundry or in the stores, the
letters and numbers indicate at a glance the pattern to which they belong.
Written orders are sent with each pattern into the foundry, with the date,
the order number, and the number of castings required from it. Error in
moulding, as it affects faced portions or prints and bosses, is sometimes guarded
against by the use of distinctive colours. Thus, while the patterns are pro-
tected with yellow shellac varnish, portions to be faced may be uniformly
painted black or red. Core prints may be painted one colour to distinguish
them from metal.
The Stores*—These occupy large areas, since patterns accumulate
rapidly, and many of them with their boxes are bulky. A storied building
is usual, the heavy work being in the basement, the lighter on floors, for which
tiers of shelves are provided, the widths and spacings of which have to be
in accordance with the general class of work done. Two general systems of"
storage are adopted. In standard work all the patterns of a set are placed
together, the light and heavy. As these are never altered for different orders,
they need not be checked over, but sent complete into the foundry. But
patterns that are not strictly standardized are subject to alteration from time
to time, and this renders measurement and checking for loose pieces and
core boxes necessary for all new orders. For these the practice is to put
all patterns of one class together, from which selections for casual orders
can be quickly made. The letters and numbers stamped on patterns and
their parts show for what previous orders they have been used, and with what
alterations. All the shelves are numbered, and the number of the shelf on
which a pattern is stored and the number of core boxes are entered in the
register. Metal patterns are kept in a separate place, many being hung on
the foundry walls.




FOUNDRY   WORK                     [t
Foundry Work

The Work of the Foundries: What it Embraces.—Since the major
portion of this work deals with the products of the iron foundries, these
must receive the principal attention in this article. And it must be remem-                             ]|
bered that the essential methods of the iron moulder are also those of the
steel, brass, and malleable cast-iron foundries. The details in which these
differ from one another are so important that each engages the services of
its own specially trained craftsmen, who would have much to unlearn, and
learn, if they should attempt to take service in one of the shops of another
group. But what all have in common are the fundamental facts: that
liquid metal is poured into a matrix of sand, usually prepared from a pattern;
that the moulds are all subject to the same laws that control liquid pressure
and the shrinkages of metal; that the various methods of making moulds,
with one or two slight exceptions, are employed in all foundries alike.
Metal Moulds.—While the castings poured into sand moulds include                             [
probably 90 per cent of the total quantity of castings made, inuch the larger
portion of these being of iron, there is another and a steadily growing
group, the moulds for which are of cast iron or steel. It embraces the
chilled castings, and the more recently introduced permanent moulds for
pipes, and the extensive practice of die-casting, employed for small, more
or less intricate articles made in the softer alloys, and required in very large
Subdivision of Tasks.—In the very extensive group of iron foundries,
there is much subdivision of tasks. This occurs in all the large shops, and
in many of those of small size. The product and the men are specialized.
The great subdivisions are: moulding in green sand, dry sand, loam; .core
making and machine moulding, each often classified under light and heavy.
In the machine department, further economies are effected. One man
makes bottoms only, another tops, while a third will core and close, ready
for a fourth to pour. A few years ago there were craftsmen to be found
in most shops who were competent to work in green sand or loam, in
light or heavy moulds, and at core making, while when occasional casts
in brass were wanted, an iron moulder would take on the task. These




made excellent chargemen and foremen. They only survive in the ranks
of the older men. Each department now employs its own sets of hands,
producing the same classes of castings the year through. Moulders, like
machinists, are specialists. Only in the general jobbing and repair shops
do exceptions occur.
Foundry Metal.—This includes cast iron in its numerous grades, the
steels, malleable cast iron, the immense groups of brasses and bronzes,
alloys of copper, aluminium and its alloys, and the varied die-casting alloys.
Many of these are now graded by analysis and by the scleroscope, instead of,
as of old, by the foreman estimating by the aspect of fractured surfaces,
supplemented by test bar results. For melting, the cupola furnace occurs in
many excellent designs, with its equipment of fan or blower, blast gauge,
platform, weighing machine, receiver or ladles. Brass-melting furnaces
are coke-fired, or oil- or electrically-heated, with provisions for utilizing
the waste heat. Steels are melted in converters of large and small capacities;
malleable cast iron in air furnaces.
Sands.—In a foundry equipped with modern appliances, the prepara-
tion of sands is done wholly with machinery. It takes charge of them at
every stage, drying, crushing, grinding, mixing, sifting, and conveying.
Suitable mixtures have to be graded for green, dry, core, and loam sands,
and again for light and heavy moulds. They differ also for steel and iron,
and facings for the moulds are varied. For this work, a complete mechanical
plant is often now installed.
The Treatment of Castings.—This, colloquially denoted by the
terms " fettling " and " dressing ", engages, in the big advanced foundries,
a large quantity of machinery and plant, doing work that was formerly all
performed by hand methods. It includes: machines for severing the runners,
with chisels or with saws; grinding wheels; pneumatic chisels for the
removal of fin marks and roughnesses; tumbling barrels for smoothing
castings by attrition; and, in the later plants, sand-blasting machines, now
made in many designs to deal with castings of all dimensions. In the more
complete plants, dust-exhausting systems of pipes with exhausters are
To deal adequately with all the aspects of foundry work outlined in the
preceding paragraphs is obviously not practicable. Neither does it seem to
be called for. Each single subject is now highly specialized. The foundry
craftsman is only directly concerned with and responsible for the preparation
of the moulds. The sands are prepared for his use, the metal is graded,
suitably melted, and brought to him, the patterns are prepared to be moulded
in a certain way, from which no essential departure can be made, and he
has no further concern with the castings if they are turned soundly out of
the moulds. Bearing these facts in mind, it is proposed to occupy the
major portion of this article with the subject of the preparation of moulds
required for the metals and alloys, leaving the collateral matters to be dealt
with in a summary fashion.
MOULDING  IN  GREEN  SAND                    113
Moulding in Green Sand                                        11
The term " green sand " does not denote any one of the specific mixtures                                ?J
used, but it signifies that the sand is moistened and rendered coherent with                                ||
water, so that it becomes sufficiently self-sustaining to retain the shape
imparted to it by the pattern, and to resist the pressure of molten metal.
It differs therefore from moulds made in dried sand and in loam, and from
cores, which are desiccated. By far the largest proportion of moulds, large
and small, is made in green sand. As there is no drying process, fuel and
time are saved. Green-sand work embraces three systems of working:
open sand, bedding-in, and turning over, or rolling over.
Moulding in Open Sand.—This is but a crude and very elementary
method, and one which is of extremely limited application, being almost
exclusively employed for making foundry appliances, loam and core plates,
back plates, moulding boxes, and sometimes balance weights for cranes.
It signifies that the mould is not covered with a cope, and the consequence
is that the upper surface of the casting so poured is left rough and uneven
as the metal solidifies. The necessary details may be stated briefly.
A Levelled Bed essential.—If the bottom of an open mould is not level,
the thickness of the casting will not be equal all over. The bed is levelled
by bedding two parallel straight-edges—"winding strips"—in the sand of the
floor, levelling them lengthways with a spirit-level, and, in relation to each
other, with a parallel straight-edge set across them, and a spirit-level. The
sand is flat-rammed a little higher than the top edges of the bedded-in strips,
and then strickled off level by them. On this bed the mould is made,
seldom from a pattern, but usually from a skeleton frame, or as often from
sectional pieces. No venting is required as in closed moulds, and no specific
sand mixtures, the moulds being made in the floor.
The Formation of Mould Outlines.—If these are produced from entire
patterns, as core grids generally are, the pattern is laid on the levelled bed,
and the sand rammed around and within it, and strickled off. In most cases
some portions have to be stopped-off to suit various core outlines. In
others, a grid larger than the pattern is required. Here the pattern is rammed
in one position, then removed to another adjacent position, and rammed
again. Generally, open moulds are constructed with sectional pattern parts.
The outlines are marked on the levelled bed by the moulder or pattern-
maker. External portions are rammed against short sweeped pieces, moved
around and rammed in successive lengths. Large central holes are rammed
against concave sweeps. Small holes are formed with cores, measured in,
and held down with weights. Straight sides are rammed against straight
strips. In all this work the depth of the mould exceeds that of the casting
thickness by from J- in. to f in., and flow-off gullies are cut at the height
corresponding with the thickness required. This is necessary, because it
VOL. I.                                                                                                                                8

is not possible during pouring an open mould to stop at the precise thickness.
The metal is not poured directly into the mould, but into a shallow basin at
one side.

Numerous adjuncts are located in these moulds: loam plates, core
plates, and solid grids with prods or prongs distributed over portions of
their surfaces. A pattern prod carrying say half a dozen prongs is pushed
by the moulder into adjacent positions without any particular regard to

Fig. i.—Shows Pattern Frame for Box with Loose Bars moved into successive Positions
A is a diagonal to keep frame square.   B, Provision for swivels,   c, Prints covering lugs for pins.
exact spacings, the prods therefore being cast with the plate. Wrought-
iron rods of various lengths and outlines are thrust into the mould to
be cast into grids. Long rods with eyes are suspended in the mould,
so that the metal runs round them and amalgamates. Nuts are cast in
when grids have to be retained in their places with screws. Looped
handles are cast in for the lifting of cores.
Large heavy moulding boxes are commonly cast in open sand. The
work is done in successive stages with the help of a limited number of parts.
The pattern box sides are framed entire, but three bars or stays suffice if

their outlines are uniform. These are moved along in rotation, one remaining
in the sand to afford support while other two are being rammed (fig. i).
Sectional patterns for fittings for boxes of various sizes and types are stocked,
and selected as required. These include box ends with prints for the iron
swivels, and core boxes for looped handles, and for the lugs in which the
pins are fitted.

Moulding by Bedding-in.—This embraces a very extensive volume
of work done, the essential characteristic of which is that it is moulded in
the foundry floor instead of in a box. The mould is covered and closed with
a cope, through which the metal is poured. As there is no bottom box with
locating pins, the cope has to be set with stakes driven into the sand of the
floor. Bedding-in is mostly adopted in the largest work. The reason is,
that the turning over of massive boxes with their contained sand would be
very inconvenient, and in some foundries impracticable. The cost of the

Fig. 2.—Strickling Facing Sand on a Mould Bed
A, Thickness of facing.
boxes also would bear too high a proportion to that of the castings, which
are seldom wanted in large numbers. And provided reasonable care is
exercised, castings can be made as satisfactorily by bedding-in as by turn-
ing over.
Variations in Details.—Methods of procedure are modified by the shapes
of patterns. If these have level lower faces and broad areas, a levelled bed
is strickled under the guidance of winding strips, as described in the making
of open moulds. The vents from the large areas must be driven down into a
cinder bed, to be brought away through large vent pipes extending from the
bed to the outside of the mould. Instead of using winding strips for levelling,
the horizontal edge of a sweeping board can be worked round a central bar,
a method which is adopted when central bosses and annular facings have
to be produced in the bottom. On the bed, first prepared, the pattern is
set and rammed. This may either be complete, or a skeleton outline, against
the outer faces of which the mould is rammed, leaving the interior to be
formed with cores.
When patterns have irregular outlines, and parts projecting into the
bottom, such as deep flanges, ribs, bosses, and lugs, each portion has to be
treated in detail if lumpy castings are to be prevented. If the pattern is
very diversified in outline, a level bed is of no value. If its main web is
flat, the bed is required. In each case the floor sand is prepared by digging
and flat-ramming, and over it a thickness of i in. or more of facing sand is

sieved. A fairly even quantity may be ensured by laying two strips on the
floor sand, say i in. thick, and strickling the facing sand level with these
(fig. 2). The pattern is bedded on this, and driven in with blows of the
mallet, applied with sufficient firmness to leave its outlines impressed on the
bed. Where the projecting portions are beaten down, the sand is rendered
harder than elsewhere, and the casting might become scabbed in conse-
quence. The pattern is removed and the hard sections of sand are loosened
with the trowel and the pattern bedded again, frequently more than once.
More facing sand is added where necessary, loose portions are tucked under
with the hands and the pegging rammer, and any necessary venting is done
with the pricker. In plain patterns having narrow sections or semicircular
outlines, the whole of the work may be done by tucking the sand under,
without removing the pattern. The pipe in. fig. 3 can be moulded by tucking

Formation of the Cope.—Any plain top box part of a size suitable to

Fig. 3. — Iron Pattern of Pipe with Flanges of Wood suitable for Bedding-in by tucking Sand under

cover the mould is selected. It is set in its position with four stakes, that
take a bearing against joggles on the box sides, or against its lugs. It is
then rammed on the pattern, removed, turned over, the mould finished
and the box replaced, guided by the stakes. The cope is then loaded with
weights before pouring, since there are no box pins to be cottered. As the
top is plain, the stays stopping short of the joint face, the sand in any deep
recessed portions of the pattern has to be carried with lifters hung from the
stays, or, in other cases, when the deep portions have large areas, grids are
suspended from the stays to carry the sand. When moulds are so long that
they cannot be covered with a single top, two boxes are laid side by side.
Long, flimsy patterns give trouble when bedded-in, because it is difficult to
prevent bending and winding. The progress of the work is therefore
checked constantly with straight-edges and winding strips.
A large volume of work is done in the floor when, instead of complete
patterns, frames or sectional elements only are used for the exterior portions.
This, though not strictly a process of bedding-in, is allied to it, since a bed
has to be prepared, and vented down to a cinder bed, and the mould is
covered with a plain top. Bedplates of rectangular and circular outlines are
often made in this way. When copes are not plain, but contain bosses and
facings, and there is no complete pattern with a top boarded over to ram

on, the cope is rammed first on a dummy sand face.   A level bed is
corresponding with the mould joint, the various pattern pieces are


set on

Fig. 4.—Half Fly-wheel Mould made without Pattern, using a Sweep Piece and Cores

this by measurement, the cope is rammed over them and removed, the sand
in the bed dug out, the mould
made, and finally covered
with the cope, which is
guided into its original posi-
tion with the stakes.

Though these large
moulds are made in green
sand, the surfaces are often
hardened slightly by the
process of " skin-drying".
A devil containing burning
charcoal or coke is suspended
in the mould, which drives
off a portion of the moisture.
But for dried moulds, a
different mixture of sands is
necessary, and these are con-
tained wholly in boxes.

Figs. 4 to 9 illustrate ex-
amples of work made in the
floor. Fig. 4 is a half fly-
wheel mould. The rim has
been formed with a sweeped
piece, and the arms with cores. The joint faces of the rim and the boss
are closed with pieces of loam cake. Fig. 5 is a portion of a fly-wheel.

Fig. 5.—Portion of Fly-wheel Mould
A, Half cores closed.   B, Half core open.

' ii

made with a sweeped piece for the rim, rammed against outer and inner
curves, and having the arms formed with half cores jointed in their middle
plane. The halves are closed at A, the lower half core is open at B. The
projections seen are the sides of the grids. Bosses only have to be bedded
in bottom and cope.

Fig. 6 is a casting moulded without any pattern portion except a sweep
that forms a circular print, shown at A. Half a dozen cores (fig. 7) are made
in the box (fig. 8), which, when laid on a level bed, produce the holes for

the lever bars, and the
spaces adjacent for light-
ening. A central boss
has to be set in the
bottom and cope.

Cores,   bosses,   and
facing pieces often have

Fig. 6.—Casting of Capstan Head to receive the Bars
A indicates a circular print and the joints of cores.

Fig. 7.—Core for Capstan Head

to be set in by measurement. In some cases a templet is useful. An
example is given in fig. 9, used for making print impressions in the
bottom of the mould for a boilermaker's levelling block. The holes are
first pitched and bored correctly in the templet, then the print, having
a shoulder to determine the depth, is thrust into each hole in succession.
The half holes round the edges of the templet are laid against the cores
already inserted.
Moulding by Turning Over.—This method requires at least two
box parts, a top and a bottom, within which the mould is wholly contained,
and a middle part is frequently included. It is, of course, the ideal method,
because both faces of the pattern are treated exactly alike, each being
subjected to direct ramming of the sand against it. This is therefore the


most universal method of moulding. Its value is most evident when patterns
have very intricate outlines, undercut portions, deep projections, loose pieces,
bosses, ribs, and so on. These can be evenly rammed directly, and parts
which are troublesome to deal with when bedding-in are more accessible.

In the usual practice, that is,
apart from the employment of joint
boards or of plates, the pattern is
embedded in a body of sand in the
top box, which is thrown out after
the joint face has been made. But
it is only shovelled in and made
sufficiently hard to ram on, so that
the loss of time is not of much im-
portance in the work of the general
shop. The following states the typical
sequence of this work. The top box
part is laid with its top face on the
floor, is filled, and roughly rammed
with sand up to its joint edges, and
the top side of the pattern is bedded
in this, until the joint edges of the
pattern coincide with the sand joint,
which, whether plane or irregular, is
shaped and sleeked with the trowel.
Parting sand is strewn over this, and
the bottom box part is placed over
the upstanding pattern, and cottered to the top box. After ramming and
venting, the two are turned over together, the bottom being brought to a
level bearing on the sand floor. The temporary sand is knocked out of
the top box, then replaced on the bottom, and rammed permanently. If
a middle part is used, this
being interposed between the
top and bottom, an addi-
tional joint is required.

The advantages of direct
ramming are secured in
other ways where boxes are
not turned over, but the

treatment properly belongs
to plate and machine-mould-

Fig. 8.—Core Box for Capstan Head

Fig. 9.—Templet for setting Core Prints

Although it is usual to joint patterns in the plane of the mould joints,
the practice is far from universal. The smaller the patterns are, the less
frequently are they jointed. Brass moulders seldom use jointed patterns
except in the larger sizes, but lay solid patterns in odd-sides. There is less
risk of the occurrence of lapping joints than when top and bottom portions


are located with dowells, and in long patterns like that in figs. 3 and 10,
if made in timber, warping in the transverse and longitudinal directions

Fig. io.~Pipe Pattern not Jointed for Turning Over.   Shows method of fitting flanges loosely in grooves

is less liable to occur than in a pattern that is made in two portions. But
as the top box must then be lifted off the pattern, it is better to leave the
flanges loose, because they have vertical faces.

Fig. ix.—"Jointed Bracket Pattern

Fig, 12.—Jointed Bracket Pattern


Three examples of brackets in which the jointing coincides with that of
their moulds are given in figs, ir to 13.   In the first the joint is along the

middle plane of the web, in the second
along its top face. If these were not parted
as shown, the lifting of the cope sand off
the vertical rib would fracture the sand, to
avoid which the rib alone is often left loose,
and the upper portion of the boss and of
the foot made fast. In fig. 13, the upper
boss and its bracket, shown dowelled, are
often made fast. Each of these patterns
might alternatively be moulded sideways,
that is as they lie on the paper.

Fig. 14, the radial arm of a drilling
machine, should properly be jointed in
the plane aa, and moulded by turning over, but bedding-in offers no
special difficulty. The prints for the column core are dotted at A, A; the
overhanging portions of the facing for the saddle must be loose, as at B, B.

Fig. 13.—Jointed Bracket Pattern


Fig. 15 is a fusee drum for a derrick crane, having the ratchet cast at one
end. The ratchet is made in a core, the print for which is outlined at A.
Fig. 16 is its mould, cored ready for pouring. Reference to some of its
details will follow presently.

j        J
 rv i



	i :


L.Jll.J     -
Fig. 14.—Radial Arm of Drilling Machine
aa shows jointing.    A, A, Prints.    B, B, Loose pieces.

Details of Green-sand Moulds.—Although some details given here
concern moulds made in dried sand and loam, others do not, or only in a
lesser degree, so that the present section is the most suitable for their con-

Provisions Made for the Support of Sand.—Owing to the fragile nature

pig> 15>—Section of Fusee Barrel for Derrick Crane
A, Print for ratchet core.
of green sand, abundant support is required. All boxes, except the smallest,
those say of from 12 in. to 15 in. across, are bridged with stays or bars (see
figs 16 and 36), spaced at from 6 in. to 8 in. apart. Those in the bottom
box are flat, since the sand there cannot fall out, being supported by the
floor on which the box rests. Those in the top are vertical, extending down


to within f in. of the joint face in a " plain top ", or to within f in. of the
pattern when the box is made for special work only. In the latter case the
bottom bars are shaped similarly, as in pipe and column work. The bars
retain the sand by their proximity, and the friction of their rough surfaces,
assisted by an application of clay wash before ramming. Middle parts
seldom have any bars, but narrow flanges are cast within top and bottom
edges to assist in sustaining the sand, and to carry rods laid on them and
disposed close to the pattern. Small flasks for brass moulding have internal
flanges, and some have their sides recessed to a very obtuse angle to prevent
risk of the sand falling out.

Fig. 16.—A Corcd-up Mould for a Fusee Barrel


Lifters, Rods, and Nails.—When deep portions of moulds or recessed
pockets of sand extend considerably below the lower edges of the bars, these
receive support with " lifters " or " S-hooks " (see fig. 16), suspended from
the tops of the bars, and going down into the sand rammed around them.
They are bent at both ends, one to rest on the bars, the other to assist in
holding the sand. They are wetted with clay water.
When portions of sand extend out horizontally, they would break down
by their own weight, or be washed away by the inflowing metal, unless
supported with rods in the larger sections, or cut nails (sprigs) in the smaller.
In each case the length must be sufficient, when enclosed in the sand, to
counterbalance the portion that overhangs. All weak portions of sand have
to be treated thus, so that a rather large amount of " sprigging " has to be
done in some moulds.
Venting.—With the exception of some small moulds, made with an open
self-venting sand, and most loam moulds, venting done with a rod or wire is
necessary. The vents are driven from the outside of the moulds close to
MOULDING  IN GREEN  SAND                     123
the surfaces of the patterns. The air and gas escape directly through the
top of the cope. That from the drag is brought out through large horizontal
channels driven between the bottom of the box and the sand floor on which
it rests. Vents that come into the joint faces are led into shallow gutters
cut in those faces, and surrounding the mould, to come out through the box
joints. The vents from large bedded-in moulds in the bottom are taken
down to a cinder bed, to be discharged through pipes outside. A similar
practice is adopted when very large masses of sand occur in closed moulds.
Bodies of clinker or coke are introduced. Into these the vents are led. The
gases are discharged quietly through these without risk of partial explosion
and shock to the mould. Vents from large moulds are generally ignited with
a hot skimmer. The amount of venting in moulds varies. Close, loamy
sand, and portions that are rammed hard require the maximum amount.
Insufficient venting is productive of blow-holes, and of scabbing.
Delivery and Mending-up.—Some rapping is necessary to loosen patterns
for delivery. Its severity has to be greater with increase in depth, in the
proportion of vertical faces, and in area. The point of the bar is inserted
in the pattern, or in a hole in a rapping plate, and the bar is struck with a
hammer laterally from all sides. During the early stages of delivery, the
top of the pattern is rapped slightly with a wooden mallet to assist its detach-
ment from the sand. The edges of the mould all round adjacent to the
pattern are swabbed with water to lessen risk of the sand being torn up.
But some fractures, except in the case of semicircular and allied outlines
that are most favourable to delivery, almost invariably occur. These have
to be repaired by a process of mending-up.
When a mould is very badly damaged, it is better to put the pattern back,
and re-ram the parts, but this is not practicable when it is of unwieldly
dimensions. Portions of the pattern are sometimes detached for making
broken sections good; often supplementary pieces are prepared to avoid
such removal. In most instances the moulder mends up with any odds and
ends suitable—straight strips, sweeps—or he bends sheet lead to outlines.
Nails may be thrust into the broken sections, and swabbed to assist the
coherence of the sand, and a stronger sand, skin-dried, will often be useful.
In mending-up, it is not easy to preserve correct dimensions and outlines.
Often this is of little or no consequence, but it is so when work has to be set
in fixtures for machining. This is one of the reasons why machine-moulded
castings should then have preference.
Pouring Arrangements.—Many moulds free from faults in the making
have produced damaged or waster castings because of improper methods
of supplying them with metal. Molten iron, steel, and brass are heavy,
and sand is fragile. The ideal method is to bring the metal in, in a position
which varies in different classes of moulds, and to let it distribute itself, and
rise quietly, instead of rushing and beating against weak sections of sand,
against cores, or parts that have to be machined. Moulds of moderate
depth are generally poured from the top (figs. 16 and 36), the metal being
brought into the thickest portion of the casting, such as a central boss, the



ingate and runner being one. Deep moulds are treated differently. The
metal is either led in somewhere down the side, to lessen the height of its
fall, or at the bottom, to rise quietly without any cutting action or splash.

The pouring basins for small moulds are of simple cup-shapes, moulded
in iron rings set on the top box, through which the metal, skimmed, passes
directly into the mould. But when large moulds are poured from a ladle
slung in the crane, some little time is spent in tipping and adjusting the spout
of the ladle. The first driblets, therefore, are not permitted to fall directly
into the mould, but into a deeper depression of the basin to one side of the
runner. The metal then being poured into this in a full volume overflows

into the runner, slag being kept back by
the skimmer. When a mould is filled
through several adjacent ingates, they are
supplied from a common basin. Long
moulds are poured from opposite ends,
to avoid the chilling effects of a too-
prolonged contact with the cold sand.

No rule can be stated for the cross-
sectional areas of runners and ingates.
These have to vary with the degree of
fluidity of metals and alloys. The only
rule is to have them large enough to fill
the mould before any chilling effect can
occur. Also, the thinner the metal the
more numerous must the ingates be made,
until for the thinnest a spray of runners
is used, fed from a common ingate. In
castings of medium thickness, a runner of
oblong section, much longer than its thick-
ness, is used. Oblong runners are also
better than large round ones for the fettlers,
because they are more easily severed, and
are less likely to cause a depression in the casting if broken off.

Skimming chambers are provided when metal has to be scrupulously
clean. Ordinarily metal is cleansed by " dead-melting ", and by baying
back the scoriae in the ladle with the skimmer. When small articles have
to be machined all over, centrifugal action is enlisted to send the heavier
metal to the circumference of the chamber (fig. 17), whence it is directed
into the mould, leaving the lighter impurities about the centre, to remain
there or to float up into a riser.

Risers and flow-off gates resemble the plain cup-shaped pouring basins,
and their functions are (a) the relief of excess of pressure and strain on the
top part of the mould, and (b) the discharging of an excess volume of metal
with which dirt and air bubbles might have become entangled. Pouring
is therefore continued for a short period after the mould has been filled.
The relief of strain is important in the case of moulds having large areas.

Fig. 17.—Skimming Chamber
A, Chamber.   B, Ingate.   c, Riser.
MOULDING IN  DRY SAND                        125

The liquid pressure will often cause the top part of a casting to " gather "
J in. or more in thickness. Risers relieve this strain by providing openings
into which the metal, otherwise confined, rises quietly. The risers are
closed with a ball of sand or clay during the pouring, and are floated off
by the filling of the mould.

The object of " feeding " or " pumping " is to supply additional hot
metal to compensate for the shrinkage of heavy masses. It is done through
a pouring basin, or a specially made opening and cup. Molten metal is
poured in, and a J-in. or f-in. rod inserted and pumped up and down until
the metal becomes too viscous to permit of further movement.


Moulding in Dry Sand

Moulding in dry sand is reserved for some massive castings that are
required perfectly sound, and free from minute specks and blow-holes.
Its principal applications are to steam and hydraulic cylinders. Only
strong mixtures of sand can be dried. This excludes all the green sands,
which, however, are frequently baked on the surface—" skin-dried ". The
porosity of the sand in a dry-sand mould when dried largely takes the place
of the venting with the wire done in a green-sand mould. The presence of
moisture even in small quantity in a mould imperfectly dried is therefore
a source of risk.

Practically all dried-sand moulds are enclosed wholly in boxes, and turned
over. They are put bodily into the stove to be dried. Since the sand is
very fragile after drying, all moulds are t: finned " in the joint faces previously
to or immediately following the delivery of the pattern, that is, they are
pressed down hard with the trowel for a distance of an inch or two back
from the mould boundaries, so that when dried and closed they will not
fracture. A slight fin is formed, but this is of no importance. Moulds
made in dry sand will bear harder ramming and more swabbing than those
of green sand. They are coated with wet blacking, while those of green
sand are dusted with plumbago powder.



Moulding in Loam

Moulding in loam is essentially an application of the form or profiling
principle on a large scale to the making of moulds.   The loam mixture

used is a strong sand, mixed
with horse manure. The mix-
ture, having been rendered plas-
tic with water, and thoroughly
mixed in a mill resembling a
mortar mill, is swept in this con-
dition with the bevelled edge of
a board, which is attached to a
vertical bar and rotated. The
profile of the mould in vertical
section therefore corresponds
with the profiled edge of the
board, and its diameter is set by
the radius of the board, measured
from the centre of the bar. The
mould has to be dried after
sweeping. Large central cores
are swept like the moulds, but
small cores are rammed in boxes
and dried.

Methods of Affording Support
to Loam. — Since the treatment
of a plastic material, when swept,
differs entirely from that of sand
rammed within flasks, suitable
methods of supporting it have
to be provided. All the load of
a mould (fig. 18) is carried on
massive plates, or rings of cast
iron. These are from 2| in. to
3 in. thick, studded with prods
all over the face that receives the
loam, and provided with three or
four lugs to receive slings for the
purpose of lifting the moulds,

.•»                       r           ,     t

or, in the case or central cores,
with long rods cast in, with eyes.

The vertical walls of moulds are swept against tiers of common bricks,
built up in a somewhat rough fashion of whole bricks and broken frag-

. 18. — A Loam Mould with Central Core in situ, Pouring
Basin, ingates, and Flow-off Gate


ments, but always breaking joint.   The bricks, dipped in clay wash, are
embedded in loam, and finely broken cinders  or  coke are inserted at
intervals in the larger spaces to assist in carry-
ing off the vents.    At about every third  or
fourth course, a layer of headers is laid in to
serve as binders.    When a mould is very deep,
an iron ring, inserted about half-way up, will
lessen risk of distortion of the brick walls.

The daubing on of the loam by hand, and
its sweeping with the board proceeds with the
building up of the bricks. From i in. to i£ in.
of space is left between the faces of the bricks
and the edge of the board. Coarse loam is
used for the greater portion of the thickness,
a finely ground mixture for the facing. On the
completion of the work, the mould is dried,
and blackened with wet blacking, which is
afterwards dried. Venting is not required in
the same degree as in green-sand moulds, be-
cause the loam, when dried is, like dry-sand
moulds, largely self-venting, and the precaution
is taken to occupy all roomy spaces between
bricks with fine cinders. But where large
masses of loam occur, which often happens in
non-symmetrical castings when pattern parts
have to be set in by measurement, the vent
wire is used freely. After these loose parts are
put in position, loam is daubed against them,
still supported against brickwork, and they have
to be left in situ during the drying of the
mould in the stove (figs. 19 and 20). There
is risk, unless care is exercised, of these loose
pieces becoming shifted during the sweeping,
and of their warping in the stove. They must not be varnished, but oiled.

Fig. 19.—Pattern Work for Steam
Chest to be embedded against a swept-
up Loam Mould

0.—Pattern Work for Cylinder Foot to be embedded against a swept-up Loam Mould


The loam that lies in contact with them delivers badly, and has to be
made good by mending-up.

Jointing.—Since moulding boxes are not available, jointing can only be
done in the actual moulds. The positions of joints are determined by the
shapes of moulds. All flanges involve joints and frequently extraneous
fittings. The bricks and loam above a joint must be carried with a ring
(fig. 18). The cope of a mould is carried on a plate having holes for the
ingates, and is turned over, the bricks being retained with rods and plates,
though many plain tops only require loam swept directly on their prods.
All joints except that of the cope are plane faces, and the cope may be the
same when it has, no boss or other part that requires exact centring. In
this case the joint is provided with a check (fig. 18) that renders it self-
centring when lowered into place. The
difference between this and other joints is
that these can be seen and set while the
mould is open. This cannot be done with
the cope which closes the mould.

Pouring and Shrinkages*—Loam moulds

Fig. 21.—Skeleton Pattern of Pipe Bend

Fig. 22.—Skeleton Pattern of S-pipe with bpaces filled
with Sand

are poured from the top, usually through a circle of ingates in an annular
basin (fig. 18). Moulds must not be closed until shortly before pouring
is done, since they absorb moisture. In deep moulds, the pressure is
so great that the bricks alone would be liable to yield, and they are
therefore rammed in the foundry pit, enclosed with sand walls, or with
iron rings. The shrinkages in large moulds would cause fracture of the
cooling castings if measures were not taken to enable the mould to yield
before them. A layer of loam bricks is used under a top flange, which
become crushed under the pressure. Often the labourers break away
some of the common bricks under a shrinking flange. Large loam cores
which would hinder diametral shrinkage have a perpendicular insertion
of loam bricks which yield before the shrinking cylinder. The interior
of a large core is filled with cinders to receive and carry off the gases.
The gases, from the exterior mould are brought out at the top and sides,
the latter being formed with large vent channels arranged in a circle outside
the mould (fig. 18), made with iron rods rammed in the encircling sand, and
Non-symmetrical Work.—This relates to loam moulds taken from skeleton







fi   IS u

£ o .

i °1

'£  -si

M     c"u

patterns. The pattern is broadly similar to the casting, being formed of
strips having the same diameter as that of the casting. The spaces between
the strips are filled with sand to give a continuous surface, and the core is
built within and dried, and the outer mould constructed. This is dried,


and removed in sections, the pattern is unscrewed and taken away, leaving
the core to be removed. An advantage of this method over the making of
a separate pattern and core box is that the correct thicknesses are ensured.
But the real reason for the adoption of the method is economy of timber and
pattern-maker's time. It is reserved therefore for the larger castings.

Loam Patterns.—These, swept in loam, are used instead of those made
of wood, to be rammed in moulds of green or dry sand. This is a rather
large and important section of foundry work, the object being, as in loam
moulding, to save the prohibitive cost of complete patterns of wood. It
includes symmetrical work, revolved against the profiled edge of a board
fixed on the core trestles, and non-symmetrical articles, formed as half pat-
terns with strickles, the longitudinal movements of which are controlled by
guide irons, or by the edges of contour plates on which the pattern halves

Fig. 27.—Pipe made in Loam
A, Guide iron.    B, Core grids,    c, Core.    D, Strickle.    E, Mould with core and its chaplets.
are swept. The longitudinal shapes are determined by the character of the
castings required. They may have regular or irregular curves, or curves
combined with straight portions. Instead of using loam patterns, it is often
cheaper to make a rough skeleton pattern of wood, with outline ribs, fill
the spaces with sand, and ram it in the mould. Fig. 21 shows a skeleton
pattern for a pipe bend as sent from the pattern-shop, and fig. 22 one
for an S-pipe, having the spaces filled with sand.
Figs. 23 to 26 illustrate the making of a pattern, and core for a loam
bend. A is the guide iron, set with weights, B is a slender body of loam
which forms the vent channel of the core, E, a part of which, D, is seen
roughly daubed on the grid c in fig. 24, with its vents, and which is com-
pleted in fig. 25. In fig. 26 the pattern "thickness" F, corresponding
with the thickness of metal in the casting, has been laid on, and the
standard iron pattern socket G and spigot H set, completing the half
pattern. After the pattern has been moulded, the thickness is stripped
off, leaving the core ready, when blackened, for insertion. In fig. 27 the
core strickle, controlled by the guide iron, is seen bridging the core, and
the mould is shown to the right, with the core inserted.


The foregoing descriptions relative to the making of moulds apply sub-
stantially to the preparation of cores.
That is, these may be (a) rammed in
green or preferably in dry sand, (b)
swept in loam with revolving or with
fixed boards, or (c) made with strickles.
Generally, the same provisions have to
be made for cores as for patterns, in
the shape of taper, in the employment
of loose pieces and prints for inserted
cores, and for shrinkages. Taking suit-
able precautions, there is no casting so
intricate that it cannot be produced
with the help of cores. As the support
of a moulding box is not available, a
large amount of detail is associated with the supporting elements around
which cores are rammed. These are round rods and wires in the smallest,

Fig. 28.—A Core being swept against the edge
of a Board on Trestles

Fig. 29.—Core swept on Bar for Fusee Barrel, fig. 16.   The section of the casting is drawn on the board
for the information of the moulder

and grids of multifarious forms in those of large dimensions.   For loam
cores made by rotation against the edge of a board, stiff cylindrical bars


are revolved on trestles (figs. 28 and 29), the latter showing the section
of the core inserted in fig. 16, and the casting section on the board.   These

r *

Fig. 30.—A Grid with Rods and Lifting Eyes cast in.   The relation of the core rammed on it is
indicated by dotted outlines

carry the loam and hay bands, and are perforated for the discharge of the
gases generated during pouring. In the absence of a containing flask, pro-
vision, in the form of rods with
eyes, and of extensions of grids,
has to be made for lifting all
except the lightest cores.

Cores rammed in boxes are
always liable to increase slightly
in dimensions, and to " gather "
as the saying is. The box sides
yield before the

Fig. 31.—Grid with Vertical Rods cast in to
afford support to a deep Core having Vertical

Fig. 32.—Core Box for Ratchet End in fig. 16
A, Ratchet.   B, Shrouding,   c, Line of joint in core.

ramming, and when removed the core swells a little. This is the reason
why so many of the larger cores have to be " nibbed " by the moulder to
reduce their dimensions. A careful pattern-maker will counter this by


making the box sides as rigid as possible, and by slightly reducing the interior
ciimensions. These precautions should always be made in standardized
^^vork, and provision made for the
taper of core prints in the boxes,
t:o avoid rubbing the taper on the
oores. Another reason why cores


Fig. 33-—The Ratchet Core in fig. 16

Fig. 34.—End Core without Ratchet in fig. 16
A, Line of joint in core.

should be made slightly below size is that, when dried, they are so hard

a.nd rigid that they retard the

shrinkage of the casting, so that

the  interior  either comes out

too   large,   or  in   some   cases

fracture occurs unless the core

is  loosened while the casting

is cooling.

Details of Core Formation.
—Generally this work is done
by the core makers, a class of
men apart from the moulders.
But this is merely a matter of
economy, a useful division of
tasks, since moulders can pre-
pare their own cores, and do so
frequently in the small shops.
Referring first to those cores
•which are rammed in boxes,
the work is substantially that
of dried sand moulds, with the
difference before noted, the
employment of an interior sup-
porting Skeleton, the "grid",                    Fig.35__ Core Box for Bevel Wheel

in  place of an exterior flask.

The first thing, therefore, which has to be decided is the form and dimen-
sions of the grid.    This both carries the load of sand, and affords the





Fig. 36.—Cored Mould for Bevel Wheel

means of lifting it into the stove and mould (figs. 30, 31, 35, and 36).
When practicable, the core is rammed in the same position that it has

Fig. 37.—Core Box, with Vent Rods

Fig. 39.—Shows Vent Rods in Core

Fig. 38.—Core with Stiffening Rods

Fig. 40.—Vent Strings in Core

to occupy in the mould, turning it over being generally avoided. Eyes
therefore come in the upper part of the core. The outlines of the grids
must follow approximately those of the core, so that a suitable grid has to


be made for every core.   These are cast in open sand from patterns kept
in stock, the moulds  being stopped off to any outlines and dimensions
required.    A large proportion of core grids can only be removed from the
interior of their castings by breaking them up and extracting them in frag-
............-.• «•*••-.,           ments, for which reason they are not made of

" ;-:.v'.           sections stouter than are necessary to sustain

••*.•.;•';;           the load of sand.    And generally in the deeper

': y.;'           cores, the cast grid occupies the bottom only,

Fig. 41.—Cores inserted in Drop Print
Impressions, bottom and top parts

Fig. 42.—Core inserted into Drop Print Im-
pression and moved along into a boss

support for overhanging masses being afforded by wrought-iron rods of
from J in. to f in. diameter cast into the grids (fig. 31). For small weak
sections, nails are embedded in the sand as in the similar situations in
moulds. Grids for cores therefore assume an infinite variety of forms.

Having the core box set on a level surface of
iron, wood,  or  sand,   and   the  grid   prepared, a    .•::.'';

Fig. 43.—Core Box, which includes two cores in drop prints

Fig. 44.—Setting a Core diagon-
ally with bottom print only

stratum of core sand is sieved over the bottom, to a depth, say, of about
i in., and the grid, well swabbed with clay wash, is bedded on it. More
sand is sieved or shovelled over the grid, and rammed over the grid and
against the sides, using the pegging rammer. Then, in all cores except
those which are shallow, a portion of the sand is scooped away from the
centre and heaped against the sides, and rammed with additional supplies
*3<S                                      FOUNDRY  WORK

until the top of the box is reached. The vent wire is now used freely, being
driven from the central open space to the box sides. The interior is next
filled with broken cinders or clinkers just lightly consolidated with the
rammer, a piece of tube inserted to receive and convey away the vents, and
the core is completed with sand rammed over the cinders to the top of the
box. The edges are swabbed with water, the box sides detached and removed,
leaving tne core standing ready to be put into the drying stove.

g, 45.—"Jetting a Core diagonally with
bottom, and top prints

Fig. 46.—Pattern with a long top print
for fig. 45

Referring to fig. 15, it will be noticed that the ratchet cast on one end is
shrouded or capped, which involves making a joint in the core. The box
is shown in fig. 32, the ring core being in halves for insertion in top and
bottom moulds, the core for the ratchet in fig. 33, and the remainder in fig. 34.
Fig. 35 shows the core box for a bevel wheel, with the core completed,
the grid, and central mass of cinders being indicated, and the strickle that
produces the curve corresponding with the edges of the vertical arms. Fig.
36 illustrates the mould, cored and closed for pouring.
Cores that are curved and thin, like those for the passages of cylinders,
have to be stiffened with rods, and vented with channels. Fig. 37 shows
a core box, ready for ramming, with vent rods inserted; fig. 38 a core with
stiffening rods; fig. 39 shows vent rods in a core previous to its removal from
the box; and fig. 40 the filing of grooves where the rods cross, for the
insertion of core strings, the portion filed being filled after, and the string
Fig, 41 illustrates the fitting of cores in drop print impressions; fig 42,
the thrusting of a core along into a boss, the space behind to be filled with
sand; fig. 43, the inclusion of two cores made in one box, with drop prints
common to both; figs. 44 to 46, two methods of setting round cores diagonally.
f,                                                 MOULDING SANDS                             137


1                                                CHAPTER V

Moulding Sands

The vast majority of moulds is made in sand mixtures.   The methods

.                   that lie outside of these are of a special character, as chill casting, casting in

permanent moulds, and die-casting, which, though of growing importance,
bear but a small proportion to the large volume of work made in sand. This
material is of pre-eminent utility because it is easily rammed or moulded
into any outline, it is so highly refractory that it is not fused by the
temperature of molten metal, it is adhesive enough to retain the shapes
imparted to it, is porous enough to permit of the escape of gases generated
in the mould by the molten metal, and, being quarried in many districts,
its cost is low.

Sand is never used in the crude raw state in which it arrives from the
quarries. It is wet, lumpy, non-homogeneous, and has to be subjected to
preliminary treatment in machines. And few sands are employed alone

)                   without admixture, though some are used thus because of their self-venting

properties.   The judicious mixing of sands to secure the best results for

if                       different classes of moulds is one of the tasks of the foreman, who has

generally to work with those kinds that are obtained locally.

i                        Facings.—The essential mixture is the facing sand.   This is prepared

to line the mould for a thickness of 2 in. to 3 in. next the pattern.   Elsewhere

:                   the flask is occupied with the " black " or <c floor " sand, which occupies the

foundry floor to a depth of about a couple of feet, and which consists of the

•                     accumulations of years from former moulds.    It has lost its original properties
I                   by repeated bakings, but when riddled and moistened with water it is used

for box-filling, serving as a backing to the facing sands. Broadly, these are
grouped as being " weak " or " strong ". The difference is that the first
contains a smaller proportion of heavy clayey material than the second,
also less coal dust. The function of the latter material is to prevent the
occurrence of " sand-burning ". While the infusible silica is the basis of
sands, a proportion of alumina is essential to provide the bond of coherence.
Oxide of iron is also present, and both these substances are fusible at pouring
temperatures. The coal dust lessens risk of resulting roughening of the
" skin " of the casting, by forming a film of one of the oxides of carbon
between the sand and the casting, a result which is assisted by the plumbago
facings dusted or brushed on the moulds. It follows that the larger the

•                     proportion of clay present in strong sands, the larger must be the quantity
J                  of coal dust.    The amount will range trom one of coal to six or eight of sand

in the strong sands to one in fifteen in the weaker mixtures.   Large moulds

in which the metal remains hot for a long time require more coal dust than

|                  small moulds that cool quickly.    The determination of the strength of a

,*                  mixture for a given mould is one of much importance.   Different grades are

•                    desirable for different parts of the same mould.   Areas subject to great





I     »'

liquid pressure, as large copes and the bottoms of deep moulds, should be
rammed with stronger mixtures than the sides. But venting must be more
thorough, or the casting will be " scabbed ".
Green, Dry, and Core Sands.—The feature which these have in common
is that they are consolidated with the rammer while in a moistened condition.
They are never wet, but sufficiently damped to retain a shape imposed when
squeezed in the hand. The retention of the form produced during ramming
depends partly on the coherence of the sand, but largely on the means by
which it is sustained in flasks, and on grids. Green sands cannot be dried,
except slightly on the surface, without losing their coherence. Sand, to be
dried, must be of a strong clayey character, and be mixed with horse manure,
which, by its carbonization during drying, counteracts the close texture of
the mould, favouring venting. But the vent wire must be used freely too.
Coal dust is also used. Core sand is mixed with clay wash, peasemeal, or
beer grounds; and generally, dry sand mixtures are suitable for cores.
Loam Mixtures.—These are made with strong sands, vented with horse
manure, with which generally a large proportion of old loam is mixed, the
whole being ground in a mill with water, and swept thus while in a pasty,
plastic condition, to be dried subsequently. It is used in coarse and fine
grades, the first for embedding the bricks in, and for the rough coats, the
second, finely sieved, for the final coats.
Chemical and Mechanical Analysis.—During recent years, attempts have
been made to grade moulding sands by chemical analysis, supplemented
with microscopical examination of the grains. These are helpful when new
sands are concerned, but their value is discounted when, as is usually the
case, large proportions of old sands are mixed with new. It is important
that the percentages of silica and of alumina should be known, and also the
quantities of iron oxide, lime, magnesia, and alkalies, which tend to lower
the fusing point of a sand, and flux it. Silica, the refractory element, must
be present in more than 80 per cent, alumina in from 7 to 10 per cent, the
proportions varying for weak and strong mixtures. But it is held that the
texture of a sand when passed through sieves of different meshes is of more
importance when deciding its suitability for a certain class of work than
chemical analysis, and that mechanical testing affords an approximate index
of the cohesive character of a sand. Weak sands have fine grains, and least
alumina. The strong sands possess coarse grains, and a large proportion
oi alumina. Castings with smooth skins can be obtained with the use of
coarsely grained sands. Fine grains are suitable for dry mixtures, cores,
and loam, with a large proportion of alumina.

Castings  made in  Metallic Moulds

Castings made in metallic moulds are embraced in three groups: (a) chill
casting, (£) die-casting, (c) casting in permanent moulds. These have nothing
in common beyond the fact that cast iron forms the whole or a portion of
the moulds. The conditions which control the pouring of liquid metal into
moulds of porous sand and those of iron are so different that the foundries
using metallic moulds are entirely separated from the sand foundries.

Chill Casting.—The fact is familiar that the effect of pouring liquid
metal in contact with a cold metallic surface is to harden—"chill"—the por-
tion that comes in immediate proximity with it. This is utilized in portions

•'•$ yjiiJLii "'•"•'•'•••


Fig. 47.—Chill Mould for Trolley Wheel
of numerous castings that are subjected to severe wear, as the treads of trolly
wheels of all kinds (figs. 47 and 48), in the rolls (fig. 49) for the iron and steel
works, for plough points, for mining stamps, stone breakers, balls and rollers
used in crushing and grinding mills, the bores of some wheel boxes, &c.
In all these cases the mould is of a composite character, being composed of
metal over the areas that have to be chilled, and of sand elsewhere.
Composition of the Metal to Chill.—The grey iron used for ordinary cast-
ings, in which the carbon is nearly all in the graphitic condition, will not chill
beyond a surface hardness of the thickness of stout paper. This is of no
value for service. An average thickness is generally required of from -J in.
to | in., extended in massive articles to i in. To produce this, it is necessary
to select a highly mottled iron, in other words, one in which a considerable
proportion of the carbon is in the combined condition, and the total carbon
Content high. And as silicon tends to throw out carbon in solution into the
graphitic state, the proportion of this element must be kept low. Sulphur
and phosphorus should be higher than for grey iron castings, since they
intensify the chilling effect. Manganese, below one per cent, is beneficial.


It is desirable to take test bars when making mixtures from new brands of
pig or selected scrap. Thorough melting is essential, and more coke will
have to be used than for the more fluid grey irons. The precautions to be
observed in venting, gating, and pouring for sand moulds are required here.
The Design of Moulds to Chill.—Only that portion of the mould which
corresponds with the area to be chilled is of cast iron, the remainder being

Z3£s&±yKtfr—i h\ i""r ' k'vt '   "i\ V«" V v'411

Fig. 48.—Chill Mould for Roller

Fig. 49.—Chill Mould for Roll

rammed in green or in dry sand. Success mainly depends on the mass of
metal in the chill. It must be large, in order to enable it to carry off the
heat from the casting poured, with sufficient rapidity to produce the necessary-
depth of chill. If this action were delayed too long, what would happen is,
that the cementite would have time to break up into iron and graphite, thus:
Fe8C = jFe + C. Cementite or iron carbide, Fe3C, is unstable when
cooled slowly. The walls of a chill therefore range from 4 in. to 8 in. in
thickness, depending on its diameter. The risk attendant on thick walls is
that of fracture, since the inner zones, expanding most, are tied by the outer.
and so tend to burst them. The precaution, therefore, is often taken of
bonding chills with a wrought-iron ring, shrunk on.
Shrinkages.—When a chill mould is poured, two shrinkages occur, that
of the casting inwards, and that of the mould in the contrary direction, so
that a space is quickly left between the two of J in. or more. Attempts have
been made to control and minimize this result, but the practice of dead-
melting the metal is usually adopted, that is, allowing it to cool slightly                            *
before pouring. Metal thus treated will lie better to the chill than that ]
which is in ebullition.                                                                                                            'l
An effect of the large amount of shrinkage that is consequent on chilling                           I
is that the portion cast in sand is weakened if not suitably proportioned. A
wheel rim having light arms must almost certainly snap in cooling. Hence
these are either made with massive arms, curved lengthwise, or the centres
are solid, having a " dished " or corrugated section. Chills do not have a
very long life. Though they may not fracture, the surfaces against which
the metal makes contact become roughened by the formation of minute
cracks, the result of repeated expansions and shrinkages. The metal too
deteriorates, approaching the condition of " burnt iron ". A new chill
must be cast, and finished by boring. Plumbago is used for facing at the
time of casting.
Die-casting.—This is a development, less than a dozen years old, of
the linotype castings. Originating with the white metal alloys, those having
a basis of lead, tin, or zinc, it now includes those with an aluminium base,
and efforts are being made to deal with those of copper. The phenomenal
demand for, and the immense supply of these castings is in response to the
call for those smaller mechanisms of universal use. These include type-
writers, telephones, gas meters, electrical instruments, speedometers, as well
as parts of engineers' mechanisms, lubricators, oil cups, bushes, small gear
wheels, &c.
Die-castings are made in metal moulds of steel, the liquid alloy being
subject to a pressure of 100 Ib. per square inch or more, which is maintained
until it has set. The result is that the castings do not require machining,
being correct to size within a thousandth of an inch, so that they will fit other
parts tightly or with sliding allowances and external and internal screw
threads will match perfectly. The teeth of gear wheels will mesh. Letters
and figures will come out sharply as though engraved. If finely threaded
screws or hard contact pieces are required, these can be cast accurately into
the softer alloys. Die-casting in steel moulds is used for-many small intricate
castings which can neither be made economically in sand, nor drop-forged,
and many for which the cost of machining would be prohibitive. Though
these dies are always expensive, their cost increasing with complexity and
the limits of accuracy insisted on, the outlay is relative. A rather com-
plicated die may cost from £50 upwards, but it will endure 50,000 casts of
a white-metal alloy, and a slightly smaller number for an aluminium-base
alloy. And the advantages and economies just stated are secured by its use,
at the cost of a fraction of a penny per casting.
The Die Moulds.—These are made in mild steel for the white metal
mixtures, but in one of the alloy steels for those having an aluminium base.
No portion of the mould is made in sand. Cores are of steel, and they have
to be drawn endwise with a lever from the casting. Sliding undercut parts
are similarly treated. Jointing is done when necessary for the removal of
the castings. Vent channels are cut. Ingates are severed while the casting
is in the mould. Means are provided for the mechanical ejectment of the
castings. In the more complicated moulds, where several operating levers
are involved, fool-proof methods are included to prevent cores and other
sliding pieces from being moved out of their proper sequence. Dies are
cleaned after casting with compressed air directed through a hose.
Furnaces.—These are essentially troughs of cast iron in which the alloy
is kept molten with a gas flame. The mould is usually carried above the
furnace, often on a tilting table. The pressure is put on with a piston in
a cylinder immersed in the metal. But many patents have been taken for
other methods, with the object of avoiding the blow-holes which are a frequent
cause of wasters. Some employ air pressure, others, centrifugal force, with
a vacuum, the idea being that blow-holes are due to the entanglement of air,
which is doubtful. The case is not analogous to that of green-sand moulding.
The cause would appear to be the chilling of the metal against the walls of
the mould, forming an unyielding shell before the interior has solidified.
The remedy is, to have an ingate large enough to fill the mould rapidly, to
bring the metal in where the sections are heaviest, and to inject under adequate
The Castings and their Alloys.—In the selection of metals to form alloys
for die-casting, shrinkage is the predominant factor. For, although casting
is done under pressure which is not released until solidification has set in,
some shrinkage must occur. Allowance must be made for this in making
the dies, or means provided to counteract it. Further, the strength of an
alloy to resist elongation by reason of the shrinkage stresses set up during
cooling has to be known with some approximation to correctness, because
otherwise, by using an unsuitable alloy, fracture may occur in the mould.
The case is different from that of ordinary moulds. The cores, being of
steel instead of sand, will not yield, sc the metal must have strength to
elongate, or it will rupture. And this varies with the1 proportions of the
elements, and with the temperature. This therefore is a matter for ex-
periment. Antimony is used to lessen the amount of shrinkage of alloys.
Only a small quantity, from i per cent to z per cent, is required in the zinc-
base alloys, but in the lead-base group it may be alloyed up to 25 per cent.
Classification of Alloys.-—Die-casting alloys are grouped as those having
low melting-points, below about 800° F., and those that fuse above that
temperature. The first are by far the most extensively used, comprising
the numerous white metals; the second are the aluminium and copper alloys.
Alloys are classified according to their bases, signifying by this the metal
which occurs in the largest proportion, and so determines the leading charac-
teristics of the alloy. These are zinc, tin, lead, and aluminium. A very large

selection of alloys is essential because of the multifarious uses to which the
castings are applied. In some cases the expansion of an alloy under high
temperatures would preclude its use. In many cases steam, oil, alkaline
liquids, sea water, and corrosive fluids would disintegrate some alloys, while
having no effect on others. Some alloys are too brittle for certain services,
others are too soft, while in some, an element will sweat out from the mass.
These facts indicate the difficulties which have to be surmounted by the die-
In the zinc-base alloys this metal may be used in a range of from 50 per
cent to 80 per cent, tin from 5 per cent to 30 per cent, and copper and alu-
minium from a mere trace to about 5 per cent. Antimony may be present
from i per cent to 5 per cent, its function being to reduce shrinkage and
impart hardness. Only a small quantity is necessary, since zinc is hard,
and does not shrink so much as tin or lead. These alloys melt at from 800°
to 850° F. They are affected by alkaline and salt waters. They are the
easiest to cast, and the strongest castings are obtained by keeping the tin
and copper at, say, from 2 per cent to 5 per cent. Tin is liable to sweat
out below the temperature of fusion.
The tin-base alloys contain from 60 per cent to 90 per cent of the metal
with from 3 per cent to 7 per cent copper, and about the same proportion of
antimony. These alloys are excellent. They are softer than those with
a zinc base, and produce castings of good finish, but the price of tin makes
them expensive. Babbitt is composed of tin 89 per cent, copper 3-7 per cent,
antimony 7-3 per cent with a trace of bismuth. These alloys melt at from
200° to 300° F. lower than those having a zinc base.
In the lead-base alloys the proportions of that metal are high, but their
uses are almost confined to the bearing metals. Lead may range from 60
per cent to 90 per cent, tin from 2 per cent to 20 per cent, antimony from
4 per cent to 25 per cent. The alloys lack strength, and are heavy. The
tin increases the tenacity and toughness of the lead, while the latter renders
the tin more malleable and ductile. The higher the percentage of tin, the
better is the surface of the castings, being smoother and brighter. Shrinkage
is reduced. Antimony increases fluidity and imparts hardness. The
maximum hardness is imparted with 17 per cent of antimony. Up to 13
per cent it expands the lead. Lead will not alloy with zinc, because segrega-
tion occurs during cooling.
When aluminium-base alloys were required for some parts of machine-
guns, pistols, grenades, binoculars, &c., difficulties were encountered because
of the higher melting-point. The temperature of any alloy must not be
higher than would prevent it being melted in an iron pot. Aluminium exerts
a solvent effect on iron, and a small percentage is found in the castings. An
excess over 3 per cent renders the aluminium alloy useless, causing it to
become viscous by the raising of the melting-point. A standard mixture is,
aluminium 92 per cent, copper 8 per cent. Small quantities of zinc, nickel,
and manganese may be included.
Attempts to die-cast brass and bronze have not been crowned with


commercial success. Both the temperature of pouring and the coefficient
of expansion are high. With the increased shrinkage the dies are strained
badly, and castings crack. The cost in any case is prohibitive. To produce
at a profit, the life of a die should be equal to that of 10,000 casts. It has
not been found possible to exceed 1000 in the brasses. To pour iron would
also destroy these moulds. Yet under different conditions this is done, as
described in the next section.
Permanent Moulds.—These are made of cast iron, and iron castings
are produced in them. The advantages gained are: (a) the saving of time
otherwise spent in making a sand mould for every cast; (b) the more rapid
removal of the castings from the moulds when set. The development,
almost wholly confined at present to the United States, is a remarkable one.
Its applications are chiefly to pipes required in large numbers, the cores
for which are also made in iron. The same grades of iron may be used for
castings made in permanent moulds as for those in sand, but iron that would
not give satisfactory results in the latter will do so in the first named. Also,
harder or softer castings can be obtained from the same metal, depending
on the time during which they are permitted to remain in the moulds. A
surface chill can be imparted, or the casting may be soft throughout. In
any case, cored castings must be removed before they shrink tightly on their
cores, otherwise they must be broken up.
The mass of metal in a permanent mould must be large, because a thin-
walled mould would become heated so quickly that a rapid succession of
castings could not be produced. For castings of 15 Ib. weight and upwards,
the mould should be of about seventy times the weight of the casting.
Castings are removed soon after the outside has set. This will usually occur
in from six to ten seconds, but the time will depend a great deal on the
weight, the degree of hardness, &c., required. That this can be varied,
though using metal of the same chemical composition, is one of the valuable
features of these moulds. If a casting is allowed to remain long in contact
with its mould, it becomes chilled, a large proportion of the carbon remain-
ing in solution, in the combined form; but if the casting is removed at a
bright yellow colour while the interior is still viscous, the exterior will
become annealed, and the casting will be soft, the carbon passing mostly
into the graphitic state.
It has been proved that iron which is unsuitable for sand casting is excellent
for permanent mould work. An iron with a percentage of phosphorus as
high as 1*5 per cent, and of sulphur o-i per cent, and silicon 2-5 per cent, is
as strong as one with smaller proportions. The explanation is, that these
remain in normal solution, not having time to separate out.
Permanent mould work has its limitations, due to the fact that iron cores,
which must be drawn out endwise, are used. This limits the forms of pipes
to those with straight or regularly curved cores. The same hindrance
occurs in die-casting. As the castings have to be removed quickly and the
moulds are massive, a good deal of mechanism is necessary for rapid and easy
handling. In general, from one to two castings are poured and removed
CASTING  THE   METALS  AND  ALLOYS             145

per minute.    Chilled car wheels, gear wheels, projectiles, and pipes are the
principal articles made in these moulds.

Casting the Metals and  Alloys
Although the principles and the general methods of making all sand
moulds are similar, yet some details have to be varied with the character of
the metal or alloy used. These are so important that the work of the different
foundries is carried out by different sets of men who have become specialists.
Each of these departments would admit of extended treatment, but the leading
facts only can be stated here.
The Iron Foundry.—This, which embraces the largest proportion of
cast work, is in a sense the standard to which the practice of the other depart-
ments is referred, and with which they are contrasted. The shrinkage of
iron is moderate, averaging -| in. in 15 in. The metal is poured mostly into
moulds made of green sand, the ingates and runners of which need not be
very large, since the metal flows freely, appearing when thoroughly melted
nearly as liquid as water. The thinnest pipes and plates can be poured,
no trouble arises from the segregation of the elements, and, generally, the
conditions under which the work of the iron foundry is performed are
satisfactory. The pouring of moulds has to be modified with the grade of
iron used. The grey irons, with say 3 per cent of graphitic carbon, remain
fluid longer than the mottled grades with about half the carbon in the com-
bined state, and therefore the runners for these have to be dimensioned
to fill the mould more rapidly. Another fact is, that the effects of shrink-
age are more severe, with liabilities to fracture if shrinkage is hindered.
The Steel Foundry.—The difficulties of the steel founder are those
consequent on the high temperature of casting, and the large amount of
shrinkage. While the temperature of molten grey iron is about 2250° F.,
that of steel is about 2800° F. As the melting-point of silica sand is in the
neighbourhood of 3200° F., partial fusion of the mould is liable to occur.
This is the reason of the rough skin seen on so many steel castings. Hence
these are seldom made in green sand, but in dried moulds with a sand mixture
high in silica. Only new sand which has not been damaged by heat is used
for facings. The chief trouble has always been the shrinkage. This, which
amounts to about TV in. per foot, coupled with the high temperature of pouring,
inevitably produces cracks, warps, hollow places, and fractures in castings
that are badly proportioned. Patterns designed for the iron founder cannot,
as a rule, be used for steel. Runners have to be much larger, feeding heads
and risers are necessary, large fillets are inserted to strengthen adjacent
parts, and thin sections must not be tied. In some cases these precautionary
provisions will add from 50 per cent to 100 per cent to the weight of the
VOL. I.                                                                                                                       10
146                                 FOUNDRY  WORK
casting required. And even then the conditions of internal strain are so
severe that prolonged annealing of the castings is necessary for the sole
purpose of relieving these strains, and lessening the hardness of the metal.
Malleable Cast Iron.—This is a white iron, having the whole of its
carbon in the combined state. It is poured into sand moulds, and annealed
subsequently in ovens for about sixty hours. This changes the carbon
into the graphitic state, rendering the castings soft and extremely ductile.
White iron is necessary, because a grey iron would produce spongy castings
after annealing. The amount of combined carbon must never be lower
than 2-75 per cent. As the white irons, which are viscous when poured,
are used, the runners have to be large. The shrinkage allowance is also
greater than that for grey iron, and generally precautions similar to those
when making steel castings have to be taken, in the form of large shrinkage
heads, and the provision of fillets.
The Brasses and Bronzes.—In all these alloys the shrinkage is
large, being about J in. in 10 in. The metal is not so fluid as grey cast iron,
and it sets very quickly. Large runners and large shrinkage heads are there-
fore necessary, not with a view to prevent fracture, which rarely occurs, but
to avoid " draws " and hollow places in the more massive sections. Feeding
is necessary in almost all moulds, even more so than in iron, because the
shrinkage is greater. The metal in the pouring cup chills quickly, so that
fresh metal must be supplied if the mass of the casting is large. Special
care must be taken in making the dispositions of the runners. These must
be brought into the heavier sections. It is well in deep castings to pour
from the bottom. A very large volume of brass work is made with odd
sides, or is alternatively plated. In each case numerous small patterns,
which may be like or dissimilar, numbering, say, from half a dozen to twenty,
are moulded in one flask, and poured from a common ingate. In these
cases the runners must be of sufficient area to fill the moulds farthest from
the ingate before the metal has had time to congeal, and little or no feed-
ing can be done. Both green and dry sands are used for moulds, the
first, as with iron castings, predominating. Generally, the moulds should be
rammed harder than those for iron, and well vented.
Aluminium and its alloys are usually poured into moulds of green sand.
The shrinkage of the metal is about double that of brass, and large runners
are required. The melting-point is rather low, being about 1160° F. The
metal must be poured quickly, as it sets rapidly. The pouring basins are
large, to act as head metal. Metal rods are frequently inserted in moulds to
hasten the cooling of the thicker parts of castings. The alloys of aluminium
are numerous. The chief elements employed are copper, zinc, manganese,
and magnesium. Moulds of green sand are used, rammed more loosely
than those for iron or brass, to prevent shrinkage cracks. The sand may
be finer than that for brass, as relatively little gas is given off, and it need not
be very refractory. The moulds can be dusted with black lead or French
chalk. Green sand cores are desirable, but if dried, they must not be too
hard, or they will check shrinkage.



The Effects of Shrinkage in Castings

All the metals and alloys in common use shrink in cooling from the
molten state. Although the amount per foot of length may not appear large,
the fact is responsible for the deformation, the lack of homogeneity, the
weakness and the fracture of a very large proportion of the " wasters " made.
The evils arise from the different rates of cooling in large and light adjacent
masses, the very small capacity of cast metal for elongation, and the method
of its crystallization. The weakness of cast metals in tension, and their
very small percentage of elongation before fracture, when compared with
the similar physical properties of forged and rolled materials, are the causes
of these results.

The Case of Unequal Adjacent Masses.—Many designs that emanate from
the drawing office have to be modified to suit the foundryman's point of

Fig. 50.—Illustrates the Camber of Castings produced by unequal Shrinkage
view. Regarded from his aspect, the ideal casting is one in which thick-
nesses are approximately equal, with the result that all portions cool and
shrink simultaneously. The more intricate the casting and the larger the
amount of coring done, the greater is the need for preserving uniformity
of sections. Familiar examples are those of steam and motor cylinders,
in which the percentage of wasters is often rather large. In these and
other castings cores are frequently inserted, or prolonged solely to avoid
the occurrence of masses of metal in corners and angles.
But in many designs of machine and structural parts it is not practicable
to avoid great disparities in the masses of metal in parts that are contiguous.
The moulder then minimizes the evil results, first by " feeding " fresh,
hot metal into heavy masses, to prevent the formation of " draws "—hollow
places—due to internal shrinkage, and second, by uncovering the massive
section, exposing it to the air, in order to cause it to cool within about the
same period as the thinner portions adjacent. A great deal of this is done
in the case of central bosses for bedplates and heavy pulleys.
Curving, Camber.—Distortion without weakening or fracture is a common
result of unequal shrinkage. It is particularly troublesome in long and
flimsy castings, as bedplates, gutterings, and similar objects in which there
is an excess of metal, not necessarily large, on one side. The casting in cool-
ing becomes permanently concave on that side. In the group of figures,


fig- 5°> the section at A being symmetrical will be straight when cold.
B, c, and D will become concave along the wider flanges, but in different
degrees, c less than B and D, because its top flange is wider. E, where the
flange is very wide, will not curve. A wide web resists the effect of flange
shrinkage because it is rigid, and it acts as a carrier of heat to the
shrinking smaller flange, delaying its setting. The gutter sections in the
next group, fig. 51, will all become concave on the solid sides. While

A and B will have a
curve   in   one  direc-
tion, c will be curved
A                              B                    C         both  on  the  bottom

Fig. 51.—Illustrates the Camber of Castings produced by unequal Shrinkage     and   the Vertical   side.

The difficulty which

confronts moulder and pattern-maker is how to counteract the effects of
shrinkage in unequal sections. No possible rule can be stated, and ex-
perience of similar classes of work is the only guide. The greater the
disproportion, the more flimsy the casting; and the greater its length, the
larger will be the departure from lineal accuracy. A moulder will sometimes
uncover a casting or a portion of the same while at a red heat, to hasten the
cooling, and so prevent curving. But that is not always practicable, nor
is it a sure method. Generally, the pattern-maker imparts camber to the


•til Hiiii1

Fig. 52.—Crystallization in Cooling
pattern in the opposite direction from that which the casting would assume.
Uncertainty, when work is repetitive, is avoided by making one trial cast-
ing, noting its amount of camber, and altering the pattern accordingly.
Crystallization.—The needle-like crystals of cast metals arrange them-
selves normally in relation to the surfaces of the mould. In fig. 52 the
strongest form is shown at A, and the weakest at B. The cylinder at c,
terminated with a semi-sphere, is much stronger than one terminated with
a flat end. These are commonplace axioms, but they have infinite applica-



tions in all castings made. If the pattern-maker does not put a radius, the
" hollow " or " fillet " in a keen angle, the moulder rubs one, as at D, thus
altering the weak crystallization of B to that shown. Additional strength is
afforded by the bracket at D, common in flanged structures, and which steel
makers often insert when not done in the pattern, to prevent cracking of the
casting. It is better to fit a bracket as shown at intervals, than to make the
fillet very large, because the result might be a " draw " (a cavity in the casting)
due to internal shrinkage, such as is seen at E, where three ribs meet with
large fillets. This would be prevented, and the casting be stronger if the
radii vvere smaller, which, while favouring suitable crystallization, would
reduce the mass of metal in the corner.

Some Common Precautions.—Castings, apparently sound, not infrequently
fracture during machining or subsequently. This is because they are in a
condition of internal tensile
stress, dangerously equal to
that of the ultimate strength
of the metal. Inspectors test
roughly for this condition with
hammer blows. Hard sand
cores and portions of dried
moulds interfere with shrink-
age, and a careful moulder will
break these up as soon as the
metal has congealed. At the

best the shrinkage is only lessened, but this in large castings may be
sufficient to counteract the allowance for tooling. Bars in flasks adjacent
to flanges (fig. 53) will check shrinkage, requiring the breaking away of the
intervening sand. Pulley arms are commonly curved, because they will
accommodate themselves to the pull of a shrinking boss instead of fractur-
ing. Pulleys with wrought-iron arms must have the boss cast after the rim
has become nearly cold. Large runners and risers will interfere with shrink-
age, and the moulder often knocks these off so soon as the mould is full.

Fig-S3-—Illustrates Shrinkage of Flanges

Right hand, Flange and weak.    Left hand, Flange reinforced
with brackets.

The  Furnaces

The furnaces include several types with many variations: for melting
iron, steel, the brasses and bronzes, and malleable cast iron. A large amount
of plant and machinery is associated with the operation of each, on which
greatly depend not only the economies of working, but the soundness and
strength of the castings produced.
The Cupolas.—With many differences in details, the essentials of a

Fig. 54.—The " Thwaites " Cupola

A, Shaft. B, Brick-lined charging door, c, Air-
belt. D. Tuyeres. E, Receiver. F, Slag hole.
G, Tapping spout. H, Hot-air pipe to receiver,
j, Fettling nole. K, Drop bottom. L, Blast pipes.

cupola furnace for melting iron are these
(figs. 54 to 57).   A tall cylindrical shell,
built   of wrought-iron  or  steel  plates,
lined with fire-brick, daubed with fire-
clay for each cast;  a charging door near
the top;   an air-belt encircling the shell
at a height of a few feet from the bottom,
whence blast under pressure is directed
through  tuyere  openings  to   iron  and
coke supported on a deep bed charge of
coke.    The furnace stands on columns,
and has a hinged bottom to permit of
the dropping out of the residuary coke,
metal, and slag at the termination of the
day's cast.   Peep holes with mica win-
dows are fitted opposite the tuyere holes
through which the furnacemen observe
the progress of the melting, and open-
ings are furnished for the removal  of
the slag, and the tapping of the metal.
Many cupolas include a receiver, a cir-
cular vessel into which the iron, passing
down through the bed charge of coke,
trickles and collects, remaining perfectly
liquid until it has to be tapped out for
pouring.     The   internal   diameters   of
cupolas range from about 18 in. to 6 ft.;
the first will melt about f ton per hour,
the last, about 12 tons.   These are ex-
tremes, the first being of value chiefly
for occasional light casts, and for mak-
ing tests of metal, the last being too
large  for  general   service,  for  which
internal diameters of from 3 ft. to 4 ft.
are preferable.
A cupola is worked as follows: after
re-lining the interior with fire-clay each
morning, the bed charge of coke is laid
in, extending to from 18 in. to 20 in.
above the tuyeres. Over this succes-
sive layers of pig or scrap, lime-stone,
and coke are placed, there being three
or four repetitions in this order until
the charging door is reached. The
fire is lit, and the interior warmed
before the blast is put on. In about

fifteen minutes the metal begins to run down. As the charges sink, suc-
cessive additions are made in the order named. Melting is facilitated by
breaking the pig and scrap into small pieces. As fusion is confined to the
area immediately above the tuyeres, extending therefrom to a height of
about 30 in., metal of different grades, harder and softer, can be charged in
the same cupola at the same time if separated with charges of coke. The

k, N. *                metal  accumulates in the

vxv>*                bed charge, and must be

tapped  before  it rises to
the tuyere holes.

The Melting Ratio.—
Most of the modifications
that have been made in
cupola design have for
their object an increased
melting ratio, which is

Fig 55.—Cupola, with Air-belt A and three rows of Tuyeres B arranged spirally.   C, Receiver.
D, Slag holes.   E, Tapping hole.   F, Drop bottom.
accomplished by supplying enough oxygen in the right place to secure the
nearest approximation possible to complete combustion. If a ton of iron
is melted with from 2 to 3 cwt. of coke, that represents good average practice.
To use less than 2 cwt. of coke is exceptional. This is only possible in
lengthy fusions, using: (i) clean iron that throws out little slag; (2) good
furnace coke; (3) a deep bed charge; (4) suitable proportioning of fuel and
iron; (5) an adequate supply of blast at proper pressure and volume, with
variations made when necessary as the melting proceeds.
Since the supply of oxygen in the right locality is the master key to
economical melting, this explains the very numerous variations that have


been made in the arrangements of tuyeres. Briefly, these usually consist
of upper and lower rows, receiving the air from the belt, and discharging it
through openings equally spaced round the circle. This disposition has
taken the place of the older method of bringing in blast through pipes into
two openings on opposite sides, which, with the low cupolas then common,
permitted a large proportion of the gases generated from the fuel to pass








Fig. 56.—The " Colliau" Cupola

AF Air-belt.  B, Flaring tuyeres,   c, Non-conducting
space filled with sand.

Fig. 57.—" Newten" Cupola

A, Air-belt.   B, Differential tuyeres.   C, Drop

away out at the charging door and at the top, unconsumed within the furnace.
When carbon is burned, 14,647 B.Th.U. are given out per pound, the carbon
uniting with the oxygen to form carbon dioxide, C02. This is called com-
plete combustion. If, however, the combustion is incomplete, due to an
insufficient supply of oxygen, carbonic oxide, CO, is formed, and if this is
allowed to escape, about two-thirds of the heat is wasted, since the burning
of carbon to CO evolves only 4415 B.Th.U. The object of the upper row,
or rows, of tuyeres which have assumed bizarre forms in some designs is to


supply the additional oxygen to the CO formed lower down by the com-
bustion of the coke. The same result is accomplished by additional height,
since a more prolonged contact of the carbonic oxide with the heated blast
is assured. For it must be remembered that the blast is cold when it enters
the furnace, and its oxygen must be highly heated before it will enter into
combination. Much heat is wasted in warming the upper charges, and the
large proportion of inert nitrogen in the blast.

Blowers and Fans.—The first named (fig. 58) are used now more often
than the second, because the action is positive, the air being driven out under
definite pressure. Good results are alsd obtained from fans if they are
selected and used with judgment, but generally they are more suitable for
the lower pressures, say not exceeding 8 oz. per square inch. The fan has
to revolve at a very high rate of
speed; that of the blower is
moderate, and the pressure and
volume are under better control.
The speed of a fan cannot be in-
creased beyond that for which it
is rated without absorbing power
that increases with the cube of
the number of revolutions. Hence
one of large diameter should be
selected to allow for contingencies.
In either case the supply pipes
must be large, free from quick
bends, and of minimum length
possible from the machine to the
cupola. A blast gauge is necessary
as a check upon the working. It
reads to z Ib. pressure, and is sub-
divided into ounces. It is necessary to regulate the blast at different stages
of melting. This is done by varying the amount of opening of the blast
gate. At the normal pressure of from f Ib. to i Ib. per square inch, the
blast must supply from 30,000 to 40,000 cu. ft. per ton of iron melted per
hour. The makers of blowers and fans give the capacities for different
sizes. From 3! to 4 b.h.p. per ton melted per hour are required.

Ladles.—These, up to about 3 cwt. capacity, are carried by hand, by one,
two, or three menu hence termed " hand shank ladles". Larger sizes are
slung in the cranes, or run on carriages on rail tracks. All are tipped when
pouring. Fig. 59 shows a common form, where the tipping is done through
bevel and worm gears. It is effected similarly in fig. 60. This type can be
run on tracks, or lifted in a crane. Both have two pouring lips, to be tipped
to either side. The bodies of ladles are formed of pressed steel plates,
stiffened with belts. Capacities are reckoned inside the fire-clay lining with
which they are daubed each morning. A cubic foot of ladle capacity is the
equivalent of 3 cwt. of iron.

Fig. 58.—-Section across Pressure Blower

Pig Breakers.—A great deal of pig is still broken up with the sledge.
Fig. 61 shows a machine used for the purpose. The pressure is not direct,
but operates through a lever arm that is pushed up by a hydraulic ram,

Fig. 59.—Crane Ladle, double geared

forcing the short arm down on the pig. Its valve is actuated by the treadle
seen at one side, and counterweighted. The pig is broken at the end that
overhangs, but some machines fracture it centrally. Some machines are

Fig. 60.—Carriage Ladle, double geared
driven by belt, others are electrically driven. Scrap is broken with the sledge
if light, and by the dropping of a ball from a crane if heavy. Labour may be
saved by using a lifting magnet slung in a crane. This is lowered on the
ball, the current turned on, the ball lifted, the current switched off, and the


ball drops. Balls are usually of half a ton weight. Heavy scrap is also cut
up with the oxy-acetylene flame.

Steel-melting Furnaces.—These are only used to a very moderate
extent outside the great steel works. Some of the larger iron foundries
make what steel castings they require in preference to sending away for them.
Special designs of furnaces are provided for such cases. Instead of the
great open-hearth furnaces which will melt 50 tons, or the immense Bessemer
converters, small " Baby " con-
verters are used, the Robert,
one of the earliest, and the
Tropenas being most common.
The small furnace can be used
for casts as low as 10 cwt.
The melting is so rapid that
two successive melts can be
poured into the ladle for a
single cast. Ferro-alloys can
be added in the ladle to pro-
duce just the amount of recar-
burization desired. The waste
of metal is rather large, and
the upkeep costly.

These converters are made
in capacities of from |- ton to
2 tons, and they are made to
tilt for pouring the. charge.
The blast is brought in at one
side only through tuyeres, and
is directed through the metal,
or over its surface. A pressure
of from 3 to 4 Ib. per square
inch is necessary. This is
supplied from a blower. A
cupola supplies the molten

metal, which must be melted much hotter than that for the iron foundry,
besides which more heat is required to melt the scrap steel included.
The latter may amount to from 25 to 50 per cent of the charge.

Brass-melting Furnaces.—While few iron foundries possess a steel
plant, there are not many of fair dimensions destitute of a department for
the melting of the brasses and bronzes. Castings in these alloys enter into
nearly all constructions, and the delays and risks attendant upon getting
castings from distant firms render the brass foundry a most valuable annexe
to that of iron. A few years ago there was little choice in the matter of
furnaces, now they rival the cupolas, both in variety and increased efficiency.
Natural draught with coke fuel, blast, oil fuel, and electricity, each with
many variations, are now employed regularly.

Fig. 61.—Pig Iron Breaker


Furnaces fed with the natural draught of a chimney, and burning coke,
the earlier and most common design, are not economical, but for small casts
they are not to be despised. The best in this design are Carr's (fig. 62),
where the fire bars are placed below the bottom, leaving a space above,
through which most of the air passes. The melting is rapid, and the crucible
does not sink. The brick lining is carried on' a flange within the furnace
above the air-space, and a non-conducting backing of broken bricks fills the.
space between the lining and the outer casing of iron. Furnaces are built
to take one or more crucibles. Several furnaces can communicate with
flues leading to a common chimney, as in the ordinary brick furnaces.

Fig. 62.—The " Carr " Brass-melting Furnace
A, Solid lined furnace.   B, Non-conducting space filled with broken brick,   c, Chimney 30 to 35 ft. high.
D, Ash-pit.   E, Pit for taking out ashes.
Improved designs of coke-fired furnaces in extensive use include pre-
heating of the metal, tilting of the crucible while in the furnace for pouring,
and the employment of artificial blast. In the first, the metal is placed in
a crucible or other annular vessel, above the melting crucible, where it is
warmed by the heat escaping from the fuel below, before it drops into the
lower crucible. The latter is not removed from the furnace, but both are
tilted for pouring. The preheater can be swung to one side when fresh
coke has to be charged. This design permits of the employment of larger
crucibles, and the attendant suffers less discomfort than when the crucible
with its charge has to be lifted out with tongs from above. The employment
of blast results in a great saving of coke, when, as is sometimes done, the
blast is warmed during its passage by the waste heat from the furnace. It
is possible in some of these designs to melt 1000 Ib. of brass in one charge.
THE  FURNACES                              I57
For large installations, the preference should be given to furnaces that
are fired with oil or gas, or with a mixture of each. There is no large con-
sumption of fuel in the preliminary heating up, and no waste of fuel unburnt,
at the end of the melt, as there is with coke. The temperature is under
precise regulation, and ashes have not to be removed. On the other hand,
oil storage reservoirs and supply tanks have to be installed, with pipes,
cocks, gauge glasses, and, if the oil is sprayed under high pressure, a supply
of compressed air is necessary. If used with a low pressure, a fan or a blower
is employed. A low-pressure burner works with air at about 12 oz. per
square inch, a high-pressure one at from 20 to 25 Ib. per square inch. In
the American " Rockwell " furnaces the pressure for the oil is 5 Ib. or more,
and that for the air is 2 Ib. per square inch. Gas may be used instead of
oil, with burners and pipe connections modified. 100 Ib. cf brass can be
melted with from 2 to 3 gall, of oil, and after a furnace has been heated
with a first charge, 400 Ib. of metal can be melted in about 45 minutes.
Electric furnaces are being used in increasing numbers when large
quantities of brass are being melted, but chiefly in the United States.
Whether they are more economical than the oil-fired designs depends mainly
on the relative costs of power, attendance, and upkeep. But there is one
important fact in favour of the electrical designs, that the metal is melted in
a closed vessel, in a non-oxidizing atmosphere, and that there is then hardly
any loss due to the volatilization of zinc, of dirty borings, and of fine scrap.
This loss often amounts to 5 or 6 per cent in the fuel furnaces. As a result,
alloys can be graded and duplicated with such precision that the average
deviation is only about 0-25 per cent. Electric furnaces will not deal eco-
nomically with small charges, since, with their necessary equipment, they are
costly to install, so that, like the oil-fired designs, they are only suitable for
the large foundries.
Electrical energy is applied to the melting of brass by two methods: by
means of the electric arc drawn between electrodes, or by the resistance
offered to a current by its passage through liquid metal, on the same principle
as that of the heating of an incandescent lamp. Each design has its advocates,
and each has its application in several furnaces that are in successful operation,
melting quantities that may range from 200 to 2000 Ib. weight. Some of
the furnaces are stationary, some tilt for pouring. A few are rocked through
an arc to maintain a uniform temperature, and prevent surface superheating,
while a perfect mixture of the metals that form the alloy is produced.
Generally, the mixture is contained in a bath in the bottom of the furnace.
This method is better for heavy charges, but for moderate casts a crucible
design of furnace is made, in which the metal is melted by the passage of
electric currents through the crucible walls.
The arc furnaces may have the arc drawn between two electrodes of
graphite, or of amorphous carbon, provision for the adjustment of which is
made by hand or electrically. The heat is transmitted to the metal below
by radiation chiefly, although in one design it is directed downwards by
a third electrode, placed vertically above, which forces the flame of the arc
153                              FOUNDRY  WORK
down on the charge. Or electrodes are inserted perpendicularly, and the
arc is drawn between these and the bath of metal, or the slag or carbon in
the trough.
The furnaces that operate by electrical induction must be so designed
as to counteract what is termed the " pinch effect". When the molten
metal lies in an open channel in a horizontal plane, a break occurs in the
current at an early stage, interrupting the circuit at the point of smallest
cross-section, and checking the melting. This pinch effect, which does not
occur in furnaces melting steel, has to be counteracted by producing a violent
circulation of the liquid metal in secondary channels or loops situated below
the charge. This is effected in different ways, in which the electric energy
is converted into heat, with rapid movements, sufficient to prevent inter-
ruption of the circuit.
Furnaces for Malleable Cast Iron.—Frequently, these are air
furnaces of the reverberatory design. To a very small extent, cupolas and
open-hearth furnaces are used. As the white iron used has to be melted
very hot, the reverberatory furnaces are built of great length, and the metal
is tapped where it is hottest. This occurs near the fire bridge, and the bed
is sloped towards this part. The fire grate is located at one end, and the
chimney at the end opposite, or to one side. The flame passes over a bridge
next the hearth, and is deflected on the metal by the low roof, which is
usually arched. To facilitate the charging of the metal, the roof is generally
made in separate sections, " bungs ", each consisting of an iron framing,
enclosing fire bricks. The sides of the furnace are built of steel plates,
reinforced with binders, and the foundation is concrete. The lining is of
brick, enclosing fire brick, also used for the roof. The working bed is of
siliceous sand, and is relined when it becomes burned away.
The annealing of the castings is done after they have been fettled, with
the result that the combined carbon is nearly all changed to graphite, and
the castings, instead of being intensely hard and brittle, have their strength
and ductility greatly increased, so that they have acquired the general pro-
perties of iron forgings. The castings are packed in boxes, " saggers ",
with hammer scale or haematite ore, piled in furnaces, and subjected to a
prolonged temperature of from 800° to 900° F. in annealing ovens. The
designs of these are numerous, though the principle is simple. The boxes
of castings, luted to exclude all air and piled in the oven furnace, are sub-
jected to the heat from solid fuel burnt in a grate at one end, or from gaseous
fuel. Flues are arranged beneath the floor, frequently also at the sides and
roof, designed with the object of delaying the escape of the hot gases until
they have rendered up all their useful heat.
Essential Machines and Appliances
The more advanced foundries of the present day employ labour-saving
methods to an extent that would have been deemed impracticable a few
years ago. Yet in too many shops wasteful ways, which are a financial
handicap in competitive efforts, are retained. It seems desirable, therefore,
to give attention to this particular aspect of foundry work, dealing with
the preparation of the sands, with machine moulding, with fettling, and
the lifting and transport systems.
The Preparation of Sands.—Sand when new from the quarry is not
suitable for moulds without preliminary treatment. This is performed in
isolated machines, or in one large plant, which is only installed in the big
foundries. New sands are wet and lumpy, often having pebbles intermixed.
Drying is necessary. In small shops this is done in the core stoves, the sand
being spread on iron plates. In the bigger foundries, drying cylinders,
which measure about 6 ft. in diameter by several feet in length, are em-
ployed. They are either disposed with the axis horizontally, or at a slight
angle. The sand, fed through a hopper, is carried along the interior of the
revolving cylinder with spiral plates, and thrown against baffle plates, which
bring it into intimate contact with the hot gases from a furnace that traverse
the cylinder. The rotation is slow, being about i r.p.m. These machines,
in different capacities, will dry from 10 cwt. to 3 tons of sand per hour.
After drying it is necessary to crush, pulverize, and grade the sand.
The machines used for these processes are edge runners, disintegrators,
riddles, and sieves. Crushing is only necessary with the coarser, harder,
clayey sands, and is not adopted with the finer qualities, but instead the
lumps are triturated. In small foundries they are broken with a punner,
and the product with the ordinary mass is put through a riddle. The
machines that crush (fig. 63) are also used for mixing wet loam, hence termed
" loam mills ". They are similar to mortar mills. The lumpy sand is
ground between revolving runners and the bottom of the pan, which is
commonly fitted with removable chilled plates. The runners are frequently
chilled, or they are steel-tyred. Scrapers are fixed at an angle to heap up the
sand in front of the runners. These revolve on their shafts, and are at the
same time rotated around the pan on a central vertical shaft. In some cases
the pan revolves under the runners. Driving is done through belt pulleys
and bevel gears, and the pulverized sand is discharged through a shoot at
the bottom of the pan. Many pans used for mixing loam have their rollers
deeply indented like huge cogs. These throw up the loam, and amalgamate
it very thoroughly. One of these is often used with a smooth roller on the
opposite shaft. Some runners again are deeply grooved, in annular fashion.
The next process is the trituration of the sand to bring it into a fine,


loose condition preparatory to passing it through the riddles and sieves.
This is done in the disintegrators, which consist essentially of annular rows
of prongs carried on a disc, which revolves at a very high speed. The sand

Fig. 63.—Grinding Mill
A, Runners.   B, Pan.   c, Chilled bottom plates.   D, Scrapers.   E, Shoot for discharge.
is beaten and thrown about violently. The prongs stand vertically ^ (fig.
64) or lie horizontally in different designs. They are carried on a single
disc, or two discs face each other with the prongs on one entering the spaces
on those of the other, the rotations being in opposite directions. Or

one may be stationary while the other rotates. The shaft of one disc is
hollow to receive that of the other. They are driven with separate belt
pulleys, or a bevel wheel on a pulley shaft drives a similar wheel on each
disc shaft, in opposite directions.

Riddles and sieves are used to grade sands into coarse and fine varieties,
to separate portions imperfectly pulverized, and, in the case of old sand, to
get rid of cold shots and nails. The former generally consists of a frame with
parallel rods, leaving open spaces of J in. or so, while a sieve has a reticulated

y     iWT  V

Fig. 64.—Sand Mixer and Disintegrator
A, Hopper.   B, Revolving prongs electrically driven.
mesh of crossing wires. Hand-operated riddles and sieves reciprocated on
a horse are too slow in action. Any machine is far more economical. The
simplest is that in which the ordinary round sieve is attached and locked to
a light iron frame reciprocated with a belt-driven pulley and crank. This
can make 800 reciprocations per minute, and deal with as much sand as a
man can shovel into it, an output of 3 tons per hour being possible at an
expenditure of from J to |- h.p. Larger machines have sieves and riddles
made to interchange in a rectangular frame driven by cranks and connecting
rods, and sloped at a slight angle from the horizontal to throw the lumps that
will not pass the meshes out at one end. Machines of this class will deal with
quantities ranging from 3 to 14 tons of sand per hour, with J-in. mesh, the
output being less with finer grading. To deal with larger quantities machines
have the sieves arranged on six sides, enclosing the sand, and rotated on a
VOL. I.                                                                                        11


central shaft, making about 30 r.p.m. A jarring action is produced by the
contact of cams, which assists in breaking up the sand, that is also thrown
about by internal stays,

Coal is ground to dust in mills provided with heavy rollers, or balls, the
first being used within closed cylinders, the second (fig. 65) in open pans.
The balls, of cast iron, about 10 in. in diameter, are rotated in an annular
path having a concave section of rather larger radius than that of the balls.
The same mills may be used for pulverizing sands.

In the largest foundries these units are associated in one automatic system

Fig. 65.—Coal Mill

for continuous treatment. In general, the arrangement is as follows: raw
sand is thrown into a hopper at the base of an elevator, which discharges it
into a drying oven. Thence it goes into the grinding mill, afterwards into
a polygonal sieve, and then to a mixing apparatus, where the coal dust is
added in the correct proportion. The old sand is treated in another part
of the plant, conveyed for admixture with the new, the product elevated into
a disintegrator, mixed, and stored in bins for use.
Machines for Moulding.—It is not possible to describe here, even in
barest outlines, the leading types of these machines, of which the useful
varieties must now be numbered by hundreds. The only way to treat this
immense subject is to state with brevity the forms and utilities of the principal
elements in their designs, with comparisons of the methods and economies
of their operations.


Mention has been made on a previous page of the loss of time involved
in the preparation of a dummy box of sand, on which the parting joint is

Fig. 66.—Valve Body Patterns of Wood mounted on Wooden Plate.   The plate, cleated at the ends, has
open joints and strips of hoop iron to secure box pins.

made in moulding by turning over.    This wasteful method is avoided in
all machine moulding, as it is also in all odd-side work, and in the plating of


Fig. 67.—Cock Body Patterns of Metal with Ingate and Runners mounted on Iron Plate
patterns, an immense amount of which is done without any assistance from
machines. As these are more widely utilized the value of the odd-side lessens,
while that of plating grows. Its basis is the plain bottom or joint board


independent of its pattern, which
is laid on the face of the board.
To this the bottom box or drag
is pinned before ramming. The
time otherwise occupied in making
a temporary sand bed on which to
ram the drag (to be afterwards
thrown away) is saved, and the
board provides a true joint plane
without strickling and sleeking it
with the trowel. From this to the
permanent mounting of a pattern
or a portion of a pattern or more
than one pattern on a plate of
wood (fig. 66) or of iron (figs. 67
and 68), where pattern portions
are attached on opposite sides of
the same plate, is a natural de-
velopment, as is also their trans-
ference from the floor or the work
bench to the table of a machine.
Economies do not cease here, but
they increase when joints are of
non - plane shapes, combining
slopes and curves, and when several patterns are mounted on one

plate, each requiring a separate
runner. In these cases it is
usually preferable to cast pat-
tern parts, plate, and runners
all in one piece, than to adopt
the method common with
plane plates of preparing the
patterns separately, and at-
taching them to their plates
with screws or rivets.

Obviously, the moulding
table is the first important
element in any machine, since
it is the plate to which the
pattern parts are attached
directly, or to which the
patterns, already mounted on
their plates, are secured.
Tables either turn over, to

Fig. 69.~-Turn-over Table Machine, with presser head above t^ing each faCC Uppermost
and carrying-off table that runs on tracks below. Pattern parts /£/*« Ar\ «mr1 nr\\ nr thp»v sir A
are mounted on plate attached to the table.                                  Vn5s- °9 ano 7% or mey drc

Fig. 68.—Brake Blocks mounted on Plate



fixed (fig. 71), in which case only the top face is used. In a relatively
small group, top and bottom faces of fixed tables are used, by pressing
boxes of sand simultaneously against pattern parts mounted on each face,
these being worked hydraulically (fig. 72). Using a turn-over table, the
sand is rammed (fig. 73) or pressed (figs. 69 and 70) over the pattern portion
on the upper face. After being turned over, with the box, the latter is with-
drawn downwards, and the other portion of the pattern, on the opposite face,
being brought upwards (fig. 74), is rammed. The closed mould is seen in
fig. 75. The majority of machines of small and medium dimensions have

Fig. 70.—Hand-moulding Machine with Turn-over Table
A, Turn-over Table.   B, Plates to secure patterns.   C, Sand frame.   D, Presser head.
tables of this kind. The large machines must generally have fixed tables.
In these, the mould is lifted off its pattern with rods or " stools ", or with
power. In some designs the table is rocked over to permit of the lifting of
the pattern out of the mould, or, in a very large number of cases, the pattern
is withdrawn downwards through a stripping plate (figs. 71, 76, 77, 78),
this being necessary in all those patterns which have deep perpendicular
sides, and desirable even when depths exceed 3 or 4 in., being beyond the
limit at which delivery can be assisted by rapping.
After plating, the two important details in the moulding operations are
ramming and delivery. Mechanical aids are provided for these in most
machines, but not in all. The cost of hand-ramming increases with the
dimensions of the mould, and with the intricacy of its details, so that several
hours may be occupied thus in moulds measuring several feet across. Here


the machines afford great economies, since they will " press " or will
" jar-ram " the largest moulds within their capacity in a few minutes, the
time spent depending chiefly on the rapidity with which the sand is thrown

into the box part. The amount
required for compression is mea-
sured within a sand " frame " of
wood or metal (fig. 70). Except
in the deeper moulds, and under
loose, projecting pieces, no pre-
liminary peg-ramming is required,
but two or three squeezings with
the presser head suffices. In the
jar - ramming machines a few
bumps consolidate the sand in the

Fig. 71.—Radiators and Flanges of Motor Cylinder
drawn through Stripping Plate in a Fixed Table A

deepest moulds. These there-
fore, after the plating, afford the
chief economies of machine work.
Delivery of patterns by hand
is only the work of a minute or
two. The advantage of using a
machine is therefore that it sub-
stitutes an accurate mechanical
lift for the unsteady action of
the hands, and that in very many
instances the employment of

stripping plates prevents breaking down of the sand, and consequent
mending-up, with inaccurate results. Some machines do not include this,
but their utilities are confined to the shallower patterns, and those whose
shapes favour delivery. The mechanical withdrawal is furnished by means

Fig. 72.—Boxes Pressed and Delivered in Unison on a
Non-turn-over Table A


of guides that rigidly control the downward movement of the pattern away
from its mould, or the upward lift of the box off its pattern.   A little rapping

Fig. 73.—Bottom Box rammed over Pattern, then tabled, turned over, and mould delivered


Fig. 74.—Top Box rammed over Pattern, embedded in Plaster of Paris, then turned over and delivered


Fig. 75-—Mould Closed and Completed
Figs. 73-75.—Cone Pulley moulded on a Darting & Sellers' Machine
A, Pattern.     B, Its mounting,    c, Table of machine,    D, Bottom box.    E, Top box.
is done, which does not sensibly enlarge the mould, like lateral hand rap-
ping, but loosens the contact of the sand slightly. Either a hand mallet is
used, striking blows on the table, or a pneumatic piston produces vibration.


The most remarkable fact in connection with machine moulding after
that of the very numerous variations in the designs in use is that of the

IX XV   X   I                                                                                                                   KX V  XX

Fig. 76.—Portion of Pattern A being drawn through Stripping Plate B lined with White Metal.
C, Pattern plate.    D, Frame of Machine.

enormous growth in their dimensions and capacities.   Patterns, the castings
from which weigh several tons, are moulded on jar-ramming machines, and



Fig. 77

Figs. 77 and 78.—Brake Shoe, pattern part moulded
A, Pattern part for bottom.   B, That for top box.   C, Pattern plates.

those of several hundredweights on hand-operated kinds. Though the
movable parts are heavy, their mass is counter-weighted with weights or
springs, and movements are rendered easy with levers and gears. For power


:hines compressed air or pressure water are employed. Associated with
se machines are conveying systems for sand, flasks, and finished moulds,
iltiple moulding (fig. 79), where moulds are poured in piles, is sometimes
ipted for small castings made in quantities.

Machines for Fettling.—In small foundries the castings are cleaned
h little or no aid from machinery, the value of which grows with output.
ten a casting is taken out of the sand in the morning, nothing is done to
>n the moulding area, but it is transported to the fettling shed, where the
es are extracted, the runners and risers cut off, and a general examination
de to ascertain whether it is entirely sound before doing any work upon
If satisfactory, runner marks and fins are removed, together with all
plus lumps and adherent sand. This is done by hand with chipping
sels, coarse files, and scratch brushes of wire. But better methods are

Fig. 78

. Stripping Plates on " International" Machine
rame of machine.   E, Stripping plates lined with Babbit.

In the larger foundries, runners are cut off iron castings with a circular
•^ for those in brass, a git cutting machine is used, which consists of two
Dosed chisels actuated through a reciprocating slide, with power. These



IfMT      tffi






are not only more rapid in action than the severance by hand, but they involve
no risk of tearing out the metal below the surface of the casting. This is
liable to occur when runners are knocked off with the hammer. This may
be prevented by nicking all round with a cold chisel before using the hammer.
In any case the surface has to be smoothed by chipping and filing, which is
avoided when a machine is employed for severance. Fins occur more or
less on all castings, following the mould joints, and the fitting of cores in
their print impressions. These are laboriously removed with hammer and

chisel. The pneumatic chipping
chisels are far more efficient.
Much economy of time results in
this kind of work, and in smooth-
ing lumpy and rough portions,
when emery grinding wheels are
installed. The larger sizes are
mounted on a floor stand, the
smaller on a work bench. To
deal with castings that are too
large to be handled and presented
to the fixed machines, wheels are
mounted on suspended arms to be
swung by the workman into any
required position.

When large quantities of small
castings have to be smoothed, this
is done, after preliminary grinding
for the removal of fins and ex-
crescences, in a tumbling barrel
or nimbler. This is a cylindrical
vessel from 18 in. to 36 in. in
diameter by from 30 in. to 60 in.
in length, rotated round its longi-
tudinal axis about once in a
second. Within this the castings
are tumbled in contact with small

" stars ", and are smoothed and polished by the mutual friction set up.
Iron is tumbled dry; brass, with water. The driving is done with a belt
direct, or through gears, or the drums run on rollers. Its axis is horizontal
or inclined. Polygonal drums are made, with chilled lining plates.

Since large castings cannot be put in tumbling barrels, these, in the more
advanced foundries with a sufficient output, are treated in a sand-blasting
plant, for which an air compressor, giving a low blast pressure ranging
between 5 and 25 Ib. per square inch, is necessary. Having a suitable plant,
castings of any sizes and weights can be cleaned. The castings are placed
in a room, constructed of sheet iron, having a perforated steel floor and a
glass roof, well ventilated. The sand, propelled by the air pressure, is





Fig 79.—Multiple Moulds poured in Files through
Ingate A
directed through a nozzle held by the operator to any portion of the castings.
The sand falls through gratings in the floor into a hopper, to be drawn by
an exhauster back to the sand-supply. The attendant is protected with a
helmet of felt or leather, covered in front with sheet rubber, which prevents
dust getting into the lungs, and material from striking his face. Air for
respiration enters through a hose at the top, the expired air passing out
between the lower part of the helmet and the shoulders. Glass is not used
for the eyes, since it would become obscured, but fine wire gauze instead.
Quartz sand is used, but chilled iron sand is better. It is prepared by
atomizing a stream of molten iron with jets of steam projected into a tank
of water. There are several designs of sand-blasting plants now in use,
Lifting and Transport Systems. — The calls for hoisting and
transport are incessant in the foundry. Much time will be wasted if the
provisions made for these are inefficient. The overhead travelling crane
is the best machine to install, because it will command the entire area of the
shop. Its power must be rated by that of the weight of work being done.
The most economical design is the three-motor crane, in which the motors
are respectively rated for hoisting, travelling, and cross-traversing. Cranes
of different powers are installed in different areas, to suit the work being done.
It is well to supplement these with a few swinging jib cranes located in areas
where mould parts are likely to monopolize the crane service for considerable
periods. As these are attached to the columns that support the roof, they
do not block any shop area. Very light moulding makes few demands on
cranes, and overhead tracks, from which depend pulley blocks, or light
hoists are often provided for these departments. An equally good alternative
is a light overhead traveller, worked with a dependent rope from below.
Many of these are driven electrically.
The overhead cranes transport as well as lift rapidly, taking moulding
boxes and castings along the shop, and transporting ladles of metal. But,
since their movements are confined by the shop walls, they have to work in
association with extramural tracks, entering a few feet within. These are of
standard gauge to communicate with the yard tracks. Here the question
arises of employing floor tracks throughout the length of the foundry. When
these are laid down, as they often are, the gauge is 18 in. or 24 in. But they
are only desirable in the departments that deal with the lighter castings.
Where heavy moulds are being handled, the floor tracks are of less value than
the overhead travellers, and it is difficult to keep them clear of mould parts
in the morning, when these are laid open for cleaning and coring. In long
shops, devoted to the light work, they are useful for general service, even
though a light traveller is employed.
172                                FOUNDRY  WORK
Shop Arrangements and Organization
Only one design of foundry is regarded with favour now, a rectangular
building, parallel, with unobstructed roof light, and comprising one bay, or
more often two or three, each bay with its own roof, but without obstructions
in the shape of separating walls.   A clear area is thus included between the
'                           outer walls that permits of ready intercommunication and efficient super-
!                          vision.    Within these bays, the work of different departments is carried on
in strictly localized areas, each being served with the cranes and tackle that
s                          are specially adapted for the work to be done.    These departments in the
I                          majority of foundries include heavy green sand, light green sand, plate and
f                          machine moulding, often subdivided further, to locate castings made in
quantities by themselves, loam moulding, and core making. These classes
of work are done by separate groups of men who seldom handle any other
branch, having developed the faculties of experts. In addition, the melting
of metal is the exclusive task of the furnaceman and his helpers; sand grinding
and mixing occupy other hands; fettling is done in a separate room. Crane
operators are required, and there is a large proportion of loading, carrying,
and attendance on the moulders that engages the services of a body of
j                          unskilled labourers.
I                               Roof spans may range from 30 to 40 ft., depending on the bulk of the
j                          work done.   A height of about 25 ft. to the spring of the roof is suitable.
If symmetry is desired, spans should be equal and heights uniform, so that
future longitudinal extensions are simplified. The enclosing walls should
be of brick. The internal roof columns may be either of cast iron or built
up of steel bars and rolled sections. In each case attachments can be made
to receive the pintles of swinging jib cranes. The main section usually
terminates with the runways for the overhead cranes, and a separate smaller
section is carried up to the roof principals. A ridge roof is usual, with a
ventilating louvre -surmounting. The principals are of steel, formed of tee
sections and bars. It should be covered with slates, laid on felt, spread on
f                         boards. Illumination is provided by a continuous skylight along each
I                         ridge, or along the north ridge only.   An alternative is the saw-tooth roof,
j                         with north light, but this is not nearly so common as the symmetrical design.
j                         Puttyless glazing should be used, and a thick glass.   Windows are not
I                         necessary in the brick walls, but they relieve the otherwise depressing effect.
J                         As in most cases, the cupolas and the machines, together with the sand and
I                         coke stores, are located outside and close to the main building.   This pre-
!                         eludes the employment of windows there, but they can be inserted in the
f                         opposite side and in the end walls.
I                              The minute subdivision of tasks that is familiar in the big machine shop
i                         does not exist in the foundries.   Men are occupied in one or other of the
*                         leading sections previously mentioned, beyond which they seldom go.   But
the more specialized the firm's manufacture becomes, the greater is the
scope for the introduction of separation of tasks within the great subdivisions.
Thus, different sets of men will be engaged on the making of large and of
small cylinders, on pulleys, gear wheels, railway chairs, or any other articles
that are produced regularly in large quantities. This becomes amplified
in some machine work where the production of a complete mould is the
combined result of the labours of several men, none of whom are moulders
in the sense of being craftsmen. And while in the general class of foundries
the work is mostly done by the day, in all specialized tasks payment by the
piece is adopted. One foreman, with an assistant and a clerk, suffices for
the supervision of all the departments, but each has generally a leading hand,
who, by virtue of his experience and reliability, is placed in charge of it,
directing the routine, while he himself is engaged in the general work of the
The Machine-shop

The Work of the Machine-shop
Changed Aspects.—This department of the engineers' factory shows
changes more extensive than those which have occurred in any other, even
though the foundry and the smithy have been considerably remodelled. To
an old craftsman the changes are remarkable. A few years ago, the cost of
machining was so high that it was avoided as far as possible, and the work
of the foundry and the smithy was arranged to this end. Those were the
days of cored holes and " black fits ", when the pattern-maker and
smith were brought to book if undue allowances were left for machining. It
was the period, too, of weak machines, of single-cutting tools, mostly made
of carbon steel, with only a meagre proportion of " Mushet " tools. To-day
opposite conditions rule. Coring, black fits, and scanty allowances for
machining are discouraged. Holes are drilled in the solid; pieces of fairly
large dimensions are turned, bored, and cut off from solid parallel bars; and,
instead of being made from expensive forgings, articles are shaped from
blocks of metal first severed in the ubiquitous hack-sawing machine, then
chucked and reduced rapidly.
The Causes of Changed Practice.—These reforms are due to several
causes. There is an intelligent distrust of some of the old long-standardized
methods of dealing with certain articles on certain machines, and new methods
of solving problems of machining are justified by results. Another improve-
ment is the increased strength of machine-tools, which is accompanied by
more rational design and a general speeding-up. There are also improve-
ments in the forms and in the materials of the cutting-tools used, and of the
very numerous appliances, such as jigs, fixtures, and devices for multiple
and continuous machining. A rigid system of gauging is now in vogue.
These and other influences have brought the machine-shops of to-day into
strong contrast with their forerunners.
The manufacture of small arms, the smaller machines, motor vehicles,
aero-engines, and so on, in large numbers, has called for the production of
thousands of similar parts with fine tolerances. This work has had to
VOL. I.                                                           177                                                               12
178                             THE  MACHINE-SHOP
be done with a minimum of skilled attendance, and has been accom-
plished at an almost fabulous reduction in costs. The special machine
and its set-up of tools is the dominating fact. It secures the degree of
accuracy desired with continuous production, and eliminates the special
fitting of parts which used to take place, and substitutes " assembling " for
this costly operation.
An immense number of such special machine-tools has been designed.
Concurrently with this developing, new appliances have been schemed to
assist and extend the purposes to which the machines may be put, and to
economize the time of the attendants, and incidentally to relieve them of
responsibility. Always, the essential point is that accurate work is produced
in large quantities, while its cost is greatly lessened.
Machining Elementary Forms.—If the elementary geometrical forms
machined are observed, they will be found to be simple and few in number.
They comprise plane, cylindrical, and helical surfaces, though in great
variety. These are practically all; yet the types of machines built to form
these very simple shapes are numbered by the score, and the individual
designs run into hundreds. Yet the numbers constantly grow—hardly a
week passes but some new machine, possessing some special feature, is placed
on the market.
The role of producing any one of these simple geometrical forms is not
confined to one method or to one kind of machine. A plane surface can
be produced in the planer, shaper, slotter, drilling or boring machine, the
lathe, the milling machine, or the grinder. A cylindrical surface can be
machined in the lathe, the boring and turning mill, or the grinder, and, with
some limitations, in the drilling machine, the shaper, and the slotter. An
internal cylindrical surface (bore) can be made in the lathe, the drilling or
boring machine, the milling or grinding machine. A spiral or helical
surface can be produced in the lathe, the screwing machine, the milling
machine, or the grinder. Special shapes may be cut in lathes provided with
" forming" slides, and with suitable tools; in milling, grinding, and
broaching machines, and in gear-cutters. Similar forms are produced on
several kinds of machines, and this fact has had a vital bearing on the changing
practice of the present day. The machine chosen is the one which will do
the work required best and most cheaply. Three considerations arise: (i)
the selection of the best method or machine; (2) the dimensions of articles
and the relative positions of the parts to be machined; and (3) the degree of
accuracy desired.
i. TH!E SELECTION OF MACHINES.—It is not always easy to choose the
best from several possible machines. Machines naturally fall into groups,
and some machines of each group are better fitted than others for the per-
formance of certain tasks. It does not follow that because a planer will
work on short pieces it is the best tool for dealing with all short articles.
The shaper or slotter may be better. As a general principle a reciprocating
machine-tool should not be employed if a rotating one will produce satis-
factory results. Nor ought short screws to be produced on the screw-
THE  WORK  OF THE  MACHINE-SHOP             179
cutting lathe if other machines designed specially for cutting short screws
are available. Certain jobs are allocated to certain machines because they
have proved most suitable in practice.
2. DIMENSIONS.—The dimensions of articles to be machined naturally
determine the size of the machines used. The size of the machine is specified
in different ways, depending on the kind of machine. The lengths of
planer and other beds, the dimensions of tables, the swing, and the
centre-distance of lathes, the sizes of chucks, mandrels, and so on, determine
the " size " of the machines. The machine may take either a single large
piece or several smaller pieces; thus a series of articles may be put in tandem
on a machine-table, or be disposed around a chuck, or two or more articles
may be. placed on a mandrel.
3. ACCURACY.—The degree of accuracy desired is the factor upon which
the interchangeable system of manufacture depends. Certain " tolerances "
are allowed, and if parts which are to fit together comply with these
tolerances, any part A will fit any part B: for example, suppose a |-in. spindle
is to fit a hole approximately J in. in diameter. If the hole is drilled so that
its diameter is less than 0-505 in., and more than 0-495 in-> and the spindle
I                   is turned so that its diameter is less than 0-495 in. and greater than 0-49 in.,
then any spindle turned to these tolerances will fit any hole bored to the
|                   tolerances stated for it.   The tolerance allowed in the hole is 0-505 — °"495
= o-oi in. The tolerance allowed for the spindle is (0-495 — 0-49)
= 0-005 i*1- Now it is clear that the finer the tolerances the more difficult
and costly is manufacture. What then is the advantage of fine tolerances?
The advantages are that a noiseless smoothly-running machine can be built
I                   which will have great freedom from wear, because the moving parts have
no room in which to knock themselves to pieces.   The contrast between a
1                  Rolls-Royce and a Ford car engine is largely one of contrast between toler-
ances. In one case we have an expensive, smoothly-running car which is
cheap to maintain, in the other a cheap car, but one with higher maintenance
The Machines.—In a study of the machine-shop, some knowledge of
the standard machine-tools must be assumed—we are here chiefly concerned
with the later developments which have followed the changing practice of
the present day. It does not harmonize with that practice to deal with the
machines, as of old, in watertight compartments. The work of allied groups
frequently overlaps. What is of moment now is the modern way of regarding
the vast subject of machining, the reaction of this view on machine design
•I                 and selection, and on shop practice.
i8o                          THE  MACHINE-SHOP
The Tools
Single-edged Gutting-tools.—These are used in the lathes, planers,
shapers, and slotters. They are so termed because each tool has but one
edge, which distinguishes the group from the reamers and milling-cutters,
which have several edges acting in quick succession. They are an obvious
survival from the period when tools were presented by hand.
Formerly these tools were ground solidly with the shank or bar on the
end of which they were forged. Later, they have been more frequently
made separately, as tool points, to be gripped in holders, of which there are
many scores of designs. The expense of a higher grade of steel can then
be incurred for the small tool point, and in many instances tool points can
be disposed and operated to much greater advantage than when they are
forged solidly on long shanks. These holders occur in several machines,
but principally in the newer lathes, in automatic turning machines, and
turret lathes. They reach their highest developments in the latter.
Tool Angles, Rake.—The term " single-edged" includes some
dozens of ends and edges shaped differently, some being true cutting-tools,
others scrapers only. The essential difference between cutting-tools and
scrapers is that the first has top rake, the second has none, that is, in the
first the top face of the tool makes an angle of something less than 90° with
the surface of the work, if plane, or with its tangent if circular; while in the
second the angle is 90°.
The " tool angle ", the angle of " clearance ", and the angle of " top
rake ", are shown in figs, i and 2. The tool angle is a measure of the ability
of the tool to resist the pressure of the cut, and it is therefore maintained as
large as possible. The clearance of 6° (figs, i and 2) need not vary much,
since this clearance is provided merely to prevent friction and heating between
the tool and the surface of the work. It may range between 3° and 7°,
though many tools that are hand ground have a larger amount, by virtue of
which they cut more freely, but at some sacrifice of endurance. The angle
of top rake is varied with the material to be tooled in order to give a good
cutting action, and to permit the chips or the shavings to come away freely.
The tool angle ranges from 50° to 85°, both being exceptional. Keen
angles would give an easy cut, but the edge would not be permanent. Two
standard angles have emerged, roughly 70° for the softer steels, and 80° for
the harder steels and cast iron.
Figs, i and 2 show standard Sellers' tools. Two sets only of angles are
adopted (fig. i), the " blunt tools " for cast iron and the harder grades of
steel, and (fig. 2) the " sharp tools " for wrought iron and the softer grades
of steel. Both are made as right- and left-handed straight tools, or as right-



	PLAN           —


Fig. i.—Standard Round-nose Tool









	\        \u
	SIDE                       T+

	ELEV.                             <5


Fig. 2.—Standard Kound-nose Tool



Fig. 3.—Double-edge Roughing Tool

and left-handed bent tools.   The angles are stated on the drawings.    The
difference lies in the side top rake.

Side Top Rake.—When considering top rake it is necessary to bear
in mind the direction in which a tool is fed in relation to the work. If it is
traversed laterally, as in turning, then a straightforward nose with front rake
only is not the best possible, because the tool angle is not in the line of travel

A                                 and the lateral strain on the

j^                                 tool is increased.    But if

rake is provided in the
direction of travel—" side
top rake "—the tool can be
fed more easily, will cut
more freely, and the chips
will be deflected away from
the tool support. This
explains why the majority
of roughing-tools have side
top rake, and why, when a
straightforward tool is em-
ployed, it is generally set at an angle for traverse cuts, and why so many
tools with top rake are bent at the points right and left to correspond
with the direction of their traverse. The blunt tool in fig. i has 14°
of side top rake, and the sharp tool (fig. 2) has 22°.

Plan Outlines, Roughing and Finishing.—The curvature of the
nose of a cutting-tool is important. Figs, i and 2
show " round-nose " tools. These gouge-like tools
remove material with the maximum of effect. The
amount of convexity varies considerably, and generally
those with the longer radii are used for the heavier
duties. These are termed roughing-tools, notwith-
standing that they are often retained for finishing.
The distinction between tools for roughing and
finishing is not observed to the same extent as of old.
The spring-tool, so long a favourite with turners, is
obsolete. Tools with double edges (fig. 3), such as
are commonly used in lathes, rough with the leading
edge, while the small following radius leaves a smooth

Knife Tools.—The knife or shaving-tools are
employed extensively in turret lathes. They rough
and finish. They cut normally to the knife edge
(fig. 4), and remove broad shavings with fine feeds,
leaving a finished surface on the work. They are made straightforward,
left-handed, and cranked. A clearance of 6° and a side top rake of 12° is
suitable. Allied to the knife tools are the narrow parting-tools, used for
severing pieces of work. These are made straightforward, left-handed


A -



Fig. 4.—A Knife Tool


and cranked. They have no top rake, but only front and side clearances.
There is also a slight clearance from the front backwards.

Tools with Profiled Edges.—These form a very large group, used
only for finishing without traverse. They have no top rake, are used in-
differently for all materials, largely in the lathes, and only to a very limited
extent in reciprocating machines, since the work is done better with pro-
filed milling-cutters. A familiar form is the " vee " tool, employed in
cutting screw threads. As these have to traverse, 8° is a suitable side clear-
ance for them. They are often made right- and left-handed, with a larger
clearance on the leading edge. They are also straightforward and bent.


The Drills.—Few drills are used now except those of the twist design.
None of these are strictly standardized, except in the practice of individual
manufacturers. The true drills have two cutting lips
only—single edges in balance. Those with three or
with four lips link the drills with the reamers, and
are used for finishing holes.

Drill Angles.—Twist-drills are true cutting-
tools. The old flat drills were scrapes. -The straight-
fluted drills used for brass are scrapes. With these
exceptions drills are right-handed cutting-tools, only
a few for special purposes being made left-handed.
The helix angle—that imparted to the flutes (fig. 5, B),
corresponds with the top rake of the common single-
edged tools. An average is 25°, but in some designs
it is as high as 30°. In the " increase " twist-drill
the angle changes as the lips are ground, becoming
less acute. The exit of the cuttings is facilitated,
and the thickness of the web increases. The in-
creased thickness of web provides additional strength
to resist torsional stress. A slight disadvantage is
the reduction in the cutting angle. The increased
twist in drills standardized by different firms varies
(fig. 5) from 26° at A to 21° at B, and from 32°
at A to 27° at B, at extremes. The angle c of the
lips of drills varies only slightly, ranging from
58° to 61°. The usual angle is 59°. The clearance
angle or backing-off, D, varies from about 6° to 15°. Fig. 5.—Elements of Twist DHII
A usual amount is 12°, but it need not be so large.

It should increase from the periphery towards the centre. On this clear-
ance depends the angle which the " chisel edge" or drill point makes with
the flat portion of the flute, which is properly 135° (fig. 6, B). If this angle
is much larger, as at A, the point is too keen for endurance; if obtuse, as
at c, the edge will not cut but will rub only.

184                              THE   MACHINE-SHOP

The point of a drill must be exactly central with the shank, and the lips
of equal length and angle; otherwise the work will not be shared equally, nor
will the hole be true to size, nor can it be drilled at maximum speed.
Thinning the lips in the larger drills (fig. 7) contributes to efficiency, espe-
cially as the tools wear back. Longitudinal clearance is the slight reduction
in diameter from lips to shank, which enables the drill to clear itself in its
hole. It ranges from 0-00025 to 0-0015 *n- Per *nch 'm length. Peripheral
clearance (fig. 7) is that round the circumference of the drill, starting from
the " land " a (compare with fig. 5), which backs up the cutting edge and
preserves the diameter.

Speeds and Feeds.—The performances of twist-drills vary greatly,
the controlling conditions being the quality of the drill, the degree of accuracy
of the clearances, the care exercised in grinding, the nature and the amount

A                  • B                       C   '                                   't—'
Fig. 6.—Effect of Angle on Chisel Edge or Drill Point                          Fig. 7.—End View of Drill
of the lubricant used, the build of the drilling machine, and the mass of
the work being drilled. Published tables of performances afford but a
general guide, to be accepted with caution. These performances may be
exceeded by as much as 100 per cent in exceptionally favourable conditions.
Speeds are usually stated in terms of carbon-steel drills, to be doubled
when tools of high-speed steel are used. Peripheral speeds are stated per
minute. Average speeds are: for cast steel, 20 to 30 ft. per minute; tool
steel, 30 ft.; malleable cast iron, 45 ft.; cast iron, 40 to 50 ft.; brass and
bronze, from 60 to 200 ft. per minute. Feeds of from 0-004 to °'°O7 in-
per revolution are employed for ^-in. drills, increased to from 0-005 to
0-015 in. for those of larger sizes. Generally it is better to increase speeds
than feeds.
Lubrication.—The efficiency of a drill depends on proper lubrication
more than on any other factor. Generally soda water, soapy water, or
emulsion are used, cast iron and brass being the only substances which are
drilled dry. Many recipes exist for making up an efficient lubricant, and
results are so largely dependent on an abundant supply of the lubricant
being provided, that a good many drills, especially those used in turret
work, have oil passages through which the lubricant is forced under pres-
sure to the lips. Sometimes oil grooves are formed within the body of the
drill, a more satisfactory method than letting tubes into grooves cut around
the periphery between the flutes. Such grooves provide ample room for
the escape of the chips. Drills with internal tubes are fixed in a turret,
}                                                          THE  TOOLS                                     185
A;,                 and the work revolves.    A cup is screwed into the shank to receive the
connection from the oil supply, this connection being usually a flexible pipe.
I                        Twisted Twist-drills.—These are being used in increasing numbers
j                  in preference to those in which the flutes are cut by milling.   The demands
I                  of high-speed work are partly responsible for this design, which is a return
to the primitive twist-drills made by twisting a flat bar of steel.
|                        Drill Shanks.—These are standardized both for tapered and parallel
shanks. The tapered shank is used with drill sockets, and the second
when the drill is held in a chuck or in a turret. There are seven sizes of
Morse tapers. One size can only be used for a small range of drill diameters
differing by a few eighths of an inch. Adapter sleeves or sockets are then
employed, the first for shanks larger than a machine-spindle takes, the second
for those of smaller sizes.
Boring-tools.—Boring is distinguished from drilling not precisely
because bored holes are usually larger than those that are drilled, but
the term signifies the enlargement of a hole which has been already
" drilled ". Though drilling may be done up to 5 or 6 in., and boring
so small as 2 or 3 in. diameter, yet the latter operation is mostly associated
with holes that range, say, from about 3 in. to 20 or 30 ft.
Boring-cutters.—The single-edged lathe boring-tool is the type on
which all boring-cutters are designed. The single cutter is retained in many
cases for roughing. The lathe tool itself has but a limited use in the boring
i                   practice of to-day.   The solid shank of the tool is a cantilever that chatters
if it overhangs much, or if the pressure of the single cut is unbalanced.
1                   For long holes two or more cutters in balance are used, either inserted in
*                   slotted bars or carried in heads, which are either fixed or are fed along
their bars. Different shapes and cutting angles may be used for roughing
and for finishing, but frequently no difference is made. In minor details
the tools follow the usual practice in tool design which has already been
described. The boring-tool is a tool point of an expensive but hard
steel, which is gripped in a bar or holder of common material.
Cutters in Bars.—Only the smaller holes are bored with cutters that
fit in slots in bars. They are single or double, and are differently secured.
A wedge is common (fig. 8) but is liable to shift, so is a round tapered pin,
flattened on the side next the cutter (fig. 9). Neither alone would provide
for setting the cutter to exact radius, which must be done by gently tapping.
j     .             Many single- and double-ended cutters are therefore "self-centred" with
a notch fitting over the diameter of the bar (figs. 8, 9, 10), and then they
'                   cannot shift.   Single cutters are adjusted radially with light hammer-taps,
1                  and are then tightened with set-screws (figs, n and 14).   They may be set
with a grub-screw at the rear, and clamped with a set-screw (fig. 12).    A
very common method is that in fig. 13, where the head of a cheese-head screw
/                  entering a notch in the shank of the cutter adjusts it finely.   Another method
applied to double cutters is shown in fig. 15.   A grub-screw with a conical




Fig. 8.—Single-ended Self-centring Cutter
secured with Wedge. Longitudinal Sectional


\ —

	I •

— £2 ^C
	^ ----------
	— H — \\

	1    1


Fig. 9.—Double-ended Self-centring Cutters secured
with Tapered Pins

SECURING NUT X>--------------

LONGITUDINAL ELEV.                                                                       END   ELEV.

Fig. 10.—Double-ended Self-centring Cutters secured with Nuts


Fig. 11

(\       \OF CUTTEf*




Fig 12.—Single-ended Cutter adjusted
with Grub-screw



Fig. 13.—Single-ended Cutter adjusted
with Cheese-head Screw and Clamped
with Tapered Pin

Fig. 14
Figs. IT and 14.—Examples of Two Single-ended Cutters
boring at one time


point moves the two cutters
outwards simultaneously,
after which they are pinched
with the set - screws in-
serted from the front.
Variations made in these
elementary fastenings are

Pilots  are  employed   to
centre and steady the action         „.         ^  ui ^             j^-u^.i^j

•'     t                                 Fig. 15.—Double Cutters expanded with Conical-ended

of   cutters.     A   pilot   may                                 Gmb-screw

enter a bush in the table of

a machine or receive guidance from a hole already bored.   The method is

very common in turret lathe work.   Some preceding figures show multiple-



Fig. 16.—Round-nosed Cutter adjusted and
clamped in Boring Head

Richards' Cutters in Boring Heads

Fig. 18




	tl W   =
 *a 3?







	_J — „

cutting.    In fig. 9 shouldered cutters are seen, one for roughing the other
for finishing, each being secured with a tapered pin.   In fig. 10 two cutters
.                      are secured with circular nuts that bear against their

HI                    faces.    In fig. n provision is made for three cutters

/Tl                   in a bar, two slots being occupied; in fig. 14 a similar

provision is made. Counter-bores are tools that pro-
duce shouldered recesses in holes already bored.
They are centred and steadied with pilots, often
with provision for changing pilots.

Cutters in Heads.—These are either flat- or
round-nose tools (fig. 16) set out with a conical screw
and clamped. Figs. 17 and 18 show two of the best
methods. The tools being set diagonally cut sweetly.
A large range of adjustment is provided for, and the
set-screws clamp the cutters securely. The heads fit
easily on the bars over half the bore only, and are
held securely with set-screws.


Reamers.—A reamer is used to finish accurately
a hole previously drilled, since no drill leaves a hole
correct to fine limits or perfectly straight. Though
the reamer removes an exceedingly minute amount,
two passages with tools of different sizes are often
necessary for the finest tolerances. The reamer has
many cutting-blades which counterbalance each other.
The old D-bit and the rose reamer (fig. 19) cut by
their leading edges; present-day reamers cut with the
whole length of their blades. The ends are slightly
tapered to enable them to enter easily. The blades
are often spaced irregularly in order to prevent chatter
and risk of " cornering ", due to the fact that if blades
are pitched equally they come round to exactly the
same place in each revolution, so that any initial in-
accuracy will be perpetuated. But the evil is lessened
by imparting a small amount of clearance. Blades are
straight, or spiral; in the latter the spiral should run
contrary to the cutting edge in order to avoid the
tendency of the reamer to " draw " into the hole.
Reamers are either solid with shanks, or are shells.
They are parallel, or tapered. They are made with
blades solid, or adjustable (fig. 19).

Clearances.—Though a reamer is a scraping tool—the faces of the
teeth being disposed radially — it will not operate well unless suitable
clearances are provided. A very slight longitudinal clearance is necessary



Fig. 20.—Tapered Roughung Rtamer
Left hand not to pull in.   Serrated spirally to break up chips.

as in the twist-drills, the tool tapering towards the shank. This prevents
the rear end from rubbing in the hole. End clearance on the lips of the teeth
enables the tool to start the cut sweetly. The side or radial clearance pro-
duces a smooth and true surface. Without this the edges would rub hard
and not cut at all, and the hole would not be true. Generally the radial
clearance is a straight face, lying at an angle greater than that of the actual
cutting edge, which is very narrow, like the " land " on a drill. The edge
so formed lasts longer than it would if it were left keen.
Flutes.—The sectional forms of flutes vary. They may be straight,
concave, or convex, the first being most common as it is more readily re-
ground than the others. The flutes of tapered reamers are straight, or spiral
in the longitudinal direction. When used for roughing, the flutes are either
notched or they have a spiral groove running all round the teeth to break
up the chips (fig. 20). Some of the chucking reamers have straight flutes,
while a good many have three-grooved spirals with oil grooves for the passage
of the lubricant. All the solid reamers have shanks either parallel or tapered
to standards. Shell reamers
fit on arbors, and are only
used for the larger holes.
Floating Reamers. —
These are used in some of
the finest operations. They
accommodate themselves to
the holes which they finish. They may float perpendicularly and at an
angle. They are employed extensively in turret work, for which special
holders are provided.
Adjustable Reamers.—These are in some- degree a result of the
growth of the limit system of gauging, in which minute differences in the
diameters of holes for tight, push, and easy fits have to be made. If solid
reamers are made to deal with certain sizes of holes, they lose their dimensions
rapidly with regrinding. There are many differences in the details of
fitting and adjusting the blades in these tools. They may be classified as
1.  Reamers having a solid body with splits, to be expanded by an
internal tapered plug (fig. 19), which is either drawn or thrust inwards with
a screw or driven with a hammer.    Only a very slight amount of expansion
is obtainable with these, but they are suitable for jobs where only fine
cuts are required with little variation in size.     They are used extensively
on turret lathes.
2.  In this group, loose blades are fitted in recesses in the body, and
expanded by the insertion of packing strips beneath them.   The one ad-
vantage of this design is that the blades bed solidly on the packing, and that
packings of increased thickness can be substituted as the blades become
worn.   They are also cheap, having few fittings, and they cannot readily be
tampered with.   Tin-foil is used for packing, the thinnest strips of which
measure 0-0005 in. thick.   These designs are used in turret lathes.
igo                              THE  MACHINE-SHOP
3.  The blades are fitted in recesses, and are expanded with wedges driven
beneath them.
4.  The blades fit in inclined slots, and are expanded by driving them
inwards towards the higher ends, with or without using locking nuts for
their retention.   The blades are ground in place while the body is mounted
on centres.
5.  In this group the blades are fitted in inclined slots, and are moved
up with nuts coned on the inside to retain the blades, with or without lock
nuts.   This design is much to be preferred to the last, because the movements
imparted to the blades are simultaneous and more precise, and regrinding
is not necessary after the setting.   Many of the best reamers are made in
this way.
6.  Blades in slots rest upon a central tapered plug or " cone bolt ",
which, being forced inwards, expands all the blades equally.   The locking
is effected with nuts.   In the " Vickers " design the expansion is imparted
without longitudinal movement of the blades.    In a sub-group the blades
are expanded with two cones, reversed, which are drawn towards each other.
A large range of diameters can be obtained with these.
7.  In some designs an eccentric or cam bolt has a series of cams like very
shallow ratchet teeth, which by their partial rotation cause the blades to move
Milling-cutters.—These have gone through a larger evolutionary
growth than any other single group of cutting-tools. They range from J in.
diameter to several feet; include true cutting as well as scraping teeth; can
be used to rough and finish work; and produce not only plane surfaces but
combinations of horizontal and vertical faces, and curved and irregular
Teeth, Speeds, Feeds.—Milling-cutters have very little in common
with the single-edged cutting-tools, since their teeth operate in quick suc-
cession over broad surfaces. In the edge mills taking deep cuts the angles
of presentation will change, and the teeth will rub on the leaving edge. Also
the chips will become entangled between the teeth and cause friction. The
teeth of all the early cutters were pitched too finely to permit of their use as
roughing-tools. Coarser pitches are imparted now, and roughing-cutters
may be had, but generally the same cutter is employed for both functions.
For roughing, the teeth are often notched to break up the chips, and all
except the narrowest cutters have spiral teeth which effect a gradual cut.
A cutter must not be run at a high speed, since its teeth would become choked
with chips, but the feed should be coarse. Feeds have been increased
amazingly. They are stated in terms of advance in inches per minute, or
in fractions of an inch per revolution of the cutter. A more practical test is
the number of cubic inches of material removed per minute. The end
mills are more efficient as roughing-tools than the edge cutters are, since
there is no change in cutting angles and the chips get away freely. As a


rule the teeth have no front rake, but clearance only, generally with two

or three facets, one being the " land " for grinding.    Formed teeth are

numerous, in which the backs of the teeth are struck from a centre eccentric

in   relation   to   the

cutter centre. These                          '"*

are reground on the

front faces only, and

retain their sectional

shapes   until   worn

thin.   These belong

to the profile group,

used for gear teeth

and allied shapes.

The Forms of
Cutters. — These
include edge, side,
end, and formed
cutters, both angular
and curved, solid
tools and those with
inserted teeth.

Edge and Side
Mills. — In these
(figs. 21, 22) the

teeth are cut on the peripheries, and on one or both ends respectively.
When cut on both ends they are used for slitting and grooving.    The



Fig. 21.—Two Interlocking Slotting Cutters clutched so that Teeth overlap
and Width can be preserved after Wear



Fig. 22.—Interlocking Cutters in which each Alternate Tooth interlocks
cutters, in figs. 21, 22, have provisions for preserving the width. Edge
mills when over i in. in diameter are provided with spiral teeth, usually
at an angle of 10°, except in some cutters for roughing, in which this angle


is considerably exceeded. Since, as a rule, cutters are used for general
service, the speeds, depths of cut, and feeds are varied to enable them to
work with efficiency on all materials. Only in the most general terms can
these be hinted at. Soft steel can be cut at peripheral speeds of from
90 to 150 ft. per minute; the harder steels from 65 to 75 ft.; cast iron,
Ho to 100 ft.; the brasses and bronzes up to 1000 ft. per minute. Depths
of cut may range from iV, to J in. in one traverse. Feeds, formerly so
low as 2 to 3 in. linear feed per minute, are frequently now from 12 to



, a5,--Two Cuftefi in €**ng, with Diiunce-piece

20 in. The metal removed in a minute with a cutter 8 in, wide working
on mild has amounted to 23 c. in., and on cast iron to 48 c, in.
End Mills*—In          (fig. 23) the end teeth cut, and those on the side
smooth the surfaces. The teeth are straight for working on brass, for other
materials they are spiral This provides a cutting rake, or, when the spirals
arc* left-hand, the rake is negative with the tendency to hold the tool hack
in its spindle. These tools only cut on the Inner ends of the teeth in the
11 centre-cut mills **, which have teeth on the inside, so that in these the tool
can be vertically to the required depth, and then traversed. End mills
are provided with               taper             or they are          fitting on arbors.
Form Cutters.—These include               angular              for cutting
and veefcl             the tee-slot               angular cutters for produc-
ing the           of            the very                             employed for grooving
a           groupy employed for fluting the drills.,                and




















g ----- 1








Fig. 26.—Examples of Milling Operations




milling-cutters, and typical of those used for gear teeth. A concave cutter is
used for producing beads, and combinations of these and similar outlines
are often made for special work.

Gang Cutters.—These, of which figs. 24, 25, 26, 27 are typical, are

Fig. 27.—Four Cutters in Gang, with Distance-pieces

used extensively on all the horizontal milling machines, but mostly on the
piano-millers. Single cutters are built up to suit requirements. Fig. 24
shows three for cutting faces and edges simultaneously. These are fre-
quently interlocked, with provisions for taking up lateral wear. Fig. 25

illustrates two, fig.
27 four on an
arbor, with sepa-
rating distance-
pieces. The group
(fig. 26) shows vari-
ous operations. At
A an edge mill is
tooling the bottom
of a bed. At B
gang mills are tool-
ing faces and edges,
c, D, E show three
sets of operations
on a bed. At F
and G cutters are
H shows the milling of a






Fig. 28.—Inserted Cutters, set spirally, held with Pins flattened to bear
against Cutters

at work on the bottoms and faces of bearings,
long strip on five faces.
Inserted-tooth Cutters.—When a cutter exceeds a few inches in
diameter it cannot be hardened and tempered like the smaller tools. Inserted
teeth of high-speed steel are fastened in bodies of cast iron or of mild steel.
Cutting points are often identical in shape with the single-edged tools,
are set straight-faced or spirally, and are fastened in many ways,
shows spiral blades held with flattened pins. Side cutters are located





with shoulders and held similarly. Roughing-cutters in fig. 29 are each
adjusted with a grub-screw and locked with a nut. Cutters are set spirally,
and tightened with tapered pins in splits. Wedge bushings and screws are

FJg. 29.—" Wrigley " High-speed Cutters for Aluminium, adjusted with Screw and locked with Nut

used for tightening,
feet in diameter.

Many of the heads with inserted cutters are several


Grinding Wheels.—The old term " emery " wheels applied to this
group has long been abandoned, since emery is employed to a limited
and ever-lessening extent, having been replaced by more effective grinding

Emery and Corundum.—The difference between these is one of
purity. Alumina is the chief constituent of each. Corundum contains a
higher proportion of alumina than emery, and its grains split, leaving sharp
edges; while emery wears smoothly, with a glazed surface. Both materials
are impregnated with oxide of iron, which, when present in large quantities,
reduces the cutting capacity. On the other hand, emery wheels produce a
high finish.

Carbide  of  Silicon Abrasives.—These are prepared in electric



furnaces from coke and sand.     These  abrasives  include carborundum,
crystolon, carbolite, corbolon, carbowalt, and corex.

Aluminous Abrasives.—These are prepared in electric arc furnaces
from bauxite, a clay that contains a high percentage of aluminium oxide.
It is a soft light-yellow earth, and is the purest form of aluminium oxide
found. Only in the electric furnace can the nearly pure alundum be separated
from the foreign matters present in the earth. The abrasives obtained in

this way are: alundum, alowalt, aloxite,
borocarbone, carbo-alumina, corowalt,
oxaluma, and rex.

Applications. — Although several
of these abrasives are employed for
similar purposes, yet some are more
suited to certain duties than others.
Broadly, the wheels used for materials
of low tensile strength, such as cast
iron, brass, and aluminium, are not em-
ployed for the steels which have high
tensile strength. In general a carbide
of silicon abrasive is used for the first,
and an aluminium oxide abrasive for
the second,

Grain or Grit. — The number
that designates the grain signifies the
number of meshes to the linear inch in
the grating forming the bottom of a
sieve, through which the grains will pass.
The numbers in common use range from
about 20 to 60. Usually all the grains
in a wheel are of the same size, but
" combination " wheels are used, with
the object of enabling them to cut fast
and finish smoothly, and so avoid a
finish grinding with a second wheel.
Grade or Bond.—The efficiency
of a grinding wheel for a definite duty
depends on what kind of material is
employed to cement the grains together. Wheels are " hard" when the
grains are not easily dislodged from their matrix, "soft" when they are
readily torn out. But the size of the grains has a modifying influence,
since a wheel with the same bond is harder if the grains are fine than if
they are coarse. Generally a harder wheel will be used on soft steel than
on the same steel if hardened. The harder the material is, the softer the
wheel should be. The reason is that a hard material will blunt the grains
more quickly than a soft one, and therefore they should be torn out more
rapidly to allow fresh grains to come into action. An exception occurs in



Fig, 30.—Edge Grinding Wheel with Bevelled
Safety Flanges


the brasses, which require a soft wheel in order to prevent clogging or
glazing of the wheel with particles of metal.

Bonds.—The three bonds commonly employed in the order of their
importance are: the vitrified, the silicate, and the elastic. The first is
composed of clays, properly a pure grade of kaolin. The wheels are
moulded and subjected to a prolonged heat to partially fuse the bond. The
wheels are of a reddish-brown colour, are very porous and free-cutting,
and are not affected by water, oils,
or temperature, and the bond is hard.


Fig. 31. — Edge Wheel mounted per-
manently on Flanges for accurate Replace-
ment and Wheel-changing on Mandrel

Fig. 32.—Cup Wheel Mounting, with Safety

But the risks of cracking do not permit of making these wheels beyond about
30 in. diameter. These are suitable for general grinding. For the silicate
bond the silicate of soda is chiefly used. The process is less prolonged than
that for vitrified wheels, and larger sizes can be manufactured. These are
not used much for cylindrical grinding; their function is that of wet grinding
of tools. The elastic wheels are mostly bonded with shellac. Vulcanite
wheels are bonded with vulcanized rubber. Both can be made very thin,
and be run in water. Vulcanite wheels can be used with oil or caustic
soda; elastic wheels cannot. These are made thin for cutting off materials,
for grinding saws, and sharpening cutters.
Wheel Shapes and Mountings.—Fig 30 shows an edge wheel, used
for cylindrical grinding, with one method of mounting.   The flanges are
ig8                           THE  MACHINE-SHOP

dished to suit the section of the wheel, and they only bear against it with
annular seatings, which do not tend to crush the wheel, and, if it should
fracture, the pieces are prevented from flying off. Wheels of parallel thickness
are also gripped with annular contact. Another essential is that the wheels
fit loosely on their arbors and tight only in the flanges, to avoid risk of their
being burst. Fig. 31 illustrates a wheel gripped with washers of leather,
rubber, or cardboard. But the principal feature is that the wheel is mounted
permanently with a screwed flange, to be removed from and replaced bodily
on the tapered end of its arbor, where it is held with a circular nut. Fig. 32
shows a face wheel. The mounting includes an encircling safety ring,
which is set back as the wheel wears. In the Blanchard wheel, used on
the firm's vertical-spindle machines, the principal feature is the provision
of holes in the flange to direct water to the face of the wheel.

The Essentials of Economical Machining
Lubrication.—The efficiency of cutting-tools depends on the lubri-
cation and the cooling of the tool point, and of the surface of the work being
cut. Cast iron and brass are usually excepted. Formerly the chief attention
was directed to the cooling of the tool; now the view-point is changed,
consequent on the increased severity of cutting, with the more rapid
generation of heat. Instead of the drip-can, the cooling liquid is delivered
in a stream, frequently under pressure, and directed with pipe nozzles or
spreaders all over the surfaces being cut.
Cooling Fluids.—With these changed views the practice has undergone
great changes. Special lubricants are now used for certain classes of heavy,
medium, and light work. As of old the best all-round lubricant is lard
oil, but the high cost of it handicaps its general use. The best substi-
tutes contain a mineral oil with a certain quantity of lard, and are termed
" mineral lard oils ". The proportions of lard are varied for different
kinds of work. Soda or potash mixed with a mineral lard oil forms soap.
The soap holds the oil in suspension, and prevents it from floating on the
Distribution and Recovery.—Instead of the drip-can a system of
supply pipes is laid down in modern shops, and each machine is provided
with its own particular equipment for distribution through jets or nozzles,
with means for the collection and return of the liquid. In some cases a
gravity supply is installed. A feed tank is placed in the roof, and the machines
drain to a sump below the floor. More often now the cooling fluid is
delivered by means of a pump. In a few shops different groups of




machines are provided with special supplies of cooling fluid particularly
suited to the kind of work done in them.

The floods of lubricant supplied are mostly recovered. For the collection
of liquids at the machines, tanks or trays are now fitted. In these the lubri-
cant is drained through a grating, leaving the chips and dirt behind. At


	i !   £

	1 !
 IL.  A

	r .
	i. ii


	•*"•*• »*' ---------------- ***-#
	JT- -7 '•-"•;,'

Details of Filter for Lubricating Oil




Fig. 34.—Enlarged Details of Oil Piping System
intervals the liquid is drawn off to be treated in centrifugal machines, or in
filters. Fig. 33 is an illustration of the arrangement of oil piping and a
filtering system in the American Tool Works, as laid down by the Richardson-
Phenix Company, of Milwaukee, Wis., U.S.A. The full lines show the supply
pipes for clean oil; the dotted lines, the return drain pipes. The first,
starting from the centrifugal pumps, A, diminish from zj-in. bore to f-in.
bore at their terminations. The drain pipes increase from i-in. to 3-in. bore
where they terminate at the filter B, which is shown in detail in fig. 34.
The drain line flushing arrangement is shown at c, fig. 34, the object
of which is to prevent sediment from accumulating and impeding the free
movement of the liquid. The oil is sterilized in the vessel D before being
filtered. The machines are lettered as follows: fig. 33, E indicates flat urret
lathes of various sizes; F, lathes; G is a centring-machine; H, one for testing;
J, one for cutting off stock; K, a Jones & Lamson lathe; L, various automatics;
M, is a stock rack.
Speeds and Feeds.—The speed of a cutting-tool, relatively to that
of the piece of work on which it operates, irrespective of whether the tool or
the work moves, is expressed by the number of linear or peripheral feet
passed through per minute by the tool or the work. The feed in turning
and planing is the lateral distance traversed between each cut; in drills and
face-milling cutters it is the depth of penetration estimated in some minute
fractional part of an inch per revolution of the tool; in edge-milling cutters
it is stated usually as the linear distance travelled by the work under the
cutter per minute; in grinding wheels it is the depth of cut given by each
setting-in of the wheel.
There are standard speeds memorized in the shops, just as there are
standard tool angles for different materials. But they are more honoured
in the breach than in the observance, and are exceeded in favourable con-
ditions. There are no commonly recognized feeds. But, with the increasing
stiffness of machine-tools and with improved lubrication and suitable tool
angles, feeds are generally very much coarser than of old, notably in high-
speed turning, in drilling, and in milling.
Relations of Speeds and Feeds.—There is no hard-and-fast rule as
to whether high speeds and fine feeds, or low speeds and coarse feeds are
preferable. In drills, for instance, it is more economical to increase speed
than feed. In edge-milling cutters the best results are secured by low
speed with coarse feed. In turning and planing, high speeds and coarse
feeds may go on simultaneously. The old speeds for carbon tools were:
cast iron, from 15 to 20 ft. per minute; steels, from 15 to 30 ft.; wrought
iron, from 25 to 40 ft.; and brass from 50 to 100 ft. These are now
generally exceeded, except in the harder qualities, and tools of high-speed
steel will cut at double these rates. But any general statements can only be
approximate, since results are controlled by many variables, as tool angles,
depths of cut, rate of feed, grade of material, the rigidity of the machine, and
the volume of lubrication—often the largest factor of all. Because of these
facts, no ratios of speeds and feeds could be tabulated that would be of any
general value.
Depth of Cut.—This may range from o-ooi in. in grinding wheels to
i in. in cutting took. An increase in depth of cut involves a reduction in
cutting speed and feed, because the capacity of a tool is measured by the
area of cut plus the feed. Heavy cuts at slow speeds are more economical
than light cuts at high speeds. But the horse-power required is greater,


which explains why machine-tools at the present time take much more
power to drive than their immediate predecessors did. The weight of
material removed in a given time is the real test of the cutting capacity of
a tool. The endurance of a tool is the true measure of its efficiency, since
one that has to be reground at short intervals is not economical. A single-
edged tool should endure for at least an hour, while milling-cutters and
those set up in boxes for turret-lathe work should last for a day or more.
The longer the time occupied in regrinding and in resetting, the stronger
is the reason for maintaining the endurance of the edges.
Setting and Securing Work.—Broadly there are two methods
employed for holding articles to be tooled. In one the piece is gripped
either directly on the work table of the machine or in a chuck, or on an
arbor, with the help of appliances that are in common use for a multitude of
various jobs. Here in general the
pieces are set and held singly even
though many are identical in shape
and dimensions. In the other
method they are not attached
direetly to the table or other work-
holding element, but to an inter-
cnoss SECTION                        mediate appliance, the fixture, or
Fig. 3S-—Clamping Plates set Diagonally                £O special adaptations of chucks Or
arbors. The first is the older
practice, necessarily retained for all classes of work that are not highly
repetitive. The second is the later method, essential to and inseparable
from mass production, and an interchangeable system.
Work held on Tables.—This chiefly concerns the planer, shaper, and
slotter groups, and the drills, boring machines, and allied forms. The
feature common to all is the level table, provided with grooves of tee-section
to receive the bolt heads for clamping work. The grooves are also used to
hold stops, angle plates, vee blocks, and so on.
The surface of the table provides the accurate datum for ensuring that
the clamped work will occupy its correct relation to the cutting-tool. Hence
the first care is to get the work to bed truly on the table. When practicable
it is well to take a rough cut off one surface of the work in order to secure
contact. If this cannot be done, then a rough surface must be packed care-
fully with wedges where it is out of contact with the table, or the clamping
bolts will pull and spring the work, and it will not be true when the machining
is done and the pressure of the bolts released. Though the effect is more
pronounced in thin pieces, it is present in all except the most massive chunks.
Hence, a safe rule is never to tighten a bolt except in opposition to a machined
surface, or, with a rough surface, near a packing.
Thin, and Substantial Articles.—When dealing with very flimsy

pieces, it is not permissible to clamp directly on their upper surfaces.
Lateral pressure is adopted in such cases, and also for those where the
upper surfaces have to be machined all over. The clamps are better if set
diagonally (fig. 35) to exer-
cise a downward pressure.
Often it is necessary to
set a stop against one end
of the work in the line of
direction of the cutting-tool
in order to prevent the occur-
rence of slip by reason of
the pressure in the longi-
tudinal direction. For hold-
ing substantial pieces, direct
clamping is adopted. But,
since tall bolts are apt to be
unstable under the stress of
heavy cutting, advantage is
taken of the presence of
suitable lower sections on
which to bed the clamps.
These may be stout flanges,
bosses, bores, or recessed portions, to be utilized by the judgment of the

Cylindrical and Bored Work.—This is located and held in vee
blocks, which ensure parallelism of the work with the table.   Parallel shafts


Fig. 36.—Large Vee Block and Vee'd Clamp

Sectional Elev
Fig. 37.-—Vee Blocks for Two Shafts, Clamp Plates adjusted with Grub-screws and thrown off with Springs

S,   Throw-offSprings
Gt Grub Screws for Adjust-
ment of Clamfs

and tubular portions of castings when located in vees of equal heights will
lie parallel with the table. If shafts are put through holes, and laid in
vees, the holes will be parallel. The clamps are variously set, according
to outlines, dimensions, and the avoidance of portions to be machined.


Fig. 38.—U- or Hair-pin Clamp

In most cases the clamps are vee'd in the grip (fig. 36), to ensure a better
hold and to shorten the length of the upstanding bolts.

In fig. 37 a vee block is made specially to hold two shafts, to be key-
grooved. The outer clamp plates are made to grip the shafts by the
tightening of their grub-screws. A useful provision is included, that of
the insertion of coiled springs surrounding the bolts, which throw the clamps
clear when the grip is slackened.

Clamping Plates and Packing.—
In general these are distinct and sepa-
rate, the packing being of wood or metal,
cut or selected to suit the height of the
clamp plates, the latter being kept strictly
horizontal. But as work becomes more
repetitive the packing is included with

the plate to avoid the loss of time involved in handling loose pieces.
Clamps have the form of plates, with slotted holes for bolts, or they are
of U shape (fig. 38), which affords a larger range of longitudinal adjust-
ment. They are single, or double, the latter to grip adjacent pieces, in
which case packing is not required (fig. 37). In this figure the grub-
screws fulfil the function of packing. Small screw jacks often serve as
adjustable packing. Another group comprises stepped blocks (fig. 39)

to give a range of heights.

Intermediate Attach -
ments.—Many articles have to
be fastened to an angle plate
instead of directly on the table.
This occurs when a piece must
have faces machined at right
angles, and when the shape is
such that it cannot be held on

the table without involving awkward packing-up. Faces that occur at other
than right angles are dealt with on tilting or swivelling angle plates.
Articles of another kind have to be held on machine centres, carried on
a machine table. These are used when machining has to be done in angular
relations, such as the machined splined grooves in shafts, and in the
drilling of holes from various angles. The work is carried on the centres
directly, or on an arbor, and the angular positions are set with pins in
holes, or latches in recesses, or by means of a circle divided into degrees.
The machine vice is admirably suited for holding small articles. It is used
chiefly on the shaper, the milling, and drilling machines, and occurs in
many forms, to hold parallel or bevelled pieces, to be machined in parallel or
angular relations.

£« 39-—Stepped Packing

Jigs and Fixtures.—A jig is an appliance that guides and controls
the location of a tool relatively to the work. A fixture is one that locates and
secures the work being tooled. The one may be used without the other.
If both are employed, they may be entirely separate and distinct, or be com-
bined in one jig-fixture. Both are intimately associated with the standardiza-
tion and interchangeability of the parts of machines and mechanisms. They
eliminate the need for the tedious, separate lining-off and setting of single
pieces, and they lessen the errors that occur in machining.
The Jig.—The original of the present-day jigs in their myriad forms
was the drilling templet. The majority of jigs are employed still in the
work of drilling and boring. In these the bushes are the vital elements,
because on their accuracy the correctness of results depends. They are
made of steel, hardened and ground, and provision is made for their ready
renewal when they become worn by the friction of drills, reamers, and
The simplest bushes are those which are a press fit in the jig. These are
only removed when worn out. A better and more accurate method is to
have a permanent lining bush to receive a removable one, the two being
fitted by grinding. It is well to make bushes with a collar, to prevent them
being pushed down too far in their holes. The edge of the bore where the
drill enters is slightly convex. Bushes are sometimes screwed in where they
must come into contact with the work. Lockings are employed to prevent
bushes from turning. A set-screw or a button is fitted to a slot in the collar,
or flats are made on collars of adjacent bushes. A bush may contain two or
more holes in close proximity. A simple bush is slightly longer than its bore.
A small bush will be of greater length, relatively, than one of large diameter.
All dimensions are usually standardized in shops where the system is a per-
manent one, and each size of bush has its own reference letter or figure.
Fixtures.—The employment of fixtures is the only alternative to the
practice of bolting articles directly to the tables of machine-tools. This is
a tedious process in the case of those of awkward shapes that require
packing, and have to be set by careful measurement. This often occupies
more time than the actual machining does, and distortion is liable to occur.
From the point of view of interchangeability it is hardly possible to set two
pieces precisely alike. The fixture is designed both to locate and to hold
the article, or often several, in the same exact position, so that each article
will be machined in the same way.
In good designs provisions are made to lessen the time occupied in
setting and in holding to a minimum. Often, as a result of high economies,
it becomes necessary to duplicate fixtures. One is unloaded and reloaded
while the other is on the machine-tool.
Jig-fixtures.—The highest developments are reached when the
fixture and jig are combined. The jig is generally hinged in some way to


Fig. 40.—Plunger for Sensitive Feed Lever







11   I L

I ;     r
f ? -I. .1



r     i


Fig. 41.—Open Fixture for holding Six of the Rodt
the fixture* to be thrown back during the removal and insertion of work, or
it may be merely lifted off like a cover. Hook bolts or swinging clamp plates
the two*   The fixture may be used for different machines on which

different kinds of operations are performed, or the jig plates may be changed
for the operations of drilling or milling. The fixture may be rigid, or it
may be made movable so as to present different faces of the work to the




Fig. 42 .—Fixture for holding Feed-pump Check Valve

tools.   Stop pins and locking devices provide for precise settings on one or
on several pieces of work held in tandem.
Example of an Open Fixture.—Fig. 40 illustrates a plunger for a
sensitive feed lever, and fig. 41 an open fixture for holding six levers while
their ends are being milled to bevels. This is by Messrs. James Archdale
& Co., Ltd., Birmingham. The levers rest in vees at one end, in which they



. x  X




are clamped in pairs, and lie parallel in recesses at the opposite end, where
they are brought up against an abutment piece. The ends to be bevelled
project beyond the vees, where they are milled with a cutter, divided, in order

to permit of readjustment
with packing as the edges

Example of a Box Jig-
fixture.—Figs. 42 and 43
give the principal elements
of a fixture by Messrs.
Ruston & Hornsby, Ltd., of
Lincoln, used in machining
the body of a check valve.
Its characteristic feature is
the provision made for drill-
ing a large number of holes
at different angles. The
valve, enclosed in the fix-
ture, is located by a flange
which enters the shallow
recess in the bottom, and
is secured by the jig cover.

Fig. 43.—Drilling Bush set at Angles in a Locating Block           The   inside   of the   COVCr  IS

recessed to receive a flange

on the opposite end of the valve body. The four bushed holes arranged
in a circle in the bottom and in the cover guide the drills for the bolt
holes in the flanges. Two other holes are drilled at angles through the
bushes A and B, to permit of which the fixture has bevelled feet at a and b.

During the vertical drill-
ing, bevelled packing
pieces are inserted under
these edges. The round
hole c receives bushes
that interchange for drill-
ing and tapping holes.
The oblong recess D re-
ceives the locating block,

Fig. 44.—Hinged Cover for a Fixture containing Four Drilling             ,            ,         ,      ,  ,£         >.

Bushes at different Heights                                  Shown detached (tig. 43 ),

that   carries   a   drilling

bush at angles in two directions. The fixture is then stood at the required
angles on a bevelled support. The holes E and F receive bushes to guide
drills for holes in the body.

Fig. 44 is the cover of a fixture through which holes are drilled at different
heights in the body of a feed pump. It is hinged at the left, and clamped
at the right with an eye-bolt in an open slot. Nuts are not run off their
bolts, only slackened.



Measurement and Gauging.—The present system of measurement
is precise and positive, and is effected rapidly. The micrometer and vernier
tools are used for taking precise measurements, and the fixed gauges check
machined dimensions to predetermined limits.

Micrometer Calipers.—In micrometric measurement the pitch of
a fine screw thread is subdivided by means of graduations on the periphery
of a disk which revolves with it. In the English caliper (fig. 45) the screw
usually has 40 threads to
the inch, and the " thimble "
—the rotating element—
has 25 divisions. Since a
movement of the screw
through one revolution cor-
responds with a longitudinal
movement of •£& in., one
partial turn of the thimble
through one division moves the screw through -£§• of -^V in. = TUTTF in. To
enable the exact longitudinal movement of the screw to be read, the
barrel or " sleeve "—the cylindrical body—is divided in a line parallel with
the axis of the screw into 40 parts, but only every fourth division is stamped
i, 2, 3, &c., from zero, corresponding with o-i in., 0-2 in., 0-3 in., &c. Each
of these subdivisions thus represents 25 thousandths of an inch. To read
the caliper, therefore, multiply the number of divisions visible on the scale

Fig. 45.—Micrometer Caliper

Fig. 46. — Vernier Caliper

on the barrel by 25, and add the number of divisions on the scale of the
thimble reckoning from zero.
Vernier Calipers. — A vernier is fitted to instruments made for making
the finest measurements. An inch is usually divided (fig. 46) into tenths,
and a vernier, of length equal to nine of the divisions, is divided again into
ten parts. Each subdivision on the vernier is therefore jku in. shorter than
one division on the rule. When thousandths have to be read, each tenth
divisipn on the rule is subdivided into four, giving forty to the inch. Twenty-
four of these parts are taken on the vernier and subdivided into twenty-five
VOL. I.                                                                                                                              14
2io                          THE  MACHINE-SHOP
parts. Each subdivision on the vernier is thus shorter than those on the
rule by TWIT in. Hence the rule: Note how many inches, tenths, and parts
of tenths the zero point on the vernier has been moved from the zero on the
rule. Count upon the vernier the number of divisions, until one is found
which coincides with one on the rule. This division will correspond to the
number of thousandths to be added to the distance read off on the rule.
Fixed Gauges.—Some fits in mechanisms must be tight and others
easy. Differences made in dimensions for the various kinds of fits are termed
" allowances ". The very minute variations that are permissible are called
" tolerances ". The term " limits " includes allowances and tolerances, and
gives the name to the " limit " gauges, which are generally guaranteed to be
correct within o-oooi in. Usually the hole is taken as the basis for measure-
ment, and the allowance is made on the shaft, but this is not invariable.
For many years after the introduction of gauges, the Whitworth cylindrical
forms only were used—the " plug " and the " ring ".    No attempt was
,                                 made at first to include limits.    The plugs fitted their rings exactly on the
f                                 application of the merest film of oil with the finger.    Tight and easy fits
!                                 were made by the exercise of judgment.    These have largely given place,
!                                 except for tapers, to the flat " snap " gauges, partly because these show a
dimension more finely than the others, and also because they can be used
k                                 on pieces that are not cylindrical.    In some forms there is a gauge, fixed by
\                                 two opposing jaws, at one end of the instrument that should pass over the
{                                work, and at the other a pair of jaws that must not.    They are called " go "
I                                and " not-go " gauges.    In large gauges the instruments are separate or
f                                combined.
|                                      The Johansson System.—In this system, end measuring blocks of
t                                rectangular  shapes   are   employed.     A   set  comprises   eighty-one  blocks
'                                divided into four series.   The first ranges from o-iooi to 0-1009 in. by
I                                increments of o-oooi in., the second from o-ioi to 0-149 *n- ^7 °'OOi in.,
f                               the third from 0-050 to 0-950 in. by 0-05 in., the fourth measure i in.,
^                               2 in., 3 in., and 4 in.    The blocks in the first series will divide up the spaces
f                               between those of the second series, and series three and four can be divided
\                               by the first and second series.    By means of combinations of the eighty-one
I                               gauges, 80,000 different sizes can be obtained.   These combinations are of
I                               much value in providing a ready method of checking the accuracy of a number
?                               of fractional dimensions.    They are used both for checking work directly,
I                              and for testing other measuring instruments, as calipers, limit gauges, measur-
I                              ing rods, jig parts, &c.    Various holders are provided.   The most remarkable
feature of these gauges is that the blocks adhere to each other by reason of
the fine accuracy of their surfaces.







The Work of the


The Lathes. —These in-
clude some forty to fifty groups,
each having well-defined spheres
of operation. They range in size
from very small to mammoth
dimensions, while extreme ma-
chines have little in common
except the fact that the work
revolves between centres or in
chucks. The prototype of most
of these is the standard, " self-
acting, sliding, surfacing, and
screw-cutting lathe "—the all-
round machine-tool, the econ-
omic value of which gets less
and less as specialized manufac-
ture increases.

Short screws, and those of
which large quantities are re-
quired, are now manufactured
on turret lathes, screwing
machines, and brass - finishers'
lathes. The longer screws and
stays are still made in screw-
cutting lathes.

The later lathes, fig. 47
being an example, nearly all
differ from the earlier in the
provisions made for speed and
feed changes. Stepped belt cones
are now almost entirely super-
seded by all-geared heads. When
cones are retained, with back
gears, speeds are arranged in
carefully chosen ratios instead of
in a haphazard way. All-geared
heads, being driven from a single



7     8

TI1H   M

pulley running at a constant speed, can he driven equally well by a belt
drive from a countershaft or by a motor. Feeds arc seldom taken now
from a '* hack-shaft ", hut from a " feed-shaft " in front of the lathe.

Another innovation the hollow spindle-•• allows stock bars to he passed
through iron) the rear of the headstock to be* gripped in a chuck at the front.
It has caused many changes, in the design of spindle journals and bearings,
which are of value,

In all the common lathes a single tool is mostly used. This is so
severe a handicap on production that a large- number of lathes have been
built for multiple-tool cutting. Some of these are automatic in opera-
tion. The distinguishing feature oj these is the mounting of a battery of
tools in the holder of the slide rent to cut simultaneously or in rapid suc-
cession. These arrangements are used chiefly in the manufacture of articles
in which several different diameters occur, with shoulders and faces.

Turret Lathes and Screw Machines.-.........-The difference between a
turret and :i capstan lathe IH that in the former the tool holder is mounted
on a saddle that slides along the bed; in the hitter, the tool holder slides
along a saddle that in fixed to the bed. The practical result is that the range
of movement of the turret in more extensive than that of the capstan. The
first is also made* in larger dimension}* than the second. The difference
between a screw machine and a capstan lathe in that the first is fully auto-
matic in action -•• hence often termed an 4* automatic "......while the hitter is not.
The movements of the first art* canned by cams mounted on drums and on
disks. The movements of the second are produced generally by gears,
feed rods, &e. The screw machine may or may not be equipped with a
turret. The work-holding spindle is most commonly single, but many
lathes now have four, five?, or HIX spindles, each carrying its piece of work
which in brought round in turn to the tools.
Although many common lathes are fitted with turrets, the ** turret lathe "
in a distinct type.   It has a hollow spindle fur bar work, and in many
has a chuck for face work, though the tendency now in to allot these functions
to distinct lathes*   It has a cross-slide, with a tool         at front and at rear.
A chasing saddle ia frequently included for cutting            of greater length
than can be done conveniently from the turret. The hexagonal turret, with
mounted on face, arid with its rotational movements synchronized
with flume nf the work, far outdistances flit* common lathe in speed of pro-
duction. It is usual to scheme the* operation* in such a way that a complete
cydk* turning, drilling, minting with rough and finishing cuts, tapping, &c.
' can he finished on a            piece during ow* rotation, In the simpler
article**) more than one piece can be tooled during         rotation.
Stops.   A feature common to all turret           is the fitting of          to
the            and                of the                  machined. This avoids
Originally a                          was used.   This is


still retained in some instances, and is fixed at the rear of the turret. The
one stop serves for every tool, so that each tool has to be adjusted by it.
This is abandoned in the better class of lathe in favour of a separate stop
adjusted to each tool. For a six-sided turret, six stops are fitted. For
turrets that have cross-traverse movements, similar stops are included to
determine diameters. The cross-slide again has its stops for the front and
back tool posts. Another kind of stop is included in the setting of the
turret tool itself. One determines the precise longitudinal position of
the bar thrust through the hollow spindle. Others, in box tools, set the

Fig. 48.—Automatic Turning Machine

length of a cut, or the throw-out of an opening die, while vee and roller
steadies fix diameters. In the screw machines the setting of the cams
determines the lengths and the diameters of cuts.

Automatic Turning Machines.—It is a remarkable fact, illustrative
of the present trend of machine-shop practice, that just as the turret lathes
and automatics have taken much work away from the common lathes, so
the turret lathes and automatics in turn are being hardly hit by other machines
possessing simpler and more restricted functions. This is largely due to the
fact that economical production requires the dividing of certain classes of
work between distinct machines. Generally heavier cutting can be done
and a larger number of tools brought into action. As a result, many lathes
are now fitted with very substantial rests for holding multiple tools. One
group, represented by several designs, is the automatic turning lathe (fig. 48).
The functions of this group are restricted, but it out-distances the turret

, I


lathe group in some classes of work, owing to the simplicity of its functions,
the ease of setting-up, and its substantial build.

Illustrations of Turret Work.—Some examples of this kind, done
on lathes by Messrs. Alfred Herbert, Ltd., are given in succeeding figures.

Fig. 49 shows the distribution case of a rotary aero-engine being produced
on a combination turret lathe. The tool seen in operation is a counterbore,
the one swung round towards the front is a trepanning tool, which cuts a
recess of lo-in. bore, irV in. wide by 4 in. deep. These two tools are used

Fig. 49.—Distribution Case of a Rotary Aero-engine being turned on a Combination Turret Lathe
with feeds as coarse as 8.8 cuts per inch, and between them they remove over
100 Ib. of metal.
Fig. 50 illustrates the second operation on a propeller boss for an aero-
engine. The work is held on a face plate form of fixture, and located
from the tapered bore with a spring tapered peg. The boss is turned with
an allowance for grinding, faced and counterbored, and the hole threaded
with a collapsing tap,
Fig. 51 shows the rough-turning of the fins of an air-cooled aero cylinder,
one of which is seen on the turret. The work is being done with a gang of
tools similar to parting-tools mounted in a special tool holder at the back
of the cross slide, all operating simultaneously. The piece is chucked
with an expanding arbor, and steadied with a revolving support carried
in the turret.


Fig. 50.—Second Operation on Propeller Boss

Fig. 51.—Forming the Fins of an Aero-engine Cylinder



Drilling Machines.—The practice of drilling commonly includes
operations allied to drilling, such as reaming, tapping, facing, arboring,
bossing; and the machines range from the sensitive high-speed group to

Fig. 52.—Heavy Drilling Machine




problem of attendance is linked. The practice of drilling single holes from
a single-spindle machine is adopted less frequently than it used to be, and,
instead, the practice of multiple drilling is resorted to where possible. In the

Fig1- 54'—Three-way Horizontal Multi-spindle Machine
majority of cases, jigs or fixtures, or both in combination, are now employed
in drilling and reaming operations.
Selected Machines.—Fig. 52 illustrates a stiff drilling machine by
Baker Bros., Toledo, for boring motor cylinders. One motor firm has 350
" Baker " machines in its plant, many being arranged in gangs. The machine
shown is electrically driven, being started, stopped, and reversed with a
push button. In all the series the speed ranges are eight in number, and
feed ranges twelve, the numbers being suited to the sizes of machines.
THE  WORK OF THE  MACHINES                 219

Capacities range from J to 4! in. diameter in steel, and from 4 to 6 in.
diameter in cast iron. A i-in drill can be fed in steel at the rate of 14! in.
per minute, a i-in. drill in cast iron at the rate of 24 in. per minute.

Fig. 53 shows one of the Asquith radial machines, electrically driven,
having a central thrust to the spindle. The firm's universal tilting table is
a valuable adjunct, since it enables each side of a piece of work (except that
in contact with the table) to be machined at a single setting. The table pivots
through a complete circle on trunnions, and carries two independent tables
on opposite faces, each of which can be given a rotary movement by hand.
These tables have tee-grooves for the attachment of work. Drilling, reaming,
and facing can be done at different angles.

At present, opinion is divided concerning the best uses to which single-
spindle machines disposed in gangs, or multi-spindle machines, may be put.
The spindles are disposed in gangs, or in clusters. Fig. 54 shows a highly
specialized design to deal with work to be machined from three faces, without
resetting it. The machine shown is by the National Automatic Tool Company
of Richmond, Indiana. Many of the multi-spindle tools have been evolved
for motor work, for drilling crank cases, cylinders, gear cases, cylinder heads,
connecting rods, &c. They produce a large number of holes simultaneously
instead of singly. They also ream, counterbore, and face the holes.

Boring Machines.—The difference between a machine that drills and
one that bores is that the latter deals with larger holes, which fact influences
the design and the operating mechanism. Though in very many machines
boring is included with drilling, only holes of small diameters and of moderate                            {
lengths can be bored in these machines.    Since these machines are for                            J
general purposes, tapping, facing, and often milling are included.   Here a                            *
large range of speeds and feeds is essential.    A modern machine of this                            ^
class will have as many as eighteen spindle speeds, ranging from 7 or 8 r.p.m.                            }
to 200 or 250 r.p.m., and say nine feeds, which, given in inches per revolu-                            j
tion of the spindle, range from 0-006 in. or 0-007 in. to 0-115 in. per                            |
revolution.    If tapping and milling are not included the range need not be
so extensive.
An excellent example of a horizontal-spindle design is the " Pearn-
Richards " combined machine, the functions of which include drilling, boring,
tapping, surfacing, milling, and, with a suitable attachment, screw-cutting.
Thirty-two variations in speed are provided, and eight rates of feed, applicable
to the longitudinal, transverse, and vertical slide movements. The illustra-
tions are nearly self-explanatory. The machine is manufactured by Messrs.
Frank Pearn & Co., Ltd., Manchester.
Fig. 55 shows one of the Crossley gas-engine beds being bored and faced
with tandem cutters in the bar. The bar is driven from the head and sup-
ported in the hinged bearing on the stay at the right hand. The square table


on which the bed is carried can be rotated on a pin to present different faces
to the work. It is also detachable. Fig. 56 shows the same bed turned round
90° to have the cylinder end faced and bored to receive the liner. After
this, an edge-mill machines the water inlet and outlet and cam shaft

Vertical-spindle Machines.—These are the most popular types at
present for dealing with motor cylinders and those of the smaller gas engines.
The spindles are massive to enable them to withstand heavy cuts in bores

Fig. 55.—Boring and facing Crank-shaft Bearings of Gas-engine Bed
(Pearn-Richards Horizontal Combined Machine)
that range, say, from 4 to 6 in., and in lengths up to about 16 in. The
analogue of this class of spindle is that of the horizontal snout boring machine,
introduced originally to deal with cylinders of small bores and having either
one or two spindles. This is being supplanted by the vertical design, to
which multi-spindles are more readily fitted, while the workman has a better
view of the operations, and the cuttings fall clear away at once instead of
choking the action of the tools. It is also easier to design and handle fixtures
for the vertical spindle machines than for the others.
With the rapid extension of automobile work the vertical-spindle machines
have been subject to many changes and improvements. Single-spindle
machines are ranged in gangs, three or four comprising a working unit.
The cylinder, held in a suitable fixture, is rough-bored under one spindle,
Fig. S6.—Facing Cylinder End of Gas-engine Bed.   Followed by boring and milling searings

(p. 2?0)


finish-bored on the next, and reamed in a third, with possibly a finish reaming
to follow. The fixture retains the casting accurately, and a jig locates and
coerces the boring bar.

When a casting includes more than one cylinder bore the same system is
adopted. But here the single-spindle machine is at a disadvantage. Twin
and multi-cylinders, therefore, cast en bloc, are better dealt with in machines
having as many spindles as there are bores, all operating simultaneously.
The boring or reaming of two, four, or six cylinders occupies no more time

Fig. 57.—Two Distinct Pieces of Work being tooled on one Machine

(p. 223!

than that of one. Then, to avoid loss of time in changing of tools and altering
speeds and feeds, work if held in a fixture can be transferred between adjacent
machines for rough- and finish-boring and reaming.
Boring and Turning Mills.—These are strictly lathes in which the
axis of revolution of the work is vertical. They afford conveniences relating
chiefly to the chucking of work on a horizontal face plate and to the very
large diameters that can be dealt with thus. The advantages are most
apparent when a number of separate pieces have to be set up, and when a
piece of work requires loose packing, bolts, and clamps instead of being
gripped in chuck jaws. And, when articles are not concentric, counter-
balancing necessary in the common lathe is not required in the vertical
machine. Another point in the vertical machines is that the work tables
are well supported, and provision is frequently included when doing light


turning for running them on the spindle only, and for massive work to
support them on an annular ring nearly as large as the diameter of the table.
Very many machines have two work-holding tables. At the opposite ex-
treme, machines of large dimensions will take pieces from 30 to 40 ft. in
diameter. On all, a cross-slide, much like that of a planer, receives the
saddles that carry the tool slides. Frequently a turret is mounted on a slide,
carrying a battery of tools. Boring and turning are performed simultaneously,
and turning may be done from two tool-holders on opposite sides of a diameter.
A photograph of work being done on the machines by Messrs. Webster &
Bennet, Ltd., of Coventry, will serve to indicate the utilities of the boring
and turning mills equipped with turrets. In fig. 57 two distinct castings
are being tooled on one machine, bored, turned, and faced, in charge of one
attendant. Loose chuck jaws hold the work in each case.


Milling Machines.—These are all derived from the Lincoln millers,
to which they bear no resemblance beyond the fact that they all employ
rotating cutting-tools with many teeth.

The Lincoln Machine.—This is used for plain horizontal and face

Fig. 58.—Slab Milling on a Plain High-power Cincinnati Machine.   Material Steel, width cf cut 5 in.,
depth £ in., feed 19 in. per minute.   Material removed 24 c. in. per minute.


I !
1 *


milling. Generally the bed is of the lathe type, and receives the saddle on
which the work-holding table has a cross-traverse movement. In some cases
the bed resembles that of a planing machine, along which the work table
traverses, this giving a longer range of feed than the other. As the table
cannot be elevated, vertical movements are imparted to the spindle, which
slides in its bearings in or on the faces of housings fixed at the left-hand end

Fig. 59.—Vertical-spindle Machine

([). 230)

of the bed.   An arbor support is provided in a tail block at the right hand,
adjustable along the bed.
The Pillar and Knee Machine.—Also frequently termed a horizontal
spindle machine, this has a hollow column that carries a headstock on top,
and a knee on one face, which receives the work table and its slides. All
vertical adjustments are imparted to the knee. Machines are plain or
universal, the first being restricted to rectangular movements only, that of
the table along its saddle, that of the table alone, longitudinally, and that of
the knee vertically. The second includes in addition a spiral head, an index



Fig. 60.—Milling the Inside Faces of Universal Yokes with two Inserted Tooth Cutters

(p. 226)

plate, a sector, change gears, and a swivel table. By these additions a rotary
movement can be given to the work while the table is being fed at any angle.
Any gear-cutting, specially that of spirals, and the teeth of cutters, can be
produced thus. These machines are marked by increased strength and
stiffness of late years, and speeds and feeds are very commonly being imparted
through boxes of gears. In Fig. 58 will be found an illustration of a

VOL. i.




powerful Cincinnati " manufacturing " machine. The rigidity afforded by
the overhanging arm alone steadies the cutter sufficiently without using the
front brace.

Vertical-spindle Machines.—These (fig. 59) in their broad outlines
suggest the common drilling machines. They have a column, arched above
to carry the vertical spindle, which receives edge or face cutters in its nose,
and is frequently belt-driven. A knee adjustable vertically carries the
work table and slides. Numerous variations occur in the details of these
machines, one of the most valuable being that of adaptation to profiling.

Piano-millers or Slabbing Machines.—These were the latest to
be developed, but they are being employed increasingly. They are built
on the planer model, with a long bed and work table, flanked by vertical
housings, carrying an adjustable cross-rail, with spindle heads. They often
successfully rival the planers, since a single cut is taken over a wide face
during the table travel, instead of requiring a large number of reciprocating
movements. Their utilities are enhanced by the fitting of horizontal
spindles on one or both sides in addition to those on the cross-slide,
sometimes also provided with angular settings, while some machines have
circular tables on the one that reciprocates. The sphere of these machines
lies chiefly in massive work, much of which is arranged in tandem, fre-
quently with the help of fixtures. Edge and face milling are both done,
and a large proportion of gang milling.

Continuous Milling.—This, the last development in this kind of
machining, includes that done on piano-millers, but it is generally understood
to refer to that performed on the rotary tables of vertical-spindle machines,
and is nearly invariably associated with the employment of fixtures. Fig. 60
illustrates a Becker machine machining the inside faces of yoke pieces, em-
ploying two y-in. inserted tooth cutters. Thirty-six pieces are held in the
fixture, and the production is 160 pieces per hour. Connecting-rod ends are
milled on their faces, with pairs of inserted tooth cutters, on a double-spindle
machine. They are set diagonally in place in the fixture to lessen the space
left for " cutting wind ".


Reciprocating Machine-tools.—These include the following tools:
(i) The standard planing machine with bed, work-holding table, housings,
and cross-rail, and tool-boxes; the derived machines are: the open-side
planers, pit planers, well planers, portable machines, and key grooving and
broaching machines. (2) The shaping machines having single or double
rams, and tool-heads. The portable shapers are a small group. Gear-tooth
planers are shaping machines of short stroke. (3) The slotting machines, in
which the tools reciprocate vertically, one or two tools being carried in the
ram. The tables are simple, with rectilinear movements, or compound,
to include a circular motion for circular slotting.

Widely though these machines differ, they are properly grouped as

in r



on tl

in ra
on t!
the .


reciprocating because the cutting only occurs on one stroke. The return
stroke simply brings the work or the tool back to its original starting-point
in readiness for another trip.

The common planer is a machine for general purposes. It takes any
work within its capacity. The functions of the shaper and the slotter are
extremely limited, since they only deal with small surfaces. The portable
machines are employed to perform their functions on massive articles in
situ or on floor plates on work that cannot be set on machines. The key-
grooving and the broaching machines are specialized designs that cut narrow

Fig. 61.—Motor-driven Planer

slots in bores and elsewhere. In some of their functions they resemble the
slotting machines, but they deal with lengths impossible on the slotter, and
produce sections at one stroke that could only be done much more slowly
on this machine.
The principal improvements in the later planers have been the following:
(i) An increase in cutting speeds, and provision for effecting several changes
in rates suitable for different metals and alloys. (2) A rapid rate of return.
(3) The cushioning of the reversal with springs to absorb and give out power
on the return stroke. (4) The employment of a light aluminium alloy for
the driving pulleys to lessen the inertia at reverse. (5) Driving at high
speeds with narrow belts, using separate fast and loose pulleys for driving
228                           THE  MACHINE-SHOP
and reverse, and pulleys of different sizes for the two functions, instead of
trains of gears. (6) Employing a large " bull wheel " for driving the table
rack instead of a small pinion. (7) A vast extension of electric driving, with
a corresponding multiplication of speeds, reverses, and feeds effected by
switches. In consequence of these improvements, modern planing-machines
hold their own in face of the keen rivalry of the piano-milling machines that
operate on the same classes of work.
An example of a motor-driven planer is given in fig. 61. Taking 84 in.
by 84 in., a 45-h.p. motor is required. It is one of a series by the Cincinnati
Planer Company of Cincinnati, Ohio, U.S.A. It is termed a " rapid power
traverse machine", because each tool-head is moved rapidly from one
position to another with power, derived from the motor mounted on the
arch, instead of slowly by hand. Power is transmitted down through a
splined shaft to a gear box at the side, provided with lever handles. Forced
lubrication is supplied to the table vees. The table is boxed, and open at
the sides so that dirt and chips can be drawn out.
Gear-cutting Machines.—Broadly all these fall under one of two
groups. In the first the teeth are shaped directly or indirectly from a
pre-existing form: directly, when the cutter has the section of the tooth
space; indirectly, when a reciprocating planer arm carries a single-edged
tool, controlled in its lateral movements by the edge of a former having the
desired tooth curves to an enlarged scale. This method is suitable for all
teeth, whether with single curves (involutes), or double curves (cycloids),
and spurs or bevels. But, since the degree of accuracy obtained depends
on the accuracy of the form, it is open to error. Though this may not be
wholly eliminated, the gears so made are good enough for most commercial
manufacturers. But they do not meet the very exacting demands of the
high-speed gears used in automobiles and the best machine-tools.
In the second group the teeth are generated from the basis of the involute
rack-tooth with straight sloping sides. A cutter having the section of a
rack-tooth is used for generation, or one flank only of a rack-tooth, or several
complete rack-teeth combined in one cutter, or, a pinion-like cutter, is gener-
ated from a rack basis, or a hob—a worm, with teeth of rack section, is
fluted in milling-cutter fashion. In some machines the rack-tooth is not
embodied in the cutter at all, but in the mechanism of the machine itself by
means of " roll cones " in one design, and in another by certain controlled
movements of slotted links.
Pressure Angles.—In order that all generated involute teeth shall
mesh together, the pressure angle must be the same for all. This corre-
sponds with the diagonal path of contact of the teeth to which the sides of
the rack-teeth on the pitch points are normal. This is 14^° in the B, & S.
system, the one until recently almost universally adopted. Its disadvan-
tage is that small pinions are much undercut below the base line, to

THE  WORK  OF  THE  MACHINES               229
oid which the rack-teeth in this system are slightly rounded, and the
:ms of the cutters for pinions below thirty teeth have two curves instead
one. Undercut can also be prevented by increasing the length of the
dendum of small pinions. But other views now obtain, chiefly in conse-
ence of the growth of generating methods, of the increasing employment
the short " stub " teeth, and the desire for the closest approximation to
ithematical accuracy. Pressure angles are now increased to 18°, 20°, and
sn 25°. Gears can thus be produced without undercut down to twelve
Machines using Form Cutters.—The type of these using rotary
tters is the Brown & Sharpe. One group is used for spurs only, another
:ludes the cutting of bevel gears. Later machines include provisions for
iltiple cutting. Form planing of spur and bevel gears is represented by
5 Gleason machines. These are made to be pitched by hand or automati-
Machines for Shaping  Gear Teeth.—The " Bilgram " was the                           j
ginal machine.    It is now made for shaping spur as well as bevel teeth.                           !
is, and the Robey-Smith, employs planing tools, the movements of which                           I
: controlled by links.    The Fellows machine cuts spurs, internal gears,                           j
i helical teeth.    It employs a pinion-like cutter.    The Sykes machine                           j
ploys two cutters, which operate simultaneously.    They produce spur                           \
I helical teeth.    The Gleason planer shapes the teeth by means of a yoke,                            j
the inside of which a segment is bolted which has the same angle as that                           I
the gear to be cut.    The Sunderland machine cuts spurs and spirals,                            j
ng a reciprocating cutter containing six rack-teeth.    The machines that                           j
pe by means of hobs, cut spur, spiral, and worm-teeth.                                                       i
DIVISION VIII                                                                     ;
Grinding Machines.—Grinding has invaded the  old territory of                            |
ning, boring, and facing.    The lathe is now often a mere satellite—a                            !
ghing, a first-operation machine, playing second fiddle to the grinder.                            {
:ut is taken with a coarse feed that leaves marked spiral ridges on the                            f
face of the work.    Then the grinder performs the second operation,                            I
nely, that of fine-finishing to precise limits.    The lathe reduces with                            I
ater economy than the grinding wheel, but the latter imparts a finish in
tere fraction of the time that would be occupied by the turner in producing
precise results.   When machining allowances are slight, the grinder takes                            >
rge of the entire work.    It is not necessary to pickle, as it is when milling                            I
ters have to remove small amounts.    The grinding wheel can operate                            !
b. allowances of iV in. or less, which would give trouble to the lathe man                            |
) has to get under the skin.                                                                                               ,
Cylindrical Grinding.—This represents by far the largest volume of
t done.   The common method, to which there are exceptions, is to                            \
ite the wheel and the work, and to traverse the wheel.   The object of
traverse is to get the maximum amount of duty from the wheel, and to
230                           THE  MACHINE-SHOP

retain its truth as long as possible. To use a traverse feed only slightly less
than the width of the wheel is more economical than to employ a feed that
bears a small proportion to the width of the wheel. The peripheral speed
of wheels is usually about 5000 ft. per minute, that of the work from 20
to 25 ft. The wheel speed is constant, that of the work is changed when
desirable for making differences between roughing and finishing. Chatter
and vibration are prevented by the employment of a large number of back

Surface Grinding.—This has been largely favoured by the employ-
ment of the magnetic chucks. These hold flimsy and awkwardly shaped
pieces, which would give vast trouble if clamped on work tables. Reinforce-
ments in the shape of stops and rings are necessary to prevent side-slip. A
large number of small pieces can be held and operated on thus. Fixtures
are also largely employed. The machines are built in two types, one in
which the work table has linear movements, the other with rotary motions.

Machines for grinding cylinders, for form grinding, and those for tools
and cutters include a large number of designs. The machines for grinding
the cylinders of automobiles and gas-engines have developed with startling
rapidity. The spindles have a planet or eccentric motion, so that while they
are revolved at high speeds they are rotated slowly in a circular pathway,
the diameter of which is increased to impart the feed. The work is carried
on a table that can be adjusted transversely to bring bores in alignment with
the wheel. The work table is fed towards the wheel with changes of travel
for roughing and finishing cuts. Wet grinding is provided for by a pump
and tank and pipe.

Continuous Grinding.—This relates to the treatment of numbers of
small pieces arranged in tandem, or in a circle, to be ground with face wheels.
Much of this work is done on magnetic chucks or in fixtures. The more
awkwardly shaped and the smaller the pieces are, the greater are the econo-
mies of continuous grinding. Often the choice lies between this method
and that of milling done on lineally or circularly moving tables.

The Shops
Organization.—This must be based on a rigid cost system, from which
the price of work in all its stages can be ascertained, and leakages detected
from day to day. The old method of adding men's time in the aggregate
and lumping contingent expenses and profits on that is no longer followed
in competitive firms.
In order to fix costs at all stages a routine system is essential.   For this

8 §2

Fig. 6a.—The Heavy-turning Shop
Fig. 63.—Spur Gear-cutting Department

THE SHOPS                                  2"
an intimate knowledge of the nature and scope of the operations performed
on hundreds of machine-tools is necessary. This devolves on the shop
manager, and on the foremen who have charge of the groups of machines,
as lathes, automatics, planers, gear-cutters, milling machines, grinders, and
so on. Each foreman must know the capacities and limitations of each of
the machines in the group of which he has charge, and must see that they
are operated to the fullest advantage. He will consult and discuss with the
shop manager respecting the best methods of machining certain articles. The
manager will decide the question of economies that may result by the trans-
ference of work from one group of machines to another, as from lathes to
turret lathes, from planers to milling machines, from lathes to grinders, and
so on. The foremen and manager jointly consider the question of the design
and employment of fixtures and jigs, and the relation of the expense which
they bear to the product. Detailed drawings are made in the office from
sketches supplied.
When the methods of machining have been determined, the details are put
on a definite basis by the foreman or the rate-fixer. Sketches are prepared,
or cards are written, stating precisely the nature and sequence of the several
machining operations involved, the tools to be used, the speeds and feeds, and
limits. Generally it is possible, as the work proceeds, to effect slight speeding-
up, which on a piece-work basis, or a bonus system, is to the advantage of
the machinist. But it does not lie with the attendant to make changes in the
general routine previously determined. That can only be done by suggestion,
with the consent of the foreman, or manager.
This organization includes all details. The grinding of tools of all kinds
is done in the tool-room, and they are checked out to the men, and returned
when they have become dulled with use. Gauges, jigs, and fixtures are treated
similarly, and they are corrected or renewed in the tool-room. In this
system nothing is tabulated by name. Every item, however insignificant,
has a number, or a letter, inserted on the drawings, and is checked out and
in by that.
The Tool-room.—-This is a necessary growth, consequent on turret
practice, and on the employment of the multiple-edged cutters used on
milling machines, gear-cutters, and elsewhere. The set-up of boxes of tools
for turret work entails elaborate constructions and delicate adjustments.
The grinding of cutters can only be done on universal machines. Drills
are ground on machines. The standardized grinding of single-edged cutting
tools is done on machines. These functions are relegated to the men in the
tool-room, who also construct the smaller jigs and fixtures. Hence the
tool-room is a machine-shop in miniature, a microcosm complete in itself.
It contains a few machines of every class, in which universal designs are in
evidence, so that, having castings and forgings and bars supplied, the whole
of the work of tool-making in its widest sense is performed within its pre-
cincts, and tools are ground, repaired, set-up, and kept in working order,
ready for use in the shop.
Fig. 64.—Department for Milling and Testing Worm Wheels
THE  SHOPS                                 235
Illustrations of Shops.
i. Messrs. David Brown & Sons (Huddfd.), Ltd., Lockwood, Huddersfield.__
This is a large works, occupied solely in the production of gear wheels.
The extensive shops are laid out on the ground floor exclusively, and all are
arranged in parallel. The works are self-contained, including pattern-shop,
foundry, smithy, and hardening-shop. There is a heavy machine-shop, and
a heavy erecting-shop. The principal sectional departments are: the raw
material stores, the tool stores, the cutting-off, the light fitting, and milling,
automatic, double helical and bevel gear, spur, worm, and spiral departments.
Also heavy and light turning, planing, boring, drilling, capstan lathe, and
grinding departments, the tool-room and inspection. The bays range from
120 to 310 ft. in length. They are served with overhead electric travelling
cranes. Skylights in ridge roofs give ample light, and arc lamps provide
artificial illumination. The machines are all driven electrically, the smaller
in groups, the larger with separate motors.
Fig. 62 shows the heavy-turning shop. Heavy boring and turning mills
are seen on the right, and a number of chucking lathes .on the left, served
with an overhead runway and pulley blocks. Fig. 63 is a view taken in the
spur gear-cutting department. A catholic selection of machines is apparent,
They include the Gould & Eberhardt, the Brown & Sharpe, the Sunderland,
and the Fellows gear generators. Much of the work is of a massive nature,
requiring the service of the overhead travelling crane. Fig. 64 is the shop
in which worm and spiral gears are milled and tested. The machines used
were designed and built by the firm. The machines for grinding worms
after cutting and hardening are also made by Messrs. Brown. The heads
of the grinding wheels are adjustable to suit the gear angles. In the general
grinding-shop, cylindrical and vertical spindle machines are installed, and
trays disposed down the centre hold the work.
2. Messrs. A. Harper, Sons, & Bean, Ltd., Dudley, Worcestershire.—
The works of this firm at Tipton are built for the construction of auto-
mobiles, the various departments of which are illustrated by the photographs
following. Precision tools are made at Dudley, and drop forgings and
pressings at Smethwick. The foundry is at Tipton.
Fig. 65 is a view in the milling department. An Ingersoll machine
occupying the centre of the shop is dealing with a row of crank cases. It is
machining the timing cover face, the cylinder face, and the ends of the feet
simultaneously. Two vertical machines on the left mill the sump face,
and the magneto, starter, and lighting faces respectively. In the foreground
at the right is seen the milling of the vertical face for the magneto cradle.
These machines are laid out in line, arranged for each operation in sequence,
and the component parts are passed along to the machines on a roller type
of conveyor, with ball-bearings, occupying the centre of the shop.
Fig. 66 shows a line of seven multiple spindle " Natco " (National Auto-
matic Tool Company) drilling-machines. The third, fifth, and seventh—the
Fig. 65.—Milling Department for dealing with Crank Cases


Fig. 66.—Crank Cases, in Fixtures carried along Tracks, being: Drilled and Tapped under
" Natco " Multiple-spindle Machines
foremost—are being used for tapping in all 44 holes in the crank case,
and the others drill between them 72 holes.   The method used for locating
*the case is the truck fixture on trunnions, transported along the tracks.
The case is loaded at the beginning of the track, and is unloaded after passing
under all the machines.   The trucks run on to turn-tables, and are returned
by way of a similar track at the rear.   Two or more fixtures are used.   As
each operation is completed the chips are blown out by compressed air.
Figs. 67 and 68 are two views of the methods employed in milling the

cylinder faces on an Ingersoll machine.   Four spindles are operating together
on two components at a time.    On the left side the crank case and manifold




faces are machined—these two faces being used as registers for machining
the top face, and the inspection cover faces on the other side of the machine.
The fixture will take twelve castings, so that it will be seen that six are
pleted at each setting.


Fig. 69 shows an assembled frame in position in a frame-drilling jig.
Any errors in the frame are taken care of by the locations being made^self-
centring and compensating. Two Hammond double-arm drilling machines

cover all the holes, the machines being bolted on channel irons. Special
shanks are used for all the various drills to bring them to their correct levels,
and thus avoid vertical adjustments of the machines. The shanks are used
jn conjunction with quick-acting " Gronkvist" Swedish drill chucks, which
allow the drills to be removed without stopping the machine. The drilling


jig consists of two cast-iron trunnions, on which swing two indexing plates.
Channel irons which connect these plates carry all the bush plates, and
quick-acting clamps, the compensating beams, and locations. The con-
struction is such that the revolving parts are light, yet rigid, so that no de-
flection can occur during drilling. The whole jig can be revolved easily,
and indexed to allow for drilling from four sides. Foot treadles are arranged
in convenient positions for starting and stopping the machines. The result
of these economies is that the 112 holes in the frame are drilled in twenty


G.   M.   S.   SICHEL,   B.Sc.

VOL. 1.                                                                  241                                                                           16
Fitting and Erecting of Heavy

The assembly and erection of heavy machinery of all types calls for
not only the skill and care which the handling of large plant requires, but,
to a very large degree, for the judgment which is partly intuitive and partly
the result of wide and all-round experience. So many problems arise in
the erection of plant on site, as compared with its erection at the makers'
works, where all the usual facilities exist, that success or failure depends
very largely on the ability to size up a difficulty correctly, and then devise
ways and means of producing the best possible results. It is not possible,
therefore, in an article of this kind, to give complete directions for the assembly
and erection of all kinds of heavy plant, as the conditions to be met with
vary so greatly; the aim of the article will be to deal with the kind of problems
that arise and the various means taken to meet them. It will probably be
conceded that, if the problems which arise in the erection of a large steam-
turbine electric generating set and condensing plant be considered, the
ground will cover most of the problems that arise when handling less com-
plicated plants. The article will deal, therefore, with a plant of this de-
scription, and some general notes will be added on the application of the
principles to the erection of special plants.

Foundations.—These are almost invariably made of concrete nowa-
days, though in certain cases brick is used for cheapness and where the
weights to be supported are not very heavy and are not subjected to shock.
In general, it is advisable to make the lowest part of the foundation block
in the form of a concrete float or raft on which the main foundation block
is built. The dimensions of this concrete raft depend on the nature of
the subsoil; where this is soft or friable, the area of the raft must be corre-
spondingly large in order to lessen the weight, per square foot on the raft,
of the superimposed machinery. Where the subsoil is particularly soft, it
will probably be necessary to drive a large number of piles first, round the
heads of which the concrete raft is built. It should be the first duty of the
engineer in charge of erection of machinery to satisfy himself regarding the
suitability of the foundations and the subsoil. In general, the foundations
are provided by the customer to the drawings of the contractors who supply



and erect the machinery, but no contractor will accept responsibility for the
foundations which are built to his drawings, as he cannot be expected to be
familiar with local conditions, or the peculiarities of the subsoil.

The datum line, or level from which all vertical dimensions are taken,
is usually the finished engine-room floor-level, and the foundation-block
height is carried up till, with an allownace of i in. to 2 in. for packing plates

Fig. I.-—-2O,ooo-kw» Turbo-alternator in Course of Erection, showing bottom half of Turbine Cylinder
in position, with Condenser and Valve Chest coupled up
under the bedplates, the top of the latter is at the required height with refer-
ence to the datum level. In some cases it is becoming the practice to build
short pieces of H-girders into the top of the foundation block to support the
bedplate; these girders are spaced every 3 ft. or so, and are carefully levelled
so as to present a smooth metal surface on which to level up the bedplates.
Where provision has to be made in the foundation block for foundation
bolts, it is advisable to make the holes big enough to allow at least 2 in.
lateral movement of the bolt in every direction; this allowance will take
care of any inaccuracies between the drawings of the bedplate and the actual
Fig. 2.—looo-kw. Turbo-alternator Set partly erected, Generator Rotor being lined up to Turbine



casting, more particularly as regards the spacing of the holes for the founda-
tion bolts. As it is the general practice to grout-in bedplates, care should
be taken to see that the top surface of the foundation block is left rough,
so as to allow the grouting material to obtain a grip or bond with the founda-
tion. Where short H-girders are built in, as explained (p. 244.), the level of

Fig, 3.—i, Concrete float.   2, Ground or basement leveL    3, Foundation block.   4,Short steel girders
lilt in.    5, Surface of block left rough for grouting.    6, Air inlet to generator.     7, Space for condenser.

built in.    s,__________

8, Foundation bolt holes.

the concrete should be left at least 2 in. below the top of the girders, in
order to allow the grouting material to obtain a good grip of the girders.
Bedplates.—Before putting the bedplates in position for carrying the
prime mover and generator, care should be taken to see that any heavy parts
of the plant, which are situated underneath, are put in position first, in
order to obviate trouble and difficulty later on. In the case of very large
steam condensers this is essential, but where the weights are not excessive,

Fig. 4.—i, Condenser.  2, Feet of condenser.   3, Wood baulk not necessarily secured to condenser feet.   4. Wood
baulk or steel girder for runway.   5, Steel rollers.   6, Cross batten to tie together supports (3).
and the dimensions reasonable, the condenser may be slid In under the
turbine, after the latter has been erected, and then jacked up or lifted by the
crane, one end at a time, and packed up till in its final position. To do
this, it will probably be necessary to build a suitable cradle, or support of
girders or timber baulks, on which to rest the condenser during this operation.
Let us assume, then, that the condenser and other heavy parts, e.g.
atmospheric exhaust valve and pipes, &c., have been, placed approximately
in position. The turbine bedplate is thea put on the foundations, prefer-
FITTING AND ERECTING                        247
ably on steel packing plates at least i in. thick, placed underneath the heavier
parts, so as to leave a gap below the soleplates, which will facilitate the
insertion and drawing-out of wedges or flat packing plates used in levelling
up. The bottom portion of the turbine cylinder can then be placed in posi-
tion and levelled up, both axially and transversely.
It may be accepted as a good general rule that, before placing together
two machined faces, the faces should be lightly rubbed over with a smooth,
flat file, more particularly all round the edges of the machined faces. This
not only ensures that the faces are clean and free from burrs (particularly
round the edges of drilled holes), but it also immediately shows up any bulges
or bumps on the faces. These burrs are often caused by the links of chain
slings pressing into the machined faces, when chains are used for lifting the
castings. Instead of putting down the bedplate first, the bottom half-turbine
cylinder may be put directly on the bedplate, bolted down, and the whole
then lifted in one piece, where the capacity of the overhead crane or lifting
tackle is large enough. In the majority of steam-turbine plants, guide or
director keys are provided between the bedplate and the cylinder, which
prevent movement in a transverse or lateral direction, while allowing free
movement in an axial or vertical direction. This is done in order to allow
the cylinder to " breathe " or expand and contract with varying temperature,
without upsetting the alignment of the set. It is important, therefore, to
see that these guide or director keys are not only properly fitted, but when
fitted are secured against the possibility of working out. The turbine
bottom half-cylinder and bedplate may now be levelled up; it is often found
when levelling up large castings that if the piece be levelled, say in a trans-
verse direction, by means of a level applied to one side of the casting, and
the level be then tried in the same direction on the other side of the casting,
it will be found to be out of level. This is nearly always due to the casting
having " sprung ", due to internal strains in the casting easing themselves,
more particularly when the skin is broken by machining. Another cause
Is the manner in which the casting has been bolted or cramped down on
the boring-mill or planer-table, when being machined. If the casting has
been sprung before machining, then, when the bolts or cramps are released,
the casting will spring back, and the result will be that the machined faces
are not true. This difficulty is got over by placing a steel or stiff wood
straight-edge right across the bedplate or cylinder, and putting the level on
the straight-edge and setting the casting or piece level in this manner. The
levelling up is done by inserting steel wedges, preferably 3 in. wide, under
the soleplate, and, when wedged up, adjusting the height of the parallel
packing pieces. In this way a true surface is prepared (represented by
the top of the packing pieces), on which the bedplate may be moved laterally
without upsetting the level of the piece. The next operation is to check
the height of the centre of the turbine shaft or spindle above the engine-room
floor-level; if the centre is too low, then it will be necessary to increase the
height of all the parallel packers by the amount by which the centre is too
low. Conversely, if the centre is too high, the thickness of the packers
248                            HEAVY   MACHINERY
will have to be reduced, either by the use of thinner packers or by having
the packers machined. Having thus levelled, at the proper height, the bed-
plate and bottom half-cylinder, the next operation is to set the cylinder
central on the axial centre line. For this purpose it is usual to use a length
of fine piano wire, stretched tightly by means of weights between the two
outer pedestals. The wire is very carefully centred at the extreme edges
of the bearing pedestals, and the bottom half of the cylinder, plus bedplate,
is then jacked over on the packing pieces below the bedplates, so as to bring
it central on the steel wire. The final test is made with an inside micrometer,
behind which is held a piece of white paper, in order to show clearly when
the end of the micrometer is just touching the wire. This adjustment makes
the cylinder right for position sideways and vertically. Its position end-
wise is usually taken from the centre line of the turbine exhaust, and this
line, as well as the axial centre line, is determined beforehand for building
up the foundations, and is retained for definitely fixing the position of
the turbine.
When the piece is finally set, the level should again be very carefully
checked, and if necessary readjusted, before the bedplates are grouted-in.
There is a difference of opinion among engineers regarding the best time
during erection for grouting-in the bedplates.    Some men prefer to erect
the whole plant complete before doing any grouting; others prefer to grout-in
the bedplates immediately they have been finally set and checked over, and
before any weight, e.g. other portions of the plant, is put in position on
the bedplate.   The arguments used in favour of the former method are,
that if the whole plant is completely assembled first, any errors in the draw-
ings, which might make it impossible or difficult to fit the various parts of
the plant together, can be adjusted without cutting away the foundations
or undoing a lot of work made permanent.   Against this advantage must
be placed the disadvantage of liability to spring the castings and soleplates,
due to the concentration of the weight of the whole plant on the compara-
tively small area of the packing pieces between the soleplates and the founda-
tion block.   On the other hand, the number of cases where a complete plant
has to be taken up and re-erected, due to some oversight in the layout draw-
ings, is so remarkably small as to be almost negligible, and a good deal can
be said in favour of grouting-in immediately the soleplates and main struc-
ture have been assembled and checked for position.   The whole of the area of
the underside of the soleplates is thus available for distributing the weight
of the plant, and in consequence the liability to settle and get out of level
is very much reduced;  further, it is usually possible to make a very much
more satisfactory job of the grouting-in process before the whole plant is
assembled, on account of the greater freedom and space to get at the job
when the bedplates and lower parts  only of the plant are in   position.
Instead of using parallel packing pieces, which are left in and grouted-
up, some engineers  prefer to use steel wedges, about  3  in. wide and
3 or 4 in. long, tapering down in thickness from J in. to nothing.   The
wedges are driven in under the sole plates until the latter are levelled up,


and, after the grouting has been run in and has set, the wedges are withdrawn.
The use of wedges is a much quicker job than with parallel packers, but it
is obvious -that the contact of the wedge with the soleplate is more or less
a line contact, as compared with the surface contact obtained with parallel
packers, and it is therefore essential when using wedges that the soleplates
be grouted-in before any weight is put on.

Care should be taken in mixing up the grout to see that it is thin enough
to run easily under the bedplates;  it should have the consistency of very




	yj.f f ]• {.t(j& 1
	-3                 2               ^ /                    \

	iri ...... ------- ez
	-------------- J^.
	>-T\             rr1/

	^ ___ i

Fi£ 5 .—Bedplate in Position before Grouting-up, showing parallel packers with surface contact and
45                                                   wedges with line contact
i  Steel girders built in.   2, Parallel packers.   3, Wedges.   4, Top surface of foundation block,
'                                            left rough.    5, Grouting level.
thin cream, and should preferably be more liquid than otherwise. The
proportions of cement and sand used are as follows: one part by volume
(bucket or barrow) of cement, two parts fine sharp sand. Before groutmg-
up a dam of stiff cement, or of boards or bricks, should be built all round
the bedplate to a height of 2 or 3 in. higher than the finished level of the
erout and the surface of the concrete foundations should then be thoroughly
wetted with several bucketfuls of water, in order to prevent the water in
the grout being rapidly absorbed by the concrete foundations The grout-
nj material is then run in, and, when the surface of the liquid is above the
bottom of the soleplates, the grout should be well worked under by means
of a short length of thin, flat iron (hoop iron i in. X * m. thick does very
250                               HEAVY  MACHINERY
so, after part of the contraction has taken place and the grout partly settled.
With regard to the length of time required to set, this varies with the nature
of the cement; some slow-setting cements take two to three weeks to set
properly, but where ordinary cement is used in mixing the grout, and the
depth of the grout is not great, sufficient setting should take place in a week.
A simple test is to stab the surface of the grout with the tang end of a file
held in the hand; setting should be allowed till the point of the file marks
but does not enter the grout. It is usual to carry the grout up over the top
of the bottom flange of the soleplate in order to secure a good grip; in
some cases the whole of the hollow interior of the bedplate is run in solid
with grout; in such cases it is advisable to have some holes drilled previously
in the bedplate to let out the air, and allow the grout to fill the whole solidly.
When properly set, the dam of brick or wood is broken away, and the
grout projecting from the side of the bedplate is dressed off. The holding-
down bolts, which secure the turbine cylinder and bearing pedestals to the
bedplate, should be carefully examined and adjusted, and dowel-pins between
the cylinder and bedplate properly fitted. A word might usefully be added
here on the use of dowels and holding-down bolts in such cases. Due to
the wide range of temperature through which a steam turbine has to work,
i.e. from the temperature of the atmosphere, when starting up, to the tem-
perature of the steam when on loads and also on account of the high tem-
perature throughout that may be reached, if the vacuum on the condenser
is lost and the set goes over to atmospheric exhaust, the expansion and
contraction or " breathing " of the cylinder and shaft may be very consider-
able, and amount to J in. or -f in. in large sizes of plant. The amount of this
expansion can be calculated from tables of the linear coefficient of expansion
for various metals, though the actual movement may differ, in certain cases,
from the calculated amount, on account of the shape of the casting, &c.
Provision has to be made to allow this expansion to take place, and at the
same time the cylinder, bearings, Sec., have to be properly held down. It
is usual, therefore, to definitely fix a datum level from which the vertical
expansion and contraction can take place, and also to fix a definite transverse
line from which the axial expansion and contraction can take place. The
datum level for vertical expansion is naturally the top of the bedplate sup-
porting the cylinder, and it is an advantage to have this level as near as
possible to centre line of the shaft and cylinder, so as to make the expansion
or " lift " of the cylinder top half equal to the downward movement of the
cylinder bottom half, and so keep the cylinder under all conditions central,
vertically, on the turbine spindle. In some turbines the datum level is
several feet below the shaft centre line; in these cases it is usual when setting
the turbine shaft to put it some ten-thousandths of an inch high, so that,
as the cylinder expands or lifts with heat, it makes itself central on the
turbine spindle. The transverse line for fixing the datum, from which the
axial expansion takes place, is secured by means of two stout dowel-pins,
one on each side of the turbine cylinder, set half in the cylinder feet and half
in the bedplate. These dowels should be a nice tapping fit in order that,


while they definitely fix a line, the cylinder is free to expand along the dowels
as it gets hot. These dowels are usually placed about midway along the
cylinder length, and the expansion thus takes place in both directions from
this line. In other cases, the dowels are fixed at the exhaust end of the
cylinder, and the whole expansion thus takes place in the one direction
viz. towards the H.P. end. It is necessary, of course, to have a fixedrpoint
on the cross transverse datum line, so as to control definitely the direction
of the sideways expansion. This fixed point is formed by vertical keys at
both the H.P. and L.P. ends of the cylinder, set half in the bedplate and
half in projections from the cylinder on the vertical centre line. These
keys enable the cylinder to breathe vertically, but keep the cylinder central
sideways under all conditions, and thus compel the expansion (lateral) to
take place equally on both sides. The dowels should be made an easy tap-
ping fit, and the holes through the
cylinder feet, through which the
holding-down bolts pass, should be
at least J in. bigger in diameter than
the bolt, so as to allow the cylinder
to expand and move laterally when
heated. The bolts, therefore, have
to be of special construction; they are
known as collar or shoulder bolts,
and are shown, together with the
corresponding type of stud, in the
accompanying sketches. A special
washer is always used under the
head of the bolt or under the nut
of the stud, and it will be seen that

when the bolt is tightened hard down on the shoulder, and the length from
shoulder to under side of head is just correct, that the cylinder, while de-
finitely held down, can move or expand sideways as required. In order to
get the exact length of the bolt, from shoulder to head, the bolt is tightened
down with the collar in position, and the amount of slack between the collar
and the head is then measured with feeler gauges, and the length from
shoulder to head is then reduced by the figure obtained with the feelers,
with the exception of two thousandths, which is left on in order to provide
a very small clearance between the bolt head and washer, and thus allow the
cylinder to expand.

Before jointing up any steam-pipe,-or the valve chest to the cylinder
bottom half, they should be carefully examined for any loose material, e.g.
nuts, pieces of steel, borings, &c., that may have lodged in the steam passages;
in the case of steam-pipes it is very advisable to either draw a heavy chain
through them repeatedly, or to tap them all over the external surface with
a heavy hand hammer, or to do both, in order to loosen any scale or rust
that may have formed inside the pipes, due to the " weathering of the
hard skin on the inside of the pipes. In particular, the steam-nozzles should

Fig. 6.—Holding-down Shouldered Bolt to allow for
Expansion and Contraction of Cylinder Feet relative
to Bedplate
x, Bedplate. 2, Cylinder feet. 3, Bolt shoulder.
4, -002 in. for sliding clearance. 5, Clearance round
bolt for expansion of cylinder relative to bedplate.


be carefully examined, especially the nozzle-box or space behind the nozzles,
which appears to be a favourite place for foreign matter to collect. Any
foreign matter not removed will be blown through by the steam, and may
seriously damage the turbine blading. As a rule, the joints between the
steam-chest and the turbine cylinder are dowelled, in the manner shown
herewith, in order to definitely fix the position of the steam-chest. The
dowels are put in at the joint in order that they can be easily withdrawn or
knocked out when the joint is broken. If the dowels are put in at right
angles to the joint, as in the other sketch, there is a danger of the dowel
being " burned in " by the prolonged action of the heat of the steam, more
particularly when the steam is superheated. On the other hand, great care
should be taken to see that the dowel used in the joint is made an easy tapping
fit after the joint is bolted up tight. If the dowel is too tight, there is a danger

of the joint being forced open at the
dowel-pin, and thus causing serious
steam leakage. A note will be added
later, on the making of joints for
steam-, water-, and oil-pipes.

Assuming, then, that the nozzles
and steam-chest have been carefully
examined and cleaned out and jointed
up to the bottom half of the cylinder,
the next step is to put in position

headed dowel-pins at right angles to joint.   On high-    the   bottom   halves    of   the    Stationary

temperature steam-pipes these dowels are liable to      «.      ,                    r>   r         1 •      •      j           1

"burn in", and have to be drilled out, unless they are    uiapnragmS.       JoeiOre  tillS   IS  UOne  the

made a fairly easy fit initially.                                  outside    Surface    of    the    diaphragms,

which  fit   into   the  grooves   in  the

cylinder, should be carefully rubbed over with a little flake graphite to
prevent them rusting in in course of time. After the diaphragm bottom
halves are in position, the drainage of the cylinder should be tried in order
to see that water, condensed steam, &c., cannot collect in the cylinder,
and not only cause rapid deterioration of the blading, but also be the
cause of the turbine shaft " whipping ", due to the wheels running in
water at the bottom. The effects produced in this way are sometimes
very serious, and have been disastrous. The best way to test the drainage
is to open all drain-cocks and run water from a hose in between the dia-
phragms, and into all pockets where water may lodge. As a rule, the dia-
phragms are so arranged that any water in the cylinder automatically drains
away to the exhaust end, and thence into the condenser. This is accomplished
by the design, or, where necessary, a small hole, say f in. diameter, is drilled
through the bottom of the diaphragms in an axial direction, the hole being
increased at the L.P. stages to | or f in. diameter. A small quantity of
steam blows through this drain hole and keeps the cylinder clear of water,
the loss of steam being quite insignificant. Previous to putting the bearings
in position, the bottom halves should be carefully scraped and bedded on
the journals of the shaft they have to carry. This is done by smearing a

Fig. 7.—Dowel-pins in Joints

i, Dowel-pins drilled radially.   Joint can easily be
split and dowel-pins removed.    2, Holes for hex-


^*^" --'


Fig. 8.—Turbine Spindle, weight 35 tons, of 2o,ooo-kw. Turbine


little red lead mixed with thin machine oil on the journal, and then rubbing
it uniformly over the journal till almost dry. The bottom half-bearing is
then put on the journal and rocked backwards and forwards a few times;
the high or " hard " spots of the bearing will be marked with red lead, and
must be carefully scraped down with a curved scraper; the red lead should
be smeared uniformly over the journal before the bearing is again marked.
This process is continued till the bearing is marked pretty uniformly. The
top half-bearings are also tried for marking, and the hard spots removed,
but the marking process is not carried so far, or is so complete, as in the
case of the bottom halves. (An exception is made in the case of bearings
for reciprocating plant, where the pressure comes on the top and bottom

half-bearings alternately, and the
necessity, therefore,1 exists for
the marking and scraping of both
halves to be done very care-
fully.) After all the bearings
have been scraped, the bottom
halves are put in position in the
pedestals or housings, and the
centre line, previously used, is
again stretched through in order
to align the actual bearings.

The bearings of high-speed
turbine plant are invariably pro-
vided with means for adjusting
the bearing relative to the pe-
destal or housing, both vertically
and horizontally. The vertical
adjustment is usually made by
means of liners, both at the top

and the bottom of the bearing: the adjustment sideways is made either by
liners or by two tapered steel wedges on either side. The bearings, also,
are frequently made self-aligning by the provision of spherical seatings
in the housings, but this degree of self-alignment is slight, and is provided
simply to allow the bearings to take up a comfortable position on the
journals, and to remove stresses due to slight inaccuracies of alignment,
or due to alteration of alignment, caused by stresses in the castings or
settling of foundations, &c. One of the most convenient and widely used
bearings is the padded bearing, on which there are packets of thin liners
at the top, and bottom, and sides, consisting of sheets of steel, varying
in thickness from -005 up to -025 in., each set of liners being covered by
a steel pad, through which screws pass and secure the pad and liners to
the bearing. By removing a liner, say -005 in., from one side pad to the
other, the whole bearing is moved over -005 in.; in the same way the bearing
can be raised or lowered by very fine stages.

Before the turbine spindle is put in position, it is advisable to raise the

Fig. 9.—Padded Bearings, showing Adjusting Liners
behind Steel Pads


bottom half-bearings, say -gV in,, by means of the bottom pads, previously
described; this is a wise precaution to prevent damage to the brass labyrinth
packing strips which line the H.P. and L.P. glands, diaphragm collars, &c.,
to prevent leakage of steam out, or air in. Some oil should be poured over
the shaft journals and into the bottom half-bearings; the shaft should then
be carefully lowered till resting in the bearings, special care being taken
during this lowering operation that the wheels are clear of the fixed diaphragms
between them. The thrust block bottom half should then be put in position
and wedged up temporarily with wooden wedges, to prevent any axial move-


II Jill:

Fig. 10.—Top Half of Turbine Cylinder turned over for fitting of Top-half Diaphragms, Glands, and Nozzles
ment. The shaft and wheels are then slowly revolved by hand, or if neces-
sary by means of the crane pulling on a rope previously wrapped round the
shaft. If the brass labyrinth strips are touching the shaft, they will mark
the shaft with a fine yellow line. Before lifting the shaft to scrape the laby-
rinth strips, it is advisable to fix the vertical guide-strips in position to prevent
the wheels fouling the diaphragms, and the shaft is then raised and the brass
labyrinth strips scraped and sharpened up again where the shaft has rubbed
them flat. The shaft is then lowered again into its final central position by
previously removing the ^V in. of liners added to the bottom pad under the
In order to put the half-diaphragms in the top half-cylinder, the latter is
turned upside down and the diaphragms then put in position and secured by
" keeps " held in by screws. Before the diaphragms are put in position,
.;}                             256                              HEAVY   MACHINERY
• iii
\\l                                   any nozzles in the top half should be bolted in position and properly jointed
'j}j                                  up.    The completed top half-cylinder is then turned back into the upright
sj                                  position, and carefully lowered down into place on the bottom half.    Guide-
$1                                  rods are usually provided for guiding down the top half, and it is advisable
',ifj                                  to put strips of sheet metal -^ in. thick and 3 or 4 in. wide at several places
'';                                  round the main horizontal joint.    This is done in order to try the spindle
(!f!                             for being free before the top half is right down in position.    The spindle is
||!                            then pulled round once or twice by hand or by the crane, and if quite free,
'^                            the pieces of sheet metal can be removed from the horizontal joint, the
jj                            weight of the top half being meanwhile taken by the crane.    The top half
I                            is then lowered right down and the spindle pulled round again.    The cover
is then finally lifted, and the edges of the brass labyrinth packing strips sharp-
ened up. The setting of the thrust block, which determines the position
axially of the turbine spindle, should now be carefully adjusted, and the
axial clearances between the wheels and the fixed diaphragms very carefully
checked over at both sides of the turbine. A record should be kept of these
clearances, and if they are less than the clearances required by the particular
type of turbine being erected, it may be necessary to have the fixed diaphragms
further machined; special care should also be taken to observe and accurately
measure the clearance between the nozzles and the blades on the first wheel.
r J                                 When the spindle has been finally set, the permanent collars for securing
**                              the thrust block in position axially can be machined to the required thickness,
and either pinned on to the thrust block by two or three countersunk screws,
or left loose and tapped round into position.
Having prepared the main horizontal joint of the turbine cylinder and
spread the jointing material uniformly, the top half is lowered into position,
dowel- or steady-pins driven in to fix the relative positions of top and bottom
half-cylinders, and the bolts through the joint then put in, and either banged
up tight with a large spanner and heavy sledge-hammer, or pulled tight with
a heavy spanner over the stem of which is passed a piece of heavy pipe
several feet long, for additional leverage.
When the supply of steam is available, it is advisable to go round all
joints which reach a high temperature under working conditions, and
tighten up the bolts after the joint has been heated up. The effect of heat
causes the jointing material in the majority of cases to " give ", and this
give should be taken up on the bolts through the joint, otherwise the joint
will very probably begin to blow in a very short time.
The next operation is to erect the alternator and couple it up to the
turbine. In the very great majority of cases the alternator stator is built
in one piece, i.e. not split horizontally, the exceptions being few and far
between. The advantages of making the stator without a horizontal break
are very great, both from the electrical as well as from the mechanical stand-
point, so that even the largest alternators are made in one piece. This
means, therefore, that the generator rotor has to be threaded through the
stator, and as this is an operation which has puzzled many engineers, a
description is given of the method employed.





the                                            is roughly levelled up on the
or               in case the engine-room floor Is not designed
to                                  The journals of the generator rotor are carefully
and the fop and             half-bearings for each end bedded down,


as described previously.   The shaft journals are then cleaned, a little oil
rubbed round the journals, and the rotor put in position in the outboard

ist position—Rotor with central lift lowered into outboard pedestal and packed up at turbine end

2nd position—Pedestal cover bolted on. Pedestal forms an out-of-balance weight which enables
rotor to be slung out of the geometric centre. Chain blocks can be hung on crane hook and help
to support pedestal if necessary

3rd position—Sling up against end of generator stator.    Outboard pedestal lowered on small steel rollers

and turbine end packed up

	----- ,

	i —
	— 1

	i — f
	j i

	J     i-
	i L_,
	i -a!n~"s_- _=• j= •s'.rvnp
	1 ,-J
	-U:      ^_^

	*~W                        j/"

4th position—Turbine end raised, rotor jacked or pulled on rollers into final position
Fig. 12
pedestal, the other end of the shaft being packed up on baulks of timber,
&c., as far as possible dead level. The top half-bearing and pedestal cover
are then put in position on the outboard pedestal, but the bolts securing
the cover are left about •&• in. slack. The rotor, together with outboard
pedestal, is then slung on the crane, care being taken to arrange the sling
(a wire rope is best) in the manner shown.   It is
forms an out-of-balance weight, which brings

Fig. 13.—Stator Shell of zo,ooo-kw. Turbo-alternator, single casting, weight 28 tons, showing
steel trunnion pins for turning casting on end
the geometrical centre towards the pedestal, and thus leaves one end specially
long, as shown in the sketches. Great care must be taken to get the whole
piece dead level when hanging on the crane, a spirit-level being used for the
purpose, held on the journal or parallel part of the shaft, the point of support
VOX..   I                                                                                                                                                                           IT" O
260                               HEAVY  MACHINERY
(i.e. where the wire sling grips the rotor) is moved until4he balance is obtained.                   '?
Sometimes, in order to achieve this, it is necessary to hang a set of chain
blocks on the crane hook and support the pedestal slightly by tightening
up the chain blocks. Having thus got the rotor slung level, it is an easy
matter to thread the piece through the stator until the rope sling supporting                   ,
the rotor, &c., is nearly up against the stator winding. A support of steel
beams or baulks of timber should previously have been arranged to support                   i
the turbine end of the generator rotor, and the rotor is then carefully lowered;
some ten or twelve, or more, steel rollers should be placed under the out-
board pedestal, and the latter lowered down on to them. These rollers
should be made of \ in. diameter steel rod, in lengths a little greater than
the width of the pedestal. The sling is then removed from the body of
rotor, and the turbine end of the rotor shaft is supported by the crane, and
a slight endwise pull applied by the crane. The pedestal end will roll on
the small rollers, and, when the generator and turbine couplings are together,
the generator bearing (turbine end) can be put in pgsition and the rotor
lowered into it. It only remains to lift the outboard pedestal end of the
rotor with the crane and remove the small rollers. It is the practice nowa-
days to place a sheet of insulating material, e.g. fuller-board, leatheroid, &c.,            . ,
under the outboard pedestal, and to insulate from the pedestal, by means
of insulating tubes and washers, the bolts which hold the pedestal down                   I
to the bedplate. This is done to prevent the circulation of stray currents
through the rotor shaft, pedestals, and bedplate. Under certain circum-
stances these stray currents reach high values, and the effect on the generator %
is to cause pitting of the journals and white-metal bearings, and the breaking
up of the oil passing through the bearings, with the formation of acid, which
in turn causes further corrosion.
Before closing up the bearings and bolting down the bearing pedestal                   \
covers, it is most important that the clearances between the bearing and
journal for oil be accurately measured, and if necessary increased to a safe
figure, and at the same time the fact be definitely established that the cover
is actually binding down the bearing inside it. The white-metal lining of
the bearings should be scraped away carefully at the sides (see fig. 15) for
a sufficient distance down, so as to leave the actual bearing area—that area
contained in an angle of about 120°; this side clearance should not be less
than five-thousandths of an inch, and it should be possible to get a feeler
gauge down on each side, all along the bearing. The clearance between
the top of the journal and the bearing is obtained by putting two or three
strands of soft lead wire across the journal and bolting the top and bottom                   j
half-bearings tightly together; on opening out again, the lead will be found
to have been flattened out to the exact clearance on the top of the bearing,
and this thickness can then be accurately measured by a micrometer gauge.
This clearance varies with different makers; an average figure is about
i mil. ('ooi in.) per inch of journal diameter, and if necessary the inside of
the top half-bearing should be scraped away carefully to obtain the necessary
uniform clearance.
Fig. 14.—2O,ooo-kw. Turbo-generator in course of erection, showing special method of lifting 7o-ton
stator by two cranes, as neither crane by itself could lift the piece
262                              HEAVY  MACHINERY

In order to determine whether the pedestal cover is binding down the
bearing, immediately before the cover is put on, a strand or two of soft
lead wire is laid across the top pad or outside machined surface of the bearing
and the pedestal cover then bolted down; if, on lifting the cover again, the
lead wire is not flattened down " to nothing ", additional liners must be put
under the top pad, and the test repeated till the desired result is obtained.
Unless this precaution is taken, there is liable to be a considerable amount of
vibration when the plant is running, and this will result in " hammering "
of the bearings and the running of the plant will get rapidly worse. In in-
vestigating vibration troubles in high-speed plant, it is always wise to
examine the bearings first of all, for clearance and for tightness in the

Before finally closing up the bearings, oil should be pumped through
the lubricating system in order to see that each bearing is receiving an ample
supply of oil. When tunning up a set such as described above, the utmost
care should be taken. As soon as the spindle just starts to move round,

the engineer in charge should have a quick run
round in order to locate any unusual noises,
sign of smoke, or evidence of heat, and be pre-
pared to shut down instantly if necessary. The
running-up for the first time frequently takes
several hours, during the greater part of which
the set is being run at slow speed, which is
very gradually increased; this gives any trouble
time to show up at lower speed, and gives the
man in charge a better chance to avert trouble

Fig. 15.—Bearing showing Sides          .         •/• ^i        i   r    ^ •      i                      ,rn            i

scraped away for Oil Clearance         than if the defect IS Shown Up at full Speed.

Joints.—In considering the best type of

material for making any given joint, while the principal point is the
tightness of the joint under working conditions, due regard must be paid
to the time and labour involved in breaking the joint, removing the old
material, and remaking the joint when such a course becomes necessary,
e.g. opening up or dismantling machinery for inspection. There are some
materials which make excellent joints, but which are removed only with
the very greatest difficulty.

There are a large number of jointing materials on the market to-day,
for all of which special advantages are claimed. Some of these materials
are in sheets, others in powder or paint, and some in metal, wire, sheet,
and net, and nearly every engineer has his own particular method of making
any given joint, which he claims is superior to any other method; only a
general statement, therefore, can be made as follows:

For Steam-joints (flanged).—The joint faces are very carefully scraped,
bedded on a small, portable, plane table, and, before being bolted together,
are wiped over very lightly with graphite or some graphitic paint, or even
left without any jointing material at all. Alternatively, a Taylor corrugated
joint ring is used, which has previously been filled with one or other of the



jointing materials in paint or putty form. For joints in pipes carrying super-
heated steam,the joint rings should be made of corrugated nickel. Alternatively,
a joint ring can be cut from a sheet of jointing material and, before being
put in position, painted on both sides with thin graphite paint—or a ring
can be cut out of thin copper gauze, and the latter then thoroughly filled
with red-lead putty, or other jointing material, before being put in position.
For low-pressure steam joints the latter is a favourite method, the addition
of a strand or two of lead wire threaded round the gauze adding to its effi-
ciency. In the steam systems of collieries, where low-pressure steam is
used without superheat, ordinary rubber joint rings are frequently used
with success on systems up to 100 Ib. per square inch.

The main joint between the top and bottom halves of a turbine cylinder
is usually made by smearing jointing material of the consistency of thick
cream on the bottom half, and
then adding a strand or two of
lead wire at the low-pressure
end, and soft copper wire, about
No. 27 gauge, at the H.P. end,
and then bolting the top half
down solidly.

It is advisable in the case of
all steam-joints, or joints where
the temperature is likely to be
high, to go round all the bolts in
the joint as the temperature is
being raised. This is known as
" following up " the joint, and it
is invariably possible to get an
extra turn or half-turn on the
bolts and nuts when the joint is
heated up. If this is not done,

there is a danger of the jointing material being blown out and the joint
having to be remade, and in some cases necessitating a shut-down.

For Joints in Water-pipes, rubber insertion is used mostly; the rubber
joint ring should be put on dry; some men smear the rubber with tallow
or grease, hoping to make a more effective joint, but grease and oil only result
in rotting rubber and should therefore not be used.

An excellent joint for flanged water-pipes can be made by a ring of thin
copper gauze filled in with red-lead putty, and a strand of lead wire threaded

Oil Joints.—Special oil-jointing material in sheet form makes the best
joint; this consists of a strong paper boiled in soft soap and caustic soda.
Alternatively, ordinary steam jointing material of the asbestos-sheet type is
frequently used, but care is taken to paint the joint ring with shellac dissolved
in methylated spirit, immediately prior to being bolted up. For large, flat
surfaces, soft soap smeared thinly over the surfaces, and a piece of lead wire

Fig. 16.—Copper Gauze Joint Ring with Lead Wire
woven in and filled with Red-lead Putty
264                             HEAVY  MACHINERY
laid on, makes an excellent joint. For jointing together large, flat, machined
surfaces of condensing plant, red and white lead, thoroughly mixed together
into a thin cream by the addition of gold size, is excellent, but it is essential
that the ingredients be thoroughly mixed together; the joint can be improved
by laying in a strand or two of tubular cotton packing. For very large,
flat joints a ribbon of asbestos from i to 2 in. broad, laid on a smearing of
red and white lead applied to both top and bottom surfaces, makes a very-
tight joint.
Use of Cranes and Lifting-tackle.—For light lifts, hemp-rope slings
are the handiest to use; it is also an advantage to have several lengths of
rope of different sizes (not slings), so as to be able to make a sling to any
given length, but in tying the two ends together care should be taken to
always insert a wooden pin (preferably 3 to 4 in. diameter and tapering
down) in the knot; unless this is done it may not be possible to open the
knot again, once a heavy weight has been lifted and the knot pulled tight;
the taper-pin is easily knocked out and does not lessen the security of the knot.
For heavier lifts, wire-rope slings should be used in preference to chains;
the latter are liable to crystallize when in constant use, and eventually snap
off short; this is particularly the case in frosty weather. If chain slings are
used, they should be annealed at regular intervals, e.g. every three months,
by heating to a dull red and allowing to cool slowly. Another objection to
chain slings is that the links are liable to press into and damage the machined
surfaces of the plant being erected, and in this respect particularly wire-rope
slings are much to be preferred, though even wire ropes will press into and
indent highly finished surfaces, such as the journals of heavy shafts, and it
is usual, therefore, in designing such shafts, to make special provision of
space at each end where the lifting slings should be placed.
In lifting a heavy piece the following precautions should be observed:
*i. When the crane or lifting-block has been tightened up, so as to just
put some tension on the slings, see that the crane ropes are vertical when
viewed both from the front and side; unless this is done the piece will swing
to one side when lifted, and, in the case of a heavy lift, considerable damage
may result.
2.  The slinging should always be arranged so that the centre of gravity
is below the point of support, i.e. the point or points from which the piece
is suspended.   If the centre of gravity is above the point of support, the
piece may capsize and fall out of the slings when being raised or lowered.
3.  If one end of the piece lifts before the other, the sling at the end
which lifts first should be lengthened, or the other end shortened.   The
most convenient way to lengthen a sling is by means of a series of shackles
each with removable pin.   For shortening a sling, pieces of timber or wooden
wedges can be placed between the sling and the piece.
4.  Whatever type of sling is used, it should always be protected from
damage where it passes over sharp edges, corners, &c.; fillets of sheet iron
or thick lead are used, or several thicknesses of sacking.
5.  When the piece has been lifted, say, an inch clear, the lifting operation


should be stopped in order to see that the crane brakes are in order, or that
the chain lifting-blocks will not run back. The piece should then be raised
another inch, in order to see that the crane can start lifting with the load
on it. It is unfortunately too often the case that the controlling arrange-
ments of an electrically operated crane will not allow the crane to start against
heavy load.

6.  The piece should then be lowered an inch, to see that the lowering
can be stopped, and that the piece is thus properly under control.     By
carrying out a few simple precautions of this kind much trouble and damage
can be averted, as, if the crane should fail to hold the piece, it can only run
down an inch or two at most.

7.  When everything is satisfactory, and the brake blocks, &c., adjusted
if necessary, the piece can be lifted into position, though care should always
be taken when lowering to see that the crane and lifting tackle are not subject
to shocks due to suddenly stopping.