Full text of "Spot and arc welding"

See other formats

Google

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project

to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the

publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to
prevent abuse by commercial parties, including placing technical restrictions on automated querying.
We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for
personal, non-commercial purposes.

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the
use of public domain materials for these purposes and may be able to help.

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find
additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner
anywhere in the world. Copyright infringement liabili^ can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at |http: //books .google .com/I

^ East Eng
Library

SPOT AND Arc

WELDING

H.'AC HORNOR, B.A.

ILLUSTRATIONS

PHILADELPHIA AND LONDON
J. B. LIPPINCOTT COMPANY

PRINTED BY J. B. LIPPINCOTT COMPANY
EAST WASHINGTON SQUARE PRESS
PHILADELPHIA, U. S. A.

\
I

To Mt Wife
MARIE M0R3E HORNOR

WITHOUT WHOSE AFFECTIONATE AND UNTIRING
AID THIS BOOK WOULD NOT RAVE BEEN WRITTEN

I I

PREFACE

Although electric welding has been used for many
years for repair work, there exists to-day a hesitancy in
applying it to new construction, especially to the joining
of heavy steel parts. It is the purpose of this book to
endeavor to dispel this apprehension. The data of tests
made in the spot welding of heavy steel plates by the
Emergency Fleet Corporation is furnished in full with
the expectation that in making this information avail-
able to shipbuilders and others, they will be able to adopt
this process to their own manufacturing advantage. In
like manner the underlying question of arc- welding proc-
esses is fully discussed with the intent of reassuring
any one who may doubt the ability of these methods to
supersede the rivet.

The author wishes to acknowledge the courtesies ex-
tended to him both by those who have contributed to the
text and those who have permitted the reprinting of
valuable published material.

Minerva, N. Y. H. A. HoRNOR.

March, 1920.

CONTENTS

CHAPTER PAQlJ

I. Materials 1

II. Electbic- Welding Systems 14

III. Spot Welding 24

IV. Demonstration of Heavy Spot Welding 38

V. General Applications of Arc Welding 97

VI. Discussions on Arc Welding 105

VII. The Arc Welder 134

VIII. The All-Welded Ship 160

IX. Theories of Electric Welding 178

Appendix I 218

Appendix II 220

Appendix III 224

Appendix IV 231

Appendix V 253

Appendix VI 276

Appendix VII 285

SPOT AND ARC WELDING

CHAPTER I

Materials

The essential consideration for those who practice
the art of electric welding is the character of the mate-
rials to be joined. As a matter of fact, this principle has
been the basis of all successful engineering achievement
and a lack of it the cause for many vital failures. The
wide range of processes combined with ingenious designs
permits the broad statement of probability that all the
metals and their alloys may in varying degrees be held
together by electric methods. This provides a varied
and interesting fi'^ld for the engineer although, like un-
tilled ground, there will be found many obstacles in the
way of cultivation. The success of the harvest will be in
direct proportion to the thought bestowed upon the
actions of the metals when under the treatment required
for their jointure.

It is not proposed to consider here the welding of the
soft metals such as gold, silver, copper, tin, zinc, alumi-
num, etc. It is sufficient to affirm that in small articles
processes of electric welding have been devised that sat-
isfactorily accomplish such joints as are needed for spe-
cial industries. The processes employed are fully de-
scribed in many published articles/ As metal is added to
the objects joined the method is referred to as soldering
rather than welding. On the other hand, nickel, copper.

1 a

Electric Welding," by D. T. Hamilton and Erik Oberg, 1918.
1 1

)
2 SPOT AND ARC WELDING

and aluminum wires of small diameter are joined by
means of a condenser spark, the process being called per-
cussion welding. Interesting as these applications are,
only this brief mention can be made in order to allow a
full opportunity for the discussion of the welding of
heavy steel pieces.

Definition. — ^Doctor Howe gives this definition of
steel in its specific sense: " A compound of iron possess-
ing, or capable of possessing, decided hardness simul-
taneously with a valuable degree of toughness when hot
or when cold, or both. It includes primarily compounds
of iron combined with from, say, 0.30 to 2 per cent, of
carbon, which can be rendered decidedly soft and tough
or intensely hard by slow and rapid cooling respectively,
and secondarily, compounds of iron with chromium,
tungsten, manganese, titanium, and other elements, com-
pounds which like carbon possess intense hardness with
decided toughness.^ " The reasons underlying the be-
havior of this material when subjected to temperature
changes must be sought for in the researches made by
eminent metalliu-gists. This is the broad subject of the
heat treatment of steel with which the practical welding
engineer should have some knowledge in order to escape
the error of applying electric welding to the detriment
of the original materials.

Production of Iron. — Iron ore when taken from the
earth may contain many and various minerals. The ore
is treated in a blast furnace for the purpose of removing
the oxygen which had a persistent affinity for the pure
iron. The product of the blast furnace is not c hemically

""MetaUurgy of Steel," by H. M. Howe, vol. i, chap, i, p. 1, ^d
Ed., 1891.

MATERIALS 3

pure, consisting usually of iron, carbon, silicon, manga-
nese, sulphur, phosphorus, and oxygen. Pure iron may
be obtained by careful preparation in the laboratory but
not for commercial purposes. This molten metal from
the furnace is cast into a form called " pig iron." When
pig iron is remelted in a crucible and cast into some
commercial form it is called " cast iron."

Cast iron takes two forms, depending upon its treat-
ment when poured. If the molten metal is cast in sand
it is a grey iron casting, if it is cast in metal moulds
(chilled) it is a white iron casting. The point to observe
here is the difference caused by the slow and rapid cooling
of the molten metal. The names given to these castings
are taken from the appearance of the fracture caused by
the large amount of free carbon in a case of grey cast
iron and the small amount of free carbon in the case of
white cast iron.

Annealing is effected by slow cooling from a high
temperature. If then white iron castings, which are
brittle as compared to grey iron castings, are annealed so
as to free the carbon which is in a combined state, mal-
leable castings will result. Malleable iron is free iron
with which is mixed carbon in a free state in the form of
graphite. The effect of annealing does not penetrate the
mass of metal and annealed castings seldom show any
effect much further than an inch below the surface.

When pig iron is melted in a puddling furnace at a
point where the pure iron appears to separate from the
mass of impurities it is removed and the slag that it
tenaciously carries with it is squeezed out by means of
rolls. The rolls natiu-ally form the mass of metal into the
shape of bars. This mechanical treatment of the iron

4 SPOT AND ARC WELDING

tends to produce purity and its efficiency is in direct ratio
to the quality of the resultant wrought iron. From the
mechanical treatment it receives it derives the name of
wrought iron. The purity also bears a direct relation to
the ore, and for this reason imported iron ( Norway and
Swedish) is well known for this quality. Wrought iron
because of its purity is used in the manufacture of high-
grade crucible steel.

Steel Processes, — The essential difference between
the manufacture of steel by the open-hearth or crucible
processes is disclosed by their names. In the latter case,
iron of like or unlike carbon percentage is melted in
crucibles usually totally enclosed, the molten mass is
held in these crucibles (either graphite or clay), so that
it may absorb silicon from the crucible walls, and then
the liquid metal poured into castings or ingot forms.
The open-hearth process melts the pig iron in a cupola,
transfers it to an apparatus whereby the impurities are
removed after which desired elements may be added to
obtain various properties in the finished steel. The
process is referred to by steel makers as basic or acid in
accord with the chemical nature of the lining of the
vessel used for removing the impurities. Crucible steel
is of high grade, more expensive of manufacture and
more costly. It is used for cutting tools, springs, fire-
arms, etc. The added elements in the open-hearth
process may be varied in many ways, thus producing
many steel alloys useful to the industries. For ordinary
plates and shapes the controlling elements are carbon
and manganese. Both these elements hold high favor
due to their ability to add greatly to the tensile strength
and ductility of steel.

MATERIALS 5

From the steel furnace by either process the metal is
cast either into moulds for steel castings or ingot moulds
for rolling into plates or shapes or into billets for forg-
ing. It should be carefully noted that plates, with which
the electric-welding engineer is much concerned, are
really steel castings which have received additional tem-
perature and mechanical treatment in proceeding
through the rolling mill. In applying the processes of
spot and arc welding to new steel construction it will be
such material that will require successful joining. As
heat is the prerequisite for the welding of the metals in
any case it becomes necessary to give close attention to
the theories of the heat treatment of steel, the metallurgy
of steel. This subject will be taken up in a later chapter.

Chemical Constituents of Steel. — Without proceed-
ing into a maze of theoretical argument it may be well to
consider briefly a few opinions regarding the effects of
the various elements added to steel and following that
their composite action on the weldability of steel.

" As the carbon increases, the tensile strength, elastic
limit, elastic ratio, and compressive strength increase
within limits; the fusibiUty, hardness, and hardening
power increase perhaps without limit; while the malle-
ableness and ductility, both hot and cold, and the welding
power diminish apparently without limit. The modulus
of elasticity appears nearly independent of the percent-
age of carbon, at least within the limits of carbon zero to
carbon two per cent." ^

Silicon is considered by some authorities to cause brit-
tleness and redshortness, and by others to be in many

«" Metallurgy of Steel," by H. M. Howe, vol. i, chap, ii, p. 13, 2nd
Ed., 1891.

6 SPOT AND ARC WELDING

cases harmless, sometimes even increasing the ductility
and in the presence of manganese " to counteract its
tendency to cause redshortness." *

Phosphorus is an undesirable element. It is well
known to cause " coldshort " or brittleness. The steel
manufacturer takes every care to remove as much of this
element as possible so that only a small percentage is
found in good commercial grades of steel. It is negligi-
ble as far as welding problems are concerned.

The same remarks apply to sulphur. Howe states :
" Sulphur has the specific effect of making iron exceed-
ingly brittle at a red heat and of destroying its
welding power." ^ ,

The effect of manganese on steel is still in the regions
of dispute. One of the best authorities states: " The net
effects of manganese on tensile strength and ductility
are slight." Manganese seems to promote continuity
and in this manner aids the ductility of steel. This is an
important characteristic of this element which may
throw light upon its use in the composition of electrodes
for arc welding.

Chromium gives hardness to hardened steel probably
increasing the tensile strength and elastic limit. The
weldability of steel is reduced by chromium.

The addition of tungsten tends to produce great
hardness in the steel alloy. This characteristic which
it retains at high temperature makes it useful for the
manufacture of high-speed cutting tools. Tungsten

* " MetaUurgy of Steel," by H. M. Howe, vol. i, chap, iii, p. 40, ^d
Ed., 1891.

"^ " MeMlurgy of Steel," by H. M. Howe, vol. i, chap, iii, p. 52, 2nd
Ed., 1891.

MATERIALS 7

steel is brittle and is diflSeult to forge even at relatively
high temperatures. Doctor Howe doubts if it can be
truly welded by ordinary methods.

Nickel steel surpasses the best carbon steel in its
superior tensile strength combined with elongation.
Upon its first introduction it was found difficult to ma-
chine, but this disadvantage has been overcome by im-
provements in machine tools. This alloy is also less
liable to corrosion than steel.

Vanadiima influences stieel in much the same manner
as nickel. Its first introduction was heralded by claims
which soon disappeared in the presence of tungsten steel.

Copper and iron act curiously in combination. A
httle copper with iron or a little iron with copper are
said to unite into a homogeneous mass, but if the propor-
tions approach equality they seem to split up into alloys.
The effect of copper is like that of sulphur, causing red-
shortness, brittleness, and an opposition to welding. This
tendency of the metals must be carefully held in mind
by those who seek to improve the non-corrodibility of an
arc-welded seam by the introduction of copper into the
composition of the electrode material.

Many other metals, such as zinc, tin, lead, titanium,
arsenic, cobalt, aluminum, may occur in iron but disap-
pear in the manufacture of steel. Those that have been
treated in brief detail above are to be found in commer-
cial steel and must be investigated by those who are de-
sirous of making a successful application of the practical
processes. of electric welding. For those who wish to
investigate and experiment with steel compositions for
the betterment of any special welding process there re-
mains the action of the alkaUne earths and the combina-

8 SPOT AND ARC WELDING

tion of iron with the noble metals. This latter field
attracted the attention of some of the older and illus-
trious scientists without material success, but the sign of
their failures may by modem methods be turned to a
mark of attainment.

Weldability of Steel.— The important consideration
for the welding engineer is the degree of weldability of
the metals placed before him to join. Not only is this
knowledge of great concern prior to the performance of
the work, but also is necessary for the investigation of
immediate and subsequent failures. Commercial steel
castings upon analysis show a general chemical composi-
tion as follows :

Carbon 0.35 per cent.

Silicon 0.40 per cent.

Manganese 0.80 per cent.

Phosphorus 0.05 per cent

Sulphur 0.05 per cent.

Of these five chemical elements the proportion of
phosphorus and sulphur are so slight as to make them
negligible. Among technicians there is argument re-
garding sulphur. The extreme is that 0.02 per cent,
sulphur opposes welding, the other extreme that suc-
cessful welds can be made with sulphur as high as 0.07
per cent. Sulphur as high as 0.15 per cent, is consid-
ered " quite unweldable." There remains the three prin-
cipal constituents, carbon, silicon, and manganese, of
which many scholars give carbon the leading place in
lessening the weldability of the alloy. This opinion is
largely borne out in practice. The whole subject resolves
itself into one needing at the present time much
earnest study and careful research, especially when con-

MATERIALS 9

sidered with the complications introduced by the arc-
welding process.

As regards manganese and silicon so little of their
reactions are known that no trustworthy statements can
be made other than those quoted and referred to iabove.
Doctor Howe says of silicon: " The good welding power
of crucible steel, usually rich in silicon, goes to show that
silicon is not especially injurious in this respect.'' ^ For
the benefit of the practitioner and the art as a whole it
would be well for some of our colleges to engage in this
industrially useful field of research.

The carbon content has received more careful and
substantial investigation with the result that more har-
mony of opinion exists. The same authority states: " It
was formerly thought that the presence of a little carbon
was indispensable or at least very favorable to welding ;
but this, I think, is no longer believed. Certain it is that
in general the difficulty of welding increases with the
proportion of carbon, and the welding power probably
practically disappears when the carbon rises above 1.3
per cent. The larger the proportion of other elements
present, probably the lower, in general, is the welding
power for given percentage of carbon. Thus the weld-
ing of apparently common Bessemer steel is said to be
hardly possible with 0.20 to 0.35 per cent, of carbon, and
impracticable with 0.35 to 0.50 per cent.; while to the
practiced worker the welding of the relatively pure
crucible steel is said to be easy with 0.87 per cent., and
possible, using the greatest care, with 1.25 per cent, of
carbon. Though the difference is probably much less

* " Meta;llurgy of Steel," by H. M. Howe, vol. i, chap, xiv, p. 252, 2nd
Ed., 1891.

10 SPOT AND ARC WELDING

than this rather loose wording implies, and though there
are welds and welds, it appears to be very marked.

" A reason why rising carbon lowers the welding
power is that it lowers the point to which we can heat
the metal without danger of burning, but does not lower
correspondingly the temperature at which plasticity sets
in; indeed, it seems to diminish the plasticity and ad-
hesiveness for given temperatures."^

Doctor Howe is here treating purely of the weld-
ability of steel in general and his statements, although in
a large degree apply, must not be construed as referring
to the electric weldability of steel. To indicate his agree-
ment with others and subjecting the question to a spe-
cific reference to electric welding the following opinion
is also quoted: " Little is known at the present time re-
garding the effect of the weldability produced by the
presence of most of the impurities given above, where
the electric arc-welding process is used. No data has
been published on the subject. It is known, however,
that steel containing 0.5 per cent, or more carbon is sub-
ject to " burning " at much lower temperatures than
low-carbon steels. This fact can readily be observed in
arc-welding practice, i.e.j the tendency being towards
" burnt " metal in the weld. The observations which
have been made up to the present time seem to indicate
that the tendency toward " burning " shown in steel of
comparative high-carbon content, is the only considerable
effect which is produced on the weldability by the pres-
ence of any of the impurities in their usual amounts."

Physical Characteristics. — The foregoing opinions

' " Metallurgy of Steel," by H. M. Howe, vol. i, chap, xiv, p. ^51, ^d
Ed., 1891.

MATERIALS 11

deal with the chemical nature of steel and are given con-
cretely with the hope that the practitionfer may find them
convenient in his daily application of electric welding.
Of an equal degree of importance are the physical
changes that take place in steel when subjected to high
and low temperature, sudden and slow changes of tem-
perature. Here is to be considered not the composition
nor the changes of composition of the material, but
simply its physical properties.

Steel is hardened by rapid cooling from a high tem-
perature. This is practically attained by quickly im-
mersing the heated steel in a bath of water or oil. The
kind of liquid used to quench the steel develops a greater
degree of tensile strength and different methods must be
pursued for difference in carbon content. Thus it is
stated that to give the highest tensile strength to low
carbon steel it should be quenched in water from a high
temperature; for mild carbon steel (0.40 per cent.) it
should be quenched in oil with a " rather high quenching
temperature ; " ® and for steel with large percentage of
carbon (1.25 per cent.) it should cool slowly from a low
temperature and immersed in oil. In general the hard-
ening of steel brings about the desirable physical char-
acteristics of better elastic limit, greater firmness, and
better tensile strength. The ductility tends to diminish
as well as the specific gravity. Steel may be quenched
in other media such as tallow, coal tar, and even lead,
but tensile strength is not bettered by these methods over
the use of oil.

The tempering of steel is employed to modify to
some extent the previous effects of hardening. This re-

*" Metallurgy of Steel," by H. M. Howe, vol. i, chap, ii, p. 19, 9nd
Ed., 1891.

1« SPOT AND ARC WELDING

quires the reheating but to a lower temperature and then
in general cooling quickly and in some cases slowly. As
hardened steel is brittle tempering increases its ductility
and makes it much tougher. It does this without decreas-
ing the tensile strength and, in fact, it has been stated
to increase the tensile strength. Though the ductility is
bettered by tempering over that of the hardening process
it is still not as ductile as annealed steel. The advantage
of tempering which apparently is an imnecessary second
operation lies in the better control of temperature with
the result that hardness and tensile strength are not im-
paired but to them is restored the ductility lost in the
hardening operation.

The annealing of steel is to completely imdo the
effects of hardening and bring the steel to a very soft
and tough state. " It increases the ductility and specific
gravity and it generally lowers the elastic limits." ° Be-
sides restoring the tensile strength of violently hard-
ened steel annealing also restores the tensile strength
caused by cold working and internal stresses in steel cast-
ings. Much of the advantage gained by annealing re-
sides in the cooling temperature and the mediima used
for the cooling process. It may be stated broadly that
slow cooling brings about the softer and tougher quali-
ties desired, but for many purposes "both cold-worked
low-carbon steel and steel castings the cooling may be
interrupted at a certain point and the material quenched.
This procedure will not give results equal to full slow
cooling, but often will serve the purposes required.

It is thus seen that iron and steel in many shapes and
with a great variety of physical properties will be laid

«" Metallurgy of Steel," by H. M. Howe, vol. i, chap, ii, p. 24, 2nd
Ed., 1891.

MATERIALS 13

before the electric- welding engineer, and it will depend
more or less upon his familiarity with this wide range of
conditions how he will meet and solve the problems. This
brief resimie from the mining of the iron ore to its fabri-
cation for use in the mechanic arts is a necessary prelimi-
nary to the study and practice of the joining of steel by
electric power. To materials that have already been
subjected to modifications in their chemical constituents,
to changes in their physical properties by the application
of different temperatures, and to internal mutations of
their own making — to these materials are applied chemi-
cal reactions, and physical effects of like nature to obtain
a connection which will either approach or exceed the
advantageous characteristics of the original metal. It
is mainly upon this point that objectors to electric weld-
ing rest their argument : that no joint can be equal to the
original material. It is upon the same point that those
who are favorable toward and well acquainted with elec-
tric welding uphold their belief in the process because
the e^ddence has accumulated to a degree which permits
the statement: that electric- welded joints can be made
which will be stronger than the metals joined.

Summary. — This plainly places before the engineer
a new problem, namely, whether he wishes his jointures
to be more or less lasting than the materials which he is
using for a given purpose. On the other hand, with this
new abiUty to secure a more favorable joint will he not
economically and safely reduce the amount of materials
to secure the same result? These engineering problems
are closely knit with a study of the metallurgy of steel
and the modern apphcations of electricity to the weld-
- ing of steel.

CHAPTER II

Electric- WELDING Systems

Fundamental principles separate electric-welding
processes into two distinct groups: resistance methods
and arc methods. As its name implies, resistance weld-
ing is accomplished by the phenomenon of the transfor-
mation of electric energy into heat energy by opposition
to the flow of current. The arc method follows the be-
havior of the electric circuit whenever it is suddenly
opened, namely, the production of a spark which if the
distance between the terminals is maintained will pre-
serve an arc. It was this characteristic of electricity that
produced the first electric-lighting unit, the arc lamp;
and the former, or resistance characteristic, that fur-
nished the incandescent lamp.

Many processes of welding have been devised from
these two primary groups but, as will be seen, the dif-
ferences rest entirely on details generally associated with
special applications. The following diagram gives the
sub-divisions of each group :

Resistance Processes Arc Processes

1. Butt welding 1. Carbon arc welding

2. Spot welding 2. Metallic arc welding

3. Seam welding a. Bare electrode

b. Covered electrode ^

1. Gas flux

2. Liquid flux

Butt Welding.— This process consists of bringing
two pieces of metal into contact end on end and then

ELECTRIC-WELDING SYSTEMS 15

clamping these ends between two jaws of high, conduc-
tive material supplied with high current at low voltage
from the secondary of a transformer. With pressure
appUed forcing the two pieces together and the current
turned on, a localized welding temperature is provided
which, as the operation is visible at all times, may be held
on until proper fusion results. It is usual practice to
maintain the pressure for a short time after the current
is turned off. Due to the end pressure a burr will form
at the juncture of the pieces aiding the observer to make
a satisfactory weld. As the outer surface of the metals
tends to conduct heat away from the point of contact of
the clamping j aws so advantageously, the interior metal
offering the greater resistance will arrive at a Welding
heat before the exterior. This feature protects the proc-
ess from any doubt as to the soundness of the finished
weld. In blacksmith welding the reverse is true and the
finished weld exteriorly may look sound but in reality
cover poor fusion. There is a large field in the industries
for the application of butt welding, especially for the
welding of tool shanks, rods, etc.

Spot Welding. — This process is so called because
the materials are not joined together continuously, but
spaced as in riveting. Plates are lapped and then
brought to a machine and placed as in butt welding be-
tween two high conducting points. Spot welding for
this reason is frequently referred to as " point " welding.
Pressure is brought to bear upon these points, current of
high value is turned on, the materials at the points are
thus ra;ised to a welding temperature, current is then
removed, the pressure released, and the weld completed.
The intensity of the current produces a very rapid rise

16 SPOT AND ARC WELDING

of temperature which with the pressure tend to prevent
the ill effects of entrapping oxygen in the weld. This
characteristic assists in practice to blow out the slag
which may form pn the surfaces of the plates. It is
natural to see an analogy between smith welding and
resistance welding, for just as the smith heats his mate-
rials in a forge and then applies pressure with his ham-
mer, so the spot- welding machine raises the material to
welding heat and then applies pressure. It would be
reasonable to believe that the differences in the applica-
tion of pressure would have a marked effect upon the
results in a relative degree to the time difference in the
raising of the temperature, but this particular interest-
ing side of spot welding has never been fully investigated
or at least reported. Spot welding of light materials,
steel up to y^ inch, has had a remarkable development in
this country for some years. Heavy spot welding, steel
up to 1 inch, was experimented with in the last few years.
In the manufacture of automobile bodies, metal furni-
ture, and bicycle parts, it has established itself firmly.
It is especially fitted for shop work, and for new con-
struction, but has not been developed nor used for repair
work. The decided advantages of resistance welding
over arc welding rest upon the fact that a weld can be
made independently of the operator, that the work can
be done rapidly and in the open, and that it permits of
practical inspection. As will be seen later these points
of advantage give a confidence in results not accorded to
any of the practical processes of arc welding. It is for
this reason that those who are responsible for new con-
struction work are less conservative to the introduction
of spot welding on a large scale.

ELECTRIC-WELDING SYSTEMS 17

Seam Welding. — A minor application of resistance
welding is employed on very thin sheets for making a
continuous seam. Instead of clamping jaws, or points,
the current is led to the work on rollers under pressure.
It is possible that this method may be capable of exten-
sion to heavier materials but it has not been shown by
any large commercial use. The seam welded success-
fully in this way would undoubtedly provide a ready
means for obtaining jfluid-tight work and its point of
economy would then rest upon the speed with which
good welding could be accomplished. When consider-
ing fluid-tight work one point of vital difference must be
borne in mind, namely, the simple retaining of the liquid
in the vessel as against the liquid under pressure.^ It
may be stated as a general precaution that all electric-
welding processes should be carefully investigated be-
fore applying them to work involving danger to human
life. Doubtless future research will permit this work to
be done, but the applications not involving such risks
are plentiful.

Carbon Arc Welding. — In this process an arc is held
between the metals to be joined and a pencil or rod of
carbon. In all arc welding the rod used for maintaining
the arc and manipulated by the operator is called the
electrode. The carbon electrode is connected to the
negative side of a low potential circuit ( 60 to 75 volts )
and the work to the positive side. A very intense heat is
produced by the carbon arc which draws from the elec-
tric supply 300 to 600 amperes. Flanged or butted
edges of thin tank steel (1/16 inch and 3/32 inch thick)
are satisfactorily fused together by the c arbon electrode

*This refers also to gases under pressure.

18 SPOT AND ARC WELDING

without added metal,^ but steel of greater thickness re-
quires a melt rod of proper composition to supply the
filling. The process under these conditions bears a simi-
larity to soldering. Where there is a great quantity of
shop repetition work with thin materials, automatic ma-
chines equipped with carbon electrodes are feasible and
economical. The process is specially valuable in the
heating of large areas and thp filling of large holes. It
is a good tool for heavy repair work. It is also possible
to cut metals with the carbon arc, but few operators are
able to follow a sharp line and thus the imcontrolled arc
leaves a very poor working edge. It is stated that the
carbon arc is economical for the rough cutting of scrap
materials and the demoUtion of steel buildings. Un-
doubtedly there is a future for this process not only in
special lines, but for extensive application to new con-
struction when methods have been devised for the better
control of the arc and ease in manipulation.

Metallic Arc Welding. — By far the greater propor-
tion of electric welding is performed to-day with the
metalhc electrode. The arc is held in the same manner
and by the same means as in carbon welding, but the
connections are usually reversed as the metallic elec-
trode supplies the filler for the weld and a larger per-
centage of the thermal energy is conducted to the metals
being joined. The process is far more comfortable for
the operator as the heat is less intense and more local-
ized. It is a cold process. One hand is free in which to
hold a screen for eye and face protection. A low voltage

is required as in the carbon process but the current is

■.^^— ^^i^»^^»— — — ^—^^— ^— ^^-^ ^^— ^— ^— ^■^— > — ^—

• " Electric Arc Welding in Tank Construction," R. E. Wagner, General
Electric Review, December, 1918.

ELECTRIC-WELDING SYSTEMS 19

much reduced (50 to 175 amperes) . The process is very
simple. It makes a very handy tool for any shop en-
gaged in metal work either for repairs or new construc-
tion. The apparatus for metallic arc welding has its
reason in the field of commercial economics and as an
aid to the operator. An experienced operator should
be able to make as good a weld with an electric circuit
controlled through a water resistance as with the most
expensive apparatus obtainable. This is not intended to
discredit the work done in providing tools for the ad-
vancement of arc welding, but it is stated to protect
those who may interest themselves in the practical appli-
cations of this process from the assertion that the appa-
ratus irrespective of the operator will produce good and
satisfactory work. This process of welding has had ex-
tensive use in the railway repair-shops in this country
for many years, and its expansion in this line is evidence
of both its reliability for serious repairs as well as its
economic value. In like manner it has been used for
repairs to niarine boilers and other applications which
will be considered later. The process has not been ex-
tensively employed on new construction work, but its
adherents have recently given it great impetus in this
direction in connection with the hastening of the ship-
building program during the war. This application re-
quired not only its extension to heavy materials, but
also its investigation by specialists to convince conserva-
tive engineers of the shipbuilding industry.

Bare and Covered Electrodes. — ^Arc welding has its
modifications like any other process. Those who strive
to securely advance a beneficial art are certain to fall
upon some weakness which may be improved or, for spe-

20 SPOT AND ARC WELDING

cial recjuirements, some strong point which may be in-
tensified. Students of the metallic-arc method while
watching the successes and failures under varying con-
ditions hit upon the theory that the bare metal electrode
brought oxygen from the air and carried it into the weld.
Practical tests showed that the bare metal electrode pro-
duced a brittle weld, strong as far as tensile strength
was concerned, but lacking in ductility and resistance to
shock. Invention soon provided a cover for the electrode
and this practice has become general in England. In
this country the practice has been entirely with the bare
electrode ; apparatus has been developed solely on these
hnes and opinion is biased for that reason.

In England two methods are employed, one the use
of a non-conducting fireproof sleeve which leads the
molten metal fronl the end of the electrode and protects
it from contact with the surrounding air. The other
method consists of an asbestos yarn impregnated with
fluxing compound wound upon the metallic electrode.
This sleeve melts with the arc and furnishes a slag which,
while it prevents the access of oxygen to the weld, must
also be brought by the welding operator to the surface
of the weld. If additional layers of metal are necessary
to finish the joint the slag formed on each layer must be
carefully chipped off before depositing the succeeding
one. Many variations are permitted with this system as
combinations may be made with different chemical con-
stituents in the electrode as well as in the flux covering.
The differences of opinion existing as they do between
the two countries (America and England) naturally
creates an interesting discussion among welding engi-
neers. It is not proposed here to enter into this dis-

ELECTRIC-WELDING SYSTEMS 21

puted field. Certain engineers in this country claim to
have produced welds with bare electrodes superior to or
equal in ductility to covered electrode welds. In Eng-
land they are apparently unable to approach with the
bare, metal electrode any of the work done by the cov-
ered electrode process. Attracted by the claims of the
advocates of covered electrodes many American engi-
neers are experimenting wjth coated electrodes, i.e.,
simply immersing the metal electrode in a solution and
permitting it to dry before using. In this regard it is
well known that a solution of ordinary whitewash will
often improve the welding quality of an otherwise poor
welding electrode. , This experimental attitude of the
American engineer at least leads to the belief that the
use of a covering on the metal electrode is of some advan-
tage despite its cost.

Other Processes. — There are two processes of electric
welding which will receive only brief mention because
they have been developed for special purposes and were
not found applicable for the joining of heavy steel parts
such as are usual in steel ship construction. They are
called the electric blow-pipe method and the " water-pail
forge." The latter is not strictly a welding process in
that electricity is used merely to heat the metal which is
afterwards forged in the customary manner. As its
name implies one side of an electric circuit is connected
to a solution held in a wooden pail and the other side of
the circuit is connected to the metal to be heated. In the
former process the electric blow-pipe is essentially a hori-
zontal-flame arc lamp using two carbons mounted like
the letter V. Between these two carbons is placed a
powerful magnet which creates a sufficient magnetic field

22 SPOT AND AEC WELDING

to blow the arc in the direction desired. This method elim-
inates the passage of the electric current through the
work and is said to be successful in the welding of small
pieces of steel ^d brass. The voltex process is a modi-
fication of the blow-pipe method in which the carbon
electrodes are impregnated with metallic oxide which is
vaporized in the flame of the arc'

Besides these processes there are many modifications
of details and apparatus connected with arc welding.
The simplicity of the electric circuit for arc welding has
already been mentioned, and it will be seen later how
apparatus has been devised to aid the operator and de-
crease the cost of operation. Arc welding may be per-
formed with either direct or alternating current. The
advantage advanced for using the latter is the simplicity
and inexpensiveness of the apparatus.

Variations in the chemical composition of the elec-
trode material and variations in current for the welding
of different thicknesses and compositions of steel go to
make up a catalogue of modifications which give great
latitude to the designer of welding apparatus. The use
of arc welding in the shop for repetition work allows
ingenuity of arrangement and connection of apparatus.
For instance, several single arc-welding units may be
connected in series if the work will permit their simul-
taneous use instead of each unit being separately con-
nected as for general field work. These are a very few
of the many possible modifications which the process of
arc welding encourages.

Summary. — From these different methods the spe-
cialist called to advise the government upon the applica-

s u

Electric Welding," Hamilton and Oberg, p. 6, 1918.

ELECTRIC-WELDING SYSTEMS £3

tion of electric welding to ship construction selected spot
welding and metallic arc welding because they were
recognized for the joining of hght materials and logi-
cally could be quickly advanced to the state of joining
heavy steel plates and shapes. They proved that elec-
tric-welding processes, particularly spot and metallic
arc, could be utilized for the joining of steel plates of J^
inch in thickness. Further, that spot welding could be
employed for the joining of greater thicknesses and
also a nimiber of heavy pieces. This cleared the way
for the practical applications of electric welding to
ship construction.

CHAPTER III

Spot Welding

There are four important points to be considered in
either light or heavy spot welding : ( 1 ) The electric cur-
rent requisite for producing the welding temperature;
( 2 ) the time in which this current is utilized for making
the weld; (3) the mechanical pressure necessary for the
electrical contact as well as for squeezing the materials
during the application of heat ; ( 4 ) the condition of both
surfaces of the two pieces of material that are to be
joined. In the spot welding of thin (or light) sheet
steel apparently the wearing away of the electrode points
is not of prime importance, but in heavy spot welding
this feature becomes a serious matter when viewed from
a shop-production standpoint. This point of difference
as well as others will be set forth when the results of the
demonstration of heavy spot welding are discussed. In
general a heavier current must be employed for the weld-
ing of greater thicknesses of steel plates, more time must
be consumed, the pressure must be increased, although
to what degree is questionable, and the contact surfaces,
Le.j, those next to the electrode as well as those imping-
ing upon each other, must receive the attention of the
engineer responsible for the results.

Apparatus for Light Spot Welding. — For this class
of work the design of machine may take various forms
in conformity with the special nature of the product.
For the process alternating electric current is employed
because a high current at a low voltage is needed to

SPOT WELDING 25

produce the welding heat. The usual voltage supplied
to shops in this country for power purposes is 220, and
this requires a transformer usually integral with the
machine. In smaller-sized machines the pressure may
be secured by levers operated by hand or foot ; in larger
machines the practice is to employ either water or air

pressure. In like manner small apparatus may require
no water-cooling arrangement for the copper electrodes,
but in larger machines this is essential. The lighter ma-
chines may also permit of the offsetting of the electrode
for performing in close quarters or accomplishing some
special object, but in the heavier designs the pressure
must come directly in line with the electrodes. As will be
seen in Fig. 1, the secondary of the transformer.

SPOT AND ABC WELDING

Flo. i.—tS-kv. light ^ot'adding oiuchLne.

which in this case is composed of thin copper strips,
conducts the induced current from the transformer to
the electrodes. This machine is used for plate work and

SPOT WELDING 27

is designed with the necessary gap. The electrical con-
nections are arranged so that adjustments for current
may be made for varying conditions of work. Auto-
matic features may be included in the design, so that
repetition work can be done with great rapidity and
with uniform results. Fig. 2 illustrates another type of
spot-welding machine having a capacity of 25 to 30 kw.
and capable of doing fairly heavy spot welding. This
machine has a small gap and could not be used for ex-
tending over wide plates. It is to be noted that the pres-
sure operates in a direct line through the electrodes.

Applications of Light Spot Welding. — As the pres-
sure required for light sheet steel (say 1/16 inch thick)
is approximately very low, about 200 to 300 pounds, the
application of this method to the fabrication of small
articles is very large. As noted, the electrodes may be
placed in various positions and offset in the electrode
holder in any manner required by the special job. Thus
kitchen utensils, like coffee pots, saucepans, etc., are
made by spot welding the spout on to the body, thus
facilitating the finishing operation by leaving a smooth
surface. House fittings, such as doorknobs, sash pul-
leys, etc., are made by the thousand in a very short time.
Small chains for various purposes are made on special
and very interesting machines, and the use of this process
in the bicycle and automobile industries has been respon-
sible for decreased cost. Many special applications 'are
of interest, such as the welding of the two magnet bars
in a telephone receiver. The difficulty of doing this by
other methods of welding is that the temper of the mag-
net steel would be drawn and so destroy the purpose for
which it was intended.^

* " Electric Welding," Hamilton and Oberg, p. 121.

28 SPOT AND ARC WELDING

Possibilities of Light Spot Welding. — Many opera-
tions not strictly spot welding may be performed on a
light spot-welding machine {see Fig. 3). By preparing
the materials in a special way, by the use of a button
placed on the materials, by means of special electrode,
many small articles can be easily and quickly welded.
Sheets may be welded to studs, bolt heads to body in a
spot-welding lyiachine, although the operation bears a
strong likeness to butt welding.' In the same manner

Fu. S.— Sp«cu] JDbi dc

screws may be welded to sheet tubing. The spot- welding
machine may be utilized to heat rivets in place and
squeeze them " home," thus performing as an automatic
riveting machine. A wrongly punched hole may be cor-
rected by introducing a proper-gized stud and then spot
welding it in place. This also may be done with heavy
spot-welding machines.^ By arranging the edge of thin
sheets with projections which act to localize the heat, and
with special electrodes or multiple electrodes, a number

SPOT WELDING 29

of spot welds may be made in one operation. It is stated
that there are machines made with a solid electrode of
copper against which a single electrode is made to move
at designed intervals. This apparatus is capable of mak-
ing thousands of spots a minute and connecting thin
sheets. It only requires a fraction of time to make a spot
in such materials.^ It is not believed that all the possi-
bilities for light spot welding are by any means exhausted
and undoubtedly the introduction of heavy spot welding
will result in the further extension of spot welding
in general.

Possibilities of Spot Welding in Ship Construction.
— Although spot welding was applied in the industries
only for steel plates not exceeding ^ inch in thickness,
the results had led many engineers to a belief in its ex-
tension to greater thicknesses. So in 1911 the American
Car & Foundry Company built a portable spot-welding
machine with a gap of 66 inches, so as to extend across
wide plates and with which they constructed a gondola
freight car. The welding machine was equipped with
a transformer of 85-kw. capacity " having a primary
voltage of 400 and a secondary open-circuit voltage of
25. . . . Pressure was apphed by means of a hand
wheel, but subsequent machines were equipped with air
cylinders for applying the pressure. Copper electrodes
3 inches in diameter were used ^ jaw, the welding points
having a diameter of % of an inch. It was found that
perfect welds could be made with this machine through
3 sheets each % inch thick, or 2^4 inches total thick-
ness." * The results of the completed car were satisfac-

3 " Electric Welding," Hamilton and Oberg, p. 108. .

* " An Electrically-Welded Freight Car," J. A. Osborne, General Electric

Review, December, 1918.

30 SPOT AND ARC WELDING

tory in every way and reports while in service show that
no abuse has occasioned serious embarrassment to the
spot-welding process. The engineer of this work states
his opinion : " I believe that this severe test of the process
has demonstrated the fact that spot welding for heavy
structures is absolutely practical and reliable. It has
demonstrated that riveting can be replaced in almost all
instances. The ease of manipulation, as well as the
gi'eat saving, will no doubt cause this process to be
universally adopted." ^

For this particular application a portable welding
machine of large dimensions was necessary, and the
makers of spot-welding apparatus were evidently not
inchned to develop such. They had built stationary
machines of 100-kw. capacity and pressure of 50 tons
on the electrodes, but had not considered designs of this
size for portable use. The upper electrode of the 100-kw.
machine was 2i/^ inches in diameter and the lower 3
inches in diameter, the pressure coming directly on them.
This apparatus was capable of welding " two strips 1^4
inches thick." ®

A long series of tests were undertaken in 1917 at the
plant of the New York Shipbuilding Corporation at
Camden, N. J. These tests were never published in
detail, although an account of them is available. For
certain small operations a 25-kw. spot-welding machine
was purchased in 1916. This apparatus was the only
means for carrying on the investigations, and therefore
the tests could not be extended beyond the welding of
two thicknesses of J/^ -inch steel plate. The results all

' Idem.

« " Electtac Welding," Hamilton and Oberg, p. 118.

SPOT WELDING 31

assured the use of the process in the fabricating shops
with machines of proper size and design. There are
many pieces of the ship's structure, such as deck houses,
doors, skylights, hatch frames, masts, stacks, etc., which
are put together and form quite a separate part from the
hull of the ship. These various sub-structures go to
make up a long list of what is generally called ship's
fittings. Such fittings rarely demand thick sections, as
they are not required to resist heavy strains nor undergo
excessive stresses. Even in the heavier class of vessels
such fittings will be made of 34'iiich steel, in some
cases less.

In view of this wide variety of work that could be
done by spot welding and the large saving to be secured
by its use led those who were making this investigation
to extend their trials to the building of a water-tight
bulkhead door. Fig. 4 shows the spot welding of the
3-in. X 3-in. X 5/16-in. angle frame to the door plate.
In the endeavor to secure water-tightness with a great
nimiber of spots and the use of flux about 141 spots were
made. It was noted that a smaller nimiber of spots and
the edges of the angle arc welded would give more satis-
factory results. After the door was completed it was
placed with the edges of the angle upon a steel plate,
securely clamped and made water-tight by the customary
methods used on shipboard. A water connection was
then made and pressure gradually applied. A slight de-
flection in the plate at the top and middle of the door was
noticeable when the pressure reached ten pounds per
square inch, but there were no leaks. The pressure was.
increased up to a maximimi of 22 J^ pounds per square
inch without leakage, but upon examination showed that:

32 SPOT AND ARC WELDING

some of the spot welds on the stiffeners had given away.
The structure had withstood pressures far beyond those
required for this service and tiie tests were conclusive as
far as the adoption of the process was concerned.

As has been intimated there are two distinct condi-

Fra. 4.— Spot welding,

tions under which riveting is done in shipbuilding: that
which secures the principal members of the hull struc-
ture called " field riveting," and accomplished in this
country by a group of men in which one man uses an
air-driven rivet hammer, and that which is done in the
shops by semi-portable machines. These latter machines

SPOT WELDING 33

are of large size, usually suspended from jib-cranes, and
are portable in the sense that they may be brought to the
work and moved from rivet to rivet by the operator.
These portable riveting machines are equipped with air
cylinders and connected by flexible hose to a common
source of supply. They are as in field riveting depen-
dent upon the forge fire for the heating of the rivet, and
on the rivet boy for supplying the rivets at a correct tem-
perature and at the proper time. This practice in ship-
building shops differs from that in bridge shops where
the machines are stationary, usually mounted with their
jaws in a vertical position and to which the overhead
crane brings the material and moves it for the continu-
ous performance of riveting. The rapidity with which
such work is accomplished in this latter case is astonish-
ing. A close study of this problem is like the run of
spot-welded work, merely requiring as much time and
attention devoted by those who understand the process
as has been given by those who understand riveting. In
short, it is a question of shop production which in the
case of riveting has been developed to a high state
of efficiency.

All these points and many more suggested that spotr
welding machines similar to a pneumatic riveting ma-
chine could be employed with advantage in shipbuilding
and would hasten the construction of steel ships. The
idea was proposed to convert the pneumatic riveting ma-
chine as now used into spot-welding machines by adding
a transformer and the proper flexible connections. This
was the inception of the design of two machines for spot
welding which were built for the Emergency Fleet Cor-
poration of the U. S. Shipping Board and with which a

34 SPOT AND ARC WELDING

practical demonstration was made the early part of last
year. The builders of this apparatus had previously ex-
perimented with a stationary apparatus of large size to
determine if there were any obstructions to the welding of
thick steel plates, and as their researches showed no hin-
drance to the process up to the spot welding of three
thicknesses of 1-inch boiler plate, they accepted an order
for two portable spot welders and one stationary spot
welder for the purpose of fabricating the steel parts for
the hull structure of ships/ This apparatus will be fully
described in connection with the discussion in the
next chapter.

There were men who desired to put the welding
methods to test in shipbuilding, for the emergency was
pressing, and yet they did not follow the English ex-
ample in going so far as the building of small water
craft. The English Admiralty built in the early part of
this year (1918) a cross-channel barge entirely arc
welded, and the results of this craft were and are well
known in this country. It was decided to make a dem-
onstration of a large portion of the middle body of one
of the standard steel ships building for the Emergency
Fleet Corporation. The scaffolding for this work was
erected, the raw material delivered, and a five-foot-gap
portable spot welder was ordered and delivered/ The
demonstration as planned was abandoned after the sign-
ing of the Armistice. The portable spot welder was sent
to the same shop as the other spot-welding machines
and was tested as a part of the demonstration of
spot welding.

^ " Research in Spot Welding of Heavy Plates," W. L. Merrill, General
Electric Beview, December, 1918.

SPOT WELDING 35

Summary. — The spot welding of light materials has
been practiced in this country for the last fifteen years
and its performance not only has guaranteed a success-
ful product, but also has increased production and

Fb). <.— Spot-nldHl W. T. door under lot «D lb>. per iqiun iiidi.

profits. It is safe to state that its extended service in
those industries that find a use for it assure its future.
For the joining of light sheet steel it has no competitor,
but for the softer metals the process has not been suc-
cessfully developed. There are certain alloys that may

86 SPOT AND ARC WELDING

be welded by the preparation of the materials or by
changes in the apparatus. The whole subject is one re-
quiring investigation, and where the results may be seri-
ous attempts should not be made without thorough tests.^
As stated, copper has been the only available metal found
for electrodes and any metals which approximate the
current carrying capacity of copper would offer no re-
sistance to current flow, and upon this principle depends
the welding heat.

With the satisfactory results of the spot welding of
thin steel sheets placed before him the engineer surmised
that the process was capable of extension to heavier steel
structures. The industry that it was most important to
aid happened to be that of shipbuilding, but the prin-
ciples of the application are suitable to all manufactur-
ing concerns employing heavy steel plates and shapes
which are now joined by rivets.

In shipbuilding the shop rivets are used to connect
and assemble small parts of the structure and for the
manufacturing of sub-structures. The field riveting is
employed to put the assembled materials together and
form the completed hull of the ship. The shop rivets
are driven by machines brought to the work and between
the jaws of which the rivet is squeezed. The field rivet
is put in place through holes punched in the assembled
parts of the ship structure and is driven by a hammer on
one side of the work, actuating against a pressure ap-
plied to the other side of the work. The main question
of applying spot welding to the entire process of ship-
building resolves itself into two possibilities. There can
be no doubt as to the use of spot welding in shop work,

*" Electric Welding," Hamilton and Oberg, p. 95.

SPOT WELDING 37

but the question rises whether it can be applied in the
field. Either great developments must be made in the
spot-welding tool whereby it may act on both sides of
the structure without interruption to other work, or there
must be prepared special designs of ship and shipyard
whereby the spot- welding tools may be used. That is to
say, that the welding machines would not be required to
assume shapes and sizes that militate against the con-
venient working of the tool. As in other matters, this
last possibility is one for cooperation between three im-
portant factors in shipbuilding: the owner, the naval
architect, and the shipbuilder. There is always an ex-
periment going on in any shipyard, as the experienced
shipbuilder recognizes that every ship he is building is an
experiment. Perhaps it is this every-day affair that
makes the shipbuilder hesitate to accept new methods for
connecting his steel ship. The shipbuilder is accustomed
to experimenting and when he discovers the benefits to
be derived by the adoption of spot welding he will be its
firmest advocate.

CHAPTER IV

Demonstration of Heavy Spot Welding
(by the emergency fleet corporation)

The first principle laid down for this investigation
was that it should be practical so that the limitations of
practice would be shown in high relief. That these limi-
tations were a hindrance to a full report of the ability of
the apparatus to perform in a certain way must not be
interpreted as a detriment to the process nor a lack of
skill upon the part of the designer. On the contrary,
such practical handicaps as were encountered only go to
prove the field of usefulness and the profits yet to be
made by those who desire to invest their capital in a
process as certain -as the results of this demonstration
showed. No theories were acted upon in the conduct-
ance of the test and only minor modifications were made
in the nature of developments after the initial trial of
the first machine. The results of the tests speak well
enough to assure the most skeptical of the safe joining
of heavy steel. members by the spot- welding process.

Description of Apparatus. — Considering the adapt-
ability of the machines for shipbuilding service it was
decided to build three machines, two semi-portable of
moderate size and one stationary machine of large size.
The semi-portable welders took an external form similar
to the pneumatic riveter and were provided with bails for
attachment to cranes for facility of handling. After a
careful survey of the usual rim of materials in the plate

DEMONSTRATION OF HEAVY SPOT WELDING 39

and angle shops of shipyards, it was agreed to build two
sizes of portable machines to cover all conditions now
met with in practice. The smallest machine has a gap,
or throat, capacity for reaching over a width of 12 inches
and the other a gap of 27 inches. This difference in
throat not only increases the weight due to the additional
frame size, but also adds bulk by reason of the trans-
former capacity which must be greater in order to over-
come the reactance caused by the enclosing of a large
body of magnetic material in the electric circuit. Al-
though much care was given to the lightening of the
frame and, as will be seen, the transformers were
made remarkably small for their rated capacity, still
these machines are much heavier than the ordinary
pneumatic riveter.

The frames of these portable machines were cast out
of gun metal to provide against any chance of a react-
ance suflScient to counteract their welding qualities. In
what may be termed the body of the machine a recess was
left for the transformer. On the top arm was located
the air cylinder for providing the necessary pressure on
the electrodes. The under body portion of the frame was
arranged with ample projections on both sides so that
the machine could be bolted down in place and used as a
stationary tool. As will be seen from Figs. 6 and 7, the
copper electrodes are mounted in a holder insulated from
the frame, but in direct line with the pressure. The lower
electrode holder is bolted directly to one arm of the sec-
ondary of the transformer, the upper electrode holder is
connected by flexible leads of laminated copper in order
to permit the necessary movement for squeezing the two
pieces of work together.

40 SPOT AND ARC WELDING

The maximum air pressure provided was the same
for both machines. The air cylinder was 8 inches in

Flo. a. — Electric wdda — li" taut — B0-;i«S-4M-6.iE. Cspible ri wtMiglngtlhatfottttl plain

diameter and attached to it was a lever arm with a ratio
of 5 to 1, so that with a gauge pressure of 100 pounds
per square inch, 25,000 pounds per square inch could be
exerted on the work. During all the tests this was never

DEMONSTRATION OF HEAVY SPOT WELDING 41

7" mch— eO-MD-UU = UM. Capable ot wcldiiig togetha ti
pUl« M" Ibick in apots 1" in diuneler.

obtained, but good welding was accomplished at 70- and
75-pound gauge, representing 17,500 and 18,750 pounds
per square inch on the work. A gauge was installed on

42 SPOT AND ARC WELDING

the air line for the purpose of cheeking the pressures at
all times. A reducing valve was also provided, so that
the test conditions could be changed and variations
allowed for thinner material.

There are many interesting points of design in con-
nection with these two machines which can receive only
brief mention. One of them is the transformer. Per-
haps no transformers were ever designed or built within
so small a space and with so great a capacity. The
designer for this alone, to say nothing of the successful
operation of the machine, should feel proud. The capac-
ity of the transformer for the 12-inch machine at 440
volts 60 cycles is 265 kv-a, and for the 27-inch machine
at the same voltage and cycles 350 kv-a. In the large
stationary machine there are two transformers of 450
kv-a each at 500 volts 60 cycles, and the over-all dimen-
sions are 11 inches by 16 inches by 18 inches. The nature
of the work done as well as a resort to water-cooling en- ,
ables this reduction in transformer size. The making of a
spot weld takes a few seconds. The current is used almost
instantaneously. The primary windings are made of cop-
per tubing specially prepared for this purpose, and the
single-turn secondary was built of copper plates bent to
shape and fitted over each other in such a way that a
passage was left for the circulation of cooling water.

Another interesting point of design is that connected
with the electrodes. As stated before, copper seems to
be the best available material for this purpose, and its
use has astonished those acquainted with metals, for it is
so much softer than the steel which it welds. " The
severity of the conditions to which the tips of the elec-
trodes are subjected will be understood when it is con-

DEMONSTRATION OP HEAVY SPOT WELDING 43

sidered that the current density in the electrode material
at this point is approximately 60,000 amperes per square
inch, and that this material is in contact with the steel
plates which are brought to the welding temperature,
under pressures of 15,000 to 20,000 pounds per square
inch. It must be remembered, also, that copper, which
is the best material for this purpose, softens at a tem-
perature considerably lower than the welding tempera-
ture of steel. The difficulty of making the electrode tips
stand up under the conditions to which they are sub-
jected has, in fact, constituted the most serious problem
which has been met in the development of these ma-
chines." ^ This question has been practically dealt with
by providing caps and separable tips for the electrodes
as well as by maintaining a free circulation of water
through the electrode. This matter will be further dis-
cussed when considering the results of tests.

Although these rnachines could be properly operated
directly from a 440-volt 60-cycle alternating-current
source, they are provided with auxiliary transformers
and panels for regulating the voltage and current. This
allows for a wide range of work and permits great free-
dom for experimentation. The regulating transformer
panel also contains a contactor for the ease of operation
at the spot welder where it is only necessary to work a
hand lever for the mechanical pressure and a tripping
switch to actuate the contactor on the transformer panel.
The contactor functioning as a throwing-in switch for
completing the electrical circuit whereby the high cur-
rent passes through the electrodes producing the local-

1 n

Recent Developments in Machines for Electric Spot Welding," J. M.
Weed, General Electric Review, December, 1918.

44 SPOT AND ARC WELDING

ized welding temperature. This arrangement of
connections permits the selector panel to be placed re-
motely from the spot-welding machine, thus allowing
freedom for movement from spot to spot.

The large stationary machine was built with a 6- foot
throat, so that it could reach the width of the usual
run of plates used in shipbuilding, and as was thought
at the time of designing, would be able to fabricate com-
plete deck houses as well as join two steel plates of ^-
inch thickness. In order to reduce the great capacity
that would be required to overcome the effects of react-
ance in this case, the designer did two skilful things. He
provided two transformers, two pairs of electrodes, thus
doubling the number of spots per operation, and then
disposed the transformers one on each side of the work,
thus reducing the reactance to a minimum. Incidentally
this allowed the use of steel for the frame which was
constructed of two steel plates each two inches
thick. Gun metal was used for the heads carrying the
copper electrodes.

Fig. 8 illustrates this machine, showing that the same
arrangement of connecting the upper and lower elec-
trode holder to the secondaries of the transformers was
employed. The same general features of design are
carried out on a larger scale. The air pressure is in-
creased to a maximum of 30,000 pounds per square inch
on each cylinder. There are two cylinders provided, so
that each pair of electrodes may be separately operated
and also to obtain successful spots when making two
welds simultaneously. The air cylinders are located in
the body of the machine and operate through 7-foot
levers to the electrodes.

DEMONSTRATION OF HEAVY SPOT WELDING 45

The electrodes are arranged to be easily removed
and may be shifted on their bases to positions of 90
degrees, so that the spots may be made in line with the
axis of the machine or transversely. .The electrodes are

Fm. B.— Duplei electric welder—

spaced 8 inches centre to centre, but may be disposed
from 10 to 6 inches centre to centre.

The cooling of the electrodes as well as the cooling
of the transformer winding is similar to that of the
smaller machines. Hydrant water has been found prac-
tical for this purpose. The water traverses the appa-
ratus in two parallel paths, " one being through the
primary winding and the other through the secondary

46 SPOT AND ARC WELDING

and the electrodes in series." ^ Separate valves are pro-
vided for independent control of flow in the two paths.

This duplex spot welder, as it is called, is capable of
producing 50,000 amperes with 500 volts 60 cycles.
With this much current in the secondaries the current
in the primaries is 1800 amperes. The kv-a under these
conditions is 450 for each transformer, and at 440 volts
and the same cycles kv-a will be approximately 350.

As with the smaller machines the 6-foot duplex
machine may be operated on a 440-volt 60-cycle alter-
nating-current source of supply, but a regulating trans-
former was provided tar the piu*pose of connecting these
machines to a higher voltage supply as well as to regu-
late the voltage and current for different thicknesses of
material for experimental purposes. The capacity of
the transformer supplied was 350 kv-a, and as was just
stated, this machine required at least 350 kv-a at 440
volts on each welding transformer — a total of 700 kv-a.
The apparent discrepancy is removed when it is remem-
bered that the operation is in seconds, a period of over-
load too short to injure the transformer. More interest-
ing is the physical difference in size between this 350
kv-a regulating transformer and one of the transformers
of the duplex spot welder.

On the regulating panel is mounted the contactor
for connecting the electrical supply to the two trans-
formers and, as will be seen in Fig. 8, the remote control
switch for actuating the contactor is mounted in a group
with the air levers and air gauges. The entire mechan-
ism is controlled from this point and with a nicety that

^ " Recent Developmenta in Machines for Electric Spot Welding," J. M.
Weed, General Electnc Review, December, 1918.

DEMONSTRATION OF HEAVY SPOT WELDING 47

reflects much credit to its builders. From this position
both electrodes may be raised or lowered independently
or together, and by placing a spacing block of copper
between one of the pairs of electrodes the other electrode
may be used to make a single spot. This is often of
advantage in working along a seam, as the multiple of
two may not meet the requirements of the adopted spac-
ing, so that it may be necessary to make a single spot to
complete the job.

These machines are equipped with all the best and
latest features of good design. The workmanship is
excellent and the materials well calculated to endure the
heavy duty that should come to this machine in regular
production work. Intrinsically the machine is of high
value and is self-contained; extrinsically this machine
for commercial production will probably require an
auxiliary that will change the frequency of the electric
supply, distribute the power as now supplied over the
three phases, and reduce the sudden rushes of current
required by the process. As has been described of this
apparatus, it all operates on a single-phase current and
at 60 cycles. Under these conditions there is a large
unbalance to be expected in the supply as well as in the
low-power factor. By means of proper auxiliary
apparatus the power consumed as well as the other
desirable features mentioned may be obtained, and this
apparatus will then function well within the bounds of
commercial economics.

Specified Requirements. — ^When the orders were
issued for this apparatus this previous consideration was
probably never entertained because it was not estimated
that these three machines would be in simultaneous use.

48 SPOT AND ARC WELDING

If such had been the ease some form of apparatus to
correct the electrical-supply conditions would have been
required. It was thought that the work of these ma-
chines was so rapid that even our largest shipyard, Hog
Island, for which these machines were ordered, would
fail to supply them with as much work as they could per-
form. And this would have been true unless great
changes in the methods of building ships had come to
pass. These machines were never completely delivered
or operated at Hog Island, but were diverted for dem-
onstration purposes to one of the many bridge shops
which were fabricating material for the shipyard. This
work being in line with strict production output these
machines were more suitable, and their introduction was
interrupted solely because of the Armistice.

Some of the detail requirements of this apparatus
had to conform to the electrical-supply conditions of Hog
Island, so that the regulating transformers were all fur-
nished for a primary voltage of 2200, single-phase, 60-
cycle alternating current. As the electrical current at
the Island was also distributed on a low- voltage (440)
system at 60 cycles the transformers integral with the
welding machines were so specified. These points were
common to all the machines.

The 6-f oot-throat, duplex, spot- welding machine was
required as a maximum of capacity to weld two spots at
once through two thicknesses of 94"i'^ch steel plate, i.e.^,
it was to be capable of welding continuously two thick-
nesses of ^-inch steel and two welds at each operation.
No particular ship work was specified nor any require-
ments conditioned upon the efficiency of the spot- welded
joint as compared to the customary methods of riveting.

DEMONSTRATION OF HEAVY SPOT WELDING 49

In the same way the 12- and 27-inch welding machines
were required as the maximum of work to weld two
thicknesses of ^^-inch steel plate without further stipu-
lation as to performance. Other requirements referred
to the special features described above and appurte-
nances, such as flexible leads for the portable machines,
transformers, panels, remote-control switches, water-
cooling arrangements, etc.

The point to be noted is that in all fairness to the
builders this specification was their responsible guide
and legal guarantee. No more' could be sought in an
acceptance test, which was the real purpose of this dem-
onstration, than that these machines would securely weld
two thicknesses of ^-inch steel plate in the case of the
two semi-portable and two thicknesses of the %-inch
steel plate in the case of the duplex welder. It is esti-
mated roughly that 80 to 90 per cent, of riveted ship
joints are three thicknesses, and approximately 15 to 20
per cent, four thicknesses of steel plate. The question
then arises. Are these machines suitable for ship construc-
tion? It is not altogether possible to carry the demon-
stration to an unequivocal answer in the affirmative, but
it is believed that the many tests made, remembering the
limitations imder which they were conducted, show that
these machines would give a satisfactory performance
for ship fabrication under favorable shop-production
management. It should be borne in mind in looking
critically at the results that the responsibility of those
interested in this demonstration ceased when the require-
ments of the specifications were fully met.

Arrangements for Tests. — The bridge-construction
company, which was one fabricating ship material for

50 SPOT AND AKC WELDING

Fio. 9.— <KI iwitdxa tad high polmliil ui-comiiig-luie paad lot Bpot-wdding danoutmliDa.

DEMONSTRATION OF HEAVY SPOT WELDING 51

the Hog Island shipyard, selected for the demonstra-
tion was the MeClintie-Marshall Construction Co., lo-
cated at Pottstown, Pa. At the time of preparation
for the tests the new Liberty Shop, devoted entirely to
ship fabrication, was just nearing completion. It was
decided to erect the apparatus in this shop in one of the
bays designed for a group of riveting machines with the
intention of retaining them in this production position if
all went well. This location gave easy access to air and
water supply and avoided interference with other sec-
tions of work in the shop.

At a high point near the roof a temporary platform
was built with head-room for the incoming-line panels,
oil switches, and the regulating transformers for the 12-
and 27-inch semi-portable machines. Fig. 9 shows the
arrangement of this apparatus which secured safety by
keeping the high potential away from the working
spaces. Fig. 10 illustrates the mounting of the regulat-
ing transformers for the semi-portable machines. The
electric conductors were carried down in a vertical line
from these transformers to a second, or lower, temporary
platform upon which were placed the selector panels.
Upon these panels were placed the remotely controlled
contactors. This is clearly seen in Fig. 11.

This same illustration shows the electrical connec-
tions from the selector panels to the two machines. On
the left-hand side is the 27-inch welder and on the right
is the 12-inch. The leads which look like hose in the
photograph are electric wires which close the contactors
which in turn energize the machine transformers from
which the secondary or induced current is obtained for
producing the welding current. The large dark leads

5« SPOT AND ARC WELDING

are the wires which carry the main primary cmrent from
the taps on the selector panels. It will he seen from the
photograph that the taps on the panels are numbered
starting on the left-hand upper row and by means of

¥n. 10.— BcgnlAlIng tmufonocn tor \t"x W •end-portable spat-vddiag madiiiiei.

two removable sliding contacts a varied combination of
connections may be made.

In this same Fig. 11 on the right-hand side will be
seen the water supply and exhaust for the coohng sys-
tem for the 12-inch machine. These connections are
made at the lower back end of the frame just inside of
which is the transformer. Almost in a direct vertical
line wiU be seen the air connections, reducing valve, and.

DEMONSTRATION OP HEAVY SPOT WELDING 63

air gauge. The photograph incidentally shows the 12-
inch machine spot welding a part of one of the samples
used in the demonstration and to be described later.

This particxilar photograph was taken very shortly
after the apparatus was installed, and at that time it was

Fkj. ll.^-5«[ecfor pbdfIi for 12"i W lemi-partAblp spat-wcCdidfl nuchihei.

anticipated that the machine would be operated as port-
able tools. A sufficient attempt was made to do this,
but it was quickly discovered that the shop did not
have the requisite crane facilities for permitting this
arrangement as a regular method. These machines were
then transferred to the other side of the steel columns
and made stationary on steel horses. This gave perma-

64 SPOT AND AKC WELDING

nency to the machines and served the purpose of the
tests, although it required very awkward and detrimental
handling of large bulky pieces to the machines. This
latter arrangement of the small machines can in part be
seen in Fig. 12 to the left of the large 6-faot duplex
spot welder.

This same illustration indicates the size of the 6-foot

duplex spot welder, and shows distinctly the electrodes
with cooling-water connections and the arrangement for
replacing the electrode tips. As the regulating trans-
former (Fig. 13) for this machine was too great in
height and weight for the temporary platform already
supporting the two regulating transformers for the small
machines, it was placed as shown on the main shop floor

DEMONSTRATION OP HEAVY SPOT WELDING 55

just back of the spot welder. To one side of it was
located the selector panel (Fig. 14) containing also the

FlQ. 19. — Regnlatiiig tnufonner for ni'^oot FlQ. 14. — Sdedor puel for sii-loot ipot-

q»t'Weldia£ nucfauje, itj p[BC« for testing, wdding macbin« in placf for Uiting.

main switch contactor. This arrangement, although
necessitating a continuation of the high-potential leads

66 SPOT AND ARC WELDING

to the floor of the shop, permitted short connecting leads
between the transformer, panel and machine. The high-
potential leads were carefully protected. The whole lay-
out was of a temporary nature, due to delay in the de-
livery of this machine and the fact that the time allotted
for this demonstration was drawing to a close.

In the same space, but a little farther down the shop,

was placed the 5-foot portable spot welder (Fig. 15)
which was intended for use in building the demonstra-
tion section of a middle body portion of a standard ship.
Through the large gap may be seen the 27-inch semi-
portable spot welder, and to the extreme right the rear
or control end of the 6-foot duplex spot welder. This
photograph gives the appearance of great bulk and
weight to the machine, but the castings were lightened in
great measure and with a sufliciently powerfiil crane

DEMONSTRATION OF, HEAVY SPOT WELDING 57

and proper rigging this 5-foot spot welder could prob-
ably be handled. The head, as will be seen, was fitted
with a pair of short-circuiting electrodes and a pair for
making the weld. These copper electrodes were about
three inches in diameter and the intention was to use a
copper button on each side of the joint to be welded.
There were no mechanical arrangements for holding
these buttons in place in case the machine had to be tipped
to make the weld, but presumably it was intended that
the buttons would be held in place by the operator or one
of his assistants. Back of the hinge can be seen the air
cylinder and piston which provided the mechanical pres-
sure on the electrodes. Just above this cylinder and en-
closed in the upper arm was a small electric motor for
close adjustment of the jaws. The welding transformers
were enclosed in the casing just back of the head near
which on the opposite side from that shown was placed
the operating switch. This control switch combined the
movements of the jaws and the operation of the main-
line switch through the remote contactor which as in the
other designs was placed on a panel at some distance
from the machine. This spot welder was not provided
with means for changing the supply voltage, but oper-
ated on a 220-volt 60-cycle single-phase alternating cur-
rent. This supply voltage was taken from a suitable
tap from one of the regulating transformers of the
other machines.

The high-potential line was brought from the main
power-house of the plant which in turn was served by the
local central station. Every precaution was taken as re-
gards the eflFects that might be occasioned on these lines.
The central-station company was fully informed as to

58 SPOT AND ARC WELDING

the experiments to be made and the amount of energy
that would be suddenly demanded. Additional central-
station transformers were at hand in case they should be
needed. No difficulties as to power supply were encoim-
tered during the trials of the small machines, and the
voltage drop at this period of the demonstration was not
excessive. It was not necessary to operate both these
machines together, and so this was never done. When
the 6-foot duplex machine was first tried the voltage drop
was very great and additional copper conductors were
installed with available material. This reduced the volt-
age drop an appreciable amoimt, but not to a point which
would give the maximum capacity of this machine. The
time for correcting this limitation would greatly exceed
that permitted for these tests and it was considered in-
expedient to expend further time and money in view of
the fact that this machine had shown itself fully capable
of meeting the specification requirements, and undoubt-
edly with a full voltage supply would have greatly ex-
ceeded them. The results in the opinion of those most
concerned testified to the reasonableness of the
action taken.

Methods and Procedure of Tests. — In view of the
broad interest taken by Lloyd's Register of Shipping in
the application of electric welding to steel ship construc-
tion and the recent investigations of this classification
society into the processes of arc welding which were con-
ducted in England, it was deemed appropriate to parallel
some of the smaller practical tests for the sake of com-
parison. One of these tests was designed to compare the
bearing value of a short attachment lug riveted as against
arc welded. This same design was used but the attach-

DEMONSTRATION OF HEAVY SPOT WELDING 59

ment lug was spot welded with the same number of spots
as rivets. The test pieces consisted of ^-inch flat plate
6 inches wide by 18 inches long upon which was riveted
an angle lug 2^ by 2 ^/^ by % inches. The lug was
secured to the plate by four ^-inch rivets spaced 2J/^
inches centre to centre. The attachment lug was 12
inches in length. In the spot-welded test piece four
spots, located as nearly as possible to the same spacing,
secured the same-sized angle lug. This sample repre-
sents a very frequent job in the fitting-up of the hull of
a ship and the practical value of the comparison can-
not be questioned.

*The second test piece was intended in the Lloyd's
investigations to compare " the relative value of welding
and caulking under tension." ^ In the case of the spot-
welding comparison this was at first not considered,
although a sample of spot- and arc- welded joint was
later made with results that were easily foretold. This
test piece took the shape of a cross and simulated the
boundary of a water-tight compartment. It consisted
of a 20-pound flat plate 24 inches long and about 15y^
inches wide. At the centre of this plate and perpen-
dicular to it were, attached by 3 ^/^ X 3 3^ X %-inch
angles, two flat plates 24 inches long and 7^ inches
wide. The ^-inch angles were first secured to the two
small pieces of 20-pound plate by ten %-inch rivets,
and then the two angles joined by the same number and
size of rivets to the large flat plate. This required the
joining of three thicknesses of material amounting to a
total thickness of two inches. The spot-welded test

'"Lloyd's Experiments on Electrically-Welded Joints," H. J. Cox,
General Electric Review, December, 1918.

60 SPOT AND ARC WELDING

pieces were made up in the same mamier and secm^ed the
members with ten evenly spaced spot welds. As much
difficulty was encoimtered not only with modifications
of the then-available spot- welding machine, but also with
the testing machine, and, as the %-inch angles of this
piece were in excess of practice for such a connection,
the spot-welded test pieces were afterwards changed to
3^ X 3^ X J/^-inch angles. During the trials certain
other test pieces were prepared and tested in order to
determine the proper proportion of current, time, .etc.>
needed for spot welding three thicknesses of ^-inch steel
plate. These test pieces consisted of two strips of size
convenient for the testing machine placed end to end and
the joints covered by a small strip top and bottom. This
allowed two spots through three thicknesses of 3^ -inch
steel, a total thickness of 1 ^ inches.

This practice of spot welding three thicknesses was
of value in the next imdertaking which was the spot weld-
ing of a ship's floor. Fig. 16 shows the floor in process
of spot welding. This job is in line with regular produc-
tion work. In order to make comparison, material was
prepared for three sets of floors. One of these was to be
riveted in the usual manner, the next was to be arranged
with a few holes punched for assembling and then to be
spot welded, and the third was simply the materials cut
ready to be spot welded. In the latter case the materials
were assembled by arc-weld tacking. When these floors
were ready for test only one spot-welding machine was
available, the small 12-inch machine. An accident had
happened to the 27-inch welder and the 6-foot duplex
machine was not ready for shipment. In addition, modi-
fications to the electrodes of this machine were necessary

DEMONSTRATION OF HEAVY SPOT WELDING 61

both in order to weld through three and four thicknesses
of 20-pound plating, and also in order to manipulate the
apparatus so as to locate the spots in the proper position
with respect to the heel of the hounding angle. These
floors were spot welded with only two points in mind.

Fu. IS.— Spot-welded ihip Bov.

both of which were clearly demonstrated : (1) That spot
welding along a certain eircumscribed seam would
neither distort nor elongate the material, i.e., cause a
creeping efi'ect; (2) that the edge of the angle resting on
the flat plate would not be scored nor distorted to the
detriment of the mechanical caulking, i.e., that a caulk-
ing edge would be preserved. That the spot welding of
these floors also showed other features that were fair

62 SPOT AND ARC WELDING

must be considered without the intent or purpose of the
test ; that the floors did not meet with a full success when
subjected to later tests was to be expected.

Historically considered the demonstration had now
reached a point where more practical data was desired.
The 27-inch semi-portable machine had been repaired
and was now set up with modifications principally in the
protection of the electrodes. This machine as mounted
and modified was found more convenient for the next
series of tests, although the 12-inch machine with similar
modifications would have produced like results. The
test pieces now prepared were varying thicknesses of
steel plate from ^ inch to ^ inch, i.e.^ ^, ^, ^, ^, and
% inch. The pieces were ten feet long and of a width
suitable for easy handling in the tensile-testing machine.
These strips of plating were lapped and tacked together
by means of the electric arc. Continuous spots were
made at a spacing which would permit of the cut samples
entering the jaws of the testing machine. The spots
were made in various ways to detect any gain or loss by
the order in which the spots were made. After the plates
were spot welded, each spot — about thirty for each
10-foot plate — was cut by means of the oxy-acetylene.
flame, then pulled in an Olsen tensile-testing machine.
These tests were made to demonstrate the uniformity of
work that might be expected in regular practice and to
determine to what extent the operator or the machine
entered into the problem. As will be seen, the results
are very clear on these points.

To convince those who might feel that spot welding
would not be able to withstand the shocks which threaten
the destruction of a ship when she strikes a submerged

DEMONSTRATION OF HEAVY SPOT WELDING 63

rock, a sample of 20-pound plate 6 feet long and 24
inches wide with a bounding angle 3^ by 3^ by 7/16
inches was spot welded. This was subjected to a rough
test under a 30-ton steam hammer. The highest static
pressure that the hammer could exert did not affect this
sample while placed on edge. The hammer was then
raised, the sample braced on the angle and a blow de-
livered. After several blows, the sample being returned
to the same relative position after each pounding, the
angle on the top edge showed fatigue. The sample was
then turned and blows repeated at different positions in
the length. A careful examination showed that where
the angle had broken away from the plate the welded
joint had torn the original metal with it. This sample
showed distortion and it was noted how tenaciously the
spot- welded angle clung to the plating.

As a sequel to the uniformity test a number of
shorter test pieces were made up of 20-pound plating
and lapped to varying widths, spot welded with two,
three, and four rows of spots. These tests were made
as a comparsion with double-, triple-, and quadruple-
riveted joints.

These latter tests with the imiformity tests on ^-inch
steel plate brought the demonstration to an end. The
^-inch uniformity tests could be made only on the
duplex spot-welding machine, which was not available
until near the conclusion of the allotted time. It was
fortunate that this series of tests could be completed, as
it furnishes data of value to those who may wish to go
forward with this process.

Results of Tests. — Before spot welding the first test
pieces of 20-pound plating it was essential that a deter-

SPOT AND ARC WELDING

mination of the time for welding this thickness of mate-
rial be obtained. It was also necessary for the operator
to know the effects of the different voltage taps. Thirteen
samples were spot welded with a single spot and with
varying time and cmrent. These samples were all made
on the 12-inch spot welder and pulled for ultimate ten-

TABLE I.

• Samples for Adjustment or Machine.

Sample
No.

Open-circuit

Welding

intimate

Voltage

Time

Load

560

40,600

560

30,600

360

20,100

410

20,800

455

36,700

455

36,600

455

35,600

407

43,000

460

45,300

510

45, 100

510

43,600

510

40,400

510

73,100

Remarks

sile. Table I gives the results as taken. The second
column is the voltage reading of the taps on the selector
panels with the main contactor switch open; that is, no
work being done. The drop in voltage is not taken into
account. The last sample differed from the others in
that it consisted of three thicknesses of 20-pound plating.
This was to determine the conditions for welding three
thicknesses. Fig. 17 gives an idea of the appearance of
a spot weld after undergoing shear in the tensile-testing
machine. Fig. 18 illustrates the torsional strains suf-
fered by the sample while under tensile pulling. This
is caused by the overlap of the two pieces, throwing the
action of the testing jaws off centre.

DEMONSTRATION OF HEAVY SPOT WELDING 65

An independent set of electrical readings were taken
as a preliminary step, so that afterwards results might
be interpreted from them, but it was determined that the
electrical conditions did not vary to a point detrimental
to the welding require-
m e n t s. Consequently,
electrical readings were
not taken in every in-
stance. During the uni-
formity tests electrical
readings were taken as a
check and proved the cor-
rectness of this decision.
Table II gives the open-
voltage readings across
the primary winding of
the 12-inch spot-welded
transformer when the
connections on the se-
lector panel were made as
' indicated by the nimierals
on the panel. These read-
ings at once established
the practice for the selec-
tion of the taps, and so by Pm, 17.— single ^wtwdd m t-o aiclcnma of

trial of the various thick- ' "

nesses of stock material or the number of the thicknesses
required to be welded, the open-circuit voltage was ap-
proximately known. With this information those famil-
iar with the electrical design could estimate the amount
of current passing through the electrodes.

For a full investigation into the electrical conditions

SPOT AND ABC WELDING

up^''"X;«

Voltage

v^-'Z^

VolUg.

1 and 5

277

3 and S

375

1 and 6

soe

3 and 6

407

1 and 7

SS3

3 and 7

456

1 and S

410

3 and B

506

353

4 and S

427

2 and 6

360

4 and 6

460

2 and 7

410

4 and 7

510

2 and 9

463

4 and 9

560

DEMONSTRATION OF HEAVY SPOT WELDING 67

CONTACTOR-

SEISCTOR PANEL

CURR£NTTRANS.
V2O0-

HlZIh

MULTIPLYEP 4-1

comecTiONS or testing instruments

Fig. 19. — Connections of testing instruments.

WELDER

SPOT AND ARC WELDING

it was decided to take a complete set of readings. In-
struments were introduced into the circuit as shown in
Fig. 19 and the mean of four readings is recorded in
Table III. These readings were taken under load. The

TABLE in.

Electrical Readings of 12-inch Machine.

Direct

260
282
340
385
310
330
375
415
350
375
415
440
360
410
440
475
♦485

P
X 200

1.75

1.95

2.15

2.4

1.9

2.2

2.45

2.7

2.2

2.5

2.9

3.2

2.7

2.8

3.2

3.4

3.05

350
390
430
280
380
440
490
540
440
500
580
640
540
560
640
680
610

w
p

200x4

P.F.

Direct

K.W.

40.

.44

48.

.435

64.

.436

76.

.703

3.2

56.

.475

• 2.4

64.

.44

2.4

72.

.392

112

89.5

.40

3.1

64.

.415

72.

.384

105

84.

.35

1.5

122

97.5

.346

• 1.5

76.

.391

2.7

110

88.

.384

2.6

130

104.

.37

1.6

142

114.

.354

2.2

147

118.

.40

OHmections

1 and 5
1 and 6

1 and 7
land 8

2 and 5
2 and 6
2 and 7

2 and 8

3 and 5
3 and 6
3 and 7

3 and 8

4 and 5
4 and 6
4 and 7
4 and 8
4 and 8

(A X 52)
P

18 200
20 280
22 360
14 560

19 760
22 880
25 480
28 080
22 880

27 000

30 160
33 280

28 080

29 224
33 280
35 360

31 720

* This set oi' readings taken with the plates set full width of gap 12 inches.

varying connections refer to the same numerals (Fig.
11) as marked on the selector panels. The last column,
secondary amperes, is calculated by multiplying the pri-
mary amperes by the number of primary turns in the
12-inch spot- welded transformer, which is 52. The last
set of readings was taken with the steel-plate test piece
shoved back in the machine in order to introduce the full
reactance and to ascertain the effect this would have
upon the power factor.

While taking these readings it was noted that the
secondary volts across the electrodes gradually lowered

DEMONSTRATION OF HEAVY SPOT WELDING 6d

as the welding time continued. No noticeable change
was seen in the primary readings during this short period.
Table IV shows readings taken on two different sets of

TABLE IV.

Showing Drop In Secondary Volts.

V
p

P.F.

V^ (Direct)

Connec-

25
sec.

tions

K.W.

sec.

16
sec.

1.5
2.0
1.7
1.7

20
sec.

sec*

3 and 6
3 and 6
3 and 7
3 and 7

370
370
410
415

2.4 480
2.4 480
2.75 550
2.70 540

100 80
100 80
115 92
115 92

.45
.45
.407
.41

2.5
3.3
3.0
3.0

2.3
2.5

2.7
2.7

2.0
2.2
2.0
2.0

1.3
1.5
1.5
1.5

1.3
1.3
1.5
1.4

1.25

connections with the secondary voltage read every
five seconds.

Electrical readings were taken during the spot weld-
ing of the lug attachment on the 20-pound-plate test
piece. These readings are given- in Table V and by ref-

TABLE V.
Electrical Readings Taken While Welding Sample Lug Attachment.

Connections

P.F.

■

K.W.

4 and 7

425

3.2 640

120

.353

2.3

33280

4 and 7

425

3.5 700

125

100

.336

2.5

36400

4 and 7

430

3.1 620

130

104

.39

3.0

32240

4 and 7

425

3.2 640

120

.353

2.5

33280

3 and 7

415

2.8 560

110

.379

2.5

29120

3 and 7

410

2.9 580

110

.407

2.4

30160

3 and 7

420

2.9 580

104

.340

3.0

30160

3 and 7

420

2.95 590

110

.355

3.2

30680

erence to the same connections shown in Table III the
constancy of the electrical conditions may be judged.
Ten samples of this type were spot welded and tested for
tensile in the Olsen testing machine. Table VI gives

SPOT AND ARC WELDING

the results of this test and Table VII gives the compa-
rable results for the riveted sample. Fig. 20 illustrates
this comparison and shows compositely the general re-

TABLE VI.
Tensile Test of Luo Attachment.

SPOT WELDED.

(Electrode pressure 15,000. Time on spots 10 seconds.)

Sample

Open-orcuit

Ultimate Load

No.

Voltage

Pounds

1-1

455

98,000

1-4

455

105,700

1-S

455

92,000

1-4

455

89,000

1-5

407

89,000

1-6

407

99,300

1-7

407

67,000

1-8

407

63,700

1-9

455

68,500

1-10

455

73,900

Bemariu

Sanmle forced out of machine

Angle bent away from plate .176"

4 spots broke

Sample forced from machine

All spots sheared

Sample forced from machine

Spots 1 and 2 tore away

Sample forced from machine

suits of the tests. On the left is the riveted sample which
invariably sheared three or all of its rivets under an ulti-
mate tensile of 45,000 to 48,000 pounds. In the middle

TABLE VII.
Tensile Test of Lug Attachment.

RIVETED.

Sanq>le No.

Ultimate Load

Remarks

2-1

48,000

Sheared all 4 rivets

2-2

47,400

Sheared all 4 rivets

2-3

46,700

Sheared 3 rivets

2-4

47,200

2-5

49,700

Sheared 3 rivets

2-6

44,000

Sheared 3 rivets

is the spot-welded sample which would spring from
under the jaws of the testing machine after an ultimate
load of 89,000 to 100,000 pounds. To the left an attempt
was made photographically to show the twisting of the

DEMONSTRAITON OF HEAVY SPOT WELDING 71

20-pound plate which caused the sample to spring from
the jaws of the testing machine. It will be noticed in
Table VI that the last four samples tested at an ulti-

Fra. 30. — Compuuon li liieLed ind ipot-wdded test piece.

mate load of 63,700 to 73,900 pounds. It will also be
seen in Table V that the comiections were changed for
these four samples. This was done after the first
samples were pulled in order that tests mi^t be made

7« SPOT AND ABC WELDING

that would shear the lug attachments in order to enable
a view of the spot weld. The results showed that a good
weld could be made at reduced amperage.

The riveted test pieces of the cross-connection sample
were next prepared for tensile test. Before this could
be done small pads of J/^-inch plate were arc welded to
the small sides of the sample (see Fig. 21). These pads

were about 3^ inches wide and were found necessary in
order to obtain a satisfactory pull in the testing machine.
These test pieces were too heavy for the testing machine,
which though originally rated at 150,000 pounds capac-
ity, was only safe to operate up to 124,000 pounds. At
this ultimate load the ten %-inch rivets were either
sheared entirely, or the angle broke, or the rivets
stretched. The spot-welded samples were in excess of
the capacity of the 12-inch spot welder for welding three

DEMONSTRATION OF HEAVY SPOT WELDING 73

thicknesses. A number of samples were made and
pulled with interesting results as far as examination of
the condition of the spot welds after shearing strains.
Later the angles were reduced to ^-inch thickness^ but
when this change was made the uniformity tests were
in progress and it was not considered of value to con-
tinue this comparison. In view of the original intention
which was the relative value of mechanical caulking
versus electric welding, a sample was made both spot
welded and arc welded. That is, the edges of the angles
were arc welded. This test piece was set up in the test-
ing machine and resulted in breaking the head of the
machine. The strain on the sample in both a static and
dynamic sense must have been very great. This special
sample is shown on the left in Fig. 21 after the test had
been made. There were no signs of distress in any of
the joints. It has been sent to the Commercial Museum,
Philadelphia, Pa., where it may be seen by any one in-
terested. On the right in this same illustration is one of
the riveted samples in which the %-inch angle broke in
line of the rivets at an ultimate load of 118,900 pounds.
The results of the first trial for uniform strength of
spot welds in one long seam was not successful. It
brought to light the necessity of two important points
in the preparation of the steel plates and f ocussed atten-
tion on the serious problem of the electrode tips. Dur-
ing the first test no care was given as to the condition of
the material. It was spot welded as received. If it hap-
pened to be clean on the surfaces next to the electrode
tips these tips were still used ; if, on the other hand, much
scale and rust rested on these surfaces the tips were
badly burned with the result that they had to be renewed.

SPOT AND ARC WELDING

This was one of the reasons why the 27-inch machine was
now used. The builders had changed the method of re-
newing the tips which reduced the time and inconveni-
ence very considerably. In like manner in the early

TABLE vm.

GiASQOw Iron Works, Pottstown.
Feb. 24, 1919.

No.

Load Lbs.

Time Sec.

Remariu

Electrode
Changed

49,700

41,600

40,600

No. 3 (b)

41,100

41,300

43,700

46,100

No. 2 (b)

45,200

9
10

51,200
35,900

12
12

No. r (c)

41,700

40,700

44,200

No. 2 (a)

43,800

44,500

41,200

31,300

No. 1 (b)

43,700

39,400

20
21

40,800
43,400

12
12

No. 1 (a)

41,900

40,300

46,800

No. 2 (c)

42,600

* Re-Spotted

*26

43,700

Full time after 26 |

42,100

41,600

Fore plate

No. 3 (a)

44,000

48,800

Pulled spot out

IC— 10" Plate 20 lbs.— M"— «H" lap— Green operator— «7" G. E. Spot Welder.

Present at test: Green, Schrader and Homor.

Avera^ break, 42,750 lbs.

Sbeanng stress (single shear) steel rivets and steel plates:

Half-inch, 12,250 lbs.

Three-quarter inch, 25,600 lbs.

Seven-dghths, 34,100 lbs.

One inch, 43,700 lbs.
Rusted, scale.

DEMONSTRATION OF HEAVY SPOT WELDING 75

tests no attention was paid to the surfaces of the mate-
rials between the lapped portions. A few tests indicated
clearly that the reverse condition was more desirable for
welding, namely, that the surfaces between the plates to
be joined be dirty, i.e., have some mill scale or rust. This
does not mean that clean surfaces cannot be welded, but
that more successful and uniform welds are made in
shorter time, with less current, and less pressure. This
fact must be observed by those who wish to repeat the
results here shown.

The results of the second set of spots made for the
uniformity tests are given in Table VIII. These spots
were not made consecutively, as it was thought at the
time that the sequence of welds might bear some relation
to the results. This consideration was not borne out by
subsequent tests. So that the following samples were
all made consecutively. The test pieces were all of the
same nature: Two 10-foot lengths of a desired thickness
of plate; these two pieces lapped about 2J^ inches to 2%
inches, arc welded at intervals along the edge, and then
spot welded from one end to the other. The first sample
was unfortunately cut in a shearing press which so de-
formed the samples that they were awkward to place in
the testing machine. The surfaces next to the electrode
tips for all samples were cleaned by a portable grinder
in way of the spots before welding. Although no par-
ticular insistence was needed, the plates were placed so
that the surfaces between had the ordinary mill scale
and rust usual in practice. It could be easily noted when
pulling the samples which were the relatively clean por-
tions of the plates.

A series of tests were now made using )4"> H'f /^">

SPOT AND ARC WELDING

and ^-inch steel plates, including electrical readings for
each spot made. The results are given in Tables IX, X,
XI, and XII. It will be seen with what constancy the
electrical conditions were maintained and how Uttle they
affected the results. After these particular tests no elec-

TABLE IX.

Unipormitt Test J^-inch Steel.

Test: 10-foot Ji" S. Plate
27" G. E. Spot Welder

Pressure: 18,750 lbs. at electrodes
Voltage: 825. Amperes: 22,230
Time of each spot: 12 seconds

No. of Spot

Ultimate Load

1
Amperes

Remariu

29,700

23,400

Spot in line with tack weld

19,000

22,620

20,500

22,620

20,200

22,230

19,300

22,230

18,900

22,230

20,100

22,230

20,100

22,230

20,300

22,230,

19,400

22,230

26,400

22,230

Spot in line with tack weld

19,100

22,280

■

19,700

22,620

19,000

22,230

17,600

22,230

19,200

22,230

18,900

22,230

20,200

22,230

18,600

22,230

17,100

22,230

30,300

22,230

Spot in line with tack weld

19,900

22,230

20,600

22,230

20,900

22,230

19,800

22,620

22,700

22,230

19,100

22,620

19,000

22,620

20,000

22,620

26,000

23,400

Spojt in line with tack weld

Nona. — ^Present at welding test: Same as on ^" S. P.

Pulling tests made March 18. 1919, at Glasgow Iron Worics.

DEMONSTRATION OF HEAVY SPOT WELDING 77

trical readings were taken. As the ^-, ^-, and ^-inch
plates were below the designed rating of the machine the
results given are arbitrary, as the current taps may be
selected to give more or less current and successful weld-

TABLE X.

Uniformity Test %-iscn Steel.

March 13, 1919.

Test: 10-foot X" S. Plate
• 27" G. E. Spot Welder

Pressure: 18,750 lbs. at electrodes
Voltage: 395. Amperes: 27,900
Time of each spot: 12 seconds

No. of Spot

Ultimate Load

Amperes

Bemarics

42,100

28,470

27,800

27,690

27,300

28,080

30,900

27,690

29,100

27,690

30,100

27,690

32,000

27,300

30,500

26,910

32,600

27,300

33,000

26,910

39,900

27,300

Spot in line with tack weld

27,600

27,456

30,000

27,456

30,400

27,534

30,100

27,300

32,200

27,300

36,100

27,300

31,600

27,690

30,300

27,690

34,100

27,690

Spot in line with tack weld

31,000

27,690

25,700

27,300

26,500

27,690

29,100

27,300

29,200

27,300

29,300

26,910

31,100

27,300

Spot pulled out

30,800

27,300

Spot pulled out

30,900

27,610

28,000

28,470

Spot pulled out

Notes. — Present at welding; test: Sdtzer and Newell (Steamboat Inspectitm Service), Martin (Ameri'
can Bureau of Shipping), Stewart and Homor.
Pulling tests made March 18, 1919, Glasgow Iron Works.

SPOT AND ARC WELDING

ing accomplished by a relative adjustment of the time of
making the weld. This brought these sizes into the realm
of shop-production questions with which this demonstra-
tion had nothing to do.

TABLE XI.

Uniformity Test J^inch Steel.

March 12, 1919.

Test: 10-foot 3^" S. Plate
27" G. E. Spot Welder

Pressure: 18,750 lbs. at electrode
Taps set constant 4-8
Voltage: 440. Amperes: 31,200
Time of each spot: 13 seconds

No. <rf Spot

Ultimate Load

Amperes

Remarks

44,100

32,370

Spot in line with tack weld

42,000

31,200

38,000

31,200

33,500

31,200

28,900

31,200

34,000

31,200

35,000

30,810

44,000

31,200

38,500

31,200

34,100

30,810

29,000

30,810

30,700

31,590

35,300

31,200

38,000

30,810

32,600

30,810

28,800

31,590

27,600

31,590

49,200

30,810

Spot in line with tack weld

36,500

31,200

40,700

31,200

•21

38,000

31,590

33,800

31,200

31,600

31,200

23,900

31,200

30,200

31,590

30,400

30,810

30,200

31,590

51,200

32,370

Spot in line with tack weld

Notes. — ^Present at test: J. B. Stewart, H. A. Homor.
Oeui sted.
Pulling tests made March. 18, 1919, Glasgow Iron Worics.

DEMONSTRATION OF HEAVY SPOT WELDING 79

It will be seen that the time given the ^-inch samples
was 13 seconds (Table XI). It was expected that
better residts .would be attained by increasing the time,
so a smaller sample (5 feet in length) was run through

TABLE XII.

Uniformiit Test X-inch Steel.

March 13, 1919.

Test: 10-foot X" S. Plate
27" G. E. Spot Welder

Pressure: 18,750 lbs. at electrodes
Voltage: 440. Amperes: 31,200
Time of each spot constant: 18 seconds

No. of Spot

Ultimate Load

Amperes

Remarics

51,200

32,370

31,590

Sample held for other ^esfs

1^,566

31,980

19,700

32,370

Sample held for other tests

28,600

31,980

26,700

31,980

22,200

32,370

15,000

32,370

18,600

31,980

Spot in line with lack weld

23,200

32,370

38,300

32,370

24,600

32,214

26,600

32,370

31,100

32,370

33,400

32,214

31,000

32,136

29,100

32,370

■

20,900

32,370

44,100

31,590

Spot in line with tack weld

35,700

31,590

32,800

31,200

27,900

31,590

30,000

31,200

35,800

31,200

31,300

31,200

29,600

31,590

21,200

31,200

35,700

31,595

49,900

32,370

Notes. — ^Present at wdding tests: J. B. Stewart, Selbser (Steamboat Inspection Service), Newell
(same), H. A. Homor.
Pulling tests made March 18, 1010, Glasgow Iron Works.

80 SPOT AND ARC WELDING

test without electrical readings at 15 seconds with results
as shown in Table XIII. The 5^ -inch sample (Table
XII) was welded with a constant time of 18 seconds for
each spot, and as this was the maximum capacity of this
machine better results could only be obtained by increas-
ing the time. Table XIV gives the results on a 5-foot
sample with spot welds made in 25 seconds, all the other
conditions remaining the same.

These results were laid before the technicians of
Lloyd's Register of Shipping, the American Bureau of
Shipping, and the U. S. Steamboat Inspection Service
of the Bureau of Commerce. It was then suggested
that a similar series of tests be made as a check upon
what had been done and with variations in the time of
making the spots in each sample. Electrical readings
were not required for each spot, but readings of the
circuit were taken from time to time to assure that they
were holding to a constancy that would not disturb the
results. The records of these tests are given in Tables
XV, XVI, XVII, and XVIII. The first ten spots of
the J4"^ch sample were given 16 seconds each, the next
ten spots were given 12 seconds each, and the last ten
spots 8 seconds each. The average ultimate load in
pounds for the first ten spots was 16,720 pounds, for the
next ten spots it was 17,540 pounds, and for the last
series 17,560 pounds. The first ten spots of the ^-inch
sample were given each 16 seconds, with an average ulti-
mate load of 34,720 poimds, the next ten spots were
given 12 seconds each, with an average ultimate load of
32,790 pounds, and the last ten spots were given 8 sec-
onds each, with an average ultimate load of 27,770
pounds. The first ten spots of the 3^ -inch sample were

DEMONSTRATION OF HEAVY SPOT WELDING 81

TABLE Xni.

Uniformity Test ^-inch Steel.

Test: 5-foot J^" S. Plate
Pressure: 18,750 lbs. at electrodes
Voltage: 440. Amperes: 31,200
Time of each spot: 15 seconds

No. Spot

Ultimate Load

Remarica

50,000

Spot pulled out

Sample held for other tests

40,000

41,800

41,900

43,000

Sample held for other tests

38,i66

35,000

38,600

Sample held for other tests

35,100

37,700

Sample held for other tests

45,466

Welding done March 19, 1919.

Pulling tesU, March 20, 1919, PotUtown, Pa.

TABLE XIV.
Uniformity Test ^^-inch Steel.

Test: 5-foot %" S. Plate
Pressm«: 18,750 lbs. at electrode
Voltage: 440. Amperes: 31,200
Time of each spot: 25 seconds

No. Spot

Ultimate Load

Remarics

Sample held for other tests

52,100

Spot in line with tack Weld '

47,500

Sample held for other tests

49,200

49,400

64,700

Sample held for other tests

52,600

54,200

Sample held for other tests

53,700

52,900

50,900

60,300

Average, 58,409 lbs.

8£

SPOT AND ARC WELDING

TABLE XV.

Lloyd's Tests:

IC— Ji" S. P. Tensile Test.

Mabch 27, 1919.

Glasgow Iron Works, Pottstown, Pa.

Pressure at electrode constant: 18,750 lbs.
Amperes: 22,230. Volts: 325

No.

Time

Diameter

Ultimate

R^vn n rira

Spot

Sec.

Spot, Inches

Load, LtM.

JliCIUAvAJS

20,600

%''

19,000

Pulled out spot

Va"

19,800

Gas hole

18,700

16,700

Two small gas holes

15,500

7 '

'){,"

15.100

Small gas hole

13,400

Small gas hole

^K«"

12,200

Small gas hole

^K«"

16,200

17,900

Small gas hole (Mate cold — ^weld-
ing resumed)

16,000

18,500

Small gas hole

»K«"

16,700

Small gas hole

'Vit"

16,700

Small gas hole

^K«"

18,000

Small gas hole

18,000

16,900

Small gas hole

^K«"

18,100

Small gas hole •

»X«"

18.600

Small gas hole

17,200

16,700

17,300

Small gas hole

16,600

Small gas hole

%•"

16.100

Small gas hole

15,200

Small gas hole

16,800

Small gas hole

18,200

Small gas hole

21,800

Small gas hole. Rusted specimen;
slag between plates

'"if!'

19,700

Present at test: Same as ^" test.

Tests show a torsional strain on samples while in machine.
Average load first ten at nxteen seconds, 16,720.
Average load second ten at twelve seconds, 17,540.
Average load third ten at eight seconds, 17,560.

DEMONSTRATION OF HEAVY SPOT WELDING 83

TABLE XVI.

Lloyd's Tests.

10'— 5^" S. P. Tensile Test.

March 27, 1919.

Glasgow Iron Works, Pottstown, Pa.

Pressure at electrodes constant: 18,750 lbs.
Amperes: 27,700. Volts: 395

No.

Time

Diameter

Ultimate

Spot

Seconds

Spot, Inches

Load, Lbs.

• •

34,700

31,100

IHe"

34,900

IIW

36,900

34,500

35,600

34,300

34,100

34,800

iKe"

37,300

• 12

iMe"

31,700

'Hi"

28,800

32,200

34,400

36,200

33,800

30,500

32,300

34,800

33,200

25,400

'H,"

28,700 ■

%^'

25,500

^K«"

28,600

26,600

'){,"

24,600

Va"

22,000

»hV'

29,900

29,800

36,600

Remarics

Spot

Small
Small
Small

pulled out
^ gas hole
gas hole

gas hole

Small gas bole. Spot started to

pull out
Small gas hole
Small gas hole
Small gas hole
Spot pulled out
Highly rusty material
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole. Fracture around

edge of spot
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole
Small gas hole (Very slight)

Present at test: Same as %" test.

Tests show a torsional strain on samples while in machine.

Plate bent before spot puljed.

Average load first ten at sixteen seconds, S4,720.

Average load second ten at twelve seconds, S2,790.

Average load third ten at eight seconds, 27,770.

SPOT AND ARC WELDING

TABLE XVn.

Lloyd's Tests.

IC— H" S. P. Tensile Test.

March 27, 1919.

Glasgow Iron Works, Pottstown, Pa.

Pressure at electrodes constant: 18,750 lbs.
Amperes: 31,200. Volts: 440.

No.

Time

Diameter

Ultimate

Spot

Seconds

Spot, Inches

Load, Lbs.

• •

41,000

41,900

iKe"

42,800

l%" .

41,100

iKi"

39,100

iKi"

43,100

IKe"

41,400

40,700

41,600

1%''

38,100

39,000

1H«"

39,900

36,300

43,000

39,300

43,900

43,700

1K«"

41,600

41,800

IM"

50,200

33,500

31,200

^Ke"

24,300

34,000

33,100

33,900

%''

33,000

10 .

^K«"

35,500

35,400

44,900*

Remarks

End spot. 3j^"iwide plate. Plate fractured
Started to tear around rim of spot

Spot pulled out 1" diameter '

I^te fracturing in direction of grain. Spot

pulling out
Spot puUed out
Plate fracturing in direction (1%") of grain.

Spot pulling out
Plate fracturing in direction (1^") of grain.

Spot pulling out
Slight plate fracture 1" around edge of spot.
Plate fractured. Spot pulled out
Fractured around edge of spot
Heavy fracture around edge of spot
Slight fracture around edge of spot
Slight fracture around ckige of spot
Heavy fracture around edge of spot. Spot

pulling out
Shght fracture around edge of spot
Slight fracture around edge of spot
Heavy fracture around edge of spot. Spot

pulling out

Fractured around edge of spot

* Plate bent before spot pulled.

Tests show a torsional strain on samples while in machine.
Average load first ten at twenty seconds, 41,080.
Average load second ten at fifteen seconds, 41,780.
Average load third ten at ten seconds, 34,180.

DEMONSTRATION OF HEAVY SPOT WELDING 86

TABLE XVin.

Lloyd's Tests.

10'—^" S. P. Tensile Test.

March 27, 1919.

Glasgow Iron Works, Pottstown, Pa.

Pressure at electrodes constant: 18,750 lbs.
Amperes: 31,200. jVolts: 440

No.

Time

Diameter

VWimtitf

Remarks

Spot

Seconds

Spot, Inches

Load, Lbs.

• •

59,500

iKe"

53,600

IK"

53,700

iKe"

54,700

44,900

42,100

45,800

44,100

iKe"

45,500

•

43,600

36,900

1M«

49,200

38,300

34,300

•

33,900

33,200

41,100

32,600

26,700

26,900

'%,"

32,300

29,900

24,000

Gas hole in spot

28,700

21,200

29,900

18,000

'%,"

33,500

30,700

Regular spot adjacent to arc- weld
tack

i%"

29,400

Present at test: Messrs. Aspinall, Frickey (A.I.S.C.), J. B. Stewart, M. I. Eshbock, H. A. Homor.
The tests show a torsional strain on samples while in machine.
Average load first ten at twenty-five seconds, 48,740.
Average load second ten at twenty seconds, 36,310.
Average load third ten at fifteen seconds, 27,760.

SPOT AND ARC WELDING

given 20 seconds each, with an average ultimate load of
41,080 pounds, the next ten spots were given 15 seconds
each, with an average ultimate load of 41,780 pounds,
and the last ten spots were given ten seconds each of an
average ultimate load of 34,180 pounds. The first ten
spots of the ^-inch sample were given 25 seconds each,
with an average ultimate load of 48,740 poimds, the
next ten spots were given 20 seconds each, with an aver-
age ultimate load of 36,310 pounds, and the last ten
spots were given 15 seconds each with an average ulti-
mate load of 27,760 pounds.

Table XIX gives the results of pulling tests of a
5-foot sample of ^-inch plate spot welded with the
duplex machine. All the spots were made in pairs except

TABLE xix.

Tensile Test Ji" S. P.

April 11, 1919.

Glasgow Iron Works, Pottstown, Pa.

Welding done on 6-ft. Duplex Machine

No. Spot

Ultimate Load

Remarks

66,700

81,100

Single spot

• 18

66,800

61,900

Broke plate

54,400

72,300

Broke plate

61,000

64,100

Broke plate

43,300

70,300

47,800

64,500

45,400

61,400

•

Average, 61,500.

Sheer 1" rivet, 48,700.

Present at welding: Stewart and Homor, April 0, 1919.

Present at pulling: Stewart and Eshbock, April 11, 1919.

DEMONSTRATION OF HEAVY SPOT WELDING 87

Number 17, which, as indicated, was made with a single
pair of electrodes in 40 seconds. It required eight opera-
tions to complete this sample and the spots were made
under a pressure at each electrode of approximately
22,500 pounds in the following time intervals : 30 and 28
in fifty seconds ; 26 and 24 in fifty-three seconds ; 22 arid
20 in fifty-five seconds; 18 and 16 in fifty-five seconds;
29 and 27 in fifty- four seconds; 25 and 23 in sixty-five
seconds; 21 and 19 in sixty-five seconds. This sample
was very bulky, difficult with rigging at hand to hold in
the spot welder, and, in addition, the arc-weld tacking
repeatedly failed to hold the plates together. For these
reasons spot No. 30 being the end spot did not give good
results. The other spots upon examination were excel-
lent, and there was no doubt that this performance could
be repeated indefinitely with as good or better results.
Electrical readings were taken for each of the eight
operations; the calculated secondary current averaged
35,520 amperes, the open voltage averaged 481, and the
closed voltage averaged 369. The average voltage drop
was 112 volts. This was an excessive amount, but all
the available materials had been exhausted for reducing
this drop which initially was over 150 volts. The results
under this limitation proved that the apparatus when
supplied with the correct voltage and current would
more than meet the specified requirements. Also that
the element of time in making the spots under given con-
ditions is a relative question involving other elements
than the electrical characteristics and appertain to shop-
production methods.

The results of the final tests made are given in
Table XX and show the tensile strength of two and

SPOT AND ARC WELDING

three spots in a row. Samples were made up of ^-inch
steel plate and spot welded in the same manner as the
other samples. After spot welding they were cut in
strips and pulled for ultimate tensile. The entire series
is not shown. The purpose of this test was to investigate
and compare the lapping of the plates as in a single-,
double-, treble-, and quadruple-riveted joint. The
single-spot samples were lapped 3 inches and the results
were similar to those already given. The ^double-spot
sample was lapped 6 inches with results as shown in
Table XX. The treble-spot sample was lapped 9 inches

TABLE XX.

Tensile Test.
April 10 and 11, 1919.
Glasgow Ibon Works, Pottstown, Pa.
y 2 Spots in Y^" S. P.

No. Spot

intimate Load

Remarks

1
2
3
4

5
6

85,800
66,200
77,700
74,100
85,100
83,400

PuUed April 10, 1919
Piilled April 10, 1919
PuUed April 11, 1919
PuUed April 11, 1919
PuUed April 11, 1919
PuUed April 10, 1919

Tensiu
3 Spots in

B Test.

r J^" S. P.

No. Spot

Ultimate Load

Remarks

1
2
3
4
5
6

115,800
92,200
89,900
85,200
97,000

124,500

Pulled April 10, 1919
Pulled April 10, 1919
Pulled April 11, 1919
PuUed April 11, 1919
PuUed April 11, 1919
PuUed April 10, 1919
Limit of testing machine
(sample not afiFected)

Present at welding: Stewart and Homor.

Present at pulling April 10, 1919: Messrs. Higgins, Stewart, Eshbock,

Homor, Smith, Ker.
Present at pulling April 11, 1919: Eshbock and Stewart.

DEMONSTRATION OF HEAVY SPOT WELDING 89

with results as given in the same table. The quadruple-
spot sample was lapped 12 inches. This is the standard
" lapping " in each case for a ship's riveted joints. The
quadruple-spot sample either exceeded the pulling
capacity of the testing machine, so that no results were
possible, or the original plate material would break some
distance from the welded joint, usually about 4 to 6
inches from the last spot. These investigations lead to
the conclusion that much material can be saved in steel
ship construction when designs are based on the spot-
welding method.

What do these results mean in a broad view of the
joining of heavy steel members? The answer is clear.
Those who are familiar with ship designs know the con-
servative requirements placed upon the strength of
joints and strength of materials for the vital parts of the .
hull structure. It would seem logical that any method
of joining that would meet or exceed the specifications
of Lloyd's Register could be employed in other applica-
tions. Compare Lloyd's Requirements with the results
given above. For thicknesses of steel plate 0.22 and
under, ^-inch steel rivets are required; for thicknesses
of 0.22 and not exceeding 0.34, ^-inch rivets; for thick-
nesses of 0.34 and not exceeding 0.48, ^-inch rivets; for
thicknesses of 0.48 and not exceeding 0.66, J^-inch
rivets; for thicknesses of 0.66 to 0.88, 1-inch rivets. The
single-shearing stress of rivets of the sizes mentioned are :

14 inch steel rivet 12,260 pounds

% inch steel rivet 18,300 pounds

% inch steel rivet 26,600 pounds

yg inch steel rivet 34,100 pounds

1 inch steel rivet 43,700 pounds

90 SPOT AND ARC WELDING

According to these loiles, 3^ -inch rivets woiJd be re-
quired for J4"iiich steel plates. Referring to Table XV
it will be noted that the average ultimate tensile of the
spots made was approximately 17,000 pounds. For %-
inch steel plates ^-inch rivets are required. Table XVI
shows an average ultimate tensile of 31,760 pounds for
a spot in ^-inch steel plate. For ^-inch steel plates ^-
inch rivets are required. Table XVII gives an average
ultimate tensile of 39,010 pounds for a spot in ^-inch
steel plate. For 5^ -inch steel plates %-inch rivets are
required. Table XVIII gives an average ultimate ten-
sile of 37,603 poimds for a spot in ^^-inch steel plate.
These averages are over the whole test, including the
short and medium time given two-thirds of the spots.
The averages would be excessive as compared to the
single shear of the appropriate rivet if only the best spot
welds were taken. For %-inch steel plates 1-inch rivets
are required. Table XIX gives an average ultimate of
61,500 pounds for a spot in ^-inch steel plate. If this
data were composed of isolated tests then questions
might be raised as to the process, but these tabulations
are repetitions with no special considerations other than
that of assuring uniformity.

Take the average of spots in Table XX, represent-
ing double shear: in the case of steel rivets a double
shear is not twice the single shear, but for argument
agree that it is. This would mean that the shearing
stress of two ^-inch steel rivets was 51,200 pounds. The
average of two spots in ^-inch steel plate is a little more
than double the single spots, 78,683 pounds. The ten-
sile strength of two spots in ^-inch steel plate is more
than the tensile strength of three ^-inch rivets based

DEMONSTRATION OF HEAVY SPOT WELDING 91

on multiplying the single shear of a ^-inch rivet by 8.
It is not known whether the shear of three ^-inch rivets
has ever been definitely established, but comparative
tests made during this demonstration would lead to the
opinion that the ultimate tensile would be much less. It
seems unnecessary to note in the same table that the
average ultimate tensile of three spots in 3^ -inch steel
plate was 100,433, although the last set of spots (No. 6)
was not pulled to its ultimate.

An interesting set of tests with three and four thick-
nesses of various size plates was contemplated when the
demonstration was brought to a close. It is reasonable
to assume that if the rated voltage had been supplied to
the duplex spot welder, this machine would have welded
three and four thicknesses of 5^ -inch plate. Undoubt-
edly the same uniformity of spot welding would have
followed the results as shown. It is needless to speculate
on what this apparatus could do, it is sufficient to record
here its performance.

One of the best practical results was the discovery
of a fairly simple and non-destructive method of in-
specting spot-welded work after completion. During
the tests it was suggested that this might be accomplished
by punching a hole, or holes, on the line of demarkation
of the spot weld. A number of experiments was made
which gave confidence in the method. In finished work
these test samples could be taken of a reasonable nimiber
of spots and with sufficient variation to assure that the
work was uniform and well done. To further empha-
size this method of inspection the ship's floors previously
mentioned were sent to the shipyard and a request made
for an inspection of this nature of independent parties.

92 SPOT AND ARC WELDING

It was known by those who had followed the work of
spot welding that there were good, bad, and indifferent
spot welds made, and the intention was to observe how
closely this method of inspection could be relied upon.
The results more than exceeded expectation. The
punchings showed all degrees of welds. The finished
article is not greatly injured by this method as the
punched hole can be refilled by either spot welding a
stud, refilling by means of the metallic electrode, or
by riveting.

General Comment. — There are two primary and seri-
ous considerations to be given the process of heavy spot
welding. They are briefly, electrode-tip protection and
the preparation of materials to be welded. These two
points are closely interlocked and greatly circumscribed
by our lack of proper metals or knowledge of known
metals. There are two secondary considerations which
may be easily improved upon and which may be stated
briefly: (1) The separation of the electrodes, and (2)
the shape of the head of the spot welder. These sec-
ondary matters are mixed with questions of the appli-
cability to shipbuilding and shop-production methods.

Fig. 22 illustrates roughly the type of electrode-tip
protection used on the 12-inch welder with which the
early tests, including the ship's floors, were made. It
consisted of a strip of soft copper which covered aftd
conformed to the top of the electrode. When this strip
was worn down it could be renewed by slipping out the
steel through-bolt. The difficulties that were met with
in practice were the " freezing " of the copper strip to
the electrode, thus requiring it to be chiseled off, and
the deformation of the electrode itself when under severe

DEMONSTRATION OF HEAVY SPOT WELDING 93

working strains as well as a result of the method of break-
ing away the strip from it. This latter necessitated
either the removal of the electrode and the machining
of same or the fiUng or truing of the electrode tip before
replacing the copper strip. This design was modified
and much improved upon for the 27-inch and the duplex
machines. A steel collar was fitted to the electrode by a

THROUGH 'BOLT

COPPER STRAP '/e

t»

^OPPER ELECTRODE

ELECTRODE WITH COPPER STRAP

Fig. 22. — ^Electrode with copper strap.

coarse thread and was removable by means of a spanner
wrench. This steel collar securely held separable tips in
close contact. As these tips were pyramidal in form the
crushing pressure deformed the tip, but did not affect
the electrode proper. The tips may be re-machined,
and as they go through an annealing process in the act
of spot welding they should do duty several times before
final scrapping. Their cost is insignificant as compared
to the electrode cost or the work which they are instru-
mental in performing. It was foimd essential in the

94 SPOT AND ARC WELDING

uniformity tests of the heavy materials to replace these
tips quite frequently, and if this is found necessary in
regular production work it raises a serious problem as
to the quick action of the process. Further develop-
ments will be needed to solve this problem, although sug-
gestions having in view automatic attachments for
renovating the tips have been made.

Naturally a great reduction in the number of tip
renewals is accomplished by the proper preparation of
the surfaces of the materials coming in contact with them.
The tests at Pottstown proved this and justified the
opinion of the designer of these machines. The subject
raises a large question. The suggestion in practical
steel shops of cleaning the surfaces of steel plates and
shapes brings down upon the spot-welding process a
cloud-burst of opposition. Three methods for cleaning
steel have been brought forward, but none are looked
upon as solving the difficulty. By pickling, that is, dip-
ping the steel in acid, the mill scale and rust may be
removed. The surfaces may be cleaned by sand-blast,
or they may be prepared as in these tests by the use of a
portable air grinder. The first proposal necessitates a
large installation expense and causes delay. The second
is a dangerous operation to perform on a large scale in
an open shop, as the flying sand or dust particles enter
the small parts of adjacent machinery as well as affect
neighboring workmen. The last proposal is feasible,
could be done while other matierial is going through the
spot-welding process and would not be excessive in cost.
The opponents of the spot-welding process cannot but
believe that this is an awkward makeshift. There is a
probable solution in a suggestion made to equip the

DEMONSTRATION OF HEAVY SPOT WELDING 95

spot-welding machine with an automatic grinder under
the control of the operator. This method would permit
the operator to be the judge as to the fitness of the mate-
rial and give him the option of saving time either by
cleaning the materials or by more frequent renewals of
the electrode tips. Besides this utilitarian advantage of
clean steel for welding there is a humanitarian reason.
When there is an accumulation of rust or mill scale next
to the electrodes a higher resistance is provided at the
point of contact. The desired place for this high resist-
ance is between the plates being welded. Accompany-
ing the high resistance at the electrodes the instantaneous
production of intense heat causes the slag to be thrown
out with great violence. This condition is dangerous for
the operator, not only for his eyes, which may in time be
affected by the radiant energy, but also to his body into
which may enter the fine slivers of molten slag.

Though secondary because ^hey appertain to me-
chanical features, yet important in that they are neces-
sary, the distance between the electrode tips in machines
designed for all-around work should be readily adjust-
able. In the apparatus just described this distance was
fixed and provision was made to fit the electrodes so
that they could be removed from their holders, thus
allowing for the placement of the machine over obstruc-
tions, and the replacement of the electrodes when the
work and machines were properly set. Undoubtedly
such an arrangement would be awkward in any steel
fabricating shop. Time is an essential element in all
production work. The correction of this mechanical
difiiculty in new designs is too simple not to be insisted
upon. The shape of the head, or more descriptively the

96 SPOT AND ARC WELDING

nose, of the machine should be such as to permit the
electrodes to function in very close quarters. The at-
tachment of angles to plates, angles to angles, two angles
on opposite sides of the same plate, are common con-
nections in steel construction. It is particularly neces-
sary for this application that the electrodes be quickly
adjustable and that they be easy to manipulate in cor-
ners and along bounding angles. Although the elec-
trodes of the 12-inch welder had to be changed in order
to bring them in the proper position for welding the
boimdary angles of the ship's floors, it proved what was
one of the greatest previous objections to the process,
namely, that a good caulking edge could be left on the
angle after spot welding. As a matter of fact, a good
weld cannot be made on edges without danger to the
operator due to the throwing of melted slag. This me-
chanical difficulty can be overcome in as simple a manner
as the adjustability of the electrodes, and should be called
for in new designs of apparatus.

CHAPTER V

General Applications of Arc Welding

Those who began an investigation into the applica-
tion of arc welding in 1917-18 were naturally surprised
at the wide use of the process in repair work and cer-
tain manufactures. Indeed, for both land and marine
repairs the success of many of the applications had war-
ranted its approval by conservative inspection bureaus
and frequently insisted upon by the owners in preference
to older and more tried methods. It was for this reason
that the United States Navy Department adopted the
process for the quick repairs made to the damaged ma-
chinery of the interned German ships, and the success
accruing from this work lent impetus to the proposals
for its extension to shop construction. Although these
applications ^ are now fairly well known and recently
have received a greater publicity, it may be well to
briefly review them.

Repair Work. — There are three interesting points
connected with this subject: (1) Ease of application,
(2) the vital nature of the repairs, and (3) the cost.
The first and third items caused the introduction of the
process, the second item was the effect of its continued
success. Engineers associated with business men will
readily put into effect an innovation that carries with it
the dual saving of time and money without involving
great danger, but a similar group will refuse to establish
as a regular industrial practice a method that will hazard

^ ** Electric Arc Welding," Lincoln Electric Company, Cleveland, Ohio.
7 97

98 SPOT AND ARC WELDING

the lives of those who as a general rule are innocent of
the means employed.

The majority of repairs are usually required for
component parts of large pieces of machinery. The
older methods of repair made it necessary to disassemble
the machinery to obtain either the broken or worn part,
or to expose it for access. In turn, indirect expense was
caused not only by the damage done by the taking apart
of the machine, but also by the reassembling of those
parts which were in no sense injured. This indirect
charge has been the most frequent factor of dispute in
the comparison of costs, not only in the case of arc-
welding applications, but in many other lines of engi-
neering. It is not an uncommon occurrence to discover
that the work of demolition has cost as much as, or more,
than the cost of the new construction. No wonder, then,
that the electric arc, which is easily brought to the work
and there produces a localized welding temperature, was
eagerly accepted. In the metallic-arc process the actual
tools required for the operator are a screen, a wire brush,
a hammer and chisel, and the electrode holder with its
flexible lead. A bundle of electrodes near at hand com-
plete the outfit needed at the work. Back of these local
tools provision must be made for the proper electrical
conditions. In American practice this takes the form of
a motor generator for direct current and a transformer
for alternating current. This machinery end of the tool
for a single operator is not so large nor heavy that it can-
not be made semi-portable and thus provide flexibility.

In very large installations the machinery end of the
arc- welding tool may be of a size to supply many welders,
in which case the equipment may be made stationary and

GENERAL APPLICATION OF ARC WELDING 99

the electrical circuit distributed to plug boards for the
individual operators. So in any desired manner the tool
may be brought easily to the work and the broken or
used parts of the complicated machinery may be re-
paired rapidly and safely in place.

Many of the large railroad systems of this country
have employed arc welding in the vital repairs to loco-
motives, which is being extended rapidly to freight and
passenger cars. This means that millions of people are
being carried across country often at high speed, around
curves, and up and down grades with the strains and
stresses of such work resisted by a jointure carefully
made by the electric arc. The boilers and stern posts of
ocean-going vessels are repaired by the same method.
In addition to this, street railways are employing the
process for the maintenance of tracks and cars, and ma-
chine-shops for the repair of machine tools. If these
instances were chance experiments there would be cause
for questioning the further use of this process, but what
is referred to here is now established practice and has
been an accepted method for many years. In the case
of locomotive-boiler repairs the records indicate that
electric-arc repairs exceed the allowed use of the original
boiler material.

As to the cost of arc-welding repairs and a compari-
son with the older methods, the preceding remarks give
an indication which is quite true, namely, that they are
very much less. In roimd figures it is roughly estimated
to save between 50 and 60 per cent. The Chicago, Rock
Island and Pacific Railroad kept excellent detail ac-
counts when introducing electric welding.^

Railway Electrical Engineer," E. Wanamaker, 1918.

100 SPOT AND ARC WELDING

As an illustration, it cost this railroad by the old
method to repair wheel spokes $1276.80, and by electric
welding $35.08, a saving of $1241.72; for repair cracks
in tanks by the old method $372.69, by the electric arc
$36.16, a saving of $337.53; for filling worn spots by
the old method $2677.80, by the electric arc $329.60, a
saving of $2348.20. These are a few items of a large list.
In a summary for the year it is shown that the cost by
other methods would have amounted to $171,279 as
against arc welding $24,912.36, a saving for the year of
$146,366.64. There may be, and probably are, other as
significant figures, but these alone are considered proof
enough not only of the cost but of the other two factors
which are just as important, the ease of the application
and the seriousness of the work performed.

Examples. — The following applications of arc weld-
ing for repairs may give an idea of the extended use of
the process. In steel foundries it is found difficult to
avoid sand spots, blow holes, and shrinkage cracks in the
finished steel castings. It is an expensive process and
not only would cause a higher cost if defective work
were scrapped and recast, but also would involve a delay.
Such defects are readily and satisfactorily repaired with
the electric arc, either carbon or metallic. By means of
the carbon arc utilized for pre-heating, it is also possible
to employ the process for the same conditions in the
manufacture of grey iron and malleable castings. Ref-
erence has already been made to railroad-shop repairs
where established use is made of this process for the
repair of engine side rods, brake fulcrum, eccentric
crank, side frames, flues of boilers, fire box, mud ring,
engine cross head, bumper beams, brake-shoe heads, pis-

GENERAL APPLICATION OF ARC WELDING 101

ton cross heads, motion frames, yokes and spokes of
wheels, building-up flanges on wheels, etc. A list three
or four times this size could be cited. Marine repairs are
usually those connected with the boilers, the rudder post
and in some exceptional cases to small portions of the
hull plating. In street-railway work much work is done
in building up the rails and worn parts of cross-overs.
Worn down armature shafts, side frames of trucks, and
gear cases are also repaired by this process. In forge
shops the metallic arc is being used not for improving
the strength of the forging, but to give it a good appear-
ance. Small defects are apt to result from the forging
operation, especially in forgings for automobiles. These
can be neatly corrected by this process. In the large or
small machine-shop the large machine tools as well as the
small hand tools may be (q(uickly put back in service and
often result in a much longer effective life when skill-
fully arc welded. Bolt holes become worn, shafting in
motors (particularly alternating-current motors of the
induction type with small air gap) wear down in the
bearings to the point of injury to the armature windings,
bearing surfaces on slides and cams wear away in the
same manner. These and many more cases are conveni-
ently corrected by building up the surfaces with the
metallic arc and then turning the pieces in a lathe to the
original dimension. " Steel mills have found it eco-
nomical to install arc welders for the purpose of repair-
ing wobblers in the rolling mill. Work is also being done
successfully in the working surfaces of the roll."

Manufacturing. — Three characteristics of this proc-
ess brought it to the attention of manufacturers : ( 1 ) Its
secrecy, (2) its adaptability, and (3) its low cost.

102 SPOT AND ARC WELDING

As the rivet was for many years the best-known
method of joining metals either very thin or of moder-
ate thickness . where a medimn-strength joint was re-
quired, this became the established practice. As
commercial competition grew the manufactm-er sought
means to reduce his costs so that his selling price would
reward him with the contract. More than this, he de-
sired that this be brought about through some process
that for a time would remain hidden to his competitors.
The electric arc allowed this because the manufacturer
could develop all the apparatus and tools in his own
shop and design them for his own special production. It
was not necessary for him to divulge his needs to the
makers of electrical machinery nor the builders of stand-
ard arc-welding apparatus. In this manner his special
method of manufacture was concealed and his secret
safeguarded. Manufacturers of arc-welding appa-
ratus will admit that of these apphcations they have
little knowledge.

More than this, the manufacturer found that the
electric-arc process was adaptable. Whatever his
methods had been the first cost of the apparatus was of
no concern in the benefits both of reduced costs and the
elimination of competition. Besides this, the arc- welding
tool, like any new tool, gave promise of greater
possibilities. It could be safely handled for repetition
work by ordinary operators, and the future promised
some form of automatic machine. Other methods such
as gas welding were tried and found both expensive and
dangerous. By experimentation with different composi-
tions of electrodes, by varying the electric current, by
the use of coating on the electrodes, and by suitable

GENERAL APPLICATION OF ARC WELDING 103

selection of electrode size for different thicknesses and
kinds of material, the manufacturer had a tool which
needed only ingenuity to uphold his production on a
paying basis.

The savings in cost of this application are impossible
to obtain. The characteristic mentioned above precludes
anything more than assumption. The fact that this
process is employed must be the best evidence that it is
the cheapest process now available. To those acquainted
with the riveting process there is no question that adding
the cost of drawing to the finished product there is no
competition with electric-welding processes. In the case
of gas welding it is to be remembered that the arc can
weld where the gas flame cannot, and vice versa. This
last expression should probably be restricted to welding
only, as it is possible to cut steel with either the carbon
or metallic arc. From this it will be seen that to compare
gas welding with arc welding it is requisite that the work
be common to the two processes. When this is done the
consensus of opinion is that the electric arc is cheaper.
Direct comparison of costs between oxy-acetylene and
metallic arc welding have been made, but such costs are
not of great value in that the conditions are changing.
After all, in general manufacturing other elements often
take precedence in the introduction of a new process, not
solely because they have the elements of money saving
so much as money making.

Examples. — The electric arc is much used in general
boiler shops. Here are built tanks, vats, tumbling bar-
rels, wagon tanks, oil stills, and several more specialties
of a similar kind. Other shops have used the arc for
manufacturing gear cases, automobile frames, street-car

104 SPOT AND ARC WELDING

entrances, garage heaters, steel bed plates for support-
ing machinery, and a long list of minor parts of appa-
ratus for oil and sugar-making machinery. The
construction of transformer tanks with the arc has for
many years superseded other methods and the combina-
tion of hand and automatic arc welding constitutes an
advance over all other methods tried. This method was
found superior because of its reliability in service and
its economy in production.^

» " Electric Arc Welding in Tank Construction," R. E. Wagner, General
Electric Beview, December, 1918.

CHAPTER VI

Discussions on Arc Welding

Aroused by the interest in the rapid extension of arc
welding in the industries as well as impelled by patriotic
feelings, a group of men gathered to discuss the funda-
mentals of the art. Although this group was engaged
in widely-separated occupations and was sincere in
wishing to cooperate for such a good purpose, it was
discerned that the differences of opinion generally ran
in parallel lines, indicating that great latitude was per-
mitted for vohtional selection. Under such conditions
group action of a determining character was not to be
expected. It was not an established body. That is to
say, each meeting constituted itself into a new meeting
to discuss the different phases of some subject and, al-
though a few members were in regular attendance, new
members were admitted, which reacted not only to re-
open previous discussions, but also to stay the pursuit of
investigation by new suggestions. This explanation is
made so that proper values may be placed upon the fol-
lowing outline of discussions which took place in 1918.

Eoctension to Thick Steel Plates. — ^Doubt existed in
the minds of conservative experts whether uniform re-
sults could be expected in the welding of 3^ -inch steel
plates of the composition used in shipbuilding in this
country. Certain members of the group reported that
compositions of steel had been experimented with and
they had found it practically impossible to successfully

105

106 SPOT AND ARC WELDING

weld. Other more radical supporters of the process
made light of this with the remark " that you can elec-
trically weld anything." The need was clear for a de-
termination of the fact that ship's steel of a thickness of
Yi. inch could be successfully welded. With this funda-
mental question was coordinated the query whether firms,
who performed welding, could turn out equal results
despite the different methods pursued. A sub-committee
was appointed to follow the practical details which con-
sisted in obtaining ship's steel plates ^ inch thick, have
them prepared for arc welding (double V, butt joint),
and make observations of the welding. These firms were
given entire freedom in the methods, materials, oper-
ators, electrodes, current, etc. In some cases two
samples were made, one with a reinforced weld and the
other with the reinforcement machined down. This col-
lection of samples was sent to the Bureau of Standards
for physical test — ^tensile, torsion, vibration, and bend-
ing. The results confirmed the opinion that successful
welds could be made in this material. Many of the
samples exceeded the yield point of the original material
as well as ultimate strength. In a number of cases the
elongation in two inches expressed in per cent, ap-
proached fifty per cent, of the original material, and one
test piece was made with alternating current at 25 cycles
bent in the weld to an angle of 78 degrees. The rein-
forced welds all showed a higher ultimate tensile than
the machined-down test pieces. The yield point in
poimds per square inch of the original plate was 38,400.
Some of the reinforced samples gave yield points of
42,460, 40,280, 40,480, 42,200, 39,000, 46,440, and
44,700 pounds. Five machined-down samples showed a

DISCUSSIONS ON ARC WELDING 107

yield point of 39,000, 39,000, 38,400. 42,400, and 41,800
pounds. The original plate tested to an ultimate tensile
of 64,700 pounds per square inch. Reinforced welds
showed results as follows: 65,470, 65,400, 66,480, 66,400.
None of the machined samples equalled nor exceeded
the ultimate tensile of the original plate, but three
samples reached 62,600, 62,800, 62,700. These tests

Fn. tS. — Cbt coupkr iB«d m itSmiy woik bdote uc wdding.

were looked upon as practical, formed the base line for
further argument, and suggestions were made both for
improvement in testing and for extension of results.
The outcome was the preparation of a more elaborate
program of tests with like materials, but with more uni-
form welding conditions and the elimination of some of
the many variables. It had been diflicult to carry out
the foregoing tests without delay. The more elaborate
tests were by their nature subject to greater postpone-
ment, and as a consequence were halted before their

108 SPOT AND ARC WELDING

entire completion. Broadly viewed it is questionable
whether such investigations are of great value in that the
arc- welding operator is a serious link in the chain. All
the tests of manual welding made by others would prove
little to the man who wished to use the process.

Composition of Electrode. — The material used for
electrode wire had long been the subject of investigation
by users of this process. Not only chemical analyses of
the electrode wire but also similar analyses of the de-
posited metal in the weld were made. Side by side with
the chemical inquiry ran the metallurgical. It was true
that certain alloyed steels gave better results with cer-
tain compositions of steel plate. In fact, it was sug-
gested that for the best interests of shipbuilding a differ-
ent composition of steel might be required on the basis
of its weldability. This suggestion would be difficult in
view of the commercial conditions which likewise dic-
tated the composition of the electrode materials. Doubt-
less an increased demand for electrode wire for special
applications might ease the situation, but under the con-
ditions the user must accept what was on the market.
That certain desirable results in welding could not be
attained with the electrode composition as furnished is
true, and one manufacturer was unable during this period
to duplicate results made elsewhere. There resulted
from this discussion a practical specification for elec-
trode wire. This instrument was issued as a guide to
shipbuilders who wished to purchase such material. The
chemical composition was such as to include all of the
manufacturers, and the test requirements were made to
invite the manufacturers' attention to the need of a dem-

DISCUSSIONS ON ARC WELDING 109

onstration of his product before the completion of the
sale. The chemical composition is as follows :

Carbon Not over 0.18

Manganese Not over 0.55

Phosphorus Not over .05

Sulphur Not over .05

Silicon Not over .08

Design of Weld. — This was the subject of special
debate because of its bearing upon the design of an all-
welded ship. The type of joint, the design of weld, the
position of weld, kind and type of weld, all required in-
vestigation before the naval architect could make draw-
ings for the ship. Early in the proceedings the strap
joint was considered to be 100 per cent, efficient and re-
mained so until word was received from England that
the butt joint was better. The strap joint was then
questioned as to the order in which the three seams of
welding should be performed which led the discussions
into metallurgical theories. Not quite the same fate was
reserved for the angle of bevel. This was a subject upon
which many differed. Evidently no practical action was
taken to settle this argument, so finally it was left to the
choice of the designer. As the amount of deposited
metal from the electrode is relative to the size of the
bevel, and as the goodness of the weld depends on the
ease given the operator to fuse the original metal with
the deposited metal, the biting-in effect, this point bears
no small ratio to the final results. In the use of covered
electrodes practice might permit a different angle of
bevel or with special electrode, where the cost was ex-
cessive, a reduction of the needed deposit of electrode
wire would greatly affect the final cost.

110 SPOT AND ABC WELDING

Associated with this question was the number of
layers of deposited metal. Though the generally ac-
cepted position was that for j4-inch steel plates, one
layer was not the proper method for securing the best
tensile strength of weld, yet the tests showed welds made
in one run which gave results as high as 62,600 pounds

Fid. M. — Cu couplet lued in nilwBT woA atUr tit welding mth ouUd dectrode.

per square inch ultimate tensile. There were very few
welds made in two layers which exceeded this figure.
This was an important point, because it directly affected
the speed of welding, one of the main factors of its
economy. Before a second layer of deposited metal can
be run in, it is necessary to clean thoroughly the top sur-
face of the first layer. If much slag has been brought
up to the top of the weld, which must be done by the
welder, it requires a chisel and hammer to fully clean
this surface. In the case of slag-producing electrodes,
this deposit must also be scrupulously removed before

DISCUSSIONS ON ARC WELDING 111

beginning the second run of metal. It is claimed, and
the claim has elements of justification, that the second
layer in its act of deposition partly anneals the first
layer ; but curiously it is found, except with special proc-
esses, that annealing does not greatly improve the quali-
ties of the weld. If there is no virtue in adding layer
upon layer of deposited metal, and if one layer will pro-
duce a reliable and satisfactory weld, time and labor
would be wasted. This question has been left to
the designer.

In a like category were the discussions on the posi-
tions of the weld, i.e., flat, horizontal, vertical, and over-
head. The extremists held that overhead welding should
be done only by specially trained men. More than this,
that the ordinary man should not be trained to do over-
head welding. The intention of the extremists was that
in order to train operatives quickly it was a waste of
time to expect them with a brief training to make suc-
cessful welds in the overhead position. Experience in
the training schools for welders showed advantages for
overhead welding as a method of practice in that the
student was more confident in all the other positions
after having mastered the difficult overhead conditions.
At the other extreme were those who, having experience
with handling the electrode, asserted that the position
was not as distressing to the operator nor as detrimental
to a good weld as generally considered. The other posi-
tions, though not receiving the same prominence in the
argument, were not as easy for the operator as supposed.
The flat position is the most comfortable and conveni-
ent, although welders may be found who prefer the
vertical position. The horizontal position is most awk-

11« SPOT AND ABC WELDING

ward and in difficult places requires the operator to
be ambidextrous.

From the tests above cited attention was called to
the difference in ultimate tensile strength in the rein-
forced weld and one that had been machined. No com-
parisons were made with a flush weld, i.e., one made flush
by the operator. The tests show quite convincingly the
opinion held that the reinforcement of the weld added
strength to the joint. With this point estahhshed it is

necessary to go one step farther and determine the
amoimt of reinforcement requisite for a certain strength
of joint or for a particular application. In this case the
particular application was shipbuilding, and some limit
either maximum or minimum of reinforcement was
essential. In this as in the case of the angle of bevel,
the question is of importance in that it means consump-
tion of the electrode and the time of making the weld.
Covered Versus Bare Electrodes. — The bare-metal
electrode process was introduced about 1895 by a Rus-
sian named Slavianoff. The covered-electrode process

DISCUSSIONS ON ARC WELDING 113

bears the trade name of Quasi-Arc and is the invention
of Mr. Arthur Strohmenger, of London. Both systems
have been already described. The Slavianoff system has
been used in this country for many years and it is stated
by one authority " that he is not aware of any user " in
England.^ The Quasi-Arc, or covered-electrode sys-
tem, was only recently introduced to American practice.
It was natural that those who were familiar with the
working capabilities of the bare electrode should insist
upon its equal performance to the covered electrode.
When a physical test of a covered-electrode weld showed
superior qualities, naturally advocates of bare-electrode
systems hastened to exhibit welds that would equal
or surpass the new competitor. To the full appreciation
of the discussion must be brought the commercial atti-
tude because this affects directly one of the principal
technical points. The customary practice for direct cur-
rent was to provide a " striking voltage " of 60 to 75
volts. Upon this practice, standard apparatus in this
country was designed and built. This " striking volt-
age " corresponds approximately to an arc voltage rang-
ing from 15 to 25 volts. The covered electrodes required
a " striking voltage " of at least 100 volts, and prefer-
ably a little over, giving an approximate arc voltage of
35. Very few manufacturers of arc-welding apparatus
allowed for any possible adjustment of the voltage of
the welding generator. This condition reacted severely
on the rapid introduction of the covered electrode. De-
spite this condition test results both from England and
in this country indicated very clearly that for alternat-

1 ((

Electric Welding," Thomas T. Heaton, The Journal of the Inst, of
Mechanical Engineers, London, February, 1919-

114 SPOT AND ARC WELDING

ing stresses and ductility there was a superiority in the
use of a covered electrode. The experiments of British
Lloyd's in electric welding were all made with this typie
of electrode, and the process was approved by this classi-
fication society. The characteristics of resistance to
shock, a reasonable ability to withstand fatigue, an in-
creased bending angle are important considerations in
the applicability of a welding process to ship construc-
tion. The claims for the covered electrode were based
on the fact that the covering provided a slag which pro-
tected both the electrode and the deposited metal from
oxidation. The result was that in the hands of a skilled
operator there was less porosity and a more ductile weld,
retaining at the same time good tensile strength.

The reported results soon turned the attention of
investigators to the advantage of some form of protec-
tion to the electrode. Experiments easily performed
showed that electrode wire that was not smooth running
or would not produce good welds, if heat treated, dipped
in acid or alkaline solution, would become better or
worse, depending upon the methods used. These trials
were'in line with Kjellberg's invention which provided
an electrode coated with a fusible silica. This coating
formed a flux which was converted into a gas by the
heat of the arc and therefore left no slag as is the case
with the Quasi-Arc covered electrode. Coating of
the electrode now came into style and the results of
tested welds above referred to showed one sample in
which half the electrode was coated with a special solu-
tion giving unusual bending characteristics. The covered-
electrode sample, though not giving results as high
as this particular sample, were next to it and exceeded

DISCUSSIONS ON ARC WELDING 115

all the others. The decision of British Lloyd's in ap-
proving this process of covered electrodes, although it
threw great weight in its favor, has not caused the ad-
herents of the bare-metal electrode to relinquish
their position.

Although it may be possible to successfully weld mild
steel for ordinary purposes with the bare-steel electrode
and thus avoid the expense of covered electrode, where
toughness is a desirable or necessary characteristic of
the weld, or for the welding of steel alloys, a special coat-

F^. M^ — Baliter uard m nilwof worlc aha mK wdding wibh anted dHtrodb

ing will give better and in many cases the only successful
results. One of the foremost electric-welding engineers
in this country has lately experimented with and has now
achieved much success with coated electrodes. He states :
" Regarding the chrome steel would advise that we have
received and successfully welded with chrome steel,
nickel, vanadium, manganese, and carbon. Have also
welded with bronze when using our electrode coating.
We have received patents on this process but have not
as yet placed any coated electrodes on the market, largely
due to the fact that it was diflSciJt to obtain alloy elec-
trodes during the war (Fig. 26).

116 SPOT AND ARC WELDING

" Recently we ran some tests on four bare mild steel
electrodes, each made by a different manufacturer.
When used bare it is impossible to secure the weld, but
with our coating the weld is very successful. Outside of
the possibility of using alloyed steel, and all results at-
tendant with the use of same, our chief aim has been to
make a weld in which the added metal would be compara-
tively free from oxidation. This would give us a weld
possessing much greater toughness than that possessed by
the bare-steel electrodes. It is needless to say that this
toughness imparts a quality very much to be desired, and
has a very important bearing on the success of electric
welds to withstand fatigue." ^

Direct Current Versus Alternating Current.— Apart
from the fact that many engineers believed that arc welds
could not be made with alternating current, their argu-
ments attacked the application of alternating current
from three sides : (1) Its newness, (2) its wasted energy,
(3) its difficulty of operation. Direct current had long
held the field and hence the apparatus and tools were
familiar to both engineers and operators. The design
of transformer^ imlike power transformers, required a
large leakage reactance in order to stabilize the arc. Due
to the character of alternating current it was impossible
to weld without holding a short arc. This is a requisite
for successful welds with direct current, but in this latter
case the apparatus gave a certain tolerance to the oper-
ator, whereas with alternating current the arc position
was dependent and must be held by the operator. This
required more practice, greater skill, and more fatigue

* Communicated to the Author by Mr. E. Wanamaker, E. E., Chicago,
Rock Island & Pacific Railway, September, 1919.

DISCUSSIONS ON ARC WELDING 117

for the operator. If the operator repeatedly lost his
arc, the weld would be porous and filled with blow holes,
and the opponents of alternating-current welding
claimed that by the nature of things this would be true.

Of this latter point the advocates of alternating-
current welding made much. They insisted that with
it you could only weld and never " not weld," as in the
case of direct current. They held that the process showed
distinctly a " biting-in " eflfect, i.e.y better fusion of the
original metal with the deposited metal. As to the oper-
ator, it was not a difficult matter to train those who
handled the direct-current arc and that a number of old
operators claimed a preference for alternating current.
As to the second point of the argument, they believed
that upon development the low-power factor, or poor
efficiency, would be greatly- improved and in some in-
stances claimed that the power factor was not so low as
to put the process out of the running as compared with
other systems. To this end they made tests to prove that
good welds could be made with any frequency and at the
lower frequencies the power factor would be better. In
answer to point three, while admitting the " newness,"
they claimed equality as far as the application to new
construction work was concerned, and then endeavored
to show the advantages from the standpoint of economy
in first cost, continuous operation, and upkeep. As it
was only necessary to have a transformer connected to
the electrical-supply leads this obviated the necessity for
rotating machinery and gave the apparatus a more
practical portability.

As in the case of the other discussions these interest-
ing points were never clearly determined. No compara-

118 SPOT AND ARC WELDING

tive data is at hand to indicate whether direct-current
welding is faster or slower than alternating current;
whether this practical advantage is benefited in either
case by flux-covered or coated electrodes ; whether it is
more difiicult, or impossible, for the ordinary operator
to weld in all positions, including overhead, in the one
system than the other; whether the practical losses in
motor generator and resistance in the direct-current sys-
tem, other things being equal, compares favorably or un-
favorably with the alternating-current system and
whether there are physical obstructions to the training of
operators that would make the alternating-current sys-
tem improbable of industrial acceptance without assur-
ance that the system gave promise of being capable of
improvements which would overcome such impediments.
Individual investigators may have settled all of these
questions to their own satisfaction, but the general prac-
titioner looks in vain for independent authority. Many
more claims than are cited here are made by both parties
to this controversy, but they lead into the field of theories.
Testing of Welds. — ^Perhaps no subject received as
much consideration both by suggestion and experimenta-
tion than the discovery of some practical method of test-
ing a welded joint. Though the question was emphasized
by practical men who wished a practical method, all the
exertions were toward theoretical or laboratory methods.
Exhaustive tests were made with delicate apparatus
upon samples containing purposely poor and good weld-
ing and by maintaining certain characteristics constant.
Methods were suggested, such as measuring the mag-
netic permeability, the change in hysteresis, drop in volt-
age, resistance. X-ray photographs, etc. The practical

DISCUSSIONS ON ARC WELDING 119

methods were by hammering the welds or chipping out
small portions for examination or by wetting a cleaned
portion with kerosene. This latter test, due to the pene-
trating qualities of kerosene, would indicate porosity.
The theoretical desire was to determine positively the
physical characteristics of the weld: the practical desire
was to convince or assure those who inspected the work
that the weld was sound. A little more was needed. The
men who serve as inspectors are responsible to those
above them in authority, and it is necessary that a prac-
tical method be established so that individual responsi-
bility, or opinion as to workmanship, is not relied upon
for serious applications. No practical method of testing
long seams, such as those that would be encountered in
ship construction, have as far as is known been devised.
To fill all the compartments of a merchant vessel for the
purpose of testing joints would be out of the question,
although for bulk-oil vessels this is now done. The only
practical suggestion advanced is that the inspectors
should be trained just as they were trained to inspect
riveting work. If a man knows how to hold an electrode
and can make a sound weld, no other man could deceive
him either as to his ability as a welder, or the quality of
work that he was performing. Much can be said of the
comparative merits of the practical methods of testing
rivets. But it is not necessary to extend the argument
because there will be a method forthcoming as soon as
electric welding is established practice.

Current and Electrode. — ^Although the manufac-
turers of apparatus give tables showing the current and
electrode diameter for varying thicknesses of mild steel,
they accompany such information with words of precau-

120 SPOT AND ARC WELDING

tion to the operator that such figures are only approxi-
mations. The electrode size is related to the amount of
current and the class of work. The amount of current
is not necessarily relative to the thickness of the plate,
although this is a good practical guide for mild-steel
plates under ^ inch. The design of joint has some bear-
ing upon the current requirements, as, for example, the
lap-joint, which undoubtedly will be better made with a
gt-eatly increased current over that necessary for the
simple double-bevel butt-joints.

In the tests of sample welds it was noticed that some
of the best results were secured with increased current,
and this observation aroused much interest. Investiga-
tions by individuals showed that increased current,
through a range of 80 to 275 amperes with all other con-
ditions constant, improved the tensile strength and duc-
tility. The next question was where the effects of
increased current terminated. It is not known whether
succeeding experiments have determined this point.

The largest-size electrode suggested in practice is
3/16 inch in diameter. A 5/32-inch electrode being a
popular size for currents ranging from approximately
100 volts to 190 volts, and used for mild-steel plates of
from % inch to ^ inch in thickness. In heavy welding
the metal is deposited in layers, sometimes two or three,
left to the option of the designer. It is claimed that each
succeeding layer anneals the one beneath it and if a rein-
forced layer is placed on top it will complete the anneal-
ing process and may without affecting the joint be
machined down. This method appealed to many engi-
neers as a long and tedious process, and the question
arose whether a weld could not be made with larger-

DISCUSSIONS ON ARC WELDING 121

diameter electrodes and accomplish the work in one rmi.
In addition to the explanation of the annealing effects
of the layer method it is believed that with an electrode,
say of ^-inch diameter, that the increased cm*rent would
be so great that it becomes a cutting current, and con-
trol of the arc not within the skill of the operator. That
is to say, that the arc characteristic would be such that
smooth movement at short-arc length would not be pos-
sible, and smooth running of the electrode is an essential
of good welding.

Rigid Versus Non-rigid Assembly. — The question
of the best method of preparing long plates for seam
welding fell into two groups : those whose practice war-
ranted the belief that welded seams could be made when
the two plates were rigidly connected either mechani-
cally or by means of widely-spaced tack welds, and those
whose practice, though not denying the possibility of
rigid assembly, warranted the belief that, by giving the
plates room for expansion and contraction, the cooling
stresses (locked-in) would be greatly reduced and that
more uniform success would result. The non-rigid sys-
tem provides for a tapered separation between the
plates, the welding beginning at the small end. Clamps
are inserted between the plates and hold the proper dis-
tance. The operator upon releasing them observes the
rapidity or slowness of the expansion and acts in accord-
ance therewith, i.e., if the opening closes quickly he
hastens his welding, or if it is slow in closing he waits.
The effecti^ of expansion and contraction are observed
and cared for in the rigid system in much the same way
with this difference, that usually the seam is not made
continuously but in sections which permit of a distribu-

12? SPOT AND ARC WELDING

tion and equalization of the cooling stresses. The marked
effect of the discussion was its relation to ship construc-
tion, for it is not conceivable how the non-rigid system
could be applied.

Both for practical evidence of the ability of arc-
welded seams in ^-inch steel plate to withstand shock
and fatigue comparable to those met in ship design, as
well as to put the non-rigid system to test, a 12-foot
tank ^ of J^-inch tank steel was built and tested. Inci-
dentally complete records were kept of the cost, time,
metal deposited, quality of electrode, etc. The designs
of joints were patterned after those already suggested
for an all-welded ship and included individual designs
of which there was doubt. After the tank was finished
it was filled with water and alternately subjected to 15-
poimds pressure and 22-inches vacuum. The designer
and builder reports that " after the first 12 cycles had
been completed, a break occurred at one end of the box ;
the break was confined principally to the solid-end and
bottom plate. . . . After 42 cycles the end patch
began to leak and had to be welded along one edge.
. . . After repairs had been made the breathing test
was continued and has now been carried to 200 cycles.
There is now a slight break on the patch and one in the
centre of the bottom seam."

It was suggested that a riveted tank similar to this
12-foot box be built and tested in the same way, but ship-
builders advised that their experience with riveted tanks
showed that such an undertaking was a waste of time
and money in that no riveted tank could be kept water-

» " Electric Arc Welding in Tank Construction," R. E. Wagner, General
Electric Review, December, 1918.

DISCUSSIONS ON ARC WELDING 123

tight nor stand the abuse given the all- welded tank. A
tank of the same dimensions and materials has been
assembled and built on the rigid system. This tank has
not yet been tested.

Ductility Versus Strength. — Closely connected with
the discussion on covered and bare electrodes was that
of the results from these instrumentalities. A good
strength weld was possible in mild steel with the bare
electrode, but the flux-covered or even thinly-coated
electrodes could produce a weld with very much greater
ductility. It was claimed broadly that strength welds
were not all that was desired in a ship joint. The ship
subject at all time to the forces of waves and wind,
affected by continual vibration of her own propelling
power and fatigued by the creeping action of many com-
plicated moments of forces — all these must be insured by
joints that would bend and not break; that would strain
and not leak;. that would creep and not snap away; and,
in short, would act in all respects like rubber. Upon
question it was admitted that riveted ships hardly ap-
proached this desideratum and that in practice the riveted
joint was comparatively a rigid joint. The suggestion
then followed that as ductile welds were expensive for
electrodes, time, etc., and not always to be assured,
strength welds of 85 to 90 per cent, of the plate, or
greater than the plate strength, be designed and em-
ployed. To the practitioner this was reasonable and
permitted the work of shipbuilding to proceed; but to
the theorists such an argument was alarming, and they
advised that the whole subject of the application be re-
turned to the laboratory for further investigation.

Cast Iron. — From time to time claims are made that

124 SPOT AND ARC WELDING

cast iron can be welded to cast iron, or cast iron to cast
steel. Like many arguments, deductions were made
from misleading, if not entirely wrong, premises and
often the parties to the controversy were both right. In
a large way it was heralded that the engine cylinders of
the interned German ships which were of cast u-on had
been electrically welded. The correct statement was
that they had been admirably repaired by the instru-
mentality of the metallic arc in combination with me-
chanical skill. There is no doubt that the method
employed was superior both in point of economy and
excellence of result. Briefly, the method was to make a
cast-steel patch to fit the broken part. A series of studs
were tapped into the cast-iron cylinder. From these
studs metal was deposited from the electrode which was
carefully played about the seam locally so as not to cause
dangerous over-heating. When the V of the seam was
filled the deposited metal was continued until it covered
a broad band. The finished weld was then hammered
with the intention of improving the quality of the weld
as well as stopping leaks. Samples of this work were
examined both for physical tests and for fusion of the
metals. The weld, as expected, was always stronger
than the cast iron and invariably the samples broke at
the joint. A small piece of deposited metal on the cast
iron could be easily broken off. This carried with it some
of the cast iron, but also indicated a brittle structure at
the jointure. It would seem that the cast iron was weak-
ened by the reactions which take place in the heat of the
arc and the chemical changes caused by the constituents
of the electrode material. From such large work as
engine cylinders and the method used for their repair, no

DISCUSSIONS ON ARC WELDING 125

reasonable deduction can be made that small pieces of
cast iron can be arc welded. Fundamentally, cast iron
is a cheap material with a rating in this country of ap-
proximately 17,000 to 18,000 pounds per square inch
tensile strength. Though some engineers state that cast
iron may be welded as well by the arc as by any other
method, they are quick to restrict this statement by the
clause, " but the results are always uncertain."

Automatic Arc Welding. — In line with the prophe-
cies that manual arc welding would be superseded by
some form of machine, one of this group of specialists
devoted himself to the practical solution of this problem.
The methods he pursued and the present results which
he has attained not only tell a story of achievement but
alsQ reflect much light of importance to arc-welding
operators. Here is his own description of his first
assumptions and how they developed :

" Early in his investigations, the writer * concluded
that a substantial equilibrium must be maintained be-
tween the fusing energy of the arc and the feeding rate
of the welding strip; and it soon became evident that if
the welding strip is mechanically fed forward at a uni-
form rate equal to the average rate of consumption with
the selected arc energy, this equilibrium is actually main-
tained by the arc itself, which seems to have, within cer-
tain circumscribed limits, a compensatory action as
follows : When the arc shortens, the resistance decreases
and the current rises. This rise in current causes the
welding strip to fuse more rapidly than it is fed, thereby
causing the arc to lengthen. Conversely, when the arc
lengthens, the resistance increases, the current falls, the

* Harry D. Morton, Secretary-Treasurer, Automatic Arc Welding Co.

126 SPOT AND ARC WELDING

welding strip is fused more slowly than it is fed, and the
moving strip restores the arc to its normal length. . . .
While this compensatory action of the arc will maintain
the necessary equilibrium between the fusing energy and
the feeding rate under very carefully-adjusted condi-
tions, this takes place only within relatively narrow
limits. It was very apparent that, due to variations in
the contour of the work, and perhaps, to differences in
the fusibility or conductivity of the welding strips or of
the work, the range of this self -compensatory action of
the arc was frequently insufficient to prevent either con-
tacting of the welding strip with the work or a rupture
of the arc due to its becoming too long. The problem
that arose was to devise means whereby the natural self-
compensatory action of the arc could be so greatly accent-
uated as to preclude, within wide limits, the occurrence
of marked arc abnormalities. There was ultimately
evolved, by experiment, such a relation between the fus-
ing energy of the arc and the feeding rate of the welding
strip as to give the desired arc length under normal con-
ditions; and tendencies towards abnormalities in arc
conditions, no matter how produced, were caused to bring
into operation compensatory means for automatically,
progressively, and correctively varying this relation be-
tween fusing energy and feeding rate, such compen-
satory means being under the control of a dominant
characteristic of the arc. In their ultimate forms, the
devices for affecting the control of the arc are simple
and entirely positive in aiction, making discrepancies
between fusing energy and feeding rate self -compen-
satory throughout widely-varying welding conditions." ^

'^ Journal of Amer, Imt, Mining Engineers, p. 818, 1919. Discussion by-
Harry D. Morton,

DISCUSSIONS ON ARC WELDING 127

The inventor of these machines has made many ex-
periments to illustrate the " compensatory action of the
control," by using varying compositions of electrode
material and work material as well as varying the volt-

Fio. «T.— Mortgo snni-llutoiiKilJc nnt«llic-ebclTOde «re-"rlding muhine.

age supply and changing the contour of the work. He
has developed two types of machines which he designates
automatic and semi-automatic. The latter appears to be
a practical shipyard tool, resembling a portable drill
(Fig. 27).

Many interesting points have been observed in the
operation of these toolsi Those o^a practical nature are:

128 SPOT AND ARC WELDING

(1) The importance of the angle of inclination of the
electrode to the work. " An annular variation of 5 de-
grees will sometimes determine the difference between
success and failure. . . . About 15 degrees from the
perpendicular works well in many cases. In welding
some materials the electrode should drag, that is, point
toward the part already welded rather than toward the
unwelded parts of the seam." (2) The affinity between
electrode materials and work materials. " Generally
speaking, the Swedish and Norway iron wires seem to
produce more quiet arcs and, possibly, a more uniform
deposition of electrode material than do other wires.
. . . To date, no steel has been tested on which appar-
ently satisfactory welds could not be made. High-speed
tungsten steel has been successfully welded to cold-
rolled shafting, using Bessemer wire as electrode mate-
rial. Ordinary steels varying in carbon content from
perhaps 0.10 to 0.55 per cent, have been welded with
entire success." (3) The electrical-supply variations.
" So far, electrode wires % inch in diameter have been
chiefly used in the machines* Successful welds have
been made with current values ranging from below 90
to above 200 amperes at impressed voltages of 40, 45,
50j 55, 60, 65, and 80. Under these varying conditions,
the voltage across the arc has been roughly from 16 to
22. The machines have thus far been run only on
direct current." (4) The short arc. " While undoubt-
edly it is difficult, if not impossible, to maintain in manual
welding an arc shorter than this (0.1 inch), the writer
has frequently, with the automatic machines, made con-
tinuous and strikingly good welds with arcs of much
less length." (5) Rate of doing work. " With the auto-

DISCUSSIONS ON ARC WELDING 189

matic machine, black drawing steel 0.109 inches thick has
been welded at the rate of 22 inches per minute. A De-
troit manufacturer welded manually with oxy-acetylene
at the rate of four per hour a large number of mine floats
10 inches in diameter, made of this material. The auto-

matic machines made the welds at the rate of forty per
hour. . . . The productive capacity of the machines
so far made has been from three to ten times that of
manual weldipg methods." (6) Type of electrode.
" Bare wire only has been used in the automatic ma-
chines, and the results obtained seemed to indicate that
the covering of the electrodes is an expensive superflu-
ity." (7) Cleanliness of materials. " The writer has

180 SPOT AND ARC WELDING

repeatedly welded with wire showing evidence of pipes
and seams, as well as with rusty wire and with wire
covered with dirt and grease. In this connection it may
be said that no pains are ever taken to remove rust, scale
or slag from the work material — even where welds are
superimposed. Apparently under uniform conditions
of work traverse, arc length and electrode angle of in-
chnation, such as are possible in the automatic machine.

Pra. «».— Two H" iliip pl»l« sulomsliiHUy arc welded.

impurities vanish before the portion of the work on which
they occur reaches the welding area of the arc."

Many of these practical observations of automatic
are welding will change former opinion, but their great-
est good will result in the attention given them by manual
arc welders. In experimenting with guch machines it
has been discovered that great differences in the welding
results come about from the location of the ground con-
nection in relation to the location of the arc. Doubtless
this is a phenomenon of magnetism or conductivity. It
is well that the arc-welding operator be acquainted with
such an observation, although in large work, long seams.

DISCUSSIONS ON ARC WELDING 131

or heavy materials this phenomenon may not seriously
affect the goodness of the weld. Besides these practical
observations the inventor of automatic arc-welding ma-
chines takes note of the theoretical side of the " con-
trolled " arc which will be considered in connection with
the theories of electric welding.

The Training of Operators. — Unanimity of opinion
places the arc- welding operator as the chief factor of the
making of a sound and perfect weld. His participation
in the process has been estimated at 80 to 90 per cent.
Early in the debate the statement was made that the
apparatus, the electrode, or the work materials had little
to do with a successful weld when compared with the
man who makes it. The operator who could not make a
weld despite the opposition of these elements was, at
least, not a skilled welder. This does not mean that
there were not combinations that could for a time
baffle his skill, but it does mean that the skilled welder
if not interfered with could produce excellent work
without specialized apparatus or with elaborately
prepared electrodes.

This was not the essential question. The timidity of
conservative advocates of arc welding was occasioned by
this very high percentage of operator and the inconsist-
ent results of his work. In other words, the conserva-
tives were not willing to risk their reputation because of
the non-uniformity and instability of this personal equa-
tion. It is nonsense to say that the operator could not
hide poor work, for this he could do and more — he could
place good and bad work side by side. For this reason
the first requirement claimed for the operator was that
he be conscientious.

On the other hand, the radicals adhered to their prac-

132 SPOT AND ABC WELDING

tical view that men who had some knowledge of steel
jointm-e either in blacksmith shops or boiler shops could
be made good welders. The work of such had been in
use for many years for much serious work, and this was
a sufficient guarantee of the results. They pointed to
other long-tried methods and asked why demands were
not made to destroy such work, as it too only received
exterior inspection. For example, the large varied use
of cast iron for many purposes connected with danger
to life, so steel castings and forgings: who knew what
was inside of these finished articles ? It was not that they
did not agree with the conservatives that for special
work a competent man should be employed, but that
there was too much stress laid on the importance of the
operator for a large run of work.

In the eagerness of the desired application it was
natural that the time of training welders should be given
most attention. There were engineers who did not hesi-
tate to state that arc welders could be trained in a few
days. A little experience on the part of any one with
the handling of the electrode would assure that this time
was too short. Any one with a steady hand and a com-
fortable adjustment of the electrical supply may make
the first time a deposit of metal from the electrode ; but
he makes a great error if he leads himself to believe that
this is all that is required to produce uniformly good
welds. As previously noted there is a great difference
in the training of a man for one single operation and
making the same man capable of applying his knowledge
and training to a large variety of work.

Further details of the training of operators will be
treated in the next chapter.

Summary. — The endeavor of this chapter has been.

DISCUSSIONS ON ARC WELDING 133

to rehearse briefly the salient discussions on are welding
and its bearing on the application to the industries.
Questions here alluded to have been selected for their
interest to the general practitioner and to illustrate the
many-sided nature of the opinions. The following list
gives an idea of the investigations suggested :

The supply and distribution of electric power.

Vanadium coating on electrodes.

Titanium coating on electrodes.

Boron sub-oxide coating on electrodes.

Magnesium coating on electrodes.

Titanium-core electrodes.

Boron-core electrodes.

Charcoal-core electrodes.

Aluminum-core electrodes.

The effect of the height of deposited metal in weld.

The value of extending the projection of the de-
posited metal in the weld.

Determination of the separation distance between the
metals to be joined.

The proper size of lap in a lap-joint.

The proper width of strap in a butt-strap design.

The proper thickness of strap for varying thicknesses
of plate.

The advantages of change of current for different
layers of deposited metal.

The effects of increased current on different sizes
of electrode.

The limits of layers, i.e., how many for varying thick-
nesses of steel plating.

Effects of the elements, i.e., rain, wind, snow, etc.

Eye protection for the operator.

CHAPTER VII

The Arc Welder

If successful arc welds are dependent upon the man
behind the electrode it is fair to assert that those who
wish the best of this process should consider him care-
fully. Experience in the training of miscellaneous men
has shown that no amount of training will make a man a
welder. Some men are so constituted, or have been so
molded by other occupations, that they cannot hope to
acquire skill iti the handling of the electrode. Electric
welding is in this respect no different from any of the
other vocations and has its degrees expressed by the
competency of the man. This comparative scale does
not mean exclusion, but for special applications it signi-
fies the importance of selection. To the employer
undoubtedly all degrees of proficiency will have their
field of usefulness and the lower degrees will work up-
ward through the fostering of ambition. This is the
keynote of selection. An experience with metals, a prac-
tical knowledge of the effects of heat treatment, even a
few years' use of the electrode in repair work — all these
may aid the arc welder in seeking a job, but they will
not assure the employer that, even with months of train-
ing, the man will be a welder upon whom he may rest
the responsibility of a new application. On the other
hand, a man who knows nothing of these things but who
displays a spirit of conquest and a willingness to accept
failure that he may gain success, this fellow will make
not " just a welder," but a skilled craftsman.

134

THE ARC WELDER 135

Traimng. — Forced methods are not conducive to the
best results. Some men can right themselves under the
confusion of haste, but others cannot get their bearings.
The object of the Emergency Fleet Corporation was to
aid the shipbuilders in hastening the construction of
ships. The procedure adopted was a notification to the
shipbuilders that training would be furnished to such
men as they cared to select with the option of welding
instruction, and in addition, an intensive course in prac-
ticalmethods of imparting knowledge. This latter gave
the shipbuilder the opportunity of setting up his own
training school after a few men were equipped. Upon
the introduction of this system a restricted bonus was
offered to those who immediately accepted the proposal.
This arrangement reacted advantageously to the train-
ing, as it permitted the retention of the men in the
schools until they were proficient with the electrode. By
making the training exclusively for the shipbuilding in-
dustry another constant was established, namely, the
Svol*k material. In this respect the training of arc
welders was a specialty. Although a broad view was
taken of the benefits of a liberal education, this feeling
was held in check by the demand for men who could be
trusted to weld certain parts of the ship without endan-
gering the structure or discounting the advantages of
the process.

Another constant could have been instituted but
opinions pointed to the necessity of future development.
This refers to the question of the kind of electrode.
Guided by American practice the bare mild steel elec-
trode could have been made the standard and the men
trained only with this type ; but persuaded by argument

136 SPOT AND ARC WELDING

and the examples of what had been done in England
the men were also trained with the covered electrode.
That in practice this policy did not warr^it the exertions
made, nothing can be said, as the application has not yet
been put to the full test of shipbuilding in this country.
It cannot be known whether the jSrst welded ship will be
built with either one or the other type of electrodes, but
very likely both would be used.

Endeavors were made to provide at every school
various types of apparatus so that the student would be
familiar with the special features and characteristics of
operation of different machines. This was done for two
reasons : ( 1 ) That no question of commercial preference
could be raised, and (2) that the men would be prepared
to operate any apparatus in the event that their company
wished to experiment with a number of types. Although
it has been stated that arc welders did not need to know
anything about electricity, the intention being the science
of electricity, it is of importance for him to be able not
only to start his motor generator and adjust the current
for welding, but also to understand enou^ of the work-
ings of the machinery to exercise discrimination in the
case of derangement of the apparatus. In some of the
schools the equipment was not so diversified as it should
have been, with the result that a number of men, although,
well equipped as welders, did not obtain a proportionate
knowledge of arc-welding machinery. This is a handi-
cap for good welding, as many of the manufacturers do
not agree in their fundamentals of design with this result
to the welding operator that frequently he wUl make
adjustments on a machine that will not produce the
desired results because of his lack of familiarity with

THE ARC WELDER 137

this particular apparatus. The Emergency Fleet Cor-
poration was able to give this additional training on
many different machines through the courtesy of various
manufacturers throughout the country.

For the purposes of standardization a list of symbols
was proposed and approved. This list has been adopted
as a standard by all users of the process throughout this
country. It led, as will be seen from the last page, to the
important economic point of the simplification of detail
drawings. Those familiar with riveted structures appre-
ciate the necessity of calculations, allowances of spacing
of rivets, and dimensioning of such drawings. ,Here
was a uniform set of standard symbols which, when
understood by the arc welder, were placed on an outline
drawing and then sent to the yard. These symbols were
purposely made elaborate so that the entire ground
would be covered, but it was intended to simplify them
as the application settled into established practice. The
men under training were all drilled in the symbols as
here given, as will be noted on the sample record of the
student (Fig. 30).

In view of the fact that this training was special, a
detail of each practice lesson would be of little value. A
good general course of lessons for commercial welding
will be found in the Appendix. As a matter of fact, time
did not permit the establishment of standard methods of
instruction throughout the schools. The procedure of
the training may be of interest.

First, the student was provided with such necessaries
as gloves, screens, electrode holder, etc., and then
assigned to a welding booth. He was then shown by the
instructors how to hold his electrode, both at the proper

star AND ABC WELDING

THE ARC WELDER 139

angle and the correct distance from the work. He was
then allowed to practice on depositing one layer of metal
in rows on a flat plate. Following this he placed a second
layer, then a single layer between the first row and then
completing the sample by a second layer. The sample
was then to be retained by the instructor as a record and
inspected for the determination of rating. This same
exercise was repeated with the work material set up at
an angle of 45 degrees from the welding table. Follow-
ing this practice the same exercises were carried out
in the vertical, horizontal, and overhead positions.
The first course was done entirely with the bare mild
steel electrode. They were repeated with the slag-
covered electrode.

It was expected by the time the student had per-
formed these exercises he would gain sufficient confi-
dence in his ability to imdertake the more difficult
joining of plates. If in the judgment of the instructor
his deposited samples did not give evidence of such con-
fidence he would repeat such of the exercises as he consid-
ered adequate. The student was then given a graded
course in joining small samples of J4"ij^ch steel plate
with different types of joints in all positions. These
exercises were repeated with i/^-inch plate. In the same
way as with the deposited metal sample, first the sample
joints were made with the bare electrode and then with
the covered. Coupled with these exercises and in some
cases interrupting the regular course, production jobs
were given the more proficient student. Although
in a way these exercises were laid out as routine les-
sons the exigencies of the training were such that
it was considered beneficial not to make the instruction
too monotonous.

140 SPOT AND ARC WELDING

The time of the student in the schools was never
limited. If the man were conscientious, speed was not a
consideration, but the quality of his work permitted his
discharge that much sooner. As a matter of general
interest the average time of the average student in be-
coming proficient in the handling of the electrode was
approximately eight weeks. That is to say, that by con-
sistent attention the average student could go through
the course laid down in this time. There were some men
in attendance for three months. There were others who
finished the course in five or six weeks.

. Home-office records were kept of the available arc
welders in order to supply information of this kind to
the shipbuilders. The form for this purpose is shown in
Fig. 31. From this accumulated data a list was com-
piled filing additional information that would both aid
in the proper selection of men and would assist the in-
structor in case any of the men were sent to one of the
training schools. Following this system a card was filed
(Fig. 32) for each student. On this card the designa-
tions of the different schools ( called training centres )
are given and space is provided for showing transfers
from one school to another. The shifting of the student
was not the usual procedure, although in some cases it
was deemed advisable. At times better instruction was
given in certain details in one school than at another,
physical opportunities sometimes decided the transfer,
and rarely the transfer was requested by the employer.
Although this filing card shows seven welding schools,
the work never proceeded to more than five schools, as
follows: One in Schenectady, N. Y.; one in Cleveland,
Ohio; one in Brooklyn, N. Y.; one in San Francisco,

THE ARC WELDER

141

s.
1

•s

c
c

"§

a S
^ 1.

2.
t

S
1

r
I

I
f

142 SPOT AND ARC WELDING

CaL ; and one in Philadelphia, Pa. The two latter schools
were only just opened when the electric- welding activi-
ties of the Emergency Fleet Corporation ceased. In
conjunction with the home-office record of the students,
a weekly report was forwarded from each head instructor
of the schools ( Fig. 30 ) . As will be seen, this report
was filled in by the student up to the column headed
" actual welding time," from this column it was filled in
by the instructor. It will be noted that the standard
list of symbols is used with slight modifications due to
the fact that the students were ixot working from draw-
ings, so that drawing symbols would only confuse them.
These weekly reports gave a check on the instructor as
well as the student, and also formed a reference in deter-
mining the results of the examinations for certification.
It was determined to certificate those men who after
several months' actual work in the shipyards showed
themselves capable of performing successfully with the
electrode. This required a further examination of the
student. The system, though exacting fairly rigid re-
quirements, was based on broad, practical lines following
the course of instruction given. As a guide for deter-
mining whether the man was entitled to certification
these points of observation were required of the examin-
ing instructor : ( 1 ) Ability to weld in all positions, flat,
horizontal, vertical, and overhead. (2) Ability to weld
in all positions with both alternating and direct current.
( 3 ) Ability to weld in all positions with both bare and
covered electrodes. (4) Ability to maintain an arc not
over }i inch long, without more than three breaks in a
12-inch rim of the electrode. (5) Welding by observa-
tion must be smooth and even. (6) A sample weld when

THE ARC WELDER

143

c
B

»^ •■»» «

2. S. S. S.

8
I

rr M

cr.

?
ff

o o o o o o

p^fc F^^ ^'^ ^^^ ^'^ ^^%

8 S 8 ^ _ _

o
p

I i

S-.

O Ov ^ CO IG 1^

^ ^ ^ ^ ^ ^

rt St o

144 SPOT AND ARC WELDING

cut by hack saw must show perfect penetration with
work materials. (7) The weld must be free from slag
and blow holes. (8) The operator must display a knowl-
edge of the effects of too high or too low open voltage.
(9) The operator must display a knowledge of the cor-
rect current adjustment for the electrode size and work
material. ( 10 ) The operator must display a knowledge
of the proper polarity for various welding conditions.
(11) The operator must not be examined unless he has
had at least four months' experience in production work.

Differing slightly from these requirements for cer-
tification the following points were laid down as a guide
to the examining instructor for a refusal to recommend
certification: (1) Unsteady habits of the operator. (2)
Inability to hold an arc ys inch or less. (3) Slag or
blow hole3 found in the welded sample. ( 4 ) A rough or
uneven weld. ( 5 ) Lack of penetration when the welded
sample is cut with hack saw. (6) When the sectional
area of the weld is equal to the sectional area of the work
material, the ultimate tensile strength of the weld must
not be less than 75 per cent, of the work material. ( 7 )
Lack of knowledge of current adjustment, size of elec-
trode, kind of electrode, and voltage conditions. (8)
Lack of knowledge as to the proper polarity for the
type of weld. (9) Less than four months' experience
in production work.

With these instructions the examiner was sent to the
plant employing the student and made his notations on
a record card ( Fig. 33 ) . These cards were kept on file
as a reference when viewing the sample welds. These
latter were brought to headquarters for final examina-
tion and tests. The system was planned to avoid inter-

THE ARC WELDER

145

o
p

i
p

OD
CO

§'
5'

e
ft

•1

3
i

Cn CO

00 O)

OO^OCn^OOtQi^

Flat

Horizontal

Vertical

Overhead

Position

owowowow.

p
. p

1 f

>
o

146 SPOT AND ARC WELDING

ferenee with the operator and, in addition, to furnish him
with further instruction if he so desired. Fig. 34 illus-
trates the form of certificate.

Needs of the Operator. — The process is not danger-
ous. Low voltages are employed so that there is no fear
of serious electric shock. The temperature of the arc is

Fu. Si.— Puna d( artiGotc tor irc wdda, Emcrsenc; Hcct Corpoimlicia Scboak '

high and the metal in the vicinity of the weld hecomes
and remains hot for some time after the weld is made
and should be handled either with some form of tongs
or with non-inflammable gloves. The arc is usually ac-
companied with sputterings and sparking which requires
that the operator wear gauntlets so that the spark will
not burn his arms. It is not the general custom to wear
a leather apron, thou^ in some positions and classes of

THE ARC WELDER 147

work this should be insisted upon. Usually the are
welder wears a pair of overalls, but if a piece of slag
accidentally falls or is blown from his weld his overalls
will not protect him. Accidents have been caused by
molten metal falling into the shoe-top, and it is wise to
provide shoes with special tongues.

The most necessary protection is that for the eyes.
Many investigations ^ have been undertaken in order to
supply the operator not only with correct lenses in the
sense of preventing the harmful invisible rays, but also
with proper lenses that will not reduce the light intensity
so low that vision is difficult. The arc welder must see
as much as possible what is going on while the electrode
is depositing metal in the weld. Protective lenses may
be mounted in three ways. For inspection work a pair
of close-fitting goggles is all that is necessary. They
must be screened at the sides, because the invisible ultra-
violet and infra-red rays are reflected in the same way as
visible rays. Goggles do not protect the operator as
well as a hand screen. He can quickly lay this down
when he has broken his arc and it gives him a ready
means for viewing quickly his cooling metal. The screen
covers his face and chest, giving him protection on all
sides. Some operators claim a liking for the screen on
the basis of its steadying quality ; that is, by holding the
left arm tightly pressed to the body gives added confi-
dence to the movements of the right arm. Metallic arc
welders rarely use a helmet, but in carbon arc welding
this is necessary both for the added protection to the

1 i(

Eye Protection in Iron-Welding Operations " W. S. Andrews, General
Electric Review, December, 1918; Bulletin 93, Technologic Papers, Bureau
of Standards, Washington, D. C.

148 SPOT AND ARC WELDING

neck and shoulders from the intense heat, and also for
the freedom of the left hand, in which the melt rod must
be manipulated. There are many designs of helmets
and screens which are subject to the personal preference
of the welder. In some welding shops combination of
glasses for making up lenses are left to the selection of
the operator. Only glasses are provided whose combina-
tions will guarantee safety. Other employers are more
rigid and require only one type of glass to be used. A
clear glass is placed outside of the colored lenses to pro-
tect them against pitting. The cheap clear glass saves
the expensive colored glass.

Three Important Lessons. — Too much stress can-
not be laid upon the primary points in the instruction of
an electric welder. All training practice must begin
with the laying-on of metal from the electrode. The
thi'ee following lessons explain clearly how the beginner
should practice in (1) making beads, (2) spreading the
deposited metal, and (3) padding.

((

Lesson I

((

BEADS

" As shown in Fig. 35, Lesson I, Running Beads.

" The electrode must be held close to the work, i.e.,
a close arc for successful welding. The natural ten-
dency is to draw a long arc, but in this case too much of
the heat is lost by air radiation and there is too great an
area for oxidation of the metal in passing from the elec-
trode to the work and there is too great an area provided
for the air to get in and form oxide and nitride, both of
which are very undesirable impurities in the weld. There

BEADS 149

is a further chance with a long arc that the electrode
material will not be deposited in the parent metal made
fluid by the arc, but spattered outside or overlapping, in
which case no weld results.

" As shown, the electrode material must pass into the
crater or fluid bowl of the work made so by the arc, and
the electrode must be held in such a position that the
metal can pass nowhere but into this fluid bowl.

" The fluid bowl only appears at the point the arc
strikes, and the arc must be kept striking just ahead of
the deposited metal partially in the parent metal or, to
say it in another way, the arc must be kept at the advanc-
ing edge of the puddle. If the arc is aUowed to draw
back on the bead simply a bridge of metal welded from
the weld to the point where the arc next strikes the par-
ent metal will result. This exercise should be made with
both flux-coated wire and bare wire. It will be noted
that with flux-coated wire with the bare side advancing
as it is designed, the flux coating will follow along,
covering the molten puddle and preventing the arc back-
ing up on the weld providing a close arc is held.

" To get the correct rate of heat for any given condi-
tion of electrodes or work the following test should be
applied, and it is the best, in fact, the only test yet de-
vised. Run a bead and note whether the edges are
undercut as at Fig. 35b, overlapped as at Fig. 35c, or
perfect as at Fig. 35d. If they are slightly undercut it
can be told by looking at the bead, but the perfect weld
and the overlapped weld can only be told by chipping
off the beads. Different electrodes melt at different
rates of heat, and different kinds of parent metal show
small and large fluid bowls or craters, depending on their

150

SPOT AND ARC WELDING

make-up, so that even with the correct length of arc this
test should be applied for each new condition. The

CORRECT
CONDITION

UNDERCUT

DEPOSITED
METAL

CRATER

OVERLAPPED
(C)

SATISFACTORY

(d)

CRATER, SMALL, AND NOT ]/^HERE
METAL DEPOSITS

RESULTS OF LONC ARC
EXCESSIVE OVERLAPPING LAST DROPS

y»

(f)

Fig. S5. — Running beads.

undercut in general results from a larger bowl than the
electrode deposited to fill it, and the overlapped bead

BEADS 151

may result from this or from a long arc. Where the
bowl is made smaller with an arc over y^ inch long with
certain current densities the bowl may disappear en-
tirely, showing results as at Fig. 35/^ g, and e^ which
also show how the results shown may be had.

" The bead should be run until proficient with the
three main different styles of electrodes, that is, the com-
pletely coated electrode in which the arc exists inside of
a viscous tube deposited at the same rate as the electrode,
the idea being to keep the oxygen and nitrogen of the air
from combining with the steel when molten and when it
has its greatest chemical affinity. With this electrode a
slag is left around and on top of the weld which must
be chipped off for a clean start if the arc should happen
to get out or in restarting to continue the bead.

" With the flux-covered electrode, bare side advanc-
ing, it will be noted that there is no danger of slag inclu-
sion and that the slag can be crushed off and that in
restarting there is no need for chipping the weld clean.
The best way of cleaning the weld is by the same action
which brings the slag and dross to the surface of the
weld with this electrode. With bare wire there is natur-
ally no danger of slag inclusion, but a continuous amount
of the oxide and nitride of iron are included as the time
for bringing them to the surface, i.e., while the metal is
too short to allow them to disentangle and come to
the surface.

" After proficiency in running beads horizontally is
attained the beads should be run on a vertical surface
from the bottom up. Here it will be noted a close arc
must be held or metal will run away. The next step is
to run the beads horizontally on the vertical surface, and

152 SPOT AND ARC WELDING

lastly, from the top down. By this time a close are will
be so natm-al that overhead welding can be attempted,
but if not successful no worry need result, as it will come
naturally in time. The most important things to be
learned in running beads are : ( 1 ) Close arc, i.e.^ not so
close as to sputter, but close enough to crackle. (2)
Keep the arc at the advancing edge of the puddle. Beads
are used commercially to seal cracks, caulking edges,
and as the first layer in butt and fillet welding
for ductility.

Lesson II

((

" SPREADING

" Here the electrode method is applied by moving tlie
electrode from side to side at the same time advancing,
Le,j back and forth in a zig-zag manner as shown in Fig.
36a. The electrode must not be moved rapidly, i.e.,
must not be moved faster than the arc can make the
liquid bowl or crater and must be moved at an even rate,
so that this crater becomes a trough into which the elec-
trode material is deposited at an even rate. The rate of
speed back and forth of the electrode must be the same
rate as that used for the advance in rimning beads, and,
in fact, each successive layer in spreading is simply cross-
wise beading done continuously and shortly, the welded
metal becomes heated to such a point where a greater
speed is attainable than at first. It will be noted that
the completely-covered slag-coated electrode lends itself
admirably to this method. It is much easier to apply
this electrode by spreading than by any other method,
as the successive movements from side to side keep the
puddle of deposited metal in the centre as a river of

SPREADING 153

metal with banks of slag, and as advance is made the
banks of slag close over the metal a short distance be-
hind the arc. This electrode in starting or restarting
the cold slag must be chipped away, and it should be
noted that the slag must be cold, i.e., black before it can
be chipped away. In merging the spreaded beads in
these electrodes the edge to be merged must be chipped
clean, else there is almost a certainty of slag inclusion in
the weld. In spreading with a flux-coated electrode the
bare side should be kept pointed towards the parent
^etal, and it will be noted that the flux coating forms
an insulating slag, so that with a close arc it is a natural
tendency to advance at the edge of the previous layer or
bead, and not to back up on the already deposited metal.
With bare wire there is no such aid and special care
must be taken to not bridge over or leave voids in this
method of spreading, i.e., bare wire is least applicable to
this method and completely-coated wire is most appli-
cable, while the half -coated flux-covered lends itself to
either spreading or padding by beads. Correct spread-
ing is shown at Fig. 36& with the welded metal sunk
in, or, as the experienced welder terms it, bit into, the
parent metal. Spreading with too long an arc results
in the deposited metal simply lying on the parent metal
as at Fig. 36c, with no biting-in effect. Fig. 36e shows
how padding can be accomplished by spreading, care
being taken, of course, to merge the spread layers into
each other, and it will be noted that this method is more
applicable to a small amount of beading whereas beads
are more applicable wherever the surface has to be raised
higher. In general more heat and hence with the same
length of arc more current can be used in spreading

154

SPOT AND ARC WELDING

than by any other method, and with speed or rate of
deposition of the metal with the arc at constant length

f')

/m^/^ <:mi

(t>)

CORKBCT SPREADING

SPREADING AT TOO 6R£AT
_^ „^^ ■>• SPEED

(d)'THIN LAYER PADDING BYSPREADlNQ

wcm^m

WELD FROM
BOW SIDES

ORIGINAL BUTT JOINT

BEFORE WELDING

y//yyj<^/jyyxyy^, . ^x^c/y^yyywx/yx/y/v:

yjryx^A/ZM/'^yzAyA . •/•/yzvx^z^yy^

(e)

(9)

WELDING BilTT JOINTS BY SPREADING ON THIN WORK-
'/fs'*" %/' STEEL, */a **'- ^6** CAST IRON AND
AUTO-SPRING HIGH CARBOtt STEEL,.

(f)

(h)

SPREADING WELDING OF THIN STEEL BUTT JOINTS WITH

COOUNO BACKING

y/z/z/z/m///^

WELDING FROM ONE SIDE ONLY
Fia. S6.— Spreading.

depends on the current; greater speed can be made by
spreading than by any other method. This is because

SPREADING 156

the parent metal is advanced over at a more rapid rate
and hence a greater rate of heat in the arc can be taken
care of by the weld conducting the heat of the arc into
the parent metal. It is this action of heat conduction by
welding that allows of the high temperature of the arc
to be the correct results from mass temperature for
molten iron and other metals. The fact that a weld is
made conducting the heat away, and if a weld was not
made as would be the case in depositing steel on a cop-
per plate, the deposited metal would be burned beyond
recognition at a rate of heat, and hence current adjust-
ment with the right length of arc would be perfectly cor-
rect when welding into steel. Figs. 36g and / show an
application of spreading to a butt-joint on thin work.
By thin work is meant from 1/16-inch to 5/32-inch steel
and ^-inch to 5/16-inch cast iron, or high-carbon steel,
such as an automobile spring. Work thinner than 1/16
inch should be backed up as in Fig. 36^, with a cold
mass such as another piece of steel or a water pad, so
that the metal when molten for welding will not fall
through. If the piece is of such thickness as in e and /
that the spreading will not melt the edges so as to fall
through, a spread can be put on both sides which will
merge in the centre as shown, and weld without any prep-
aration of the joints. Automobile springs being about
3/16 inch to 34 iJ^ch thick can be welded successfully in
this manner, and the excess metal of the spread ground
off. The reason spreading is used for this comparatively
thin work is the phenomenon that in proceeding with a
bead the edges become too hot and fall through, whereas
in lacing back and forth, as in spreading, more of the heat
is carried into the parent metal.

166 SPOT AND ARC WELDING

" Lesson III

" PADDING

" Padding in general is a succession of beads run
parallel to each other and offers a great field for useful-
ness in building up worn parts or parts machined down
too far for subsequent machining to correct size. In
laying these pads parallel the electrode must be held so
that the arc bites into the preceding pad and the parent
metal at the same time, as shown in Fig. 37a. If the arc
is held as in Fig. 37b a good joint will be had to the
parent metal, but little or no joint to the first bead, and
in machining these welds voids will be found as black
slag spots, as can be imagined from Fig. 37/. Fig. 37c
shows these beads correctly merged, and Fig. 37d shows
the beads perfectly welded to the plate but no merging,
and hence no good for subsequent machining. In com-
mercial padding the choice of running the beads length-
wise or crosswise of the work is to be had, and in general
lengthwise is the more desirable, as the piece has more
time to absorb the excess heat of the arc before return-
ing to supply more heat. In either case the outer edge
should be gone around with the bead for each layer, so
that a sort of trough can be formed while the metal is
cool and shows the least disposition to run away. For
quick rough work this trough could then be filled in by
spreading, but in general successive lines of beads is the
more reliable method. After the method of a reasonably
smooth pad on a horizontal flat surface, if attained, mak-
ing a pad on a vertical surface is next in order, running
a bead horizontally with successive beads directly over
it and later making the pad both from the bottom up in
vertical layers horizontally and from the top down. In

PADDING

157

making these pads overhead it will be noted that the
only hard bead to put up there is the first one, and when

FIRST BEAD

"> wmm///////M

t RATER

CORRECT

CORRECTLY MERGING BEADS

(C)

ELECTRODE

(h)

fARC

CRATER

^^^

INCORRECT

INCORRECT-NO MERGING

(d) ^

SUCCESSIVE LAYERS
CORRECTLY MERGED

^^ M ^ INCORRECT

^:^:^A^

NO WELD
^t^ ^RS OF OXIDE

(Q) ^^^^'/yy^^ ^^ EFFECT ON PADDING

Fig. 37.— Padding.

welding against this first bead and the parent metal the
student will be able to notice correct conditions for
drawing the arc equally from both, as in that case the

LAYER OF
OXIDE

158 SPOT AND ARC WELDING

metal shows the least disposition to fall. The overhead
welding has no function of negative or positive polarity,
but simply the fact that one liquid drop will remain on
the ceiling, excess over one drop falls, leaving one there,
is the secret of overhead welding. It has probably been
noted before this that in welding no drops should appear
at the end of the electrode, i.e., when the arc is held the
right length and the metal flows evenly, the metal of the
electrode passes through the fluid bowl as a gas or vapor
and condenses on the parent metal as a liquid, rapidly
changing into a viscous and then a solid mass. The rip-
pling appearance of the finished weld being successively
frozen ripples of the liquid pool due to the magnetizing
effect of the current in the arc when just at the point of
freezing. In general; overhead welding requires more
current for the same conditions, as the arc must be held
closer, which means slightly less voltage, and henqe to
make up the same heat the current must be raised. As
stated before the one trick in overhead welding is in
getting started and holding an ajbsolutely unbroken
close arc. A great aid in holding this close arc is for the
student to rest his body and left elbow, if he is right-
handed, against the piece to be welded, steadying his
right hand holding the electrode holder,

" Another way is to put the right elbow up against
the overhead surface and to use the elbow as a centre
and guide for bringing the hand and electrode up at an
even rate. In overhead welding the student will best see
how the electrode must be fed into the weld at the exact
rate the electrode is being melted. Another method still,
in order to feed the electrode at an even rate, is to use the
elbow of the welding hand resting on the knee and rais-

PADDING 169

ing the toe at the rate desired. Another permissible
method for welding and especially useful in overhead or
vertical welding is to use a stick similar to an artist's
maul-stick. Overhead welding even when unsuccessfully
tried shows what is required for first and successful pads
more quickly than any amount of verbal or written in-
structions. Padding by beads should only be done with
bare-wire or flux-coated electrodes, as coated electrodes
are only successfully applied by spreading, as shown in
the second lesson. The one thought necessary in suc-
cessful padding is the merging of the pads with the
parent metal and with each other." ^

"ITiese lessons are published in full by the Electric Arc Cutting &
Welding Co., Newark, N. J. They were conununicated to the Author by Mr.
C. J. Holslag, chief engineer of the company.

CHAPTER VIII

The All- welded Ship

The first and last resolution of the group of experts
gathered to advise the Emergency Fleet Corporation on
the use of electric welding in the ship program urged the
building of an all- welded ship. The last resolution
carried with it the request for the organization to be
formed and the money to be appropriated for this pur-
pose. Such requests were never followed by authoriza-
tion, and the all-welded ship, like other innovations, was
for one reason or another never started. The advocates
of the process never halted in their endeavors to suggest
means of obtaining permission to build such a vessel and
many progressive shipbuilders offered to undertake the
construction if formally approved.

In the early deliberations it was not felt that the pro-
posal to build a welded ship would be immediately
acceded to, but when certain practical tests and demon-
strations were completed there would result no hesitancy
on the part of those in authority. This view was con-
sistent with the times as greater innovations were readily
sanctioned. So that no time might be wasted, the inves-
tigational work, both practical and theoretical, was
pushed with all haste, data in great quantities gathered,
and the educational work was laid down in some syste-
matic form. Without a great deal of effort, it was found
that the shipbuilders were using the process very spar-
ingly and efforts were made to find the reason why the

160

THE ALL-WELDED SHIP 161

non-essential parts of the riveted ship were not are
welded. Some shipbuilders were doing a little of this
work but wished to do more, and the question of permis-
sion rested with the classification societies. This excuse,
if it were such, was soon expunged by a joint approval
by Lloyd's Register and the American Bureau of Ship-
ping of a list of ship's fittings which could be arc welded
(see Appendix) . It was objected that this list referred
only to jobs with which ordinarily classification societies
did not concern themselves. These societies had gone fur-
ther by a clear statement that upon submittal with full
information they would consider for approval proposals
for arc* welding " other parts of the vessel." It was this
last expression that gave rise to the probable extension
of arc welding to a standard riveted ship with the result
that a special committee was appointed to give views on
this subject. .

Welding a Standard Riveted Ship. — ^Realizing fully
the greater benefits that would accrue in a welded ship
designed from the point of view of this process, still it
was believed that the standard riveted construction could
be more expeditiously put together with the electrode
than by riveting, riveting work being at that time the
reason claimed by shipbuilders for the delay in ship de-
liveries. The steel shapes and plates were being deliv-
ered by the mills in ample time; and quantities of
material unfabricated could be assembled in the way
selected. This naturally would not show the saving in
cost nor the reduction in materials and weight that was
inherent in the electric- welding process when used to the
f idl ; but still there would be a small percentage of sav-
ing. The main assumption was not conimercial economy
11

162 SPOT AND ARC WELDING

but saving of time, the serious problem then facing the
country. The report of this sub-conmiittee showed that
speedy construction would follow the omission of water-
tight stapling required in riveted ships, the omission of
all cementing of decks to the shell, as the electrode could
guarantee water-tightness, the omission of all laps, liners,
jogglings, straps, etc., in the shell plating above the
bilge by employing butt-welded seams — the same thing
was recommended for decks. Riveting of shell plates to
frames was suggested as a conservative measure. Spot
welding of the brackets to frames or beams and the spot
welding of floor plates to frames, reverse frames and
clips, was considered feasible if done in the shop and
these assembled parts arc welded to the other members
in the ship. Spot welding was also recommended for the
attachment of cargo battens and other fittings. Addi-
tional suggestions included all non-strength members,
all deck erectiotis, smoke pipes, up-take, ventilators,
ducts, combings for hatches, and man-holes, door frames,
separately-built tanks, lockers, and racks. The masts,
booms, shaft and pipe tunnels, and similar cylindrical
work usually riveted in two pieces were to be welded
throughout, omitting the straps. Cast-steel fittings
were to be welded to plates. Oil tanks were to be welded
" even in conjunction with riveting." Swash plates
could be tack welded to decks and bulkheads omitting
the flanges in order that these plates might be washed
away without endangering the hull proper. The recom-
mendations also include the welding of all stanchions
both head and heel, pipe railing, and all flanges on steel
pipe and tubing.

It is easy to see that this detail list covers the welding

THE ALL-WELDED SHIP 16S

of approximately 90 to 95 per cent, of the ship. In fact,
it only reserves for riveting the main strength members
of the hull. This suggestion was never put into effect.
It occasioned a full discussion on the merits of combin-
ing arc welding with riveting and established the
opinion that an arc- welded and riveted joint was
not satisfactory.

For emergency repairs certain riveted joints had
been arc welded along the edge instead of caulking.
The damaged part had destroyed the caulking edge.
These jobs were carefully watched. The riveted joint,
if not reinforced by a strength weld, i.e., if the edges
were simply made water-tight with a fillet weld, due to
the creeping action, would throw the strain upon the
welding. The water-tightness of the joint would then
be impaired. If a strength weld were made on the edge,
all strains would be taken off the rivets and they would
serve no good purpose. In repair work it was found de-
sirable to weld on a thin-flanged piece over the entire
riveted joint. This flanged piece allowed sufficient play
to the riveted joint and yet would not tend to open the
welded joint. In this manner a good water-tight job
could be obtained.

Welded Craft. — ^As far as is known the first water
craft to be partially electrically welded is the Dorothea
M. Geary ^ a motor boat built of steel and 42 feet in
length. She was launched about the end of November,
1915, and is used for ship repairs in Ashtabula Harbor,
Ohio. Although the small size of this boat and the thin
plating (about 3/16 inches thick) used for shell and

^"The First Electrically-Welded Boat," John Liston, General Electric
Review, December, 1918.

164 SPOT AND ARC WELDING

decks may not assure the conservatives that 10,000-ton
ocean-going cargo vessels may be built likewise, yet the
record of the boat in service is remarkable from the
standpoint of the strength and fatigue-resisting quali-
ties of electric welds. " On December 17, 1915, shortly
after the launching, a call for repair work was received
from Fairport Harbor, distant about 30 miles, and al-
though Lake Erie shipping was practically suspended
at that time owing to weather conditions, the welded boat
was at once headed into the lake which was covered with
floe ice and made the run to Fairport in about three and
one-half hours. When the harbor was reached it was
found to be covered with four inches of solid ice, and
into this the welded boat was rammed, breaking her way
through, at reduced speed but without a stop, to the
pier where the ship she was to work on was laid up. After
the return to Ashtabula careful inspection of the boat
failed to show that any injury had been sustained." ^

And again: "In the following year an accident
occurred which threatened to destroy the welded boat.
Work was being performed aboard the freighter Aleccis
Thompson in the Superior Slip at Ashtabula, the welded
boat lying alongside, when the freighter C. Russell
Hubbard, which, moored at the opposite side of the slip,
broke adrift, swung across the slip, literally squeezing the
small craft between the two large freighters. . . . The
sides of the welded boat being crushed into a maximum
distance of 18 inches amidships. . • . This damage
Was repaired by means of jacks which were used to force
the sides back into normal position. . . . Leaks were

*"The ElectricaUy-Welded Boat," John Liston, General Electric Be-
view, December, 1918.

THE ALL-WELDED SHIP 165

started, but investigations showed that they were again
due to loosened rivets and not to any failure of the weld."

The edges of the plates were V'd for butt welding
and a metallic electrode was used of approximately the
following constituents: Carbon, 0.10 per cent.; man-
ganese, 1.87 per cent.; and a trace of silicon. The keel
was electrically welded, but the hull plating was riveted
to the frames and keel, and " the structure above the
deck line was riveted and strengthened with angle iron."
The welded seams were left reinforced, and after welding
were " pneumatically hammered."

In 1917, a 60-foot section of a 1200-ton bulk-oil barge^
was electrically welded in this country. The remainder
of the barge was the usual riveted construction. This
craft was for service in Mexico. She was 165 feet long,
38-foot beam, and about .8 feet 6 inches in depth. She
was constructed of ^-inch plates for decks and shell,
and the transverse members were of 5/16-inch plating.
It is stated that this barge carried nearly a full cargo
from New York to Tampico, a distance of 2500 miles
on the ocean, without harm to hull or cargo. " No re-
pairs have thus far been required, notwithstanding the
barge has been in service about twelve months as a bulk-
oil carrier on the Panuco River, where the rapid currents,
wind, and tide, combine to make navigation difficult."

Reference has already been made to welding activi-
ties in England. In the early part of 1918 there was
built an all- welded -cross-channel barge.* This craft had
a deadweight carrying capacitj^ of 275 tons. The frames
were 2>^ X 2^ X >4-inch steel angles, the floors w ere

' Report of Capt. James Caldwell to the U. S. Shipping Board, 1M8.
* Report of Capt. James Caldwell to the U. S. Shipping Board, 1918.

166 SPOT AND ARC WELDING

7 X 3^ X 7/20-inch angles, and the shell and deck
plating ?^ inches. The edges of the shell plating were

joggled in order to provide flat welding and to reduce
overhead welding to a minimum. Holes were punched
for assembling the vessel in the regular manner. These
holes were spaced about 10^ inches to receive the service
bolts for erection purposes. After the welding was com-
pleted, the holes were closed by the electrode. The shell
seams were full welded on the exterior, but tack welded
on the interior. She was divided by three water-tight
bidkheads. The hull was flat and there were four strakes
of plating to each side.

Though minor leaks were discovered on the first
loading, the barge has been in successful use across the
channel and has shown no signs, according to last re-
ports, of anything detrimental to the process.

Careful records of time and cost were kept for com-
parison with riveted barges of the same size and type.
It is interesting to note that the total cost of electric
welding was $1500. Of this item $310 represented cost
of labor, $300 the cost of current, and $890 the cost of
electrodes. This latter item is high, due to the use of the
slag-covered electrodes, which is the approved practice
in England.

Next of interest is the time of welding. At the com-
mencement of the work, the average was about 4 feet
an hour, but towards the completion this increased to
about 7 feet an hour. During the work a maximimi of
14 feet an hour was attained.

In 1918, arrangements were completed for the elec-
tric welding of a battle-towing-target keel, to be built at

THE ALL-WELDED SHIP 167

the Norfolk Navy Yard. This structure is built of steel
shapes and plates, and functions under water to support
the wooden target which is destroyed by gun fire. Al-
though not rightly classed as a water craft, still the
shock, strains, and stresses that are so much the concern
of those who design and build steel vessels will be suf-
fered by this keel. It is unnecessary to detail the con-

Fm. SB.— BitUe^owing-tuget k«j.

struction elements. A very good idea of the keel is
given in Fig. 38. This structure is now completed and
awaiting the wooden super-structure. When this latter
is built and their combination effected, the entire target
will be tested at sea. FuU reports of the trial of this
keel in service will probably be available after the navy
has completed its investigations.

Patented Designs. — During the war, two American-
patented designs for welded ship constructions were
patriotically offered for the use of the Emergency Fleet

168 SPOT AND ARC WELDING

Corporation, As will be seen, these designs were not
accepted, and drawings were prepared for a proposed
welded ship by a special sub-committee. Without enter-
ing into a mass of patent claims, one of these American
designs ^ was characterized by omitting many of the
small connection pieces used in regular riveting con-
struction and thus reducing weight and cost. By tack
welding of clips and lever fulcrums temporarily, the
usual run of badly twisted shapes and plates could be
straightened preparatory to welding. By this same
method it was claimed that an entire ship could be easily
assembled without the use of assembly bolts, thus avoid-
ing the expense and damage to the work material by
punching holes. As the fulcrum clips and handling
attachments were only lightly welded, they could be
readily knocked off the original metal with little loss of
time and no damage. /

The other American design^ was based funda-
mentally upon the conception that angle bars, which are
necessary to a riveted connection, were unnecessary for
the welded connection. With this in view, the construc-
tion of a vessel was reduced to the use of plating through-
out. The design carried the requirement that one edge
of the plate be flanged. Thus in the case of the shell
plating the flange served as a continuous longitudinal
member. The design embodies the best strength quali-
ties with reduction of weight of both the transversely-
and longitudinally-framed vessels. By cutting notches
in the straight edge of the plates at desired intervals,
bolts could be used to draw up and secure the work for

*Capt. James Caldwell's Report to the U. S. Shipping Board, pp. 109
and 93, 1918.

THE ALL-WELDED SHIP 169

welding. This method of assembly was not essentially-
different from present practice. The main objection to
this design was the fact that the plates must be flanged at
the mill, as the run of ship structural steel would not
permit of cold flanging. As the larger percentage of
material was required to be flanged, it would seem from
the standpoint of standardization that the design
merits consideration.

An English patent shown in this country differs
from the last American design in that it clung to the
angle-bar connection. Instead of employing continuous
welds along the edges of the angle, it preferred to notch
out the angle flanges with desired spacing and then arc
weld at the notching, if the boimding bars were simply
for holding together non- water-tight members. If the
compartment was to be water-tight, then one edge and
the notching of the angle bars were welded. Straps
were prepared with elliptical holes, so that they would be
welded as well as the edges. The whole design was con-
sidered from the point of view of the fusion of all the
parts requiring to be joined. In such places as the inner
bottom of the ship, where lengths of moderately- thin
plating would affect the jointure by panting or fatigue
stresses, brackets formed by angle bars were welded as re-
quired. Evidently this design had in mind the reduction
of welding, but seemingly the preparatory work on the
original materials would exceed the saving in arc-
welding labor and materials. In view of the American
design discarding the angle bar, and the results of tests
of both butt- welded joints in flat plates and cross con-
nections, apparently there is little merit in this
patented construction.

170 SPOT AND ARC WELDING

Emergency Fleet Corporation Design.^ — This de-
sign was made under pressure due to the exigency of
the time. No more than three weeks were given for the
preparation of complete drawings of both the ship and a
proposed yard in which to build her. At that particular
time, no shipbuilder would consider the proposition of
building any more ships. There was a growing shortage
of labor, and riveters were scarce and costly. The objects
sought in this design were a ship that could be quickly
manufactured, not built, a method of shop procedure
that would not make further demands on a depleted labor
market, and yet a restriction of design that would not
exceed the possible workings of the economic law.

For general outlines, accommodations, propulsive
machinery, etc., a standard ship was taken as a guide.
The majority advice of welding experts was taken in
the adoption of, and the approximation of, percentage
strength of the electrically- welded joints. At that time,
the consensus of opinion was that the strap-joint with
three full-strength welds was the strongest joint. This
design of joint was used in all connections of principal
members. Where plates met at right angles in cruci-
form, full-strength welds were provided. At the time
of design, it was considered conservative to provide a
bearing strip to distribute the thrust where a plate was
connected at right angles to another imsupported plate.
After discussion on this point, other methods were sug-
gested for overcoming this difficulty without resort to
the bearing strip which in the minds of same experts was
a detriment to the connection.

Perhaps the most radical depa rture from a ship-

" Supplement of Notuticus, June 1, 1918, N. Y.

THE ALL-WELDED SHIP 171

design viewpoint was the use of transverse plating in-
stead of the customary longitudinal plating. When the
design was first discussed the question was asked : What
is the function of the shell plating of a ship ? The reason-
able answer is contained in the familiar name " the skin
of the ship," Le., it keeps the water from coming in and
prevents the cargo from going out. In short, the shell
of the ship has but a small share in the strength of the
ship. In this design, the strength of the hull was more
than conservatively designed. The keel, the centre keel-
son, rider, bilge plate, shear strake, and upper-deck
stringer, were planned in long lengths and with few
joints. These joints were carefully distributed. The
calculations of the naval architect showed that a section
through the cargo hatch as designed was stronger by 10
per cent, than the riveted vessel. The practical purpose
served by the transverse plating was that it gave a method
for manufacturing ships.

Adopting a 6-foot-wide plate, the design provided
for putting the ship together in sections the width of the
transverse plate. Each plate was fitted in the shop with
two frames and their connections. There were two of
these pieces, one for each side of the vessel. Two sections
of the floor were likewise built in the shop and two 6-
f oot deck sections with two beams attached, one for each
deck, completed the pieces, making up one 6-foot sec-
tion. The plan of operation was to build these pieces
under shop-production methods and so arrange the work
that each day one 6-foot section of the vessel would be
completed. Thus plans could be made ahead of time
to follow a strict schedule. As these details would be-
come standard procedure, more time on the part of

172 SPOT AND ARC WELDING

executives, superintendents, etc., could be devoted to the
delays usually connected with the fitting up of the vessel.

To meet this.manuf acturing scheme, a yard plan was
devised. This consisted in. the establishment of one, and
only one, way to build the ship, from stern to stem. A
crane was designed to move up the ways, handling the
pieces from the shop, imtil they were securely welded,
and then moving on for the next section. This was more
than a crane, it was a moving shelter for the men and
shops with the addition of flexible scaffolding. The esti-
mates showed that a 6-foot section could be erected and
completely welded an one working day of eight hours;
which meant that when such a shipyard was working
under its best efficiency that a standard ship of 410 feet
could be completed in seventy-five working days. This
estimate is on the basis of a single shift of operators.

The designs as submitted were widely criticised, and,
curiously, on one point, namely, that it was not a rivet-
less ship. The frame brackets were riveted to the beams.
This was considered a conservative measure by the design
conmiittee. Opinions were pronounced for an all-welded
ship, based, no doubt, on the sentiment that an innova-
tion must stand on its own feet. Then, as time went on,
shipbuilders claimed that their works were progressing
so rapidly that shortly they would have vacant ways.
There would be no necessity for a special yard in which
to build the first welded ship. At once with this argu-
ment, all the fundamental reasons for its being deprived
this welded-ship design of further consideration. No
longer did there exist a shortage in the labor market, no
longer were people at large aroused at the shortage of
ships, no longer did the foreign cables cause sensations.

THE ALL-WEIJ>ED SHIP 173

and no longer did the shipbuilder worry over the de-
livery of completed tonnage. And soon to follow this
were the straws of the Armistice.

Isherwood Welded Ship. — ^Just before the news of
the cessation of hostilities, Mr. J. W. Isherwood brought
to this country from England a design of a 3900-ton
deadweight ship which could be electrically welded. The
application of welding was made to his well-known
patented longitudinally-framed ship. The main objects
sought were : ( 1 ) Adaptation to existing shipyards, i.e.,
all the pieces fabricated in the shops could be easily
handled on the ways by the crane service provided for
riveted ships ; ( 2 ) a careful reduction of overhead weld-
ing, especially in the field work; (3) the use of service
bolts as customary for the assembling of the hull mate-
rials ; and ( 4 ) the small amount of welding necessary in
the field. Full details of this interesting design will be
found in the Appendix. Mr. Isherwood stated that:
'^ This design was prepared by me in London with the
cooperation of Mr. W. S. Abell, Chief Ship Surveyor of
Lloyd's Register of Shipping." This gives added
weight to the design, if indeed it were needed, as the
inference is that the rules of Lloyd's for electrically-
welded ships have been fully considered. The designer
has provided broadly for any play of conservatism that
might question the electrically- welded process, and he be-
lieves as a matter of economy that, where many service
holes would be required for proper fairing, the
punching of a few additional holes and riveting that
particular connection might result in an increased speed
of construction and a reduction of cost. In general, he

174 SPOT AND ARC WELDING

has taken the very best and safest points of the welding
process and combined them with the best and most cus-
tomary shipbuilding practice. As a careful study of his
design will show, every eventuality in electric welding
has been considered and as well an inclusion of the pos-
sible developments of the art.

With this design as a basis, it was not difficult to in-
crease the size of the vessel to 5000 tons deadweight.
And as by that time the majority opinion was favorable
to the larger-sized vessel, negotiations for the com-
mencement of work on a design of this type were about
to be effected through the formation of a proper organ-
ization when the activities in electric welding were post-
poned, due to the close of the war.

Lloydfs Rules.^ — Before the issuance of require-
ments for the use of electric welding in ship construction,
Lloyd's Register made specific investigations of a tech-
nical and practical nature. That is, the tests were made
on fairly-large samples and careful measurements with
delicate instruments were taken. Realizing that many
tests of the ultimate tensile strength of electric welds
had been made but that other and more important char-
acteristics of the joint had not been considered — these
being essential to the construction of ships — experiments
were performed to determine if there were any differ-
ences in the modulus of elasticity of the weld and ad-
jacent plate; or at least a difference that would prohibit
the use of the process. Experts in the testing of steel
were not surprised at the results which showed that " the
difference in elasticity between the weld and the plain

^ See Appendix.

THE ALL-WELDED SHIP 175

plate is negligible." ^ This deduction was made upon
the completion of the welded sample, but the same
authority states : " To obtain further information regard-
ing the properties of this deposited material, small test
pieces were prepared entirely composed of it, and the
modulus of elasticity thus determined was foimd to be
11,700 tons per square inch as compared with about
13,500 tons for mild steel and 12,500 for wrought iron."
Although ultimate strength tests were made for a
comparison between treble-riveted lap-joints and lap-
welded and butt- welded joints, the most important tests
from a structural viewpoint were those subjecting the
joint to alternating stresses. The results indicated that,
whereas the welded joint would break down imder re-
peated stresses (5,000,000) with 6 tons per square inch,
the unwelded test pieces would not break down imtil
they had reached 10 tons per square inch. Alternating-
stress measurements were taken on various types of
joints, all of which led to the following conclusion:
" These tests showed generally that welded material will
not withstand a very large number, say several millions
of alternations, if the applied stress is greater than about
plus or minus 6 tons per square inch. The capabihty to
resist alternating stresses is, of course, of the very great-
est importance in shipbuilding materials, and would
appear for the present, at least, to limit the application
of welding to vessels in which the stress is not exceeded.
The calculated stresses in ship structures are in the case
of large vessels rather in excess of this figure, although,
from such information as is available on the subject, it

« ** Electric Welding for Shipbuilding Purposes," W. S. Abell, Journal of
East Coast Institute of Engineers and Shipbuilders, 1918.

176 SPOT AND ARC WELDING

would appear that the stress actually experienced by the
material in a ship is considerably less than that calculated
under the usually-assumed conditions. It is well, how-
ever, to proceed cautiously in the application of a novel
method of construction like electric welding, and it would
probably be wise to limit its application meantime to
vessels whose length is not greater than about 300 feet."®

Specialists in electric welding, though interested in
all these investigations, were disappointed for one reason,
namely, that the welded samples were all made by one
process. In a practical sense this may be looked upon as
unfortunate, but in a technical sense the maintenance of
a constant electrode material, current adjustment, and
skill of operator was more important. One of the diffi-
culties to the introduction of electric welding has been
the claims of those who prefer some special system or
appliance. And confusion must result where all the
different modifications to the simple practice are inter-
mingled in elaborate tests. Let it be squarely said that
Lloyd's Register has announced rules which require a
certain process to be employed. They have also an-
nounced that they will undertake to approve any process
of electric welding which meets their rules and which in
the judgement of their technical staff qualifies it for the
joining of ship's steel.

Summary. — From the preceding outline of the de-
velopment of the all-welded ship, it will be seen that in
this country the proposals did not advance beyond the
point of design. In England it proceeded to the actual
laying down of a coasting vessel 150 feet in length. The

•"Electric Welding for Ships," W. S. Abell, Journal of East Coast
Institute of Engineers and Shipbuilders, 1918.

THE ALL-WELDED SHIP 177

latest advice concerning the building of this vessel indi-
cates that the work is proceeding at a slow rate. The
underlying causes are concerned with commercial mat-
ters which cannot be discussed here. With the removal
of the impedimenta surrounding the technical ability of
the welding processes to joint heavy structural pieces,
there can exist but optimistic feelings that practicaF
obstacles will vanish in the future.

CHAPTER IX

Theories of Electric Welding

It is proper to state that unfortunately the nomen-
clature of electric welding is not clear. This makes
doubly difficult an exploration into the field of theories.
In the matters already treated the terms used have been
those customary to practice. It may be better for what
is to follow to explain the general conception of welding
and what is meant by "autogenous soldering" or
" autogenous welding."

The word " weld " in accordance with the dictionary
is of Anglo-Saxon origin, probably related to the verb
" well," to gush. " Welding," as a term for the modem
processes of joining metals, is in dispute. Some experts
hold that it is not fully descriptive of what takes place
in the process ; others claim that used as a general term
it creates confusion ; and still others in order to conserve
time immediately divide " welding " into two parts, call-
ing one " pressure welding " and the other " autogenous
welding," and proceed to define these branches. His-
torically, welding as performed by the blacksmith was
the union of two pieces of metal with or without an
external source of heat or without fusion. That is to say,
that metals in a cold state can be united by pressure
derived by striking with a hammer, although the force
of the blows or friction might impart some heat to the
union. This is an approach to a generic definition
of " welding."

178

THEORIES OF ELECTRIC WELDING 179

Much confusion has grown around the use of the
words " soldering " and " autogienous." It is generally
accepted that when a metallic joint is affected by means
of some external unionizing medium which adheres to
the pieces to be joined and closes up the gap between
them, this is " soldering." The unionizing medium is
usually a softer metal or alloy melting at a lower tem-
perature than the parts to be joined. " Brazing " does
not complicate matters because the term is used restric-
tively for a hard " solder," one that melts at a relatively
high temperature. But the word " autogenous," said to
have been introduced by the French, gives much trouble.
The word is from the Greek, and means self -generated.
Evidently considered applicable to the carbon-arc
process and oxy-acetylene process when joining thin
sheet metal, as the soldering rod was not requisite,. If the
term " autogenous " gives the conception of furnishing
or generating its own heat, then those who claim that arc
welding with the metallic electrode should be called
" autogenous soldering " are not without justification.
On the other hand, if the conception of metallic arc
welding resides in the electrical view that the heat is pro-
duced by the arc and that this metallic vapor is furnished
by an external source of energy, there can be nothing
" autogenous " about it. In the same sense this word is
held by the advocates of the oxy-acetylene process.

Forgetting terms for the moment, there is marked
distinction between the joining of metals by the two
electric methods. In the former, called " welding,"
pressure is used and the heat is locahzed by the resist-
ance to the flow of electric current ; in the latter no pres-
sure is applied and the heat is localized through the

180 SPOT AND ARC WELDING

action of the electric arc. These characteristics are fully
differentiated by the accepted nomenclature, the former '
termed generically resistance welding, and the latter arc
gelding. Of these genera, two species have been
selected for special investigation, respe;ctively, " spot
welding '' and " metallic arc welding."

Spot Welding. — There are very few theories to be
found for the effects produced by spot welding. This
is specially the case with the spot welding of heavy ma-
terials as the development in this line is of very recent
date. It is a fertile field of investigation and should be
undertaken while the process is on the threshold of in-
dustrial acceptance. Some work was done in the related
process of butt welding,* though the investigations were
cut short at an important point. The heat treatment of
steel has received careful study because the employment
of this metal is of necessity to the industries. It is well
known that the grain structure of steel can be materially
changed after being strained by annealing. It is an
every-day affair in manufacturing lines to thus remove
the harmful effects of cold-worked steel or to soften,
strengthen, or harden, rolled or forged steel in order to
adapt its physical properties to specific uses. But little
is known of the effects produced by strains at the inune-
diate moment of heating and this is the crucial question
connected with spot welding. Those who have watched
the practical o*peration of making spot welds question
the need of high pressures except for making good eon-
tact at the electrodes, and for holding the lapped pieces
firmly together. A poor contact between the pieces is sl

' " A Study of the Joining of Metals," J. A. Capp, Qenerbi Electric Be-
vieto, December, 1918.

THEORIES OF ELECTRIC WELDING 181

desirable condition because a high resistance at that point
localizes the heat. If an excessively high pressure is not
required, or is a disadvantage to the grain structure for
the right quality of weld, then the designer of spot-
welding apparatus would welcome the news, because the
pressures now employed for heavy steel sections work a
hardship on the electrodesT^^-^

There are two theories of spot welding that have re-
cently come to notice, though it should be remembered
while considering them that it has not been possible for
either of the investigators to enter into a consideration of
all the questions involved. The first investigation was
made with rather medium material, about ^-inch mild-
steel plates, and the deductions made were all of a metal-
lurgical cliiiracter. The second was made upon the
hurried request of the author who submitted two samples
of spot welds in ^-inch mild-steel plate. The second
investigation reduced the matter, at least for the pres-
ent, to the ordinary behavior of the structure of steel
under the effects of heat. The former thaorj^ for the
sake of exposition will be^ealled the metallurgical theory
and the latter the heat theory.

Metallurgical Theory. — ^This theory was advanced
by its author Mr. E. E. Thum,^ in an article entitled
" Electric Welds " which appeared in the September
15th issue of Chemical and Metallurgical Engineer-
ing, 1918. It is interesting not only as a metallurgical
observation but also as one of the first contributions to a
study of the dninges in the structures of mild steel when
submitted ^nultaneously to high temperatures and
great pressure.

* Assoc. Ed., Chemical and Metallurgical Engineering, N. Y.

182 SPOT AND ARC WELDING

Mr. Thum describes a comparative test of the
strength of a rivet as compared with that of a spot weld.
" Four 3 X 8-inch bars were cut from ^-inch stock
structural steel plate. Two of these were riveted to-
gether with two ^-inch rivets, one driven 1% inches
from either end ; while the other pair were spot welded
at corresponding points with a machine designed to give
a weld J/s inch in diameter. The bars were then laid flat
on end supports, bent through about 45 degrees by a
concentrated central load and sawed lengthwise through-
out, cutting through the centre of the connection. . . .
The bars riveted together slipped past one another,
shearing the rivets, while the bars welded together showed
absolutely no movements at the ends nor did a micro-
scopic examination of the weld show any indication of
plastic yielding."

It is to be noted that small blow holes were observed
in the weld which is explained on the basis that " the
sheets were taken from a stock pile and no attempt made
to clean them of rust or scale before welding. At the
time of welding little chance was given for any extrusion
of hot metal, owing to the continuous lateral stmport of
the heated area; in this manner any impurities which
originally existed on the surfaces of the plate would be
trapped and retained."

Although Mr. Thum observed a peculiar structure
" to a greater or less extent in all the spot- welded low-
carbon-steel plates," which was not noticeable in the
butt welds examined, he found this pieculiar structure
best developed " near the outer edge of the spot- welded
structural plates " used in the above experiment. The
structure referred to lies next to " spheroidal mass at

THEORIES OF ELECTRIC WELDING 183

the centre of the welds " and " grades into the iinaflFected
original stock less abruptly." The micrograph illustrat-
ing his article shows " parallel striations " and foliations
which he believes suggests pearlite " which it evidently
cannot be, since the original metal is ordinary struc-
tural steel, whose micro-section gives the typical hypo-
eutectoid appearance." He next suggests the resem-
blance to martensite, but puts this aside on the basis that
" low-carbon steel would not be expected to develop such
large quantities of martensite nor would martensite be
expected in separating zones of sorbite and pearlite."
He then considers a possible explanation on the assump-
tion of annealing twins, mechanical twins, X-bands or
slip bands, but these conditions under which the structure
forms eliminates all of these suggestions. This leads to
the question of the " suppression of the carbonaceous
areas so prominent in the centre of the weld and in the
original stock."

After tracing in detail the temperature in the various
ellipsoidal zones forming the spot weld and cleq,rly de-
lineating the growth and subsequent contraction of the
heated areas, he then forms his opinion which follows:
" The austenitic spheroid, being directly between the
dies of the welding machine, is under considerable com-
pressive stress, but is unable to flow any measurable dis-
tance owing to its uniform side supports by relatively
rigid metal. Extrusion of a fin as in a flash weld is evi-
dently impossible, but the highly-stressed crystals of
austenite develop their characteristic octahedral cleav-
age planes exactly similar to those accompanying the
surface Islip bands appearing in a polished surface after
its imderlying metal has been severely strained. On

184 SPOT AND ARC WELDmG

cooling through the transformation range, the austenite
tends to precipitate its excess ferrite — each crystal re-
jects the sorbite to its boundaries, hence the ferrite tends
to gather along the cleavage planes. Thus the central
position of the weld is largely of dark-etching troostite or
sorbite, but close examination shows a well-defined hair-
like precipitation of the excess ferrite fringing many of
the allotriomorphic crystalline boundaries, and nimierous
very fine, white, parallel striations crossing at 60 degrees
appear in one of the dark areas near the centre of the
original micrograph. Very definite ' fringes ' of larger
parallel ferrite needles extending inward from the grain
boundary are seen near the edge of the sorbitic zone.

*^ Just outside the darker-etching zone (in the region
of eutectiform structure) a peculiar conjunction of pres-
sing and temperature occurred. During welding the
temperature passed the transformation range, and the
pearlitic areas passed into austensite. Migration of car-
bide, to equalize the carbon contents of these original
austenitic crystals, must have been extraordinarily rapid.
The austenitic crystals were at the same time fractured
along their octahedral cleavage by the compressive stress
of the welding dies, exactly as indicated in the discussion
of the imderlying dark-etching areas, and on cooling,
after the electric current had been interrupted, each little
lamina bounded by the paraDel cleavage planes acted as
an independent crystal in expelling ferrite to its surfaces.
Hence the ferrite marks the sub-microscopic cleavage
planes ; or rather, the thinnest plate of eutectoid remains
at the nucleus of each crystalline lamina. After polish-
ing and etching, the eutectoid films etched dark as a
series of straidit lines crossing from boundary to bound-

THEORIES OF ELECTRIC WELDING 186

ary of the original crystalline entity. Evidently the
process of extruding the ferrite from the original aus-
tenite laminae was just completed when the cooling of
the zone in question prevented the further molecular
mobility necessary for the agglomeration of the thin
plates into balls of less superficial areas, which is the nor-
mal appearance of low-carbon steel."

He completes his interesting paper with a compari-
son of this " peculiar structure " to the appearance not
of the Widmanstattian structure as ordinarily illustrated,
" but the close-packed W-bands ... on a lower
magnification," and then states that, " Howe and Levy
observe the same general appearance in over-strained
and heated austenitic manganese steel, caused by the
same train of events, that is to say, first fracturing the
oFiginal austenitic crystals of the quenched metal along
their cleavage planes and thus forming a multitude of
new crystalline entities."

Heat Theory. — Reference to Table XII, Chapter
IV, giving^e results of tensile tests on thirty spots made
in one 10-foot length of lapped ^-inch structural-steel
plates will show that " spot No. 5 " was held for other
tests. The same will be noted for " spot No. 8 " in Table
XIV, Chapter IV. These samples were called respec-
tively " bad weld " and " good weld." They were sent
to Mr. S. W. Miller, proprietor of the Rochester Weld-
ing Works, who kindly examined them and prepared
the micrographs (Figs. 1, 2, and 3).

The " bad weld," No. 5, in Table XII, lay between
spots 4 and 6 which show an ultimate tensile of 19,700
and 28,600, respectively, hence the name given the
sample. The " good weld," No. 8, in Table XIV, lay

186 SPOT AND ARC WELDING

between spots 7 and 9, which show an ultimate tensile
of 64,700 and 52,600, respectively. This was considered
practically " a good weld," because these tests were to
convince those interested that uniform spots could be
made that would exceed the strength of a single rivet
shear of the size required for the stock material by ship
classification societies. Lloyd's for ^-inch steel plate
would require a J^-inch rivet which is calculated to shear
at 34,100 pounds per square inch. So the '' bad weld "
was only about half as good as the required rivet and the
" good weld " was about half again as good as the speci-
fied rivet. The mechanical pressure, the current, and
Voltage, were held approximately constant; thus main-
taining only two variables, time and condition of mate-
rials. As to the latter, the slag and mill scale was
undisturbed on the lapped surfaces of the plates, but the
surfaces next to the electrodes w'ere ground down with a
portable hand grinder. As the tables show, the " bad
weld " was given 18 seconds of both mechanical pressure
and current, and the " good weld " was given 25 seconds
of the same treatment. Although no chemical tests were
made to insure the exact composition of the ^-inch steel
plate, this material was taken from stock which was pur-
chased under the specification requirements of the
American Society for the Testing of Materials, which
usually conforms to the following analysis :

Sulphur 0.04 per centl

Carbon 0.22 per cent.

Phosphorus 0.01 per cent.

Manganese 0.41 per cent.

Figs. 1 and 2 show a section cut through the spot
welds and magnified to double the size. Mr. Miller's

THEORIES OP ELECTRIC WELDING 187

notes on Fig. 39 showing the " bad weld " are as follows:
" SmaU defects in centre probably due to shrinkage.
Grain at ' A * is very coarse, at ' B ' very fine, and at
* C ' coarser than original, but not bad. ' A ' and ' B '
are shown on continuous micrograph (Fig. 41), but ' C *
is not shown. ' D ' shows columnar grains due to high

Fn. 39.— Bid mid nugnified about twin. Snull delecti in ontn pmbabl; due lo ibrlnkiigb

A mhJ B sbowa in Fig- 41. D ihowa columnu gniaa due to hi^ t«np«nture and npid cDoling,
Grain at C equiaxed due to slower ooobag.

temperature and rapid cooling. Grains at ' C ' equiaxed
due to slower cooling. Referring to Fig. 40 showing
' good weld,' the notations are the same as in Fig. 39.
' E ' crack at end of weld probably due to shrinkage.
In course of time these might be dangerous. ' F ' spot
of oxide in plate not caused by welding, original defect.
Columnar grains ' D ' very clear in this specimen. Re-

188 SPOT AND AEC WELDING

ferring to Fig. 41 this is a continuous micrograph at 100
diameters taken from the edge of ' had weld ' (Fig. 39)
towards the centre. The ' good weld ' is the same except
in degree. The union in both welds is perfect."

Mr. Miller explains the development of the steel
structure as shown by these particular samples as fol-

Tio. 10.— Good weld DugnifiRl twkf. Gnin st A nry cohik; Hi B, rtr; fine: al C, couwr thu

origiaaJ. but wt bad. £, cnck at vud of weld probably du« to jhrinka^. Id ooune of lime these

might tie dangeiwia.

lows: " I have spent considerable time in examining the
structure of these welds under high power. I was very-
much interested in examining them in view of a letter
which I received from Mr. Thum, western editor^ of
Chemical and Metallurgical Engineering, in which he
claimed to have found in spot welds (made, I think, in
lighter material) a very peculiar structure which he

*Now Associate Editor.

THEORIES OP ELECTRIC WELDING 189

accounted for in a very peculiar way. I have been un-
able to find any evidence whatever of the structure to
which he refers, or anything resembling it.

" The changes in structure which occur in spot welds,
using the two you sent me as a basis for my statements,
are only such as would be expected from the heat condi-

190 SPOT AND ABC WELDING

tions under which the work is done. When the plates are
just touching, the heavy current heats up the spots that
are in contact, and the temperature becomes very high
locally, due to the resistance. The amount and extent
of this heat will vary, of course, between different welds,
but not seriously. The heat becomes less as the outside
of the plate is approached because of the conductivity of
the plate and the cooling effect of the cooler outside
layers. There are certain well-defined temperatures of
steel of any given carbon content at which changes in
structure take place when heated, these resultant struc-
tures depending on both the amount and time of appli-
cation of the heat.

"^ In the centre of a spot weld the temperature is very
high, probably near the fusing point. The cooling is
rapid and the result is a structure in the centre of the
spot weld inside of the ellipse in which the grains are
columnar and perpendicular to the outline of the ellipse.
This is just what occurs in electric arc welds and in the
cooling of steel castings in molds. In the case of a weld,
the long axes of the columnar grains are perpendicular
to the sides of the Vs. In a casting, they are perpen-
dicular to the sides of the mold. Further inside of the
ellipse the grains are coarse, due to the very high tem-
perature. Just outside of the ellipse the temperature
has not been high enough to permit of the formation of
columnar grains, but the grains are coarse; and due to
the comparatively high temperature and rapid cooling,
pearlite has not had time to form, so that the structure is
confused and the ferrite and cementite are rather
indiscriminately mixed.

" Still further out in the dark zone the heat has been

THEORIES OF ELECTRIC WELDING 191

high enough and the cooling rapid enough to produce
sorbite in the grains, surrounded by a fihn of ferrite.
This is also a typical structure in electric and oxy-
acetylene welds, and is simply another degree in the
transition stage from melted metal to normal steel. The
temperature in this sorbitic zone has been so high that
the grains have had time to grow. They are, therefore,
very large, although equiaxed and not columnar.

" Just beyond this zone is one in which the tempera-
ture has been just above the upper critical point, the
Acs point. This temperature was sufficient to refine the
grain as much as can be done by heat treatment alone.
This also is characteristic of electric and oxy-acetylene
welds. Outside of this last zone the material gradually
changes to that of the original metal unaltered by the
heat. As stated before, I can see no difference between
the structures in and around spot welds and those in oxy-
acetylene and electric welds except in degree.

" It seems to me that these statements are entirely in
accordance with the theory of heating, and I cannot find
any evidence so far that the pressure has anything to do
with the structure, although it is possible, as stated in niy
former letter, that it may have some effect. This, hbw-
ever, would require considerable investigation before
drawing conclusions. I think it can be safely said that
there is no reason to be apprehensive of the strength of
spot welds ; and the two that you sent me appear to be of
excellent quality, although probably the second one is
somewhat better than the first." *

It is to be expected that these two theories of s^ot

welding will lead to further inquiry into the nature of

— _ — . __ — . , — I I • t ,

* Personal letter from S. M. Miller, dated May 7, 1919.

192 SPOT AND ARC WELDING

the reaction which takes place in the steel structure. The
important practical deduction is clear that the process is
a sound one for joining heavy structural members.

Practical Aspects. — In this connection there were a
few points observed during the demonstration at Potts-
town that may aid in dispelhng certain preconceived
notions. On the other hand, they may lead to more
interesting doubt which often ends in greater benefit to
such a process.

The first point had to do with the size of spot mdi-.
cated after tensile pulling. It was generally held that
the size of spot had a great deal to do with its physical
strength. In the earliest tests attempts were made to
compute the diameter of the spot and reduce the ulti-
mate tensile strength to poimds per square inch of the
observed spot. This was reasonable in view of the method
pursued in the case of a riveted joint, but the punching
of the holes for riveting had removed good metal usually
replaced by one of poorer quality. In spot welding the
original metals are there, but changed in structure, and
logically the imion of the metals could be made continu-
ous by the proximity of the spot or to the overlapping of
same. It was conceded during these tests that the diam-
eter of the spot was merely an estimation depending
upon the observer. Allowing hberally for this inaccuracy
it will be seen from the tabulations in Chapter IV that
the diameters of spots so recorded do not consistently
indicate the strength characteristics of the union. The
data of Lloyd's tests (Tables XV, XVI, XVII, and
XVIII) in this respect gives composite estimates made
by several observers. The difficulty of judgment of the
spot size is complicated by the change in form of the

«••

THEORIES OF ELECTRIC WELDING 193

spot. As will be seen by comparing Figs. 1 and 2 the
impression is at times of a circle and then of an ellipse.
Further, the effect of the cooling process makes for more
confusion in determining the ring or boundary of the
adhesion. It is possible, also, that these lines of demarka-
tion may be attributed to the surface strains occasioned
by tensile pulling. A good example of the range of
ultimate tensile for constant observed size of spot is seen
in Table XV, Chapter IV. From spot No. 7 to spot
No. 20 the data gives the spot diameter as 11/16 inch.
The first four of these spots were made in 16 seconds,
the next ten were made in 12 seconds. The ultimate ten-
sile ranged from 18,600 pounds to 12,200 poimds.

The next point of interest, and often of surprise to
those who had not seen it, was the fracturing of the stock
material and the complete adherence of the union at the
spot. This is remarkable in the J^-inch plate samples as
shown by the results in Table XVII, Chapter IV. The
samples were visibly twisted in the Olsen testing ma-
chine and in the majority of cases the yielding of the
material surrounding the spot was clearly observable,
taking many seconds before complete rupture. In some
instances it was necessary to wait quite a time, and in
^ others to subject the test piece to further loading. The
implication is not unwarranted that the ultimate rupture
is more than safely removed from the initial yield point.
This statement should not be confused with the impres-
sion so often concluded that, because the stock material
fractures without the region of the weld, that the weld
is better than the stock material. As a matter of inves-
tigation this may be a serious defect of the process. The
important consideration is how far the application of

194 SPOT AND ARC WELDmG

heat and pressure may be employed so as to preserve
the integrity of the original structure and yet obtain a
tensile strength sufficient for practical purposes. Table
XVII, Chapter IV, shows a much higher uniform tensile
strength of spot welds than is required to compete with
the rivets required for a similar jointure.

The third and last point observed has to do with the
heat conductivity of the plate material and its relation to
electric conductivity. Unfortunately, this observation
was most clearly noticed in the last series of tests which
could not be extended with sufficient fulness to make
definite conclusions. In the earlier tests it was noticed
that the end spots in a continuous seam, if given an equal
time with the intermediate spots, would show an increased
ultimate tensile. If the end spot were given slightly
more time, the ultimate tensile was very much increased ;
and in many cases, if the time were less on the end spots,
the ultimate tensile was as good or better. This action
can be noted by reference to the tabulations in Chapter
IV. The first samples made with the object of adjust-
ing the machine were two narrow strips of steel with a
single spot. Undoubtedly, the results from this test
which fixed the adjustments and constants used in weld-
ing the ship's floors gave a wrong assumption which the
uniformity tests afterward corrected. There can be no
doubt that in making successive spot welds, either in
single, double, or triple rows, the temperature varia-
tions in the stock material bear a close relation to the
resultant tensile strength of any given spot. It is sug-
gested that, in line with the heat cycle of each individual
spot, the residual or recuperating structure of the stock
material may interfere with its natural changes. Con-

THEORIES OF ELECTRIC WELDING 195

versely, the heat cycle of an individual spot may, through
the action of conductivity, or convectivity, affect the spots
already made. During the tests it often happened that
work was interrupted and seams were left to cool all
night in a rather cold shop, but continuation of the work
did not indicate that such interruptions were detrimental
to uniformity.

In the last practical tests two and three rows of spots
were made in i/^-inch steel plates. The first samples were
made by continuous spots first on one row and then on
the next row, and the time was held constant for all spots.
As the imif ormity tests for this thickness of material in-
dicated good results at 15 seconds, this time interval was
adopted. The tensile pulling of all these samples were
far below expectations and a second attempt was
immediately made.

Fig. 42 shows the order in which the spots were now
made. It is to be noted that this order of spot put them
in line with the tensile pulling. The time intervals were
varied, i.e.^ in the two-row sample the end spots were
given 15 seconds and the intermediate spots 20 to 30
seconds; in the three-row sample the end* spots were
given 20 seconds and the intermediate 25 to 30 seconds.
Fig. 42 attempts roughly to show the relative size of
spots as observed on the samples after tensile pulling.
The Roman numerals are the strips as cut for the pulling
tests. Nos. I and VI of the two-row and Nos. I and VI
of the three-row samples, though given reduced time,
exceeded, except No. V of the two-row sample, all the
intermediate spots.

As speculation is only possible, it is assumed that, as
the metal is locally heated by the electric resistance

196

SPOT AND ARC WELDING

which increases with the temperature, the surround-
ing metal tends to conduct away the heat. The end
spots would have the advantage of no plate material or,

comparative: size of spotS'2 spots 6 UP

%/ me %s %/2 %/s /^

JT izr -^ar

COMPARATIVE SIZE OF SPOTS -3 SPOTS- 9" LAP

Fig. 42. — Comparative aze of spots — ^S spots = 9" lap.

at least, a very small area of conductance on one side and
a large area on the other side. The intermediate spots
would be in varying degrees circumscribed in this re-
spect. It will be noticed in Fig. 42 that the No. II strip,
which is a combination of spots 3 and 4, is not compa-

THEORIES OF ELECTRIC WELDING 197

rable in tensile strength to strip No. V which is a com-
bination of spots Nos. 9 and 10. These latter spots
being made after the entire piece of material was well
heated. It is observed in continuous spot welding that
the spot adjacent to the one being made glows to cherry-
red heat. Further, though not so common, a number of
spots in the vicinity of the weld being made glow at vary-
ing degrees. This observation would indicate that the
metal of the plate is conducting some of the electrical
energy from the electrode and doubtless aids in the results
of decreasing ultimate tensile as shown by these tests.

Arc Welding. — ^Whereas the theories of spot weld-
mg are scant, those for metallic arc welding are very
numerous. It is not to be inferred that numbers in this
case give finality to conclusions; on the contrary, the
light is very dim in the region of the weld, and besides
the phenomenon of the arc leads to various aspects of
the problem.

Metallurgical Views. — The deposit from the metallic-
arc electrode is defined and treated as an unannealed
steel casting. In the making of this casting the stock
materials joined play an important role, because the arc
must fuse the adjacent metal and mix with a portion of
it to form the weld. Accepted as a casting, the efforts
of the metallographists have been to discover why the
structure displays brittleness with good tensile strength
and what means may be taken to make this casting more
ductile. These are questions which when fully answered
will standardize materials and apparatus, and also re-
lieve the operator from much responsibility.

With the aid of the microscope, investigators have
found, upon the examination of the weld, whether by

198 SPOT AND ARC WELDING

fracture or by cutting, a structure which contains a large
number of lines or plates. The determination of the
composition of these plates has brought with it differ-
ences of opinion and modification of views which make
conclusions impossible. One specialist finds these plates
at the grain boimdaries as well as in the grains, and fur-
ther observations have caused him to believe that the
grains are bounded by thin films which are sufficiently
tenacious to preserve a high tensile strength in the weld,
yet will rupture under shock or alternating stress. His
observations aflSrm that crystallization cannot take place
through these films, but he is not ready to state what the
composition of the films may be. He suggests the im-
probability of their being cementite in the case of metallic-
electrode welds, " as the carbon is almost entirely burnt
out," ^ but he allows the possibility of iron oxide, or
nitride of iron.

Another expert concludes that the plates or lines
" are not cementite, or martensite, or any similar carbide
product, but most probably nitride of iron." ^ This con-
clusion was arrived at after careful investigation of a
specimen in which the deposite4 metal was analyzed and
showed a carbon content of 0.04 per cent, carbon.

Still another specialist calls these plates " Nitrogen
lines," showing " the presence of a considerable percent-
age of nitrogen, but their absence does not always mean
that nitrogen is not present."^ He states further:

6 (i

Path of. Rupture in Steel Fusion Welds," S. W. Miller, Trans. Amsr.
Inst, Mining Engineers, February, 1919.

« " Micro-structure of Iron Deposited by Electric Arc Welding," G. F.
Comstock, Amer, Inst, Mining Engrs,, Feb., 1919.

» " The Metallurgy of the Arc Weld," W. E. Rudder, General Electric
Review, December, 1918.

THEORIES OF ELECTRIC WELDING 199

" Nitrogen is one of the most effective elements for mak-
ing steel brittle. As little as 0.06 per cent, will reduce
the elongation on a 0.2-per-cent. carbon steel from 28
per cent, to 5 per cent. It is contained in regular steel
only in very small amounts, varying from 0.02 per cent,
in Bessemer steel to 0.005 per cent, in open-hearth.
Under ordinary conditions of fusion, nitrogen has little
effect upon iron, but under the conditions of the electric
arc the nitrogen becomes more active. This is probably
due to the formation and decomposition of nitrogen-
oxygen compounds with a consequent liberation of active
atomic nitrogen. The fact that these lines do appear in
welds made in nitrogen gas alone suggest that the oxy-
gen need not be present, the nitrogen molecule being
split up by the arc stream, or perhaps the iron vapor
combines directly with the nitrogen. There is some evi-
dence that it does."

This opinion leads to the subject of occluded gases,
i.e.y gases that are absorbed by steel. Theoretically, these
occluded gases would play some part in the reactions
that take place imder the intense heat of the arc vapor,
the temperature changes following the passage of the
arc, and the cooling effects produced by the conductivity
of the mass of steel acted upon. For this reason the arc
operator learns to localize his arc and thus prevent the
stock materials from attaining a high degree of heat.
Such a condition, if permitted, causes too slow cooling
which results in a coarse grain and probably traps gases
which react to cause brittleness in the finished weld. The
field of research for the comprehension of the nature
and action of occluded gases is as important as any of
the other investigations tending to a solution of the

200 SPOT AND ARC WELDING

•

questions which now cloud a complete knowledge of
this process.

An interesting set of experiments with the object of
analyzing and measuring the occluded gases in iron alloys
was performed by Gellert AUeman, professor of chem-
istry, Swarthmore College, Pa.^ The results of this in-
vestigation are so closely akin to the theories of arc
welding that it is recommended that those desiring fur-
ther information on this subject carefully consider them.
It is not possible to quote the whole article, but the con-
clusions that have a bearing upon the metallurgy of arc
welds are as follows: " (3) It appears that the gases p^re
evolved in the following order: Hydrogen is most
readily set free, carbon monoxide comes next, and nitro-
gen seems to be held most tenaciously.

" (4) Whether oxygen is the result of the decomposi-
tion of various oxides of iron or the disassociation of car-
bon monoxide or carbon dioxide has not been determined.

" (5) We have shown that ferrous alloys may occlude
relatively large volumes of gases — in some cases equal to
about two hundred times the volume of the metal.

"(6) We suggest that in addition to the ordinary
fimctions of metals like aluminuni, tungsten, chromium,
manganese, titanium, silicon, etc., when placed in fer-
rous alloys, these elements may act as a catalytic agent;
and either prevent the occlusion of large quantities of
gases or aid in the elimination of such gases at lower
temperatures than would ordinarily take place.

" (7) We have shown that the removal of gases from
ferrous alloys markedly changes the microstructure and
increases the density of the alloy."

* ** Occluded Gases in Ferrous Alloys," Gellert Alleman and C. J. Dar-
lington, Journal of The Franklin Institute, 1918.

THEORIES OF ELECTRIC WELDING 201

Arc Characteristics. — "The arc is the welder's essen-
tial tool. It functions to transform electrical into highly
concentrated thermal energy. At the terminals this
energy serves to melt the parent and filling metals : in
the stream it stabilizes the arc and surrounds the fluid
pencil metal with a protecting mantle of hot inert gases,
usually oxide. A short arc length obviously assures
greater protection to the transferred metal than a long
arc, as the path traversed by the incandescent metal is
shorter and the liability of convection currents disturb-
ing the protecting envelope less.

" The mechanism producing crater formation and
transference of metal from the pencil electrode to the
parent metal has not been definitely isolated. However,
it kppears reasonable to assume that the propulsive force
projecting the filling metal across the arc is largely ob-
tained from the rapid expansion of occluded gases and
vaporized metal, the impact of the conveyed gas, vapor,
and liquid on the molten surface of the parent metal
producing the familiar crater.

"It is well known that all metals absorb gas, and
that the quantity of gas occluded varies with the char-
acteristics, preparation, and exposure of the metals.
Iron has been analyzed containing, at atmospheric tem-
perature, as much as 50 volumes of gas. Upon forming
an arc the extreme end of the pencil electrode is partly
liquefied and partly vaporized and then conveyed across
the arc stream. The concentration of energy at this
terminal, approximating 1500 watts for a 150-ampere
welding current, produces a rapid increase in tempera-
ture, and therefore a corresponding increase in the ex-
pansion of the metallic liquid and vapor and absorbed

202 SPOT AND ARC WELDING

gas. This expansion would tend to follow the path of
least resistance which extends through the arc terminal
and into the arc stream. Its effect would be to produce
a metallic blast. The force of this stream has been ob-
served to vary with change in electrode analyses, char-
acter of occluded gas, current density, arc length, and
the use of direct or alternating current. An additional
expansive force is probably secured by the union of elec-
trode materials with occluded and atmospheric gases.
However, as the temperatures in the arc stream are too
high to permit the existence of such compounds their
formation must occur in the surrounding envelope with
the result that the force developed is partly absorbed in
scattering hot metal.

" The metallic blast is of particular interest because
it appears to. be the basis for overhead welding. How-
ever, to properly utilize this phenomena it is necessary
for the welder to maintain obviously a short arc and to
adjust welding conditions so that the electrode end im-
mediately below the arc terminal remains comparatively
cool. If a globule of molten solder is dropped on the
inclined surface of a cold plate, it will run off at a slower
rate than if it strikes an inclined hot surface, due to the
difference in strength of the film produced by both sur-
face tension and the congealing of the metal. Similarly,
if the electrode end is maintained at a high temperature,
the molten metal will run down the side of the electrode,
thereby greatly increasing the difficulty of utilizing the
vapor blast. Such heating of the electrode also causes
leakage of occluded gas through the hot wall with con-
sequent diminution of the force produced by gas expan-
sion at the arc terminal.

THEORIES OP ELECTRIC WELDING 20S

" Excessive electrode temperatures are usually ob-
tained as a result of repeated attempts to start the arc.
When a number of false starts are made increased heating
results, due to the greater current flowing with the arc
short-circuited. If the operator develops the necessary-
skill to deposit metal after but a single start, the pencil
electrode end will remain quite cool, facilitating thereby
the continued transference of metals to an overhead weld.
Other aids to the maintenance of the proper electrode
temperature are :

" 1. Use of a lower current density than that em-
ployed for flat or downward deposition.

" 2. Use of an electrode having a melting point
higher than that of the parent metal.

" 3. Use of a direct-current supply circuit having
such characteristics that the arc short-circuit current
does not exceed greatly the operating current.

" 4. Use of a thin coating on electrode to facilitate
starting the arc." '

Physical Views. — In general the theory is accepted
that after the metallic arc is struck a vaporous stream of
metal from the electrode is formed. Surrounding this
vaporous stream is a clearly-Wsible flame indicative of
ordinary combustion and hence considered " a flame of
oxides." An observer working upon the problems of
automatic arc welding holds this theory: " As the result
of thousands of observations of welds produced auto-
matically (wherein the personal equation is entirely
eliminated) , the writer ^° inclines towards the theory that

•Communicated to the Author by O. H. Escholz, Oct- 4, 1919. See
Electrical World, Sept. 27, 1919, et seq,

"*" Discussion on Welding Mild Steel," Harry D. Morton, Transactions
American Sac. Mining Engineers, 1919.

204 SK)T AND ARC WELDING

the molten electrode material passes through the^ arc in
the form of globules, and that where ^-inch electrode
material is employed with a current of about 150 amperes
these globules are deposited at the rate of approximately
two per second. The passage through the arc of each
globule apparently constitutes a specific cause of insta-
bility in addition to those existent with slowly-consumed
electrodes. This hypothesis seems to be borne out by
ammeter records, together with the fact that the elec-
trode fuses at the rate of about 0.2 inch per second.
Moreover, the globules appear to be approximately
equal in volume to a piece of wire 0.125 inch in diameter
and 0.1 inch long."

In the same article this last observer gives this vivid
picture of the arc at work: "What seems to occur is
that the molten metal in the crater is in a state of violent
surging, suggestive of a small lake lashed by a terrific
«torm. The waves are dashed against the sides of the
crater, where the molten metal of which they are com-
posed quickly solidifies. The surgings do not seem to
synchronize with, nor to be caused by, the f alUng of the
globules of molten metal into the crater, but seem rather
to be continuous. They give the impression that the
molten metal is subjected to an action arising from the
disturbance of some powerful force associated with the
arc — such, for instance, as might result from the violent
distortion of a strong magnetic field. Altogether, the
crater phenomena are very impressive; and the writer
hopes ere long to be able to have motion pictures made
which, when enlarged, should not only afford material
for the most fascinating study, but also throw light upon
some of the mysterious happenings in the arc."

THEORIES OF ELECTRIC WELDING 205

Other opinions have to do with a study of the spec-
trum of the metallic arc in which department of research
apparently little work has been done. Such investiga-
tions would require special apparatus whereby the arc
would be not only observed under varying electrical con-
ditions, but also recorded in small increments of time.
Following one set of electrodes and stock materials many
combinations would need examination before a complete
spectroscopic theory could be approached.

Electrical Views. — Much of the electrical theory is
concerned with voltage drop. It has been well stated
that " up to the present ho satisfactory proof of any
theory has been put forward. It is by some ascribed to
a back e.m.f . produced at the point of volatilization of the
carbon and by others to the energy absorption required
by this volatilization." In view of this opinion, it is in-
teresting to note the experience of a careful observer as
to the action of the welding voltage. The results of his
investigations are given in full :

" After careful compilation of about five-hundred
readings of voltage across the arc and current through
it, I was forced to the conclusion that the current does
not affect the voltage enough for an ordinary meter to
notice it. With an oscillograph the rapid variations in
voltage due to changes in the current can be traced ; but
the voltage as shown by an ordinary meter, no matter
how delicate, is constant for any set of given conditions.
The conditions which do change the volts across the arc
are : First, length of arc ; second, type of electrode ; third,
gases in the arc such as would appear from coating on
the electrode or flux used on the job.

" There is a voltage below which an arc cannot be

206 SPOT AND ARC WELDING

held. And without starting an argument, as to whether
this is a counter e.m.f., a C. R. drop, or a combination
of the two, we are calling it a minimum arc voltage.
This lies between 10 and 11 volts. There is added to
that always a constant drop, depending on one of the
three conditions named above, and this drop varies from
one to two volts in bare wire at ordinary welding cur-
rents to 15 to 20 volts for heavily-coated wires. There
is the added resultant due to the average value of the
superimposed guardian, or puncture voltages, the peaks
of which are shown only on the oscillograph, but the
average value, of course, adds to the other two, making
the voltage across the arc as we have found it to be as
follows: For bare wire from 11 or 12 volts to 22 volts,
the lower value being an inhumanly steady operator
holding the closest arc possible with a wire that has no
carbon content nor any covering or jBux. The other end
of the scale is the carbon arc, about which much has been
pubhshed, which voltage varies from 40 to 55. The
highest metallic-arc voltage is tljat of a very heavily-
coated electrode, such as the English Quasi- Arc. The
voltage across this varies from 27 to 40. The voltage
across a completely-coated gaseous-flux electrode is from
22 to 35. The voltage across a half-coated electrode
varies from 15 to 30, the variations being as aforestated,
the sum of the variable-length voltage to the constant
ordinary resistance drop and the necessary voltage. The
only voltage that is variable is that due to the man's hand
in holding different lengths of arc.

" I think it is admitted by now that the short are is
desirable ; in fact, a short arc is absolutely necessary for
good work. With a short arc there is less chance of the

THEORIES OF ELECTRIC WELDING 207

metal being oxidized, less chance that it will fall on cold
work, less chance that nitrogen will be raised to such a
temperature that it will combine with carbon and steel
to form harmful compounds, and less chance of oxida-
tion. And, a point that has probably not been mentioned
before, the total heat is kept within reasonable control ;
namely, every welder knows that with a short arc he has
the metal under control, but with a long arc the heat is
raised by the voltage having risen across it. If there is
any decrease in the current and the temperature, it rises
imtil the steel is ' wild.' And, whereas the transition
from a short arc to a long arc is easy, the transition
backwards is very hard, and the tendency is, once having
started the long arc, to hold it at least until the end of the
electrode. Unfortunately, when no check is made on the
quality of the metal deposited, the long arc being easier
held — depositing the metal faster — is used ; and in some
rare cases, such as filling in castings, where the mass of
metal is great enough to satisfactorily dispose of the in-
creased rate of heat, it can be used to advantage.

" There has been much promotion of constant-
current apparatus, and equal promotion of constant volt-
age, but it can hardly be denied that a constant rate of
heat is what is desired. For a necessary change in length
of arc due to various physical imperfections in the circuit
or the electrode, this change being within the working
range between too long an arc and too short an arc, there
is considerable variation allowable — we would say }i
inch to 3/16 inch;> whereas anything over 3/16 inch
begins to be too long an arc, and rapidly tends to be-
come ^ or ^ inch long, in which case no welding can be
done, the metal of the electrode melting so much more

208 SPOT AND ARC WELDING

rapidly than the work that there is not crater enough in
the work to receive the electrode metal; and, unless the
crater in the work is as large or larger than the deposited
electrode metal, no welding can result. For instance,
steel cannot be welded by pouring molten steel on it at
any temperature except molten.

" The conditions, then, for good welding are: First,
control of the arc length ; second, constant rate of heat ;
third, a good operator ; because, with the first two condi-
tions ideally realized, we are still at the mercy of the man
that he guide the arc to weave the desired joint together.

" In limiting the voltage across the arc some appa-
ratus has been equipped with relays which cut the arc
out, or cut resistance in, or give other notifications that
the arc is too long. Other efforts to limit the voltage of
the arc have been made by reducing the open-circuit
voltage or guardian voltage until only a certain length
arc could be held. It is interesting to note here that an
arc cannot be held with a supply at the voltage across
it; in fact, the values begin to be twice the voltage of the
arc before any arc at all can be held. This is true of
alternating current or direct current. The minimum
voltage at which an arc can be held is aroimd 31 or 32
volts and this is obtained with bare wire and with no flux
or coating. While the gases of coated wires help to hold
the arc, they also raise the necessary voltage, and hence
raise or hold about the same the minimum voltage at
which an arc can be held. With just as low an open-
circuit voltage, an arc can be held on alternating current
as on direct current. With alternating current the
guardian voltage is supplied partially by inductive
kicks or transformer characteristics, and the normal

THEORIES OF ELECTRIC WELDING 209

open-circuit voltage can be lessened. On direct current
a reactance gives a like action, but in much less degree,
naturally the average voltage across the arc showing the
effect of these also, but the open-circuit voltage, at least
the quite open-circuit voltage, does not show it. The
only method so far discovered of limiting the length of
arc, without moving parts and without losing the neces-
sary open-circuit or guardian voltages for puncturing
through dirt, oil, slag, and giving good penetration,
is by the alternating-current special transformer for
arc welding." ^^

No satisfactory technical explanation has been offered
for the ability of the arc to deposit metal from below.
That is the phenomenon of overhead welding. The
same expert, fully cognizant of the theories put for-
ward, looks upon the matter in this common-sense way:
" Overhead welding depends neither on the machine nor
electrode, except in the case of co/npletely-coverefl elec-
trodes where a special covering, thinner and harder, i^.^
freezing more quickly, is resorted to in order to keep the
metal and slag from dropping. Overhead welding is
not a function of polarity, or the metal being carried in
one direction by the flow of current, but simply a case of
capillary attraction of the molten parent metal for the
molten electrode metal, and the electrode must be moved
ahead at a steady constant rate, so that but the equivalent
of one drop is molten at a time, and this drop draws up
into the parent metal instead of dropping. For instance,
one drop of water will cling to the ceiling, but more than
that will fall, leaving one drop remaining. A very close,
but not too close, arc must be held for overhead welding,

" Communicated to the Author by C. J. Holslag, Septemjber, 1919.
14

210 SPOT AND ARC WELDING

and the voltage is then reduced, and hence the current
must be increased to give the requisite amount of heat.
On the alternating-current machine this change is auto-
matic and inherent. On other systems the current can
be arbitrarily increased. On alternating current, also,
the progress must be faster, as more electrode is melted
with given conditions."

The Physical Behavior of the Welding Arc.^^ — "The
phenomena of the electric arc furnish an interesting field
for careful study. The production in so restricted a re-
gion of a temperature so high as to volatilize any known
substance is truly marvellous. The means by wjiich the
temperature is automatically held and regulated is still
more wonderful. Before the days of electric welding the
remarkable temperature effects of the arc furnished the
chief interest. The art of electric welding raises a new
question: the manner of transference of the material
through the arc. On this point, the experiments of
the joint Research Committee of the National
Research Council and the Electric- Welding Committee
of the Emergency Fleet Corporation have yielded
considerable information.

"1. Data culled by a sub-committee on the physics of
the arc from the experiments of the Research Committee,
together with results obtained by Professor R. G. Hud-
son, of Massachusetts Institute of Technology, and
results obtained by Professor C. F. Hale, of Albany
Teachers' College, have shown conclusively that the mode
of transfer of material through the arc is totally different
from the mode of transport through liquid electrolytes.

" The amount of material transferred through the arc

" Conrniunlcated to the author l^ Prof. Allison W. Slocum, University of
Vermont, November, 1919.

THEORIES OF ELECTRIC WELDING 211

was in some cases three times as great, and in other cases
only one-thousandth part as much as would be trans-
ported thrpugh an electrolyte by Faraday's laws.

" ' Positive and negative ions/ as used by Faraday,
have no meaning in the phenomena of the electric arc.

" 2. In the electric arc there is a potential difference
between the electrodes which may be measured with a
voltmeter. This potential difference proves the exist-
ence of an electrostatic field between the electrodes and
a consequent electrostatic pull upon the surface of the
electrodes tending to remove the molten metal. A simple
calculation of the magnitude of this force shows that it
cannot be rehed upon to transfer the metal across the
arc to the plate. It could not even remove the metal
from the electrode against the force of surface tension
alone unless the surface be so near the boiling point that
the surface tension becomes practically negligible. By
no possibility could it remove the metal and carry it up
against gravity as in the case of overhead welding.

" Moreover, when the temperature of the surface is
close to the boiling point of the metal, the electrostatic
pull would draw the metal from the surface in threads
like glass wool. This is the probable explanation of the
sharp point left on the electrodes which have been used
for welding with alternating current.

" Professor R. G. Hudson has proposed a theory in
which he avoids reliance on the inadequate electric force
by supposing the formation of a gas in the electrode
beneath the surface which by its expansive pressure ex-
plosively propels the molten metal across the arc gap.
This theory is published in the Electric Welding Journal
for November, 1919.

212 • SPOT AND ARC WELDING

" 3. Though the electrostatic field of the electromo-
tive force between the terminals of the arc is ineffective
in pulling metal from the electrode, it is, indeed, the all-
important factor in the heat-producing effect of the elec-
tric current- Electricity produces heat only while
moving in the direction of an electric force. This is
true whether electricity behaves like an incompressible
fluid in its flow or flows like a stream of electrons

or thermions.

" The fall of potential, which exists between the ter-
minals of the electrodes, occurs chiefly at the surfaces of
the electrodes. This ' electrode fall ' is the most impor-
tant feature of the behavior of the arc. For it is here that
the heat is chiefly produced.

" This electrode fall is in part conditioned by the elec-
trostatic force required to pull the electrons or thermions
from the metal of the electrodes. At very high tempera-
tures thermions are spontaneously emitted in copious
streams. This.phenomenon is well known in the behavior
of tungsten filament lamps. At such high temperatures
the electrode fall is small, a matter of a few volts. At
lower temperatures the electrode fall is larger. It is very
important to note that the electrode fall diminishes with
rising temperature and that in the neighborhood of the
boiling point of such metals as iron it changes very
rapidly with relatively slight changes of temperature.
It is this behavior of the electrode fall that gives to the
electric arc its extremely stable automatic regulation of
the temperature of the surfaces of the electrodes regard-
less of the conditions behind the surfaces.

" The rate of heat production is proportional to the
product of the current strength and the electrode fall.

THEORIES OP ELECTRIC WELDING 213

When the arc is playing steadily the rate of heat produc-
tion equals the rate of heat dissipation. If any change
occurs which tends to increase the rate of dissipation, this
change instantly tends to cool the surface. The ten-
dency to cool the surface is rapidly checked by the in-
crease of the electrode fall and the consequent increase
of heat production at the surface.

"4. When an electric current is flowing through an
arc it is surrounded by so-called lines of force which be-
have in some respects as stretched elastic bands exerting
a normal pressure inward upon the surface and through-
out the interior of the material of the arc. This effect is
known as the pinch action of the current upon itself. A
calculation of its magnitude in the case of a welding arc
shows that its pressure is comparable to that of the sur-
face tension of a stream of water flowing slowly in a long
thread from a faucet. The regulative influence of the
surface tension on the stream of water is small but ap-
parent. The regulative effect of such a pressure on a
stream of vapor of one-thousandth part of the density of
water would be one thousand times as great.

" The pinch action of the current may be considered
as having the effect of a kind of tube of considerable
stabihty through which the material of the arc is flowing.

"^ In accordance with the considerations stated above
and the general data reported to the Research Com-
mittee by various experimenters, the sub-committee on
the physics of the arc suggested a vapor theory of the
electric arc. Strong evidence for this theory is
constantly accumulating.^^

** Hagenbach and Landbein (Archives des Sciences^ Jan.-Feb., 1919) have
shown that for current intensities, not too small, the anodes of metallic arcs
( Ag. Cu. Fe. Na. W.) are heated at the tip to the temperature of ebullition.

214 SPOT AND ARC WELDING

" According to the vapor theory of the arc, its be-
havior consists in a process of boiling of the surface of
the electrode; the transfer of the vapor through a kind
of tube furnished by the pinch action of the current ; and
the condensation of the vapor on the surface of the plate.
In the interior of the pinch-action tube there is a core of
pure iron vapor which flows through the core exactly as
steam flows through the pipes from the boiler to the
radiator in a steam-heating plant after the air has been
completely expelled.

" In the case of the boiler, the heat is supplied
through the bottom of the boiler and causes the boiling
commotion throughout the volume of water. In the case
of the arc the heat is applied to the boiling surface and
the boiling commotion, when it exists, occurs only in the
regions immediately beneath the surface. This is the
phenomenon of the spluttering of the arc.

" The boiling of the surface is accompanied with the
absorption of heat as latent heat of the vapor. This heat
is again set free when the vapor condenses upon the
plate. The flow of vapor across the arc is limited by the
rate at which the conductivity of the plate can take away
the heat liberated by condensation.

" The temperatures of the surfaces of electrode and
plate are automatically regulated to approximately the
boiling point of the metal by the peculiar action of the
electrode fall. The electrode fall at the plate is regu-
lated by the force necessary to supply the positive charges
to the arc. The electrode fall at the electrode is regu-
lated by the force required to supply the negative
charges to the arc. And both electrode and plate
are automatically held steadily at the right temperature
for the purpose.

THEORIES OF ELECTRIC WELDING 215

" The vapor behavior of the arc suggests the follow-
ing rule for welding: Choose the largest current per-
mitted by hub conductance of the plate and choose the
smallest electrode that will carry the chosen current.

" The vapor theory is suggested as a tentative theory.
Whether it will completely stand the rough and tumble
of careful experimenting remains to be seen. In any
case, it serves to direct the attention of the welder to the
most important factor of successful welding — the heat
distribution and the temperature gradients in the neigh-
borhood of the welds."

Practical Aspects. — In reviewing the work of spe-
cialists who have investigated the science of arc welding,
one very encouraging opinion stands out clearly, that up
to the present time no obstacles have been observed that
would deter the use of the process for the joining of
heavy structural members. The general caution always
exercised by technical men centres on the skill of the
operator. As previously shown, this responsibihty may
be relieved by the future development of automatic arc
welding. Until that time it will be the duty of those in
charge of important electric-welding progress to select
and train the best men .as arc welders. Practically, the
art of arc welding reduces to the man who makes the
weld, but this man must have behind him the work and
encouragement which comes from those whose advocacy
of the process is guided by common sense and sincerity.

For purely practical purposes a test that may prove
interesting was made at the Pottstown demonstration to
show the effect of the combination of spot and metallic-
arc welding. Fig. 43 shows the four sample pieces.
No. 1 was a single spot weld made in 12 seconds with the

216

SPOT AND ARC WELDING

full capacity of the 27-inch portable spot welder. No. 2
was a lap-joint arc welded with direct current on both
laps and made in three layers. Nos. 3 and 4 were similar,
but the order of the welding was reversed, Le.j No. 3
-n- y / ;a sample was arc welded

S/N6LB SPOT Yz PLATE.

IZ SEC. AMP, 3/200 /8T50 LBS.

UWfMATS T£NSiLE LBS, 46500

ARC WELD r- 3 LAYERS -D, C,
UmMATE TENSILE IBS. 106800

ARC WELD St SPOT WELD
APPRO, DIAM. SPOT '^/tS "
ULTfMATE TENSILE LBS. 106500

first and then spot
welded, and No. 4 was
first spot welded and then
arc welded. The results
would indicate little ad-
vantage of the combina-
tion, but the arc weld
parted at the joint under
an ultimate tensile which
slightly exceeded the ulti-
mate tensile in the com-
bination joint which broke
the plate material and left
the joint intact. The ulti-
mate tensile for each
sample follows:

Ab.4^

No.

1..

. . 46,500

No.

2..

. . 106,800

No.

3..

. . 106,500

No.

4..

. . 106,500

SPOT WELD ic ARC WEED
APPRO. DIAM. SPOT 1 916 "
ULTIMATE TENSILE LBS,l0650O

Fig. 48. — ^Ezi)eriment in combination of arc and
spot welding.

Nos. 3 and 4 joints
were afterwards pried
open with a wedge in or-
der to examine and measure the size of spot. In the
former this was estimated to be 13/16 of an inch, and in
the latter 1 5/16 of an inch in diameter.

THEORIES OF ELECTRIC WELDING 217

Summary. — The differences of opinion of tech-
nicians might make the practitioner experience a feeling
of doubt. This is a natural stage in the develop-
ment of any art or practice, and must be so considered in
discussing and applying the theories advanced. The only
condemnation is for those who obstruct advancement for
the sake of gain, or who enhance some trivial character-
istic of the process to the detriment of a better imder-
standing of the whole. The work of theorists is mainly
to find defects in order to cure them and, with the instru-
mentalities now available, it is certain to follow that the
results of their work will advance the practice to a higher
state of efficiency.

APPENDIX

APPENDIX I

The Classification Societies Have So Far Considered and
Approved of the Application op £lectric Welding to the
Following Parts of Vessels:

Deck-Rail Stanchions to Plating

Clips for Detachable-Rail Stanchions

Continuous-Railing Rods (Joints)

Attaching Deck Collar (L Rings) around Ventilators

Attaching Deck Collar (L Rings) around Smokestack

Attaching Cape Rings around Smokestack^ Pipes^ etc

Attaching Galley Fixtures to Plating

Attaching Bath and other Fixtures in Officers' Quarters

Attaching Cowl-Supporting Rings to Ventilators

Bulwark Rail Top Splicing and End Fittings

Skylights over Galley

(a) Engine-room Stairs and Gratings

(6) Boiler-room Stairs and Gratings

Attaching (a) and (6) to Plating Grab Rods on Casing
All Stairs and Ladders, including Rail Attachments
Door Frames to Casing, Hinges, Catches, Holds, Coach-hooks, etc.
Clips for Attaching Interior Wood Finish to Casing
Entire Screen Bhd
Also Coal Chutes
Butts of W.Tv and Q.T. Boundary Bars on Bulkheads or Floors in

Double Bottom
Ventilator Cowls
Stacks and Uptakes
Bulkheads (that are not structural parts of the ship), partition

bulkheads in accommodation
^18

APPENDIX 219

Framing and Supports for Engine and Boiler-room' Flooring or

Gratings
Cargo Batten Cleats
Tanks (that are not structural parts)
Shaft Alley Escapes

Steel Skylights over Accommodation Spaces
Engine-room Skylights

Grab Rods on exterior and interior of Deck Houses
Deck Houses not covering unprotected openings through weather

decks
Reinforcing and protecting angles round manholes
Joints of W.T. Angle Collars at frames in way of W.T. Flats
Other parts of a vessel in which electric welding is proposed must be

submitted for consideration.

March 25th, 1918.
For Lloyd's Register of Shipping, J. French
For American Bureau of Shippings Geo. G. Sharp

"-A

APPENDIX II

Electrode Material for Metallic Arc Welding *

Welding Circular No, 1,

Chicago^ August 1, 1919.
To All Concerned:

It is intended to standardize our materials used for metallic arc
welding. A copy of our specifications for welding materials is
attached herewith.

You will note that there are now five different electrodes shown
as Rock Island Nos. 1, 2, 3, 4 and 5. Any additional electrodes
which it may become necessary to purchase and use in the future
will be given a Rock Island number, and supplements covering same
issued to all concerned.

It is desired that Mr. Sedwick see that each lot purchased meets
the specification, except for the actual flowing and weldability of
the metal, which will have to be determined by an expert welding
operator who shall be designated by Mr. Wanamaker or Mr. Penning-
ton — preferably an operator at Silvis.

When the material is O.K.'d, it can then be placed in Silvis
stock, it being assumed that all metallic-arc-welding electrodes will
be delivered to the Silvis Store and distributed from that point.

The electrodes should come in boxes plainly marked, showing the
Rock Island number and kind of material.

The Store Department should prevent any confusion or mixing of
the different specifications or analysis of electrodes, preferably by
having their racks or bins divided into different sections, one section
for each electrode number. This would hold true not only for the
Silvis Store, but all stores where metallic-welding electrodes
are handled.

* Courtesy of E. Wanamaker, E. E., Chicago, Rock Island & Pacific
Railway.

220 i

APPENDIX 221

These electrodes, at least for the present, will all be received
" bare," as called for in the specification. At Silvis, a sufficient
percentage of the Rock Island No. 1 electrodes will be coated to
meet the demand from the different points. All of the Nbs. 2, 3, 4
and 5 electrodes will be coated.

The coating specifications will be furnished by circular letter,
detail instructions regarding same to be furnished by Mr. Wana-
maker or Mr. Pennington.

All electrodes which are coated Jat Silvis shall be bound into
l6-lb. bundles — each bundle to be tagged, showing the Rock Island
number and kind of electrode material.

The electrodes may be coated on shop order and returned to
stock, care being used to see that the different numbers do not be-
come mixed or confused.

Copies of all circulars to date affecting the handling of coated
electrodes are attached herewith.

W. J. TOLLERTON.
SPECIFICATIONS FOR ELECTRODES FOR METALLIC ARC WELDING.

The following specifications to be used for purchasing electrode
material for metallic arc welding:

Rock Island No. 1 — Mild Steel — electrodes, to be used for all
ordinary purposes.

Chemical Composition:

Carbon Not over 0.18

Manganese Not over 0.55

Phosphorus Not over .05

Sulphur Not over .05

Silicon Not over .08

To be furnished in 3/32"— 1/8"— 5/32" and 3/l6" sizes.

Rock Island No. 2 — Medium-High-Carbon Steel — electrodes, to
be used for purposes where a medium-high-carbon-steel property is
desired — preferably for driving-wheel flanges or rail work.

222 APPENDIX

Chemical Composition:

Carbon — 0.65 - 0.75

Manganese — 0.60 - 0.90

Phosphorus — Not over .05

Sulphur — Not over .05

Silicon — Not over .08

To be furnished in 5/32" size.

Rock Island No. 3 — Nickel Steel — electrodes, to be used where
strength' and elasticity is desired for strength members, such as
frames, shafting, axles, etc., or any case where metal of such quality
is required

Chemical Composition :

Nickel — Not less than 1.50 or over 2.0

Carbon — Not less than 0.20 or over 0.50

Manganese — Not less than 0.28 or over 0.60

Phosphorus — Not to exceed .05

Sulphur — Not to exceed .05

Silicon — Not to exceed .08

To be furnished in 5/32" size.

Rock Island No. 4 — ^Manganese Steel — electrodes, to be used for
all hard-wearing surfaces, especially so where extreme toughness and
medium hardness are required, for instance — ^track steels, steam-
shovel dippers, dipper teeth, frame jaws, and any part of a machine
structural work, pressure vessel work, etc., such as would require
metal in the weld of extreme toughness.

Chemical Composition :

Manganese — 10.0 to 14.0

Carbon — 1.0 to 1.25

Phosphorus — Not to exceed .05

Sulphur — Not to exceed .05

Silicon — Not to exceed .08

To be furnished in 5/32" size.

APPENDIX 228

Rock Island No. 5 — Medium-Carbon Steel — electrodes, are pri-
marily of value for axles, forgings, piston rods, etc., or in any case
where sufficient carbon content is desired to limit abrasive wear.

Chemical Composition:

Manganese — 0.30 - 0.60

Carbon — 0.88 - 0.52

Phosphorus — 05

Sulphur — 05

Silicon — .08

To be furnished in 5/32" and 3/l6" sizes.

Material:

The material from which the wire is manufactured shall be made
by best-approved process.
Physical Properties:

Wire to be of uniform homogeneous structure, free from segre-
gation, oxides, pipes, seams, etc.
Test:

The commercial weldability of the No. 1 wire electrodes shall be
determined by means of tests by an experienced operator, who shall
demonstrate that the wire flows smoothly and evenly through the
arc without any detrimental phenomena.
Surface Finish:

To be absolutely free from oil and grease, and to have a dull, flat
finish, free from polish or scale.
Packing:

All wire shall be straight and cut to 14-inch lengths and shipped
preferably in boxes or kegs not exceeding 300 lbs. net weight. If
shipped in bundles, wrapping material must be free from oil or
grease. Each box, keg or bundle must be plainly marked, showing
the C. R. I. & P. number, together with the kind of material
and weight.

APPENDIX III *
Lesson I

THE ARC-WELDING MACHINE

It is important that the operator become familiar with the welding
machine before attempting to use the arc for welding operations.
Two drawings are reproduced showing the names of parts of the
welder set. It is not necessary for the operator to memorize the
names of the detail parts except that he should understand the loca-
tion and purpose of the essential parts as follows: — Brush, Brush-
holder, Commutator, Exciter Commutator, Field Coils, Motor, Exciter,
Grease Cup, Ball Bearings, Shaft, Bracket, Frame, Poles. Any
electrician can point out these parts on the welder set if the operator
is unable to do so.

The arc-welding generator is electrically separate from the motor
which drives it. A welding generator may be driven by either a direct-
current motor or an alternating-current motor or by a steam or gaso-
line engine. The source of power to drive the welding generator has
nothing whatever to do with the behavior of the welding generator;
provided of course, it is furnished in sufficient quantity and turns the
welding generator at the proper speed. The motor end of the welding
machine is like any other motor of the same rating.

The principle of operation of the welding generator is very simple
to the man who has had some experience with direct-current gener-
ators, but is difficult for any one else to understand. For the benefit
of the man who has had electrical experience, it is sufficient to state
that the welding generator is merely a specially-designed separately-
excited generator with a differential compound winding and that an
inductive ballast is used in the arc circuit. It is desirable for the
operator to understand the principle of operation of the welding set

* Courtesy of Lincoln Electric Co., Clevelatid, Ohiow

224

K--

as well as the electrician understands it, but it is not absolutely neces-
sary. The accompanying cut shows the volt-ampere characteristic
and the wiring diagram of the welding generator.

The welding outfit should always be installed by an electrician.
All cables are labelled and the direction of rotation is marked so that
no difficulty will be experienced in installing the outfit without the use
of a wiring diagram.

The stabilizer is made tap of coils of wire around a
laminated steel core and its purpose is to make tbe
arc steady and easy to operate.

An electrician should explain to the operator the
proper method of starting the outfit.

The control panel contains the apparatus with which
the operator controls the behavior of the welding gen-
erator, adjusting it to give the proper amount of heat
for welding. Two cuts are shown showing two types of
control panel used. The portable type accomplishes
the same thing as the stationary type. The voltmeter
and ammeter are left off the portable type on account
of the fact that they are too fragile to stand the rough
use to which they would be subjected on portable
equipment.

Fig, 47 shows the ordinary equipment used by the
operator, and welding table. Referring to Fig. 59,
the proper clothing for an operator is shown, — it con-
sists of black cap, unionalls, cotton gauntlet gloves,
"■""■. "■p^d."""' split-leather apron.

P MACHINE

1. Open main switch and control switch on panel,
S. Start welding set.

3. Turn rheostat as far as it will go to the left

4. Close control switch into position marked 100, (In this posi-
tion the current in the arc will be approximately 100 amperes.)

5. Put a piece of 5/32" welding wire in the metal-electrode holder.

APPENDIX 287

6, Place a piece of boiler-plate scrap on welding table to prac-

7- Cloae main switch on panel.

8. Sit down on stool in front of welding table. Take hand shield
in left hand, metal-electrode holder in right hand. With shield held
in front of- face, touch boiler plate with end of welding wire. The

Flo. *S.-PorUble p™*l.

result will be a spark and the welding wire will stick to the boiler
plate. Let go of electrode holder and open main switch on panel,

9- With a new piece of welding wire, and face shield in front of
face, scratch welding wire side wise on boiler plate to get spark, then
draw welding wire about an eighth of an inch away from the plate.
Hold welding wire vertical to boiler plate, otherwise arc will be diffi-
cult to start.

Repeat the above operation until an arc can be maintained as long
as desirable. The beginner should burn from 7S to 100 pieces of
welding wire at this practice, observing through the shield what hap-
pens in the arc. As the operator becomes, more skilful, he should try

«28 APPENDIX

. to hold a shorter arc. The proper length is about an eighth of an
inch. The operator should spend about 15 hours on this kind of
practice. The amount of current or amperes required for welding
depends principally upon the size welding wire used. Three-
sixteenths-inch welding wire requires about 160 amperes. (Turn

Pja. 47. — Opentor'i took.

rheostat as far to left as it will go and close control switch into 150-
ampere position.) For points in between 100 and 150 amperes, turn
rheostat to right, with control switch in 150-ampere position.

Lesson II

STARTING THE ABC

This exercise deals with the proper method of starting and stop-
ping an electric arc. The be^unner usually draws an arc and starts
to weld at whatever point the arc happens to start operating properly.
In other words, the beginner usually welds where it is possible for
him to weld rather than welding in a predetermined place. The pur-
pose of this exercise is to give the operator sufficient control of the
arc to enable him to weld at any place he may decide upon.

1. Place a piece of scrap boiler plate on the welding table. With
a piece of soapstone mark a line across the plato. Now weld & bead
as nearly as possible y^" to the right of this line. Make the bead as
straight as possible. Repeat this operation until a perfectly straight
bead 1/^" from the predetermined line can be laid down.

2. In this exercise the operator should print his initials on a piece

APPENDIX 229

of scrap boiler plate and weld a bead over the linea. Having produced
a perfect set of initials in this manner, take another piece of scrap
boiler plate and make the initials the same size without previously
printing them with soapstone. The operation should be repeated
until the operator can reproduce his initials without following the
lines. The purpose of this exercise is to train the operator to control

Pia. «.— Sample of Be»dB.

an arc and lead it in a predetermined direction. It also involves the
J training of the operator's eyes to see where he is leading the arc.
This will be difficult at first, owing to the fact that the operator can
see nothing but the arc itself through the protective glass.

3. The operator should now take hammer and chisel and examine
the beginning of several beads which he has made. It will be found
that the beginning of the bead is usually not securely welded to the
plate. This is due to the fact that the arc was held too long at the

2S0 APPENDIX

instant the bead was started. The operation of starting the arc at
the predetermined point should be repeated with this fact in view
until a satisfactory weld is made at the beginning of the bead.

4. The end of the bead is quite as important as its beginning. In
referring to beads which the beginner has previously made, it will be
found that a considerable crater has been left at the point at which
the arc was broken. The objection to this crater is that it is difficult
to start welding at this point when it is desirable to continue the
bead. The crater may be filled before the arc is finally broken by
merely crowding down the arc until the desired amount of metal is
added, and breaking the arc suddenly by pulling the wire sharply to
one side. The operator should practice this operation until he is able
to finish a bead, leaving a crater of not to exceed 3/1 6 of an inch
in diameter.

5. The exercises outlined in the preceding four paragraphs should
occupy at least ten hours of the operator's time. The following
sample is to be made as to the record of the operator's ability to start
and stop an arc properly :

Material required: one 12"xl2"xV2" piece of boiler plate; three
sizes of electrode are required— S/l6", 5/32", 1/8".

No marking with soapstone is to be done on the plate. Referring^
to the photograph reproduced herewith, the first three rows of bead's
arc to be made with 3/16" wire, using approximately 150 amperes.
Each bead should be one inch long. The beads should be three-quar-
ters of an inch apart. They should be straight and parallel. Each
bead should have a perfect weld at its start and a very small crater at
the finish. The next ^ve beads are to be made using 5/32" electrode
and the next two, using %" electrode, with about 125 and 100 amperes,
respectively. One side of the plate should be completely welded in
accordance with the above instructions. The plate should then be
turned over and the operation repeated and perfected on the other
side of the plate.

APPENDIX IV
Lbsbon III

BDILDINO-UP OPERATION

The purpose of this exercise is to show the operator the prop-^-r
method of building up several layers of welded material. It is as-
sumed that in Lesson II the operator has learned to deposit metal
from the welding wire on a piece of boiler plate and have it entirely
welded along the line of fusion. Until the operations outlined in
Lesson II are completely mastered, it is useless to proceed with the
exercise of building-up operations.

Material required: One 10"xl2"xl/^" piece of boiler plate. One

Fn. W.— Built-up pada of deposited metiil.

size of electrode, 5/32 of.an inch, is required. The current should be
about 125 amperes.

Referring'to the photograph reproduced herewith, three pads are
to be built up on the face of the plate. These pads are to be 6" long,
2" wide, l" high. The first pad starting from the left-hand side of
the plate is to be built up without any particular design or pattern,
and without brushing or cleaning of the oxide-covered surfaces.

The nert pad is to be built up following the definite pattern. First,
brush the spot on which the second pad is to be built very thoroughly
with a wire brush. Second, build up a single layer of metal the width

231

2S2 APPENDIX

of the pad, using a series of beads laid along the 6" dim
always starting at one end and finishing at the other end. Having
deposited the first layer, the oxide-eovered surfaces must be brushed
thoroughly with a wire brush. Eaeh layer should be brushed at least
three minutes. The seeond layer of the pad should be built up so
that the beads run at right angles to the beads of the first layer, i.e.,
the beads are parallel to the 2" dimension of the pad. This practice
is conuuonly called " lacing." The second layer to be as thoroughly
brushed as is required upon finishing the first layer. Each succeed-
ing layer should be thoroughly brushed.

The third pad is to be built up in exactly the same manner as the
second pad, with the exception Oiat in place of brushing the work
with the wire brush only between each layer, the oxide must be
entirely cleaned off by the use of the hammer and chisel. It will
be noted that the oxide may be removed by comparatively light
blows on the chisel. It is not necessary to cut away any metal to
knock the oxide from the top of the layer with a chisel. The wire
brush may be used to brush the oxide off the metal after it has been
cut away with a chisel.

The operator has now completed three pads. The first pad il-
lustrates how welding should not be done. The second pad illustrates
a fairly satisfactory practice. The third pad illustrates the best
practice. If possible, the operator should have this sample sawed

APPENDIX 23S

diagonally through the three pads. It should then be set up on a
grinding machine and a fine surface ground on the cut section of the
pads. This can be done in a tool room. The ground surface should
then be painted with diluted sulphuric acid or tincture of iodine. It
will then be easy to compare the quality of the metal in the three
pads. The operator should also observe carefully the line of fusion
between the pads and the original plate. This fusion must be per-
fect if the weld is of any value. The photograph reproduced here-
with illustrates the appearance of a good line of fusion.

Lesson IV

PLATE WELDING

This exercise is one of the most important of the series because
the welding of plate is the most frequent application of the electric
arc-welding process. The welds which must be made in structures
made of plate^ such as tanks^ are not always horizontal^ so that the
operator must learn to weld not only in the horizontal position but
also in the vertical and overhead positions. Three samples are to
be made as the record of the operator's ability to weld in the hori-
zontal position and the vertical and straight overhead positions.

Material required: Six 10"xl2"xl^" pieces of boiler plate
beveled 45 degrees on one 12" edge, 5/32" electrode with 125 to
150 amperes.

1. The operator should spend approximately ten hours in pre-
liminary practice. Several pieces of scrap boiler plate should be
beveled and tacked together as shown in the accompanying photo-
graph. These plates should then be set up vertically and welded,
starting at the bottom and welding up. The operator should use
his own resourcefulness in arriving at the best way to make a weld
in this position, trying several different methods and observing the
following points: Does the weld extend completely from the inner
to the outer edges of the plate ? Does the heating of the plate cause
sufficient expansion and contraction to affect. the character of the
weld? Does the expansion and contraction caused by the heating of

234 APPENDIX

the plate produce a warping or buckling P After the operator has
satialied himself on these points, two pieces of scrap hoUer plate
should be beveled and placed in position, ready to weld straight
overhead, and the operator should try to weld them together in this
position, welding from the under side only. The operator should put
the pieces approximately one-sixteenth of an inch apart for this
esercise. This kind of welding is very difficult and requires a con-

siderable amount of practice to master. It will be found that the
operation will be somewhat easier if 150 amperes is used on 5/S2"
electrode at first. In welding beveled plates, the operator should
remember that the welding wire or electrode should be held as nearly
perpendicular to the surface being welded as possible, and that good
melding can only be accomplished when a short arc it maintained.
The operator should pay particular attention to the difference in
sound between a lung arc and a short one. A long arc sputters
and has a distinct hissing sound. It is impossible to weld with such
an arc. A short arc has a rapid-fire metallic click which may be
readily distinguished. The operator should maintain a short arc on
all classes of welding. Where possible, an electrician should be asked

APPENDIX iSS

to connect a low-reading Toltmeter across the arc, so that the voltage
may be read while the operator is weldingi. The Toltmeter siiould
read fron^ 15 to 18 volts while the arc is in operation. The greatest
amount of heat b obtained on Ijie work when the electrode holder is

Tra. Al^ — Horizontal, r«rtical And overhead Aampln.

negative. This is the proper connection for both metal and carbon
electrode work.

While the arc is in operation, there will be a circular spot of
molten metal upon the work. The operator should concentrate his
attention upon the side of this molten spot of tnetal which is in the
direction of motion of the electrode. Tihis may also be described
as the forward edge of the circular spot. The arc should be directed
on this point, since it is at this point that the greatest amount of
heat is desirable. It i* possible to make an electric weld only when

236 APPENDIX

the globule of molten metal from the welding wire is thrown into
molten metal on the piece being welded. If the globule of metal
drops on metal which is not molten^ it may sticky but it will not be
welded. The operator should study the action of the metal in the
heat of the arc very carefully. The operator should begin to realize
at this point that merely holding an arc is not necessarily weldingi
but that the art of welding is 90 per cent, brain work and 10 per
cent, manual labor.

2. Place the horizontal sample of welding in position on the
welding table. Put a 5/32" electrode under each plate in a position
parallel to the beveled edges and about %" from the lower edge of
the bevel. This will raise the beveled edges higher than the square
edges and give the sample a ridge through the centre. The object
of this practice is to allow for the warping of the plates by the
heating of the arc. After the sample is welded it should be straight
with the two plates squarely in line. Place the edges % of an inch
apart all the way across. Tack the pieces together as shown in
Fig. 51. Now with 140 amperes and a 5/32". electrode, weld one
layer in the bottom of the bevel in about 3" sections. By this is
meant that the operator should weld 3 inches, skip 3 inches, weld
3 inches, skip, etc., until he has gone all of the way across the
plate, then go across the plate again, filling the three-inch gaps.
This is to minimize the effect of the heating. The plate will then
be welded with one layer all the way across. The operator must
manipulate the arc in such a manner as to weld the lower* edges of
the plate completely together, i.e., the metal from the electrode must
run clear through the plates and be firmly welded on the edges. The
operator should then take hammer and chisel and clean the oxide
from the surface of the welded metal very thoroughly. The second
layer may now be welded into the bevel, starting at one end and finish-
ing at the other end. This layer should be thin and .should not extend
higher than the upper surface of the plates. Chip oxide from sur-
face of welded material, and put the third and finishing layer on
the weld. The third layer should extend about 3/1 6 of an inch beyond
the edge of the bevel on each plate, and Vs" above the upper plate

APPENDIX 237

surfaces. The plate should now be turned over and a reenforce-
ment of equal width and thickness put on the other side. The pur-
pose of this practice is to make the section of the weld equal on
both sides of a centre line through the metal of the plate. If the
weld were reenforced on one side and not on the other the stress
would be concentrated on the side which was not reenforced when
the weld is put in tension.

S. The two plates should be tacked together as in first exercise,
but in this case the .beveled edges are to be set vertical, as shown

Fia. as. — Flfltes welded in vertical posiUon.

in Fig. S3. The weld is to be made according to a definite pattern,
starting at the bottom and finishing at the top. This pattern is
triangular. The operator should start on the right-hand plate at a
point of about 8/16 of an inch to the right of the beveled edge, hold-
ing the welding wire as nearly perpendicular as possible to the
snrfaee being welded. The movement should be along the beveled
edge of the right-hand plate toward the farther edge, then along
the beveled edge of the left-hand plate toward the nearer edge,
extending to a point 3/l6 of an inch ta the left of the bevel on the
left-hand plate, then across to the starting point. Five-thirty-sec-

238 APPENDIX

ond electrode with about 125 amperes is to be used. The operator
must pay particular attention to see that the farther edges of the
plates are securely welded together. A considerable amount of
metal should be run through the edges to make this certain.

4. For the sample of overhead welding, the plates may be tacked
together as shown previously, except that the opening should be
approximately % of an anch. The two plates are to be welded in
the overhead position after they have been tacked. Several pieces
of plate Ys of an inch thick, 1^" wide and 6" long are to be cut,
and a 3/32" electrode should be stuck on extreme edge of one of the
corners so that the electrode stands out perpendicular to the piece.
The purpose of the electrode is to serve as a handle. This %" piece
is to be pushed through quarter-inch opening between the plates
from the under side and to be brought into position so that it will
form a backing for the weld. Fig. 52 shows the position of this
plate. After the plate has been placed in position it may be tacked.
The use of this plate makes the overhead welding somewhat easier
than welding without its use. Start the overhead weld at the centre
of the job and weld toward one end. A definite pattern should be
followed. Start at the lower edge of the right-hand plate at a
point 3/1 6 of an inch to the right of the bevel. Continue along the
beveled edge of the right-hand plate up to the backing plate, across
the backing plate and down the beveled edge of the left-hand plate
to a point 3/1 6 of an inch to the left of the bevel. This will form the
first bead. Now start the second bead at the beveled edge of the
right-hand plate and on top of the first bead, and fill in, as far as
possible, the opening formed by the beveled edges of the plates. A
third bead will be required to complete this operation. The operator
now has two surfaces to weld on, the surface formed by the welding
material, which should be approximately vertical, and the surfaces
of the plates to be welded. The pattern of the first pad should be
followed out from this point on welding at the junction of the
previously-welded material, and the surfaces of the plates being
welded together so far as this is possible. This makes the weld
more a vertical weld than an overhead weld and considerably sim-

APPENDIX 2S9

plifies the operation. The operator should use about 150 amperes to
start with, cutting it down to 125 or less as the plate warms up.
Having completed one end of the weld in this manner, the other end
may be welded in exactly the same way. It will be found that the
backing plate will warp and tend to gjet out of contact with the
beveled plates. This will not interfere with the welding and will
enable the operator to reenforce the weld on the top side, which
is very desirable.

Lesson V

THIN-PLATE WELDING

This exercise is to give the operator some experience on thin^
plate welding. The difficulties encountered in thin-plate welding
are comparatively simple of solution, and the operator is left to use
his own resources to k considerable extent in making the sample.
The great difficulty in welding thin plate arises from the tendency
of the arc to burn through the thin plate, owing to the great inten-
sity of heat. Practically all thin plate is covered with a heavy scale
of blue oxide, and it is necessary to get this oxide cleaned off in
order to make a good weld. This may be done with hammer and
chisel or a sand-blast. The operator has already found that it is
necessary to have clean metal in order to make a good weld. The
quickest and best way of getting clean metal is to sand-blast the
surfaces to be welded. This applies to metal of all thicknesses.
The reason blue oxide gives the operator trouble is that it is a very
poor conductor of electricity, and it is hard to get the arc started
on an oxide-covered surface and also that the oxide gets into the
metal of the weld.

Material required : One piece of 24" x 30" sheet steel approxi-
mately l/l6 of an inch in thickness; l^" electrode with 90 to
100 amperes.

1. The operator should study the drawing reproduced (See Fig.
56) and lay out the pieces to be cut in order to make the sand-
blast pot shown. This will leave some scrap material around the

9M APPENDIX

edges which should be cut with a hack-saw into pieces approsimat«ly
2" I 4". The operator should practice welding these scrap pieces
by laying them down on the welding table and welding a straight
seam. One sample should also be welded with the two pieces per-
pendicular to each other as shown in accom-
panying cut. (Fig. 54.) Approximately two
hours should be spent on this practice.

3. The operator should now cut the plates
necessary to form the sand-blast pot and weld
them together. It is suggested that the heads
be made smaller than the shell so that they fit
on the inside. They should set back from the
edge of the shell about 14 "■ One small hole
should be burned through at the location of
Pio M— TwopiBMBweid^ '^^ °^ ^^ fittings in order to allow the heated

one perpendicular.

air to escape while the welding is being done. The fitting can be
put on the sand-blast pot at some later time by the operator.

This exercise is in the nature of a test of the ability of the
operator to make a solid homogeneous weld which is properly and
thoroughly done. A great many electric welds are subjected to
steam or water pressure and, unless they are properly made, they
will show leaks, and will fail at a point below the pressure for which
they were designed. It is very important that the operator should
know when he is making a good weld. If he does not know this, his
work is entirely worthless. He is as poor a workman as the jeweler
who must smash an expensive watch in order to find out how it was
made. A skilful operator, who has a reasonable degree of judg-
ment and intelligence, knows when he is making a good weld. If
he has made a section of a weld which is not good, he should either
cut that section out and reweld it or inform the man responsible

APPENDIX 241

for the job of the fact that a particular section is faulty. A man
who will lie to himself in regard to the quality xif his work will
lie to the man who is responsible for its quality, and is worse than
worthless as a skilled operator.

Material required: One 18" section of 8" wroug^it-iron pipe
or seamless tubing, two %"'thick boiler-plate heads to fit on the
inside of the pipe or tube. These heads should be beveled 49
degrees on the circumference, & pieces of l" black wrought-iron
pipe 6" long, one piece of %" or l" pipe according to the size

Fm. m.— Boiler flue welding.

water pipe used in the shop where the welding is done. This pipe
is to be connected to the water system so that the completed sample
may be tested under pressure. Six holes are to be drilled at inter-
vals of 2" into the 8" pipe to take the six l" pipes. One hole is to
be cut to take the %" or 1" pipe.

1. The heads are to be welded into the pipe as shown in the
accompanying cut. (Fig. 35.) The operator must be careful to
hold a short arc and so far as possible keep the electrode perpen-
dicular to the surface being welded. The surfaces which are to
be welded must be clean and the oxide must be removed from each
layer of metal before the nest layer is welded, by the use of sand-
blast or hammer and chisel. The l" pipes are spaced close enou^.

APPENDIX

MORSE 2 TfiPCR
1

NOZZLC

6L0NQ

^COUPtlNq

Fig. 56.

APPENDIX 243

together so that some difficulty will be experienced in making a good
weld between pipes. This is done purposely because it is a diffi-
culty frequently encountered in practice. The operator should
mark with chalk the spots where he believes, owing to the manner in
which he welded the sample, that the leaks will occur. Weld the
ends of the six l" pipes shut.

2. The operator should connect the sample to the water system
of the shop and test it for leakage. (It is advisable to pour the sample
full of water before the connection is made, so that it will be entirelv
filled with water when under pressure.) If leaks are found, the
operator should cut out that part of the weld, examine the weld
and find, if possible, the cause of the leak. The defective Sfiots
should be rewelded and the test repeated.

Lesson VII

MISCELLANEOUS JOBS

The object of this exercise is to give the operator an idea of
a few of the many different kinds of applications of the process.
A great deal depends upon the operator's natural resourcefulness
in planning a job. One of the difficulties is in knowing how to
go about a job so that it may be done with the least possible exer-
tion. The more highly skilled the operator is, the easier will be
the way which he chooses to perform the operation. This involves
careful planning of the operation before it is started. The operator
who cannot plan in advance exactly how he is going to do the job
will -have little success in doing it. As has been stated before,
success in welding depends more upon the use of the brain than
upon the use of the hand's. The operator should be able to tell
exactly how he proposes to do a certain job, and explain the reasons
why he intends to do the job in that particular way.

Material required: One riveted section as shown in Fig. 57, Qne
angle-iron section as shown in Fig. 58. These two samples need not
conform to any specified dimensions.

2M APPENDIX

1. For preliminary practice, the operatiw should take two pieces
of 1^" scrap boiler plate, and tack them together in the form of a
lap joint. This sample should then be set up in the vertical position
and a fillet welded on the wider side of the lap, similar to Fig. 57.
This operation should be repeated until the operator is able to get

a good weld and the fillet has a uniform appearance. The operator
should calculate the number of feet per hour of this work he can
do. This work is similar to the operation encountered in the weld-
ing of a caulking edge on the riveted seam of a steam boiler. It is
necessary to weld only one bead to form the fillet; 140 to 150 amperes
should be used. The operator should cut across the seam and

APPENDIX 245

the fillet to determine whether or not be has made a
good weld.

S. With a piece of scrap hoiler plate set id the vertical position,
the operator should weld a number of circular beads approKimately
xm' in diameter. After eight or ten of these circular beads have
been welded, the operator should clean the oxide from the surfaces,
and weld a second bead around the first bead. This is an operation
similar to that of welding around the head of a livet. One of
these circles should be cut and the weld examined to see that it
has been properly done and that the' second bead ia fused thoroughly
to the plate and to the first bead.
This is an operation which must be
thoroughly mastered before pro-
ceeding further.

3. This exercise consists of weld-
ing two pieces of heavy plate to-
gether without bevelling. If possible,
two pieces of %"-thick boiler plate
should be obtained for the exercise.
Each edge which is to be welded
should be set in a horizontal position

and a bead welded along the centre ^^- «.— Angle iron seoUon.

of the plate. The second bead should then be welded in top of the
first, removing the oside from the first before the second is applied.
When- both edges are thus prepared and put together, the operator
will have what amount to bevelled edges to weld together, but it will
be necessary to weld from both aides in order to complete the job.
One weld of this nature should be made and cut so that the operator
may examine it to see that fusion has taken place throughout the
entire weld,

4. This exercise is the one shown in the cut (Fig. 57) and con-
siats of welding the caulking edge of a riveted joint and welding
around the rivet head. The method of welding the caulking edge
has been previously explained. In welding around the rivet head
it is advisable to heat the rivet before welding around the head.

446 APPENDIX

With the plate in a vertical position (rivets above the caulking edge);
draw an arc on the head of the first rivet, allowing the metal from
the electrode to fall clear of the rivet head. This should be con-
tinued for about two minutes or until the rivet is thoroughly heated,
then the fillet should be welded around the rivet. The operator
should then skip two rivets and repeat the operation on the fourth
rivet. The idea of skipping rivets is to keep the heat distributed so
that contraction in the metal will not set up shearing stresses in the
rivets. By following the above practice, a very tight joint will
result when the metal of the rivets and plates cools. The result is
similar to the result obtained by putting in a hot rivet and peening
it over. When such a rivet cools, it contracts and pulls the plates
tightly together. The operator may turn the sample over and repeat
the operation on the other side, perfecting it, if possible.

5. The exercise of welding an angle-iron section is one which
illustrates a type of job which is quite common. The angle may
be cut from a straight angle section and the triangular shape cut
out with a hack saw. The triangle is cut out so that the angle may
be bent at right angles. The tip of the triangular, however, must
be cut square off in order to allow a right angle to be bent without \he
edges coming entirely together. The distance between the edges
after the angle has been bent through 90 degrees should be equal
to the thickness of the angle. The operator may then bridge cross
the two edges from one side allowing as little metal to drop down
between the edges as possible. Then the angle should be turned
over and the space between the edges completely filled by weld-
ing in one or more layers.

Lesson VIII

FLUE WELDING

This exercise deals with the welding of flues into the flue sheet
of a boiler. This work is encountered in fire-tube boilers of all
kinds. The operation requires a considerable amount of skill in
handling the arc. A preparation of the flue sheet for welding in
actual practice is usually what makes the job a success or failure.

APPENDIX 247

In practice, the proper way of preparing a flue sheet for welding
is to put the flues in exactly as if they were not to be welded. The
boiler should then be flred at least once to allow the tnbes to take
their permanent set. The flne sheet should then be sand-blasted
to clean the surfaces to be welded. If no sand-blast is avulable,
the pneumatic tool should be used to knock the oxide off the sur-
faces, after which the surfaces should be thoroughly brushed with
a wire brush, then the welding may be done. If the work ia pre-

Fn. ».— Weldinit boiler fluu; showing are welder's clothing outfit.

pared in this manner and properly welded, the results will be
uniformly successful.

Material required: Section of '^" boiler plate with four 2"
flues rolled in as shown in cut; %" electrode with 100 amperes
should be used.

1. Set the sample as shown in the photo^aph. Use head shield
and hold the electrode holder in both hands as shown in the cut.
The first flue at the top should be welded starting at the point shown
in the cut and welding one-half way around, moving from right
to left. Then the other one-half welded starting at the original
point and moving downward to the left The second flue should
then be welded starting at the bottom and welding in two halves so
that they meet at the top. The operator may then weld the other two

248 APPENDIX

flues by either of the two methods illustrated, depending upon which
the operator likes the better. One of the flues should then be sawed
in half to show the quality of the workmanship.

Lesson IX

WELDING STEEL CASTINGS WITH CARBON ARC

This exercise illustrates the kind of work done in a steel foundry
and in certain railway shops. The carbon arc is used in the same
manner as the flame of an oxy-acetylene torch. From 300 to 600
amperes are required for carbon-electrode work of this nature. The
operator must use both hands and therefore the head shield is
required. The carbon-electrode holder is held in the right hand
and the welding rod is held in the left hand. Carbon-electrode weld-
ing is usually considered easier than metal-electrode welding, but
there is considerable skill required to handle a carbon arc successfully.

Material required: One small steel casting (Fig. 60), carbon-
electrode holder, carbon electrode %" in diameter, sharpened to a
point at one end, 300-ampere welding capacity (if 300-ampere
unit is not available, two 150-ampere units may be connected in
parallel), 3/1 6" welding rod.

1. For preliminary practice, the operator should use the 300-
ampere carbon arc and cut into small pieces several pieces of boiler-
plate scrap. For this work the arc should be held approximately
a quarter of an inch long. After the operator has practiced suffi-
ciently at this work to be able to make a clean cut along a pre-
determined line, he should try welding together two pieces of boiler-
plate scrap using the carbon arc and the 3/16" welding rod to fill in
with. It will be rather difficult to control the arc and lead it in
any desired direction.

2. If 3/16" carbon electrodes are available, one should be sharp-
ened and placed in the metal electrode holder and some cutting of
l/l6" plate done using 150 amperes. The operator should be able to
cut a straight, clean cut upon completing this exercise.

3. Using the riveted sample which was used in Lesson VII, the

APPENDIX 249

operator shouH use the 300-ampere carbon arc to cut out a section
of the upper plate between two rivets. To perform this operation,
the plate should be set up in the ssnie position in which it was
welded so that when the metal is melted by the carbou arc it can
run down out of the cut. The sample should then later be welded
flush, using the metal-electrode process. After working with the

FiQ. 80. — Small eteel castiot.

carbon arc and before working with the metallic arc on this job,
it will be necessary to chip the oxide off the surface to be welded,
since the carbon arc forms a very thick coating of oxide.

4. This exercise deals with the correction of a flaw in the steel
casting due to a sand spot. This defect in the steel casting is
caused by the crumbling of the mould. It is necessary to bum the
aand spot out with the carbon arc and fill in new material from the
welding rod. If there is no sand spot on the casting available, it
will be suflicicnt for the operator to heat a spot approximately 1^"

250 APPENDIX

in diameter to the molten state, then quickly break the arc and
strike the molten metal a sharp blow with a ball-peen hammer; If
the operator had performed this operation on a sand spot, he would
have floated out; most of the sand by the heat of the arc. The sharp
blow with the hamnaer throws the molten sand and slag out of the
weld. The next operation is to fill in the defect with new material
from the welding rod. The operation must be performed as rapidly
as possible, otherwise the metal added as well as the metal of the
casting in the vicinity of the weld will be ruined by the extreme
heat. The arc should be used to cut off short pieces of the weld-
ing rod and then these pieces should be melted and puddled in the
proper place. In case the arc breaks during the operation, it should
be started again on solid metal that is not molten and the arc
brought over into the welding area quickly. If the arc is started by
touching the molten metal with the carbon electrode, it is very likely
that the weld will be hard, owing to the fact that carbon from the
electrode has gotten into the weld. As soon as the added material
has been fused into the weld, the arc must be broken. There is always
a tendency on the part of a beginner to play the arc too long on the
completed weld in an attempt to give the weld a smooth-finished
appearance ; this . results in burning of the metal. In steel-casting
work to avoid hard spots two points must be observed : ( 1 ) Some pre-
heating must be done around the point at which the weld is to be
made with the arc so that it will not be cooled too suddenly. (2)
The carbon electrode must not be brought in contact with the molten
metal as explained before.

This operation should be performed several times by the operator
until he can produce a weld which is satisfactory to him.

Lesson X

CAST-IRON WELDING

The purpose of this exercise is to give the operator an idea of
what can be accomplished with the electric arc on cast iron. The
operator will frequently hear amazing statements as to what some

APPENDIX 2fll

particni&r operator has done along the line of welding cast iron,
but it is a fact that there are only a few conunercial applications
of the process in the welding of cast iron. The difficulty in welding
cast iron with the electric arc is not due to the fact that the metal
cannot be properly fused, but is due to the fact that the sudden intense
heat of the arc over a local area results in the production of a hard
weld and the introduction of contraction stresses which often result

in cracking. Using the carbon welding process, cast-iron welding
rods may be fused into a cast-iron piece. Using the metal-electrode
process and a soft iron or steel electrode, it is impossible to make
a reliable weld between the added material and the cast iron. Using
the metal-electrode process, certain work can be done hy the intro-
duction of steel studs in the cast-iron pieces to be welded together,
so that a certain amount of strength is obtained by the bond formed
between the steel studs by the welded material.

2518 APPENDIX

Material required: 300-ampere welding capacity, 3/1 6" cast-iron
welding rod. One small grey-iron casting (Fig. 61).

A small grey-iron casting should be broken and the edges bevelled,
using the carbon arc for cutting. The pieces should then be placed
in a carbon mould so that the molten iron, when it is added, will not
run away from the joint. This is illustrated in Fig. 61. The
carbon arc should be used to pre-heat the casting. It is not necessary
to heat the piece to a red heat. The carbon arc and cast-iron welding
rod should then be used to fuse the added material to the piece.
As in Lesson IX, care should be exercised not to play the arc upon
the weld any longer than is necessary to give complete fusion. In
case the metal gets too hot and runs badly, the arc must be broken,
and an interval of time allowed for it to cool slightly to eliminate
the trouble. After the weld is completed, the piece should be wrapped
up securely in asbestos paper and allowed to cool slowly for 6 or 8
hours (larger pieces require from 18 to 24 hours to cool). As an
alternative to wrapping in asbestos paper, the piece may be covered
in previously-heated slacked lime. The idea of the lime is the
same as the asbestos, to cool the casting slowly. If the work is
properly pre-heated and welded rapidly and very slowly cooled,
the material in the weld will be as readily machinable as the balance
of the piece. No flux of any kind is required, although borax
may be used.

APPENDIX V

This design was prepared by me * in London with the coopera-
tion of Mr. W. S. Abell, chief ship surveyor of Lloyd's Register. The
design is such that a lot of work could be done at constrijctional
works, or could all be equally readily done in the shipyard. The
work is closed up by the use of service bolts.

The system of welding in view is the "arc/' although " spot "
welding could be adopted for the ground work. In general, the
position of materials does not differ from that of a riveted ship.

There are no large pieces to handle; therefore, no special lift-
ing facilities are required, but such shipyards as can conveniently
handle large sections could complete large sections on the ground,
if desired.

The difficulty of doing " overhead " welding has been strongly

emphasized by welding experts, and this has been kept in view and

♦.

obviated, or at any rate so far as the " strength " welds are concerned.

A slight departure from riveted practice is to make some of
the side and bottom shell plates much wider than would ordinarily
be the case for a ship of this size.

As it happened in this design, a length of plate of 22 ft. fitted
in admirably with the arrangements, and has the maximum area
which would be supplied from the mills in Great Britain for a plate
of the thickness, viz, : 1/^ inch without " extra."

Three longitudinals are fitted on these plates, and are butted
at the shell butts in order that the longitudinals can be welded to
the plates on the ground. There is no reason why, if desired, two
plates or large sections might not be welded together on the ground,
and the longitudinals fitted in 44-ft. lengths, but this, as pre-
viously mentioned, depends on the builders' facilities.

* Courtesy of J. W. Isherwood.

253

254 APPENDIX

The " clips " of the transverse members being in short pieces
are also welded on to the plates on the gromid.

The seams and butts of the bottom plating are " butted ** and
fitted with outside straps.

The seams of the side plating are arranged clinker fashion in
order to obviate overhead strength welding.

The vertical stiff eners to the floors are fitted in the shop and
in cases like this, viz.: of stiff eners and not strength connections
and where only an odd hole or two could be saved by welding, it is
suggested that riveting be adopted. The same remarks apply to
the horizontal " clips " on side transverse and so on.

The longitudinals and " transverse " attachment angles are welded
to the plates in the shop similarly to the bottom plates. The
gunwale bar in addition to the longitudinal and transverse attach-
ments is welded to the sheerstrake before erection.

The tank-top margin angle is also welded to the shell plate
abreast it before erection.

The side transverse members and transverse deep beams are
assembled and welded on the ground in such sections as are con-
venient to the builders, that is, the beam part can be secured to the
side transverse frame either before or after erection as desired.

The tank-top plates and deck plates have both the longitudinals
and transverse attachments welded to the plate in the shop. It
will, of course, be appreciated that where the longitudinal forms
the seam strap it can only be welded to the edge of one of the plates.
The completion of the connection is, however, a very simple matter,
being only an ordinary downward butt weld.

The process of erection and the welding to be done in the ship
might be briefly described, bearing in mind that no special facilities
of any kind are required for handling the material beyond what
would be in use in an ordinary shipyard which would ordinarily
build a riveted vessel of the same dimensions.

The keel plates are laid as usual, they have already welded to
them the angle bars to take the centre girder and the outside straps

APPENDIX 255

in way of the seams and also at one end of each plate to complete
the butt with the next plate.

The plates for the next strake are brought into position^ they
have in addition to the longitudinals and transverse framing clips^
edge strips welded along one edge to form the seams for the adjoin-
ing strake and edge strips at one end of each plate to form the
butt of the next plate.

This procedure is adopted until all the bottom plates are laid
in position, the seams and butt strips forming convenient rests for
consecutive plates and the service holes for holding the plates in
approximate position.

The floors are then dropped into position (it might here be
remarked that the floors are widely spaced ; in this particular example
they are 5 ft. 6 in. apart) and secured to the longitudinal by bolts
through the outstanding flanges of the vertical angles on the floors.

When a sufficient number of tank-top plates, with the longi-
tudinals and transverse clips already welded on, are laid in position
so as to adjust and fair the bottom components, the welding opera-
tions on the ship can be commenced.

As soon as the margin plate of tank top is adjusted to its
proper position, the transverse frames, which are 11 ft. apart, are
erected and shoved in place.

The sooner the sheerstrake, which has already welded to it the
longitudinal frame and giunwale bar, can be brought into position,
the more readily can the side plating with its framing work be
erected and faired.

Welding can, of course, be commenced on the side as soon as
a sufficient section is faired and secured.

The deck transverse beams, if not already lifted into position
with the side frame, are then placed in position and secured. The
deck plates, with the longitudinals already welded on, are dropped
into position and the deck faired, when all the welding now to be
done in the ship can be proceeded with. I wish to draw attention
to the very small amount of welding to be done on the ship.

256 APPENDIX

DOUBLE BOTTOM

Full welds: Butt seams bottom plating.

Butt straps of bottom plating and butts of longitudinals.
Connection centre keelson plate — ^top and bottom.
Seams of tank-top plating.

Butts of tank-top plates and butts of longitudinals.
I might here remark that Lloyd's chief surveyors have intimated
their approval of the butting of the tank-top plating at the centre
line in order to obviate overhead full welding.
Light welds: One edge of outside seam strap.

Top and bottom of floor plates to transverse clips; floors 5 ft.

6 in. apart, giving ample room for working.

If the service holes in the seams and butts are filled with plugs

and welded and the small connections at the intersection of the

longitudinals with the floor plates are welded, then all riveting in

the double bottom on the ship is obviated.

It is assumed that the service holes for the work done in the
ship were filled up before the material was brought to the ship
and before erection.

SIDE PLATING AND INTERNAL WORK

Full welds: Outside edges of shell overlaps.

Butts of shell plates and longitudinals.

Connection of transverse plates to shell clips.

Connection of transverse to inner-bottom plating.
Light welds: Inside edges of shell overlaps.

DECK AND INTERNAL WORK

Full welds: Seams of deck plating.

Butts of deck plating and longitudinals.
Heel of gunwale bar.
Connection of beam to transverse.
Light welds : Connection of transverse beam plates to clips already
on deck.
Pillar and detail work.
The same remarks apply in regard to the filling of service holes.

258

APPENDIX

tKAOTHt**

<^ 7»..

73.3J

tXM&lMtt^

3o»

•,

Teoza .9* 1 ««.«.» M«

ill ario&p Bit .

-a

I* w^

APPENDIX

259

MIDSHIP atCTlON

t>.HC^lonm.->-t.»l<TM B P? so S -Q Hgr APTM Vt» Ay:.9 ng^T M MW» ga 1

t • ■»'• 11' tfa«»P « f «.'ll-C

..,^<tU.-'2m^-;Aif^^

in WFito!n!rTl^

iO iOBiiOliOpr:

) mi 1

! \r~

f^^^^^=;--^

^ ®

APPENDIX

261

""^— •"fmriimmi

i i n -m i iiiiiiwi

■Mct»«. ..« ■■HOC n»M»<,tm^ ~ ^< «, «.T1»M «.-WVM

-^^ *"•

ma^ilui 1 tia

262

APPENDIX

SHIP ag;r^,T|ffM

THttr A»a-rf a»g.>tw.,te» ^:^' fff„ ^^ „ .»

I aCALe ik\.^f'

^otr hi rotJE^ajL

*»»*'-^

.S£.

"^^^ mirt ifl^

-J

•r.ivWft

[t %'■ t^H

"••"r.-W

L.Di_aZb

■r.fi

^* •'-'•nri? trTut TnrTip

itP»..5T»^t* Hurra. 1^^,

c w^ttc But^ -^f^, ^,» rrrtwn

^'»'->^

^'"'•'3'

JDtlUtt,

P> PLATb^ t^TT^. ftBffyn >1^-T«gM^.^|Y,T£g=

J=bA

% *»m,LL atnt^f^ At t...^

APPENDIX

£63

■flitWIM ^

lbbLjAIL.

II 1 iMrnwn n

■— "y-^"tVMIinM mni M^[*^inilwil

im m^-iimiMi iftiit

■^ L..WT».IDI» ■■»> « TM.

gCLtaw TNit B.«i 1

tMHM.nrrrr bw«t

A o

fCs^-'T^

)

\ Lit irttf

y*iH|nf ■wiVWiltffiflfiii

■* * *^*" *-~ ' — tma

iwwi tiffc «mAa-

•nn ■f- —

264

APPENDIX

IN5TRUCTI0M CH/tR

ESKiP 1
icHmiotq —

WITH 6TANDARD SYMBOLS

HAhAL OP
ARTICLE.

/||OU«MT
I WOW

OUT

ROW

e I

M 5.

— I —

5 ft

Ct»

BRASS | 2

6M.V>UillE

FINI6H — ^^Si^

SIZE or

MATERIAL

»IN«Lt

D A

] — ISTRAIMn

5TRAP| — fpUTT [ — fLAP | — ^q^^^^^ — |nLLETH PLUG H '^^^ I

OOUOLC

DCSICM OP

weuD

SIM6LE I

aevet.

rrLAr"— ^S^^—

TACK

VavcK*

POSITION OF
WELD

VERTICIL

OVCRHCAD

}

KINO OP
WELD

tr »gW ftTM
~rAXEF»

COMPOSITE

CATCK*

|RfiiiirflR(ai| —

rLU6H —

TYi^E OP
WELD

ORU&H

CHISEL.

CONCAVE

JSTCEU I — OR0N2C —

PRCPARATIOI
POR WELD

APPENDIX

265

TYPE OF JOINT

STRIP weld is ona in whioh tliB seam of two
adjoining plataa or aorfacaa ia reinforoad by
any form or ahapa to add atrength and atability
to the Joint or plate. In thia form of veld
the aeam oan only be welded from the aide of tbB
work oppoalte the reinforoement, and the rein-
foroemant of whaterer ahapa moat he welded from
the aide of the work to li^oh the reinf oroement
ia applied.

BUTT

BOTT weld ia one in whioh two platea or
aorfaoea are hroaght together edge to edge and
welded along the aeam thoa fomiad. The trwo
platea whan ao welded, form a perfectly flat
plane in thamaalrea exoluding the poaaible
project ire oaaaed hy other indiridual ohjeota
aa framaa, atrapa, atiffenera, eto., or
the building ap of the weld proper.

LAP weld la one in whioh the edgea of
two platea are aet one abore the other and the
welding material ao applied aa to bind the edge
of one plate to the faoe of the other plate.
In thia form of weld the aeam or lap forma
a ralaed aorfaoe along ita entire extent*

APPENDK
TYPE Of JOINT c..".>»">

APPENDIX
K.MH OF WU.D

SHWLtV "^

iffi'

M dailpiar. Ts M Ufl >tun CM "T"

» Tiiiti 1> to ta ■ laiilBia "■trtDeth"

■ plat* lattlDg *artla>lly to tM

^JglnlBg BMlBr, ud onll •bm tl»-

•stnd* uo b» ■pplltd froB Mth lU** or tli»

tut of u idlal]

oouBu KvcL ";^ „*;:=;«»¥»

WBi-w M. OKE «id« ooly to im Bngl., tm dtEnai or whloh

^-"> <'^ \^ ^ -» !•" to th. i..lg»r 10 K u«d -M«

f N.IL' r^^ ""IMl- ■tr.nglh 1. rtnalrM, Mi «*« ■Us-

. 268

APPENDIX

POSITION ' or WELD-

BULKHEAD

rap or deck .

^.^ OeCIC HATING .

PLAT position is determined when the welding material is applied to a
surface on the same plaiie as the deolc, allowing the eleotrode to be held in an
uprigfbt or vertical position. The welding surface nay be entirely on a plane
with the deck, or one side may be vertical to the deck and welded to an adjoining
member that is on a plane with the dedc.

HOKIZQNTAL position is determined when the welding material is applied
to a seam or opening, the plane of which is vertical to the deck and tbe line of
weld is parallel with the deck, allowing the electrode to be held in an inboard
or outboard position.

YBBTICIL position is determined when the welding material is applied
to a surface pr seam, whose line extends in a direction from one deck to the
deck above, regardless of whether the adjoining members are on a single plane or
at an angle to each other. In this position of weld, the eleotrode would also be
hel4 in a partially horizontal position to the work.

OVEPHBi^ position is determined When the welding material is applieid
from the under side of any member whose plane is parallel to the deck and
neoessitates the eleotrode being held in a downright or inverted position.

APPENDIX
KIND or- WILD.

t b> i^soitm bj

270

APPENDIX

TYPE- or WELD

fetlN FORCED 8V«t«i./"^ _^.^ovtr%9 Mwf •wowrTi MinwH immo

RBIMFCSiGBD la a term applied to a weld when the top layer of the weldiOB
material la Uillt up above the plane of the sorroundlng material as at Fig. "A** or
Pig. "F* ahore, or when need for a oomer as In Tig. "C**. The top of final layer
ahoald projeot above a plana of 45 degrees to tha adjoining material. 2*hi8 45 degrees
line is shovn 'Motted** In Pig.^C** above. This type is ohiefly used in a "Strength" or
"CSonqpoaite** Idnd of weld for the puxpose of obtaining the maximum strength effloienoy.
and should be speoif led by the designer, together with a minimum number of layers of
welding material.

rLU5H

PLUSH is a term applied to a weld when tha top layer is finished perfectly
flat or on the' sane plane as on the adjoining material as shom at Pigs. •«D'* and "B** above
or at an angle of 45 degrees when used to oonoeot two surfaces at an angle to each other
as at Pig. "P** above. This type of weld is to be used where a maximum 'tens lie strength
is not all Important and must be specified hy the designer, together with a minimum
number of layers of welding material.

CONCAVIL

••TTW LlNtS SMOW TIM MfSN m«MC«. •

CONGi.VB.is a term applied to a weld wlran the top layer finishes' belcw the
plane of the surrounding material as at Pig. "G" above, or beneath a plane of 45 degrees
at an angular connection as at Pigs. "H" and *'J" above.

To be used as a weld of no further importance than filling in a ^eam or
opening, or for atrictly cauUcing purposes, when it is found that a miniiman amount of
welding QAterial will suffice to sustain a specified pound square inch pressuro without
leakage. In this^Type of weld'* it will not be necessary for the designer ordinarily
to specify the number of layers of material owing to the Uck of structural importance.

APPENDIX

271

.COMBINATIONS- OP • 5YMB0U6.

STRAP WCLQ REINFORdC^
COMl^>«ITt OFdLAYffR9>
VERTICAL, STRAIGHT.

"■33^

,dmwr .

STRAP ^

Tbis aketob and syaibol ahows a strap
holding two plates together, setting vertical-
ly, witli tbB welding naterial applied in not
less than three layers at each edge of the
strap, as well as between the plates with a
reinforced con|>osite finish, so as to make
the welded seams ahsolately water, air or
oiltigfat, and to attain the oBxinam tensile
atrength. fPhe edges of the strap and the
plates are left in a natural or sheared
finish. This type of welding is used for
most particular icind of work where jBexinaio
strains are to he sustained.

STRAP WEtP» FLUSH*
STRENGTH OF 3 LAYERS,

HORIZONTAL, FLAT ANP
OVCRHEAR POUBLC ALVCI

This illustrations shows a strap holding
two plates together borisontally, welded as a
strength member with a minimum of three
layers and a flush finish. Inasmuch as the
strap necessitates welding of the plates
from one side only, both edges of the plates
Bre boTelled to an angle, the degrees' of
which are left to the discretion of the de-
si0ier. The edges of the str^p are left in
a natureO. or sheared state, and the w a xImHrn
strength is attained by the mode of applying
the welding material, and through the sec-
tional area per square inch exceeding the
sectional area of the surrounding material.

STRA^ TACK.. OVERHEAC^
8- CENTER TO CENTER^
4* L0N6, BUTT, REIHPORCEP,

composite: or 3 layers^

FLAT, 6TRAI«>4T.

W«k9

This symbol represents two plates butted
together and welded flat, with a codecs ite
weld of not less than three layers, and a
reinforced finish. A strap is attached by
means of orerhead tacking, the tacks being
four inches long and spaced eight inches
from center to center; In this case, the
welding of the plates is of mszimum strength
and water, air oroiltight, but the tacking
is either for the purpose of holding the
strap in place until it may be continuously
welded, or because strength is not essential*
111 the edges are left in their natural
OF sheared state.

#Vtf|HBA» ^»9m4

272

APPENDIX

COMBINATIONS OF 5YM50LS JtoHTinuc^

TtF

butt v/elp, concave^
caulk^in^ of z layetrs,
flat; straight.

Tb» symbol sboini represAnts a Batt
Weld between tn plates with the welding
material finished concaved and applied
in a minimum of two layers to talce the
place of caulking. The edges of the
plates are left in a natural shear out
finish. This Symbol will be quite fre*
quently used for deolc plating or any
other place where strength is not e8s«>>
tial, bat where tbe material must he
water, air or olltight.

dTBAlQHT

eUTT WCLO, REINFORCE^
STRENGTH or 3 LAYERS,
VERTICAL, OOUBLt VEC,

ftOUBl^ VEC

I m t

PLATE. .^ PtATC

et^

'VfcRTICAU V^CLl

This Symbol is used where the
edges of two plates are vertically buttM
together and welded as a strength mambeif*
The edges of the adjoining plates are
finished with a •T)ouble Yee" and the min-
imum of three layers «f welding material
applied from each side, finished with a
convex surface, thereby naicing the sec-
tional area per square inch of the welil»
greater than that of the plates. This
will be a conventional Symbol for shell
plating or any other members requiring
a maximum tensile strength, where the
welding can be done from both sides of
the worlE.

BUTT WtLD, TLUSH, ^^^^
COMPOSITE, or 3 LAYERS

FLAT, double: bevel.

^4.AT . win.0

This Symbol shows teo plates batted
together in a flat position where the
-welding can only be applied from the top
surface. It shows a weld required for
plating where both strength and water-
tightness are to be considered. The braid-
ing material is applied in a minimum of
three layers and finished flush with the
level of the plates. Both edges of thft
adjoining plates are bevelled to- an angle,
the degrees of irtiich are left to the dis-
cretion and judgment of the designer,
and should only be ased when it la itt-
possible to well from both sil^a of th«
wortc*

APPENDIX 275

.COMBINATIONS- OF-SYMOOL5-(toMT.«u«(9.

LAP WtUCJ COMCAVE,
CAULMN6 OF Z. LAYERS, , ,
OvtRHCAf AhV FLAT, ?l,lll

aTRAI&MT

LAP WELCJ REIMFORCCD

iTRENfiTH or 3 LAYERS '

AND TACMN6, iVCENTtR J

TO CtNTCR., e LONG, o

VCRTICAL, 5TRAIGMT t

PLUfi ANP LAP VltLP,
1STRCN6TH OF a LAYtRSi
[i FLUSH, FLAT, I

•4B9S an

•trmngtll -I

mail u»i

f lan»l

WldlBf

274

APPENDIX

C0MI^INATI0N5 • OF ^ SYMBOLS* fro-T.-.«o .

FtU6 ANP PILLtT WtLC^
REINFORCCPy 3TRtN^TM O^
3 LAYER^t PtAT ^IN^LC
PCVCI-, ANP ST^AUHT

PILLPT vaetPL RCWFORCEJR
COMPOSITE OF 3 LAYCRa^

plat: vertical anp

OVCf^HCAP^ STRAiaHT

ViCMTICAl.

evsmMtAV

SKCriON
TMHU AA

PILLET VlEtP, rLU6H,
6TREW6TM OF 3 LAYCR6,
FUAT, 6TRAIQHT,

ft, AT FIIXKT ^tUJO

The adjoining sketch ehovs a pad eye
attactaed to a plate by msaas of a fillet
weld along the ^dge of the fixture, and
further strengthened by plug welds in two
oountersonk holes drilled in the fixture.
The welding nateriAl is applied in a flat
position for fa strength weld with a min-
inoin of three layers and a reinforced
finish* The edges of the holes are bev-
elled to an angle, which is left to the
Jt;d0nent of the designer, but the edges
of the fixture are left in their natural
state. This method is used in fastening
fixtures, olipe or accessories that
wiuld be subjected to an excessive strain
or Tihratioi;*

This illustration shews a fixture
attached to a plate by mans of a con^site
weld of not less than three layers with a
reinforced finish^ The fixture being
placed vertically, necessitates a coi»-
bination of flat, vertical and overhead
welding in the course of its erection.
Although a fixture oi this kind would
never be required to be watertight, the
coiq>osite symbol is simply as a possi-
bility of a combination.

This symbol represents a fixture
attached to a plate by a strength fillet
weld of not less than three layers,
finished flush. The edges of the fixture
are left in their natural state, and the
welding material applied in the corner
formed by the vertical edge of the
fixture in contact with the face of the
plate.

APPENDIX

275

.CoriDINATION5 • OF-$YMDOL5*«oHfiN««x

.S3 r

TEC VfEL.P, FLUSH.
6TRENQTH OF 5 LAYERS

FLAT, ftiNatc vet.

fte adjoining stotoli illustrates ths
edge of a plate velded to ttas faoe of aaottasr
plate, as intha case of the bottom of a trans-
verse bulidMad being weld.ed against the deck
plating. To obtain a msxinnm tensile strength
at the Joints the edge of the plate is out to a
"Single Vee<« and welded on both sides with a
strength weld of not less than three layers, and
finished flush. This would be a ooorenient way
of fastening the interoostals to the keelsons.
In this particular oase, the welding is dona
in 9, flat posit iont

.63 V

Tee WELi^ ReiNFORcei^

STRENGTH OF 3 LAVeRQ^
{TICAL, dl,NQLC vet.

This symbol shows another oaae of Tee
Veld with the' seam setting in a vertioal
position, and the welding material applied from
both sides of the work. The edge of the plate
is finished with a "Single Yee" ani a minimnm
of three layers of welding material is spplied
from each side, finished with a oonrex surfaoe,
thereby making the seotional area, par square
inch of the weld, greater than that of the plate,
allowing for a msximum tensile strength in the
weld.

STRAP AND TEC Vie LP,
FLAT^ RClNPORCeC^ TACX^
ta** CENTER TO CENTER,
6" LONG, SINGLE BEVEL,
OVERHEAD, ATRCN6TH OP
:^ UAVERS,.. PLUSH.

The illustration herein shoim, represents
an exan^le of the possible ooQbination of
symbols. An axigle iron is tack welded to the
plate in the form of a strap or stiffener,
though in aotual praotioe, this might never ooour.
The tacks are spaoed twelve inches from center
to center, and are six inches long, and ^plled
in a fiat position, with a roinforoed finish.
As the strap prevents welding the plate. from
hoth sides, the edge of. the plate is bevelled,
and the welding material applied for strength
in not less than three layera in an overhead
position and finished flush,- Note tlmt in
specifying tack welds, it U essential to give
the space from center to center of weld, and
length of wled by use of figures representing
Inolies placed either aide of the oiroumsoribiag
•ynbol of the^ combination*

APPENDIX VI

{Lloyd's Register of Shipping,)
Application of Electric Arc Welding to Ship Construction

introductory remarks

Although electric welding in various forms has been employed
for many years for ship repair work, yet, in practice, owing to
many factors, its use has been practically confined to those parts
of the structure which are not likely to be exposed to important
structural stresses.

It is only in recent times, commencing from the early days of
the war, that appreciable progress has been made in the developments
of electric welding which would appear to justify the extension of
such methods to replace the usual riveted connections of heavy

structural work.

•

The aim has been to secure reliability and regularity of operation
in the welding process, and to assist the workmen by improving the
means of control over the work. Just as in the case of application
of steel to shipbuilding, it was necessary to devise means for the
production of the material in large quantities and of constant quality,
so also is it necessary that the welding electrodes should be manufac-
tured with the greatest possible degree of uniformity.

Reliability of operation is also facilitated by adjusting the density
of the electric current to the size of electrode used, and, further, the
size of the electrode should reasonably vary directly with the thick-
ness of material to be connected.

Research was necessary to discover means for minimizing the
burning of the deposited material in the direction of preventing
oxidation. In the early days of welding, the molten electrode was
exposed to air throughout the whole time of deposition and con-
sequently oxidation was more or less certain to occur. With coated
276

APPENDIX 277

metal electrodes, burning is reduced to a minimum by the use of a
slag which envelops the molten steel and floats on its surface after
contact is obtained with the material to be connected. Even in this
system, skilled workmanship is essential, as the production of a long
arc obviously increases the chances of burning.

The composition of the material of the electrode in relation to
the nature of the steel to be connected is obviously a matter of
importance. What the composition is to be can only be gauged by
experiment and by wide experience, and it is in devising the physical
tests for work of this nature that the greatest difficulties arise.

It is commonly accepted that the tests imposed on manufactured
material do not in any way represent • the strains which may be
experienced in practice. Such tests are rather based on simple means
for determining the average reliability of the material. Thus also
is this case no one particular test is likely to determine whether the
welding process under trial is sufficient for the work it is likely to
have to do.

It is therefore necessary to approach the problem rather on the
basis of circumstantial evidence and to decide from a number of
different types of experiments whether, on the whole, the perform-
ance is satisfactory.

The more particular problem in shipbuilding is the connection
of mild steel containing a percentage of carbon of about .15. This
material in the form of plates and section bars has considerable
work done to it during the process of manufacture, with the con-
sequence that it possesses a fine structure and a ductility which is
uniform in any direction. The finished material may be said to be
practically free from fibrous structure.

With electric welding, molten metal is attached to the mild
steel and from the extent of the cooling surface the deposited material
is rapidly lower in temperature, with the consequence that the weld
tends to become deficient in ductility.

The problem therefore is to select the material of the electrode
so that the general elastic properties of the structure are not
unduly depreciated.

278 APPENDIX

The investigations were undertaken to determine the possibilities
of the application of electric welding to shipbuilding and as it was
desired to obtain as good a knowledge as possible of the physical
properties of the combination of rolled and welded material, highly
skilled operators were employed.

It must therefore be realized that the results of the experiments
which have been made represent skilled practice, and that in general
such performance can only be equalled with good workmanship and
efficient supervision.

NATURE AND DESCRIPTION OF EXPERIMENTS

The general scope of the experiments included: —

(a) Determination of modulus of elasticity and approximate

elastic limit.
(h) Determination of ultimate strength and ultimate
elongation.

(1) rotating specimens,

(2) stationary test pieces.

(d) Minor tests, such as —

(1) cold bending of welds,

(2) impact tests of welded specimens.

(e) Chemical and microscopic analysis.

Tests were carried out on specimens as large as possible, par-
ticularly in respect to the static determination of elasticity, ultimate
strength and elongation, some of the test specimens being designed
for a total load of just under 300 tons. The advantage of these
large specimens was that the effect of workmanship was better
averaged and the results were more comparable to the actual work
likely to be met with in ship construction.

With alternating stresses the specimens were relatively of small
size. For the rotating test pieces, circular rods, mainly machined
from a welded plate, were used, the diameters selected being 1 inch
and % inch. These bars, about S feet in length, were attached to a
lathe headstock and a pure bending moment in one plane was applied

APPENDIX 279

by means of two ball races to which known weights were attached.
The material of the bar was thus exposed alternately to maximum
tension and to equal maximum compression once in each revolution.
• The machine was run at about 1,060 revolutions per minute.

Bars of identical material were tried in pairs, one specimen
welded and the other unwelded, and the numbei' of revolutions before
the specimens parted was observed for various ranges of stresses
varying from ±15 tons to ± 6 tons.

In the second series of alternating stress experiments, flat plates
were used of three thicknesses, vis. — y^ inch, % inch, and % inch.
These specimens were tried in groups of four, each giroup consisting
of one plain, one butt-welded, one lap-welded and one lap-riveted
plate. The specimens, which were about 14 inches long by 5 inches
broad, were clamped along the short edges, so that the distance
between the fixed lines was 12 inches. Each plate was also clamped,
near the middle, to the end of a pillar, which by means of a crank
arm was caused to oscillate and to bend the specimen equally up and
down by adjustable amounts (the maximum total movement in any of
the experiments tried was 5/1 6 inch). The machine was run at vari-
ous revolutions (not exceeding 90 'per minute) and the number of
repetitions at which the specimen parted was observed.

Minor tests of various kinds were undertaken, of which the
principal ones had reference to the suitability of the welded material
to withstand such bending and shock stresses as might occur in
the shipbuilding yards. The experiments on bending consisted of
doubling the welded plate over a circular bar of diameter equal
to three times the plate thickness, and comparing the results with
those of the plate of the same material but unwelded.

In the impact tests, heavy weights were dropped from various
heights on to the welded portion of a plate 5 feet in length and '2
feet 6 inches in bireadth, the weld being across the plate parallel
to the shorter edge. The deflections were noted and the condition of
the weld was examined after each blow.

The chemical and micrographical examination followed the
ordinary practice.

280 APPENDIX

SUMMARY OF EXPERIMENTAL RESULTS

1* Modulus of Elasticity and Approximate Elastic Limit,

(o) In a welded plate the extensions in the region of the weld are
sensibly the same as for more distant portions of the unwelded plate.

(fe) With small welded specimens containing an appreciable
proportion of welded material in the cross-sectional area, the rela-
tion between extension and stress is practically the same^ up to the
elastic limits as for similar unwelded material.

(c) The elastic limit (or the limiting stress beyond which exten-
sion is not approximately directly proportional to stress) appears to
be slightly higher in welded than in unwelded material.

{d) The modulus of elasticity of a small test piece, entirely com-
posed of material of the weld, was about 11,700 tons per square
inch as compared with about 13,500 tons for mild steel and about
12,500 tons for wrought iron.

2, Ultimate Strength and Ultimate Elongation.

(a) The ultimate strength of welded material wiiih small speci-
mens was over 100 per cent, of the strength of the unwelded steel
plate for thicknesses of % inch, and averaged 90 per cent, for plates
% and 1 inch in thickness.

(h) Up to the point of fracture, the extensions of the welded
specimens are not sensibly different from those of similar un-
welded material.

is less ductile than mild steel, and the ultimate elongation of a

welded specimen when measured on a length of 8 inches only
averages about 10 per cent, as compared wiiih 25 to 30 per cent.

for mild steel.

S, Alternating Stresses.

(a) Rotating Specimens (round bar),

(1) Unwelded turned bars will withstand a very large number
of repetitions of stress (exceeding, say, 5 millions) when the range

APPENDIX 281

of stress is not greater than from 10^ tons per square inch tension
to 10^ tons per square inch compression.

(2) Welded bars similarly tested will fail at about the same
number of repetitions when the range of stress exceeds ± 6% tons
per square inch.

(h) Stationary Test Pieces (flat plate),

(1) Butt- welded specimens will withstand about 70 per cent, of
the number of repetitions which can be borne by an unwelded plate.

(2) Lap- welded plates can endure over 60 per cent, of the
number of repetitions necessary to fracture a lap-riveted specimen.

4, Minor Tests.

(a) Welded specimens are not capable of being bent (without
fracture) over the prescribed radius to more than 80 degrees with
^-inch plate, reducing to some 20 degrees where the thickness is 1
inch. Unwelded material under the same conditions can be bent
through 180 degrees.

(b) Welded plates can withstand impact with a considerable
degree of success; a half -inch plate of dimensions already quoted
sustained two successive blows of 4 cwt. dropped through 12 feet,
giving a deflection of 12 inches on a length of about 4 feet 6 inches
without any signs of fracture in the weld.

5, Chemical and Microscopic Analysis,

(a) Chemical A nalysis,

(1) The electrode was practically identical with mild steel, but
there was a greater percentage of silicon.

(2) The material of the weld after deposition was ascertained to
be practically pure iron, the various other contents being carbon .03,
silicon .02, phosphorus .02, and manganese .04 per cent, respectively.

(6) Microscopic Examination,

(1) The material of the weld is practically pure iron.

(2) The local efl^ect of heat does not appear to largely affect
the surrounding material, the structure not being much disturbed
at about 1/1 6 of an inch from the edge of the weld. The amount of
disturbance is still less in thin plates.

282 APPENDIX

(3) The weld bears little evidence, if any, of the occurrence
of oxidation.

(4) With welds made as for these experiments, i.e., with flat
horizontal welding, a sound junction is obtained between the plate
and the welding material.

6. Strength of Welds {La/rge Specimens),

(a) Butt Welds have a tensile strength varying from 90 to 95
per cent, of the tensile strength of the unwelded plate.

(fe) Lap Welds,

(1) With full fillets on both edges the ultimate strength in
tension varies from 70 to 80 per cent, of that of the unwelded material.

(2) With a full fillet on one edge and a single run of weld on
the other edge the results are very little inferior to those where a
full fillet is provided for both edges.

(c) Riveted Lap Joints, For plates of about 1/2 "^^h in thick-
ness, the specimens averaged about &5 to 70 per cent, of the strength
of the unperf orated plate.

OBSERVATIONS ON EXPERIMENTAL RESULTS

(i) Static Elasticity, It will be observed that the statical tests
made to determine the elasticity indicate that, in general, the com-
bination of welded and unwelded material behaves practically homo-
geneously up to at least the elastic limit. Moreover, the experiments
show that the process of welding is such that the stress is distrib-
uted practically uniformly over the weld, and also transmitted uni-
formly to the adjacent plates.

The material of the weld is practically pure iron, and from the
tests made on a specimen composed entirely of the deposited material
of a weld, it will be seen that for a given stress the weld stretches
slightly more than mild steel. This property will enable any undue
occurrence of load being transferred in a proper manner to adjacent
portions of the structure.

When, however, the stress exceeds the elastic limit and is so great
that the extension grows continuously without increase of load, the
welded material fails sooner than mild steel. This disadvantage is.

APPENDIX 283

however, of little practical importance in shipbuilding, and may be
regarded as negligible in the particular problem under consideration.

(2) Dynamic Elasticity, In a structure, such as a ship, which
is exposed to variations and reversal stresses, it is extremely important
to know whether the material to be used is likely to break down
rapidly under such alternations and ranges of stress as are likely to
be experienced. The modified Wohler tests employed in the experi-
ments certainly indicate, if considered solely by themselves, that
whereas for a given number of alternations mild steel would withstand
a range of stress of, say, ± 10^ tons, the welded material might be
expected to fail at about ± 6^ tons, a figure which is more nearly
experienced in ordinary ship construction.

It would appear to be necessary to design the welded joints in
such a manner that the amount of work likely to be thrown on the
joint is as small as possible, and to meet such a condition a welded
joint requires to be either lapped or strapped.

It will be noticed that the material in the weld appears to be
nearly pure iron, and experiments of repetitive stress show that
wrought-iron bars are likely to fail under a range of stress of per-
haps ± 7 to 8 tons as compared with mild steel at ± 10 to 11 tons.
The weld has to be deposited electrically and is subject to variations
in workmanship ; it would consequently be considered satisfactory if
the material could withstand a range of stress of, say, ± 6^ tons.

Consideration of the dynamic-elasticity properties appears to show
that in any case the welded material can experience as large a num-
ber of repetitions of stress as wrougjht iron could do, and it is always
recognized that although iron could not approach the tests? for mild
steel, yet it was a satisfactory material for shipbuilding purposes.
Further, attention to design of details will increase the performance
of the welded joint, and in addition it must not be forgotten that
5,000;000 repetitions of stress is perhaps more than equivalent to 10
years* good sea service.

(3) Physical Nature and Properties. It has been mentioned that
the welds experimented with are to be regarded as having been
produced under most favorable conditions, and that throughout

284 APPENDIX

the experimental welds were made with the specimens horizontal and
below the operator. In- practice, welds will require to be made ver-
tically and overhead as well, consequently extreme care will be
required in such operations.

The physical examinations indicate that the materials of the
electrode and the system of welding adopted were suitable and reli-
able. Moreover, there was little apparent oxidation and the material
in the neighborhood of the weld was not affected to any pre-
judicial extent.

(4) Strength of Welds and Minor Tests, Broadly speaking, the
tensile strength of butt welds was as great as the unwelded material,
but it is considered that greater reliability of workmanship is obtained
with jomts which are either lapped or strapped.

It was also found that the lapped joint was practically as strong
as a riveted lapped joint and would probably remain tight when
subjected to more trying conditions than are necessary to disturb a
riveted lap joint.

In view of the satisfactory results of the extensive and exhaustive
trials which have been carried out on electric arc welds, the Com-
mittee have decided to adopt, as a tentative measure, the following
Provisional Rules for classification in Llbyd's Register Book of
vessels electrically welded, subject to the notations " Experimental "
and " Electrically welded."

The approval of the Society will be given to any system of weld-
ing which complies with these Regulations and consideration will
be given to any alternative constructional arrangements which may
be submitted for approval.

APPENDIX VII

TENTATIVE REGULATIONS FOR THE APPLICATION OF ELECTRIC ARC

WELDING TO SHIP CONSTRUCTION *

(a) system of welding and WORKMANSHIP

(1) The system of welding proposed to be used must be
approved and must comply with the regulations and tests laid down
by the Committee.

(2) The process of manufacture of the electrodes must be such
as to ensure reliability and uniformity in the finished article.

(3) Specimens of the finished electrodes, together with specifi-
cations of the nature of the electrodes, must be supplied to the
Committee for purposes of record.

(4) The Committee's officers shall have access to the works
where the electrodes are manufactured, and will investigate, from
time to time as may be necessary, the process of manufacture to
ensure that the electrodes are identical with the approved specimens.

(5) Alterations from the process approved for the manufacture
of electrodes shall not be made without the consent of the Committee.

(6) The regulations for the voltage and amperage to be used
with each size of electrode, and for the size of electrode to be employed
with different thicknesses of material to be joined, are to be approved
by the Committee.

(7) The Committee must be satisfied that the operators engaged
are specially trained, and are experienced and efficient in the use
of the welding) system proposed to be employed.

(8) Efficient supervisors of proved ability must be provided,
and the proportion of supervisors to welders must be submitted
for approval.

* Issued by Lloyd's Register of Shipping.

285

286 APPENDIX

(b) details of construction

(9) The details of construction of the vessel and of the welds
are to be submitted for approval.

(10) Before welding, the surfaces to be joined must be fitted
close to each other and the methods to be adopted for this purpose
are to be approved.

(11) All butt and edge connections are to be lapped or strapped.

(12) With lapped connections the breadths of overlaps of butts
and seams and the profiles of the welds to be in accordance with the
following table: —

Thickness Width of overlap Throat

of plate. seam and butt. thiduiefls.

Inches. Inches Inches.

.40 and under 21/4 -28

.60 2^^ .38

.80 2% .48

1.00 3 .50

Intermediate values may be obtained by direct interpolation, and
for thicknesses below .40 the throat thickness is to be about 70 per
cent, of the thickness of the plate.

(IS) A "full weld" extends from the edge of a plate for a
distance equal to the thickness of plate to be attached, and the
minimum measurement from the inner edge of plate to the surface
of weld is the throat thickness given in the table above.

(14) A "light closing weld" is a single run of light welding
worked continuously along the edge of the plate. Such a weld may,
however, be interrupted where it crosses the connection of another
member of the structure.

(15) An " intermittent or tack weld " has short lengths of weld
which are spaced three times the length of the weld from centre to
centre of each short length of weld. Such tack welding may vary
in amount of weld between a " full weld " and a " light closing weld."

(16) The general character of welds is to be in accordance with
the following table: —

APPENDIX

287

Inside Outside
edge. edge.

(a) Butts of shell, deck and inner-bottom plating'
(6) Butts of longitudinal girders and hatch coam- -
ings J

(d) Butts and edges of bulkhead plating j

(e) Frames to shell, reverse frames to frames and

floors

(f) Beams to decks

(g) Longitudinal continuous angles
(h) Side girders, bars to shell, intercostal plates,

floors and inner bottom
(i) Bulkhead stiffeners

F = full weld, L = light weld, and T = tack weld.

L F

Toe. Heel.

(17) AH bars required to be watertight are to have continuous
welding on both flanges with tack welding at heel of bar.

(18) The welded connections of beam, frame and other brack-
ets are to be submitted for special consideration.

(19) The Committee may require, when considered necessary,
additional attachment beyond that specified above, and the welding
of all other parts is to be to their approval.

288

APPENDIX

, OS/ ^(/u/o

C O? St 0/ £-

siz/o/!
000 e oooiT 000/

> "k -^

S 9> Q;

•—I

o
u

o
o

V.
o

• •

H
H

APPENDIX

289

171

■d

m
m

IVi!

ft/5

V^
At

kli

3 O

_ i

«•

i^A

/?/^

ere

9/fi
trie

■M

'2.

J!!^

*fc

r
1

■^

«i

41
4

irffi

1?^

•*

i!*^

^~"

•^

!t!^^

1^"

^^-

■^

S£

££

us.

ug/

1 ^'

fffi

££j

•

■■n# ** J

'•^

V %» ^ j«,ibJ^ 1

£/

.E^

rr/

lie ARC CUTTING £, i^BUOINiT CO.

EMM

^9^

INDEX

AbeU, W. S., 175, 176

Air pressure for welding machines,

40, 44
AUeman, Gellert, 200
Amer. Inst. Mining Engineers, 126,

198, 203
Angle of bevel, 109
Annealing, 2, 3

of steel, 11, 12
Archives des Sciences, 213
Arc, characteristics, 201

long, 148, 207

short 128, 148, 201, 206, 234

starting the, 228
Arc welders, 131, 134-5, 137-8, 140,
142

protection of, 146

lessons, 148-159, 224, 227
Arc welding, 180

alternating current, 116, 208, 211

apparatus, power for, 289

beads, 148, 229

boiler flue, 246

carbon arc, 17, 248

cast iron, 123, 250

cost of, 166

current, 116, 120

design of weld, 267

electrode position, 149, 234

examples, 100, 103

flexibility of tools, 98

in repair work, 97

kind of weld, 269

layers, 110, 236

manufacturing, 101

method of assembly, 121

Arc welding, miscellaneous jobs, 243

oil tanks, 162

overhead, 157-8, 202, 209, 211, 238,
253

padding 153, 156, 231

position of woric. 111, 128, 151,
156-8, 268

preparation of, 233

pressure work, 240

railroads, 99, 220

rate of heat, 149, 155-6, 199, 201,
207, 212, 235

ship joints, 163

spreading, 152

steel castings, 248

steel plate, 233
symbols, 137, 264, 271

theories, 197, 203, 210, 213

thin materials, 155, 239

time of, 128, 166

type of joint, 265

type of weld, 270
Assembly methods, 121
Autogeneous welding, 179
Automatic arc welding, 125, 203

Beads, arc welding, 148, 229
Boiler flue welding, 246
Brazing, 179

Bureati of Standards, 147
Butt welding, 14, 180, 182

Caldwell, James, 165, 168
Capacity of spot welding machines,
44, 48, 49

291

292

INDEX

Capacity of transformers for spot
welding, 46

Capp, J. A., 180

Carbon arc welding, 17, 248

Carbon content of steel, 5, 9, 183,
198

Cast iron, 3, 123, 250

Caulking rivets with arc, 246

Chemical and metallurgical en-
gineering, 181

ChicagOi,* Rock Island & Pacific
Railway, 99, 220

Commercial Museum, 73

Comstock, G. F., 198

Conductivity of heat, spot welding,
194

Connections of testing instruments,
67

Costs in repair, 100, 166

Cox, H. J., 58

Current, arc welding, 116, 120
in spot welding electrodes, 43
large spot welding machine, 46

Design of ships, 167, 170-1, 173, 253,
285

Design of weld, 109, 267

Diameter of spot, 192

Direct current, 224

Duplex welding machine, descrip-
tion, 44

East Coast Inst. ^Engineers and

Shipbuilders, 176-6
Electric Arc Cutting and Welding

Co., 159, 288
Electrica;lly welded ship, 16b, 161,
162.

design of, 167
Electric blow-pipie method, 21
Electric weldability of steel, 10
Electric welding, processes of, 14

Electrodes, arc welding, 148, 201,
206, 212

bare and covered, 19, 112, 129, 151,
152

coated, 115

composition of, 108, 128

current, 116, 120

gaseous flux, 20

liquid flux, 20

position of, 149, 234

practice with, 227

size of, 120

specifications of, 108, 221
Electrodes, spot welding, adjust-
ment of, 45, 95

cooling of, 45

current in, 43

for spot welding machines^ 42

protection of tip, 92

Weed, on design of, 43
Emergency Fleet Corporation spot
welding machines, 42

designs of, 170

requirements of, 47

schools, 136-7
English coasting vessel, 176
English cross-channel barge, 163
Escholz, O. H., 203
Exhibit of welding, 73
Experimental spot welding machine,

Franklin Institute, Journal of, 200
Frequency, 117

Gaseous flux covering, 20
Geary, Dorothea M., 163
General Electric Co., 18, 28, 29, 34,

43, 46, 59, 104, 147, 163-4, 180,

198
Generator, adjustment of, 226

operation of, 224

INDEX

293

Hagenbach and Langbein, 913
Hamilton and Oberg, 1, 22, 27, 29,

30, 36
Heat, rate for arc welding, 149, 166,
156, 201, 207, 212, 235

conductivity for spot welding, 194

theory of spot welding, 180, 186
Heaton, T. T., 113
Holslag, H. J., lessons of, 148-159

theories of, 209
Howe, H. M., 2, 5-7, 10-12, 185

effect of carbon, on weldability of
steel, 10
Hudson, R. G., 211

Inspection of spot welding, 91
Institute of Mechanical Engineers,

London, 113.
Investigations suggested, 133
Iron, production of, 2

welding of cast iron, 123, 260

wrought, 4
Isherwood, J. W., design, 173, 253

Kjelberg, 114

Lafigbein, Hagenbach and, 213

Lincoln Electric Co., 97, 224

Liquid flux covering, 20

Liston, John, 163

Lloyd's Register of Shipping, rules
for welded ship, 174, 276, 285
ship's parts welded, 161, 218t, 256
spot welding tests, 58, 192
uniformity tests, 80

Long arc, 148, 207

Materials, cleanliness of, 75, 129
fracturing, 193
position of, 151, 156
preparation for arc welding, 233
preparation for spot welding, 94

Materials, thin arc welding, 155, 239

thin spot welding, 24
McClintock-Marshall, tests of spot

welding, 49
Merrill, W. L., 34
Metallic arc welding, 18, 113
Metallurgical theories, arc welding,
197

spot welding, 181
Miller, S. W., 186, 188-191, 198
Morton Harry D., 126-6 203

Nauticus, 170.

Nitrogen content of steel, 198, 201,
207

Occluded gases, 199, 201

Oil barge, 165

Operation welding generator, 224

Operators, training of, 131

Osbom, J. A., 29, 30

Overhead welding, 157, 202, 209 211 ,

238, 263
Oxidation, 148

Padding, arc welding, 153, 156, 231
Physical theories, arc Welding, 203,

210
Pig iron, 3
Pinch effect, 213
Position of weld. 111, 268
Power for aro welding machines,

289
Pressure welding, 240

Quasi- Arc, 113

Reactance in spot welding machine,

39, 44
Regulating panels for spot welding

machine, 46

294

INDEX

Reinforced weld, 107
Resistance welding, 180
Riveting in shipbuilding, S2, 36

comparison of, 72
Rivets, caulking with arc, 946

shearing stress of, 89
Rudder, W. E., 198
Rules, Lloyd's, 174

Seam welding, 17
Schools, 136, 137, 140
Ship construction, designs of, 167,
170, 173, 263

5-foot spot welding, machine for,
56

pounding test, 62

spot welding in, ^, 37

spot welded floor, 60

transverse plating, 171
Shipping Board, reports to, 165
Ships, electrically welded, 105, 160,
161 169, 165

combination joint, 163

welded motor boat, 163

English coasting vessel, 176

English cross-channel barge, 165

first welded motorboat, 163

Lloyd's rules, 174, 218, 276, 285

oil batge, 165
Short arc, 128, 148, 201, 206, 234
Slavianoff process, 18, 113
Slocum, A. W., 210
Soldering, 179

Specifications for electrodes, 108, 221
Spot welding, 14, 15, 182

application of light spot welding,
27

diameter, 192

fundamentals of, 24

heat conductivity, 194

in ship construction, 29, 37

Spot welding, inspection of, 91

making single spot, 47

of railroad cars 29

pounding test, 60

preparation of materials, 94

ship's floors, 60

surfaces, 76

tests at New York Shipbuilding
Corporation, 30

test results, 63

theories of, 180, 186

theories, metallurgical, 181

thin materials, 24

uniformity tests, 62, 80
Spot welding machines, arrange-
ments for tests, 51

air pressure of, 40, 44

capacity of duplex, 44, 48

capacity of small machines, 49

capacity of trens formers for, 46

current in electrodes, 43

current in latge machines, 46

electrodes for, 42

experimental machines, 34

5-foot portable, descriptio;i of, 6G

for shipbuilding demonstration,
56

high tension supplied for, 57

making single spots, 47

operating voltage, 43

reactance, 39, 44

regulating panel for transformers
for, 46

small portable, description of, 43

tests, 49, 61

transformer design, 42

transformers for, 46

voltage, drop of, 68, 87
Spreading, arc welding, 152
Starting the arc, 228
Steel, 2, 3

INDEX

£95

Steel, annealing of, 11, 12

carbon in, 5, 9, 183 198

chemical constituents of, 5

crucible, 4

definition of, 2

hardening of, 11

joining of, 89, 22S

manufacture of, 4

nitrogen in, 198, 201, 207

open hearth, 4

plates and shapes, 5, 105

tempering of, 11

weldability of, 8, 10
Steel casting, 248
Strohmenger, Arthur, 113
Surface for spot welding, 75
Symbols for arc welding, 137, 264, 271

Target keel, 166

Tests, arc welding, 174-5

connection of instruments, 67

Lloyd's spot welding, 58, 80, 192

of spot welding machines, 49, 51

of welds, 118

pounding ships' floors, 62

results of spot welding, 63

ships' plates, 105

tabulations, 64, 66-9, 70, 74, 76-9,
81-6, 88

tank welded 122

uniformity, spot welding, 62, 80
Theories arc welding, 197, 203, 210,
213

electrical, 206

heat 180, 185

metaUurgical, arc, 197

metallurgical, spot, 181

overhead, 202, 209 211

physical arc, 203, 210

spot welding, 180, 186

vapor, 213

Thum, E. E., 181

Time of arc welding, 128, 166

Training of arc welders, 131, 134-5,

137, 140
Transformer capacity for, 46

design for spot welding machines,
42

regulating panel, 46
Type of weld, 270

Vapor theory, electrical arc, 213 '
Voltage, across arc, 288

drop for arc welding, 205, 211-12

drop for spot welding, 58, 87

impressed, 128

minimum arc, 206

open circuit arc, 208

operating, 43

striking, 113
Voltex process, 22

Wagner, R. E., 18, 104, 122
Wanamaker, E., 99, 116, 220
Water-pail, forge method, 21
Weed, J. M., design of spot welding

machines, 43
Weld, design of, 109, 267, 270

ductility of, 123

position of. 111, 268

reinforced, 107

strength of, 123

testing, 118

Weldability of steel, 8

Howe on the, 10
Welders, arc, 131, 134

fitness of, 142

lessons, 148-159, 224, 227

protection of, 146

selection of, 134

training of, 131, 134-5, 137-8, 140

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Spot and arc welding"

See other formats