’0

 

A STUDY OF WELDING AND WELDED
DESIGN

Submitted by
Philip Ii 11131:: “gagglmg

TH E515

 

m; 4 '57

ACKNOELEDGHENTS

I wish to express my uincaro appro-
oiation to Dean Dirk: and the Engineering
faculty, at liohigan Stat: College. for
their thoughtful and friendly counsel and,
ospeoiully. to profeusor L. 0. Miller for
hi- hoarty 000peration. and in u like man-
ner. to Hr. Bart Sangster for his willing

and helpful advice.

26483 4!

REFERENCES

The Procedure Handbook of Arc Welding
Design and Practice.

Lincoln Electric Co.
Designing for Arc Welding.
Lincoln Electric Co.

Manual of Electric Arc Welding.
E. H. Hubert

Arc Welding and Cutting.
Westinghouse Electric Co.

Manuals and papers frcm,

Air Reduction Sales Corp.

There are six general Olnqans of welding,
namely:

(a) Forge woldinr

(b) Therrit welding

(c) Resiotonce welfiing

(d) Goa walling

(6) Atomic hydrogen welding

(f) Electric arc welding

Thane varioua proceaaos will be diacuooed
briefly so thot the rolntir: im~ort1nce of each
can be better underatfiod.

Forge weliins iv just whtt ’ho name states
that it in. The metah in hanted in * fire or
furnace until a fuoicn to yornturc is touched and

is than withdrwx on: forged until the field is com-

L)

plated. It was “9E3 ottnnoivoiy by the oli time
blacksmitha 9nd 3123 in many iniuvtriol pltnta, but
'9 *hc nolfiinv method: have ir¢1¢voi it hon been
largely re locefi Fy coma othe: metficfl. The coat Cf
forge wo111n~ is hith anfl the procoq is slow "a
cowoored to the other tyroo. Howovmr, thorn ore 1
{on scecifil fielio where this method is otill used
anoneaafully.

Thorwit molding i9 uaod to repair on} figgewhlg
heavy p1"t3 ouch on rails <nfi r ilwuy Equi;m3nt.

It 13 coco filiahed by huilfiin: ~ mold ar:uni the

pl*cc to be welded and than filling th mold
With molten steel. Tho stool fusna on to tho
frocturod momberc and aolidifioc, rcculting in
o goofl union betwccn tho ports. The molten
steel is derived from a chooionl reaction be-
two n iron oxifio an: gendered aluminuo. The
two are mircd and ignitofi, canning the oluminum
to combine with the oxygen and leave the reiucod
atoel at a very high to rercture. This high

to oorcturc is nocecafiry to mtko sure thit good
fuoion will rooult. Tho ayrlicntion 0' this
mothofi is r=thwr cu *oraoo but is uno’ui for
some kinic of heavy work.

Rocictfncc weldinr m he: uoo c” on electric
current to heat tho joint to the {union point.
Tho current in pqoooi through the piece: to be
wcliod, caucing a high tcmyernturo in the joint
duo *o the high rcaictnnoo. The pieces ore then
precaod togethnr anj fillOEod to cool. This rré~
02cc is very fact *ni woken reasonably good wrldo
at a low cont. It is particulqsly guinea to butt
welding, on! 53:10? wail-3M", ciao some kin}; of 8"".“..;t's
wells. Tn main diff culty is that the coat of
the equipment is hich wni ucuollv boo to be c»—

aifinrd to suit the job. This fact limits the

proc’fi": t3 proa‘13t:¢n u;~_:‘3' 6103273 f3:- al fee! gmgrgl

0

~ ‘1399.

Gen welding is dona with e hivh torpernture
flame directed by a specially deairned torch. The
high temperature is produced by 83$“ly1n: the gases
unier precaure an: ucin: pure oxyren for the fuel
gee to combine with. The fuel is unufilly acetylene
althouoh hyirogen is cometimoe ueefi. The weliing
is accomplinhod by hc-ting a portion of scam or
fracture until a ”gudile' is produced and then
feeding in the filler motel from a too eenecielly
for this yurposo. This proceee is ueei more than
any of Lho oth r exoupting the tleotric'nro. It
is very flexible uni hxe c 10* first ecst, but the
coeretina cont 1! rather hich. It has some oyeciel
eiventeree in eenll work such on tubing on! thin
cheete, alto it require: no outeiie poeer of ony
kind to uze this equipment.

The game torcheo that are used for Voicing
can be fittei with en eiertor and ueei for cuttinc.
Special cutting torch a are nleo combined With
controlling mochvniema to use on {reduction work.
Flame cutting is a very haniy method to use find
is often the most economical too if the shepeu
to be out are irregulor. This process in used
extensively to cut cnd trio ports f0? wolici fob»
ricetion.

Atomic hydrogen arc weliinz is ver= similar

to gas welding as far as the manual part is concerned
but tho heat is proiucafl ty on alternatinz currrnt arc.
Two tungsten electrodes maintain the are independent
of the piece to be wolied and a stream at hydrogen gas
is passed through tho arc vni on to the place to be
welded. The heat in the gos atrcof is intence and
hence waldo very fast. This method is more ooatiy
-thqn arc waldinz and iq not used a gte't deal.

Electric arc welfiio: is the most widely uoed
of any of the ueliinz methoia in use at the present
time. It is cheaper cud faatc“ for mcny kin‘a o?
work and con to aiapted to al oat any coniitiuns.
It is comparatively new in its douchpment but'haa
mwic rapid advancement in tho last ten yours. The
proccsa is suitcblc for either monutl or automatic
welding and “193 eithcr a curb n or w mat llio
electroda.

The bore metallic yoi aq* uacd alwcst altrgether
at first but is no: largely realmccfi ty 3 covvred
rod, canning a saving in time and also hattor welda.
The bare rods will give acids which are were porous
and not a3 strong as salts made with o ahiclflod arc.
This ia Sue tovthc att3ck of the air on the unpro—
tected cctel uhil? in the uolten state. This difficulty
is largelv overcome by the covered rod by provifiing
an inert atmosnhare rr uni *h3 are and alao the dc-

roaition Of 3 1???? 0? 31%? over the weli metil. Sowe

of the slag gets mixed into the weld but it is still
more olefin than the mctcl deposited from a bare rod.
The biggest advantages 0’ the shield are are th~t
it is fawter, due to th: elimin~tion of 'blo¥,' find
th‘t the slag keeps the metal from cooling cc fast
as the unprotected metal would; also the comgcciticn
of the weld metal is more nearly the same 33 the rod
was hecauce the air has hai lacs chance to use up some of
the conntituents.

The Carbon rod is used a great deal in automatic
machines cni 3110 on sheet metal for manual work.
It can be ucefi with or Without a filler roi éepcnding
on what the Job requires. The plain carbon arc is a
cheap method find is satisfactory for com“ kinia of work.
It can be made suitable for almoat any work if an
nutczenizer is used. Thic it a material thct oo'ts
much as the ccverinc on the electroie does. It may he
cithcr a powder, a paste, or a prepfircd corfi. A1-
thwuzh this mat a wold: which are suitable to most
work, it 13 limited to hori'onz l welfiing because the
molten metal runs off of a vertical or ovcrhead w ld.
The coat of using 1 tillcr rod and an autcgenicer
makec tho shielded carbon arc almost cs expensive as
the motulic arc, but it is more easily used with auto—
matic machincc.

The chcrcctcriczics of the welding gcncrctor

have an import nt bearirg on the eaae "oi speci of

making a good weld. 'Chtrocterietics” is me at

to include such factors as the shape of the volt
ampere curve, th» eaee with which the 810'6 is changed
and aloe the range. Other feature would be ltability
or the case with which the machine oouli handle

oufden change.witbcut causing dioturhoncea in tho

erc and the cite of the machinW. The size in in-
tended to mean the current rouge of the generator

and not the dim noions

The shape of tho curve has much to do with the
ease of striking an arc and also on the kinds of work
which c*n be handled by the generator. For example,

a hich cprn circuit voltage is necessary for winiy
conditions: or where blow is encountered. A lower

Op n voltere anfi a flttter curvc are used for vertical
and overhead welding. Thin 1" because it ie.necesaery
to hold a short orc wnfl still hove late of heat to
cause good penetration.

The controls on ;he machine should to handy and
eaoy to Operate cni should give on indication of r nae
of work thot c n be hcndled at any given setting. Th9
size of the machine must be suited to the gonertl run
or work which it is exyected to handle, that is, a
shoot metal welder con not be used on heavy work and
vice versa.

The prone: current find voltage have a definere

effect on the weld metal. They affect the ease
and retifiit* of devisition on! also the phyaiool
prepertinc. Weld metal that is overheatefi will be
burned ani brittl on: metal which eaa not hot
enouzh will have poor penetration, both resultin?
in p’or welds.

There are men? varivtiona in the types of
jointa used, but moat of then, exooyting the plug
welds, are similfir to one. or a combinotion. of the
typoe shown in figures 1, 8, and 3. These types
were inoludei in the test specimens so th't a com-
.1riscn of :h= merit: of each might be mode. The
different typea that were broken are listed in the
data by the numb r and letter of the Joint to which
they are similar. Thin {not make: it easy to cam—
pare the average strength of each type. in examination
will show that type No. a (a), gevo the meet oon~
eietently high results, with Ho. 3, an the next most
'ooneietent. although not quite a: strong. No. 1,
includes the whole refige from the poor Joints to the.
very best, depending on the quality of the weld.
This make. the joint rather uncertain unless i good
welder does the Job, followed by a ridgid inspec-
tion. Under these conditions the joints are eonsie—eo'
tontly good and very strong. They have the further
advantage that they will work into almost any kind

of design while the lap joints will not.

The lap joints all broke in the parent metal but
the side a lded had a tendency to ten: near the weld
and thus lower the strength somewhfit. Th so Joints
are only cuitshle for narrow pieces because the edges
are all thot is held find could not sup ort s wide
Joint euch as a plate. They are used a great deal
in etructursl work, in a modified forr, for holding
the stay bracing on to the he"vy members.

The lop Joints with the fillet welds gave good
results and broke the parent metal. st the ultim~fie strength
of the etr p, severe- inches sway from the weld. There
is st lenst on" precaution, however, which should be
considered in lap welded Joints; nnmely, Penetration.
in are which will give good results on butt welds is
often not hot enough for lap joint welds and, conse-
quently. poor penetration and e wear weld will result.
This is due to the greets: volume of parent metal which
must be he~ted because of the double thickness.

Before Le‘ving the subject of joints it would bo
well to say something about 'end cffeot.‘ This is a
term used to designate the difference, in structure and
propertiea,in the weld metal where the ore is storted
and where it is st03pod, es compared to the middle part
of the Weld. Thea; eni portions ore usually much weaker
than the rest of the weld and, therefore, fail first.

This p*rtisl'brestdown puts a heavier lood on the

remainder of the eel? find a failure is quick to follow.
The causes of 'end effect' mey be mony, but the meet
cbvioue ere probably the cost imoortent. When the

arc is first struck. the electrode aterte to deposit
but the parent metel ie lerger uni has not had time

to heat below the surface very much. This makee a
Joint which d-penda more on the nth eive quality of the
bee? than on the fusion into solid piece. Obviously.
it ie not very etronz. When th* arc in extinguished.
at the end of the head, it always letvee a crater which
is very hard to fill up complete.y. The metcl in the
crater 13 extcccd t3 the 'rdy' QtTOQFhGT? beccuec the
protective entwlope of inert gases furnished by the
electroie is no longer present. This alloee e deterior-
ation of th'5 metal and also allows it to cool repiily,
thus being here brittle than it'ehould be. The
cumulative effect is that this port will not stand
bending because it is too brittle end that it has 3
smaller cross section and is e130 weaker than the rest
of the weld. Th laet end rf the weld is usually the
weekeet pert in the whole beai.

End effect is of smell imccrtcnce in long welds
because the retio of the length of poor bend to the
length of good heat it very email. As the weld becomee
shorter, however, the ratio gete lwrger and more im-
rcrtent. In very short welds it ie a major factor to

conei?er and may make it neceeaory to use a special

arrangement to get dependable rewulte. It should be
noted that a.l of the pieces broken for this paper
were of the short weld type and, therefore, the results
represent the least that ehould be eXpeoted from average
welding, and not the beet. It is entirely possible to
break the parent metal every time if a long joint is
welded and the test etrip in taken out of the midile.
Good preparation of the work is just as im ortént
to good welds as the welding operation itself. The
parts to be welded must be clean and nooeseibie to the
arc. Oxideq, dirt, end oil or water will often spoil
en othe wise good job. If the weld doesn't have to
be of the very best it may be ohe~per and faster to
burn through these obetnolee than to remove them, but
this is not usually the oeee. Special care must be
taken to make lure that the are will reach the bottom
of the joint or poor penetrntion and even unwelded
areas will result. This trouble is removed by placing
the ends of small pieces a little ways apart at thwt
the heat of the arc can act t7 the whole area of the
pieces. Large pieces must have the joint edgeo out
back at the outer surface as in the shape of a 'V',
so thet\the electrode can be held ne rer to the bottom
of the joint to meke sure that the penetration is gooi
over the whole arei. Severq- be1i3 ere often laid

into a Joint of this kind if it is large enough to

require that muohfiller m tel. In that case it is
necessary to thtroughly clean the 3123 from the
preceding layer before the next is put on or there
will be large voids in the weld metal and the re—
eultont low strength. 81nd blasting is n very quick
end thorough way of doinu this but more expensive
than the usual method of merely chippinr it out by
hand or an nit hammer.

The right type of ore must be very carefully
chosen to suit the work. This is s.connon stumbling
block for the inexperienced, or poorly informed.
welder; sspsoislly if several machines of different
types sre used. The different neohinss nuke the
prOblcn more complex and confusing because they src
likely to have different characteristics and bshsws
quite differently for the same 'nc lond' settings.
If the solder is familiar with the Inchins he is
using it will eliminate that source of uncertainty,
but he sill still have to choose the right type of
src if he expects to get good.results. tbs in‘
stance; it is well known that s.hsary job requires
more heat then s light one. and that on src which
is subject to 'blow' must hows n‘higher open voltage

than one which isn't.

These are general rules and wouldn't be of
much V1.ua unless the welder had had eXperience
with such mr‘r and kner many othere rules beside .

The arc length central: tho voltige drop be-
'tsaen the electrode and the work. This factor is
important in aevwrn; different ways: Fir*t, it
has a bearini on the quality tni rapidity of the
weld; second, it influence? penetrntirn and
grotto: loos; third, it mun: be united to the
electrode and the 50h. II! a ehort arc is used

th hoot will be concentrwted on'u smaller area
than it would with a longer arc. 'This will do-
pooit metul factor and make it possible to use
lose power ena still have good penetration. The
smaller amount of pow r is desirable both as o
SCVIRQ uni a; protocti n againat worrinr. The
short are 1e yes the molten metal in contact with
the atmosphere 3 shorty: time and cuts down the
deterioration fro: this a urce; it ale“ nroviiez
u more direct path for the met 1 to travel vnd
‘doesn't =llo: as much :3 upotter off to the sides.
It would b: of emall value to any that n short arc
in hrneficial unleoe we underatood whet woe mount
by a short ere. ‘Generwlly a short fire is token to
moon the are which giyea the bent results with the
electroie which is beinu uqod. Lorge rode use

longer gape then smart rode becauae tho are in more

forceful and the glovuleo of metal are larger
end all; 'hridge over' e smell g~p. Covered rode
also use somewhat lonq r arcs than here rods. The
reozone ore that the shielded arc is protected
from the air oni also the coverinq directs the
stream of motel so that it doecn't blow a; boily.
Perhqpe thie is a wood place to mention that
are blow it a megnetic disturb nco which so for
hag been 31 not imroasible to control. It 19
worse on ho'vy pieces than on lirht one: and Gn~
peciolly near the lust end of the weld. This 1%
probably due to the building up of the magnetic
field no the weld proceeda. The quality of the
rod (bore rod) seems to have on ePfect also. Some
roia have some foreign ingrefiitnta which cause
trouhle; theae are probably in the nature of was
inclusions Which expend when h eted 3rd coueo file—
turb nee. An alternating current "rc has the
adrentoge over a direct current arc that it is
not bothered by blow whatsoever. The direction
of the current charreo so r“:1dly that a magnetic
£191! doean't have time to build up and dance
trouble. There welders deforit e rend cieen bead
find give exce-;ent penetration. They have the
further fi3V“nt1?€ th't they work directly throuzh
a trcnoformer instefifl c? 1 motor-generitcr set.

This gives thea a much better efficiency but also

introduces n atone: which mwny‘weidc;s are afraid
0’, namely; the roanitility of 1 hre~i46wn in the
transformer inoulwticn rhich wouii allow the ful.
Ila" voltfigr t“ tench the operator. This fiance:
is small cut 13 really not ih~ moin diariVQntnge.
Th main troubio; ore that they are only suited
for gown weldinf uni c n't be used on a portable
outfit.

There are three main trpce c9 103?”Ot10n for
relied jointa; Vieucl. eieo‘r“-meznetic 1ni X-rey.
Vieuwl inspection is diviied into inorecticn durinfi
welding ini insgscfion after welfiin . By cheervinz
du;ing the welding procec? en_ exo rienoei in-
egec2_r can tel- consiforo‘ly “bout the condition
of the jfint, the penetration oni fucion chi the'
fovm~tion of the hood. The fact: find *ha eouni of
the arc, no wrli no tho rotc o” burn off of the
electrcflc, till enéhlc him to decifle whetho” the
weli it satisfactory or not. The intraction
oft‘r Weldin' shouii note the size and shape of
the hewi, the icootion t? cratera "ni any undercut
or overlap, 13 Well as the density of thn role
metal. Undercut week he the joint and ovorlng
ie 3 eirn of soot penetration. I? the overlno is
very bad it touii he weilto take a blunt nonefi tawl
ant chip olong the edgee to see if the weld metal
con be knocked off. The locotion and number of-
craters will tell something about the welfier‘a

chill and as wee mentioned earlier thfi longer

bands will hnve less 'end effect.‘ It is not
necsas ry to inspect every joint excegt on very
exact werk. For the most part it :11. be suffi-
cient to pick on; " representative number and
judge the job on the basis of those reeulta.
Electra-magnetic inareotion makes use of
magnetic field uni iron filings or an in-
stxument o pfibLO of detecting chnnqae in the fiux
density. The iron filinga are not sensitive to
fiaxa which are dee: *n the W‘ld an: the surface
must be horizont l nnfl emortd if this method is
uaei. The instrument methni has not been usefl
much teceuee the instruments ere extensive and
eeaiiy damaged. The X-rey method is the most
meantwte but is 3190 very eXpensive. It is need

em very im-ortent work such a; pressure veaaeis

{9

me 501: Einia of machinery. The K-r~ye will
detect flame which are deep in ‘he metal and could
not he found by any other metho‘.

Another method which 19 sometimes used on Jointa
which mus: be tight against 16“k ye in to mix an
011 sciu bi? dye in kerosene ani paint it on one
surfwe' of the weld. If the kerosene own penetrate
thraagh the weld 13.3111 stain the ether 31*e and
she-'7 the iOO'tirn of the firm.

The atrag steel need in the te~t piecea ilsted
1n the data was of e goofl ocmmerciel Ernie. One

FIEOB (place H”. 1) W33 bréken Without 1 “614
so thqt fihfi “9149 00911 be onwpered to the strength

of the steel; and else to determine ch“ elastic
limit find yield point. It chauld be noted thwt
even the pooreet welds broke above the elaetic
limit and that sixteen of the twenty-two welfls
brcke above the yieli paint, five cf which
equcled the ultimfite strength of the cteel.

Pieces Nos. 3 & 3 were welded cn Both sides
(1: figure No. l ('3) with 3 coverefi rod. Th9
average strength «an about 53,000 lbs. per sq. in.
sci g:ula have been 5 little hi her except for
the 'end effect' on such short wclie, as m “tirned
earlier in this paper. Pieces No. ll & Ho. 13
were made by e ccmmerci"l eelie 4n? were the same
93 R03. 2 & 3 excep: frr the tinker currcnt used.
Theee piecee broke at 53.3fifii its. pr: 31. in...
p'ficticnlly the ultimflte strength cf the steel.
These result: show that e butt jcint properlv
wc;dei from bctn si’ee will be we eetiefnctfiry an
q @0115 pi cc if strength is the main concern.
I? the uelfi must itenfi shock er bending it will
scmetime- be necessary ‘0 h? t tr’et it. but
usually not if a gccfi weliin~ prccedure ic followed.
Figure No. 19.92cwe the m13ncfied crass secti-n
of e dcubl“ welded butt j~int find figure No. 4
share a arcphic cu'line of th" same thinfi.
Firure No. l0 is e micrograph of welfl metal as it

is de.ceited in w single beol. The structure is

martenaitic and “rt to L9 brittle

cl in t.13 coniiticn wzrlfl be.
shows the acme weld metal, insiéc
zone, fifter the be“! Has been Nu:
siic cf the jci.t. The he‘t fro;
hP-fl ~nnc ls ’He steel in ‘ amfill

‘ le'ves a much h~ttcr,

genscua structure. The grain size
first taxi in: the edjiccnt pfrcnt
fine} :ni LOIQ ductila. F aura c

parefit fietfil a3 it i? ”utaiie of t
Zine. Theee gicturet help TC Gin
1: so strong in the fusion zone in
93 right be exrccted. The fiflCGS

brckfiu For tfiiq re “It showcfi thtt
cc:urrcd in tha welt mata; or in
an} practically never where the tw
picture; 313% Show that the use of

or seccni p“33 has a beneficial cf

anj Will lc've it in trughcr, more
This trcfi mant accomrlichcs alflott

th~t q hart treatment veuli and it

quicker. It has at laa~t fine diqa

tha h at ttfatmcrt, hcwcver, ncmcl

give ne'rly 39 5003 stress relief

QT
. k

a a uncle becauge i? is too loccl

he

as any c031

Figure NC. ll

of the influence
or. the ether
the accnnd
area arcund
tore hemo-

in 70th tha

m

DA

tal 13 re-
9 shows the

he influence

win why a wrld
etctd cf wefiker
WHICh were

the Frctkg uzuwlly

cfirent ctnl

:-

3 grined. These

'I

'wash'
fact on the weld
useful ccndition.
thc axmc rQSUlta
is chcéper wni
dvzntage cvcr
jccsn't

it

Yo

Pieces No. 6 & 7 were d’ubie welded butt
joints filBP, but bare rod W19 used. The data
shows that bare rod #111 give quite good results
if used by a tnoroughly ex crienced welder, but
if net it is likely to be inconsiwtent and a
90‘5 factor of eafnty should he need. The main
objecti‘ns tc the bare rod are thfit it is harfier
ta us” eni in like v to have more inclueions
and flees in the weld metal. The head is not
protect-d by a seating of slaw we the coated rod
head is “n4 conqequentlr coolq much teeter. This
factor probably has much to fin witq the brittio
condition ani cracks that she: up in bare rod welds.

Pieces No. 8 & 9 were butt joints welded on
one sifle with covered rod. They were ibout 801
as strong as the dcuhie welded joint: and need
practically the same amount cf red but only took
abnut two thirda 9*: lone to make. Pieces No. 4 8: X5
were the same excent thfit bare rod was usefi. They
show a marked reduction in strength, even as com-
pgrefi t. the reduction in the case or the double
weided jointa. Thane rnaulta seem to shcw that the
bnre rod is much more in nePd of annealinq than the
slaw protected covered rod an? ihqt a 'wesh' be“%
1% decideily beneficiai.

Piecas N0. 14 & Nd. 15 were forgnfi a9 the
welds were ooclinv fin? showed ub*ut 5% increaqe in

strength over pieces Noe. fi-f‘3 made at th* same

tine :xcept unforged. Fijurea Nos. 5 i 6 Show

tee views cf rne of these forged welde and

figures Ho. 7 & NC. 8 Shea two views of 1 weld
which eqan't forged. Nate the even em*oth
fracture and the small unfused'epwcea in the
center of the forged weld as crmpéred to the
other. The 'end effect“ is also quife npp~rent

in beta welds; this illustrates the difficulty
that we: discussed efirlier in the report. The
hifher strength nnfi finer grqin site in the forgwd
weld is due t4 rh¢ 301* we kin* the metal received
5ft r it hefi created the lower critical temperature.
This exqule shnws that pa¢nin¢ " ”elfl has won
beneficial effects; but unleee thw weld rod used
ie suitabl», the gain in strergth will have to

be balanc d asainet a legs in tou;hneee and duo*«
ility 29 shaft by vessel frficture in fhe unforged
well an! Lhe smvcth even fracture in the forged
piece.

Pieces he. 13 & No. l7 me e 0? type 30. 2 (a),
lea joint: weliei on both eiie~. Thfiy broke the
teat bore abcut three inche~ ~¢cy from the weld qt
the ultinwte strength of the steel. En bad fi”Tcote
were notice xi h these join: wuc: 33; twistina
o: tearing. The main chjectinne efiuld prohcbly be

t‘ §

the hulkinesa ani the off~ae "g tne membere.

t"

Pieoeq N0. 17 & 19 were thfi sr-tu‘n‘a 33 Nos. 13 & 17,

except welded only on one side. The strength waa

much lees (about 70% th—t of the double welded
joints) but only half as much rod wee us d.
These joints would be stronger if they didn't
dintcrt when the load is npylied. The off-eat
causes the free end to curl rut away from the
Other menher nnd puts a hendin» streee an well
an teneion in the weld. Thin typ~ in quick ani
cheen to mek , hOWBVfir, «n1 wouii answer well
enouqh for m"ny purposes.

Pieces No. 33 & no. Bl were lap jointq of
th” qifie welded type. These pi oee hrnke in the
strep nt nb‘ut 9"; of the ultimate etrnnzth of
the steel. The break occurred tirht us t" the
joint ani wa~ caused b « teerinv not! n, or
heniinr atreee, iniuced by the tendency of the
atrers t“ pull int‘ 2 straight line. These
join" are th* enaiest tn make on amn-l atrrpa
heoquee the trruble of hevin: the arc blow through
the str"" at the edges is elimin ted. If their
limitfiti;n (93% ult. strength) is reaii‘ed, eni
allowefi for, there shouli he no serious objection
to the use of this type.

Pieces 3. 2? & Ho. 93 were welded by
acetylene simply to show 'hnt this method will
also product good welds. However, it isn't 33'
fast or cheap an 1hr are; even allowine for the
loWer first coat. Aoetvlnne eeldine is prob~h1y

the neat neeful for weldinz light gaune metel and

is preferred over the are by many welders for

thia we lio~t§afi.

This report falls far shcrt of covering
the whale subject of wexding; but it is hoped
th”t it is at 1 ~2t a step in the right direc*
tioh qnfl may 3?rve an arm' sm~11.1rd103tion of
a fienernl survey "n6 perhqpa be an 311 to any-

one who wishes to use it.

 

 

 

 

(a)

 

 

Z?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(b) O
NO. 1 A butt Joint weld.
is Welded on both tides.
b fielded on one side.
I
f
l
(a)
L (b)
No. 3 A lap Joint weld.

a) Welded on both sides.
b) Welded on one side.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W
l
l
lb. 3 A lap joint, side welded.
and bond
’13
\ ) a)
parent ,- , pun!“
metal ’- “ - notnl
(A) (A)
/
(4) ‘5) (a)

 

 

 

 

V

1“ bead

lo. 4 Th. orou auction of a weld as seen
under n.310r0300pe.

‘3 Influence ions of the lat bead.
Influence non. of tho and bend.

(see illustration No. 13)

 

 

lo.

 

6

\‘4:H .. . J.

2 '3'. _ . ,
., 03$,” “1.-

‘ 1.?‘72 r
3.1-1: (1....

Ind view of the forged butt weld freoeure

.‘

 

la.

The fracture of e plain but: weld.

 

Ind view of plain weld fracture.

lo.

10

 

Hierogreph of perent eetel. (100:)

 

lierogreph of the let weld‘bead.

 

Iierogreph of the let weld bead
efter the 3nd weld bead bee been
depoeited. (100:)

(100:)

--- - * *mmv -

-W‘ n‘” -‘.—. _ u.
‘ .

.- o-nerg— _-¢.—-._. —--?

 

J

we. 1...}...3.
. r.\..evpuv.\.q'.ilmfm§4u{
r’o,$.lu.lfi..\. y u

.1 C‘” e
h... . .,.r\ L. -

 

 

 

 

 

The crcee eection of a butt Joint welded

13

h.

 

from both eidce. (10 x)

The let bead ie on the bottom ee in Fig.

la. 4.

n? indicated

influence cone?

lotice the

etrncture as

9. lb

4. end the different

in Fig. lo.

ehcwn in the micrographa in Pig

.L
l
n.

D

10

9

The grain etrueture ineide of the influence

Kc. ll.

Fig.

ecne ie practically the came as in

DATA

All welds were made with a 5/33. Fleetweld

’7 rod, 130 amperee. end 30 - 25 volts; un-

leee etated ctherwiee.

All etreeeee

eree.cf the etrep

Gold rolled etrep
Broke at

Ultimate etrength
Elongation
Kinetic limit
Yield point

.type 30. l (e)
Broke et
Ultieete etrength

Type lo. 1 (a)
Broke et
Ultimate etrength

Type ‘0. 1 (b)
Broke et
Ultimate etrength

Ultieate etrength

Type lo. 1 (e)
Broke et
Ultimate etrength

were comguted on the baeie of the

before welding.

Piece No. 1

1/4 0 x 1 1/2 0
33,000 lbs.
A 58,700 lbs./ eq. in.
asepercent
31,000 lbe./ eq. in.
41.500 ' ' '

Piece Ho. 3
13,470 lbe. (in the weld)
49,300 lbe./ eq. in.
Piece No. 3

19,220 lbs. (in the weld)
51.eoo lbe./ eq. in.

Piece No. 4
bare 5/33 ' rod
13,500 lbe. (in the weld)
36,000 lbs./ en. in.
Piece Ho. 5

here 5/33 ' rcd
13.450 lbe. (in the weld)
33.200 lbe./ sq. in.

Piece No. 6
here 5/333 ' rod

16,780 lbs.
44,700 lbe./ ea. in.

Piece No. 1

Type no. 1 (e) bare 5/33 ' rod
Broke et 19,100 lbs. (in the weld)
Ultimate etrength 51,000 lbs./ eq. in.

Piece Ho. 8
Type No. l (b)
Broke et 17,900 lbs. (in the weld)
Ultimate etrength 47.700 lbe./ eq. in.

Piece HO. 9
Type no. 1 (b)

Broke et 17.300 lbs. (in the weld)
Ultimate etrength 46.100 lbe./ eq. in.

.OQ‘OOOQCOOCOOOOOOOOOOOO00......OOOOOOOOOOC.OOCOOOOOOO’

W, made with 150 emperee and 35 volte.
Ecluding piecee Hoe. 10. ll. 1”. 13. ideal-duaele

Piece lo. 10

Type Bo. l (b) . 5/33 ' Tleetweld '7 rod.
Broke et 18.850 lbe. (in the weld)
Ultimate etrength 50,300 lbe./ eq. in.

Piece 30. ll ‘

Type 80. 1 (e) 5/33 ' Fleetweld '7 rod.

Broke at 31.930 lbe. (in the weld)

Ultimate etrength 58.t00 lbe./ eq. in.
Piece 30. 13

Type 30. l (b) 5/33 ' lleetweld '5 rod

Broke et 16.130 lbe. (in the weld)

Ultimete etrength 43.100 1be./ eq. in.
Piece 30. l3

Type I0. 1 (e) 5/33 ' Fleetweld ‘5 rod

Broke at 31.870 lbe. (1n the weld)

Ultimate etrength 58.400 lbe./ eq. in.

‘OOOOOCOOOOO....O..OOQQOOOOOOOOOOOC0.00.00.00.0000300.

Piece No. 14

Type no. 1 (e) weld forged while coolin .
store at 30.370 lbe. (in the weld
Ultimate etrength 54.300 lbe./ eq. in.

 

P10°° Rae 15

Type No. 1 (e) Weld was forged while cooling.
Broke at 3l,l40 lbe. (in the weld)
Ultimate etrength 56,400 lbe./ eq. in.

Piece R0. 16
Type No. 8 (e)
Broke at 22,020 lbs. (in the etrep)
Ultimate etrength 58.800 lbe./ eq. in.
Piece no. 17 ‘
Type No. 3 (e) .

Broke et 31,950 lbs. (in the strep}
Ultimate etrength 58.500 lbe./ co. in.
Piece Ho. 18

Type Ho. 3 (b)

Broke at 14,300 lbe. (in the weld)
Ultimate etrength 38.100 lbe./ eq. in.
Piece No. 19

Type lo. 3 (b)
Broke at 16.060 lbe. (in the weld)
Ultinete etrength 42.800 lbe./ eq. in.

Piece Ho. 30

Type No. 3

Broke et 30,270 lbe. (in the etrep)

Ultimate etrength 5 54.100 lbe./ eq. in.
Piece Ho. 31

Type lo. 3

Broke at 30,740 lbs. (in the etrep)

Ultimate etrength 55,300 lbe./ eq. in.

Acetylene lelding
Piece no. 23
Type no. 1 in) Same etrap and e i 6 tip.
Broke et 81.600 lbe. (in the strep)
Ultieete etrength 57,600 lbe./ eq. in.
Piece No. 23
Type Ho. 1 (e) Same etrep end a i B tip

Bro: at 18,890 lhe. (in the weld
Ultimate etrength 49,800 lbe./ eq. in.

5:va

A

83E

'

.3

3' 1L1? I

'1‘

l
V
r.

r’v m e
v' 1‘ 5

I