006

1*

,“1.

INVESTIGATIONSWELDED FLAT

,

HARMER E.

TECHNICAL REPORT

ON

OF BRITTLE CLEAVAGE FRACTURE OFPLATE BY MEANS OF A BEND TEST

BY

DAVIS, G. E. TROXELL, EARL R. PARKER

ANJJ A. BOODBERG

University of California

Under Bureau of $hips Contract NObs-31222

1’

COMMITTEE ON SHIP cONSTRUCTION

DIVISION OF ENGINEERING AND INDUSti~AL RESEARCH

NATIONAL RESEARCH COUNCIL

fY?TRBLIBRARY

.’

,’

ADVISORY TO

SHIP STRUCTURE COMMITTEE

uNDER

Bureau of Ships, Navy Department

Contract NObs-34231

DATE: MARCH 10, 194S

NATIONAL RILMARCH COUNCILWashington 25, ~. CC

Chief. Bureau of

kiarch10, 1948

ShipsNavy be~artmentWashing~on 25, D. C.

Dear Sir:

Attached is Report Serial No. SSC-6 entitled lTlnvesti-gation of Brittle Cleavage Fracture of Welded Flat Plate byMeans of a Bend Test.1’ This report has been submitted by thecontractor as a technical report on one ,phaseof the work doneon Research Project SIR-92under Contract NObs-31222 between theBureau of Ships, Navy Department and the University of Cali-fornia.

The report has been reviewed and acceptance recom-mended by representatives of the Committee on Ship ConstructionjDi.Ti~ionof En~ineerj_ngand Industrial Research, NRC, in accw’d-a.ncewith the terms ofShips, Navy Department

the contract between theand the National Academy

Bureau ofof Sciences.

Very truly

., .....4---’--)

Enclosure

yours,

(!-+-:w&m”a ‘“;;.-4--:)L%Frederick M. Feiker, ChairmanDivision of Engineering.and

Industrial Research

..+

thisthis

PREFACE

The Ship Structure Committee through the Bureau of Ships is distributingreport to those agencies and individuals who were actively associated withresearch program. This report represents a part of the research work con-

tracted for under-the section of the Committees directive to ‘Idissemi.nateperti-‘ nent information to all parties having an interest in the building and operating

of ships,

The distribution of this report is as follows:

Copy No. 1 -Copy No. 2 -

Copy No. 3 -Copy No. 4 -Copy No. 5 -COPY No. 6 -Copy No. 7 -Copy No. 8 -Copy No. 9 -Copy No. 10 -Copy No. 11 -copy NO* 12 -Copy No. 13 -Copy No. 14 -Copy No. 15 -cOpy No. 16 -

Copy No. 16 -Copy No. 17 -Co$?yNo, 18 -Copy No. 5 -copy No. 6 -Copy No. 7 -copy No. 8 -Copy No. 19 -Copy No. 10-copy No. xl.-Copy No. 12 -Copy No. 13 -Copy No. ~ -Copy No. 15 -Copy No. 20 -Copy No. 21-Copy No. 22 -copy No. 23 -copy No, 24 -copy No. 25 -

Chief, Bureau of Shipsj Navy DepartmentDr. D, :J.Bronk, Chairman, National Research Council

Committee on Ship Construction

V. H. Schnee, ChairmanJ. L. BatesH. C. BoardmamPaul FfieldM. A. GrossmanC. H. Herty, Jr.A. B, KinzelJ, M. LessensG. S. MikhalapovJ. OrmondroydH. W. PierceE. C. SmithT. T. ‘.iatsonFinn Jonassen, Research Coordinator

Members of Project Advisory Committees SR-87, bR-92SR-96, SR-97, sR-98 and SR-1OO

Finn Jonassen, ChairmanR, H. AbornL. C. BibberH. C. BoardmanPaul FfieldM. A, GrossmanC. H. Herty, Jr.C. H. JenningsJ. M. LessensG. S. MikhalapovJ, OrmondroydH. r. PierceE. C. SmithT. T. WatsonA. G. Bissell, Bureau of Ships - LiaisonMathew Letich, American Bureau of Shipping - LiaisonE. M. MacCutcheon, Jr., U, b. Coast Guard - LiaisonCapt. F. W, Myers, Jr,j Transportation Corps Board - LiaisonE. Rassman, Bureau of Ships - LiaisonComdr. R. D. Schmidtman, U,,S. Coast Guard - Liaison

—_-

copy No. 26-Copy No. 2’7-cOpy No. 28-copy Iio, 29 -

Copy No. 30 -Copy No. 31 -COpy No, 32 -Copy No, 4 -Copy No. 33 -copy No, 3 -

Copy No, 34 -Copy No. 35 -Copy No. 36 -COpY No. 25 -Copy No. 37 -COPYNO, 23 -copy No. 38 -Copy No. 21 -Copy No, 22 -Copy No. 39 -Copy No, 27-Copy No. 40 -Copy No. 28 -COpY NO. 29 -COpy NO. 16 -Copy No, 41 -COpY No. 42 -Copy No. 43 -Copy No. & -

Copy No. 45 -Copy No. 20 -copyNo, 46 -Copy No. 47 -Copy No. 48 -Copy No. 49 -Copy No. 50 -Copy No. 51 -COPY No. 52 -Copy No. 53 -

T. L, SOO-HOO, Bureau of Ships - LiaisonJohn Vasta, U. S. Marittie commission - LiaisonR. E. Miley, Bureau of Ships- LiaisonJ. L. “WilsonjAmerican Bureau of Shipping - Liaison

Ship Structure Committee

Rear Admiral dllis Reed-Hill, USCG, ChairmanRear Admiral Charles D. Mheelock, USN, Bureau of ShipsBrigadier General Paul F. Yount, ‘.~arDepartmentJ. L. Batesj U. S, Maritime CommissionD. P. Brown, American Bureau of ~hippingV, H. Schnee, Committee on Ship ~onstruction - Liaison

Ship Structure Sub-Committee

Captain L. V. Honsinger, IJSN,Bureau of Ships, ChairmanCaptain R. A, Hinners, USN, David Taylor Wdel BasinComdr, R. H, Lambert, USN, Bureau of ShipsComdr. R. D. Schmidtman, USCG, U. S. Coast Guard HeadquartersCol. Werner t?.Moore, USA, Office Chief of Transportation,War Dept.Capt. F. W. Iviirs,Jr,,” Transportation Corps BoardHubert Kempel, Office Chief of Transportation,War DepartmentldathewLetich, American Bureau of ShippingE, M. MacCutcheon, Jr., U, S. Coast Guard HeadquartersR, M. Robertson, Office of Naval Research, U$NJohn Vasta, U, S. Maritime CommissionI, J, Wanless, U. S. Maritime CommissionR. E. ~iiley,Bureau of Ships, USNJ. L. Tlilson,American Bureau of ShippingFinn Jonassen,-Uaison Representative, NRCd, H, Davidson, Liaison Representative, AISIPaul Gerhartj Liaison Representative, AISIWm. Spraragen; Liaison Representative,.WRCy ..

Q~n ? T“~Y *&& )4,? ‘w, LZJe&i

Navy Department

Comdr. 8. S. Mandelkorn,USN, Naval Administrative UnitA. G. Bissell, Bureau of ShipsL i. J-0&+&s, l?u~ “ )&klWxkt.4&Noah Xahn, New York Naval ShipyardI. R. Kramer, Office of Naval Researchi;~,R! Osgood, David Taylor Model Basin

N, E. Promisel, Bureau of AeronauticsK. D. llilliams,Bureau of ShipsNaval Academy, Post Graduate SchoolNaval Research Laboratory

Copy Nos. 54&55 - U. S. Naval Engine&rin gExperimentStationCopy No, 56 - New York Naval Shipyardj Material’LaboratoryCopy No. 57 - Philadelphia Naval ShipyardCOpy NOS. 58& 59 - Publications Board, Navy Dept. via Bureau ~f Ships, C~de 330cCo~y Nos. 60 and 61 - Technical Library, Bureau of Ships, Code 337-L

U. S. Coast Guard

COPY No. 62 - Captain R. B. ~nk, Jr*~ USCGCopy No. 63 - Captain G. A. Tyler, USCGCopy NO. 64 - Testing and Development Division

U. S. Maritime Commission

copy No. 65 - d. E. Martinsky

Representatives of American Iron and Steel InstituteCommittee on Manufacturing Problems

Copy No. 66 - C. M. Parker, Secretary, General Technical Committee, AmericanIron and Steel Institute

Copy No. 18 - L. C, Bibber, Carnegie-Illinois Steel CorporationCot)yNo. 8 - C. H. Herty, Jr., Bethlehem Steel CompanyCo;y No: 14 - E. C. Smith; Republic %eel Company

l’ield~gResearch Council

Copy No, 67 - C. A. Adams COpy NO. 69 -Copy No. 68 - Everett Chapman Copy No. 43 -

Copy No. 70 - F, M. Feiker, Chairman, Div. Engineering

LaMotte GroverWm. ~praragen

& Industrial Research, NRCCopj No. 71 - Clyde Wi.llitis,Chairmanj Committee on Engineering MaterialsCopy No. 3 - V, H. Schnee, Chairman, Committee on Ship ConstructionCopy No. 16 - Finn Jonassen, Research Coordinator, Committee on Ship constructioncopy No. (!+4- ‘h. G, “dilson,Technical Aide~ Cohttee on ShiP constructioncopY No, 72 - Harmer-E. Davis, Investigator, Research project ~R-92CQpy No. 73 - G. E, Troxell, Investigator, Research Praject SIG-92Copy No. 74 - Earl d. Parker, Investigator, Research project SR-92Copy No, 75 - A. Boodbergj Investigator, Research Project W-92Copy No. 10 - J. M. Lessens, Investigator, Research project ~-lolcopy No, 76 - Scott B. Lilly, Investigator, Research project =-98Copy No. 77 - C. H. Lorig, Investigator, Research Projeot W-97COpy No. 78 - J. R. Low, Jr., Investigator, Research Project SR-96Copy No. 79 - Albert Muller, Investigator, Research Project SR-25Copy No. 80 - George Sachs, Investigator, Research Project SR-99Copy No. 81 - C. d. sims, Investigator, Research F?rojectSR-87Copy NO. 82 - C. B. Voldrich, Investigator, Research Project SR-1OOCopy No. 83 - Clarence Altenburger, Great Lakes Steel CompanyCopy No. 84 - A. B, Bagsar, Sun Oil CompanyCOpy NO. 85 - E, L. Cochrane, Massachusetts Institute of TechnologyCopy No. 86 - George Ellinger, National Bureau of Standardscopy No, 87 - M, Gensamer, Carnegie-IllinoisSteel CorporationCopy No. EU3- S. L. Hoyt, Battelle Memorial InstituteCopy No. 89 -Bruce Johnston, Fritz Laboratory, Lehigh UniversityCopy No. 90 - N. M. Iiewmark,University of IllinoisCopy No, 91 - L. J. Rohl, Carnegie-Illinois Steel Corporation

.———

copy No. 92 - “i. P. Roop, Swarthmore CollegeCopy No. 93 - Saylor Snyder, Carnegie-Illinois Steel CorporationCopy Nos. 94 thru 118 - 1. G. Slater, British Joint Services kissionCopy ifo.119 - Carl A. Zapffe, Carl A, Lapffe laboratoriesCopy No. 120- Transportation Corps Board. Ehmoklyn, New YorkCopyNos. 121 thru 125 - Hbrary ~f Congress via ~~eau of Ships, Code 330cCopy No. 126 - File Copy, Committee on bhip constructionCopy No. 127

OOpy No. 128COPY No. 129copy No, 130copy No. 131COpy No. 132Copy No. 133Copy l?o.134Copy No. 135COpy NO. 136Copy No. 137Copy No. 138Copy No, 139

.. Copy No. 140copy No. l&lCopy No. 142Copy No. 143Copy No. 144Copy No. 145Copy No. 146Copy No. L47Copy No. 148copyNoe 149Copy No. 150

- *“ ~~”ll;qv,o,d= * , ~44X. +4($-

(Copies 128 thru 132- Bureau of ShipsCopies 133 thru 150- File

Total No, Copies - 150)

:A

TABLE OF CON’MNTS

P*

Conte@s, ...$.,....e......... . . . ..eS. @.. +..oe00~4*m9aQaQ •QS~O i.

List of Tables,................................*..d*o.****aO*ii

List of Figures...............~..............................iii

Abstract..................................................,.. iv

Introduction,,..,.....~..............................c....... 1

Summary of Results of Tube Tests..,,..............~....? 3

Considerations in Choice of Test........................ 5.,

Testing Procedure.,,,..,,~...,........~......?...a.?............ 7

fipertiental Res~ts .........,,........................~.=. ,.. 10

1.

2,

30

k.

5“

6.

7*

8.

9*

10,

Bend Tests on Samples Cut from Tubes...............* 10

~periments

Experiments

Experiments

Experiments

with

with

with

with

E6010 Electrodes................... ~

Unionmelt and Steel-A,...,......... 13

Low-Alloy Electrodes & Steel-A.,.., 15

Stainless-Steel Electrodes & Steel-A15

Effect of Composition and Microstructure of BasePlate......... .

Effect of Preheating and Postheating...,t..........t

Evaluation of the Effect of Residual Stress...,,.,...

Effect of Arc Voltage,..aa..........p..=.=...9.....

Evaluation of Metallurgical Factors.................

Conclusions........=......... [email protected].@.........

16

1$

20

21

22

24

26

— —.

ii

LIST OF TABLES

Table 1, --

Table 2. --

Table 3. --

Table 4. --

Table 5. --

Table 6. --

Table 7. --

Table 8. --

Table 9. --

Composition, Treatment,killed Steel Plates

Summary of Test Results

and Physical Properties of Semi-

of Test on 20-in. Diameter Tubes

Results of Bend Tests on 10-in. Square by 3/4--in. ThickPlates Welded with Fleetweld 5 (E601O) Electrodes

Summary of Bend Test Results on 10-in. Square by 3/4-in.Thick Plates Melded by Unionmelt Process

Chemical Analyses on Typical Weld-Bead Deposits

Summary of Test Results on 10-’in.Square by 3/4-hThick Plates with Beads of Shieldarc 85, Murex 80,Murex AWL and 25-120Stainless 8teel Electrodes

Summary of Test Results on lo-in. Square by 3/4-in*Thiak Plates with Unionmelt Beads

Results of BendThick SpecimensPostheating

Results of BendThick Specimens

Tests on 10-in. Square by 3/4-in.Showing the Effect of Preheating and

Tests on 10-in. Square by 3/4-in.Showing the Effect of Welding Voltage

Page

27

28

29

30

31

32

33

34

35

— —.

iii

LIST OF FIGURES

Fig. 1. --

Fig. 2. --

Fig. 3. --

Fig. 4. --

Fig. 5. --

Fig. 6. =---

Fig. 7. --

Fig. 8, --

Fig, 9, --

Fig. 10. --

Fig. 11. --

Fig. 12. --

Relationship between True Stress and Natural Str~in forCou ons Cut from Plate and Weld of Tube J, at 70 F and

4-40 *

Bend Test Specimens Cut from Tubes I and J.

Variation in Bend Angle at Maximum Load with Temperatureof Test, E601O Electrode (Fleetweld 5).

Variation in Bend Angle at Maximumof Testj Unionmelt Beads and Welds

Variation in Bend Angle at Maximumot Test: Alloy Electrodes.

Variation i:lEs:I(2Angle at MaximiJmOf TesT,for TIlreel>~ateMater:eis,,

Photomicrograptm of Steels A, B as

Variation in Bend Angle at Maximumperature (Single Weld Beads).

Variation in Bend Angle at Maximum

Load with Temperature



~%ge

36

37

3$

3$

39

39

ro.lledjantiB normalized. 40

Load with Preheat ‘lem-41

Load with Preheat Tem-perature (Three Types of Electrodes)., 41

Effect of Heat Treatment after Welding on Ductility in theBend Test. 4’2

Variation in Bend Angle at Maximum Load with Preheat Tem-perature for Different Arc Voltages and Two ‘Typesof Weld. 42

Typical Cracks Formed in

Odd numbered figures from Figs, 13 -

Even numbered figures from Figs,,14-

Weld Metal During Bend Test. 43

79. -- Cross-sections of Weld Beads &-120

80, -- Results of Hardness Surveys 4.5-21Z

.

ABSTRACT

conducted

deposited

Reported herein are the results of several series of bend tests

on lo-in. square by 3/4-in, thick plates upon which had been

weld metal. The experimentswere carried out in conjunction with

tests on large welded tubes of ship plate. The work was begun as part of

NDRC Project NRC-75 and completed under U. S. Navy Contract NObS-31222,

This report presente a description of the test specimens and the

testing technique, Both double-V butt-welded and single-bead specimens were

tested as simple beams supported on a 6-in, span, centrally loaded. The weld

or bead was so placed that it was stressed in tension by the applied load.

Tests were made at various temperatures ranging from room temperature to

-40*F on specimens welded with E601O electrodes, low-alloy steel electrodes

25-20 stainless

tility at -40°F

steel ele~trod~sj and with Uniolnmelt,

of preheating to various temperatures

The effects on duc-

ranging from @°F t~

500°F, of postheating to llOO°F, and of

effect of arc voltage and the effect of

at various temperatures were determined

normalizing were determined, The

the weld contour on the ductility

for E601o and Unionmelt deposits.

Chemical analyses of typical weld deposits were made and the microstructure

of mdst of the specimens was st~died, and microhardness surveys were conducted

on about half of the specimens tested+ Included in the report are the follow-

ing data: welding conditions, testing temperature, bend an~,e, maxinum

elongation~ maxtium load and type of failure, microstructure, and hardness

values.

The cracks were

the heabaffw+,ed zone of

found to originate in the weld metal and not in

the base plate even though the heat-affected zone

v’

was in many cases much harder than the weld metal. The ductility of the

bend test specimens was found to be improved by preheating, postheating,

low arc-voltage and by the use of stainless steel electrodes. The ductility

was decreased by low preheat temperature, low testing temperature, fast rate

of electrode travel, high arc voltage, and by the use of electrodes having

high hardenability,

.

The

andU.S. Navy

The

Technical Report

U,S, Navy Research Project(Causes of Cleavage Fracture

NObs-31222in Ship Plate)

INVESTIGATION OF BRITTLd CLEAVAGd FRACTUREOF l.!ELDiDFLAT PLATd BY MEANS OF A BNTD T&ST

lday 1946

From: University of California, Berkeley, CaliforniaM, P. O~Brienj Technical Representative

Report prepared by:Harmer d. DavisG. E. Troxell Engineering Materials Laboratory

‘Earl R, Parker University of CaliforniaA, Boodberg

Introduction

work reported herein was

Contract No. ijObs-31222,

formally stated purposes

conducted under OSRD Contract OEMsr #1418

vihichbecame effective August 31, 1945.

of the general investigation are:

(1) determinations of the behavior of steel under conditions of multiaxi@_ stress,

(2) determination of factors responsible for the brittle cleavage fracture of

ship plates without manifestation of appreciable ductility or plastic flow. The

experimental work

effects of stress

fail in a brittle

was divided into two main sections: (a) investigations of th~

and temperature conditions on the tendency of various steels t.a

manner, in which the principal tests are made on test specimens

in the form of tubes and flat plate, and (b) investigations of the behavior of

built-up weldments, principally in the form of sections made to simulate welded

——

-2-

hatch corners~ tostudy the influence of design as well as of materials on behavier

under load. The subject of this report is concerned with some special studies made

in conjunction with section A of the general investigation

In

subjected to

raised as to

the course of the investigation on large welded tubular specimens

various conditions of hi-axial stress, a number of questions were

the effect of metallurgical changes

the mode of failure. In an attempt to interpret

on the large tubes, some bend tests were made on

caused by the welding process on

some of the results of the tests

samples taken from the welded

portions of certain of these tubes. The apparent successful.correlation of the

results of these simple bend tests, as regards certain conditions of failure, led

to an amplification of

the mode of failure.

Owing to the

the study to cover

lack of success in

failure of weldments by means of standard

the effect of various weld conditions on

previous attempts to explain causes of

tensile and impact tests on small

specimens taken from a weldment, it was considered desirable to utilize a test

specimen and a procedure where the interaction of the weld metal and parent metal.

would be closely similar to that existing in an actual structure or part. Further-

more, since metallurgical examinations have shown that a gradation in metallurgical

structure may take place within very short distances in certain zones, a test which

would include the effect of the anisotropy seemed preferable.,

Because questions raised by the experiments on the large welded tubes

are the reason for undertaking the bend tests, and because the tube experiments

form a general background for interpreting the bend test studies, the principal

. ———

-3-

and pertinent results of the tests on the tubular specimens are summarized in

the following paragraphs.*

Summary of Results of i’?elded-l’ubeTests. -- The 20-inch diameter

tubular specimens were made by forming two pieces of 3/4-inch thick ship plate

into half-cylinders each 10 feet long. These pieces were then welded together

by two longitudinal seams 180° apart to form cylinders. The chemical composition

of the steel used for the cylinders~ designated as Steel A, is given in Table 1,

E6020 electrodes were used for the welding, Ten tubes were tested, two of which

were heated 6 to $ hours at ll(X)°F.after welding (l?post-heated~t).The remaining

8 tubes were given no heat treatment after welding. Tests were conducted at 70°F

a~d -40°F~ using various ratios of longitudinal to transverse stress,

.The results of the tests are summarized in Table 2. All the tubes

tested at 70°F exhibited considerable ductility prior to fracture. All the tubes

tested at -40°F except those post-heated to llOO°F, were relatively brittle. The

reduction in wall thickness of the tubes tested at -40°F ranged from 2 percent

to 6.7 percent for the tubes tested as welded, while the post-heated tube had a

wall-thickness reduction of 31 percent before failure. The tubes tested at 70°F

were more ductile, Reductions in thickness ranged from 7,5 percent

A study of the fractures revealed that the origins of the

to 32 percent,

fractures were

almost invariably in the welds even though the welds were sound. The welds

appeared to be acting as crack starters thus causing premature failures of the

tubes. In view of the nature of the failures obtained with the tubes, it was con-

sidered advisable to conduct some supplementary tests on the weld deposit of the

———.- —-%. E, Davis, G. E. ‘B:uxell?E, R. Parker and M. P. O~Brien, I!Behaviorof SteelUnder liulti-AxialStresses and Effect of Welding and Temperature on This Behavior’l.~hl~ FlaLs %-” !.{1vs306),OSRD Report 6365, Serial Ni-5f@, Dec. 6, 1945..

—

tubes. This was

of each cylinder

-1+-

possible because two $ in. lengths had been cut from the ends

in preparation for testing. From these rings were machined

standard 0.505-inch tensile test bars.

Tensile tests were made at 70 and -40°F on specimens cut from the tube

walls and from the welds. Typical true-stress natural-strain curves of these

materials are given in Figure 1. Both the weld material and the plate material

were very ductile. Contrary to the behavior of the tubes, the tensile specimens

showed greater ductility at -40°F than at 70°F. This apparently inconsistent

result might be due to one of a number of causes, The 0.505 inch diameter tensile

bars were cut from the center portion of the weld; thus the last passes deposited

at the surfaces of the welds were not tested. It is the surface layers in the

multipass weld which are likely to be the least ductile. AI-1underlying passes

are reheated by subsequent passes with a consequent improvement in ductility,,

The surface layers in the weld deposit might be the source of the cracks in the

tubular specimens. Another factor which must be considered in relationto the

discrepancy between the results of the tensile bars and those of the tubes is

the difference in the stress conditions. The tubular specimens were subjected

to biaxial loading while the tensile bars were loaded in simple tension. Further-

more, in the tubes the stress condition was complicated by the rough surface of

the weld so that triaxial tensile stresses may have existed near local irregular-

ities. This difference in thelstate of stress presumabl~-might account for the

difference in behavior of the material in the two tests. The picture was further

complicatedby the fact that the weld zone in the tubular specimens had a residual

tensile stress estimated to be about 45,000 psi. No residual stress was present

in the 0.505-inch diameter tensile bar.

—

-5-

It became apparent that there were many variables which should be

evaluated before a critical analysis of the tube test results could be r.lade.

The evaluation of the separate factors bymenas of the tube tests was considered

impractical because of the cost and because of the time necessary for such a

study. To facilitate the investigation, a bend test was chosen and used for

the studies.

Considerationsin Choice of Test. -- Considerable thought was given

to the problem of choice of an appropriate test, and a number of preliminary

experiments (including some on other types of tests) were conducted to study

the merits of various conditions of test. The criteria by w}liehthe various

possible test procedures were judged were as follows:

la The outer layer of the weld deposit should be included in the strained material.

2, The specimen must be of such a nature that biaw”.alstresses can be applied.

3. The specimen must be capable of retaining residual stresses of a magnitudecomparable with those in the tubular specimens (18 inch long specimens wereused for determining effect of stress).

l+. The specimen must be relatively cheap, simple to make and easy to test.

The bend test seemed to satisfy these conditions. The outer layer of

the weld undergoes the maximum strain; when wide specimens are bent as simple

beams there is developed a transverse stress equal to one-half of the longitudinal

stress (a stress ratio of’1,:1is produced in a cupping test); bend test specimens

can be made large enough to retain large residual stresses (18!tlong or more);

the cost of making and testing such specimens is relatively small.

Although the bend test seemed ideally suited for the proposed expcri..

ments, its suitability n:aded to be evaluated experimentally, This was done by

—

-6-

making bend tests on specimens cut from the tubes. The specimens contained the

weld zone and about three inches of the adjacent plate material, The results

obtained with these bend test specimens correlated very well with the results

of the tube tests. (Details are given later in report under Expertiental Results).

Consequently, the bend test was adopted for evaluating the influence of the various

factors.

—.

~7-

Testing Procedure

u bend test specimens used in the

cut from Tube I and J) wsre 10 in, square by

liostof the specimenswere made by

experimentalwork (except those

3/~ inch thick.

depositing beads along the center

line of one face and parallel to the direction of rolling of the plate. Some

of these specimens were subsequently tested in the as-welded condition, i.e.

with the beads not ground off, while others were tested after the beads had

been ground flush

made on specimens

of ‘si~glebeads.

with tihesurface of the plate. Additional bend tests were

of the same size but containing double-V butt welds instead

The butt-welded specimens were made by welding together two

5-in. by

then cut

50-in. pieces of 3/4-in. thick plates. The long welded pieces were

into 10-in. lengths to form the square bend-test specimens.

In some series of experiments, the plates wore heated ~r cooled to

various temperatures before the beads were deposited. Tho heating was done with

an acetylene torch and the temperature of the plate was measured by a contact

pyrometer. The temperature of the plate just ahead of the arc was maintained

within ten degrees of the desired temperature during the welding. Som2 of these

specimens were subsequently post-heated in controlled temperaturemuffle furnaces

to determine the effect of post-heating upon the ductility,

All specimens were loaded as simple beams on a 6-in. span so that the

direction of the maximum tensi.1.estress was parallel to the bead or to the butt

weld. The supports extended across the entire width of the specimen,, The load

was applied at the midspan through a loading member which also extended across

the entire width of the specimen, The loading member was rounded,to a radius

-8-

of 1/4 inch on the edge where it contacted the test specimen. The fixtures

which supported the ends cf the beams were also rounded to a 1/4 inch radius

and were recessed toaccomodate the weld bead so that the plate rested on the

supports across the entire width.

Tests were conducted at various temperatures ranging from room tempera-

ture (about 70°F) down to -/+OOF. The specimens were cooled with dry ice to a

temperature slightly below the desired testing temperature, A

in the testing jig and the test was begun when the temperature

thermocouple attached to the surface of the plate) reached the

specimen was placed

(measured with a

value recorded in

the tables of test results. The specimens were loaded as rapidly as poss~ble

to mum load. The loading operation required from15 to 30 seconds depending

upon the ductility of the specimen. Because of the rapid loading, the tests

were adiabatic rather than isothermal. The temperature rise which occurred

during the test varied from a few degrees for the brittle specimens to as much

as 25°F. for the most ductile specimens Since the results of the tests are

comparative, only the temperatures of the plates at the start of the tests are

reported,

During the course of preparation of the specimens, measurements of

the plate temperature were made prior toweldi.ng, and the welding currents,

voltages and rate of electrode travel were recorded. During each test, the

temperature,maximum load, bend angle and surface elongation were measured. The

elongation was measured parallel to the weld over a one-inch gage length and in

the region of

the action of

is available.

maximum elongation. Unfortunately, it was impossible to observe

the weld during the test so no record of when the first crack formed

Most of the specimens tested at low temperatures fractured completely

across the plate when

the weld zone usually

-9-

the first crack formed. At room temperature, however,

cracked in several places and often the plate did not

fracture. In such cases, the cracks formed at bend angles which were sometimes

much smaller than those recorded in the tables. The bend angles at the high

temperatures thus measured the ductility of the plate material rather than the

ductility of the weld deposit, The low temperature tests, however, provided a

better measure of the weld zone ductility because the first crack which formed

caused the plate to fracture without additional deformation.

-l f3-

Experimental Results

The studies conducted during the course of the investigation may

conveniently be divided into ten groups, as listed below.

1. Bend tests on samples cut from tubes.

2. Experiments with

3. experiments with

4. Experiments with

5. Experiments with

E601O electrodes,

Unionmelto

low-alloy electrodes,

stainless steel electrodes.

6. Effect of composition and microstructure of the base plate,

7. Lffect of pre-heating and post-heating.

$. Evaluation of the effect of residual stress (with 181’long specimens).

9* dffect of arc voltage.

10. Evaluation of the metallurgical factors,

The results will be discussed in the order listed.

1. Bend tests on samples out from tubes. The first step in conducting

the experimentswas to determine the suitabilityof the bend test as an indicator

of the performance of welded structures such as the pressure vessels studied on

NDRC Research Project NRC 75. Whether or not the bend

and ductile fractures to correspond with the fractures

an importent factor to be de~ermined,

test would yield brittle

obtained with tubes was

As described in the introduction, the tubular specimens were tested

under various ratios of longitudinal to transverse stress; hence the biuial

stress ratio of 2:1 developed on the tension side of the bend test specimens

was the same as that used for some of the tubes. The temperature of the bend

test was also adjusted to correspond to that used in the tube tests. The bend

—.. -- .-

-.u -

test thus satisfied the conditions of stress and temperature.

During the preparation of the tubes, sections about 8 inches long

had been cut from each tube. From some of these sections, pieces about 8 inches

square were cut which contained the welds. ‘1’hescrather crude spectinenswere

then bent in such a way that the longitudinal weld was stressed in tension, sir&-

lar to the manner in which their counterparts in the tubes were stressed. The

specimens wel-etested as centrally-loadedbeam on a 6 inch span. The curwature

of these specimens,

results”complicated

mens was so obvious

though slight, made a quantitative interpretation of the

but the difference between the ductile and the brittle speci-

that rdinernents were not necessary. Furthcrmorej the

behavior of th~ specimuns in the bend test corresponded with the behavior of the

tubes from which the bend test specimens were cut. Figure 2 5h0wS the results

obtained when two of the specimens were ‘oente The ductile specimen was taken

from Tube I which had been post-heated to llOO°F after welding; the brittle

specinen was taken from Tube J which was left in the as-welded condition.

It appeared from

afford a quick, economical

many complex factors which

these preliminary results that the bend test might

and satisfactorymeans for evaluating some of the

seemed to influence the behavior of loaded welded

jo$nts. .diththis belief in mind, the following ~xperiments were performed.

,.,2, :tiperimentswith E601o Electrodes One of the electrodes conunonly. —— ...—.—— 0

used for the welding of mild steel is the d6t)10type of electrode. This eiectrode

is made of low carbon steel and has a cellulose coating. A Fleetweid 5 rod was,,“’chosenfor these tests. All of the bend test samples in this series of tests

were 10 inch square and 3/4 inch thick. Most of the specimens were made of single

10 “&.chsquare plates on which weld beads were deposited, Some of the specfi~c’s,

—

. .

-12-,.,$,,:: ..,.

however; were made of two 5 inch wide, 50 inch long plates joined bya double-V,. ,,.:.

butt weld “andsubse@entAy cut to 10 inch’’squares. Only steel A was used for,.

.

these experiments-,~~ththe E601O electrodes; ‘:”:

The plates were tested as centrally->?aded:,sfiply-supportedbeams with.!..

a six inch span. The weld beads were plac,e,dint$e:lo~gitudinil direction so:.,that the maxhnurntensne stress was parallel with the direction.of the weld,

The resulting fractures thus traversed the bead. Specimms welded at room

temperature were bent at temperatures ranging from70° to -40°F. The re~ults,,

of these tests are recorded in .Table3 and Figure 3. Figure 3 show’stl%’kffdct,,.

of testing temperature on the bend angle obtained in these tests; These resuits

clearly showed the influence of the testing temperature upon the ductility.

,.!: ,,A comparison between the double-V butt-welds shows a striking contrast

.4.’.... ~,.in behavior. The weld made by a series of longitudinal passes (single pass

stringer beads) was extremely brittle while the weld made by.oscilla~ing the,, .,,

electrode from side to side so that the entire groove was almost.filled in a,:

single pass (wash weld) was very ductile. The difference ir welding technique

had a profound effect on.the ductility of the weld, The specimens were made by

welding together two 5 inch wide pieces of plate each 50-in. long, The stringer

beads were started at one end of theplate’and deposited progressively towards

the otherend~ By the time one pass was completed, the plate at the place where

the weld was begunhad cooled to room temperature. T“huseach pass was deposited

,.on a cold”plate and’consequently”cooled at a rapid rate. The wash weld progressed

,., .,.s0 slowly,“however,that the plate ah~ad of the weld was actually preheated by heat

,,. ..!which diffused)fo~vard f~stcr than the weld progressed. The primary difference

in behavior appeared to be due to the difference in the rate at which the weld zone

- L-j -

cooled. This observation led to another series

7 of this report.

The results

between those for the

bead deposits was, of

from the specimens having

two types of butt welds.

of experiments recorded as part

a single bead were intermediate

The cooling rate of the single

course, much faster than that of the wash weld. Conse-

quently, it is to be expected that the single bead deposits wouldbe less ductile

than the double-V butt welds made by the wash weld technique. The cooling rate

of the weld zone in the doubl’eV butt welded specimens made by the stringer bead

technique must have been of the samtiorder of magnitude as that of the single bead

specimens, yet the single bead specimens were more ductile than those that were

butt-welded. The difference in ductility is apparently due to the difftirencein

the contour of the surface. The single beads were ground flush with the plate

surface while the butt-welded specjmens were not ground,. This conclusion was

confirmed by later tests on single bead specimens tested in two conditions--as

welded and with the beads ground flush (see Fig,,8).

3.

employed for

desirable to

Experiments with Unionmelt and Steel A, Unionmelt welding is commonly—- —— .

the welding of flat mild steel plate. Hence it was considered

study the performance of Unionrneltwelds, In this series of exper+,-.

ments, single beads were deposited at high rates of speed to obtain deposits as

nearly comparable with those deposited by hand with the E601o electrodes. A

current of 320 amperes was used with a l/4-in. rod, The rate of travel of the

electrode was 15-in. per minute. Thus a small bead was obtained. The beod.swere

deposited at room temperature ,Ondthe specimens weru tested at various temper:ttures

ranging from 70° to -40°F,

is plotted as a function of

The results

temperature

are recorded in

in Fig. 4. ‘The

.

. .

.

- lf+-

greater ductility than did the specimens having E601O beads. This can be at

least partially accounted for by the fact that, in spite of all that could be

done to deposit metal at the same rate in the two cases, ‘thebead deposited by

the Unionmelt process was:larger and consequently cooled at a slower rate than

did the E601O bead. There are, however, other conditions which were different

in the two cases and might presumably influence the rusults. One of the most

important dtiferenc-was in the arc atmosphere. The atmosphere surrounding,,

the arc made by the E601O electrode was ’mdoubtedly rich in hydrogen gas,,.-,.,;.,.,,-

originating from the cellulose coating on the electrode. The Unionmelt flux~.,,

on the other hand, is a mineral type coating contaifing:~silicatesof calcium,....... .

aluminum, etc. and consequently generatedlittl& h~&ogen”when it was fused,.

It is well known that excess hydrogen lowerethe ductility “ofalloy steel welds,,.

and it may be that

Double-V

hydrogen has a similar effect on low carbon steel.

butt-welded speciqens @n% alsomade;by the”U&onmelt process,.

using standard welding procedure. The lo-in. square specimens were cut from a..

long piece made by welding,together two 5-in..

wide pieces of plate each 5&in..,

long. The weld was started.at one end and progressed continuously towards the

other end. One pass,...

welded was placed in

ing the test. Tests..,’ !

results are reported..’

was deposited on each side of the plate. The last side

the bending.jig so that it would be stressed in tension dur-

were made on these specimens at various temperatures. The

in Table 4 and the”bend angle is plotted as a function of

the temperature in Fig. 4.

The si~’le-bead specimens’weremore ductile than the butt-welded

specincms even though the single-bead deposits cooled more rapidly than the

butt welds. The difference in ductil.itywas undoubtedly due to the surface condi-

- 15 -

tion. Th~ single bead specimens were ground flush with the surface of the plate

w~ile the butt welds were not.

&. Experiments with Low-alloy Electrodes and Steel A, Another’series

of experiments was conducted with several types of low-alloy steel electrodes--

Shield-arc 85, Murex &pe 80 and Murex AWL. It must be recognized at the outset

that bead tests with alloy steel electrodes are likely to be misleading. The

alloy st.el electrode is diluttidexcessively by the mild stctilin a single bead

deposit. Consequently, it is necessary to know the actual composition of the

weld deposit before the results of the bend test can be intelligently appraised.

To obtain this itiormation, met~l was machined fr.m the central portion of the

beads of some of the specimens after they were tested, These fi.hipswere then

analysed along with similar samples obtained from other specimens. The chemical

compositions are reported in Table 5. The results of the bend tests are given

in Table 6 and the bend angle is plotted as a function of testing temperature

in Fig. 5. The bend angles of the specimens made with the kiurexType 80 elec-

trode were essentially independent of the testing temperature while the bend

angles of the specimens made with the AWL electrodes were roughly the same as

those obtained with E6010 specimens, The specimens welded with Shieldarc 85

were the most brittle specimens tested. The harnesses of the beads and heat-

affected zones in these specfiens were later measured with a microhardness

tester, The results obtained are discussed in detail in section 10 of this ‘

report.

5. Mperiments with Stainless-SteelElectrodes and Steel A. Experiments—,. ———.———. —

were made with columbium-stabilized,19 percent chromium; 9 percent nickel el~c-

trodes, and with 25 percent chromium, 20 percent nickel electrodes. The beads

— —.

-16-

made with the 19-9 electrodes were so brittle that

many cracks developed in the weld deposit when the

even at room temperature.

specimenswere bent through

very small angles. The 19-9 electrode material was so diluted during welding

that the single bead deposited on the plate had a very poor metallurgical struc-.,

Lure. To avoid this condition, 25-20 electrodes were used thereafter. Even with

this electrode the weld deposit was diluted by the fused base material but not

enough to cause the formation of the brittle metallurgical structure associated

with the 19-9 deposit. The composition of the 25-20 beads is given in Table 5.,.

The results of the bend tests obtained.with 25-20 beads are given in

Table 6 and the bend angles at various temperatures are,plotbed in Fig. 5,,., ,’

to’getherwith the results from tests on specim~ns having low-alloy steel beads,

It is”i~lportantto note that no cracks formed either in.the.weld zone or the ‘

plate at any testing temperature. ,!’,’

r,.:,

A microscopic examinationwak made of the weld zone to determine

,..,whether or not the heat-affected zone ‘of<the“platehad cracked, Microcracks could

,.conceivably form in this zone but might noi grow.

,..~crohardness surveys later

revealed that this zone had a maximum h’ardnessof about 500 Knoop which is equal ,

to approximately 5C10Brinell. A.,.

failed to show any cracks. This

this material.

The 25-20 type

of mild steel to prevent

careful microscopic”survey of

seemed remarkable considering

,’

. . .

the hardened zone,,

the hardness of,..

of welding electrode seemed to be ideal for the weldihg

low temperature brittleness in the weld zone. The,..,,

reason for the,re~rkable performance.,ofthe 23%20‘s~eclmens’’wasnot”~ediately,,’:’

apparent. Two factors might contribute to the favorable action of this electrode.,.

.,,..,

,,:’.

One,is the good ductility

other is the low hydrogen

.~?-

and insensitivity to notches of

content of the arc atmosphere,

had a lime-type coating which liberated relatively little

obtained on mild steel with this type of electrode are in

the weld bead; the

The 25-20 electrodes

hydrogen. The results

line with those

reportedly obtained

A word of

of 25-20 electrodes

with the same type of electrode on low-alloy steel.

warning is necessary, however, in connection ~~iththe use

for the welding of mild steel, Some later experimentswere

performed which indicate that the results obtained in the original tests must be

reconsidered. In the final experiments several identically prepared bend test

specimens were tested in bending with the bead running transversely rather than

longitudinally. The heat-affected zone ~f the plate was thus subjected to the

maximum tensile stress over the entire length of the bead. The bead was left

on the specimens and consequently acted to concentrate stress at the junct<.on

of the bead and the plate where the heat-affected zone is located. These speci-

mens, when tested at -400F.j fractured through the heat-affected zone after

bending through a relatively small angle. ‘These

25-20 electrodes for welding mild steel will not

results show that the use of

always prevent brittle fractures

from starting in the welclzone, and that the state of stress and the orientation

of the principal stresses is a factor in determining the ductility and failure,

especially when marked variation in hardness exists in the weld zone.

6. Effect of Composition and Microstructure of Base Plats Some——— — ——

limited experiments were performed to explore the effe-,tcf composition and micro-

structure of the plate metal on ductility of weldrnents. Semi-kjlled steels of

two different compositions (steels A and B) were used fc: these bend tests, One

of these steels was tested in two different conditions of heat treatment (E as

. .—

,.

,.!

.,’.

,,

.,’ :

-. 1$ -,,..;,,,,”.,

rolled, and B normalized). The compositions and heat treatments of these steels,.

are given in Table 1* Unionmel~ yeld beads were deposited on specimens of these,.

plates. The results of the bend tests are recorded in Table 7 and ’the’bend,,.-

angles at various temperatures“are“plottedin Fig, .6. As indicatedby the differ-

,.ence between the behavior of steels’A and B, the chemical composition has some

effect, although it was apparently S&U in this.caset Moye.irnportantis the

fact that the B steels, which’were chemically identical$ behaved.differently in

,,. -,,the two different conditions o-fheat treatment. There was a ~all difference

,,

in &icrostructure which can be”seen in the pho’tornicrographsshown in Fig+ 7.

The microstructure of steel A“is’given for comparison. The difference in the

results obtained with the two B series”of specimens makes it obvious th?.imicro-

structure of the base plate must be considered as an important fac+)c.”,,Yi is

.!”,. ,

not sufficient to study the effect of chemical composition d.oi~~c ! TLJ. ‘.;! c)f

.,,.

the effect of chemical composition and microstructure is in itself <.u~.:,[:.:project

so it was not considered adviaable”to extend the study at this tir~e,

,“., .,

Cracks which developed +’irst,inthe weld metal in this group of test.. .

specimens apparently acted as notches so that the behavior of the’pl”ates(after

the welds cracked) was conditioned by the notch sensitivity of the ‘basemebal,,., ,,

. .. .;.’. .. ... .... .... .,,

,. 7. Effect of Pr~heati~g~. A number of experiments were,,

performed to determine the effe~t of preheating and postheating on the behavior.. . . . .

of bend test specimens,. The plates we,re,cooledor heated to various temperatures..,.

;,., . ,...

as indicated in Table 8 before the,w,eldbeads were deposited. The belilangles,,:,,.

obtained for tests at -i+OOFplotted against the preheat temperature foi speci-

mens with E601.Obeads are shown in Fig. 8. Specimen: were’tested both with the

beads left on and with the beads ground flush with the surface of the plate.

-19.

The specimens tested in the as-welded condition (beads not ground) were consider-

ably more brittle at the low temperatures than were the specimens for which the

beads were ground flush. It is important to note, however, that even the as-

welded specimens were extremely ductile if the weld beads were deposited on hot

plates. The bead contour had no effect upon the bend angle when the plate was

heated above 400°F before welding. The remarkable effect of preheating upon the

ductility of the bend test specimens was found to be true for all types of elec-

trodes used and is undoubtedly the result of the slower cooling associated with

the preheating. Slower cooling causes a softer microstructure to result and also

allows more time for dissolved gases, particdarly hydrogen, to escapee Addi-

tional comment is given in section 10 concerning the effect mf preheating cm the

microstructure of the weld zone.

Fig, 9 shows the effect of preheating upon the bend angle at -l@°F for

specimens having various kinds of weld beads.

For some specimens, the plates were at 100°F before welding) and

sequent to welding were heated to 11.000Ffor one hour, As shown in Table

in Fig. 10, this treatment had a profiouncedeffect upon the ductility of the bend

test specimens tested at -40°F, Similar specimens were given a normalizin~ treat-

ment (heated to 1650°F and air cooled) which, in genera~.jseemed to yield results

similar to the treatment ai llOO°F. The results of some of the tests on normalized

specimens are also reported in Table 8 and are plotted in Fig. 10.

The beneI:ici:~effect of postheating is ob’~j.~us,The reason for the

improvement is net clear?;rbrought out by the tests, The llOO°F treatment is

generally called a stress--reliefheat treatm~nt and it is true that it does

relieve the residual weldicg stresses to a large degree, One important qyestion

is whether or not the rel--efof the residual stress is responsible for the i-e-

sub-

$ and

mix. le ~~~rov~’nentin the performance of the bend test specimens (and for the

_po _

improvement in the ductility in the large tubular

role played by residual stress, another series of

are described in detail in the following section.

specimens). To establish the

experiments was conducted which

8, Evaluation of the Effect of Residual Stress, Although weldments

which are given the so-called stress-reliefheat treatment are largely relieved

of residual stress, a

have also occurred in

changes are discussed

microscopic examination reveals ,thatmetallurgical changes

the heated specimens. (The details of the metallurgical

in section 10,,) Consequentlyj it was considered necessary

to,evaluate the effect of stress and the effect”of other changes separately.

Fortunately this could be done in a simple manner but 18!!long specimens had to

be used instead of the 10IIsquare type. It had been previously der,onstrated*

that when all elements parallel to a weld are elongated an amomt ci ,bcui cne

percent or more$ the residual stresses are reduced to a low order c:’.I~L gi:~;k.ude.

Hence by elongating the test specimens about one percent w s.+:em+l;~‘c~t~ng.,

machine prior to bending, it was possible to eliminate prac.ticaii;:?.:?.c< tn,e

residual stress without ~fecting the metallurgical structure. VJ2~*qL:-,i~ ~\’&~,.

donej the specimens were tested and found to behave essentially the same as i;hose

containing the residual stresses. At -40°F there was a small bu’~percepii.hl.y,..

larger bend angle for the pre-stretched specimens but the difference was very.!.

small,. The bend angle iq all.c.,aseswas relatively large so that the residual.,..:..

stress could not be expected to.sllo,~a Large’effect,.,’

In truly brittle failures where the total elongation prior to failuref,.

is only a fraction of a percent~ the strain associated with the residual stress;,: ..,. ,“.,.:..,.

would be an appreciable part of the total strain; and residual stresses would.,,

be expected to contribute greatly to.the.fracture. Brittle fractures of this‘,

——— -—.%,, P, DeGqrmoj J.L, lvier@r,.andFinn Jonassen: ~lResidualSt:-essesin Ship;’Jel.din”q“(NS-3C4)lf,OSM Report No. 4388, Serial No. M-370, Nov. 13> 1944..

,:,,..,.:: .,,“....’. ,..’

., :,’

_p~_

nature were not obtained within the range of

and consequently the residual stresses could

temperature employed in these tests,

not have played an important role

in the bend test results. This conclua,ionwas checked in several ways,

It was also found that residual stresses of high magnitude could be

induced in bend test specimens by local heating with an acetylene torch. The

maximum temperature was never allowed to exceed 500°F so no metallurgical changes

could occur in the mild steel plates. These specimens thus had the residual

stresses but had no discontinuity in the metallurgical structure as did the welded

specimens. In the bend tests, there was no difference in behavior between the

plates with residual stress and those without.

specimens

identical

preheated

Still another check was made on the effect of residual stress. The

preheated to 5000F were found to have a residual stress pattern almost

with the stress pattern in the specimens welded without preheat. The

specimens thus had the residual stress but had very different metal-

lurgical structures. The preheated specimens were very ductile compared with

those specimens not preheated.

The conclusions reached were: (1) residual stresses have little in-

fluence on the ductility of bend test specimens tested, (2) the beneficial effect

of the so-called stress relief heat treatment is primarily the result of changes

in the metallurgical structure (which may include the elimination of hydragen

which apparently causes brittleness when present in large amounts in the weld

zone),

9*

welds made at

Effect of arc voltage. It was observed by H, E. Kenuedy+$that

high voltages tend to be brittle in the bend test. Sme experiments

were conducted to check his observations. The results of the investigation are

summarized in Table 9 and the bend angle vs. temperature curves are ,plottedin

‘>. 3. Ke~nedyo—..—

IIsomeCauses of ‘Brittl.jFailures in ,eldedIild Steel Structures.”;“~eldingJour;ial,Nov. 1945’,,~eldingResearch tiupplumentjp. ,588s.

‘... .,:

,,, .,

Fig, 11, It stiouldbe noted that ,preheat~g relieves the brittleness induced in,.. . ,,

the weld zone by the highwelding voltaq,e. The m~hanism by which high welding

vdltage:influencers’.the.duetilityof th~::yeld.is not clear, It is conceivable.,,. .,, #.. .‘, ...:,..,

that$he silicon content of the Un>-onmelt,weldis increased by the high voltage.,. .::,.:-’,;,,’.,::’

Th~s wotid account for the increa$ed brittleness of the Unionmelt weld but not. ,’...::,. ,..!...:~.1’..,’,

for that of’tliehand weld; The iqcreaqed brittleness of the specimens made with: ;..:: ; “... ......!4.

the”Fleetwe”ld5 rod might be explained by the increase in the nitrogen and hydro-.,.. , ;: ,,i:”:’ ,..

gen contents which presumably res~lt.from the high voltage welding. No sin~le,.,,,,1, ,., ,,, ,,:,;,..

expatiation seems adequate at thepresent time and it is possible that several,.,, ,,,: ,,..-,.,,.,

f“actors”contribute to the britt~enessof the specimens,welckdat high voltage.,,’ .. .., 10, Evaluation.dme~allurgical factors. The most striking results,. .,,,. ,: .“:. ~,

in thebend test experiments were,those,obtained by preheating and postheattig.,,,; ,,,,..,:,:, ;

The ef;fectswere ,o~iously not related to the residual stresses bu+ .,~peamd to.,. ,,

be closely associated with the metallurgical structure of the weld ad tie:,+..,) .,.>.,

affected zone. Microhardness surveysi~eremade of a number of +;hebend +,est,.

specimens to determine the relative harnesses of the weld zones The i-esdts,, .

of these.mrveysare given.in Append= A along with macrographs showing the con-.,.. :.-.

tours:of,the weld beads (which are also important @ determining the performance,., ,. .’..

Of band .test.specimens).,,.’

The results of these hardness surveys may be generalized briefly as

fbllows: the.higher the preheating temperature the softer the weld zone; post-.

heating at llOO°F softened the weld zone; the hardness was a maximum in the heat-,“

affe~ted zone.. ,Theheat-affected zone required special study because of the

highhardnesses found there.

which is rp~hly equal to 450,,. ,!’,. ,,appeared to be Bainitic. The

(~ some cases the hardness exceeded 450 Knoop.

Brinel.1).The:structure was not iiartensitic,but.. .. ::, ,.heat-affected zone is heated atmve tihutransformation

——

-23-

temperature during the welding and subsequently cools to room temperature. The

cooling is unusual in this zone, however, because it is first cooled very rapidly

by the cold plate near the weld but the cooling rate is rapidly slowed down as

the base plate becomes heated until the cooling becomes extremely slow as room

temperature is approached. This type of cooling is conducive to the formation

of Bainite which seemed to be the structure in the heat-affected zone. The high

hardness in this zone indicated that this region was the most likely place for

cracks to originate. This conclusion did not agree, however, with the obser-

vations reported for the bend test specimens having 25 Cr - 20 Ni beads which

were very ductile and did not crack in the heat-affected zone even though the

maximum hardness in this region was about 500 Knoop.

lJanyof the bend test specimens having plain carbon and alloy steel

beads were also examined microscopically to determine the origin of the cracl:s~

In almost every case, the cracks started in the weld metal. A photograph of

cracks in various stages of growth is shown in Fig. 12. A few cracks were found

to start in the heat-affected zone of some of the specimens examined but such

cracks were exceptional. It immediately became obvious that the properties of the

weld metal were the important factors governing the behavior of the bend test

speci-mens, It appears, then, that preheating and postheating improve the ductility

of the bend test specimens because of the effect of these treatments on the proper-

ties of the weld deposit, It should not be assumed, however, that the heat-

affected zone is never a source of trouble, because in several cases cracks were

observed to start in this region. The 25-20 specimens bent with the beads trans-

verse to the beam axis cracked in the heat-affected zone after onl,ymoderate

amounts of strain had occurred.

1;

2.

3.

4*

,,.,,.

‘5.

,,. .-24-

.-~NiLUSIONS

‘Bendtest speciniensmade by depositing weld

mild steel plate became progressively,,.

was lowered.,,,,

Diff&ences in welding technique made

metal of various compositions on

more brittle as the testing temperature

a large difference in the ductility of

mild steel welds. Doublb-V butt welds made with E601O electrodes using the

stringer-bead’technique beriithrough an angle of only 15° when tested at

‘40°F while similar spec&ens made by the wash-weld technique bent through. .. . . ;

70° to @q The difference in the ductility is attributed to the difference

in the cooling rates associated with the different welding practices,.,.

The cooling rate of the weld appears to be one of the most important single

factors in determining the ductility of mild-steel bent-test spticimens,,

The weld zones in specimens preheated to retard the cooling r,at.ewere found,.

to be much softer (on the Knoop microh~rd.nessscale) than sumi.lari:~gjons

in specimens which cooled at more rapid rates.

Reheating to llOO°F after weldfig produced cha~es in the properties and

metallurgical structure of the weld zone which changes were very beneficial.

However, when tested at -40°F,,specimens which had been preheated to 500°F

were found to be as ductile as specimens which had been postheated at llCO°F.

A complete normalizing treatment (air cooled from 1650°F) did not increase

the ductdity appreciably more than did the 1100% treatment.

The beneficial effect of preheating could not be attributed to the effect of,,

residual stress because the ~eheated specimens had essentially the same re-

sidual stress pattern as had the specimens welded without preheat,, The in-

crease in the ductility could be the result of the difference in the proper-

ties as reflected by the microstructure or to a difference in the gas content -

of the weld zone.

—

6.

.“’,-’

7.

8.

....

.,

10.

The so-called stress-relief

duced beneficial changes in

-,25.-..

heat treatment

the prope’iiies

not ,beattributed to the relief of stress.

(llOO°F after welding) pro-

of the ’weldzone which could

This was .demonstratedby

stress relieving specimens by stretching prior to bending (stress re-

lieving by stretching doec not chaqge the metallurgical structure,)and-.

by introducing residual stresses in an unwelded plate by local ~e.sting

to below 500°F prior to bending. The specimens which were stress r,e-

Iieved by,stretching behaved identically with those.containingresidual.-

stresses,, The specimens

no weld, behaved exactly,“.

containing the residual str,~s.eesjbut having

like si+nilarspecimens having.no!mesidualstress.

T% heat-affected zone of the base plate was found to have a maximun

hardness which in ’some”cases exceeded 450 hoop (approxhately 450..

Brinell). This zone appeared to have a Bainitic type of microstructure.

Microscopic observations established that cracks originated in the weld

metal and not in the heat-affected zone even though the heat-affected.

zone was u&ually harder than the adjacent weld metal.

Specimens’made with 25 percent CR, 20 perce~ltNi

be very ductile when tested at low temperatures,

preheating of the base plate.

electrodes were found to

even when welded without

High arc voltages were found to produce less ductile welds than Lower arc

voltages.

.2(j - .

,. .

;., APPENDIX A - RESULTS OF MICROHAJUXWSS SURVEYS ,,

Cross-sections of a numb& of typical specimens on which weld beads

had been deposited were prepared for microscopic examination and hardness.,

testing, Hardness was measured along a line’through the center of’the weld,, ,.. ‘,,

beads, across the thickn~ss of the plate,””starting at t’hetop’of the beads.,,. ,.,

The hardness values were determined as Knoop’microhardness numbers and were,,

measured with a Tukon hardness tester which was operated with’s 500-gram,,.

load in these tests. The specimens were chosen so that comparisons could be..:

‘madebetween the effects of “preheating,postheatingi aria“chemi-~~1composition

of weld. ,.,

Macrostructures, at a,magnification of approximately L+jare shown in,.

the .odd-numberedfigures, Fibs. ,~ to ’79. The results ofmicrohardness sur-~eys

are shown in the even-numbered figures, Figs, 14 to 80; in these latter figures,. ..,..,

are shown photographic reproductions of the region along which the harkess

indentations

(top of weld

and hardness

a comparison

. . .

were made and ‘plotsof hardness VS, distance from top of weld bead,,,,-. .

bead as shown in correspondingmacrograph). The macrostructures

plots for each specimen are given in consecutive f,ig~es so that

can be made between.structure and.hardness~

-.

-2’7-

TABLE 1. -- COMPOSITION, TREATMENT AND PHYSICAL PROF’E.RTI?LS

OF SEMI-KILLED STEEL PLATES

lJOTE: Physical properties based on tests made at room temperature on standardA.S.T.M. 0.505-in. tensile bars,and standard keyhole-notchCharpy specimens, described in A.S.T.II.SpecificationE8-42.

~ominal ~hemi~alcomposition

Yield stress, ksi.

Nominal tensile stress,ksi,

Percent elongation in2 in,

Reduction in area, percent

Ener=g absorbed in Charpyimpact test, ft. - lb,

Hardness, Rockwell “B”

Steel A“—As Rolled

0.25 C!0.47 N?n

35.8

58.3

43.5

61.4

W*6

60-62

Steel BarAS Rolled.————.

0,18 C0.72 Mn

31*9I

56.7

~t~● 7

69.3

48.0

—Steel Bll

Normalized..—,—.—_. ——.-

J.13 c0.72 Mm

35.4

58’0

‘;569

66.7

46.7I

58-;3 , 59-60

..—-— ‘——-...-.—

TAB13 2. -- SUMdARY OF TEST rrssums - 20-IN. DIAMSTER TUBES—— .—. —

-—

Ener.7–-

.--—Reduction. [Potential

——— —.—

+

W:l at Nature of FractureThickness, Fraoture,

_% ft.-lb.h—-

T

_._ ——---

Initis.tedby shear in plate about 4 in.

30.0 133,000from and parallel to weld near mid-section; after shearing for about 5in., crack propagated over consider-able length of tube by cleavage.

32.()

_.1

14 R,500 Practically same as for tube A.

—— --L_. ——— —-—-

—..

T—-

Conventional NominalStress Stress.b

—————.Average -

True Stresscat Fracture.

_—— —-Maximum

PercentLoading

Conditions

—— --

InternalPressure

,,

——

xial load,nt.press.

n

!1

,,

n

,t

FromPlateNo.

.—

13970

1F991

13993

——

13970

13973

1399s

13973

13973

13995

13970

13994

13970

Stressa

?;’

TestTeinP.

OF.

—-

70

70

——

-42

Speoi-men

——

A

~j

Bd

-..——

cd

Dd

E

F

~j

J

——

G

!?Longi-tudinal—.

26,250

24,900

25,660T~mi

Tram- Frop.verse Limit

52,500 30,000

49,800 30,200

51,300 40,000

—— ——

51,300 30,000

53,100 so, 000

51,000 30,000

i.ztrmate—.

59,003

S2,000

54,000

tionElomin !

,ong.

41

<1

——

<0.2

ps..—Lonz.—..--..—

42,$00

45,000

30, Oooe

——

-e

-~e

6yx)r_)

——-

*~

85 500—~—

61,000——-

——.

50 000—L___

—- —..

—— --Trans.———

85,000

9&OQ

——

60,000e

in.

Trans.——-—

15.0

21.0

-—.—-

3.0

4.4

4.0

10.6

2.0

18.7

3.0

—_—-

0.3

2

2

2 3.0 110,000i

Cleavnge fracture entirely around cir-cumferential end weld.

------+--+-;~~;;~~mt”re);, ~fter,e9.0 ilo~,sdo ; pair by welding, shear for 4 in., then

deava.ge.g

Cleavage fracture originating in weld

7.5 102, COO abo!ut48 in. from mid-section and prop-agating spirally arOund tube.

~Cleavage fracture originating in weld20.0 I 143,000 I 2 in. from mid-section and propagating

completely around tube. No shattering.:

_.___. +---

-i-

-- —————

Cleavage fracture originating in weld3.5 1 86,500 , 24 in. from mid-section. Specimen shat-

1 tered into many pieces.

I201,000Cleavage fracture originating in weld

31.0 4 in. from mid-section. Specimen shat-tered into mmy pieces.

I160,500Cleavage fractme originating in weld

6.7 28 in. from mid-section. Specimenshattered into many pieces.

-—-- —. —

20 l::k

—

,Cleamge fracture originating at defectin plate 90° from weld 44 in. from mid-section and propagating around specimen.No shattering.

70

70

70

55,100

56,600

56,800

55,000

56,500

51,500

%4,500

55,0@J

59,000

—- —-.

19,500

-—

59,500e

56,0006

72,000

4.4

3.5

9.2

1.$3

17.4

3.0

-——

2.0

1

1

1

1

1

1

_--—

1/2

-44

-39

-3B

——

-44

———

45,400

62,500

59,700

—— ---

60,800

44,500 40,00C

51,400 40, Ooc

59,300 25,000

t— . .

1

24,790 42,00C

—— -- --—

45,500

88,500

64,500

—--- —-.

25,000xial load,nt. press.

a - &T = oiroumferentialstress; @L = longitudinal stress.

b- Nominal stress computed on basis of original w1l thickness but with respect to greatest diameter obtained at the designated load condition.

c- Computed as lomd on a given section divided by actual cross-sectionalarea, except as noted in footnote~ for specimens failing prematurely. Underlined values indicate direction

of stress presuned to govern failure.

d - Failure occurred at or nea~ ends or end connections,presumably due to high complex stresses caused by bending induced by end restraint.they indicate average stress levels attained before looalized conditions caused failure.

However, results are significantin that

e - Values given are those for mid-section of tube at instant of failure.

f- Shear fracture 6 in. long, crossing longitudinalweld at 1/8 in. from cirownferentialend weld. Fracture started inside of tube, in circumferentialweld.

g - Fracture 4 in. long (apparently shear) at root of circumferentialweld, starting about 6 in. from nearest longitudinalweld;men at angle of about 70° from axis.

then cleavage fracture extending completely around speci-

h- Compre6sion energy in liquid and concrete plugs, and elastio energy in epecinen.

i - Computed on basis of original diameter and ori~ thiokness.— University of California

.i- Tube heat-treated for 8 hrs. at llW°F. , after weldins. Engineering Materials Laboratory

I

1

TABLE 3 ● — RESULTS OF BEND TESTS ON 1O-IN. SQUARE BY 3/4 IN. THICK PLATES OF STEEL AWELDED wITH 5/32 IN. DIAM. FLEE~@ ~ (E601O) ‘W~OD~

Arc Voltage 26-28, Welding Current 150 Amps.

Speci- Rate of Elec- Test MaIc. Bend

Type -f Weld Material

Elong.

trode Travel, Temp. Load Angle, in 1 in. RemarksmenNo. in. per min. oF Kips. Degrees percent

1

2

None

Single bead, ground flush with plate 10

2 tt 11 tt n N 11 10

4567

8

9

10

U

12

13

14

15

16

17

18

19

20I

21

n tf n M M !1 10

Double V butt-weld, stringer beads, cooled to lo700F between passes, as welded

ditto 10

ditto 10

ditt$a 10

ditto 10

Double V butt-weld, wash weld technique, cooledto 70%? between passes, as welded. 2

Ctitt@ 2

ditta 2

Ctitt& 2

ditto 2

Double V but~weld, wash weld technique, cooled ~to approximately 25@F between passes, as tided

ditto 2

ditto 2

ditto 2

dittrz 2

-40 9070 85.332 80.20 79.8

-20 76.1Jo 71.0

70 81.6

32 81.3

0 79*3

-20 7494

-~ 70.6

75 92.0

32 89.5

0 92.0

-20 92.0

-~ 92.6

75 91.0

32 90.5

0 92.0

-20 92.6

-40 95.0

78

78

60

59

33

24

55

~2

32

19

14

82

$2

72

81

59

83

79

77

m

78

28

24

19

20

14

40

22

15

12

7

6

28

28

25

27

23

28

25

27

n

27

No cracks

Cracks in weld onlyn r! tl It

t! 11 tt 11

Cracks in weld and plate

tt 11 II !1 tl

Cracks in weld onlyI

Cracks in weld and plate%

tt u tl tf ItI

t! tt 1? tl It

Tl ft It tt It

Cracks in weld only

#t It tl 1!

!! 11 t? II

?1 II t! tl

Cracks inweld and plate

Cracks inweld only

tf 11 1! II

n 11 tt W

11 tf 11 1!

NO cracks

~ABIE L. — SUMMARY OF BEND TEST RESULTS ON 1O-IN. S2UfJlEBY 3/’4-IN.THICK PL&TEX OF STEEL-A—— -

WELDED BY UNIONMELT PRCCESS+$-.—-

Electrode Diameter 1/4 in., Arc Voltage 26-28

Speci- 19elding Rate of Elec- Test Max..—.—

Bend Elong.—-’men Type of Weld Material Current, trode Travel, Temp., Load Angle, in 1 in.,No.

RemarksAmps● in. per min. ‘F. kips. Degrees percent —-

Cracks in weld only.


Cracks in weld and ,

I

I

22

23

24

25

26

27

28

29

30

31

Single bead, ground flush with plate

ditto

ditto .

ditto

ditto

320

320

320

15

15

15

70

32

-20

92.0

81.0

85.5

75s5

77

68

42

38

66

44

51

56

34

23

24

20

16

13

a

15

18

21

11

plate.o

Cracks in weld and iplate,

Cracks in weld andplate.





Cracks inweld and

15 -40 83.0

-60320 15 82.5

Double V butt–weld, cooled betweenpasses, as welded

ditto

ditto

ditto

ditto

800

800

800

12

12

12

75

32

0

92,1

89.5

92.6

800 12 -20 94.8

-40 94.012plate.

-~Oxweld 36 rod, Grade20 ffi~

.31--

Specimen%ectrode Type ‘$~a~ - ;Percent of EleZGnt Shown

—

No. lf~ Si Mo (jr Ni——. —..——

2

64

70

100

32

76

42

37

47

22

25

52

56

58

62

94

95

Fleetweld 5 26-28V

11 26-281?

11 26-28V11 30-32V

Shie13.arc85 26-28V

f! II 26-28V

L~~i~,l,r 22-24V

Nurex 80 22-24V

25Cr-20Ni 26-28V

thi(mmeltcmSteel A 26-28Vt! !1 lt “ 26-28V

!1 “ Steel Bn 26-28Vfl II !1 “ 26-~.871

It r: Steel BW26-28V1! [I 11 “ 26-28V1! 34-36V1! 34-3~7T

0.18

0.17

0.13

0.15

0023

0.17

0,24

0.24

0,18

(),20

0.17

0.13

0.13

0011

0.12

0.13

0.21

O*46

0.44

0.43

0.45

9.45

0.40

0.58

0.49

0.42

0,44

0.62

0.61

0.60

O*54

0.36

0.38

0.02

0.04

0.10

0.10

0.11 0.61

0.11 0.26

9.06 3.13

0.09 0.09

16.40lL.18

0.43

0.27

0.37

!)..33

0,32

0.33

0,49

0.52— ..—

aSee Tables 3, ~, 6, 7, 3, and 9 for details.

~BLE 6. — SUM ‘~!RYOF TEST RESULTS ON 1O-IN. SCJU.RXBY 3/4-IIT:THICZ FIATE OF STHL-A

WITH BEADS OF SHIELDARC 85, MUREX TYPE 80, ljU_REXA.W.L. KND 25-20 STAINLESS S~IEL ELECTRODES

Electrode Diameter 5/32 in., Welding Current 150 Amps., Rate of Electrode Travel 10 in, per min.

Speci- ‘T= Maximum ‘——~ .——

Bend Elong,.—..—

Type of Weld MaterialArc

men VoltageTemperature Load, Angle, in 1 in., Remarks

No. GF kips. Degrees percent ..—

32

333k35

36

3738

3940

u4.2

43

U

b5

/!+6

47

.48

49

50

51

Shieldarc 85 electrode;ground flush with plate

ditto

ditto

ditto

Murex No. 80, ground flushwith plate

ditto

ditto

ditto

Murex A.W.L., ground flushwith plate

ditto

ditto

ditto

25 Cr - 20 Ni Electrode;ground flush with plate

ditto

ditto

ditto

26-28

i!

II

tt

?1

22-24II

T!

nItIt

II

tl

11II

26--28It

11

1?

u

70

32

0

-20-40

70

32

0

-20

-4070

32

0

-20-40

70

32

0

-20

-40

84.0

78.0

66.0

55.0

59,0

80.0

80.8

80.4

79.0

83.0

84.6

74.7

77*3

76.9

64.5

85.6

86.9

87.3

89.2

88.4

66

45

19

96

73

70

59

7370

75

38

k9

431971

71

69

70

66

2/!+

16

9

71.25

23

23

20

23

2323

13

15

15

720

18

26

22

20

Cracks in weld only

Cracks in weld only




Cracks in weld only

Cracks in weld only

Cracks in weld and plate ‘u

Cracks in weld only ‘wt

Cracks in,weld and plate

Cracks in weld only

Cracks in weld only

Cracks inweld and plate



No cracks

No cracks

No cracks

No cracks

No cracks

I

TABLE 7. -- sWARy Cl?TEST MSULTS ON 1O-IN. SQUARE BY 3/4-IN. THICK PLATES

WITH UNIONMELT !+3ADS

Electrode Di~eter 1/4 in., Arc Voltage 26-28, Welding Current 320 Amps., Rate of Electrode Travel 15 in. per min.

Speci- Test Maximum Bend Elong.

men Type of lWeldYlterial Temperature, Load, Angle, in 1 in.t Remarks

No. OF kips. Degrees percent

22

23

24

25

26

52

53

54

55

56

57

58

59

60

61

62

63

Unionmelt, single bead, groundflush with plate, Steel A

ditto

ditto

ditto

Unionmelt, single bead, groundflush with plate, Ste61 Bar

ditto

ditto

ditto /.

ditto

Unionmelt, single bead, groundflush with plate, SteelBn

ditto

ditto

dttto

ditto

70

32

-20

-40

-60

70

32

0

-20

-40

-60

70

32

c1

-20

40

-60

92.0

81.0 ~

85;5

83.0)

82.5

73.Q~

79..0>

79.0

73.0

81.5 ‘

69.(D

85.0;

8~;Q.

80,0

84.5

85.0

87.0

75*5

77

68

42

38

76.5

66

69

38

50.5

37

71

7?3

86

76.5

57

70

23

24

20

16

13

22

22

21

14

16

13

21

21

26

22

16

20



Cracks in weld and plate.

Cracks in weld and plate t

Cracks in weld and plate. ~~>>

Cracks in weld only. I

Cracks in weld.and plate.


Cracks in weld and “plate.





No cracks.




TABLE 8. — RESUL’ESOF BEiNDTESTS ON1O-IN- SQUARE BY 3/L IN. THICK PLJTES OF STEEL-A

SHOWING THE EFFECT OF PREHEATING AND PCETHEATING——

Arc Voltage 26-28

lSpeci- Elec- Welding Rate of Elec- Test Plate Tap. Temp of Max. Bend Elong.I men Type of Weld Material trode Current trode Travel Temp. before weld-Postheat Load Angle in 1 in.Fai.lureNo. diam, in. Amps, in. per min. ‘F ing, OF. ‘F. kips. D6grees percent

6,465

I 66‘ 67

g

70

E?37475’76

;:7980818283848586878889

E9293

Fleetweld 5, bead ground offdittodittodittodittoditto

Fleetweld 5, bead left ondittodittodittodittoditto

Shieldarc 85, bead left ondittodittodittodittoditto

Unionmelt, bead left ondittodittodittodittoditto

Fleetweld 5, bead left onditto

Shieldarc 85, bead left onditto

Unionmelt, bead left onditto

5(32

tl111!11tf11RuIttl

t!

M

lf

II

ft

174t!

M

1!

tt

ft

5!32

tf

1;4u

150II!f

11

?1

II

t!nH

?1

!!

H

II

ft

lf

u1?

3;0ltItItfltt

15011H

3;0t?

10It

f!tf?1!1}!

M

t!

!1

tf

15Nf?ttt!It

10tf

-401!ftt!t!n

tt

II

11

It!t

1!II

lt11ft

II

f!

u

11

tf

tt

tf

tf

11

tl

-77t!11

lf

o None100 ff200 1!

300 If

400 tt

500 1?

o 11

100 t!lt

300 rt

400 t!

500 It

o tl

100 t!

200 1!

300 lt

400 11

500 tfo !!

100 R200 tt~o !!400 tf500 It

100 d100 e100 d100100 :UX? e

67.081.084.086.081.082.360.072.084.08$.488.08’7.055,070.087.08’7.08~,088,075,082.085.086.088.089.088.090.0$8.085.085.585.5

285081

;:7492649617172?244955

2;w434767

;938381868383

1020252524257

u1921252!46

1219202022121522232A25262326252828

aabbbbaaaaab~aa’aaaaaaaac

;b

:bb

NOTES: a - Cracks inwe d ni

- No cracks?- Cracks in weld only d- Postheated at 11OOOF. A.C. for one hour.

\3- Postheated a l~580~1~~d a?r-cooled for 3 4 hour.

I

TABLE 9--- RESULTS OF BEND TESTS ON 1O-IN, SQUi.i~BY 3/4 IN. THICZ Pli’iTESOF STEEL-A—. .——. —.— —.. ——..SHCT;INGElFECl’OF WELDiIiGVOLTAGE-— . .....-—— -

Testing Temperature -@°F,, No Postheating Treatment

Speci-.——

Elec-—.

Arc Weldingllateof Elec-Plate Temp. Max. Bend Elong.men Type of~eld Material trode Volt-Currenttrode Travel,beforeweld-Load, Angle, in 1 in. RemarksNo. diam.,in.age Amps. in. per min. ing, ‘F. kips.Degrees percent

—.— . . ..— —-.— ————~-—-—-——-——70 Fleetweld 5, bead left on

ditto

ditto

ditto

ditto

ditto

Unionmelt, bead left on

ditto

ditto

ditto

ditto

5/32!!

If

!1

ti

11

U4H

II

It

t!

ff

11

If

11

11

u

n

5/3211

!1

ItII

26-28 150

It !!

11 11

II II

II It

M 1!

11 320It II

u 11

It II

f! 11

If t!

34-36 “

11 1?

11 1?

11 11

1$ tl

17 t!

30-32150If 11

11 f!

11 II

!1 Ii

10fr

!1

tl

lt

t!

15It

o 60.0 9 ‘7

11


ditto

ditto

ditto

ditto

No cracks


ditto I

ditto wW

ditto 1

Cracks inweld only

ditto

100 ?2.0 2671

72

73

74

75

82

200

300

.l@O

500

0

84.0

a$.1+

88.0

87.0

75.0

49

61

71

72

30

.43

47

67

81

19

21

25

24

12

83

84

100 82.0 15

2285.0!1

II

11

.200

86.0300 23

21!$

85

86 400 88.0

87 500 89.0 91 25

7

ditto

67,60 19 Cracks in weld and platedittc~

72.0 12 ditto95 dittf-1 11 100 23

96

97

80.0 34

43

33

53

48

15 dittoditto

ditto

ditto

ditto

Fleetweld,5, bead left on

ditto

dittodittoditto

11

tt

n

If

10u

H

tl

tt

@$o

77.0, 86,0

54,060.0

ditto300400

5000

100

15

ditto98 14

20

36

151226

ditto99

100

101

102103loi+


ditto

2003004.00

83.073.085.0

402883

ditto

dittoNO cracks

-3 b-

Q

ti

8’

/Oc

80

60

40

I?o

o.,

WASH WELO

%

WELDEO HOT (N (ETWEEN EACH

1’

I

DOUBLE-V 6LJTT WELOS 00u8LE

‘ COOLEO!49s)

SINGLE BEAD

I

k/o0

60 -40 -i .?0 40 6

V BuTT WELDS;LOCOLD (coOLED

EACH PASS)

d-1

+

s? BEAOCH PASS)

80 100TESTING TEMPERATURE, “F

FIG3-VARIATIOIV IN BEiW ANGLE AT MAXIMUM LOAD WITH T@iPERA7wE

OF TEST - E 6010 EL EC TROLF. (FLEETWELD 5)-(FoR DETAILS SEE TABLE 3)

7WG. 44.E220

m

SW(ZE 6EA0, GRIIJND FLUSH ~- \

60

0/

$

g

~- 40 ~

ddY

sDLXJBLE-V BUTT b#FL4

TESTED AS WELOEO$!8

20

980 -60 -40 -20 0 20 40 60 #~~ TEhfPERATl#7~eF

FIG. 4 VARIATJOIV IN BEND ANGLE AT MAXLWM L OAil WITH TEAiPE/?A7uRE

OF TEST - LWONMEL T BEADS AND WEL OS - (FOR DETAILS SEE TA8LE 41

DWG. 44E2Z3

STA;NL ESS STEEL, 25Gt - 20Ni

MUREX 8

) ‘

v 1 1/!\ I I I 1

l---l+ W-Sio’u’i’60

r I 1 I I I i [ !-40 -20 0 20 40 60 80 /00

TESTING TEMPERA TURE, e F

1 FIG. 5 — VARIATION lib’ BEND ANGLE AT MAXIMUM LOAD WITH TEMPERATUREOF TEST — ALLOY ELECTRODES. - (FOR OE?2LLS SEE TABLE 6)

1

IA – STEEL & –S TEEL

\ /—

80

- >

(.

0hua 60Qu

> {Q

Q--102 40q t

/.

Q

zm

20

BEAD GROUND FLUSH WITHPLATE.

o-80 -60 -40 -to o 20 40 60

TESTING TEMPERA TURE, “F.

!

> FIG. 6–VARiATJON IN BEND ANGLE AT MAXIMUM LOAD WITH TEMPERA7VRE OFh TEST FOR THREE PLATE MATERIALS. [ UMIONMELTBEADS)

DWG. 44E222( FOR DE TAILS SEE TABLE 7)

4/

100

BEAD GR12uN0 OFF

80

g

% 60&l

b’ BEAD AS WELDED

x$

40

$

m4

,?0

ALL S%Y7MENS TESTED AT -40”FE6010 ELECTRODE [FLEETWEL05)

(

-700 0 100 .?004

300 400 5W 600

TEMPERATURE OF PLATE BEFORE WELDING, “F

1~f~ 8– VARIA TI(?N /ff BEND ANGLE Ar MAx/M/JM LOAD w/TH PREHEAT

I TEMPERA 7URE — SINGLE WELD 6EADS. ‘(FOR LWTAILS SEE TA8LE 8)

#VG 44E225

*

100

1

UHtONMEL r

@o

whb FL EETWELD 5~

~ 60

G

g

w >SHIELOA RC 85

Q40

$c

$?0

ALL SPECIMENS 7F6?ED Ar -40 “F.

SINGLE 6EAOS, NOT 6ROUN0.

o I I 1-100

40 100 Eoo 300 400 500 600

TEMPERATURE Of PLATE 8EFO#E WELDNV6, V

Ii FIG.9—VARIATION IN BENO ANGLE AT MAXIMUMLOAD WITH #?tHEAT> TEMMRA?URE- ZUREE TY~S OF ELECTRODES.

(FOR DETAILS SEE TA8L E 8)DWG 44E221

AS WELDED STRESS RELIEF NORMALIZED

HEAT T@EATMENT + HOUR‘1650 “6

I HOUR‘1100 “F AIR COOLED

FIG 10- EFFECT OF HISAT TREATMENT AFTER WELDING

ON DUCTILITY IN THE BEND TEST.

(FOR DEZAILS SEE TABLE 8)

Im

/“<>

/“”i+ ..

80

UNIONMELT - (.?6 ’28 V) >..q’ ,- -: ;

~ ‘ ./l~~ 60 /

#

//

tQ FLEETWELD 5- (26 ‘28V) .

w..

q 40$

) ~

g UNIONMELT - (34 ‘36V)

% $

2 0, Y / I

spECIMENS TES~ Ar ‘* “*BEADS NOT 6RoUWD

---

~ 100 0 100 200 300 m Sal 600

TEMPEffATL4?E OF PLATE BEFORE WELDIN$, ● F

FIG II - VARIATIOIV IN BEND ANGLE AT WXIAUJM LOAD BW)f P~-~) HEAT TEMPERA7UE FOR DIFFEREWT ARO VOLTAGES AND

TWO TYPES OF WELO. - (FOR DETAILS SEE TABLE 9)

DWG 44E227

-. —

– 1.’> –

WELDMWAL

HEATAFFEC7i5iZONE

PL4TEMETAL

o

a

./6

. /..

.2(

..

.

.

.

.

.

●

✎

✎

●

●

✎

✎

✎

b.

●

-.

zoo 3ZW 400

FIG /4 - SPE-CIMEN 2; FLEETWELD 5 ELECTRODE( SEE TABLE-3)

.. . ...G#— .—. .- ,, .+-.. ,. .,-, —

/,’.-..—.._ . . .. .. -—=

WELDMEZAL

HEATAFFECIE~ZOWE

PLATEMETAL

./(

./5

.2’4

.23

.34

.●

●

●

●●

●

●

D

B.

●

.

.

●

4

●

e●

●

●

P

●

.

●

.

.

.

●

,

a1

-- - —.-/w f-uu SW 400 5iw

K/VOOP HRRDVE3S

F/G Ig) --..SPEZXMEN25; UNIONMEL T ELECTRODE.(S~~ TABLE 4 )

. .. ..-.—

FIG. /7 - GROSS--SL--CT/O/l/ OF WELD BEA/!3,

SPE_C/ME/V 32 ($-INCH PLATE )

b

[;9 ..—. . .. . . . ,, ..__

WELDMETAL

HEAT/WFECTEDZONE

PATEM’TU

6

.03

./6

./’

,Zc

.23

●

I

%%

.

z●

● ,●*

●*.’”

● .●

●

●

✎

?

K/VOOP HRR’’NES5

400

FE 18 -SPECIMEN 32; SH!ELL!ARC 85 ELECTRODE,(SEE TABLE 6 )

. ... . .. . . . . . ..._,

- .==.-.= .,- 4

51 -.—

WELD .WETAL

MEATWFEC7EDZONE

0

.2.

.3

●

✎

●

●

●

☛

●

9

●

.

●

●

●

●

●

●

●

●

●

f

i

t“●

●

●

●

—.200 Xx7

KIVOOP HWDVE=

400

FIG.20 -SPECIMEN 3~ MUREX 80 ELECTRODE;( 5= TABLE6)

FIG, 2/ - CROSS- SEC TIIO/V OF WEL /!2 BEAD,

SPECIMEN 42 ( +-MH PLATE )

WELDMETAL

H’A7AFFEC7EDZONE

PLATEMETAL

●

d

.

.

.

..

.

●

●

1

1,..

9

.

.

.

$●

●

●..●

.

J....*.

K/VQOP HURUVZ3S

FIG Z2-SPECi711EN42i /MYZX AWL ELECTRODE,( SEE TABLE 6)

400 5-

-- ....,.?---..—--=---.. .. .... . .. “ . ,.- . . . ,, ,“-.,-.J.. H. —H-.— ->W--.W.

-55-——. ——

WELDMETAL

PLATEMETAL

,/0

./!5

●

.

.

..

9

●

●

●

.

.

,

.22

●

●

●

F.*.

.

. .

0

KN(20P H’AWVESS

40(7

FIGZ4-SFECIMEN 51; 25 Cr – 20 M. ELECTRODE;( SE. TABLE 6)

~lG,25- CROSS- SEC T/O/v OF WE-LD BEAD,

SPECIMEN 52 ( +-INCH PLATE )

- ;7 -

VELD4ETAL

+EA TlFFECT!Efl‘ONE

o

,05 -

9

●

.●

✎

●

s.

●

●

.

/

..

●

.●

.

.

●

●

Zou Zw 400 5m

KNOOP HBPL2VE55

FIG 26 -SPECIMEN 52; UN/ONMZ T ELECTRODE.(SEE TABLE 7)

~lG,27- CROSS- SEC T/O/v OF WELD BEAD,

SPECIMEN 58 ( +-INCH PLATE )

,.,,,. —..-.—— I

WE--DMETAL

HEATAFFECTEZONE

0

.05

.

.

.

.

.*

4

.

.

.

●

..

●

—

●

●

●

Y

:1

,

.

.

0 zoo 3“ 400 :

KNOW UBPDVKS5

o

FIG Z@-SR5GV14EN58; UNIONMEL T ELECTRODE.(SEE TABLE 7)

..’,

.——.,..7--- —-...—,-—. ,=..—,,—.

F/G, 29- CF?OS.5-SE-(?7/0/4/ OF WELD BEAD,

sPEc/MEAi’ 70 ( $-INCH PLATE )

..—

L1J. --

—.—

U&uMETAL

M5A TAFFECTEDZONE

RATEMETAL

~~.. :... .-% -. ,-’- ‘.

---- ,“- ...’! - ,’

-.,.. .,.

0

.Lu

./d

./5

.

.,

●

●✎

●

✎

✎

✎

✎

●

✎

✎

✎

●

✎

●

✎

�✍✝✍ ❐✍✍

✎

-e

●

.●

.“.

.●

.

.

..

●

.

.

●

.

.—/. zoo 3-- 400 5(W

KNOOP HRR12WSS

FIG 30 -SPECIMEN 70; FLEETWELD 5 ELECTRODE;O ‘F PREHEA To ( SEE TABLE 8 )

.. .. .— ~- .,=- s..,..,,, .+-- ..,,,- . . . . . -. . . . . . . . . . . -,.,..__. _. ... x_.___—., ..,-.,........ .—— ..—,-..

WELDMETAL

HEATAFFECTEDZONE

PLA Ti5METAL

,., ,, ,. . ?d+,. ti’.

. b,C.,

,“. ,’ ”/-.

. .

0

.05

Q ./0

Gk

k./5

i?

1.

● .

b

.

.

●

.

.

.

●

●

.

.

.

.

●

..

/’..

●.....r.

..

.

.50/m zoo 3

>.

9

KIVOOP HRPDNE55

400

FIG 32 -SPECIMEN 71j HEE7W’LD 5 ELECTRODE;100 “F RQEHEAZ (SEE TABLE 8 )

FIG. 33- CROSS- SL-C T/OIV OF WE-L0 isEA D,

SPECIMEN 72 ( ~-lNCW PLATE )

WE-wMETAL

AEATAFFECTECZONE

o

.&

./0

./5

.26

.

.●

✎

●

✎

✎

✎

✌

✌

●

●

✎

✎

✎

❞

●

✎

●

❞

b

.

.

1:\_ .

“ ,0

,.

.

●

●

.

.

,30/m zoo 3’m 400

KNOOP HBRDNESS

FIG. 34- SPEC%’MEN72; FLEETWELD 5 ELECTRODE;200 ‘F PREHgAT (SEE TABLE 8 I

“’

r ——. ----- -= . . “.- “. . . . . . . —..-. a .—— -. .

7

L--- .—!,... . . . . .,.— ,.———. ●

WELDWETAL

HEATAFFEC7E’JZaw

PLATEMETAL

,—— —.——..——— o

.2

.2:

.3

..

●

✎

✎

✎

J

..

.●

.●

..

..

.

..

●

.

..

..

4

.

.

..

.

.●

zoo 3“ 400

KNOOP HWDVESS

FIG.367-SPECIMEN 73; FLEE7WELD 5 ELECTRODE;300 “F PREHEAT ( SEE TABLE 8 )

o

———

-—.-.. .. . .. .. ---.. ........ .......—-——-——- —---

HEATAEFECZEDZONE

RATEMETAL

.2

.3t

●

●

●

✃

✼

o

●

.,

●

.

.

.

●

●

●

.1

●

i .

.

.

.

zoo 3“ 400

KNOOP HB#?DVES5

FIG. 38- SPECIMEN 74; FLEETWELD 5 ELECTRODE;400 ‘F PREHEAZ CSE’E TABLE 8 J

r—WA,., -=.-, ..’ ., . ~ . . . . . .,. . . . . . , >-, , .. e,- _., _ ., . . . . ...<+-= ~.--m , ~, _,==_

1

I —,—,. -.. . . ... .. - .,.Z..—-.

.. ,;—-—...—.

WELDWEZAL

HEAT4FFEC7EZ7ZONE

RATEMETAL

0

.05

./0

,2!

35

..

.,

I

●

.

●

●

.

,

.●

.

●

.

●

.

.

.

b

.

,

●

*

b

—

w zoo 3ZW

Klvt70P H.umNE55

400 5m

FIG @ - SPECIMEN 75; FLEE TWELD 5 ELECTRODE;500 ‘F PREHEAT ( SEE TABLE 81

F/G’. #/ - CROSS- SEC T/OIV OF WELL? BEAD,

sPLEc/ME-N 76 ( $-INCH PLATE )

.. [; –

WELDMETAL

HEATAFFEC7EJ

ZONE

PLATEME7XL

B

o

.05

./0

./5

*2O

1,.

●

a

●

●

.30/00

●

●

●

●

●

●

●

●

●

●

●

●

✎

●

✎

●

●

●

.*

4a

●

.

●

●

.*

●

●*

●

●

.*.

●

200 3“ 400

KNOW H9QDVES5

F/G. 42 -SPECIMEN 76; SHIELDARC 85 ELECTRG!G!~O ‘F PREHEAZ ( SEE TABLE 8 I

FIG. 43- CROSS- SEC T/fOfv OF WELD BEAD,

SPECIMEN 77 ( +MiCH PLATE )

4

- 7: -

WELD ●

MET&

HEATAFFECr50Z(2VE

PLATEMETAL

o

.05

.10

./5

.20

.25

.3tA

1

i

9

9

●

●

‘*●

I

0

KNOOP H..RDNESS

FIG. ~ -SPECIMEN 77; SHIELDARC 85 ELECTRODE;/00 “F PREHEAZ cSEE TABLE 81

.. ‘,’2–

~lG.45- CROSS - .5EL:T/O/v OF WELD BEAD,

sPL”C/M,EA/ 78 (; INCHPLATE )

,...;(-

—.. -. ——

WE”6?METAL

YEAT4FFECZ+E070/vE

PLATEYETAL

25

---

4

●

.

●

!

.

.

.

●

.

●

b.

t

—.—.

.

●

● !●

.—__ .●

●

. --. -9. ...* --- ----*-

*

●—

●

B

●

●

●

●

—— .

●

●

●

●

I

2“ 400

FIG. & -SPECIMEN 78; SHIELDARC 85 ELECTRODE;200 ‘F PREHEAT (SEE Z3BLE 8 J

I

1

WELDMH%L

HEATAFFECTfiZONE

PLATEMETAL

7(-, _I .’

.

.4

.*

.

-d

I

.

●

●

✎

✎

,

,

●

—-

_—

——

~“mv 400

KNOOP H/h?UN’SS IFIG. +@- SPECIMEN 79; SHIELDARC 85 ELECZ?OD~

300 ‘F PREHEA ~ { SEE TABLE 8 I I

.,, ”<:). ..

—-—....-—— — — ‘——--”. . . . .. ..- —-

—-—-— --- -------.-=-- ~-.——

—.

,-

WELQMETAL

HEATlFFECTEL

PLATEMETAL

o

.05

/c

/5

.24

.2:

..

,.

.

●

✌

✌

●

●

✎

●

●

✎

✎

✎

✎

✎

●

✎

●

●

✎

●

●

●

✎

●

✎

✎

✎

,)

.

●

.

.

●

zoo 3ix2 400

KIVOOP H9QDVE5S

FIG. So -SPECIMEN 80; SHIELDARC 85 ELECTRODE400 ‘F PREHEAT (SEE TABLE 8 I

—....-.,-. ...—-

~,.,,,—. ,— —.—..---——-”----—————

F-1(; 5/ “ GROSS.5“- .5E-c ‘f’o/v OF WELD BEAD,

sPEc//b?EN 8/ ( &NCH PLATL- ,)

— -—. —..—. .,—-. —=-....-.— .— —r— ——,—

W..LDIVETA.L

UEAT

ZONE

PLATEMETAL

o

/6

,/:

,Zi

\b

.

)

.

.

●

●

#

● .

●

●

●

I

●

●

●

●

●

●

●

●

●

4

●

●

●

e

●

●

●

●*

--zoo X42 400

KNOW HX7POVE55

FIG. S2 -SPECIMEN 8/; SH’ELDARC 85 ELECTRODE;500 ‘F PREHEAZ r SEE TABLE 8 I

~--- -. . ...+... . .. . “., , .,. . . . . . . .

.—_.__-. —.

MEATIFFECTEL7wi5

PLATEMOAL

0

,05

$,/”

h

.2

,2

.3

.

..

.

●

●

●

.

.

—

.

.

,

●

● I

●

●

●.

,

.●

✎

●

K/vooP HWDNE5S

FIG. 54 -SPECIMEN 82; UNIONMEL T ELECTRODE;0 ‘F PREHEAT CSEE ZZBLE 8 I

-e— . . . . . . . . . ,.. . . . . . . .. . .-—. —..”. -. —.. . “.., -/. ..---.4.-, -., .—.—”-

1

-,,. . ...... - ,“’ ., . . ,.--—----\

g~ .-—,—.... . . . . .. .

WELD/W~AL

HEATAFFECTED2’CVVE

PLATEMETAL

o

.05

.25

-. ’.-,-, ..,,. ,>..~”

3UJ

●

●

✎

●

.●

●

✎

●

✎

✎

✎

✎

✎

✎

✎

●

✎

●

.

.

3’CV

KNOOP HBPQNE55

400

FIG = -JY%’Z2MEN 83i UNIONMELT EECTRODE;IOO”F PREHEAZ ( SEE TABLE 8 )

..%.

~/Is.57- CROSS- SEC ~/ON OF WELD BEAD,

sPEc/MEN 84 (+ INCH PLATE )

- Syj _

———. — . .. .—...- _ _________

WELOMETAL

HEX7AFF.CZEZZONE

RAEMETAL

.2%

.25

.JW

...

..

.

a.

..

..

...

,●.

..

●..

...

.....

. .●.

.

.●

● “

..

●

.

8

—-/w zoo 3’i’i 400

K/VOOP H%?DW3.S

F/G J%- SPECI’MEA/84; UNIONMELT 5 ELECTRODE;200 ‘F PREHEA~ c SEE TABLE 8 I

.- .----,------ .—

FIG. 59- CROSS- SECTION OF WELD BEAD,

SPEC/MEIV 85 ( +-INCH PLATE )

-91-——.. .

WELDMETAL

PLAT-M~Z+AL

0

.(25

.x

.2!

3

b*

.

.

.

.

e.,●

●

.B

.

4

.

●

.

●

,

●

.

●

#

●

●

●

●

.

.

●

.

.

.

.

—.- ---

400

KNOOP WWDVESS

F/ii @ -S}’L”C:IMEN85i U7WOMM5LT ELE~RODE;300”F FF?EHEAZ ( SE’ TABLE 8 )

-.

—... .... ... .... ... ..

—-— .. .---- ... . ~.-. . “—. —.- —-.— — - ,——-

-93-.-——. —

WELDMETAL

PLATEMETAL

0

.05

/c

,/2

.2

.a

.3

●

●

●

●

●

●

●

F●.

●.●

●

J ●

●

. Q 4

KNOOP HDPDNESS

FIG 62- SPECIMEN 86; UNIONMELT ELEGTROC!Ei .400°~ PREHEAZ ( SEE TABLE 8 )

FIG, 63- CROS.S -Sk C T/O/v OF WELD BEAQ

(2 -INCH PLATE )SPECIMEN 87 ~

. . .. —--..-——

WELDMETAL

HEAT!FFEC7”LZONE

PLATEMETAL

o

.05

/G

,/$

,2{

2.

.3

—.

●

I

>

I

zoo 400

KNOW HJQI?DNE55

FIG @ -SPEGIWHV 87; UNIONMEU ELECTRODE;500 “F PREHEAZ ( SEE TABLE 8 )

/,——.

WEZAL

HEAT)IFFEC7E’GZONE

PLATE/WEZAL

o

.03

4

/m zoo SW 406’

KNOOP HRPDNESS

FIG.66- SPECIMEN 88; FLEETWELD 5 ELECTRODE;/ HOUR - //00 ‘~ POSTHEAZ (SEE TABLE 8 )

—._— -

.. .7,.’:s!.. .-. -... —

FIG. G7 -GROSS- .5[-(” T/O/v OF WELD EL12?D,

.SPECIME-N 89 @fiVCH PLATE )

,..-..,--- . ●

WELDMETAL

ALA TEMETAL

-9$’-——..—-.-... . ------ ._,-.—

v zoo Tm 400 5m

K/VOOP H!PDNESS

_ .,----- ..—.—

.. . ....-—

I

.I

.,-1

—... ...“L--.

. “___. _.. —

WELDMETAL

MEATAFFEC7ED

ZONE

PLATEMETAL

-“’P o

.Ot

.11

./:

.2i

.3d

..

●

✎

✎

✎

●

●

●

✎

✎

✎

✎

✎

●

✎

✎

✎

●

●

●

●

✎

●

●

●

●

✎ ●a.

●.●

.●

● .●

●

●

●

●

●

..

●...

zoo 3?W 420

FIG.~ - SPECIMEN 90; SHIELDAR’C 85 ELECTRODE;/ HOUR- //00 ‘F POSTHEAZ (SEE TABLE 8 )

. . ...- ..——

FIG.7/’ -~i?O.SS-SE--C’TIGW OF WE-LD BEAD,

SPECIMEN 9/ (+ ‘INCH PLATE )

-------._ . . .

WifuME7%IL

PLATEMEZiAL

6

*

,

m

0

.

ea

●

B

●

.

,

*

.

.

●

●

.

#

..

●

.

●

..

.

.

zoo A

,[

KN(70P HBPDAK5S

FIG.72- SPECWE)V 9/; SHIE’’DARC 85 ELECTRODE;$ ~~~ -1650 ‘~ POSZ4EAZ ( SEE TABLE 8 )

-... -. . ..... .. . ... . . ,.---— —. .—

.—L.

. ~~~ –,,=,.. ... .. . . . —.. — ,.

WELDVETAL

MEATAFFECTELZONE

PLATEA?EZAL.

R@----’ ‘- t?

.05

Q.L5i?

.24

..

●

✎

✎

✎

✌

●

✎

✎

✎

)

.

●

.

●

●

e●

●

●

...

●,.

●

..s.~

.B

D

--

zoo 3ZW 400

KNOOP HDPDNESS

FE 74 -SPECXVHV94; UIWONMHT ELECZRODE;O‘F I%?EHEAT ( SEE TABLE 9 )

... . . .---- ...

--—---”-””~

F1675 - CROSS- SEC T/OR OF W[L D BEAD,

SPECIMEN 95 ($-W+ PLATE )

, .-*,L .; —

.... ... .. . . ......---———...

WELDMETAL

HEAT4;;F7z-’f:TEL

PLATE/vETAL

fG

./5

.a

.2?

.34

,

—

.

9

●

✎

✎

.

..

●

✏

.

.

4

●

.

.

,

>

..

.*.

●

✎

FE 76 ~SPk”(JIMEN 95- UNIONMELT ELECTRODE”/00 ‘F PREHtAT ( SEE TABLE 9 )

m . . . . . . . . . . ----

109 –

WELDMETAL

HEA ?“AfF’’C7i?ZZONE

PLA7i5METAL

FIG.

,

t

. .●

●

●

✎

✎

✎

✎

●

✎

.

.

a9

●

.

.

.

/

..

.

.

●

——v 2s7 4

KIVOOP H~QLWESS

B -SPELYMEN/00; FLEETWELD 5 ELECTRODE;

j!.. ..

—._

●

_ i,~ -—-. .. .. . ..-....——, .— —.

WELL)MEZ’AL

HEA7AFFEC7WZONE

RATEMETAL

0

.(25

,/0

./5

.Ca

..

...

1

.

.

D

.

.

*.

..

●

✎

.

●

●

●

● I

.●

✎

●

●

●

e.

.●

.

33A

.

KNOOP HRPDVESS

400

FM& -SPECIMEN10/; FLEETWELD 5 ELECTRODE;t/0~”~ PREfiEAZ ( SEE TABLE 9 )

.

,.....=.. . . ..

Documents

006