AllA3 q34 MEASUREMENT OF VELOCITY AND DISPLACEMENT OF ATRANSVAREStRAIN TESTING MACHINE IU MAERIA S RF SARCAR ACOT VALE (AUSTRA IA NJBALDWINFT Al.MYF

N(IA',l11 MRt N 473 ( olII NuuuowuiulullED

I 111 .0 EMII.

I 03..25

-

MICROCOPY RESOLUTION TEST CHARTNATIONAL SoutAU OF SAI.OAAOS -16- A

" - - ,- , I ii ll I I4I.

W-T"73M-003-296

DEPARTMT OF DEFENCEOWO SOINO AMD TCMMOLCOGY OGS

MAATERIAL RORIAC34LSMYBE

WALW MeIR

MRL-TN-473

NEASURENENT OF VELOCITY AMD DISPLACENT OF

A TRANSVARESTRAINT TESTING NAD4INE

N.J. Balduin and B.F. Dixon

DTIC

S ELECTEFEB 15 MD1D

2 1,0_ _ _

r. - a ____IT.

DEPARTMENT OF DEFENCE

MATERIALS RESEARCH LABORATORIES

TECHICAL N=T

MRL-TN-473

MEASUREMENT OF VELOCITY AND DISPLACEMENT OF

A TRANSVARESTRAINT TESTING MACHINE

N.J. Baldwin and B.F. Dixon

ABSTRACT

A tranavarestraint testing machine has been instrumented todetermine the accuracy of the existing displacement measuring system and todetermine the velocity characteristics under test conditions.

measurements of displacement and velocity were achieved andrecommendations for machine and instrumentation improvements have been made.

Approved for Public Release

POSTAL ADDRESS: 01 mer. MueWWal Reemmieb LabwralkP.O. Sea 00. Asse VaOe. Vioud 3022. Ansoel

-inW

4""-st&- *;..;

SECURITY CLASSIFICATION OF THIS PAGE UNCLASSIFIED

DOCUMENT CONTROL DATA SHEET

REPORT NO. AR NO. REPORT SECURITY CLASSIFICATION

MRL-TN-473 AR-003-296 Unclassified

TITLE

Measurement of Velocity and Displacement of a Tranvarestraint Testing Machine

AUTNOR(S) COPORATE MTHRMoterlels Research~ LaboraoresBALDWIN, N.J. and PtO. 1DIXON, S.F. P.O. awx so,Ascot Vale. VlIcorf 3032

REPORT DATE TASK NO. "SON S

MAY, 1983 DST 81/168 DSTO

CLASSIFICATION/LINITATION REVIEW DATE CLASSIFICATIOM/RELEASE AUTHORITYSuperintendent, MELMetallurgy Division

SECONOARY DISTRIBUTION

Approved for Public Release

ANNOUNCEENTAnnouncement of this report is unlimited

Velocity Measurement Solidification CrackingDisplacement Weld Metal

4 Calibration

COSATI GROUPS 11060

ABSTRACT

A transvarestraint testing machine has been instrumented todetermine the accuracy of the existing displacement masuring system and to

determine the velocity characteristics under test conditions.

Measurments of displacement and velocity were achieved and, recaindations for machine and instrumentation improvements have been made.

Acession For

DTIC TAB 1UnInnotmoGed 0JustIf loatio-

Distribution/.--

Avalabity Codes

"SEIOIITY CLASSIFICATION OF THIS PAE

K ,,. Oi icL lasI -v"

* ~ ~ ~ Ds Special~* .~~--

CO0N TE NT S

Page no.

1.* INTRODUCTION I

2. THE MACHINE

2.1 vescripelon of operat ion 2

3. INSTRUMGTATION OF THE MACHINE 2

3.1 Calbratlon 2

4. E2PRIMMAL 3

S . RESULTS 3

6. DISCUSSION 3

7. COUMLUSIONS

* TAMLE I MSTTMTCRL ANALYSIS OF DEFPLECTION RATES FOR THREE

*DIFT WUDIVG PRCESSES 7

TAILE 2 STATISTIZM ANALYSIS Or DEFLECTION RATES FOR DIFFERING

GOUPS or D3ILECTOR VALUES 7

MEASUREMENT OF VELOCITY AND DISPLACEMENT OF

A TRANSVARESTRAINT TESTING MACHINE

1 . INTRODUCTION

h The causes of solidification cracking in ferritic steels have beendiscussed in detail [1]. Three factors which influence crac susceptibilityare composition, strain and strain rate [2,31.

The transvarestraint test is one of a number of procedures used tostudy weld metal solidification cracking (4-6). This test applies a bendingmoment to a plate during welding (See Fig 1). The crack sensitivity of thematerial is determined by the angle of bending necessary to cause cracking.Early investigators used large forces to ensure that the bending moment wasapplied rapidly and consequently assumed that variations in velocity were ofno consequence. Garland and Bailey [71 however, observed that variation invelocity can occur and, by modifying their apparatus to reduce this variation,improved the reproducibility of their test conditions.

Arata and co-workers 181 modified a small tranevarestraint machineso that they could vary its velocity. Subsequent test results on AISI type310 stainless steel indicated that reducing the deflection rate caused cracksto become progressively longer and the number of cracks to be reduced. theyalso identified a critical velocity below which no cracks formd regardless oftotal displacement

The aim of the instrumentation procedures described in the presentpaper m to measure the rate of bending due to the ran and, ubsequently, todetermine the effect of bending rate on the extent of cracking in welds laidon 19 = tickd plates of carbon manganese steel.

i*i ma e as ten'I mesure am the distnc of loading ram travel (displacement). ,•

.i i.,ifli i i i I i; ' ''

2. THE MACHINE

The tranevarestraint apparatus used in the instrumentation exercisewas constructed according to specifications supplied by The Welding Institute,Cambridge, UK IS]. The machine was designed to rapidly apply a predeterminedstrain across a solidifying weld (see Fig. 1).

2.1 Description of Operation

Air is stored in a 56 litre vessel at approximately 700 KPapressure.. A low pressure system (60 KPa) holds the test plate in contactwith a gable-shaped former, with a force of 390 kg. When the weldingelectrode reaches a position 10 ma from the finishing edge of the test plate amicroswitch activates a solenoid valve which dumps air from the storage tankinto a pneumatic ram which rapidly bends the test plate over the former via aloading yoke. The maximum force applied by this ram has been staticallymeasured at 4700 kg.

The amount of bending applied to the plate is controlled bypresetting the gap between adjustable ram-stops and the yoke. The actualdisplacement of the loading yoke is monitored by two dial gauges positionedsymmetrically on each side of the test plate.

3. INSTRUMENTATION OF TE MACHINE

The electronic instrumentation (see fig. 2) added to this apparatusconsisted of two DC.DC displacement transducers (LVDT) mounted coaxially overthe dial gauges, and one clip gauge mounted on knife edges between the ram andcylinder. The LVDT's were powered by variable direct current power suppliesand their outputs monitored by a digital voltmeter. The LYOT's werecalibrated to provide a measure of ram travel and readings were taken beforeand after each test. The clip gauge was powered by a variable divect currentpower supply and its output wee fed into a transient recorder which stored aselected portion of the output and displayed this on a cathode rayoecilloscope. The transient recorder was triggered by the same microswitchwhich initiated the ram movement so that the cathode ray oscilloscopedisplayed clip gauge output against time for the interval during which the ramwas in motion. A series of traces recorded by this technique is shown injFig. 3.

3.1 Calibration

Each LVOT was calibrated with a drum micrometer and its outputadjusted to give an output of 1 volt per millimetre of displacement.

The clip gauge was calibrated with a micrometer head and its outputedjusted to give an output of 5 mV per millimstre of displacement.

2

I L

The transient recorder and cathode ray oscilloscope were adjusted toproduce 2 divisions of oscilloscope displacement per millimetre displacementof the clip gauge (i.e. Full scale displacement of 8 dive - 4 m clip gaugedisplacement).

4. EXPURIDENTAL

In order to determine the effect of using different welding

processes on the rate of displacement a series of three transvarestraint testsinvolving submerged arc (SA), metal inert gas (NIG) and simulated manual metalarc (SNUA) welding were conducted. The three series of tests were intended

to have the same heat input, weld metal composition and surface profile. Thefull results of these including weld metal coWpoitions and welding processvariables are contained in another report [9].

In addition to the above, a subseries of four tests using SA weldingwas conducted with reduced pressures in the storage tank.

S. RESULTS

Load displacement curves recorded from the cathode ray oscilloscope

were analysed to determine "velocity" (v) and "maximum velocity" (v ) asillustrated in fig. 4. These velocities were averaged for each of'Me testseries and the results are presented in Table 1.

The velocities were then regrouped into ranges of deflection and theaverage velocity for each group is recorded in Table 2.

Values of standard deviation (s) were calculated for each groupingin Tables I and 2.

The mot significant result is that the average velocity of the HRLtranevarstraint mchine is a nominal 1.0 m/sec. The standard deviationvalues indicate significant variations in test velocities.

6. DISCUSSION

Although a three point loading rig applies a bending moment to thetest specimens, all ma emnt has been measured in term of ran deflection.The actual angle (0) through which the plate is bent is calculated by

0 , 2 artan A

where 4 - Mdisplaoeent"1 a distance between loading pins (see Fig. 1)

7- -*

As a result of early experiments, certain improvements were made tothe machine. Firstly, the original dial gauges were replaced by moreaccurate metric dial gauges mounted in a follow down configuration for gaugeprotection. Secondly, the screw threaded ram stops were calibrated toimprove the convenience and reliability of presetting the ram gap.

The electronic instrumentation of the tranavarestraint testingmachine provided an indication of the velocity and displacement of the loadingram during a number of tests.

The results of these experiments have provided useful informationabout the operation of the transvarestraint test. It had been anticipated,for example, that different welding procedures may result in different ramvelocities even though the same rate of heat input was used. As can be seenin Table I, the average velocity is 1.0 mm per second far SA and MIG, and 0.9m per sec for SMA. In addition, the value of average maximum velocity is1.2 mm/s for SA and MIG and 1.1 mi/s for SMMA. Allowing for the considerablescatter of results indicated by standard deviation values ranging from 0.1 to0.6, the variation in ram velocities are insignificant, and thereforeconsidered to be unaffected by the welding process.

In theory, the ram force is dependent on the pressure in the storagetank. In one series of tests (L3f 22-25) the maximum pressure in the tank wasregulated down to 40% of the operating pressure in the hope of introducing acontrolled, lower ram velocity. Progressively reduced ram velocities wereachieved, although the lowest ram velocity, i.e. 0.7 mm/s, had also occurredwith maximum pressure in the storage tank. The four results were thereforeaveraged and the results recorded in Table 1. Comparing the "full pressure"and "low pressure" results in Table 1, it can be seen that the decrease invelocity was almost negligible.

The shape of the ram displacement versus time curve showed, in mostinstances, that during each stroke the ram accelerated to a relativelyconstant velocity. The transition from acceleration to constant velocityoccurred at various times and displacements, suggesting that the restrainingforce in each test plate varied.

Statistical analyses summarised in Table 2 show the relation betweenram velocity and final displacement. These figures show an increasing ramvelocity (both mean and maximum) with increasing final displacement. Thisindicates that the ran may be accelerating for up to 2 am displacement.

The considerable variation in velocity identified by these testsindicates that certain modifications to the test apparatus are desirable.These modificitions are discussed below.

Firstly, the acceleration of the ram and yoke assembly isproportional to the force generated by the ram minus the restraining force inthe test plate. The low pressure tests indicate a restriction between thestorage cylinder and the ram. ecause air has to be admitted to thepemmatic ram there is a time delay before the pressure in the ram builds up

4

__ _ __ _ __ _ _.me_ _ __ _ _

from the hold down pressure to the maximum pressure in the storagecylinder. Maximum acceleration of the ram would be achieved by applyingmaximum pressure to the ram before test initiation. This requires arestraining force to prevent the test plate being bent before welding and acontrolled means of removing the restraining force to control the velocity ofthe ram.

Secondly, it is noted that many plots of displacement against timeshow discontinuities; these are probably caused by rocking of the test platewhile it is being welded. The problem may be reduced by guiding the loadingyoke throughout its travel. Rocking of this nature would also causeinaccurate measurements of velocity to be recorded at the clip gauge. It istherefore suggested that the clip gauge be relocated to a position on the sideof the loading yoke and vertically beneath the gable shaped former.

These changes would permit the use of a single displacementtransducer and, therefore, reduce the amount of work involved in processingresults.

Finally, the agreement between qtatic measurements made by theLVDT's, dial gauges and clip gauge was good. However the displacements asmeasured from the dynamic clip gauge output differed from the static clipgauge measurements, sometimes being higher and sometimes lower. Thisprobably occured because the thermal shrinkage forces associated with weldingwere not always balanced by the "hold down" forces applied by the loading

* yoke. This allowed the ram to move slightly from its preset position. Thisaffected only the clip gauge trace.

7. CONCLUSIONS

Investigations of the MRL tranavarestraint machine revealed thefollowing:

1. Averaging of the displacements as measured by the two dial gaugesprovided a reliable measure of ram displacement.

2. Discrepancies between the preset ram displacement and measured ramdisplacement were reduced to within 0.1 me by calibrating the restrainingstops.

4 3. The man of the average ran velocities for 31 test runs was 1.0 mmper second. The lowest recorded average velocity was 0.5 rm/sec and the!'I highest was 1.7 rm/sec. The man of the maximum ram velocities was 1.2 mper second.

4. Ran velocities increased with increasing ran deflection.

5. This transvarestraint machine requires modification to providereliable testing at different ram velocities. These modifications shouldinclude imroved dynamic velocity measuring instrumentation, introduction of avelocity control system and elimination of side movement in the ram.

v.

mJ

8. REFERENCES

1. Dixon, S.F. "Weld Metal Solidification Cracking in Ferritic Steels -A Review". Aust. Weld J 26, 23-30 (1981).

2. Borland, J.C. "Generalized Theory of Super-Solidus Cracking inWelds and Castings". British Welding Journal 7, 508-512 (1960).

3. Prokhorov, N.N. "The Technological Strength of Metals WhileCrystallising During Welding". Svar. Proiz, 9, (4), 1-8 (1962).

4. Brockhurst, P.J. and Muir, H. "Hot Cracking of Welds". Welding andFab. and Des., 9, 80-86 (1965).

5. Farrar, 3.C.M. and Garland, J.C. "Developments in the

Transvarestraint Test for Assessing Solidification CrackSusceptibility of Weld Metals". The Welding Institute ReportM/77/73 (1973).

6. Senda, T., Matsuda, F., Tahano, G., Watanabe, K., Kobayashi, T. andMatsuzaka, T. "Fundamental Investigations on Solidification CrackSusceptibility for Weld Metals with Transvarestraint Test". Trans.Jap. Weld Soc., 2, 141-162 (1971).

7. Garland, J.G. and Bailey, N. "Solidification Cracking During theSubmerged Arc Welding of Carbon-Manganese Steels - The Effect ofParent Plate Composition, Carbon, Manganese, Sulphur andPhosphorus". The Welding Institute Report M/75/73 (1973).

8. Arata, Y., Matsuda, F., Nakagawa, H., Yatayama, S. and Ogata, S."Solidification Crack Susceptibility in Weld Metals of FullyAustenitic Stainless Steels". (Report 111). Trans. JWRI, 6, 37-46(1977).

9. Dixon, B. "Comparison of the Solidification Crack Sensitivity ofThree Welding Procedures", to be published.

Z 77

6.

r i

TABLE 1

STATISTICAL ANALYSIS OF DEFLECTION RATES FOR

THREE DIFFERENT WELDING PROCESSES

NO. OF AVERAGE DEFLN. AV MAX. DEFLN.TEST SERIES TESTS RATE RATE

v a V m amm/s mm/s hur/s N/s

SAFull Pressure (DW01-20) 13 1.0 0.3 1.2 0.6Low Pressure (DW22-25) 4 0.9 0.3 1.1 0.5

MIG(DW26-37) 6 1.0 0.1 1.2 0.3

(DW38-49) 0.9 0.1 1.1 0.2

TABLE 2

STATISTICAL ANALYSIS OF DEFLECTION RATES FOR

DIFFERING GROUPS OF DEFLECTION VALUES

T0T SERIES NO. OF AVERAGE DEFLN. AV MAX. DEFLN.TESTS RATE RATE

V S. V 5-/, -/ m}: ./

0-0.5 am Defln. 10 0.9 0.2 0.9 0.2

0.5-1.0 as Defln. 8 0.9 0.2 1.2 0.3

1.0-2.0 m Defln. 1.1 0.2 1.4 0.3

2.0-3.0 - Defln. 6 1.5 0.3 3.2 2.3

..'., . :. . . . :7

".K 1-7", ., . . . .

T .7

0

00

02.oLe

4I

I

00

40

WAWANIN

0.5mm IDW-021

I.Omm (OW.O11

4 2-0mm IDW-03)

I * 2.Jmm (OW-4ll

rIG. 3 Dispiecesnt versus tims curves depictinq ram velocities.

-- now

z

'V-

ta tiTIME (s)

'Iaverage velocity -

FIG. 4 Mtosocacltnofaverage and maximum ram velocities.

K I 7

ML-TN-473

DISTRIBUTION LIST

MATER TI RESEARCH LABORATORIES

DIRECTORSuperintendent, Metallurgy DivisionDr C.W. WeaverLibrary (2 copies)

Mr B. Dixon

DEPARTMEN '' DEFENCE

Chief Defence ScientJist/Deputy Chief Defence Scientist/ (0 copy)Controller, Projects and Analytical Studies/Superintendent, Science and Technology Programme

Army Scientific AdviserAir Force Scientific Adviser

0 Nikvy Scientific Adviser'Officer-in-Charge, Document Exchange Centre (17 copies)Technical Reports Centre, Defence Central LibraryCentral Office, Directorate of Quality Assurance - Air Force

a Deputy Director Scientific, and Technical Intelligence, JointIntelligence Organisation.

Librarian, Bridges LibraryLibrarian, Engineering Development EstablishmentDefence Science Representative, (Summary Sheets Chly) i

Australia 'High Commission, London.Counsellor Defence Science, Washington D.C. (Summary Sheets Only)Librarian, (Through Officer-in-Charge), Materials Testing

Laboratories, ALEAEP&I, N.S.V.Senior Librarian, Aeronautical Research LaboratoriesSenior Librarian, Defence Research Centre Salisbury, S.A.Director, Weapons Systems Research LaboratoryDirector, Electronic, Research laboratoryDirector, Advanced Engineering laboratorySuperintendent, Trials Resources LaboratoryLibrarian, R.A.3. Research Laboratory'Officer-in-Charge, Joint Tropical Trials and Research Establishment

DIPARTMNT 'OF DEFREN SUPPORT

Deputy Secretary, DOSHead of Staff, British Defence Research a Supply Staff (Aust.)

10" PEAL AND SThTZ DUVAW'DhU AND IMMUEWRL1

HIMA Canberra Offie, Noden, A.C.Y.The Chief Librarian, Central Library, C.S. I.R.0.Library, Australian Atomic hberg Commission Research Establishment

(MRL-TY-473)

DISTRIBUTION LIST

(Continued)

MZCELLAUEOUS AUSTRALIA

Librarian, State Library of WW, Sydney N6WUniversity of Tasmania, Morris Miller Library, Hobart, TAS.

M33CELLINK -'OVESEAS

Library - Exchange Desk, lWtional Bureau of Standards, U.S .A.UK/USA/CAN/lU ABC. Armies Standardisation Representative (4 copies)The Director, Defence Scientific Information & Documentation

Centre, IndiaMilitary, lbval and Air Adviser, High Commission of India,

CanberraDirector, Defence Research Centre, Kuala Laupur, MalaysiaExchange Section, British Library, U.K.Periodicals Recording Section, Science Reference Library,

British Library, U.K.Library, Chemical Abstracts Service

* INSUP: Acquisition Section, Institute of ElectricalEngineers, U.K.

Engineering Societies Library, U.S.A.Aeromedical Library, Brooks Air Force Base, Texas, U.S.A.Ann Germany Documents Librarian, The Centre for Research Libraries,

Chicago Ill.Defense Attache, Australian Embassy, Bangkok, Thailand Att. 0. Ponder

-4w

DAT

*C j><gw

VC.