1-s2.0-S0924013615301436-main

Journal of Materials Processing Technology 229 (2016) 349360

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

jo ur nal ho me page: www.elsev ier .com/ locate / jmatprotec

Finite element analysis of residual stress distribujoined din

Se-Yun HDepartment of heon,

a r t i c l

Article history:Received 28 MReceived in re22 SeptemberAccepted 23 SAvailable onlin

Keywords:Thermo-elastoFinite elementWelding residWelding heat X-ray diffractiElectro-gas we

g movweldiestedech

proceeldmrce mated b

quitepost-

1. Introduction

Recentlymechanicalthan 70 mmto meet thecoaming arport contain(TEU) (Parkplates of a by joining tsuch as sub(FCAW) aremore widelthe erectioning processprocesses inheat input, ductivity duFCAW alwanormally m

CorresponE-mail add

(Y. Kim), jh lee

leading to low productivity. EGW is considered to be an alternativeprocess that may replace the traditional FCAW process due to its

http://dx.doi.o0924-0136/ , the major shipyards have begun using thermos-ly controlled ultra-thick plates with thicknesses of more

and yield strengths of approximately 460 MPa, in order strength requirements near the upper deck and hatchea of very large container carriers. These ships trans-ers from 10,000 to 18,000 twenty-foot equivalent units

et al., 2007; Han et al., 2009). The shear strakes or decklarge container carrier and FPSO vessels are fabricatedhe ultra-thick plates. Traditionally, welding processesmerged arc welding (SAW) and ux-cored arc welding

preferred for welding these thick plates. FCAW is usedy in many shipyards, particularly for butt welds during

procedure and for vertical butt welds during the join- of the side shell. FCAW has some benets over other

that it is relatively easy to automate and requires lessboth of which are key parameters for improving pro-ring the fabrication process of large merchant ships.

ys entails a relatively large number of welding deposits,ore than 15, for the thick-sectioned plates, eventually

ding author. Fax: +82 32 864 5850.resses: [email protected] (S.-Y. Hwang), [email protected]@inha.ac.kr (J.-H. Lee).

higher welding speed and better efciency.Electro-gas welding (EGW) is a continuous vertical position arc

welding process in which an arc is struck between a consumableelectrode and the work piece. EGW entails higher heat input thanother welding processes such as SAW or FCAW because it requiresonly one or two layers of thick weld bead. The high heat inputcan induce a larger tensile residual stress at the weld toe, leadingto reduced fatigue strength. Therefore, the shipping classicationsociety began to focus on the fatigue strength of a thick sectionweldment, (thickness of more than 65 mm), because of the unde-sirable consequences induced by the applied welding process (Banget al., 2009). Some authors claim that deeper investigations need tobe undertaken in terms of the residual stress and fatigue strengthbetween thick-sectioned welded joints fabricated by the FCAW andEGW processes (Teng and Chang, 2004).

As a result of the advantages and disadvantages of both pro-cesses, a hybrid FCAW and EGW process, wherein, the twoprocesses are combined, and the tandem EGW process, whereinthe two electrodes are positioned in a row, have been proposedas shown in Fig. 1 (Sasaki et al., 2004). Welding-induced resid-ual stress and its adverse effects on fatigue strength are still inquestion for these newly proposed welding processes. Therefore, adetailed investigation of the distribution of residual stress through

rg/10.1016/j.jmatprotec.2015.09.0372015 Elsevier B.V. All rights reserved.using two-pole tandem electro-gas wel

wang, Yooil Kim, Jang-Hyun Lee

Naval Architecture and Ocean Engineering, Inha University, 100, Inha-Ro, Nam-Gu, Inc

e i n f o

arch 2015vised form

2015eptember 2015e 26 September 2015

-plastic stress analysis analysisual stresssourceonlding

a b s t r a c t

A computational approach considerindistribution produced by electro-gas nite element analysis (FEA) was suggultra-thick plates. Based on a thermo-mfor EH40 thermo-mechanical control simulate the weaving motion of two won Goldaks double ellipsoidal heat souresidual stress eld on the surface treby the FEA and measurement showedprole. Residual stresses relieved by grinding, is also presented.tion in a thick plateg

South Korea

ing heat sources was introduced to predict the residual stressng (EGW) joints. Considering the two-pole tandem of EGW, a

to evaluate the thermal behavior of EGW applied to a joint ofanical FEA, the proles of the residual stresses are investigatedss (TMCP) steel plates with a thickness of 80 mm. In order toent poles, a quasi-steady heat ux model is introduced basedodel. The X-ray diffraction methods were used to measure they chemical polishing. The residual stress proles determined

good agreement, as regards the values both peaks and of theweld treatments, particularly by ultrasonic peening and toe

2015 Elsevier B.V. All rights reserved.

350 S.-Y. Hwang et al. / Journal of Materials Processing Technology 229 (2016) 349360

Fig. 1. S

advanced nnique are re

In this dimensionathat targetsbutt-weldeIn additionup and coocompared w

What difrom other(GMAW) anwelding elevolume moditional westeady movelectrode inthe difculing electrodthe two-dimstudy, X-Raof weld resiby tandem was compaplicated moconsidered.used in the

Table 1Chemical composition of EH40 (wt%).

Si Mn P S Ni

ical pr

Thick

80

cond

ng con

t (A)e (V)

(cm/mnput (eat in

teria

matscepshowut fral cotivelC

0.08

Table 2Mechan

Steel

EH40

Table 3Welding

Weldi

CurrenVoltagSpeedHeat iTotal h

2. Ma

Theless suFig. 2 mens cchemicrespecchematic diagram of two-electrode EGW (Sasaki et al., 2004).

umerical analysis and a reliable measurement tech-quired.study, a numerical analysis procedure for a three-l (3-D) nonlinear elasto-plastic nite element analysis

the prediction of the residual stress distribution of ad joint made by a tandem EGW process was developed., the actual residual stress induced by the local heat-l-down processes was measured, and the results areith the results from numerical analysis.

stinguishes a numerical simulation of the tandem EGW traditional methods such as gas metal arc weldingd FCAW is the vertical motion and weaving of thectrodes of EGW, which is the main source of the largerlten pool formation. A numerical simulation of the tra-lding process typically only considers the transverseement of the electrode, but some transient motion of the

the tandem EGW process creates difculties. Becauseties are linked to the large heat input of the weav-es, previous numerical studies have been limited toensional simulation (Bang et al., 2009). In the present

y diffraction (XRD) was used to measure the distributiondual stresses on the surface of a butt-welded joint joinedEGW. For validation, the measured residual stress eldred with the numerical analysis results, and the com-tion of the welding electrode with high heat input was

The commercial nite element program MSC.Marc was analysis.

The specrently practwo plates groove shacally arrangweaving mThe shape oing the weldgroove. Theness of 80 man angle of

Table 3 applied to svoltage, and

3. Residua

Residuamal cyclingto generatestress can eyield strengnonuniformcool-down,internal resthe residuastress eld on the theent types osuch as thediffraction,magnetic mnon-destruference betthe materia0.03 1.56 0.009 0.0013 0.01

operties of EH40.

ness (mm) Yield Strength(MPa)

Tensile Strength(MPa)

Elongation (%)

470 582 25

itions of tandem EGW.

ditions Electrode number

1 2380 38042 42

in) 2.8 2.8kJ/mm) 34.47 34.47put (kJ/mm) 68.95

l and welding conditions

erial used in this study is TMCP EH40 steel, which istible to thermal cycling during the welding process.s the weld condition and Fig. 3 shows the test speci-om welded plates. Table 1 and Table 2 summarize themposition and mechanical properties of the material,

y.imen was fabricated according to a welding process cur-ticed in shipyards. Fig. 2 shows the arrangement of theprior to being joined together, along with the detailedpe. The tandem EGW process was applied to the verti-ed plates, with two electrodes moving upward with aotion to generate a molten pool with a larger volume.f the molten pool normally depends on factors includ-ing current, speed, weaving motion, and root gap of the

size of the plates was 1250 2400 mm with a thick-m, and the root gap of the V-groove was 10 mm with

20.lists the welding conditions of the tandem EGW aspecimen fabrication. The same values of electric current,

speed were applied to both electrodes.

l stress measurementl stresses remain after welding due to non-uniform ther- during the welding process. The local heat-up tends

a stress eld during the heat-up process, and theasily reach the yield strength locally due to the lowerth at elevated temperatures. This process can leave a

permanent plastic strain eld (Joshi et al., 2010). After the induced permanent plastic strain eld acts as antraint and causes an elastic stress eld, which is calledl stress. The thermal cycle leaves not only a residualbut also a change in mechanical properties dependingrmal cycle history at a given location. Several differ-f residual stress measurement techniques can be used,

hole-drilling method, X-ray diffraction (XRD), neutron ultrasonic methods, the Barkhausen noise method, andethods. XRD and the neutron diffraction method arective methods using radiation based on the volume dif-ween the normal and distorted alignment of lattices ofl (Price et al., 2008). XRD, which is one of the most

S.-Y. Hwang et al. / Journal of Materials Processing Technology 229 (2016) 349360 351

accurate mresidual stradvantage ois determinthe changedin the anglethe two is g

2 d sin = nwhere d is diffracted XFig. 2. Conguration of the V-groov

Fig. 3. EGW specimens.

Fig. 4. Measurement of the variations in lattice spacing fro

ethods, was used in the present study to measure theesses on the surface of the welded specimen. XRD takesf the fact that the microscopic structure of the materialed by the orientation of the grains. As shown in Fig. 4,

distance between the lattices is reected by the change of the diffracted X-ray beam. The relationship betweeniven by Braggs law (Chun et al., 2007):

(1)

the lattice distance, is the angle of the incident and-rays, and is the wavelength of the X-rays. Once the

lattice distaaccordingly

Using Xparallel to wspecimens.raw specimtive) peak sa peak tenposed that surface rouequipmente.

m the diffraction angle.

nce is calculated from Eq. (1), the strain can be obtained.RD technique, longitudinal residual stresses (stresseseld direction) were measured on the top surface of the

Fig. 5 shows the measured stress distributions on theens without any surface polishing. Compressive (nega-tresses were found in the HAZ while there would besile stress in the arc welding process. It can be sup-the stress prole on the surface was distorted by theghness due to the external forces produced by welding

and subsequent works.


from

Table 4Surface treatm

Thickness

Surface

Thereforical noise wlight grindiface. The sremoved usstresses. Ta

The residimens undewith and wredistributiment techn(UP). Thesestrength buual stress dsurface cracsurface resiand ultrasoas-welded sat the locatiaway from the toe treanegative or level with rplotted in Fthe weld linFig. 5. Measured residual stress distribution with respect to the distance

ent by chemical polishing.

Before treatment After treatment

81 mm 80.7 mme, a clean surface condition without rust and mechan-as required for accurate measurements, and then theng or polishing was necessary to produce a better sur-cales/oxides on the surface of the weld joints wereing chemical etching without disrupting the residualble 4 shows the surface status of the treated specimens.ual stress distribution was also measured for the spec-r different conditions. The specimens were preparedithout toe treatments to check the residual stress

on characteristics according to the different toe treat-iques, such as toe grinding (TG) and ultrasonic peening

toe treatment techniques are benecial to the fatiguet sometimes the toe treatment also changes the resid-istribution near a specimens surface. This can affectk formation during its initiation phase. Fig. 6 shows thedual stress distributions of the as-welded, toe ground,nic peened specimens, respectively. In the case of thepecimen, the welding residual stress reached 300 MPaon of the weld toe, and decreased gradually as it movedthe weld down to the zero level. In contrast, in case ofted specimen, the welding residual stresses began withnear-zero values at the weld toe and converged to a zeroespect to the distance from the weld toe. The stressesig. 6 are transverse stresses; i.e., components normal toe.

Fig. 6. Measuweld toe after

4. Finite e

4.1. FEA pro

Due to thspatial locaear materiacontractionual stress astress predsecutive ananalysis anis the inputapproach uiteration atof both thethe entire a

Becausearc-generatshould be nism. Fig. 7imposed onthe symmeweld line. T the weld toe before the chemical polishing.red residual stress distribution with respect to the distance from the the chemical polishing.

lement model of EGW

cedure

e high heat input used in the EGW process, the extremel temperature variation along with the highly nonlin-l behavior tends to induce non-uniform expansion and

of the material, eventually leading to welding resid-fter cool-down. The numerical analysis for the residualiction of the EGW process was composed of two con-alysis procedures: nonlinear transient heat transferd nonlinear stress analysis. The output of the former

of the latter as a forcing term. A stepwise incrementalsing the NewtonRaphson method was adopted for the

every computational step due to the nonlinear nature heat transfer and stress analysis. Table 5 summarizesnalysis procedure.

the EGW welding process entails the weaving of twoing electrodes moving upward, the numerical analysisable to consider the vertical upward welding mecha-

shows the analysis model with the boundary conditions the model. The bottom of the model was fully xed andtric condition was imposed along the back side of thehe sliding copper shoe, which was used to rapidly absorb


the high temmodeled rethe specimbetween thmodel, it washoe occurrthe analysismesh with near the wesize was apThe size of t80 mm thicthe residuaFig. 7. FEA model of tandem EGW

Fig. 8. Weaving heat source simplied by quasi-s

peratures induced by the high input of the EGW, wasalistically with a mechanical contact condition againsten. Because the complicated thermal contact behaviore sliding copper shoe and the specimen is difcult tos assumed that the heat transfer into the sliding coppered without resistance. To minimize the CPU time during, the mesh size was controlled in such a way that the nea 5 mm length was applied to both the base materialsld line and the sliding copper shoe. An increasing meshplied gradually to the region away from the weld line.he entire FE model was 320 mm long, 630 mm wide, andk. The length and width of the model was chosen so thatl stress distribution would not be affected signicantly

by the remisoparametof nodes an

4.2. Heat in

The mostion is the hheat sourcethe arc genand the heheat sourcebut the spa.

teady heat ux.

oval of the material in numerical analysis. Eight nodeeric hexahedron elements were used, and the numberd elements are 7031 and 5784, respectively.

put model

t important modeling feature in a 3-D welding simula-eat input into the system. The heat input by the moving

has two main mechanisms: the heat ux induced byerated between the base metal and welding electrode,at generation induced by the molten pool. A moving

was modeled by Rosenthal (1941) for the rst time,tial distribution of the heat source was not considered


(Kim et al., be modeleddouble ellipprovides a a moving aracteristics onumerical ssteady movFig. 8.

Fig. 9 shostudy. The hi.e. one in tthe rear qudouble-ellipFig. 9. Double ellipsoidal heat source m

Fig. 10. Temperature-dependent stress-strain curv

2000). The spatial distribution of the heat source might by using the Gaussian surface heat ux model or thesoidal heat source model (Nguyen, 2004). The latter

better representation of the real physics in the case ofc for a butt weld (Kim et al., 2000). The weaving char-f the moving heat source should be considered in theimulation of the EGW process. In this study, a quasi-ement of the heat source was assumed, as shown in

ws the double ellipsoidal heat source model used in thiseat source distribution combines two different ellipses,he front quadrant of the heat source and the other inadrant (Goldak et al., 1984). The power densities of thesoid heat source, qf (x, y, z) and qr (x, y, z), describing

heat ux disource can

qf(x, y, z) =

qr (x, y, z) =

Q = V where qf anthe front angeometric pheat ux shodel.

e of EH40 TMCP.

stributions inside the front and rear quadrant of the heatbe expressed by the following equations.

6

3 f fQ

abcf

e

3x2a2 e

3y2b2 e

3z2c2 (2)

6

3 frQ

abcr

e

3x2a2 e

3y2b2 e

3z2c2 (3)

I (4)

d qr represent the weld ux rates per unit volume ford rear weld pools, respectively, a, b, cf, and cr denote thearameters dening the size and shape of the ellipses ofown in Fig. 9. Q is the heat power, is the efciency, V


is the voltagdened as grate Q splitsof the heat

ff =2

1 + cf/cFig. 11. Temperature-dependent mechanical and thermal p

Fig. 12. FZ and HAZ observed by microstructur

e, and I is the current. The fractional factors ff and fr areiven by Eq. (5). It is assumed that the total energy input

into fractions ff and fr in the front and the rear quadrantux deposited; then, ff + fr = 2 (Joshi et al., 2013).

r, fr = 21 + cr/cf

(5)

The efcientransferrederated by tdependencassumed va2007). The roperties of EH40 TMCP.

e inspection.

cy () is a quantity that represents the ratio of the power from the arc to the base metal to the entire power gen-he arc. The efciency is difcult to estimate due to thee on the welding conditions and molten depth, so anlue of 0.65 was used in the analysis (Armentani et al.,geometrical parameters a, b, cf, and cr were adjusted to


consider thin a quasi-s

The coothrough cosurface waary conditithe surrounapplied to the compleconvection in Eq. (6). InFig. 13. Heat ux distributions of two welding ele

Fig. 14. Simulated temperature distribution at w

e effect of the weaving motion of the welding electrodeteady manner.l-down of the specimen during and after weldingnvective and radiative heat loss on the specimenss considered by imposing the corresponding bound-ons on the surfaces. For convective heat transfer intodings, the natural convection boundary condition wasthe surface exposed to air. Rykalin (1974) introducedx lm coefcient, wherein the temperature-dependenteffect and the radiation effect are combined as shown

the present study, the emissivity (r) was assumed to

be 0.9, and perature. Tdue to the c

h = 24.1

4.3. Materi

Becausevariations mechanicalsis. Fig. 10 sctrodes in FEA.

elding process.

the lm coefcient (h) was set to be a function of tem-he heat ux losses on the surfaces of the welded plateonvection and radiation are given by:

104 r T1.41 (6)

al properties

a material experiences considerable temperatureduring the welding process, temperature-dependent

and thermal properties were considered in the analy-hows the temperature-dependent stress-strain curves


of EH40 TMlinear isotro(2003) andmodel withbehavior ofof the weld metal. The to be those

5. Result o

5.1. Temper

As it is sby a functiparametersFig. 15. History of temperature at the inspecti

CP material. The curves were assumed based on multi-pic hardening properties suggested by Chang and Teng

Masubuchi (1980). A nonlinear isotropic hardening a von Mises yield criterion was applied for the plastic

the base metal. The mechanical and thermal propertiesmetal were assumed to be the same as those of the basematerial properties of the sliding copper shoe were setof copper (Simon et al., 1992) Fig. 11.

f nite element analysis

ature distribution

hown in Section 4.2, the heat source model is denedon of position and time together with a number of, which affect the heat ux distribution. The param-

eters of thtwo most zone (HAZ)residual strheat sourcimate relatinput sugg(2005).

The varichanges theand HAZ zowith them odicular to thwas checkecross sectioof the tempon locations.

e heat source is essential in the prediction of theimportant zones, fusion zone (FZ) and heat affected

associated with the temperature distribution and theesses. To calculate the distribution of heat uxes, thee parameters were determined through an approx-ionship between the weld pool length and energyested by Wahab and Painter (1997) and Gery et al.

ation of the parameters of the heat source in the FEA temperature distribution therefore inuences the FZnes. The FZ and HAZ calculated by FEA were comparedbserved from the cross section of the specimen perpen-e welding direction. Thus, the validity of the parametersd by examining the microstructural distribution of then of the weldment, which enables a reverse expectationerature distribution. Fig. 12 shows the metallographic


section ovethe HAZ an

The parGaussian paiteratively HAZ contouin Fig. 12. TTable 6. TheparametersFig. 16. FZ and HAZ estimated from the temperature

Fig. 17. Transverse stress distribution on the specim

r which the microstructure was examined to observed FZ.ameters were calibrated in an iterative manner. Therameters describing the heat source were then adjusteduntil the simulated temperature produced the FZ andr same as that shown in a macrograph of the test piecehe parameters of the heat source are summarized inreafter, the heat ux of electrodes was dened by the. Fig. 13 shows the heat ux distribution generated by

the two weprocess.

Fig. 14 sculated by thermal cyctemperaturperatures olocations 1The locatiodistribution of FEA.

ens top surface.

lding electrodes running in parallel during the welding

hows the snapshots of temperature distribution cal-FEA during the welding process and Fig. 15 shows theles at six locations depicted in Fig. 14. The maximume inside the fusion zone is above 3000 C, and tem-f approximately 1800 C and 800 C were observed at5 and 30 mm away from the fusion line, respectively.n 30 mm away from the fusion line is considered to be


the heat-affrom the teHAZ is showcontours ofobserved in

5.2. Residua

Figs. 17men weldetransverse men with rFig. 17, is inual stress ddecreasing Fig. 18. Transverse stress distribution on the specime

Fig. 19. Stress distribution along the specim

fected zone. The area of FZ and HAZ were determinedmperature distribution shown in Fig. 16. The predictedn as a series of contours between 850 and 1800 C The

the FZ and HAZ shows a good agreement with those of the welding specimen shown in Fig. 12.

l stress distribution

19 show the residual stress distribution in a speci-d by using the tandem EGW process. The calculatedstress component on the top surface of the speci-espect to the distance from the weld toe, shown in

good agreement with the measured values. The resid-ecayed rapidly as it moved away from the weld toe,to the compressive stress of 200 MPa, and then con-

verged to on the top tion effect the correspues is also similar distface. The dstresses neadue to the sliding coption that thceramic baical analysithe dependmight havens bottom surface.

ens depth.

zero. The sliding copper shoe, which was arrangedsurface of the specimen, did not affect the signica-on the residual stress distribution. Fig. 18 shows thatondence between the measured and calculated val-good on the bottom surface of the specimen with aribution pattern, compared to those on the top sur-ifferent trends between the measured and calculatedr a location 100 mm away from the weld toe might beeffect of thermal contact between the specimen andper shoe, which was modeled based on the assump-e contact between the two materials is perfect. The

cking material, which was not modeled in the numer-s, might have acted as a heat absorber. In addition,ence of the measurement accuracy of the XRD method

contributed as a source of the discrepancy. Fig. 19


Table 5Summary of numerical analysis of EGW.

Geometry Size of model: 320 630 80(mm) Bevel angle, Root gap: 20 , 10 mm Element type: 8-node hexahedron Numbers of nodes: 7031 Numbers of elements: 5784

Welding Condition Room temperature: 20 C (Natural convection) Tandem two weld pass

Material Properties Mechanical property: Isotropic Heat transfer property: Isotropic

Heat Source

Analysis Opt

Table 6Parameters of

a

25 mm

shows the sness directmid-thicknponents.

6. Conclus

3-D nisimulate thtion inducestresses comIn the simuble ellipsoiinput of twsource modby EGW, anThe proposmeasured acorrespondis observedby ceramic

Before tthe surface ing. The resevaluated btoe-treatedgitudinal testrength. Ajoint was thwas similaras FCAW an

Based on the experimental and numerical results, it can bededuced that the present FEA approach is a feasible method topredict the welding residual stress for various EGW joints with dif-ferent welding conditions. Furthermore, the developed approachwill potentially be helpful to predict the HAZ in an EGW jointinvolved in a practical steel structure. It is discussed that the devel-oped approach not only successfully captured the parameters ofheat input but also can be used as a prediction method for theresidual stresses induced by EGW.

Acknowledgement

s reseshor

(MO reset.

nces

ni, E., efc

3193S., Banhnesse. J. Co.H., Teual st522.S., Sony diffr

Nav. AJ., Bibb

sourc, Long

sourcer. Pro

Hildeerial plappin

Semeprotot-Wall, Kim, sfer anhine Tchi, K.., N.T., Plasticity: von-Mises yield Criterion

Model of EGW Vertical welding direction Heat source: Quasi static heat source (Double

ellipsoidal model) Convection: Natural convection coefcient

ion Coupled analysis: Thermal-structure Iterative procedure: Full NewtonRaphson Large strain: Updated Lagrangian

heat source for FEA.

b cf cr ff fr

25 mm 10 mm 10 mm 1 1

tress distribution with respect to the specimens thick-ion. The larger tensile stresses are present near theess area for both transverse and longitudinal stress com-

ion

te element model of tandem EGW was developed toe temperature eld, and the residual stress distribu-d by the tandem EGW process. Meanwhile, the residualputed by FE model were veried by the experiment.

lation, a quasi-steady-state heat source based on a dou-dal heat source model was used to consider the heato electrodes of EGW. Although a relatively simple heat

Thifor OffAffairsversitysuppor

Refere

Armentaweld(1),

Bang, H.tougplat

Chang, Presid511

Chun, K.X-raSoc.

Goldak, heat

Gery, D.heatMat

Joshi, S.,matover

Joshi, S.,in a Thin

Kim, I.S.tranMac

MasubuYork

Nguyen

el was used to calculate the residual stress eld inducedd a good prediction was obtained in the present work.ed process was evaluated by a comparison between thend calculated residual stress distributions. The overallence between the results is good, but some discrepancy, because of surface condition of the specimens affectedbacking material and sliding copper shoe.he residual stress measurements, the scales/oxides onof the weld joints were removed using chemical polish-idual stress characteristics of the EGW specimens werey XRD measurements of the residual stresses for the

specimen and the ultrasonic peened specimen. Lon-nsile residual stress was much higher than the yield

typical feature of welding residual stress in the EGWat the region with high tensile stress was large, which

with that induced by other arc welding processes suchd CO2 welding.

Park, C.S., Sonthe shipbu482487.

Price, J.W.H., Z2008. Comthe issue o

Rosenthal, D., cutting. W

Rykalin, R.R., 1Weld., 12

Sasaki, K., SudDevelopmTech. Rep.

Simon, N.J., Drcryogenic

Teng, T.L., Chalife for but

Wahab, M.A., experimenwelding harch was supported by the Special Education Programe Plants of the Ministry of Trade, Industry and EnergyTIE) of the Korean government, and by an Inha Uni-arch grant. The authors gratefully acknowledge this

Esposito, R., Sepe, R., 2007. The effect of thermal properties andiency on residual stresses in welding. J. Achiev. Mater. Mfg. Eng. 2022.g, H.S., Joo, S.M., 2009. Analysis of mechanical behavior and fracture

KIC in EGW welded joints for high strength EH36-TMCP ultra thickmput. Struct. Eng. Inst. Korea 2 (6), 565572.ng, T.L., 2003. Numerical and experimental investigations on theresses of the butt-welded joints. J. Comput. Mater. Sci. 29 (4),

g, H.C., Im, S.W., 2007. Measurement of welding residual stresses byaction method. In: proceedings of the annual autumn meeting. J.rch., 208213.y, M., Chakravarti, A., 1984. A new nite element model for weldinges. Metall. Trans. 15B, 299305., H., Maropoulos, P., 2005. Effects of welding speed, energy input ande distribution on temperature variations in butt joint welding. J.cess. Technol. 167 (23), 393401.brand, J., Aloraier, A.S., Rabczuk, T., 2013. Characterization ofroperties and heat source parameters in welding simulation of twog beads on a substrate plate. Comput. Mater. Sci. 69, 559565.tay, C., Price, J.W.H., Nied, H.F., 2010. Weld-induced residual stressesype dragline cluster and comparison with design codes. J.ed Struct. 48 (2), 89102.I.K., Moon, C.J., Jeong, Y.J., An, Y.H., 2000. A study on analysis of heatd residual stress on the weld zone using FEM. J. Korean Soc.ool Eng. 9 (5), 96104., 1980. Analysis of Welded Structures. Pergamon Press, Oxford, New

2004. Thermal Analysis of Welds. WIT Press, Southampton UK., Y.R., Lee, J.S., 2007. A study on the application of vertical welding toilding with high deposition rate. J. Soc. Nav. Arch. Korea 44 (5),

iara-Paradowska, A., Joshi, S., Finlayson, T., Semetay, C., Nied, H.,parison of experimental and theoretical residual stresses in welds:f gauge volume. Int. J. Mech. Sci. 50 (3), 513521.1941. Mathematical theory of heat distribution during welding andeld. J. 20 (5), 220234.974. Energy source for welding, Houdrement lecture. Int. Inst.3.a, K., Motomatsu, R., Hashiba, Y., Ohkita, S., Imai, S., 2004.ent of two-electrode electrogas arc welding process. Nippon Steel

90, 6774.exler, E.S., Reed, R.P., 1992. Properties of copper and copper alloys attemperatures. NIST Monogr., 176177.ng, P.H., 2004. Effect of residual stresses on fatigue crack initiationt-welded joints. J. Mater. Process. Tech. 145, 325335.Painter, M.J., 1997. Numerical models of gas metal arc welds usingtally determined weld pool shapes as the representation of theeat source. Int. J. Press. Vess. Pip. 73, 153159.

Finite element analysis of residual stress distribution in a thick plate joined using two-pole tandem electro-gas welding1 Introduction2 Material and welding conditions3 Residual stress measurement4 Finite element model of EGW4.1 FEA procedure4.2 Heat input model4.3 Material properties

5 Result of finite element analysis5.1 Temperature distribution5.2 Residual stress distribution

6 ConclusionAcknowledgementReferences

1-s2.0-S0924013615301436-main

Documents

thicksectioned plates

traditional fcaw process

electrogas welding egw

applied welding process

higher welding speed

tandem egw process

thick sectionweldment

uxcored arc welding