Effect of Buffer-layered Buttering on Microstructure and Mechanical

Properties of Dissimilar Metal Weld Joints for Nuclear Plant ApplicationDinesh W. Rathod1,*a, P. K. Singh2, Sunil Pandey1, S. Aravindan1

* Corresponding author - email: dineshvrathod@gmail.com, Tel.: +44 161 2751916a Present address: MTRL, School of MACE, University of Manchester, United Kingdom, M13 9PL

1 Deptt. of Mechanical Engg., Indian Institute of Technology Delhi, Hauz-khas, New Delhi-110016, India2 Bhabha Atomic Research Centre, Mumbai-400085, India

Abstract

In this study, we present the metallurgical and mechanical investigation of four dissimilar welds

between SA508Gr.3Cl.1 and SS304LN. The welding processes for buttering deposition and fill-pass

welding were varied with ERNiCr-3/ENiCrFe-3 consumables. The Ni-Fe alloy buffer layer was

introduced as intermediate layer in buttering and then the joints (with and without buffer layer in

buttering) were fabricated. The effect of Ni-Fe buffer layered buttering and welding processes on the

resulting weld joints properties has been addressed. Metallurgical and mechanical properties,

fracture toughness were measured and various examinations were carried out for integrity

assessment on all the weld joints. Addition of a Ni-Fe buttering layer leads to the development of

more favourable properties than observed in welded joints made using the current practice without a

buffer layer. Control of carbon migration and its subsequent effect on metallurgical, mechanical

properties due to buffer layer has been justified in the study. Conventional procedure of DMW

fabrication has been proven to be the least favourable against the new technique suggested.

Modification in current integrity assessment procedure would be possible by considering the

properties at interfacial regions, introduction of yield strength ratio mismatch and the plastic

instability strength in the integrity assessment.

Keywords

Buffer layer; Dissimilar Welds; Structural Integrity; ERNiCr-3/ENiCrFe-3; Ni-Fe alloy;

SA508Gr.3Cl.1

1. Introduction

In nuclear power plants, the light water reactor pressure vessel is made of ferritic steel, typically

SA508Gr.3Cl.1, which must be joined to stainless steel pipelines, usually made of SA312 Type

304LN, using arc welding processes. Such dissimilar metal welds (DMW) are typically made with

Ni-base consumables as the filler metal in an attempt to mitigate the strong variation in physical,

chemical and mechanical properties across the weld [1, 2]. It is also common practice deposit a Ni-

1

base (ERNiCr-3) buttering layer onto the ferritic steel before making the joining weld to reduce

carbon migration [1-4]. The Ni-base consumables (ERNiCr-3/ENiCrFe-3) are extensively used for

DMW joint fabrication because of certain advantages of carbon migration. Despite the necessity of

DMW, and the improvement in properties conferred by the use of Ni-base consumables, the desired

design life has not been achieved [3-5] and many failures [3, 6-9] have occurred. The types of

failures and locations [3, 6-9] in DMW joints are still remain the big challenge to assess the causes of

failures. The performance of DMW joints is greatly affected by other often-associated problems [1-4,

10] like degradation of ferritic steel due to oxide notch (low oxidation resistance), metallurgical

deterioration at the interfaces, and residual stresses during buttering and welding. During any

welding procedure, thermal stresses develop which can be detrimental for structural integrity and

performance [11]. In case of DMWs the mismatch in coefficient of thermal expansion (CTE) of

austenitic and ferritic steel can cause different stress profile across the weld joint [1, 12-18]. The

chemical and microstructural variations across the joint can also be severe; the formation of a carbon

depleted soft zone and carbon enriched hard zone which forms due to carbon migration [1, 2, 7, 9,

12, 13, 15, 17, 19-22] has been implicated in failure. Controlling the carbon migration is therefore

crucial to minimising the likelihood of failure of the weld. DMW joints have varying metallurgical,

mechanical and fracture toughness properties across the weld joint, which also affect the integrity of

joints.

Although it has been shown in earlier study of Rathod et al. [7] that carbon migration can be

controlled through the use of Ni-Fe alloy buffer layer in buttering made with Gas Metal Arc Welding

(GMAW) process and ERNiCr-3. The implications of a Ni-Fe alloy buffer layer in buttering using

GMAW process on the mechanical properties of the joint have yet to be documented. Variations in

welding processes also have the potential to influence metallurgical and mechanical properties across

the joint. Standard practice is to deposit buttering layer of Inconel 82 (ERNiCr-3) using Gas

Tungsten Arc Welding (GTAW) process and a completion weld (fill-pass welding) with filler metal

Inconel 182 (ENiCrFe-3) using Shielded Metal Arc Welding (SMAW). This study compares the

microstructural and mechanical properties of joints produced using standard practice and those made

with an intermediate Ni-Fe alloy buffer layer in the buttering deposited with GMAW [7, 23]

technique. GMAW process is not commonly used for preparing DMWs due to mixing of O 2 or CO2

in argon gas shielding to maintain the arc stability. In present study, we demonstrate the use of

GMAW with pure argon shielding for buttering and completion of welds. A comprehensive

assessment of the integrity of the welds was carried out by means of 100% radiographic inspection of

weld joints, all-weld tensile test of weldment zones, composite tensile test, Charpy V-notch test of

weldment zones, fracture toughness of weld, angular distortion, chemical analysis, microstructure

evolution and the micro-hardness measurement across the weld joints.

2

2. Materials and Experiments

2.1 Materials and Welding

The quenched and tempered SA508Gr.3Cl.1 and austenitic SS304LN steel supplied in solution

annealed condition of pipe form were machined into plate form (150x50x18mm) samples with single

‘V’ groove geometry with compound bevel joint design. Four DMW joints were fabricated with two

different buttering procedures and two different completion weld procedures. For two samples, four

layers of buttering, with total of sixteen passes of 2mm diameter ERNiCr-3 TIG rod were deposited

using GTAW process onto the machined surface of two SA508Gr.3Cl.1 plates. For remaining two

samples, an initial intermediate buffer layer of Ni-Fe alloy (ERNiFe-CI) was deposited with GTAW

process, the subsequent three buttering layers were deposited by GMAW process using 1.1mm

ERNiCr-3 MIG wires as described in earlier study [7] to give a total of thirteen passes for the four

layers. The thermal expansion and the tensile properties of consumables and base metals are same as

the properties reported in earlier study [8]. Chemical composition of base metals (BM) and filler

metals (FM) used in study is given in Table 1.

Table 1 Chemical composition of base metals and filler metals

Fig. 1. Schematics of buttering deposits for (A) without buffer layer using GTAW process, and (B)

with Ni-Fe buffer layer using GMAW process

The schematics of buttering layers employed on ferritic steel plates after re-machining for groove

geometry are shown in Fig. 1A and B for the groove geometry of buttering deposits without and with

buffer layer respectively. GTAW buttering employed by 3mm diameter tungsten electrode with

straight polarity using ERNiCr-3 and ERNiFe-CI filler metals. Pure argon gas at 7 L/min was

provided during GTAW while, 14 L/min was employed during GMAW process. The contact tube to

work distance (CTWD) was maintained ~15-18 mm while the wire feed rate adopted was 4.57

3

Materials and Consumables Weight Percentage (wt %)C Ni Cr Fe Mn Nb Ti

SA508Gr.3Cl.1(BM) 0.197 0.53 0.12 96.95 1.30 - -SS304LN(BM) 0.025 8.22 18.09 70.83 0.83 0.01 -

ERNiFe-CI (FM)(TIG - 2.4mm) 0.025 53.01 0.15 43.24 0.74 0.003 -

ERNiCr-3(FM)(TIG - 2mm) 0.017 72.71 19.86 1.40 2.94 2.75 0.41

ERNiCr-3(FM)(MIG -1.1mm) 0.016 72.47 20.01 1.28 2.74 2.88 0.36

ENiCrFe- 3 (FM) (4mm) 0.042 67.17 14.09 6.83 7.51 1.99 0.45

m/min for GMAW process. The interpass temperature during buttering and welding was maintained

between 150-180oC. The process parameters employed for the buttering are given in Table 2.

Table 2

Process parameters for buttering deposition on ferritic steel plates of weld joints

Plates Layer Process Consumables Current (amps)

Volts (V)

Welding speed

mm/sec

Heat input

KJ/mmWithout

buffer layer buttering

Layer 1-4 GTAW ERNiCr3-ɸ2

mm-TIG rod 91-95 8.5-9.5 0.49 1.58

With buffer layer

buttering

Buffer layer 1 GTAW

ERNiFe-CI- ɸ2.4 mmTIG rod

96-100 8.5-9.5 0.49 1.67

Layer 2-4 GMAW

ERNiCr3-ɸ1.1 mm-MIG wire

123-126 23 2.72 1.06

To attain lesser angular distortion, the compound bevel angle provided in the joint geometry

followed by preliminary investigation. The compound bevel geometry as shown in Fig. 2A and B has

been employed for the joints fabricated with SMAW and GMAW processes respectively.

Fig. 2. Schematics of compound bevel joint design for Joints (A) with SMAW process, and (B)

GMAW process

The buttering deposits were examined for defects using Dye Penetrant test as per the criteria of

ASME Sec-V, Article 6. After finding them to be defect free, the completion welds were carried out.

GTAW was used to produce two root passes using ERNiCr-3 TIG rods and back purging. The

subsequent completion welds were made with either SMAW or GMAW. The process parameters

during completion weld are given Table 3. The weld joints were classified as A-1, A-2, B-1 and B-2

according to the buttering and welding processes used and can be seen in Table 3.

Table 3 Process parameters during Weld Joints Fabrication /weld completion

WeldJoint Buttering Pass Process Consumables Current

(amps)Volts (V)

Welding speed

mm/sec

Heat input

KJ/mm

A-1 Without buffer layer

Root GTAW ERNiCr-3 (ɸ2 mm) 110-114 8.5-10 0.86 1.11Fill SMAW ENiCrFe-3 (ɸ4 mm) 100-106 25-28 0.82 3.33

A-2 Root GTAW ERNiCr-3 (ɸ2 mm) 110-114 8.5-10 0.86 1.11

4

Fill GMAW ERNiCr-3 (ɸ1.1 mm) 106-118 24-26 1.04 2.69

B-1 With Ni-Fe buffer

layer

Root GTAW ERNiCr-3 (ɸ2 mm) 110-114 8.5-10 0.86 1.11Fill SMAW ENiCrFe-3 (ɸ4 mm) 100-106 25-28 0.86 3.17

B-2 Root GTAW ERNiCr-3 (ɸ2 mm) 110-114 8.5-10 0.86 1.11Fill GMAW ERNiCr-3 (ɸ1.1 mm) 106-118 24-26 1.14 2.46

All welding activities (GTAW and SMAW, GMAW) were carried out in manual mode as per

requirement of ASME Sec-IX. Uniform dilution was attained by adopting the weaving bead

deposition and the run-in and run-out defects were to minimised by using dummy blocks. The four

as-welded joints after removal of dummy blocks are shown in Fig. 3.

Fig. 3. As welded weld joints after removal of dummy blocks

All four weld joints were subjected to 100% radiographic inspection as per the requirement of

ASME Sec-V, Article 2 and the joints were qualified according to the acceptance criteria of ASME

Sec-III.

2.2 Testing methods and procedure

2.2.1 Specimen Fabrication

The typical length of each weld joint was 150 mm out of which specimens were extracted for

mechanical and metallurgical test. All specimens were extracted and machined using wire cut

electric discharge machine (EDM). The location of specimens on the weld joints is shown in Fig. 4.

Three specimens for composite tensile test (CTT), five sub-size specimens for Charpy V-notch test

from each region of weldment, one specimen for fracture toughness of weld metal by single edge

bend specimen (SEBN), two specimens for metallurgical investigation and remaining length was

used to extract the four specimens from each weldment region for all-weld tensile test.

5

Fig. 4. Location of specimens extracted for the mechanical and metallurgical investigation

The extracted specimens and their respective positions on weld joints are shown in Fig. 5 for all-

weld tensile test. The position of extraction of Charpy V-notch test specimens for each weldment

region shown in Fig. 6. The specimen from HAZ ferritic steel, buttering, and weld metal regions are

shown in Fig. 6A, B and C respectively.

Fig. 5. Position of specimens extracted from each weldment region (HAZ ferritic steel, Buttering and

Weld metal) for all-weld tensile specimens

The extraction position of composite tensile test specimens and fracture toughness (SENB)

specimens are shown in Fig. 7A and B respectively.

6

Fig. 6. Position of extracted specimens for Charpy V-notch test (A) HAZ ferritic steel, (B) Buttering

and (C) Weld metal region

Fig. 7. Position of extracted specimens for (A) Composite tensile test, and (B) Fracture toughness

2.2.2 Mechanical testing

ASTM E8M standard was used for machining and testing the sub-size tensile (standard sheet type)

specimens at the ambient temperature (24oC) for composite tensile test and the all-weld tensile test.

The INSTRON 5582 machine was used with 2.5 mm/min strain rate for all tensile specimens under

this test. The standard sub-size specimens for Charpy V-notch impact test were machined according

to the dimensions specified in ASTM E23 standard by considering the notch position in desired

weldment region. The V-notch dimensions were confirmed with shadowgraph profiler. The testing

was conducted on conventional calibrated machine at the ambient temperature (24oC). The specimen

fabrication and the analysis for fracture toughness (CTOD-Crack Tip Opening Displacement)

specimens of weld metal have been made as per the procedures given in ASTM E1290-02 and

E1820-09 standard. The notch and crack tip in standard single edge three point bend specimen

(SEBN) specimens were positioned in the centre of weld metal and in the direction of welding (Fig.

7B). The pre-cracking and the testing of SEBN specimens have been conducted on MTS 810

machine at ambient temperature (24oC) adopting the standard procedure given in standard. The

specimens for metallurgical test were used for angular measurement with digital image processing

and the results were confirmed by coordinate measuring machine (CMM).

2.2.3 Metallurgical testing

The microstructure of ferritic steel was revealed with 2%Nital for 3-4 seconds after the reaction

started. Stainless steel was etched electrolytically with 10% oxalic acid at 6V, ERNiCr-3/ENiCrFe-3

7

were observed with 10% ammonium persulphate electrolyte etch at 6V. The chemical analysis using

optical emission spectrometer (OES) has been conducted for weldment regions of weld joints. The

chemical compositions of Ni-base weld regions and stainless steel were obtained using the ARC-

MET 8000 (Oxford Instrument) Spectrometer and the SpectroMaxx spectrometer was used for the

ferritic steel. The average values from three locations on the specimens have been used in the

analysis. Similarly, the ASTM E415-99a and ASTM E2594-09 standards were used for measurement

of ferritic steel and Ni-base alloy respectively and the ASTM E1507-07 protocol was followed

during measurement. The hardness variations across the weldment zones have been measured using

Leica VMHT Auto instrument with 10gf and 100gf test loads. Lower load (10gf) used at the

interfaces of weld joints for more precise results and the 100gf load was used in the weldment

regions.

3. Results and Discussion

3.1 Angular distortion

Angular distortion in weld joints are shown in Fig. 8. The angular distortion [24, 25] is indicative of

presence of residual stresses. Residual stresses were not quantified but with angular distortion, it can

be qualitatively compared for the all four weld joints. The completion weld made with GMAW (A-2,

B-2) has shown same angular distortion as an effect of same heat input and associated solidification

rate.

Fig. 8. Angular distortion measured on weld joints

Higher angular distortion observed with the joint fabricated with current practice (A-1) and the lower

with buffer layered GMAW deposited joint completed with SMAW. The effect of buffer layer and

buttering process observed to be significant with completion weld by SMAW than the GMAW.

Variations in terms of heat input, dilution and cooling rate (welding conditions) are more significant

in joint B-1 and that might be sufficient for tempering of HAZ ferritic steel and buttering deposit by

multi-pass SMAW process (high dilution and heat input). Faster cooling rate attained with GMAW

processes was missed in joint A-1. Angular distortion could become the function of formed phases

and their subsequent tempering while rest of the welding parameters are remaining same. These

variations may cause to change the weld chemistry, which is diluted with buttering. The minor

variations in micro-alloying may affect solidification behaviour and resulting angular distortion. The

GTAW and GMAW have the marginal difference in heat input than the SMAW process and

8

therefore, angular distortion in joint A-2 and B-2 is same. Hence, favourable angular distortion (in

terms of lesser residual stresses) can be achieved by new procedure of buttering and welding

compared to the existing one (joint A-1).

3.2 Materials chemistry in weld regions

The chemical composition (OES) measured in different weld regions of the weld joints are given in

Table 4. It can be seen that despite dilution in the buttering deposits and weld metal due to thermal

input of the welding processes, the weld chemistry of buttering and weld metal meets the ASME

Section-II specified chemical composition for DMW joints with marginal variations owing to

dilution (increased Fe in buttering). The samples, which included a Ni-Fe alloy buffer layer, showed

a much lower amount of carbon in buttering region, although levels of carbon in the weld metal (fill-

pass) were similar, than samples without buffer layer. Carbon diffusion could be severe for buttering

region than weld metals because the carbon needs to travel (diffusion) through buttering region to

reach the weld metal, which is extremely long atomic distance for carbon to migrate in as-welded

condition.

Table 4 Chemical composition in the weld regions of all weld joints

Buffer layered samples also showed a higher percentage of iron (Fe) in buttering region which would

affect the solidification and partitioning behaviour of the resulting phases in buttering region [26,

27]. Niobium fraction is reduced in buttering and is significant in buttering without buffer layer than

buffer layer and might caused due to increased solubility of Fe in solution. In weld metal, no

considerable variation in Nb has observed. The increased Fe content in nickel (Ni) matrix could

reduce the CTE [28] and that would help in bridging the mismatch in CTEs of stainless and ferritic

steels [8]. This graded composition owing to buffer layer could provide a gradient in thermal

properties between two dissimilar metals whose CTE are significantly different and such graded

composition is not possible with ERNiCr-3 buttering alone. This variation could also affect the

residual stresses due to sharp differences in thermal properties of the dissimilar metals involved in

joints [29] and further quantification on residual stresses could also be possible and this is at present

is the limitation study.

9

Weld Joint

Zone Weight Percentage (wt %)C Ni Cr Fe Mn Mo Nb Ti Si

A-1Buttering 0.050 68.11 19.07 7.72 2.47 0.05 1.67 0.36 0.22Weld Metal 0.051 65.46 15.81 7.63 6.49 0.06 1.99 0.62 0.98

A-2 Buttering 0.041 68.24 19.02 7.59 2.63 0.06 1.60 0.36 0.19Weld Metal 0.048 65.47 20.52 7.02 3.38 0.16 2.37 0.27 0.30

B-1 Buttering 0.022 62.14 18.49 13.41 2.79 0.38 1.99 0.32 0.23Weld Metal 0.049 64.86 15.94 7.75 6.61 0.06 2.01 0.56 1.19

B-2 Buttering 0.021 62.41 18.22 13.56 2.99 0.13 2.09 0.33 0.21Weld Metal 0.052 65.88 19.67 7.10 3.16 0.18 2.80 0.39 0.25

3.3 Weldment Microstructure

The SA508Gr.3Cl.1 ferritic steel has been in bainitic structure in as received condition of quenched

and tempered pipe forging. Similarly, the SS304LN pipes in solution-annealed condition have the

equiaxed grain structure of austenite twins. The cross-section view of the weldment regions of weld

joints with the extent of HAZs, buttering region and weld metal are shown with macrograph in Fig.9.

The as-received microstructure of SA508Gr.3Cl.1 and SS304LN is shown in Fig. 10A and B. The

fusion interface between buttering and ferritic steel of the joints A-1 and B-1 fabricated with

buttering using without and with buffer layer of Ni-Fe alloy has been shown in Fig. 11A and B

respectively. The HAZ of ferritic steel microstructure of joint A-2 is similar to A-1 while that of B-2

is similar to B-1 due to identical buttering deposition in these joints.

Fig. 9. Macrograph of weld joints showing the weldment regions and base metals

The carbon denuded soft zone (the white phase field) owing to migration of carbon [2] is observed in

joint A-1 and A-2 and shown in Fig. 11A. The soft zone observed to be in width of ~50 micron. The

thin martensite at interface and the extent of decarburised region near the interface are shown by

arrows, which are formed due to sharp composition gradient [7, 30] and diffusion. The

microstructure is pearlite like bainite structure having the significantly fine grain size than parent

metal (Fig. 10A) owing to tempering caused by multi-passes.

The carbon-enriched zone (dark phase field) has been formed due to presence of Ni-Fe buffer layer

in the buttering of joint B-1 and B-2. The arrow in Fig. 11B of joint B-1 is showing the carbon-

enriched hard zone in the HAZ of ferritic steel. The pearlite-like structure with some fraction of

reformed martensite can be seen in the Fig. 11B. The grain structure is also fine compared to parent

metal owing to multi-pass tempering effect and the width of carbon-enriched zone is ~60 micron.

The formation of a carbon-denuded zone has been attributed to the affinity of carbon for chromium,

which is present in higher proportion in the ERNiCr-3/ENiCrFe-3. Carbon migrates from ferritic

steel towards the Ni-Fe alloy but builds up at the interface due to reduced diffusion rate of carbon on

Ni-Fe alloy [7]. In contrast to ERNiCr-3, which has chromium content of 18-21%, the Ni-Fe buffer

layer contains no chromium, thus carbon does not migrate from the ferritic steel to the Ni-Fe alloy

(buffer layered buttering)[7].

10

Fig. 10. As-received microstructure of base metals (A) SA508Gr.3Cl.1, and (B) SS304LN

Fig. 11. Microstructure of HAZ of ferritic steel for (A) joint A-1, and (B) Joint B-1

Hence, the Ni-Fe alloy buffer layer acts as graded composition [7, 21, 23, 31, 32] in the sandwich

pattern of DMW joints with buttering. Ni-Fe buffer layer significantly controls the carbon migration

and respective formation of carbon depleted / carbon-enriched zones in HAZ near the interface [7].

Fig. 12. Microstructure of HAZ of SS304LN for (A) joint B-1 with SMAW, and (B) Joint B-2 with GMAW

Variations in buffer layer and buttering could not affect the HAZ of SS304LN as the SS304LN

comes in focus during completion weld. The considerable variations in HAZ of SS304LN have been

observed owing to heat inputs of welding processes. The joint A-1 and B-1 fabricated with SMAW

process while joint A-2 and B-2 with GMAW process. Hence, the HAZ of SS304LN of joint B-1 and

B-2 of SMAW and GMAW processes are shown in Fig. 12A and B respectively. A clear transition

11

region, known as partially mixed zone (PMZ), is visible in both HAZ as shown with arrows. The

SMAW welded joint has smaller equiaxed austenite grains than those in the same location with

respect to the fusion line in the GMAW weld. This is due to the higher heat input of SMAW process.

The austenite grains in Fig. 12A are smaller while, those in Fig. 12B are bigger than the respective

base metal (Fig. 10B).

The interior of buttering region microstructure of joint A-1 and B-1 (without and with buffer layer) is

shown in Fig. 13A and B respectively. The buttering and weld metal microstructure of ERNiCr-3

and ENiCrFe-3 is fully austenitic and indicating lack of allotropic transformation during

solidification [7, 26, 33]. The as-solidified columnar microstructure of joint A-1 shows much

orientational variation than that of B-1 and the different colour indicates the different

crystallographic orientations of austenite phase. In joint A-1, the buttering alloy (ERNiCr-3) is only

diluted with ferritic steel close to the fusion line, in B-1 however, the buttering alloy (ERNiCr-3) is

diluted with Ni-Fe buffer layer. Whereas, the Ni-Fe buffer layer itself diluted with ferritic steel at the

interface. Therefore, the buttering deposit with buffer layer would contain higher amount of Fe

compared to the buttering deposit without buffer layer. The increased amount of Fe in buttering

deposit leads to decrease in the Nb and Ti solubility in austenite phase and their ability to remain in

solution is getting limited [26].

Fig. 13. Interior of buttering microstructure for (A) joint A-1 without buffer layer, and (B) Joint B-1 with buffer layer

Fig. 14. Interior of weld metal microstructure for (A) joint A-1 – SMAW , and (B) Joint B-2 – GMAW process

12

The considerable content of Nb and Ti (3-4%) increases the bulk solidification temperature range

[27, 34] due to presence of more amount of Fe in buttering deposit. This increases the degree of

constitutional undercooling and causing the dendrite structure to widen [35]. Hence, the primary arm

spacing and dendrite width in Fig. 13B of joint B-1, with buffer layer, has been wider than the Fig.

13A of joint A-1, without buffer layer. Owing to more amount of Fe in buttering deposit with buffer

layer than without buffer layer, the solubility of Nb in solution is decreases. Therefore, the

partitioning of Nb and Ti to the interdendritic region is increased and it would result in formation of

laves (Fe2Nb) phases [7, 23]. Hence, the significantly more fraction of laves phases (dark spheroidal

indicated with arrows) can be seen in Fig. 13B of buttering with Ni-Fe buffer layer than without

buffer layer (Fig. 13A).

The weld metal microstructure of joint A-1 and B-2 with SMAW and GMAW process is shown in

Fig. 14A and B respectively. The similar microstructure is developed in each case due to use of the

same filler metals. The use of coated electrodes in SMAW process (joint A-1) leads to very fine slag

inclusions [8] as encircled can be seen clearly in micrograph of Fig. 14A. Such slag inclusion is not

present in Fig. 14B due to flux-free GMAW process. The chemistry of weld metal for Nb and Fe is

almost similar with marginal variation in each weld joint. Hence, due to fine slag inclusion [8],

different cooling rate and heat input during welding, some marginal variations have been observed in

joint A-1 and B-2. Fine secondary phase particles at grain boundary have been shown with arrows in

Fig. 14A. The migrated grain boundary (MGB) and solidification grain boundaries (SGB) are clearly

visible and arrows indicate them. The slightly courser grains were observed with GMAW process,

and the SGBs and MGBs are more prominent. The arrows in Fig. 14B indicate the shiny NbC phase

particles. Laves phases are not observed in weld metal of GMAW process perhaps due to the faster

solidification rate in GMAW process.

3.4 Micro-hardness evaluation

Micro-hardness variations across the weldment regions of all four weld joints are shown in Fig. 15.

Hardness in HAZ of ferritic steel has been increased due to reformed martensite fraction [36].

Martensite formation at the ferritic steel and buttering interface due to weld chemistry variation [7,

30] has been evidenced. The less heat input during GMAW than GTAW process has not caused the

tempering of martensite as an effect of multi-pass buttering deposition. This caused to have the more

hardness in HAZ of ferric steel for joints buttered with GMAW process. The effect of carbon

enriched / denuded zone within ~50 micron distance in HAZ ferritic steel from interface is

considerable.

13

Fig. 15. Micro-hardness variations across the weldment regions of four weld joints

The composition gradient in buttering layers and weld metals (ERNiCr-3/ENiCrFe-3) could remain

almost constant due to filler metal chemistry [1] hence the hardness in weld metal is not expected to

vary significantly. Minor variations could be possible in weld region because of different heat inputs.

Considerable variations in buttering near the interface are observed due to significant chemistry

variation and dilution. The hardness variations across the weldment regions of weld joints in present

study are in agreement with the earlier research [7-9, 12, 13, 17, 36, 37].

3.5 Composite Tensile Test

The engineering stress strain curves for the specimens of all four joints are shown in Fig. 16. All

specimens were fractured from the weaker portion of the base metal SS304LN except one specimen

from joint B-1. This specimen has been fractured from buttering region.

Fig. 16. Engineering stress strain curves for all weld joints

The yield strength (YS) and ultimate tensile strength (UTS) of specimens from joint with Ni-Fe

buffer layer observed to be more than without buffer layer joints. The obtained results from the

composite tensile test suggest the required strength and ductility for the DMW joint requirement

along with structural integrity.

14

3.6 All weld tensile properties of weldment regions

The tensile properties YS, UTS, uniform elongation (UE) and total elongation (TE) along with

associated properties, as yield strength ratio (YSR) and plastic instability strength (PIS) were

determined.

Fig. 17. Typical stress-strain curves of weldment regions for weld joints (A) A-1 and B-1, and (B) A-2 and B-2

The typical stress strain curves for the specimens of joint A-1 and B-1 are given in Fig. 17A, while

those of joint A-2 and B-2 are shown in Fig. 17B. These curves are the example of the measurement

made for all samples and their results were used for presenting the tensile properties in this study.

The average YS observed in the weldment regions and base metals of the all weld joints is shown in

Fig. 18.

Fig. 18. Average Yield Strength (YS) for weld joints

The YS in HAZ of ferritic steel region is more than parent metal except joint A-1 (conventional) due

to carbon depleted zone and effective tempering of the reformed martensite. However, the joint A-2

has marginally more YS than parent metal due to completion weld by GMAW which was SMAW in

joint A-1. The YS of buffer-layered joints (B-1/B-2) is considerably more than parent metal and the

joint A-1/A-2. The formation of carbon-enriched zone and insufficient tempering of martensite due

to GMAW buttering have caused to increase the YS and the results are in agreement with the

hardness profiles. Yield Strength in buttering region of all joints is almost same.

15

Fig. 19. Average Ultimate Tensile Strength (UTS) for weld joints

The weld chemistry variations are considerably significant (section 3.2) for the buttering deposits

due to dilution and buffer layer. This caused to have marginal compromise in chemical variations

suggested in ASME Sec-II, particularly for Fe content (more than specified). These variations have

not made any adverse effect on the YS of buttering regions. The YS in weld metal of GMAW

welded joints is more than SMAW joints due to presence of fine slag inclusions that affects the

dislocation mechanism. The UTS of all weld joints is shown in Fig. 19. The UTS is more than parent

metal (SA508) in HAZ region and is not seen with joint A-1 as seen with YS. This caused due to the

same reasons explained for YS, which seems to be almost consistent with rest of the joints with more

strength. The UTS in buttering and weld metal region is observed with same trend as YS. The lowest

UTS observed with joint A-1 (conventional) while the rest of joint showing considerably more

strength. The average UTS in joint A-1 is more than parent metal SS304LN but, the scattering in

weld metal of this joint cannot be truly considered safe for integrity assessment.

Fig. 20. Average Uniform Elongation (UE) for weld joints

The uniform and total elongation in the weld joints can be seen in Fig. 20 and Fig. 21 respectively.

The elongation is almost consistent with all joints for the HAZ ferritic steel region, which is

marginally less than the parent metal. The elongation (UE/TE) observed to be more with GMAW

16

process with buffer layer for buttering and weld metal. The UE and TE are not favourable and

desirable against the all weld joint in conventional procedure (joint A-1).

Fig. 21. Average Total Elongation (TE) for weld joints

The variation in UE/TE is in accordance with the variations observed in YS and UTS for the same

associated reasons. The tensile properties like UTS and UE influence the deformation and type of

fracture in the material. As in case of DMW joints, these properties significantly varied across the

weldment regions. PIS values in each material zone have been calculated using the equation1.

PIS = UTS x (%UE/100 + 1) (1)

PIS is the true stress that reflects the influence of UTS and UE and the estimated PIS values for all

weld joints are given in Fig. 22. This strength is almost independent of stress concentration but

strongly depends on temperature [8, 9, 38]. It represents the resistance to local necking initiation and

used as local failure criteria for ductile materials [8, 9, 38]. Hence, the calculated values of PIS for all

joints in ductile (austenitic) region of buttering and weld metal have been examined.

Fig. 22. Average Plastic Instability Strength (PIS) for weld joints

PIS in HAZ ferritic steel of joint A-1 is considerably less than the rest of joint and is not desirable.

The favourable PIS profiles are noticed with the joints welded with the GMAW process due to the

better UE and UTS owing to process compared to SMAW. The effect due to buffer layer and

17

subsequent carbon migration can be notice with PIS variation in HAZ ferritic steel. Based on PIS, the

values with buffer layer and buttering deposition as well as welding with GMAW process is found to

be the favourable against conventional DMW joint manufacturing procedure for the ductile failure

criteria. Buffer layer and heat input variation (GMAW/SMAW) causes the favourable and desirable

effects especially in buttering and weld metal regions. This makes the YS, UTS, UTS, UE, TE, and

PIS better than without buffer layer joints. On the extent of identified properties, the excellent and

desirable properties are found in this sequence for joints B-2, A-2, B-1 and A-1. This suggests that

buffer layer introduction is adorable against the conventional procedure (DMW fabrication) while

the GMAW process suggest the better strength than SMAW process for the welds. Hence, using

ductile failure criteria, it can be suggested that the conventional procedure joint (A-1) is more

susceptible for failure compared to other joints in the severe plastic deformation regime.

The strength mismatch between the weldment regions and base metals could significantly affect the

crack driving force, crack growth resistance and the strain concentration location under fully plastic

condition [38]. The yield strength ratio (YSR) has been used [8, 9, 38] as a measure of weld strength

mismatch. Both base metals and consumables have the different tensile properties, which can lead to

vary strain concentration location in weldment regions accordingly [8, 9, 23]. The equation 2 has

been used to calculate the YSR for all joints for the ferritic and stainless steel side.

YSR = YSWM / YSBM (2)

Where the YSWM and YSBM are yield strength of weld metal and base metal respectively. The YSR

more than one (overmatching) is generally desired, and if it is less than one (under-matching) then,

strain concentration can occur in weld region and fracture can be initiated from pre-existing defects

[8, 9, 23, 38] if any presents. The effect of mismatch is significant when the strength of mismatched

material exceed 10% [38].

Therefore, the estimation of YSR would be important for integrity assessment of DMW joints. The

calculated YSR for different locations like ferritic steel side (weld metal / base metal and buttering /

HAZ ferritic steel) and SS304LN are summarised in Table 5. It can be seen that for any process the

YSR is only overmatched on one side of the weld, the SS304LN. Current welding consumables do

not provide overmatching on the both sides and this is inherent problem for DMW integrity.

Table 5 Yield Strength Ratio (YSR) for all plate joints

Weld Joint YSRWM-BM508 YSRBT-HZ508 YSRWM-BM304

A-1 0.54 0.70 1.39A-2 0.64 0.60 1.65B-1 0.56 0.56 1.46B-2 0.61 0.58 1.58

The YSRWM-BM508 observed to be more favourable with joints welded with GMAW (Joint A-2, B-2)

process than SMAW (joint A-1, B-1) and is important for integrity assessment. The YSRBT-HZ508 for

18

joints with buffer layer observed to be less than the joints without buffer layer. This indicates that,

plastic strain concentration and fracture initiation do not occur initially in ferritic steel base metal,

HAZ ferritic steel, and weld metal but rather at SS304LN and/or in buttering region of these joints.

Considering the YSRBT-HZ508 for joints (Table 5), the results are found to be consistent with composite

tensile test during fully plastic deformation (Fig. 16). Hence, the one sample from joint B-1 in CTT

has been fractured from buttering region but satisfying the required strength as per DMW

requirement as per ASME [8]. The estimation of YSR could contribute the value addition in integrity

assessment [8, 9, 38] and same thing has been justified with this study. The under-match in weld

strength is not desirable but that is inherent problem in DMW fabrication. The worst mismatch in

terms of YSR needs to be included (quantification range) in the integrity assessment procedure and

codes to ascertain the location and type of failure mechanism.

3.7 Charpy V-notch impact toughness

The average impact toughness for all weldment regions of weld joints are presented in the Fig. 23.

The impact toughness of HAZ ferritic steel region has been observed to be more than base metal

ferritic steel due to fine grain structure of HAZ (FGHAZ) ahead of the crack tip (V-notch). The

coarse grains in HAZ (CGHAZ) are ~50 micron wider (Fig. 11A and B) hence, the exact positioning

of V-notch in CGHAZ is not possible. Therefore, the notches are positioned in the FGHAZ which

having the hardness 250-275 HV100 (Fig. 15). The FGHAZ and fraction of reformed martensite as

discussed earlier caused to increase the impact toughness marginally in the HAZ of joints than the

joint A-1. The weld chemistry in buttering and the resulting phase structure has shown the increased

impact toughness of buttering region due to buffer layer and GMAW buttering process. Same thing

is also revealed in weld metal region.

Fig. 23. Charpy V-notch impact toughness/energy for weldment regions of weld joints

The buffer layer buttering process / welding process variations suggest that the joints B-1, B-2 and

A-2 have the favourable profile of impact toughness in the spatial region of DMW and this is far

19

better than the joint fabricated with the existing procedure (joint A-1). Based on spatial variations in

DMW for impact toughness, the buffer layered joint are most desirable by integrity assessment.

Fig. 24. Fracture surfaces of Charpy specimens for buttering region of weld joints (A) A-1, without buffer layer and (B) B-2, with buffer layer

The impact toughness in weldment regions is found to be consistent with microstructure and the

tensile properties in the study. The fracture surface of buttering region without buffer layer (joint A-

1) and with buffer layer (joint B-2) is shown in Fig. 24A and B respectively. Large number of small

size dimples shows evidence of ductile tearing, as cab be seen in Fig. 24A. Arrows indicate the

second phase particle nucleated micro-voids in buttering region. The multifaceted surface is

observed to be dominant than small size dimples in the Fig. 24B owing to variations caused by Ni-Fe

buffer layer. The multifaceted surface indicates ductile tearing along the grains. While, the ductile

tearing along the grains is significantly less in Fig. 24A (joint A-1) than the Fig. 24B (joint B-2) and

this is found to be consistent and in agreement with the observed results.

Fig. 25. Fracture surfaces of Charpy specimens for weld metal region of weld joints (A) A-1, SMAW process and (B) B-2, GMAW process

The fracture surface of weld metal of SMAW (joint A-1) and GMAW (joint B-2) process are shown

in Fig. 25A and B respectively. The presence of fine slag inclusion due to coated electrodes in the

ductile dimples is clearly visible in SMAW (Fig. 25A) but absent in weld of GMAW (Fig. 25B)

process. The inclusion and second phase particle nucleated micro-voids are shown with arrows in

Fig. 25A and B. The dimple size in Fig. 25A is observed to be smaller than Fig. 25B (GMAW

20

process) and encircled in the figures. These observations are consistent with the recorded impact

toughness in the joints. The dimple orientation in buttering region shows the columnar structure

when compared with the fracture surface of weld metal.

3.8 Fracture toughness testing

Typical Load - crack mouth opening displacement (CMOD) diagram for weld metal of all weld

joints has been analysed on the specimens (Fig. 26A) and the Load-CMOD plots are shown in Fig.

26B. The CMOD values are found to increase steadily with load in the elastic-plastic condition.

Hence, the CTOD ( ) corresponding to the maximum load point was employed for the evaluation

of CTOD ( ). For the computation of CTOD from the CMOD value, the total CMOD

corresponding to maximum load point was divided into two parts, first belongs to CMOD

corresponding to the elastic part (ve) and other one is CMOD corresponding to the plastic part (vp) of

crack opening. The elastic (δe) and plastic (δp) parts of CTOD were calculated [39] using equation 3

and 4 respectively.

(3)

where KI is the stress intensity factor corresponding to the critical load, E is the elastic modulus, σYS

is the yield strength and υ is the Poisson’s ratio.

(4)

where Vp is the plastic component of CMOD corresponding to the critical load, a is the original

crack length, W is the width of the specimen and r is the rotation factor which may be taken as 0.4 as

per standard.

Fig. 26. (A)Test specimen after CTOD test, and (B) Typical Load-CMOD plots with

21

obtained results

Total CTOD (δc) was calculated using equation 5 as given below;

(5)

Stress intensity factor in case of SENB test was calculated using equation 6

(6)

where fα is the function of α= (a/W) and given as

where P is the maximum load, S is the span length and B is the thickness of the specimen. For CTOD

testing, according to ASTM 1290-02, the ratio of a/W is given as 0.45 ≤ a/W ≤ 0.70. Whereas, the

a/W ratio of specimen in this study is ranging from 0.51 – 0.56. Hence, all specimens have satisfied

the a/W requirement. The CTOD values for the joints welded with GMAW process have been

observed to be significantly higher than the welded by SMAW process owing to presence of fine slag

inclusion in the weld. As an effect of buttering deposition with buffer layer/buttering process, the

fracture toughness (CTOD) has been observed to be decreased marginally in buffer-layered joints

(B-1 and B-2) than joints A-1/A-2. The resulted weld chemistry in weld metal due to different

buttering deposits and dilution as shown in Table 4, the Fe and Nb content are marginally more in

buffer layered joints (B-1/B-2) than the joint A-1/A-2. The laves phase formation at grain boundary

would be more pronounced with joint B-1 and B-2 weld metal and that may cause to marginal

reduction in the CTOD fracture toughness. The better fracture toughness can be achieved by GMAW

process than the SMAW process welding.

4. Influence of investigated properties on integrity assessment

The reliable and precise methods for structural integrity assessment of DMW joints are not available

[9, 37]. The designing of DMW joints and its assessment is made at present on the figures/facts and

information obtained from similar metal welds. The structural integrity assessment at present is

depend on the outputs from several years of experience [9, 40]. According to the integrity assessment

described in R6, European method SINTAP, and FITNET FFS[41-43], the dissimilar weld is

considered as the sandwich combination of parent metals and different weldment region metals[9].

Variations in metallurgical and mechanical properties in HAZs and the regions at fusion interfaces

are significant due to dilution and chemistry mismatch. However, these variations and its resulting

effects are not considered in existing procedures and codes [9, 37] for integrity assessment. Under-

matching and overmatching the weld strength with base metals is inherent property of DMWs

integrity and the defects or failures may appear anywhere in the weldment regions. The under-

matching is serious concern and need to be quantified as the minimum required criteria for the

22

strength under-matching and ductile failures in severe plastic deformation. The variations weld

strength in properties of buttering, weld metal and even in HAZs could occur due to the weld

chemistry mismatch and that is apparent owing to composition gradient and the active diffusion

mechanism [7].

The fracture and failure mechanisms cannot be predicted for the notch positioned in any specific

weldment region or at interface owing to micro-segregation (chemistry mismatch) of the formed

phases and intermetallic compounds. In this condition, the initiation of crack and its propagation path

will be defined by the formed phase structure and the intensity of micro-segregation in very narrow

zones near the interfaces. This can suggest that, the existing procedure and codes cannot be really

used with the desired accuracy by compromising the failures in terms of nuclear hazards. The

variations in mechanical properties have significantly affected the fracture resistance and crack

growth path [37, 41]. The different metallurgical and mechanical properties with marginal disparities

in same material region could remain exist [37, 44, 45] in weld materials. Crack initiation resistance

and its propagation can greatly be affected by heterogeneity in mechanical properties across the

weldment regions [9, 46]. Therefore, the development and modifications in existing procedures and

codes of DMW integrity assessment can be made. This can be done by considering the local

mechanical properties with due considerations of HAZs and interfacial zones by analysing the micro-

segregation, weld chemistry mismatch and the metallurgical properties in terms of formed phase

structure.

5. Conclusion

In present study, welding processes, buttering processes and Ni-Fe buffer layer variations were

analysed for metallurgical, mechanical and fracture toughness properties. The derived conclusions

from the analysis are listed here.

1. Lesser angular distortion (can be in terms of measure for residual stresses) can be achieved by

introducing the buffer layer, varying the buttering process and welding process. This distortion

is significantly less than the existing procedure of DMW fabrication.

2. The Ni-Fe alloy buffer layer in buttering causes the certain variations in material chemistry of

buttering and this is desirable for the thermal, metallurgical and mechanical properties than the

buttering without buffer layer in the weld joints. Owing to Ni-Fe buffer layer, weld metal

chemistry in buttering changes and causes the effect on solubility of Nb and Ti due to increased

Fe content. This leads to more fraction of laves phases and wider dendrite arms compared to

buttering without Ni-Fe buffer layer.

3. Carbon depleted soft zone is formed (~50 micron) in HAZ of SA508Gr.3Cl.1 of the joints

without buffer-layered buttering while, the carbon enriched hard zone (~60 micron) is formed

23

with Ni-Fe buffer-layered joints. This variation justifies the control of carbon migration owing

to buffer layer in the DMW joints.

4. The tensile properties with buffer layer joints are more desirable and favourable. The most

favourable properties are recorded with buffer layered buttering joint welded with GMAW

process. While, the least favourable properties are found with conventional joint (joint A-1).

Charpy impact toughness in the weldment regions recommends the Ni-Fe buffer layer and the

GMAW process for welding of such DMWs. Weld strength and ductility of the weldment

(composite tensile test) is marginally favourable with Ni-Fe buffer layered joints than without

buffer-layered joints.

5. YSR and PIS suggests the conventional procedure of DMW (joint A-1) is least favourable

against the buffer layered and GMAW processes employed weld joints in study.

6. The fracture toughness (CTOD) of the GMAW weld joint is significantly more than SMAW

joints due to presence of fine slag inclusions from the coated electrode in the resulted weld.

7. Effective use of GMAW process for buttering and welding along with the buffer layer

application can be possible. While, the modification in integrity assessment procedure can be

possible by considering the interfacial region properties, chemistry mismatch and the YSR/PIS

measurement criteria.

Acknowledgement

The authors acknowledge the support given by Board of Research in Nuclear Sciences, Department

of Atomic Energy (India) for present work. In addition, author also thankfully acknowledging the

technical expertise offered by Joanna N. Walsh, School of Materials, University of Manchester for

the present work.

References[1] A.K. Bhaduri, S. Venkadesan, P. Rodriguez, P.G. Mukunda, International Journal of Pressure Vessels and Piping, 58 (1994) 251-265.[2] C.D. Lundin, Welding Research Supplement, 61 (February 1982) 58s-63s.[3] R. Miteva, N.G. Taylor, in: E. Commission (Ed.) NESC report, Institute for Energy, Netherland 2006.[4] N. Taylor, C. Faidy, P. Gilles, in, Institute for Energy, European Commission, DG-Joint Research Centre, 2006.[5] F.W.B. Brust, P. Scott, in: 19th International Conference on Structural Mechanics in Reactor Technology (SMiRT 19), SMiRT 19, Toronto, 2007.[6] F.W. Brust, P.M. Scott, Y. Yang, in: 17th International Conference on Structural Mechanics in Reactor Technology International Association for Structural Mechanics in Reactor Technology Praque, Czech Republic, 2003 pp. 1-8.[7] D.W. Rathod, S. Aravindan, P.K. Singh, S. Pandey, ISIJ International, 54 (2014) 1866-1875.[8] D.W. Rathod, S. Pandey, P.K. Singh, R. Prasad, Materials Science and Engineering: A, 639 (2015) 259-268.[9] D.W. Rathod, S. Pandey, P.K. Singh, R. Prasad, Journal of Pressure Vessel Technology, 138 (Feb 2016) 011403-011403.[10] J.N. DuPont, R.E. Mizia, in, Idaho National Laboratory, Idaho Falls, Idaho, March 2010, pp. 1-68.[11] J.A. Francis, H.K.D.H. Bhadeshia, P.J. Withers, Material Science and Technology, 23 (9) (2007) 1009-1020.[12] M. Sireesha, S.K. Albert, V. Shankar, S. Sundaresan, Journal of Nuclear Materials, 279 (2000) 65-76.

24

[13] A. Celik, A. Alsaran, Materials Characterization, 43 (1999) 311-318.[14] C. Jang, J. Lee, J. Sung Kim, T. Eun Jin, International Journal of Pressure Vessels and Piping, 85 (2008) 635-646.[15] M. Sireesha, V. Shankar, S.K. Albert, S. Sundaresan, Materials Science and Engineering A, 292 (2000) 74-82.[16] M. Sireesha, S.K. Albert, S. Sundaresan, International Journal of Pressure Vessels and Piping, 79 (2002) 819-827.[17] C.R. Das, A.K. Bhaduri, G. Srinivasan, V. Shankar, S. Mathew, Journal of Materials Processing Technology, 209 (2009) 1428-1435.[18] A.K. Bhaduri, I. Gowrisankar, V. Seetharaman, in: NWS-86, India, 1986, pp. 51-63.[19] M. Sireesha, S.K. Albert, S. Sundarasan, Metallurgical and Materials Transactions A, 36A (June 2005) 1495-1505.[20] C. Sudha, A.L.E. Terrance, S.K. Albert, M. Vijayalakshmi, Journal of Nuclear Materials, 302 (2002) 193-205.[21] Y.Y. Ying, S.R. Kae, S.R. Haur, C. Chun, Journal of Materials Science Letters, 20 (2001) 1429-1432.[22] R. Anand, C. Sudha, V.T. Paul, S. Saroja, Supplement to Welding Journal, 89 (April 2010) 65s-74s.[23] D.W. Rathod, in: Department of Mechanical Enginnering, IIT Delhi, New Delhi, Jan 2015, pp. 383.[24] J. Frankel, A. Abbate, W. Scholz, Experimental Mechanics, 33 (1993) 164-168.[25] A. Abbate, J. Frankel, W. Scholz, in, US Army Armament Research, Development and Engineering Center, Watervliet, N.Y. 12189-4050, May 1993.[26] H. Naffakh, M. Shamanian, F. Ashrafizadeh, Metallurgical and Materials Transactions A, 39A (2008) 2403-2415.[27] H. Naffakh, M. Shamanian, F. Ashrafizadeh, Journal of Materials Processing Technology, 209 (2009) 3628-3639.[28] J.W. Elmer, D.L. Olson, D.K. Matlock, The Welding Journal, 61 (September 1982) 293s-301s.[29] S.H. Ryu, J.H. Park, C.S. Lee, J.C. Lee, S.H. Ahn, S.T. Oh, Materials Transactions, 50 (2009) 1553-1557.[30] J.N. Dupont, C.S. Kusko, Welding Journal, 86 (February 2007) 51s-54s.[31] J.N. DuPont, S. Babu, S. Liu, Metallurgical and Materials Transactions A, 44A (July 2013) 3385-3410.[32] G.J. Brentrup, J.N. Dupont, Welding Journal, 92 (March 2013).[33] R. Dehmolaei, M. Shamanian, A. Kermanpur, Materials Characterization, 59 (2008) 1447-1454.[34] I. Hajiannia, M. Shamanian, M. Kasiri, Materials & Design, 50 (2013) 566-573.[35] J.N. Dupont, S.W. Banovic, A.R. Marder, The Welding journal, 82, no.6 (2003) 125s-135s.[36] G. Srinivasan, A.K. Bhaduri, in: Proceedings of 63rd Annual Assembly & International Conference of the International Institute of Welding, Istanbul, Turkey, July 2010, pp. 81-87.[37] H.T. Wang, G.Z. Wang, F.Z. Xuan, C.J. Liu, S.T. Tu, Materials Science and Engineering: A, 568 (2013) 108-117.[38] J.W. Kim, K. Lee, J.S. Kim, T.S. Byun, Journal of Nuclear Materials, 384 (2009) 212-221.[39] N.K. Sharma, D.K. Sehgal, R.K. Pandey, Engineering Letters, 19:4 (November 2011) 1-6.[40] R. Chhibber, N. Arora, S. Gupta, B. Dutta, Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 220 (2006) 1121-1133.[41] R6, Procedure R6-Revision 4. Gloucester: Nuclear Electric Ltd., 2007.[42] SINTAP, British Steel Report, Sheffield. European SINTAP Procedure., 1999.[43] M. Kocak, in: D.M. KOÇAK (Ed.) 63rd Annual Assembly & International Conference of the International Institute of Welding, Istanbul, Turkey, Lecture, 2010.[44] H.T. Wang, G.Z. Wang, F.Z. Xuan, S.T. Tu, Engineering Failure Analysis, 28 (2013) 134-148.[45] H.T. Wang, G.Z. Wang, F.Z. Xuan, S.T. Tu, Materials & Design, 44 (2013) 179-189.[46] B. Younise, A. Sedmak, M. Rakin, N. Gubeljak, B. Medjo, M. Burzić, M. Zrilić, Materials & Design, 37 (2012) 193-201.

25