Effects of post-weld heat treatment on the microstructural evolution and mechanical properties of dissimilar friction stir welded AA6061+SiCp/AA6061-O joint

A research on the microstructure and mechanical properties of the dissimilar joint was carried out so as to understand the effects of post-weld heat treatment (PWHT) on microstructural evolution, microhardness, tensile and flexural properties of the dissimilar friction stir welded (FSWed) joint. The results showed that there is a sufficient and rather complicated material mixing in the nugget zone (NZ). After PWHT, the grains in the joint area except for the NZ demonstrated a grain refinement with more homogeneous and equiaxed grains. Fine and clustered SiC particles in the NZ are confirmed by scanning electron microscope (SEM). In addition, the microhardness values of nugget zones both in as-welded and PWHT condition exhibited a higher Vickers compared to that of AA6061-O. Furthermore, a mean tensile strength of 183.30 MPa that demonstrates a 52.22% increase in tensile strength is also observed in the transverse tensile tests subsequent to PWHT. Considering the three-point bending test results, it is explicit that an increase of 121.96% in the bending extension is obtained as there is no significant increase in maximum bending force after the PWHT.


INTRODUCTION
Metal-matrix composites (MMCs) are quite serviceable materials especially for aircraft, automotive and marine applications, where the weight reduction is important, when their properties of high strength to weight ratio is taken into consideration.MMCs are generally utilized for providing some significant advantages over conventional monolithic metals.These main advantages comprise high specific strength, elastic modulus, fatigue resistance, thermal conductivities, electric conductivities and abrasion resistance.On the other hand, MMCs have some drawbacks such as the properties of high density, high processing temperature or costly manufacturing process (Kaw 2006, Ceshini et al. 2007a, Cavaliere et al. 2008).
Recently, particle reinforced MMCs are increasingly finding their application in several industries such as automotive, aerospace, marine, electronics and sporting goods (Pirondi et al. 2008).Considering the global market trends, it is estimated that the MMC market will reach $431.1 million by 2020 at a compound annual growth rate (CAGR) of 5.7% from 2015 through 2020 (Shah 2016).Among MMCs, those based on aluminum alloy matrix have come continuously to the attention of various engineering fields of automotive, aerospace, marine, sports and recreation (Gopalakrishnan andMurugan 2011, Dinaharan andMurugan 2012a).Aluminum matrix composites (AMCs) have excellent specific stiffness, high strength to weight ratio, better wear resistance and superior fatigue properties in comparison with the corresponding monolithic alloys.AMCs also exhibit better high temperature properties (Kalaiselvan et al. 2014, Cioffi et al. 2013).In addition, Al-Mg-Si alloys including AA6061 have better strength, high corrosion resistance, better ductility and superior welding characteristics in comparison to other high-strength aluminum alloys (Elangovan et al. 2008, Ahmad and Bakar 2011, Ilangovan et al. 2015).Therefore, these alloys have widespread applications in marine frames, storage tanks, pipelines and aircraft parts.However, dissolution of Mg2Si precipitates is an important problem in the heat affected zone that is affected only a thermal cycle (Elangovan et al. 2008).
When AMCs and monolithic alloys are practically used as the structural components, joining of these materials to provide good mechanical performance, obtaining cost-effective and promising assemblies is an important secondary process in the manufacturing of most structures.Therefore, sufficient joining of these materials are attractive in many engineering applications and can be acceptable to enhance mechanical properties by use of AMCs only at determined locations in a structure where necessary (Sharifitabar andNami 2011, Guo et al. 2012).However, joining of these materials by conventional fusion welding processes may result in unfavorable effects such as chemical reactions between reinforcement-matrix material, presence of several inclusions and undesirable phases in matrix material (Lee et al. 2006, Feng et al. 2008, Wang et al. 2013, Prater et al. 2013).Friction stir welding (FSW) has become a unique welding technique suited for joining many hard to weld metals, especially high strength aluminum alloys due to its advantages provided as a solid-state joining method.FSW is also a strong candidate for welding of Mg-, Cu-, Ti-, Al-alloy matrix composites, some steels, stainless steels, thermoplastics and different material combinations, even though the technique was initially developed for Al-alloys (Çam 2011).
In recent years, several studies have been reported regarding the application of FSW to MMCs including aluminum alloy matrix AA6061.Prado et al. (2003) studied the amount of tool wear and wear rate at various traverse speeds in the FSW of AA6061/Al2O3/10p composite.Marzoli et al. (2006) evaluated the thermal and microstructural stability of aluminum matrix composite (AA6061/Al2O3/20p) subsequent to FSW and reported that a higher heat input is necessary at the beginning of the weld in order to obtain requisite plasticity.The effect of FSW on the microstructure evolution, tensile and low-cycle fatigue behavior of AA6061/Al2O3/20p metal matrix composite was also presented by the researchers (Ceshini et al. 2007b).Root et al. (2009) investigated the crystallographic texture and particle-matrix interaction during FSW in the AA6061/Al2O3/10p composite joints.Dinaharan and Murugan (2012b) carried out FSW of AA6061/ZrB2 composites and studied the effect of FSW process on the microstructure, mechanical and wear properties of the joints.Besides all these, considerable work was also carried out on the application of friction stir processing (FSP), which is utilized for producing MMCs, so as to improve mechanical properties compared to the unreinforced alloys (Chen et al. 2015, Pantelis et al. 2016, Tabasi et al. 2016, Narimani et al. 2016).However, few papers can be found in the literature discussing the feasibility of dissimilar FSW between AMCs and monolithic alloys (Sharifitabar andNami 2011, Guo et al. 2012).To the best knowledge of the authors, there is no study related to the influence of post-weld heat treatment (PWHT) on the microstructure and mechanical properties of dissimilar FSWed AA6061/SiC/20p composite and 6061-O aluminum alloy.
Considering the above fact, the primary objective of this study is to gain a better understanding of the effects of PWHT on the microstructural evolution, tensile and three-point bending properties, and hardness behavior of the dissimilar joint between AA6061/SiCp composite and AA6061-O fabricated by FSW.In the present work, AA6061/SiC/20p composite and AA6061-O are FSWed free of volumetric defects using a conical unthreaded tool of 4140 steel having HRC (Rockwell scale) 65.Then the microstructure and mechanical behavior of the dissimilar joint were scrutinized before and following the PWHT.

EXPERIMENTAL METHODS
Dissimilar sheet materials, namely 3.4 mm thick AA6061/SiC/20p composite and 3.2 mm thick AA6061-O were utilized during FSW process in this study.The composite used was fabricated by Materion (UK) via a powder metallurgy.Physical properties of the as-received composite material are given in Table 1 (SupremEX 620XF 2016).Chemical composition of the monolithic alloy is also shown in Table 2.The plates of size 300 mm x 120 mm were friction stir butt welded using a tool rotational speed of 1250 rpm, welding speed of 125 mm/min, tilt angle of 1°, and a penetration depth of 0.4 mm into the upper surface of the workpieces.The levels of each parameter were chosen by taking into consideration the results of preliminary experimental studies (Avettand-Fѐnoël and Simar 2016, Salih et al. 2015).In the current research, the monolithic alloy was fixed in the retreating side and single-pass welding process is performed.Temperature distribution in the weld line was also measured during the process by using the non-contact infrared thermometer (Fluke 576 CF) so as to provide a sound weld.The rotating welding tool with a shoulder having a diameter of 20 mm and an unthreaded pin of 5 mm diameter was manufactured from 4140 steel and then enhanced to a hardness of 65 HRC.Before FSW, surfaces of the plates were decontaminated from possible oxide films by a steel riffler.Thereafter, the joint was fabricated along the extrusion direction.General view of the welding process and dimensions of the fabricated tool are shown in Fig. 1.
The tensile properties of all specimens (as-welded and post-weld heat treated) were investigated in terms of tensile strength, percentage of elongation and also joint performance.Tensile tests were performed at room temperature at a constant speed of 1 mm/min using a Shimadzu AG-X (100 kN) Universal Testing Machine.Tensile test specimens were extracted from normal direction of the weld line according to ASTM: E8 and the test coupons were selected from different positions of the welding seam.At least three tensile tests were carried out for as-welded and post-weld heat treated (PWHTed) specimens so as to minimize the fudge factor.In addition, the Vickers microhardness profiles were also obtained across the mid-thickness of welds so as to evaluate the hardening effect of both FSW and PWHT processes.The measurements were performed with a loading of 100 g and a span of 0.5 mm.Moreover, three-point bending tests comprising only root test of the specimens were carried out according to the standard (EN ISO 5173) using same tensile testing machine with a crosshead speed of 2 mm/min and a span of 50 mm.Rectangular specimens having dimensions of 200 mm x 18 mm were tested before and after PWHT.Similarly, two bending tests are performed from each type.
The conventional metallographic methods (mounting, sanding, polishing, and etching) were utilized for the preparation of metallographic specimens.The characterization of the specimens before and after PWHT was performed through the Nikon Epiphot 200 inverted metallurgical microscope and the scanning electron microscope (SEM) of JEOL JSM-6060.The as-welded specimen was etched with Keller's reagent for 20 s and the specimen in the PWHTed condition was etched with HF+HNO3 solution for 120 s in an attempt to obtain the macrostructures and also to evaluate the microstructures comprising of stir zone (SZ), thermomechanical affected zone (TMAZ), heat affected zone (HAZ) and the parent materials.In this work, a PWHT was carried out by way of solution heat treatment (530°C for 1 hour), followed by water quenching (rapid cooling in a water of 42°C) and finally artificial aging (178°C for 8 hour).Diagram in Fig. 2 describes the PWHT process.Besides that, Fig. 3 also demonstrates the dimensions of the tensile test specimens and the sampling schematic for joint characterization.Samples in the PWHTed conditions are coded as PWHT specimens in the other subsections.It is worth pointing out that the diversified thermomechanical histories of the joint create different zones including NZ, TMAZ and HAZ.A sufficient material mixing in the NZ is also achieved.In addition, the onion-ring structure can also be seen in Fig. 4a.On the other hand, an explicit interface between NZ and TMAZ is not observed with regards to the transition regions in both macrographs.5 reveals the effects of PWHT on the microstructure evolution of dissimilar friction stir welded joint in terms of the monolithic alloy side.Grain boundaries of the composite and the SiC particles cannot be observed through the instrument of optical microscope (OM) as the dimensions of aluminum powders in AA6061/SiC/20p (AMC620XF) are quite small.In addition, the average size of SiC particles used in the powder metallurgy process is about submicron.Therefore, only the particle distributions on MMC side are determined through the SEM before and subsequent to PWHT.This is because of the fact that the determination of grain size in particle reinforced composites is quite difficult due to the heavily dispersed SiC particles (Bahrami et al. 2014).
It was observed from Fig. 5a and Fig. 5e that the elongated dendritic microstructure of parent material disappeared after PWHT.In addition, grain refinement with more homogenous and equiaxed grains is realized.The average grain size of the sample after PWHT is determined to be about 30 µm.When the joint is subjected to PWHT, the elongated fibrous microstructure in HAZ is coarsened and the aspect ratio of the structure decreases.The TMAZ of as-welded joint also demonstrates significant change, deformed and elongated grains are observed in this region.The orientation of the grains in TMAZ owing to the material flow in consequence of the stirring effect of the tool can be seen in Fig. 5c and Fig. 5g.After PWHT, more equiaxed grains are reported.However, the grains having a higher average grain size (~35 µm) than those within TMAZ in as-welded specimen are also observed, locally (Fig. 5b).In the NZ, dynamically recrystallized and finer scale grains are formed due to the frictional and deformational heating.It should be noted that there is no significant difference in the grain size of these fine-equiaxed grains subsequent to PWHT.Besides all these, more precipitates can be seen explicitly in the joint area after PWHT.
By the way, it should be noted once again that the friction and plastic deformation play a significant role on the generated heat during the process.Besides, it is widely accepted that the microstructure and mechanical features of joint are considerably influenced by the generated heat (Serindag and Kiral 2017).It has been reported that the tool rotational speed has more influence on the frictional energy by comparison with other major FSW parameters such as the axial plunge force and welding speed (Aziz et al. 2016).Considering the heat generation rate based on the friction within the FSW process, some parameters such as the friction coefficient µ, probe radius r and angular velocity  come into prominence.Regarding the temperature distribution in the weld zone, "V" shape temperature contour is generally observed (Chen and Kovacevic 2003).In other respects, the tool pin also makes a contribution to the generated heat through the friction on both threaded and vertical surfaces (Kovacevic et al. 2016).Herein, in order to assess the heat generation mechanisms in the modeling of FSW process, contact conditions such as the sticking, the sliding or partial sliding/sticking conditions can be thought as the probable contact states (Schmidt et al. 2004).6 and Fig. 7, respectively.From Fig. 6, it is confirmed that the SiC particles are distributed in the aluminum matrix uniformly.In other words, there is no evidence for agglomeration of SiC reinforcements in the base composite matrix.Considering the element distributions in the NZ (Fig. 7), particle rich regions are seen locally though the SiC particles are distributed all over the matrix.It is also noteworthy to say that the NZ only comprises Effects of post-weld heat treatment on the microstructural evolution and mechanical properties of dissimilar friction stir welded AA6061+SiCp/AA6061-O joint Latin American Journal of Solids and Structures, 2018, 15(8), e49 7/17 the chemical composition of the workpieces.This evidences that there is no unfavorable new elements in the NZ owing to the tool wear especially.SEM micrographs of the particle dispersion in the joint area of the FSWed specimens before and after PWHT are shown in Fig. 8.According to this figure, it is observed that the size of coarse SiC particles in the BM shows an alteration from 0.7 µm to 2 µm.With the FSW process, shape and size of the particles in the joint area demonstrates a clear difference except for the HAZ.Fined and oriented SiC particles can be seen in the TMAZ.In the NZ, it is obvious that there are fractured SiC particles resulting from the abrasion through the welding tool and due to the attrition of SiC particles with each other (Sahraeinejad et al. 2015).In addition, FSW resulted in the clustering of SiC particles in the NZ.Moreover, the final shape of the SiC particles in the NZ are observed as more spherical compared to those in other joint regions.Fig. 8d shows the NZ of as-welded joint with the particle size of the base composite being about 0.5 µm.In addition, particle poor and particle rich regions can be seen in the figure.It should be noted that this is in tight agreement with the EDX mapping of NZ, as well.Fig. 8eh exhibits the SEM micrographs of FSWed specimen after PWHT.With the PWHT, it is interesting to see that there is no remarkable difference in terms of the dispersion of SiC particles.Similar behavior is reported in the literature with regards to the distribution of the particles and also the matrix grain size (El-Sabbagh et al. 2013, Knowles et al. 2014).However, particle poor region in the NZ (Fig. 8h), which clearly demonstrates the particle-matrix interface, is reported in the specimen even as the number of particles has increased by comparison with the base composite.

Microhardness
The Vickers microhardness profiles are obtained across the mid-thickness of FSWed joints so as to evaluate the hardening effect of both FSW and PWHT processes.Hardness distributions of the joints are illustrated in Fig. 9.It is evident from the figure that there is an intricate mixing of workpieces.The complicated material patterns are thus viewed in the weld zone.Furthermore, the high energy input during FSW based upon the high tool penetration depth led to a wide joint area.As seen from the figure, the average hardness values of base materials for the monolithic alloy and the MMC are in the range of 40-44 HV and 144-158 HV, respectively.With the PWHT, an obvious increase in the microhardness at the base materials of both workpieces was observed.The maximum hardness values after PWHT on the AA6061/SiC/20p and AA6161-O are about 203 HV and 52 HV, respectively.In both conditions, the Vickers of stir zones are observed to be higher than that of the weaker material (AA6061-O).This obtained increment in hardness is associated with the microstructure evolution of the material that is in annealed condition.The decrement in the grain size and homogenous dispersion of the grains and precipitates due to the high temperature and high plastic flow results in the given recovery.However, the same trends in the NZ are not observed when the hardness of AA6061/SiCp composite is considered.The main reason could probably be related to the change of material grain size and dispersion of precipitates locally due to the blend of the weaker material.Besides that, overaging and solution of metastable phases in the NZ may have an unfavorable effect in the hardness, as referred in the literature (Sharifitabar andNami 2011, Kumar et al. 2015).When the figures are analyzed, it can be seen that there is an upward tendency in the microhardness from BM to NZ at the retreating side (RS) of aswelded specimen.In the HAZ of both sides, an explicit hardness increase can be seen compared to the base materials.The maximum hardness in TMAZ at RS is determined to be about 71.8 HV, while the hardness value in TMAZ at the advancing side (AS) is in the level of both NZ and the BM.In other words, there is no big difference between the hardness of NZ and BM at the composite side.After PWHT, it should be noticed that the hardness in HAZ at the RS is slightly lower than the hardness in HAZ at the side of AA6061-O in as-welded condition.At the same time, the hardness distribution within HAZ at the AS appears to be more disordered compared to that of RS.On the other hand, more homogenous hardness profile, which is in the level of BM, is obtained in TMAZ at the RS subsequent to PWHT.Moreover, a slight hardness loss was observed in TMAZ at the AS by comparison with the hardness of AA6061/SiCp composite.These hardness measurements clearly show that the PWHT can be used to enhance the hardness behavior of the dissimilar joint, though the process resulted in a complicated hardness distribution comprising sudden variances at the MMC side owing to the irregular dispersion of sub-micron reinforcing particles (Nikoo et al. 2016).Furthermore, it can be commented that the microhardness measurements are in good agreement with the microstructural characteristics obtained by OM.  and Structures, 2018, 15(8), e49 11/17

Transverse tensile properties
The tensile strength and elongation (%) of unwelded AA6061/SiCp composite are obtained as 309.64 MPa and 5.62, respectively.The mechanical properties of the monolithic alloy in terms of the tensile strength and the elongation (Pct) can be found in the Ref (İpekoğlu and Çam 2014).The engineering stress-strain curves of the as-welded tensile test specimens can be seen in Fig. 10.From the curves, it is observed that all specimens demonstrated a similar trend to those of AA6061-O parent material.The transverse tensile test results of as-welded specimens are also added to Table 3 for comparison.A higher mean tensile strength value, i.e., 120.74 MPa, is reported during the tests.The tensile test results also exhibited a joint efficiency of 103.64%, 103.29% and 103.17% in the FSWed specimens, respectively.These results mean that the appropriate selection of welding parameters and consequently obtaining a defect free weld resulted in a high joint-efficiency for the dissimilar joint between AA6061/SiC/20p composite and AA6061-O.In this study, joint efficiency is determined by dividing the ultimate tensile strength (UTS) of the dissimilar joint to that of the weaker material.In addition, all the test specimens fractured on the retreating side (AA6061-O) of the weld in the base material, as expected.This is in good agreement with the hardness measurements that demonstrates a significant hardness increase from BM to NZ.Failure positions of all specimens, which can be seen in Fig. 11, are also located away from the weld region.Thus, sound welding of the dissimilar workpieces is also proved.However, maximum failure strain of the specimens is about 9.25%, which represents 32.46% joint efficiency in terms of failure displacement.According to the results, it is worthwhile to mention that the successful FSW process can provide good performance in tensile strength, while reduces the elongation to failure in dissimilar joints between the AMC and the monolithic alloy.The comparison of transverse tensile testing results including tensile strength and percentage of elongation of as-welded and PWHTed specimens are shown in Fig. 12.An average of three specimens is utilized for the calculation of tensile strength and elongation (%) values of both as-welded and PWHTed joints.It is clear from the figure that the PWHT effectively improves the tensile strength of the dissimilar joint while reduces the failure elongation.The tensile strength and elongation of PWHT joint are in the order of 183.30MPa and 4.37%, respectively.It is evident from these results that there is a 52.22% increase in tensile strength after PWHT.This result clearly demonstrates that the microstructural modification owing to the dispersion of finer grains and precipitates in 6061 aluminum alloy uniformly improves the failure response of the dissimilar FSWed joint.On the other hand, PWHT also resulted in a decrease of 47.09% in the percentage of elongation.Considering the fact that the PWHT brings about a localized plasticity, this significant decrease can also be related to O-temper condition of the monolithic alloy.
In all specimens in PWHTed condition, fracture took place in the HAZ of welds at the RS.This is due to the slight microhardness loss close to the boundary between HAZ and the BM at the RS.Fracture positions of the PWHT specimens and SEM micrograph showing fracture surface of the tensile test specimen is demonstrated in Fig. 13.According to the SEM fractograph, the indications of ductile fracture mode including multitude dimples of various shapes are discovered in the fracture surface.The surface can also be characterized by tear ridges and shallow dimples that demonstrate inadequate toughness (Guo et al. 2017).From the figure, it can be inferred that the PWHT resulted in a slight increase in the maximum bending force.Maximum bending forces of the specimens after PWHT are in the order of 962.27 N and 1038.20 N, In addition, it is explicit from the results that the extension in three-point bending test shows a significant improvement subsequent to PWHT.An increase of 121.96% in bending extension is obtained in the PWHT specimens.Maximum bending extension is reported as 30.83 mm.Bending modulus of the specimens are also calculated during the tests.According to the results, modulus of 41.76 GPa and 81.76 GPa are obtained for as-welded and PWHT specimens, respectively.Thus, a substantial increase of 95.79% in bending modulus is provided after PWHT of the dissimilar FSWed joint.The bending angles before and after PWHT are shown in Fig. 15.It is evident from the figure that the bending angle increases subsequent to PWHT.This result means that a higher bending resistance is provided thanks to the PWHT.In addition, an L-shape is observed in as-welded specimens while the PWHT specimens take a U-shape during the three-point bending tests.As shown in the figure, there is another significant difference on the location of bending corners between different welds.Although bending in as-welded specimens takes place close to HAZ at the side of weaker monolithic alloy (AA6061-O), the location of bending corner is in the weld region for PWHT specimens.This can be associated with the improved mechanical properties of 6061 aluminum alloy after PWHT.

CONCLUSIONS
In this paper, microstructure and mechanical behavior in terms of tensile, flexural and hardness properties of the dissimilar friction stir welded joint between AA6061/SiCp composite and AA6061-O were investigated experimentally.The effects of PWHT on these characteristics also studied and discussed.Regarding the results, the following conclusions can be drawn: a) After PWHT, few explicit differences are observed with regards to the fine-equiaxed recrystallized grains in the NZ.b) Finer SiC particles are reported in the NZ with FSW.However, there is no big difference in the NZ with regards to the particle dispersion subsequent to PWHT.c) The hardness values of stir zone in both as-welded and PWHTed condition are increased compared to that of the weaker material.The maximum hardness is found as 203 HV at the AS subsequent to PWHT.d) FSW results in a maximum joint efficiency of 103.64% in terms of the transverse tensile strength.However, the tensile test results demonstrate a 32.46% maximum joint efficiency in terms of failure displacement, though the process exhibits a 64.59% increment compared to the elongation properties of AA6061/SiCp composite.e) A mean transverse tensile strength of 183.30MPa that demonstrates a 52.22% increase in the strength is also observed with the PWHT.
On the other hand, ductility of the dissimilar joint declines with the applied PWHT.f) In as-welded condition, fracture takes place away from the weld region at the RS of all specimens while tensile failure occurs within HAZ at the RS subsequent to PWHT.g) Even though there is no explicit increase in the maximum bending force, PWHT improves the bending extension and the bending modulus in the order of 121.96% and 95.79%, respectively.
As a result, there is an explicit need for more efforts to understand the effects of various welding parameters such as the tilt angle, tool penetration depth, axial force, material position and so forth on the microstructural evolution and mechanical properties of dissimilar joints, even though this research presents a well-rounded experimental investigation on the dissimilar butt joint between the AMC and the monolithic alloy fabricated by using FSW.Besides, the applications of FSW of dissimilar AMCs are still limited and extensive work is also needed to evaluate the feasibility of FSW between AMCs and other promising materials such as titanium and magnesium alloys in many engineering applications.

Figure 1 :
Figure 1: General view of (a) FSW process (b) fabricated tool dimensions

Figure 9 :
Figure 9: Microhardness distributions of the joints in (a) as-welded and (b) PWHT condition

Figure 10 :
Figure 10: Tensile behavior of the as-welded specimens

Fig. 14
Fig.14illustrates the comparison of bending force-extension behavior of the specimens before and after PWHT.From the figure, it can be inferred that the PWHT resulted in a slight increase in the maximum bending force.Maximum bending forces of the specimens after PWHT are in the order of 962.27 N and 1038.20 N, In addition, it is explicit from the results that the extension in three-point bending test shows a significant improvement subsequent to PWHT.An increase of 121.96% in bending extension is obtained in the PWHT specimens.Maximum bending extension is reported as 30.83 mm.Bending modulus of the specimens are also calculated during the tests.According to the results, modulus of 41.76 GPa and 81.76 GPa are obtained for as-welded and PWHT specimens, respectively.Thus, a substantial increase of 95.79% in bending modulus is provided after PWHT of the dissimilar FSWed joint.

Figure 14 :
Figure 14: Flexural behavior comparison of the welds before and after PWHT

Table 3 :
Transverse tensile test results of as-welded specimens