Science and Technology of Welding & Joining

Hydrogen attack and the associated change in mechanical properties in a 3Cr–1Mo–0.25V steel subjected to a simulated welding thermal treatment were investigated to show the influence of welding conditions on the heat affected zone. The maximum peak temperature reached in the simulated welding process affected the microstructure, which in turn is related to the occurrence of different quantities of methane bubbles.

An effect of the crystallographic anisotropy of β-tin grains on thermal fatigue properties of Sn–1Ag– 0˙5Cu and Sn–3Ag–0˙5Cu lead free solder interconnects were discussed. From an orientation imaging microscopic observation, three types of microstructures (single crystal-like, fine grain type and large grain type) were observed in both solders. The single crystal-like microstructure disappeared and the large grain type occurred by further fatigue due to recrystallisation. Because single crystal-like microstructure had the {100} plane approximately parallel to strain concentrated areas, recrystallisation could be retarded if the slip systems of {100} or {100} operate and an amount of thermal strain decreases because these slip systems have the larger critical resolved shear stress due to an anisotropic nature of β-tin. One of the reasons Sn–3Ag–0˙5Cu had longer thermal fatigue life than Sn–1Ag–0˙5Cu can be the number of the single crystal-like or the fine grain type microstructures in Sn–3Ag–0˙5Cu were larger.

Removal of the oxide film is usually mandatory before brazing aluminium alloys. Typical fluxes contain toxic and corrosive chemicals such as chlorides and fluorides, which must be removed after brazing, for instance to minimise subsequent corrosion. Therefore, developing a brazing process that is capable of joining aluminium based alloys without the use of any flux is of interest. A new method for flux free brazing of aluminium in air, using gallium, has been developed and is capable of providing strong bonds, having strengths comparable to that of the parent material. The key feature of this new method is that the thickness of the liquid gallium at the joint interface is minimised in order to reduce grain boundary attack and subsequent embrittlement by liquid gallium. Hence in this method a very thin layer of gallium is applied to the faying surfaces of the parts being joined, and these are then immediately quenched in liquid nitrogen to stop further diffusion of the liquid gallium into the aluminium. Bonding is subsequently carried out by rapidly heating the gallium treated aluminium parts, and holding them at a temperature between 400 and 550 °C for a few minutes under a controlled pressure. Bonding can be improved by a subsequent heat treatment at elevated temperatures to enhance the diffusion of gallium into the aluminium and so reduce further its concentration at the interface and grain boundaries. Such post-bond heat treatments may not be required if the bonding time is increased for a few minutes.

The influences of spot welding on the microstructure and mechanical properties of an Al–5·5Mg–O·3Cu alloy have been investigated. Results showed that dendrites were formed with porosity and cracks in the nugget. Grain boundary melting occurred in the heat affected zone and wide grain boundaries appeared. The alloy exhibited low hardness in the nugget centre. Tensile cracks propagated at the edge of the nugget and mixed rupture with dimples and intergranular fracture occurred. Fatigue fracture initiated at the edge of the nugget and propagated perpendicularly to the tensile axis. Transgranular fracture with striations was also observed.

Strong steels are usually difficult to resistance spot weld because of the tendency to form hard phases. This applies particularly to the transformation induced plasticity (TRIP) assisted steels with relatively high carbon equivalents. A new development in this context is the δ-TRIP steel, designed to retain δ-ferrite as a stable phase at all temperatures below melting. Fully martensitic regions are therefore avoided, making it possible to weld in spite of the high carbon concentration. The authors present here the first spot welding tests on the novel alloy system.

The long range diffusion of rapidly diffusing elements in dissimilar welds, which occurs during heat treatment and in service exposure, initiates heavy microstructural changes on both sides of the weld. Various carbides and nitrides either dissolve or precipitate as a function of the distance from the fusion line. The type and amount of these particles can be predicted by a numerical method for solving problems of simultaneous precipitation and diffusion. Predicted precipitate distributions for a 2·25Cr/9Cr and a 2Cr/0·5CrMoV weld are verified using experimental data obtained with analytical and energy filtering transmission electron microscopy. The phase diagrams for the high chromium side and the low chromium side of the welds are evaluated and compared with the experimentally determined precipitate distribution.

When dissimilar welds are heat treated or exposed to in service conditions, long range diffusion of rapidly diffusing elements can be induced by strong gradients in chemical potential across the fusion zone. This diffusion is usually accompanied by precipitation or dissolution of carbides and nitrides. In the present work, six 2·25Cr/9Cr and 2Cr/0·5CrMoV welds in different heat treatment conditions have been characterised by means of optical and electron microscopy. Carbon profiles have been obtained with wavelength dispersive X-ray analysis. A general thermokinetic model for simultaneous precipitation and diffusion, which is implemented in the software MatCalc, has been used to predict the time dependent phase distributions and the carbon profiles based on multicomponent thermodynamic and kinetic data. The calculations and the experimental data show good agreement.

A significance of two factors, fine dispersoids in a solder and the anisotropic nature of β-Sn, on thermal fatigue endurance is discussed using flip chips connected by Sn–xAg–0·5Cu (x: 1, 3 and 4 mass-%) lead free solders, together with Sn and Sn–1·2Ag–0·5Cu–0·05Ni, for comparison. Both 3Ag and 4Ag showed better thermal fatigue properties than Sn and 1Ag, and a thermal fatigue life of 1·2Ag with Ni was close to that of 3Ag despite of its low silver content. Microstructures of the solders before thermal fatigue tests can be classified into a single crystal-like and a fine grain type. However, this classification, which affects the amount of thermal strain by the anisotropic nature of β-Sn, cannot accurately describe thermal fatigue lives observed. On the other hand, Vickers microhardness of the solders, which was resulted from fine dispersoids, showed good relationship with observed thermal fatigue endurance.

The weldability of 1.6 mm thick 5182 Al–Mg alloy sheet by the single- and dual-beam Nd:YAG laser welding processes has been examined. Bead-on-plate welds were made using total laser powers from 2.5 to 6 kW, dual-beam lead/lag laser beam power ratios ranging from 3:2 to 2:3 and travel speeds from 4 to 15 m min-1. The effects of focal position and shielding gas conditions on weld quality were also investigated. Whereas full penetration laser welds could be made using the 3 kW single-beam laser welder at speeds up to 15 m min-1, the underbead surface was always very rough with undercutting and numerous projections or spikes of solidified ejected metal. This 'spikey' underbead surface geometry was attributed to the effects of the high vapour pressure Mg in the alloy on the keyhole dynamics. The undesirable 'spikey' underbead geometry was unaffected by changes in focal position, shielding gas parameters or other single-beam welding process parameters. Most full penetration dual-beam laser welds exhibited either blow-through porosity at low welding speeds (4–6 m min-1) or unacceptable 'spikey' underbead surface quality at increased welding speeds up to 13.5 m min-1. Radiography revealed significant occluded porosity within borderline or partial penetration welds. This was thought to be caused by significant keyhole instability that exists under these welding conditions. A limited range of dual-beam laser process conditions was found that produced sound, pore-free laser welds with good top and underbead surface quality. Acceptable welds were produced at welding speeds of 6 to 7.5 m min-1 using total laser powers of 4.5–5 kW, but only when the lead laser beam power was greater than or equal to the lagging beam power. The improved underbead quality was attributed to the effect of the second lagging laser beam on keyhole stability, venting of the high vapour pressure Mg from the keyhole and solidification of the underbead weld metal during full penetration dual-beam laser welding.

Although correlations of welding parameters with the metallurgical features of conventional fusion welds in low carbon steels are well established, information on process–structure–property relationships associated with pulsed laser welds is more limited. This paper presents results on the characterisation of weld metal and heat affected zone (HAZ) microstructures observed in laser welded AISI 1006 steel. Pulsed Nd-YAG laser welds in the bead on plate configuration were used for this purpose, both in overlapping and non-overlapping bead configurations. As very rapid heating and cooling cycles occur during laser welding, the microstructures observed in the weld metal are the result of rapid solidification producing thin columnar austenite grains extending from the fusion boundary, which transform to martensite and bainite during fast cooling to ambient temperature. The HAZ structure in the base plate can also be rationalised in terms of the rapid thermal cycling experienced. The HAZ is narrow with the intercritical reheated subzone being dominant. As microstructural development has a critical effect on the mechanical properties of welds, microstructural characterisation plays an integral role not only in the understanding of pulsed laser welding, but also in the selection of optimum welding conditions for the material of interest.

The aim of the present work is to examine the yield strength mismatch between the weld metal and the base metal, which affects the deformation and fracture behaviour of welded joints. Specimens prepared from SAE 1020 steel plates were welded using three rutile covered electrodes having different tensile strengths. Fracture toughness of the welded zones was investigated based on the behaviour of cracks caused by fatigue. The mechanical properties of the weld metal and heat affected zone (HAZ) were examined, and graphs were obtained to compare crack progress values Δa (mm) and welding material toughness J (N mm-1), which is a fracture mechanics parameter. Microhardness investigation was carried out on the weld metal, base metal, and HAZ transitions via measurements along lines on four different zones of the specimen. Micrographs of the specimens were also taken.

The present study investigates the effect of joining parameters on the microstructural and mechanical characteristics of dissimilar friction stir spot welding (FSSW) between AA 1050 Al and 22MnB5 hot stamped boron steel. Mechanical performance has been evaluated by shear and microhardness testing. Optical microscopy has been used to investigate the microstructure generated in the different FSSW regions. A macrostructural examination has revealed the creation of mechanical interlocking in the Al steel connections. No volumetric defects or any other imperfection has been found in all FSSW connections. Shear failure load has increased with increasing both tool rotational speed and plunge depth for all FSSW connections. Higher plunge depth has improved the mechanical interlocking between lower and upper sheet due to the formation of a larger secondary flash. Encouraging results have been obtained using coated WC–Co tools in terms of durability and joint performance.

The friction stir welding (FSW) of 1050 - H24 aluminium alloy was performed to investigate the mechanical properties of the joints and determine the optimum FSW parameters. The mechanical properties of the joints were evaluated via tensile tests. The experimental results showed that a distinct softened region located at the weld and heat affected zones occurred in the joints. The degree of softening and tensile properties of the joints are significantly affected by the welding process parameters, such as welding speed and rotation speed. The optimum FSW parameters can be determined from the relations between the tensile properties and the welding parameters, and the maximum tensile strength of the joints is equivalent to 80% of that of the base material. When the welding parameters deviate from the optimum values, a crack like defect or significant softening is produced in the joints, thus the tensile properties of the joints deteriorate and the fracture locations of the joints change. All these results can be explained by the hardness distributions and welding defects in the joints.

Resistance microwelding of dissimilar materials such as Pt‐10Ir and 316 low carbon vacuum melted stainless steel is becoming increasingly important for making electrical connections in medical devices. The joining of dissimilar materials increases flexibility in design while providing economic advantages, where more cost effective materials can be substituted for traditional materials. In this work, the performance of joints made using different electrode forces was studied by examining the surface morphology, cross-sections, joint break force and dynamic resistance measurements from resistance microwelding joints. Electrode sticking and excessive expulsion were observed with low electrode forces, whereas joints with undesirable cracks and notches were produced at higher electrode forces. Based on the analysis of single pulse welds, a new process variation using multiple pulses was developed, which improved the weld surface quality while obtaining a joint strength near 90% of the Pt‐10Ir wire tensile strength.

Titanium aluminide is a potential material for high temperature structural application. Recent research on Ti–25Al–10Nb indicates that the alloy possesses excellent superplasticity at high temperature and is suitable for diffusion bonding. In this study, the bonding process parameters, such as surface roughness, bonding temperature, bonding time, bonding pressure, and post-bond heat treatment (PBHT), affecting the properties of bonded joints were investigated. The American Welding Society constrained single lap shear test was applied to assess the quality of the joints. Experimental results showed that a smooth surface roughness is beneficial for the diffusion bonding of Ti–25Al–10Nb. Shear strength of joints increased with the bonding temperature, bonding pressure, and bonding time. However, in the bonding temperature range 850–950°C, shear strengths of joints were all less than 80% of that of the base metal in the single lap shear test. After PBHT at 1050°C for 120 min, the quality of joints could be improved significantly. As rolled Ti–25Al–10Nb can also be diffusion bonded and the bond interface and micro voids eliminated during annealing through recrystallisation and grain growth. Interlayer diffusion bonding using 18 μm thick aluminium foil as an inter layer could lower the bonding temperature to 700°C, but line defects caused by solidification were still present at the centre of the inter layer. Even if the alloy joint was homogenised at 1050°C for 24 h, voids remained in the bonded region. Solid state diffusion bonding of Ti–25Al–10Nb using a TiAl3 interlayer of approximate thickness 50 μm, coated on one side of the joining alloy surfaces, was also investigated. No microvoids were observed in the joint of the solid state diffusion bonded alloy after PBHT at 1050°C for 12 h.

Grain growth phenomena during post-weld heat treatment (PWHT) were examined in friction stir (FS) welded 1100 Al. The FS welds were produced at rotational speeds between 550 and 3600 rev min−1. The FS weld produced at the lower rotational speed had the finer equiaxed grain structure containing feature of the more deformed microstructure in the as welded state. It was found that the regions having more deformed microstructure in the as welded state experienced abnormal grain growth more preferentially in the stir zone during PWHT. The stir zone exposed to higher temperature during FSW was more stable at higher PWHT temperatures. When the weld was heated for a long time at a temperature that was high enough for the abnormal grain growth, the stir zone produced at the lower rotational speed had the smaller grown grains. The present study suggested that the abnormal grain growth of the stir zone during PWHT had similar features of primary recrystallisation.

Modified X2CrNi12 stainless steel, conforming to EN1.4003 and UNS S41003 grades, has been designed with less carbon (<0.015%) and impurities to improve the weldability and mechanical properties. The present paper deals with submerged arc welding (SAW) of 30 mm thick plates of this steel with austenitic and duplex stainless steel consumables. Several samples extracted from the welded joints were subjected to mechanical testing by means of tensile, bend, Charpy impact and crack tip opening displacement (CTOD) fracture toughness tests. Microstructural examination including grain size analysis, hardness and ferrite measurements was carried out. Salt spray and blister tests as corrosion testing were applied. Considering all data obtained, it can be recommended to use austenitic filler metals as an economic alternative for SAW of this steel in the areas where impact is anticipated and adequate atmospheric corrosion resistance is needed since the weld made with austenitic wires exhibited very encouraging low temperature impact toughness properties related with finer grained microstructure and adequate strength and corrosion properties.

The present study is concerned with gas tungsten arc welding of two high strength aluminium alloys, namely, an Al-Zn-Mg alloy (RDE 40) and an Al-Li based alloy of Russian grade 1441. One of the critical requirements of these two alloys is that they should be weldable. In the present work, weldability aspects of these alloys were studied in terms of solidification cracking tendency, microstructure, tensile properties, and microhardness across the welds. These studies were extended to dissimilar welds between RDE 40 and 1441 produced via conventional gas tungsten arc (GTA) welding as well as pulsed current GTA welding. It was found that RDE 40 welds were less sensitive to solidification cracking and weld metal porosity compared with 1441 alloy. The superior weldability of RDE 40 was related to the equiaxed nature of the fusion zone and a lower sensitivity to moisture pickup. It was possible to produce RDE 40-1441 welds without defects. Pulsed current welding of RDE 40 to 1441 showed improved mechanical properties compared with conventional GTA welding, and these were related to the refinement of the fusion zone microstructure.

Hardness, tensile properties, and hot cracking susceptibility of Al–Li alloy 1441 were evaluated with respect to different filler alloys AA 2319, AA 4043, and AA 5356, as well as the parent alloys. The hardness in the as welded condition was 70–90 HV and improved by 20, 30, 40, and 40 HV after heat treatment with AA 4043, AA 5356, and AA 2319, and 1441 fillers respectively. Tensile strength showed similar trends as hardness in the as welded and heat treated conditions. The hot cracking tendency was the maximum for welds deposited with 1441 filler and the minimum for welds deposited with AA 5356 filler, both values being less than the cracking tendency for autogenous welds. Hot cracking tendency was correlated with grain size, segregation distance, and distribution of the low melting phases.

The present work investigates the effects of adding a small amount of Cu to Sn–3Ag–1·5Sb solders. The present results indicate that adding 0·5 and 1·0 wt-%Cu to Sn–3Ag–1·5Sb solders causes the liquidus temperature to decrease from its original value of 233·4°C to 231·6°C and to 231·4°C, respectively. Furthermore, it is noted that the addition of 1·0 wt-%Cu reduces the difference between the liquidus and solidus temperatures. It is shown that the added Cu reacts with the Sn content of the solder to form Cu6Sn5 particles in the β-Sn matrix, which are distributed non-uniformly since the Cu content is low. The experimental results also reveal that the growth rate of the solder joint interfacial intermetallic compound layers increases at higher levels of Cu addition. Finally, it is established that adding Cu to the Sn–3Ag–1·5Sb solder not only improves the adhesive strength of the solder joints, but also reduces the rate of degradation of the adhesive strength of the joints during thermal storage.

Studies on the weldability of 17-4PH stainless steel, in the 621 degrees C overaged condition, showed that Cr-eq/Ni-eq ratio higher than 1.5 resulted in primary ferritic mode of solidification in the weld metal. Post-weld aging treatment at 482 degrees C enhanced the strength of the weld joint with corresponding reduction in impact toughness of the weld metal while post-weld aging at 621 degrees C caused marginal reduction in strength of the weld joint with significant increase in impact toughness of the weld metal.

Post-weld heat treatment (PWHT) procedures and heat input during welding of 17–4PH stainless steel, using matching chemistry consumables, have been optimised in relation to its microstructural condition before welding based on room temperature tensile properties. The 17–4PH stainless steel was welded in two different prior microstructural conditions, namely, condition A (solution treated) and condition H1150 (overaged), using three different heat inputs of 0·27, 0·48, and 0·72 kJ mm-1, and post-weld heat treated to condition H900 (aged) or condition H1150 (over aged), using different heat treatment procedures. Room temperature tensile tests were carried out to study the combined effects of prior microstructural condition, heat input during welding, and PWHT procedures.

The hot ductility behaviour of a material exhibits a good correlation with its heat affected zone (HAZ) hot cracking sensitivity. This study was aimed at evaluating the HAZ hot cracking susceptibility of 18%Ni maraging steel and exploring the relationship between hot ductility and microstructure of the HAZ in an 18%Ni maraging steel. Optical metallography and scanning electron microscopy (SEM) investigations revealed that the ductility loss was accompanied by fracture transition from ductile transgranular mode to brittle intergranular mode. Fracture specimens were examined using SEM with an attached energy dispersive X-ray spectrometer. The SEM fractographs indicate that the ductile transgranular dimple fracture (i.e. coarse) surface was the titanium rich inclusion precipitated at the grain boundary, which promotes microvoid nucleation, growth, and coalescence and leads to localisation deformation until final fracture. Results of hot ductility tests also indicate that the titanium rich inclusion can reduce the ultimate tensile strength without reducing the tensile ductility. The appearance of brittle intergranular fracture is caused by the constitutional liquation of titanium sulphide inclusions that induces formation of a liquid film penetrating the grain boundaries. The results show that titanium sulphide inclusion not only reduces the HAZ tensile ductility, but also reduces the ultimate tensile strength.

A practical repairing technique using laser surface melting (LSM) was developed to remove the stress corrosion cracking (SCC) in overlaying of Inconel 182. Influence of microstructure of different heat treatments performed during repairing process on intergranular cracking/intergranular stress corrosion cracking (IGC/IGSCC) susceptibility was discussed. The intergranular precipitate was identified as M23 C6 by TEM. The microstructure with no intergranular precipitate and refiner sub-grain after LSM process shows excellent IGC/IGSCC resistance. The stress relief heat treatment induced severe microstructure of high IGC/IGSCC susceptibility, owing to the semicontinuous intergranular precipitation. The influence of Nb/C ratio on IGC/IGSCC susceptibility of three nickel based superalloys after LSM process was also investigated. For both of the Inconel 182 alloys with different Nb content, the microstructure after LSM process and following sensitisation treatment showed precipitation free grain boundary. The results of corrosion tests also indicated that the material with higher Nb/C ratio showed higher IGC/IGSCC resistance after LSM process and following sensitisation treatment.

This paper presents the microstructural modification in a dissimilar joint of Ti‐5Ta‐1·8Nb alloy with 304L austenitic stainless steel, fabricated using explosive cladding process. The interface had a wavy nature with occasional presence of shrinkage cavities and solidified melt zones. X‐ray Rietveld and electron microprobe based analysis did not reveal the presence of intermetallic phases at the weld interface within their detection limits. Evidences for the transformation of fcc to bct phases in 304L stainless steel and formation of metastable fcc phase in Ti‐Ta‐Nb alloy, not predicted in the phase diagram are provided. These phase transformations are understood in terms of severe plastic deformation during explosive cladding process.

Solidification structure of a weld in a manual gas tungsten arc welding of 3 mm thick Ti–5Ta–1·8Nb plates has been examined using optical/scanning electron microscopy, EDX spectroscopy and electron backscattered diffraction techniques. The weld exhibited columnar grains of prior β, with a substructure of martensitic α′. The backscattered electron image revealed the solidification mode to be cellular–dendritic, and the EDX analysis confirmed the partitioning of Ta atoms between the dendrites. The crystal orientation of the product α′ obtained by electron backscattered diffraction and the Burgers orientation relationship (for β→α′ transformation) were used to predict the orientation of parent β grains in the weld zone, and the long dendritic arms were found to be aligned parallel to β. Analytical solution based on Rosenthal equation was used to calculate the temperature cycle of the welding process. The solidification mode predicted from calculated G/R ratio (where G is solidification gradient and R is rate of solidification) and Kurz–Fischer map matched with the observed cellular–dendritic solidification structure.

Gas tungsten arc welding was performed on 18Ni (250 grade) maraging steel sheet using two different filler wires: one of the same composition as the base material and the other containing more cobalt and aluminium and less molybdenum and titanium. Weld specimens were then aged under four different sets of conditions. After metallographic characterisation, mechanical properties including hardness, tensile strength and ductility, and fracture toughness were evaluated. Results showed that use of the matching filler material led to lower strength but higher ductility than in the other case; this was attributed to the presence of reverted austenite in the former (caused by segregation, especially of molybdenum) at the fusion zone substructure boundaries. In both types of weld metal, a re-solution treatment followed by aging at 480°C resulted in optimum tensile properties. Fracture toughness of the aged weldments was in general close to that of the parent material aged at 480°C; some deterioration occurred only when welds with pronounced segregation were made at high temperature.

Seven 2.25Cr–1 Mo superheater outlet headers with service exposures ranging from as fabricated through to 190 kh (21.7 years) have been examined using analytical TEM. Parent plate, heat affected zone, and weld regions have been characterised to understand the influence of service exposure on the microstructure and the carbide make-up. Post-weld stress relief heat treatment in a virgin header produced a starting microstructure devoid of M3C, and consisting of M7C3 and M23C6 in the bainite and M2C in the ferrite. Service induced aging produced coarsening of M2C and the development of denuded zones along prior austenite grain boundaries. The carbide compositions in the bainite shifted towards thermodynamically stable phases, rich in Mo and Cr. One header exhibited anomalous behaviour; parent plate compositional effects may have contributed to this, along with fabrication and/or operational differences. Most of the individual carbide phases studied were found to undergo compositional changes with parabolic kinetics. These microstructural and compositional changes may find application in the assessment of effective time at temperature.

Although modified 9Cr–1Mo (grade 91) steel is considered weldable using conventional procedures, the attainment of optimum weld metal properties often causes concern. In the current work, plates of grade 91 were manual metal arc welded using electrodes with slightly different compositions provided by two manufacturers. Post-weld tempering was performed at 730 and 760°C for 2 and 6 h in each case. The weld fusion zones were metallographically examined in detail and hardness, tensile properties, and toughness were estimated. The results showed that the weld metal produced from one of the two electrodes was distinctly superior in terms of ductility and toughness for any given welded or heat treated condition. This was attributed to the slightly higher silicon, niobium, and chromium contents in the inferior electrode, which resulted in both retention of some high temperature ferrite and a greater degree of precipitation. In this electrode, additionally, the alloying elements were introduced through the flux coating, which produced inhomogeneities in the microstructure that degraded the mechanical properties.

Modern power plant steels are post-weld heat treated at temperatures in excess of 700°C. Some of these steels are richly alloyed with elements such as nickel or tungsten and have been reported to exhibit unexpected changes in strength as a function of heat treatment and solute concentration. It is demonstrated here that the changes are a consequence of solute induced variations in the stability of austenite in the temperature range over which heat treatments are conducted. The evidence for this comes from direct experimental measurements using dilatometry and from phase diagram calculations. Excessive alloying with nickel is found to lead to the unintentional formation of austenite during post-weld heat treatment, whereas variations in the concentrations of ferrite stabilising elements render the alloys incapable of becoming fully austenitic. This in turn leads to quite large variations in the mechanical properties of the final component.

It is essential to set up the activated tungsten inert gas (A-TIG) welding process parameters to produce the desired weld bead geometry and heat affected zone (HAZ) width in modified 9Cr–1Mo steel weld joints. Therefore, it becomes necessary to develop a tool for optimisation of A-TIG welding process. Genetic algorithm (GA) based model has been developed to determine the optimum process parameters. In this methodology, first independent ANN models correlating depth of penetration, weld bead width and HAZ width with current, voltage and torch speed respectively were developed. Then, GA code was developed in which the objective function was evaluated using the ANN models. There was good agreement between the target and actual values of bead geometry and HAZ width obtained using the GA optimised process parameters. Thus, a methodology using GA has been developed for optimising the A-TIG process parameters for modified 9Cr–1Mo steel.

Modified 9Cr‐1Mo steel plates of 6 mm thickness have been laser welded using CO2 laser. The effects of beam intensity and overall heat input (168-1500J/mm) on the bead characteristics, microstructure and mechanical properties of the welds have been investigated by varying the laser welding parameters such as laser beam mode, power and welding speed. The microhardness survey carried out on the welds after post-weld heat treatment did not reveal any soft zones in the intercritical heat affected zone for welds made with a heat input of up to 420 J mm−1. The tensile strengths of the welds were comparable to that of the base material. Charpy impact tests on subsize specimens revealed that the welds have good toughness. δ-Ferrite was observed in the fusion zone of the welds made at heat input of 700 J mm−1 and above, the content of which increased with the increased heat input.

Ultrasonic velocity measurements have been carried out across the weld line in two perpendicular sections of modified 9Cr–1Mo ferritic steel weldments, in the as welded and post-weld heat treated (PWHT) conditions. The ultrasonic velocity plot is correlated with the weld profile in both the sections and with the associated microstructural features in different regions of the weldments for the two conditions investigated. The present study reveals that the weld profile can be imaged and the adequacy of the PWHT can also be assessed in modified 9Cr–1Mo ferritic steel weldments using ultrasonic velocity measurements.

Welding of modified 9Cr–1Mo(V–Nb) steel pipes has been carried out via shielded metal arc (SMA) and tungsten inert gas (TIG) welding processes. The weld joints have been produced using different preheating temperatures, followed by post-weld heat treatment (PWHT) at various temperatures. The microstructures of the weld and of the heat affected zone (HAZ) of the weld joints have been studied under the optical microscope and correlated with the preheating and PWHT. The average hardness of the weld and different regions of the HAZ, and tensile properties of the weld joints have also been studied and correlated with the preheating and PWHT. The tensile properties of the SMA and TIG weld joints produced using preheating and PWHT at various temperatures are compared and correlated with their microstructures. It is noted that a comparatively high preheating temperature of the order of 573 K is beneficial, and PWHT is necessary to reduce the susceptibility to cold cracking of weld joints of the present steel. The PWHT at 1123 K enhances ductility to fracture, but decreases the tensile strength of the base material, causing fracture of both the SMA and TIG weld joints from this region close to the HAZ. The tensile properties of SMA welds are found to be superior to those of the TIG welds, especially for PWHT at temperatures up to 1023 K.

Chromium–molybdenum steels are extensively used in the steam generator circuits of power plants. These components may require welding of the cracks that can develop during fabrication, storage, and transportation stages, or during the service life of the plant. This investigation compares repair welding methods for Cr–Mo steels, using 2.25Cr–1Mo and 9Cr–1Mo materials. To simulate aging during service, welds were heat treated at 873 K for 5000 h. Simulated repair welding of the aged welds was carried out at the weld/base metal interface, i.e. at the location at which cracks are usually reported to occur during service. Two repair welding methods (half bead and butter bead temper bead methods) conforming to the ASME Boiler and Pressure Vessel Code were used. Tensile properties, hardness profiles, and X-ray diffraction based residual stress distributions were determined for both the Cr–Mo steel welds to evaluate the simulated repair welds. Analysis of the test results showed that both the repair welding methods can be used for 2.25Cr–1Mo steel welds, although the butter bead temper bead method is much more suitable for both the 2.25Cr–1Mo and 9Cr–1Mo steel welds.

Two experimental aluminium alloys (belonging to 2000 and 7000 series respectively) were welded using the laser beam welding (LBW) technique and an Al–Si alloy as filler. Different combinations of pre- and post-welding thermal treatments were proposed. The change of strength during aging treatments was investigated by microhardness measurements carried out on the weld and on the regions of the alloys not affected by the LBW process. The microstructure of the welded specimens was studied by scanning electron microscopy and energy dispersive X-ray spectroscopy. The strength of the joints after different thermal treatments was checked by T pull tests. Different thermal treatments resulted in different final strengths of the joint.

The present paper evaluates the solderability of three discontinuously reinforced aluminium matrix composites. The tested composites were an aluminium alloy of the 2000 series (AA2014) reinforced with different percentages of silicon carbide particles (6, 13, and 20 vol.-% respectively). A similar study was carried out on the unreinforced aluminium alloy for comparative purposes. Three low temperature filler alloys of the Zn–Al system were used for soldering. Drop formation tests were performed to evaluate the wettability of the molten filler alloys and sample joints (single overlap specimens) were produced to determine solder penetration in the joint clearance. Microstructural studies of the joints were carried out to determine the effects of the solid reinforcement particles on molten pool behaviour and solidification mechanisms.

Steady-state friction stir welding of 2014 Al alloy has been modelled using an Eulerian formulation that considers coupled viscoplastic flow and heat transfer near the tool. The model equations are solved using the finite element method to determine the velocity field and temperature distribution. The grain size is related to the strain rate and temperature of the friction stir welding process in 2014 Al alloy, and the hardness is found to be essentially related to grain size through the Hall–Petch relationship in the stir zone. Using the simulation results of strain rate and temperature, the microhardness in friction stir processing of 2014 Al alloy was predicted. It was found that the microhardness prediction distribution along the friction stir zone agrees well with experimental results.

Commercial Al-2024 (T3) alloy was friction stir welded at various heat index (HI) values using bead on plate approach. Quantitative analysis using electron microscopy and differential scanning calorimetry (DSC) revealed a complex variation in the precipitation evolution of Guinier-Preston-Bagaryatskii (GPB) zone and Al2CuMg (S phase) precipitates in nugget and heat affected zone (HAZ). Differential scanning calorimetry data also suggested formation energy of GPB zone equal to −236 J g−1. Tensile properties attained a maximum for HI values close to 3·94 in the nugget region, which was attributed to corresponding minimisation of the volume fraction of coarse S phase to the profit of GPB zone. The precipitation of fine S phase precipitates resulted in higher tensile properties in HAZ as compared to those in nugget for all HI values. Experimental data were used to determine major strengthening mechanisms by using constitutive relationships.

A finite element model to predict the evolution of stress and distortion for a bead on plate tungsten inert gas (TIG) weld in aluminium alloy 2024 plate is described. The thermal model was calibrated against thermocouple measurements, and the mechanical model was validated against direct measurements of residual strain made using synchrotron X-ray diffraction. Particular attention has been paid to the clamping arrangement, the transfer of heat between the plate and the copper backing plate, and the selection of appropriate thermal history dependent mechanical properties. The validated model has been used to determine the optimal arrangement for 'low stress no distortion' welding using laser heating applied to both sides of the joint line ahead of the TIG torch. In this manner it is predicted that the peak longitudinal tensile stresses in the weld region can be reduced to 15% of their normal values.

Local melting of eutectic films and cracking is found in Al 2024 and Al 7075 alloy friction stir spot welds. Dissolution of melted films removes all evidence melted film formation in spot welds made using typical welding parameter settings. For this reason friction stir spot welding is carried out at a rapid plunge rate of 10 mm s−1 and an extremely short dwell time of 0·05 s and after tool retraction, the welded samples are rapidly quenched using a mixture of methanol and liquid nitrogen at a temperature of −80°C. Eutectic films rich in Zn and Cu are formed in Al 7075 spot welds while melted Al2CuMg particles promote the formation of α-Al+Al2CuMg eutectic films in Al 2024 spot welds. Melted eutectic formation and cracking is also observed beneath the tip of the rotating pin during Al 7075 friction stir spot welding and is consistent with the occurrence of melt wear in this location.

The microstructural evolution and resultant mechanical properties during the friction stir welding (FSW) of precipitation strengthened aluminium alloys depend on initial temper as well as FSW process parameters. Al-2024 alloy under two different initial tempers, T3 and T8, was used in the present study. FSW bead-on-plate runs were performed at different values of process parameters (tool rotation rate and tool traverse speed). Microstructure and mechanical properties of the nugget region and heat affected zone (HAZ) were evaluated. Differential scanning calorimetry (DSC) revealed that in the nugget region, presence of Guinier–Preston–Bagaryatskii (GPB) zone results from the partial dissolution of Al2CuMg phase. The microstructure and tensile properties were found to be independent of the initial temper of the material in the nugget region. In the HAZ region, tensile properties increased at higher heat index values for T3 condition, and decreased monotonically for T8 condition.

In the development of Al–Li alloys for aerospace structures, their behaviour during welding plays an important role. One way of improving weldability is to refine weld solidification structures, which can be achieved by a variety of means. In this work, a type 2090 Al–Li alloy was gas tungsten arc welded with two different filler materials corresponding to types 2319 (Al–6.3Cu) and 4043 (Al–5.2Si). Inoculation with titanium together with arc oscillation through an imposed alternating magnetic field was used to refine the weld fusion zone microstructures. Post-weld aging and tensile testing were employed to assess possible improvements in performance. It was found that the combined treatment of inoculation and magnetic oscillation resulted in fully equiaxed, fine grained structures and that this led to a noticeable increase in aging response and tensile properties, especially ductility.

Keyhole and cover pass variable polarity plasma arc welds were made on aluminium alloy 2195 with measured contamination levels of nitrogen, oxygen, and hydrogen. Contamination levels ranged from less than 10 to 500 ppm in both the argon plasma gas and the helium shield gas. It was found that nitrogen leads to more severe porosity than either hydrogen or oxygen, and that rear shielding is required for keyhole welding of Al–Li 2195 alloy to protect the weld from nitrogen in the atmosphere. Both nitrogen and oxygen contamination produced a dark surface on the weld bead, which comprised metallic aluminium particles, nucleated in the melt, that had aggregated at the surface of the weld pool.

In the present work, 7·8 mm thick AA 2219 rolled plates were successfully joined without keyholes using semiconsumable tools by filling friction stir welding (FFSW). The shoulder further effect was performed to enhance mechanical stir, and mechanical properties were enhanced effectively. The results showed that the bonding interface of AA 2219 bit and keyhole is defect free. The softened region on the advanced side is the weakest part of FFSW joint rather than the bonding interface of the keyhole. The average ultimate tensile strength and elongation are 172 MPa and 11·2%, equal to 90 and 82% of the base weld without defects respectively. Excellent bonding interface and mechanical properties of FFSW joints have been exhibited.

Underwater friction stir welding (FSW) has been demonstrated to be an effective method to improve the mechanical properties of joints. To illuminate the characteristics of underwater FSW, the microstructural evolution and its effect on mechanical performance of an underwater joint were investigated in the present paper. The weld nugget zone (WNZ) is characterised by the homogeneity of refined grain structures in the direction of thickness. The precipitate deterioration is gradually strengthened from the heat affected zone (HAZ) to the WNZ. The dislocation movement featured by different types of grain boundaries suggests that continuous dynamic recrystallisation has occurred in the WNZ during FSW. The evolutions of strengthening precipitates and grain structures have significant influence on hardness distributions and the tensile fracture features of the underwater joint.

Aluminium alloy AA 2219 is used in aerospace applications because of its good weldability, stress corrosion resistance, and mechanical properties over a wide temperature range of + 250 to - 250 ° C. The elastic-plastic fracture toughness JIc has been measured on 7.4 mm thickness autogenous alternating current tungsten inert gas welded plates of 2219-T87 alloy in the as welded condition, using a single specimen technique. The present paper elucidates the variation of JIc across the weld joint. The toughness was evaluated at three different locations across the weld, namely, in the weld, fusion boundary, and heat affected zones, and compared with parent metal toughness. It was found that the fracture toughness across the weld was higher than the parent metal toughness. The fusion boundary had the lowest toughness among the three zones across the weld joint. The variation of toughness among the different zones is explained with reference to fractographic observations.

Dissimilar welding between the aged 25Cr–35Ni (wt-%) heat resistant steel and alloy 800 is not successful due to liquation cracking in the heat affected zone of the aged steel. Microstructures, mechanical properties and weldability of the dissimilar weld were characterised using SEM and Varestraint test. The effects of a preweld solution annealing, heat input, interpass temperature and type of filler metal were investigated. It was found that the most important step to improve weldability and to reduce cracking susceptibility was solution annealing. A suitable annealing treatment was then proposed. It was concluded that dissimilar welding of alloy 800 to the aged 25Cr–35Ni steel can be successfully carried out under the conditions of solution annealing, low heat input and low interpass temperature using Inconel 82 or 617 filler metals.

Dissimilar welding of the aged alloy 800 and the as cast 25Cr–35Ni (wt-%) heat resistant steel was investigated. Microstructures, mechanical properties and weldability of the dissimilar welds were characterised using optical microscopy, scanning electron microscopy and transition electron microscopy equipped by energy dispersive X-ray spectroscopy, and Varestraint test. Since such dissimilar welding was susceptible to crack formation in the heat affected zone of the aged part, the effects of a preweld solution annealing, heat input, interpass temperature and type of filler metal on the weldability of two alloys were investigated. It was found that during the solution treatment, the precipitates produced in the service stage were decomposed and that TiC was formed. In addition, tensile strength and hardness were reduced, but ductility and toughness increased. It was concluded that the most important step to improve weldability and to reduce cracking susceptibility was solution annealing. A suitable annealing treatment was then proposed. The best weldability was found under conditions of solution annealing, low heat input, low interpass temperature and using Inconel 82 or 617 for filler metals.

The high strain rate tensile ductilities of gas tungsten arc welds in an Ir–0.3 wt-%W alloy containing 60 wt-ppm Th (designated DOP–26) have been determined at test temperatures of 900–1200°C. Within this temperature range, the welded specimens of DOP–26 exhibited tensile ductilities of 9–15%, independent of the test temperature. These values are comparable to those of unwelded DOP–26 tensile specimens tested at temperatures below 1000°C, but significantly lower than (approximately half) those of unwelded DOP–26 tested above 1000°C. Elongation measurements at points along the gauge length of tensile tested specimens indicated that ductility was fairly uniform across the base metal and weld regions. At a tensile test temperature of 900°C, fracture occurred in the base metal with a mixed intergranular–transgranular failure mode. At 980°C and above, fracture occurred along the grain boundaries in the centreline of the weld. Scanning electron microscopy of fracture surfaces revealed the presence of numerous secondary phase particles along grain boundaries in the weld region. These particles were rich in thorium and were identified as an Ir–Th eutectic phase (melting point ∼2080°C) that formed as the weld pool cooled. These particles, and the larger grain size of the fusion zone compared with the base metal, contributed to the lower tensile ductilities of the welded specimens compared with unwelded specimens. Because high strain rate tensile ductility in this alloy is strongly dependent on grain size, the grain growth behaviour of welded specimens of the alloy was also studied. In as welded specimens, the average grain diameters (measured through the thickness of the specimens in a plane perpendicular to the welding direction) in the base metal, weld centreline, and fusion zone were ∼21, 41, and 72 µm respectively. For annealing times up to 1065 h at 1400°C and up to 100 h at 1500°C, grain sizes in the weld centreline and in the fusion zone did not change significantly. For these same anneals the base metal grain size increased gradually to 45 and 58 µm for 1400 and 1500°C annealing respectively. The base metal grain sizes were comparable to previous data from unwelded specimens of this alloy. However, excessive grain growth for an annealing time of 250 h at 1500°C was observed and as yet is unexplained.