Manufacturing and Processing of Advanced Materials

A Review on the Joining of Dissimilar Materials with Special Context to Laser Welding

Aditya Purohit¹, Tapas Bajpai¹, *, Pankaj Kumar Gupta¹, Arpana Parihar²

¹ Department of Mechanical Engineering, Malaviya National Institute of Technology Jaipur, Rajasthan, India

² Microfluidics & MEMS Centre, CSIR-Advanced Materials and Processes Research Institute (AMPRI), Hoshangabad Road, Bhopal 462026, India

Abstract

In recent times, there has been an increasing demand for dissimilar metal fabrication, as this weldment utilizes the specific benefits of different metals for a particular application. In this paper, the recent trends evolving in the field of dissimilar material joining, which introduces residual stresses, distortions and formation of brittle intermetallics within the structure is discussed. As these are highly undesirable, therefore various techniques were studied by the researchers, which reduce the distortions and formation of brittle intermetallics. The use of numerical techniques in this field was also studied as they provided the researchers with an insight into the process. Mostly, the joining of dissimilar material is done using friction stir welding and laser welding, but the use of friction stir welding has constraints in terms of material temperature thus, the joining of dissimilar weldment is discussed by giving a special context to laser welding technology.

Keywords: Dissimilar material joining, Laser welding, Optimization techniques.

* Corresponding author Tapas Bajpai: Department of Mechanical Engineering, Malaviya National Institute of Technology Jaipur, Rajasthan, India; E-mail: [email protected]

INTRODUCTION

Laser welding is a state of art technique which is normally performed in keyhole mode. Unlike conventional welding techniques, this welding technique enables deeper penetration in the material combined with lower heat-affected zone (HAZ) since the focal diameter of the laser beam can be adjusted according to need. Application of laser welding varies from structural applications, automotive sector to sophisticated applications such as dentures, implants and complex electrical circuits. Dissimilar material welding is one of the fields to be explored. It's a phenomenon concerning the current industrial trend because dissimilar material welding uses the specific properties of different materials for an application that ranges from dissimilar metal electronic connection of Al-Cu in an electronic vehicle to large structures where dissimilar joints are required [1]. The problem in dissimilar welding is that because of dissimilar materials, the materials have different thermal expansion coefficients, which leads to stress application on the material due to fluctuation in temperatures, and if that stress is higher than the yield stress of the material, then the material might fail [2]. Apart from this, the other reason is the formation of intermetallic phases in the fusion zone because of less solubility between different metals. These phases have hardness variation in the weldment zone, and this variation of hardness might lead to the failure of the joint [3].

It's a well-known fact that in order to get a mechanically sound joint, it is required for the joint to be stress-free to avoid distortions and crack initiation. The major contributor to it is the different expansion coefficients of the material and phase transformation of the material. There are three components of stresses, mainly quenching, phase transformation and shrinkage. Shrinkage leads to tensile stresses, i.e., the region which is last to cool, faces tensile stresses and vice versa. Another factor is phase transformation, so the region that faces phase transformation first is subjected to tensile stresses. Whichever factor dominates the most, its effect is shown on the weldment [4]. Due to high-temperature exposure on the weldment, the temperature rises, and subsequently, a thermal gradient is set up between the weldment zone and the base metals. Due to the gradient and the difference in various material properties, residual stresses are set up in the material, and subsequently, distortions are introduced. The tensile residual stresses in the welding zone lead to a compromise in the structural integrity of the joint. If stresses are above the material's permissible limit, it might lead to the failure of the material. The subsequent distortions introduced within the assemblies are also highly undesirable since they compromise on the tolerance front, which is not desirable in the manufacturing unit. Deformation introduced in the assembly can vary from longitudinal and transversal shrinkage to angular deformation. Therefore to address the above issues and mitigate these phenomena, the researchers are working in this direction. Different aspects of dissimilar material welding were examined by analysis of various fronts of this particular field. The effects of different phenomena which lead to the deterioration of the joint were analyzed, and different researchers tried to mitigate the problem in their own way. Classification on the basis of various paths adopted by the researchers [5].

OPTIMIZATION

Optimization techniques serve as an integral part of research because various optimization techniques give the researchers an idea about the variable parameters, which should be set in such a way that the optimum results are achieved. Various techniques for researchers are available to optimize the variables such as RSM (Response surface methodology), which plots a 3D curve of output variable with respect to input variables.

Researchers as presented in Table 1 used the following optimization techniques:

1. Design of experiments (DOE) or Taguchi optimization

2. ANOVA or Analysis of Variance

3. RSM or response surface methodology

4. Combination of the above techniques

Getting an idea of how various variables in a welding process interacted with each other and the final results, helps in giving a clear outlook of the process and also helps in predicting outcomes. Bhattacharya. et al. [6] tried to study the effect of various parameters such as power, frequency, and scanning speed on the weld width and HAZ of the weldments. Centrally composite design technique was used to design the set of experiments for the experiment, and RSM was used to develop a mathematical model, keeping weld width and HAZ as the output. Specimens of polycarbonate and acrylic were taken and welded using Nd-YAG. RSM results showed that with the increase of power, HAZ and weld width increased. After attaining good weld width and strong weld, it starts to decrease with the rise in power. Results also showed that weld width and HAZ do not depend on frequency. The results of ANOVA also validated the same. One way of predicting the quality of the weld is by analyzing the bead geometry, i.e., the dimension of the bead should be kept minimum, with the weldment also serving its purpose. Juang and Tarng [7] analyzed and optimized the weld bead geometry of weldment of stainless steel prepared by gas tungsten arc welding (GTAW). Researchers narrowed down the large number of experiments down to a fixed number using Taguchi DOE technique. Since the bead geometry consists of different variables such as front height, back height and back width therefore instead of optimizing a single variable using a loss function, all the variables are optimized at the same time by assigning weighted residuals to each function according to the literature. The optimization criteria taken by the authors for that particular loss function was Lower the Better.

Toughness in weldments is an essential property as far as bridge construction and shipbuilding are concerned. Anawa and Olabi [8], fabricated sheets of 316 SS and low-carbon steel. These dissimilar metals were joined using a continuous CO2 laser. Firstly using design expert, Taguchi set of experiments was designed, keeping laser power, focus diameter, and speed as variables and impact strength or toughness as the response output. The analysis of the S/N ratio with larger the better characteristic gave us a particular set of variables for which the highest toughness of the specimen was achieved. Further analysis was done using ANOVA, which predicted the significance of the model and showed that LASER power is the most influential factor keeping in mind the impact strength of the weldment. Prabhakaran and Kanan [9] tried to weld two dissimilar metal plates of AISI 1018 and AISI316, properties were examined on the basis of central composite design (CCD), and they were analyzed using response surface methodology (RSM). The response surface plot showed the combined effect of parameters on weld strength, and it is quite evident that high power, low speed, and high focal length help increase the weldment strength.

Table 1 Optimization of weld parameters with respect to input parameters.

Kalins et al. [10], in their work, tried to weld titanium sheets using Nd-YAG lasers. Using Taguchi L9 orthogonal array, we attempted to study the effect of various parameters on the tensile strength of the joint, and finally, an optimum combination of parameters was found. ANOVA results concluded that laser power was the most influential parameter on the tensile strength. Predictive equations were also developed to predict the output.

Patil and Waghmar [11] studied the effect of current, voltage and speed on tensile strength has been analyzed using Taguchi designed experiments on AISI1030 mild steel. ANOVA results concluded that welding current and welding speed are the most influencing parameters which affects the tensile strength. Optimisation of parameters to keep residual stresses to the minimum, ultimately reducing distortions and giving the desired mechanical properties. Vemanaboina et al. [12], butt welded combination of SS316 and Inconel, Taguchi L9 orthogonal array was applied keeping residual stress as the output and a particular combination for which residual stresses are minimum were found out. ANOVA was also applied, and it concluded that the root gap was the most influencing factor in deciding the residual stress.

RESIDUAL STRESSES AND DISTORTIONS

The variation introduced within the base metals and weld path in terms of temperature both laterally and transversely introduces temperature gradient within the sample. This variation of gradient combined with the different expansion coefficients introduces residual stresses within the material. If not mitigated, they seriously compromise with the overall strength. There are various techniques to measure the residual stresses, such as X-ray diffraction (XRD) and blind hole drilling method. These techniques measure the lattice and bulk distortion by nondestructive and destructive techniques, respectively. As could be seen from the previous studies that the residual stresses over the weldment can be harmful for the weldment and its mechanical properties. Therefore efforts were made by the authors to mitigate the effect. Madhvan et al. [1] attempted welding of AA 6061- AZ31B Mg in lap joint configuration. They found that as the heat input is increased the solidification rate decreases consequently reduction in magnitude of tensile stress is observed. Roshith et al. [13] , joined thick plates of SMO 254SS by using Pulsed Continuous Gas Tungsten Arc Welding (PCGTAW) with (ErNiCrMo3) as filler and autogenous CO2 laser welding with 3032 J/mm and 120 J/mm heat input, respectively. A variation in the pattern of residual stresses was observed in both the cases, as with PCGTAW due to external filler material and increased heat input tensile stresses are induced in the weldment and compressive stresses in the base material whereas in the weldment, welded by laser welding compressive stresses were induced. Tensile stresses in the PCGTAW lead to lower fatigue life, premature failure and crack initiation.

Apart from conventional preheating and post-heating, other techniques were also studied and introduced by the researchers to mitigate the residual stresses within the weldment. Mohanty et al. [14] welded a specimen of AISI 316 using continuous CO2 laser welding. A Vibratory Stress Relieving (VSR) setup, which used vibrations of the order of 70 Hz over the workpiece, tried to relieve the induced residual stresses by vibration instead of by changing the microstructure. Results showed that after VSR, the value of residual stresses in the specimen decreased to a limit, which had a positive effect on the tensile strength of the joint, thus increasing UTS of the joint. It was observed that the hardness of the weldment decreased after the exposure of VSR, which, apart from saving energy compared to the conventional methods of increasing the temperature to refine the structure, the use of VSR also ensured no intermetallic compounds were formed in the weldment. Reddy et al. [15] studied the effect of filler wire on residual stresses distribution. Marging steel as a base material was joined using different filler materials namely maraging steel filler, austentic SS and medium alloy medium carbon steel. Peculiar residual stress behavior is observed between different filler materials as austenitic filler material showed tensile stress at the center compared to maraging steel and low alloy medium carbon steel which showed compressive stress at the center. It was also found that post weld aging results in minimizing the residual stresses because of over tempering.

Effect of welding speed on the residual stresses was analyzed by Sindhu et al. [16], in which lap joints were prepared using two speeds, namely 5.1 and 4.1 m/sec. Further they were subjected to fatigue and tensile testing. They found that higher static and fatigue strengths were observed in the samples welded with 4.1 m/sec, as compared to the samples welded at 5.1 m/sec weld speed. Also, higher residual stresses were found in specimens welded at 4.1 m/sec as compared to the samples welded at 5.1 m/sec weld speed. This is due to the higher heat input available at low welding speed. The hardness profile of the WZ, BMZ and HAZ was observed. It was found to have abrupt variation in the hardness profile. The effect of residual stresses on dissimilar material welding was analyzed by various researchers, in this chain work was done by Bajpei. et al. [17], they tried to study and mitigate the effect of residual stresses on thin sheets of Al alloys welded together by GMAW. In their work they experimentally welded the sheets using different cooling conditions. It was observed that apart from water cooled model, all the models had tensile residual stresses induced inside them which would ultimately compromise with the strength of the weldment joint.

Apart from inducing residual stresses, the difference in thermal expansion coefficient and thermal gradient would lead to distortions in the joining plates. Distortions can be of various kinds, such as transverse, longitudinal and angular. Distortions seriously compromise the tolerance in the weldment structure. Various researchers carried out their studies on this. To study or analyze an output parameter, it must be related to the input parameters of the process, therefore Shichun and Jinsong [18] related the material deformation with various criteria, namely laser parameters, geometric parameters and material properties, and concluded that bend angle increased with an increase in the number of passes, power, thermal expansion coefficient and the bend angle decreased with the increase in sheet thickness, scanning speed and beam diameter. Alternate ways of reducing the distortions were also studied by the researchers, Mochizuki et al. [19] introduced the concept of riverside preheating to minimize distortions. In their work, before the MIG welding, a riverside TIG welding whose change in position is taken as a variable, and the power input (Q) is also taken as a variable, and the combined effect of all those on angular distortions was studied experimentally and numerically. Das and Biswas [5] studied the effect of parameter variation, namely thickness, power, number of scans, scanning speed and thickness. The results were analyzed, keeping output as deformation. A combination of parameters for minimum deformation was found, and ANOVA results showed us that the bending of sheets was significantly affected by the number of passes and sheet thickness. An overall idea about the distortions in welding, their kind and types were documented by O.P. Gupta [20], who studied the various kind of deformations occurring during the welding and concluded that longitudinal shrinkages showed a regular variation along the weld line and transverse shrinkages are shown to having a tendency of being less at the starting point which leads to angular deformation of the plates.

MISCELLANEOUS

Apart from the above points, researchers also investigated various other phenomena to mitigate the problems or thoroughly understand the problem. These included the effect of offsetting the laser beam on mechanical properties and microstructure. The simulation further emulates the welding process and studies the various outcomes graphically and numerically. Apart from that, there were some other variables that would be discussed in the next section.

Offsetting of Laser Beam

The offsetting of the laser beam while welding dissimilar material usually mitigates the effect of intermetallic formations. The offsetting is usually provided away from the material having a higher reflective index. Chen et al. [21] studied the effect of processing parameters on the characteristic of Cu-SS joint using laser welding and observed that the weld mode transformed from brazing to fusion as the weld offset was shifted from SS to the interface of the base metals, respectively. The effect of various parameters such as offset, oblique angle, welding speed, and power on mechanical behavior, microstructure, and appearance was studied. The optimal value of various input parameters is to get researchers to put forward the desired output parameters. According to Chen et al. [22], laser butt welded Ti-6Al-4V and Inconel 718 concluded that brittle intermetallic phases could be reduced by deviating the laser beam on the Inconel side of the weld and crack-free welds of the alloy combination could be obtained by the use of higher power and velocity. Reduction in porosity was found when high velocity was employed because less time was made available for solidification, which thus promoted uniformity in the density of the structure.

Numerical Modelling of Dissimilar Laser Welded Joints

Researchers employed various simulating packages such as SYSWELD, Abaqus and ANSYS to simulate and study, the welding process and welding outputs.The technique usually uses temperature loads as the input for the process of generation of mechanical outputs, such as residual stresses and distortion. Attar et al. [23] performed a simulation of the mathematical model developed for the welding of dissimilar materials such as 304SS and Copper. Various heat models were used to develop a heat source for the process. After observing the optical image of the weldment and comparing it with the various developed heat models, the volumetric double conical heat source model was found more suitable than the other models. It was also concluded that material with higher yield strength would store higher stress and thus have lower distortion. Bajpei et al. [24] studied the behavior of residual stresses in thin dissimilar Al sheets. Experimental results were verified using a fem-based simulation in which goaldak's volumetric heat source was used. The high temperature was observed in AA5052 compared to AA6061, because of the latter's high conductivity. Significant longitudinal stresses were observed in AA6061 compared to AA5052 since the ultimate strength of AA6061 was higher.

Monfared et al. [25] welded austenitic steel and compared the experimental analysis with the simulated results. SYSWELD was used to simulate the welding conditions, and various hypothetical conditions were proposed, which were necessary to propose the simulation and related residual stresses with the angular deformation. Firstly thermal analysis was done, which created input for further mechanical analysis, which predicted residual stresses and deformation, according to the boundary conditions employed. A thorough study on the nature of residual stresses on the top and bottom surface was studied both in the longitudinal as well as transversal direction.

Effect of Variation of Pulse

Modern-day lasers available to us are generally of two types, namely continuous and pulsed. The effect of change in pulse shape, frequency, and various parameters associated with the laser pulsation in welding of dissimilar material was analyzed by the researchers. Lerra et al. [26] studied the effects of pulse shape, pulse distance, and pulse energy on the weld seam's mechanical, thermal and electrical characteristics. Pulsed energy was varied, keeping shape constant for different pulse separation distances. Researchers took Al-Cu as their base metal. Weld seam dimensions increased until cutting was done for increased pulse energy. Decreasing pulse distance led to higher penetration. Optimal process parameters led to low weld depth. Mathivanan et al. [27], carried out their study on Cu and Al sheets, oscillation of laser source, and pulsed nature of power source on the weldment nature has been studied. When a tensile shear test was carried out on the sample without beam oscillation, the weldment broke abruptly, confirming its brittle nature. Whereas the weldment with beam oscillation showed behaviour that confirmed induced ductility. The microstructure analysis of the oscillated samples confirmed the results, which showed dimples, confirming induced ductility.

Variation of frequency of the power source and its effect on the weld geometry because of environmental effects such as air pressure was interestingly analysed by Ghosh and Sharma [28], who studied the variation of frequency and mean current on the nature of weldment. An alloy of Mg-Zn-Al was taken, increasing current and frequency increases the hardness of the weldment. The fluctuation in current may give rise to an air aspiration effect due to the change of pressure around the weld zone, which would lead to a loss in the material. Diametto et al. [29] welded sheets of Cu using a fiber laser and observed the effect of wobbling of the laser head on the weld bead parameters. It was observed that for lower power, voids and ripples were visible, which eventually got diminished at higher speeds. Increasing the rotational diameter decreases surface voids but also decreases surface penetration. When different trajectories were compared, it was concluded that weld parameters were mostly related to the frequency of rotation as higher frequency induced overlapping, which means less and less penetration. At lower speeds, due to high concentration of heat at certain points leads to maragonia effect leading to voids and spatter.

Other Techniques

Apart from the above-mentioned techniques, other techniques, such as the effect of water cooling, use of flux, variation in clamping forces, variation in temperature during testing, etc., were employed by the researchers to analyze and study their effect on output parameters. Liu et al. [30] observed thermal expansions, and cooling contractions introduce variations in the clamping forces, which affects welding strength. Different levels of pre-set clamping forces were applied on the different thicknesses of sheets, and it concluded that samples welded with pre-set welding forces had higher UTS. The higher thickness of the sheet had a lower maximum temperature. Eventually, the force fades out as the expansion from the load cell moves away. Pankaj et al. [31] experimentally investigated butt joints of AISI304 and mild steel, prepared using continuous CO2 welding. In contrast to similar joints, the dissimilar joint had lower elongation before fracture. The fracture occurred near the mild steel side due to the presence of equiaxed dimples, as observed through SEM. The grain structure was found to be coarse near the steel side because of lower heat conductivity. Sharma et al. [32], welded AHSS(Advanced High Strength Steel), namely TRIP780 and DP980. In their study, the effect of prestraining on the sample TRIP 780 and unstrained samples were also compared. It was observed that the energy absorption in the prestrained samples was comparatively lower than in the strained sample. Prestraining also converted the austenitic phase into martensite, thus inducing hardness in the sample.

Xu et al. [33] fabricated dissimilar Titanium –Al joints using pulsed Nd-YAG laser welding. Hot cracking susceptibility and shear fracture behaviour of the welded joint was examined by varying various parameters. The HCS of the weldment increases with the increase of power, but after a point, it starts to decrease as a higher amount of heat, increases the volume of the melted pool; this reduces the cooling from the pool and enables the melted metal to repair off the cracks. Wu et al. [34] studied the effect of water cooling on the weldments prepared using Ni-SS as the base metals, filler wire of ERNiCrMo-4 and concluded that adding cooling water refines the microstructure, increases load strength, increases hardness but also decreases weld depth and increases reinforcement. Zhang et al. [35] analyzed the effect of hybrid laser and arc weld welding on low carbon and austenitic steel and concluded that zones influenced by hybrid arc had wider HAZ and deeper penetration with large grain size. Formation of austenite and martensite in hybrid and laser zone, respectively, due to the addition of filler. Good tensile and ductile behaviour was observed in the hybrid zone compared to the laser zone. Antony and Rakeshnath [36], joined Cu with SS316L welded butt joints were prepared using a CO2 laser. Two levels of samples were created with different power but the same speed. Results showed that samples made with more power showed higher tensile strength; also, a proper fusion was observed in that case. Mai and Spowage [37] fabricated Steel-Kovar, using Nd-YAG laser, using a 350 W power-driven laser without filler material used to join the material. EDX analysis showed a uniform mixture in the weld zone of steel and Kovar. Comparative analysis showed that weld depth decreased with an increase in welding speed, pores decreased, and their size increased.

CONCLUSION

The paper discussed various challenges encountered while joining dissimilar metals. Optimization methods associated with the different welding process to get the optimum