A Contemporary Review on Friction Stir Welding of Circular Pipe Joints and the Influence of Fixtures on This Process

,


Introduction
Metal joining processes are essential in most manufacturing industries since they can produce complex shapes with diferent material properties. Welding is an excellent permanent metal joining technique due to its superior joint properties compared to fastening and semijoining processes. Among the various advanced welding operations, friction stir welding (FSW) is an emerging and promising technique, that originated from the welding institute (TWI) UK in the year 1991 [1]. Te basic principle of FSW is entirely diferent from other conventional welding processes. In the conventional welding process, the base material undergoes melting followed by solidifcation [2], whereas FSW produces metal joints by the solid-state deformation principle [3,4]. Moreover, the metals Compared to other fusion welding processes, the FSW requires minimal power requirements, and preprocessing and postprocessing eforts [5]. Te weld quality aspects such as joint strength, fnishing, and metallurgical characteristics produced by FSW are far better than those of other fusion welding processes [6]. Due to the elimination of consumable electrodes and shielding gases, FSW is also highly advantageous in terms of economic and environmental considerations [7].
FSW uses a nonconsumable tool that does not involve conventional electric arc generation and avoids harmful radiation [8]. Moreover, the unique principle of FSW operation enables it to join materials with diferent physical properties [9]. Dissimilar metal joints are highly benefcial to the automotive, aerospace, and marine industries. A few examples signifying the FSW applications in industrial structures are shown in Figures 1 and 2. Friction stir processing (FSP) is a derivative of the FSW technique, bearing the diference that the former is used to modify the surface properties of the metals [13,14]. Tough FSW proves advantageous, the process itself has inherent issues with the backing of weld zone, thinning of welds, and keyholes. Tese issues are also addressed by employing proper control strategies, these issues can be averted [15]. While FSW was invented to join metallic materials and recently its applicability has been extended to join various kinds of thermoplastics as given in the reference [16,17]. In addition, research works have also explored the feasibility of making hybrid joints using dissimilar materials, such as metals [18,19] with thermoplastics by FSW [20][21][22][23]. Notable works are carried out in the joining of metals with polymer matrix composites as well [24]. Previous researchers have proposed various auxiliary assistances in the FSW process to match the  [10],and (d) space applications [11]. joint quality with industrial standards. A few examples are liquid nitrogen-assisted FSW [25], stationary shoulderassisted FSW [26,27], and ultrasonic assistance in FSW [28][29][30]. Also, FSW in recent times has taken new dimensions in the frontiers of extrusion, additive manufacturing, and grain deformation processes. FSW is employed in the extrusion process, called as friction stir extrusion has resulted in high-strength and ductile magnesium alloys [31]. FSW in additive manufacturing areas have proved to be one of the repairing tools which exhibited good performance in the repair aluminium alloy structures [32]. Similar to FSP, FSW has also been employed in the fabrication of ultra-refned grains of composites involving aluminium matrices [33]. Tese distinctive derivatives of FSW have been used in more extended applications, such as electronics, marine, railways, and aerospace [34]. NASA has fabricated a fuel tank with the help of the FSW process for a Space Shuttle [35]. Te "iMac" panels of Apple have been welded using FSW [36]. Te subframe at the front side of the "HONDA Accord" car consists of aluminium (Al), and steel material welded using FSW [37]. Al panels of rail commuters by Hitachi are fabricated with the help of FSW [38]. Te desks of the cruise ship "Te World" from Fosen Mek have been joined by FSW [39]. Based on this evidence, it may be envisaged that the FSW can be implemented in most of the industrial sectors in the near future. FSW is a "natural" technology because it does not create any hazardous materials that could cause environmental or human damage. Using FSW, various workpiece joints such as lap, butt, T, and fllet joints can be obtained. More importantly, the resulting surfaces do not require active cleaning, as FSW is a clean process compared to other conventional welding processes. Te present study reviews the diferent geometrical shapes welded by FSW and their criticality in the process. Te following sections discuss the principle of the FSW process, process parameters, and various geometries of FSW welded joints. Finally, this article is concluded by adding future scopes. Te geometrical nature of the majority of the surfaces joined by FSW is fat rather than curved or zigzag. Te majority of research works were primarily focused on getting fat joints by the FSW process followed by analyzing weld geometry, mechanical characteristics, metallurgical characteristics, and tribological behaviors. But the research works concerning the applicability of FSW to complex and circular shaped joints are very few. Tis creates the necessity to critically review the current status of the FSW techniques, primarily concerning to the nonfat surfaces such as curved, italic, or zigzag and the fxtures associated with the process. Tus, the FSW process has a promising future towards the engineering of high-strength joints with improved performances.

Working Principle of FSW
Te FSW uses the plastic deformation property of the materials for joining. Te process is comparable to the forming process where the materials are deformed to the required shape by applying pressure and temperature utilizing the plasticity property of the base material. Usually, a vertical milling machine with a rotating spindle is used to mount and rotate the tool at the desired speed. Te spindle is stationary, and the bed on which the workpiece is held gives feed motion [10]. Te tool used for FSW is diferent in construction from other tools used in manufacturing. It has a pin of the required length and a shoulder, which attributes to the deformation of the work material during their interaction. When the rotating tool interacts with the work material, heat is generated due to friction. Te frictional heat softens the work material; thus, it becomes more ductile near the interaction area [40]. Tere are four stages by which the work-tool interaction progresses and they are plunge-in, dwell, traverse, or welding, and retract [4]. Te surfaces of workpieces that need to be joined are held together rigidly using fxtures to form a line of the weld. During the frst stage of operation, the rotating tool plunges into the workpiece material at the joining line until the shoulder comes in contact with the work. At the dwell stage, the material fow due to plastic deformation is ensured at the same position.
Further, the traverse stage is initiated, in which feed motion is given to the bed where the work is mounted. Due to tool rotation and feed movement, one side of the substrate will move along the direction of the tool rotation, and the other side forces in the opposite direction with respect to the tool rotation. Te former is known as the advancing side (AS) and the latter is known as the retreating side (RS). Te material transfer happens from AS to RS. Te tool deforms the material and extrudes some from the AS, and is forged back to the RS. Tis process continuously happens till the end of tool travel. Te material fow occurred in the ring form and was called onion rings [41]. Te tool is retracted out at the end of the traverse motion, hence, producing a permanent metal joining process [42] (Figure 3).

FSW Parameters.
Te quality of the weld is majorly infuenced by the parameters associated to the weld, tool profle, materials, and geometry of the weld. Te following subsection details the efects of each parameter.

Weld Parameters.
Te parameters of FSW attributes to the joining quality are tool rotation speed, weld/traverse speed, angle of tool tilt, and axial load ( Figure 4). Te tool rotational speed (spindle speed) and the welding/traverse speed are the main parameters responsible for the generation of heat and achieving weld quality in terms of the strength of the weld joint. Te tilt angle helps to ease the plunge; hence, the deformation and transportation of material are improved. Te depth of the plunge will be the point at which the shoulder of the tool touches the surface of the workpiece, which ensures the heat generation and transfer of torque. Te tool pin length is calculated based on the thickness of the workpiece. Tus, complete contact between the shoulder and work ensures the initiation of the weld.
(1) Tool Rotational Speed. In FSW, a cylindrical tool has to rotate to produce heat and move to carry the heat along the joint. Due to this, the welding rate of the FSW process is Advances in Materials Science and Engineering relatively slow compared to other welding processes. Usually, the tool rotates in a speed range that varies from 300 to 3000 rotations per minute (rpm) [43][44][45].
(2) Weld/Traverse Speed. Tis controls the duration of the welding process. Tis is also one of the major infuencing factors for deciding the quality of the weld. Te tool's traverse speed or weld speed varies between 5 and 400 millimeters per minute (mm/min) along the joint length [46][47][48].
(3) Tool Tilt Angle. Tilting the tool to some angle from the normal to the substrate is regarded as a tool tilting angle, which is claimed to afect the material fow and heat generation by many researchers. Te resulting weld joints are approximately 2 to 4 degrees slightly lean or tilt position [49,50] from normal to the surface. Te tilt is given to prevent the efect of downward forces from damaging the joint. Tis parameter fnds its signifcance during dissimilar welding.
(4) Axial Load. Te axial load has a strong infuence on the frictional force generated. Te load varies from material to material based on their physical properties [51]. For instance, steel may require a large axial load, whereas soft metals like Al may need less. Te control automation of such axial loads is being attempted by researchers for workspecifc applications [52].

Tool Design Parameters.
Te successfulness of a weld in FSW depends on the elimination of defects in FSW [53]. Te tool design parameters play a crucial part in achieving a defect-free weld [54]. Te FSW tool ( Figure 4) parts identifed as parameters are: (i) Pin: Tis is responsible for the stirring action during the welding process by excavating the molten material from the advancing side to the retreating side. (ii) Shoulder: Tis part of the tool is responsible to exert the load over the stir zone and ensuring the creation of a friction force between the tool and workpiece. Tus, has more control over heat and torque generation.
Diferent profles of tool pins have been developed and experimentally investigated through various research works, as shown in Figure 5. Te most promising tool profles in FSW are cylindrical, threaded, conical, square, etc. [57][58][59]. Compared to conical and cylindrical profles, square pin produced an enhanced fow of material and ultimate tensile strength (UTS) at the weld joint [60]. Due to the presence of sharp edges in square pins, the pulses generated per second by sharp corners are higher than in other profles, leading to a higher ratio between dynamic and static volume [61]. Spiral-shaped pins are considered after square pins due to   the presence of ridges. Te ridges create more downward movement of material during the tool work interface than the other tools [62]. Apart from all the conventional profles of tools, a special design called "twin stir," where two tools rotate in the opposite direction simultaneously. Tis design aims to improve the symmetrical nature of the material fow, minimize the clamping force, and reduce the rework (postweld works like removal of fashes, etc.) and defects of the weld [63]. Similarly, the efects of diferent shapes of shoulders have also been reported by previous works, such as convex, concave, etc. [64,65]. Over the years, other confgurations of tools have also been studied by various researchers, such as the two shoulders, bobbin tool, and selfretracted tool to conquer the formation of keyholes [66][67][68][69][70]. Some of the experimental works conducted on diferent materials to fnd better parameter windows for successful FSW joints are shown in Table 1. Te various parameter levels and resulting mechanical properties, such as tensile strength and hardness values are also presented.
(1) Tool Material. Tool material is selected based on the strength of the workpiece material. Te tool material is required to be intact during the temperature rise and the frictional force generation that happens at the time of the welding process. Teoretically, the welding temperature will be below the melting point of the workpiece. Te common tool materials used are tool steel, high speed steel, nickel alloys like Inconel, and polycrystalline cubic boron nitride (PCBN). At times, FSW tools are also coated with ceramic materials to increase their performance based on application requirements.

Materials and Joint Parameters.
Te weld quality and process efciency also depend on the type of material used. Te work material and the bed on which the material is mounted should possess low thermal conductivity to reduce the loss of energy in the form of heat. Unlike the other operating parameters, material-related parameters can be managed ofine. In most cases, the profle of the tool is selected according to the type of material and thickness used.

Various Weld Geometries.
Te geometry of the parts to be welded plays a major role in defning the welding setup that is to be employed. Te welding principle being the same, the fxture setups are designed to facilitate the particular weld geometry. Te most common type of weld joint made is a butt joint, which is a staring seam resulting from fat plate geometry [10,[83][84][85]. Apart from butt joints, lap joints [86][87][88] and T-joints [89,90] are also employed based on industrial needs. All these types of joints require only slight modifcations in the fxture setup to facilitate the welding process. Te examples of fxture setup used for producing FSW T-joints and circular joints are shown in Figures 6(a) and 6(b). But the sizeable untapped area in the FSW joining technology is its industrial employment to the joining of circular geometries i.e. pipes. Te past decade is fnding many research attempts in employing FSW to join pipes (detailed discussion in Section 3). Te most challenging factor in facilitating FSW of pipes is the fxture design (detailed discussion in Section 4). Te various other factors infuencing the FSW of pipes and their fxture attributes must be examined carefully before taking FSW for industrial pipe joining applications.
Hence, the FSW process parameters like tool rotational speed, weld speed, and tool geometries have a major infuence over the weldment performance. Te accomplishment of various geometries welding will have the aid of fxtures for its successful facilitation.

Circular Geometries
Even though the usage of pipe dates back ages, the pipe welding process got intensifed and went commercial only after the exploration of oils, as the pipelines were crucial components in the transportation systems. Quality pipe weld joints were the call of the day as the pipeline industry     [43] 6 Advances in Materials Science and Engineering started booming. Te commercialization of pipe welding is rooted in the fling of many patents. Te frst patent on the procedure involved in pipe welding was published by the Heating and Piping Contractors National Association in 1931. Subsequently, several patents were fled on the methods and apparatuses required for the welding of pipes [91][92][93][94][95]. Te Welding Institute (TWI), since its inception in 1968, is constantly working in the area of welding pipes [96]. All the experimental studies related to the FSW of pipes invariably involved the development of their fxtures to facilitate the welding process. Te reason being, the fxtures are dependent upon the dimensions of the pipe (diameter, thickness, and length) that were selected for the study (Figure 6(c)). Also, there are high chance that the design of the fxture infuencing the process outcomes. Te FSW process ofers high mechanical strength at the welding joints, making it more advantageous than fusion welding techniques. Te very low distortion in the FSW process makes it a suitable pipe joining technique, which is not the case with fusion welding techniques. Te FSW process is also controlled by quite a number of parameters deciding the quality of weld like other welding processes. Most of the research work has been carried out in fat plate geometries for understating the outcome of process parameters on the weld joints. Tool design and tool speeds (rotational and traverse) are the primary factors for parameter optimization studies.
Apart from friction welding [97][98][99], FSW is predominantly employed in joining Al and its alloy pipes. Tere are traces in the literature where copper (Cu) and steel pipes are also joined using the FSW process.

FSW of Al Alloy Pipes.
Te past decade's history has traces of works involving attempts in employing FSW for joining of Al pipe. Lammlein et al. [100] used the FSW process to join 106.68 mm diameter Al alloy 6061-T6 pipes with a wall thickness of 5 mm and reported to achieve 70% strength that of the base metal. Ismail et al. [101,102] studied the efect of tool rotational and weld speeds over the tensile strength of AA6063 pipe joints. Tey obtained high-strength joints of 126 MPa under optimum values of welding parameters (a rotational speed of 1500 rpm and a welding speed of 1.8 mm/s). In a similar approach, Ismail et al. [103] developed a dedicated fxture for joining pipes using the FSW process (Figure 7(a)). Te fxture has a series of components that hold the pipe frmly while performing welding. It is also responsible for delivering rotary feed while welding. Welding of AA6063-T6 pipes was performed using this fxture by varying the weld speed and travel speed. Tis welding setup has produced a joint that has a smooth surface fnish at the weld zone.
Meyghani and Awang [104] performed a thermomechanical simulation and compared the FSW for fat and curved geometries (Figure 7(b)). Tey found that the  [92], and (c) a typical rotary fxture for FSW of pipes [46]. temperature and stress behavior were approximately asymmetric on diferent sides of the tool. In the tool front and back sides for both the fat and the curved models an approximately diferent behavior was observed, while, there were some similarities in the temperature behavior at the advancing and the retreating sides. Te temperature pattern in the fat model was lower at the front of the tool as compared to the trailing side (Figures 7(c) and 7(d)). Below the tool, the high heat fux between the outside of the shoulder radius was seen and the workpiece interface layer generated a "V" or 'basin" shaped temperature gradient [104] (Figures 7(e) and 7(f )).
During the FSW of Al alloy 6061-T6 pipes Doos and Bashar [105] found a tensile strength of 61.7% compared to that of base metal strength. Similarly, Ismail et al. [101] obtained a void-free and high-strength pipe joint of AA6063 under optimum parameters (at 2.4 mm/s and 1500 rpm). Senthil and Parameshwaran [106] experimented with three diferent tool pin profles: square, pentagon, and hexagon for FSW of AA 6063 pipes. It was seen that hexagonal tool pin profles at 1600 rpm could produce maximum strength. Khourshid and Sabry [107] investigated the FSW of AA 6063 pipe joints. It was found that a higher tool rotational speed of 1400 rpm had produced pipe joints of high strength. Ismail et al. [108] have studied the temperature cure characteristics during the FSW of AA 6063-T6 pipes at the plunge stage. A dwell time of 54 s was required to complete the plunging process. It was found that the temperature on the AS was greater than the RS with variations of 5% to 25% from the weld centre. Maggiolini et al. [109] performed an analysis on the crack path and fracture modes in the Al alloy 6082-T6 friction stir welded tubes subjected to tension-torsion loading. 38 mm outer diameter (OD) tubes were welded using a specially fabricated fxture system. Tey noticed that 50% of the cracks were initiated from the AS of the weld joint, around 39% were from the RS, and 11% were from the remaining positions of the weld. A weld joint efciency of 55% was achieved a similar observation when compared to fat welds. Te major problem reported was the designing of a retractable tool to avoid the pin-hole formation. Automation was recommended for joining pipes using the FSW process.
In 2009, Defalco [110] reported that FSW has a tremendously positive impact upon employing it to the pipeline industry in terms of performance and cost-saving. Tavassolimanesh and Alavi Nia [111] developed a new method to clad dissimilar pipes involving pure copper and AA 6061-T6 materials. Te maximum shear strength has been produced at a tool rotational speed of 710 rpm and a weld speed of 60 mm/min. Senthil et al. [112,113] performed a comparative study on the mechanical performances of friction stir weld produced by the same welding parameters for plate and pipe and reported around 10% variations between them ( Figure 8). Ronevich et al. [115] analysed the hydrogenaccelerated fatigue crack growth of FSW X52 steel pipes (Figure 8(c)). Among tests in hydrogen, fatigue crack growth rates were modestly higher in the FSW than in the base metal (BM) and 15 mm of-center tests. Meyghani and Awang [104] compared the thermomechanical behavior of friction stir welded fat and curved surfaces using fnite element analysis. Tey found that the temperature in the pin bottom area of the curved model is higher than the fat model. Tis is attributed to the diference in FSW zones between fat and curved models. Te measured stress along the cross-section showed an M-shape pattern, in which the fat model had higher stress in the stir zone than the stir zone of the curved model. Senthil et al. [116] employed nondestructive testing techniques to analyze the defects present in the Al alloy pipes joined through FSW. Flexural and crashworthiness studies were also performed on FSWed aluminium alloy pipes [113]. A novel method of joining the pipes using hybrid friction stir welding has also been reported, which is addressed at eliminating root defects that occur during the welding process [117]. [119] developed a newly designed mechanism to join highdensity polyethylene pipes using the FSW process. Te fxture study revealed that having an internally expandable mandrel will favor a good welding process. Te tool rotational speed, traverse speed, and tool ofset were optimized for high-strength joints using the Taguchi method. It was found that tool rotational speed had more infuence over joint strength. Te considered process parameters were optimized at 2500 rpm of tool rotational speed, 110 mm/min of transverse speed, and 3.5 mm of tool ofset. Chen et al. [114] tried to join AA 3003 and Cu pipes of small diameter (19 mm) using the FSW process under constant process parameter conditions (Figure 9). A tool ofset study showed that large Cu particles were accumulated in the nugget zone and the bulk interface. Te four weld regions of the pipe circumference, namely, the former (−40°to 90°), middle (90°t o 180°), later (180°to 320°), and overlap (320°to 360°), had diferent responses to the mechanical properties of the weld joint due to their variation in temperature exposure. Te overlap region resulted in defect-free joints due to the second pass. Te latter region showed a high strength profle with a good yield curve, but the overlap region had the highest strength among all regions with a value of 213 MPa.

Employment of Optimization.
After establishing the FSW process for pipes, researchers have attempted to optimize the process parameters to obtain sound weld joints ( Table 2). Both traditional and nontraditional optimization techniques have been used to optimize the process parameters. El-Kassas and Sabry [120] employed a hybrid RSM-fuzzy model to optimize the parameters of underwater FSW to join AA 1050 pipes. Te process parameters, namely, tool rotation speed, traverse speed, and tool shoulder diameters, were varied, and their optimal conditions were found for superior tensile strength of the pipe joint. Tese variable process parameters were optimized at 980 rpm, 200 mm/min, and 20 mm, respectively, for this welding process considered. Akbari and Asadi [121] used the FSW process to weld A356 pipes. Tey used the Taguchi method to optimize the mechanical properties of the FSWed AA356 pipes. Te process parameters viz., tool pin shape (threaded, triangular, square), tool rotational speed, and traverse speed were optimized for optimal combinations under three levels. Te responses considered were Si particle size, hardness, and tensile strength. Te 3D FEM analysis method was employed to simulate the process using a constant shear friction model as follows: where τ f , μ and σ n represent the frictional stress, friction factor, and shear yield stress, respectively. Te simulation results showed that the 0.5 mm plunge depth produced a defect-free weld. Te results of the simulation were appended to the experimental methodologies, and welds were created as per the Taguchi method's L9 array. Te optimal conditions have arrived as a tool rotational speed of 1600 rpm, traverse speed of 80 mm/min, and a square tool pin shape. Aliha et al. [122] joined AA6063 Al alloy cylinders of 140 mm outside diameter using the FSW process. During this study, the efect of tool rotational speed and traverse speed on the mechanical properties of the weld was studied. Bend tests were performed for the longitudinal (L) and transverse (T) orientations of the welded joint specimen. Higher welding speeds (1250 rpm) produced high-strength welds. Te average hardness of the FSW zones was found to be greater than the base metal hardness. Te fracture load and energy of the FSW specimens were found to be signifcantly higher than the base metal. El-Kassas [56] optimized the FSW of AA 6061 pipes using multicriteria decision-making techniques. Tey used the Technique for Order of Preference by Similarity to an Ideal Solution Advances in Materials Science and Engineering (TOPSIS) and grey relational analysis (GRA) to optimize the process parameters for attaining maximum tensile properties. A tool pin diameter of 5 mm, a tool rotational speed of 1800 rpm, and welding speed of 10 mm/min were found to be the optimal process parameters. Khoushrid et al. [123] employed regression analysis to optimize the process parameters for joining aluminium alloy 6061 pipes using the FSW process. Te regression equations, which were the results of their investigation, are as follows: where x = tool rotational speed in rpm, y = workpiece thickness in mm, and z = tool traverse speed in mm/min. Senthil et al. [46] employed an RSM-based desirability function approach to optimize the welding speed and the tool rotational speed for joining AA 6063-T6 50 mm OD pipes. Te RSM technique was used to populate the design matrix, based on which experiments were conducted. Te desirability function was employed to carry out the optimization process, as shown in Figure 10. Te optimized parameters for obtaining more than 70% of the base metal strength were a 1986 rpm tool rotational speed and a 0.65 rpm (102 mm/min) weld speed.

Closure of Exit-Hole.
One of the most shortcomings in joining pipes using the FSW process is that the exit-hole forms at the end of the process [124]. A few nascent attempts at closing this exit-hole are found in the literature. Ghavimi et al. [125] proposed a "flling FSW" method to fll the exithole that results during the FSW of pipes. Tey compared the semiconsumable similar pin with the semiconsumable dissimilar pin for flling the FSW joint between AA 5456 plates and pipes. Te optimized flling process parameters of 800 rpm pin rotation, 50 mm/min plunge velocity, with a similar pin of 8°cone angle were proposed. Also, similar flling FSW has also been reported by Han et al. [126]. Hattingh et al., [127] joined 38 mm OD Al alloy 6082-T6 pipes using the FSW process and compared them with plate welds. A retractable tool was used for performing the welding process to avoid the left-over hole. Te results from both tubular and microtensile specimens had joint efciency of 0.55. A frm shear texture was observed in the central part of the stir zone. Behmand et al. [128] have used consumable pins to fll the exit-hole of FSW joints. Te best results were obtained with appropriate process parameters, such as rotational speed and plunging time, as shown in Figure 11. New FSW derivatives-likeself-support FSW has shown their efciency in addressing keyhole issues of AA materials [129]. Tus, apart from straight geometry studies, circular geometry studies have taken the FSW process closer to industrial applications. During FSW of circular geometries, the facilitation of rotary fxtures has played a major role in establishing the welding process.

Importance of Fixtures in the Welding Process
Most of the automated manufacturing, inspection, and assembly operations require fxtures to achieve cost and time efectiveness. Fixtures are used to locate and hold the workpiece and facilitate the particular industrial process. Also, such a holding must be consistent throughout the operation and must constrain the workpiece in a secured position till that special operation completes. Fixtures are usually designed and manufactured for a particular workpiece individually. Te three major structural components of a fxture are locators, clamps, and supports. Te various factors that infuence the fxture design are the overall dimensions of the part, condition of a part material, degree of accuracy required, number of pieces to be made, loading and clamping surfaces, and fnally, type and size of a machine tool. Fixtures are widely used in machining, welding, assembly, inspection, and in testing processes. More gaps exist that are to be flled in welding fxture research areas [130]. Fixtures are employed in a welding process to obtain repeatability along with the achievement of the required targets. Residual stresses (both tensile and compressive) and distortions are the major problems associated with all welding processes. Stress corrosion cracks are caused by the tensile residual stresses reducing the fatigue life of the weldment [131,132]. Compressive residual stresses are induced when there is compressive loading, which results in a decline of buckling strength [133]. Te distortion may cause mismatching of joints and deviations in the targeted physical parameters [134]. Most of the studies on the weld quality and optimization of welding techniques (involving residual stresses and distortion) were parameterized only based on the metallurgical factors and physical efects of tools and workpieces [135]. But fxture efects were not incorporated as they too infuenced the quality of the weld and productivity [136,137]. However, few researchers have started to consider the efect of fxtures into account during their work, which is the primary attention caught up by this paper. Figure 12 assembles the various fxtures used in the FSW of pipes. Figure 12(a) depicts the fully automated fxture for FSW of similar pipes, and the semiautomatics fxture is shown in Figure 12(b). A completely automatic fxture used to join dissimilar pipes (Al-Cu) is shown in Figure 12(c). Tis employs an expanding mandrel to facilitate the stir zone support. Fixtures with bottom rollers is   12 Advances in Materials Science and Engineering used to reduce the bending moment created during pipe welding due to axial load (Figure 12(d)). Fixture for the lap joint of pipes is also employed, as shown in Figure 12(e). Figure 13 shows the primary functions of a typical fxture along with its possible parameter controls. Te desired fxture in a welding process must be the one that brings the distortion to zero and also, at the same time, must facilitate a stress-free weldment with optimal fxture elemental factors involved. Also, such a fxture must also be able to take care of the after-efects of the welding process like cooling. Te elemental factors of a fxture may include the number of clamps and locators, the location of clamps and locators, and the amount of clamping forces. Locators play an important role in orienting the workpieces with respect to the welding directions. But recent studies [137,139] reveal that the traditional "3-2-1" locating scheme is no longer valid in meeting the latest desired needs of the fxture [140]. Tis calls for more researches related to computer-aided welding fxture design [130]. A welding fxture thus developed should help in the reduction of deformation in the workpiece due to thermal expansion, hence avoiding the dimensional variation.

Fixtures in the FSW Process.
Te efect of the fxture on the FSW process is much higher than in any other welding process. Te fxturing elements involved in FSW are shown in Figure 14. In FSW, nearly half of the total mechanical energy is converted into heat energy, increasing the workpiece's temperature, and some part of the remaining energy causes deformation in the fxture elements [143]. Researchers are considering the clamping forces majorly controlling the distortion that happens during the welding process and the residual stresses developed due to the fxturing systems. Te initial and working clamping forces in an FSW process were studied by Richter-Trummer et al. [144] for butt welding of AA2198-T851. Tey designed a special fxture device that controlled the initial clamping force and measured the evolution of the clamping force during welding. A stereo-based vision system measured the distortion, and load cells measured the clamping force. Mechanical tests and microstructural analysis reports revealed that the clamping forces did not infuence the tensile properties; however, dispelling of tensile properties occurred under higher clamping forces. Distortion and residual stress measurements showed that the lower residual stress was present in highly distorted workpieces and vice versa. A higher clamping force of 2500 N led to superior weldment properties in their work. Te efect of the fxture (before and after release) on the orthogonal residual stress distributions was reported by Chen and Kovacevic [145] during their study on the FSW process. Before releasing the fxture, the lateral stress with a maximum value of 316 MPa was found to be higher than that of the longitudinal stresses.  [46], (c) fully automated fxture for dissimilar welding [114], (d) fxture with bottom roller supports [138], and (e) fxture for lap joint of pipes [111].  But both the longitudinal stress and the lateral stress decreased signifcantly when the fxture was released. Tus, both longitudinal and lateral stresses were infuenced by the fxture release. Hence, fxture control during and after welding is required to control workpiece deformation and inducement of residual stresses. Finally, they concluded that the fxturing release will afect the weld's stress distribution. Also, they remarked that further development of the FSW requires an assessment of the fxturing condition to control the stress distribution. However, the fxture system employed was not reported. Te efect of fxture release alone on the residual stress was studied by Zhu and Chao [143]. Tey modeled the fxture release efect during the cooling of the weldment area. It was observed that the residual stresses in the longitudinal directions decreased notably after releasing the fxture than that of releasing before. Hence, reports were made that the fxture release must be considered in the analytical simulations also for the residual stresses determination in the FSW process. Te same was cited by Baghel in his survey [146]. Also, Baghel and Siddiquee, in another study [147], designed and developed a fxturing system involving a fxture, clamps, and a key using mild steel for the FSW process that is to be mounted on a vertical milling machine. Tis developed fxture is best suited for a robust vertical milling machine and proved its fexibility in welding stainless steel 304 plates of various thicknesses. A fxture setup used in the FSW process by Soundararajan et al. [148] consisted of a clamp to constrain the sides of the workpieces and a backing plate to constrain the bottom portion. Te amount of clamping force used was not provided. It was found that due to fxture constraints on the sides of the workpiece, high compressive stresses were found at the welded areas because of thermal expansion. Another type of fxture was employed by Hwang et al. [149] during their study on the FSW process, but its efects on the welding process were not examined. Patil and Soman [47] carried out the FSW with the help of a specially designed fxture on a CNC vertical milling machine. Tey reported that the fxture was used to secure the plates to obtain the initial joint confguration. However, its efects and infuence on the welding process were not reported. Salloomi et al. [150] have shown that the efect between the tool and its surrounding area coupled with the efect of clamps will lead to variable compression stress along the weld direction with compression near the tool location. Tey observed that due to high clamping, the longitudinal and transverse residual stresses reach maximum values at the edges. It was also concluded that clamping constraints and locations afect the stress components through signifcant localized efects beyond the heat-afected zone. Te analysis model of Bufa et al. [151] for an FSW welding process was assured with clamping and supporting conditions. Only the  Figure 14: (a) Fixture elements in FSW [141] and (b) tool supplements in the FSW process study [142].
Advances in Materials Science and Engineering efect of fxture release (unclamping state) was included. However, the fxture considered and its efects were not presented in detail. A specially designed rotating clamping fxture was used by Doos and Bashar [105] to clamp and hold the two segments of the 6061-T6 Al pipe together for butt joint welding through the FSW process. Tis clamping fxture consisted of a gearbox with a rotating clamp, a fxing mandrel, and internal and external anvils. Teir report on the fxturing efect was limited only in preventing the workpiece assembly from any movement during the welding operation.

Fixtures in Other Welding Processes.
Other welding processes like arc welding and laser welding are also afected by the fxture systems, but the studies on their efects are only limited. Arc welding processes also experience distortions like the FSW process [152]. Researchers observed that the welding fxtures employed for the arc welding process infuenced both weld pool geometry and residual stress distribution. Te fxture developed by Kohandehghan and Serajzadeh [153] made the depth of the weld pool reduce by 21% and the transverse residual stress by 76%. Sikstrӧm et al. [154] performed an experimental and modeling approach to study the clamping forces' infuence on distorting the workpiece during gas tungsten arc welding. Tey used the specially designed fxture setup to investigate the infuence of fxture clamping forces on the structural integrity of the welded workpiece. Te designed fxture consisted of screw holes for stif fastening of the workpiece at one side and a cylinder to maintain the variable fxture force through the base support, block, and the yoke. Te replaceable plates ensured the same friction for all welding facilitating zero slip. Te workpiece was made to experience three diferent levels of clamping force, namely, stif, medium, and loose clamping with force values of 295 N, 84 N, and 0 N, respectively. Te experiments showed that the welding with loose clamping produced minimized residual deformation of 0.30 mm when compared to 0.40 mm and 0.53 mm with medium and stif clamping, respectively. Welding fxtures were also examined to prevent distortions during the cooling process. Vural et al. [155] examined the efect of weld fxtures in preventing the distortions during cooling of the workpiece after arc welding. Tey used a special welding fxture and analyzed it with two conditions, namely, cooling on fxture and cooling outside fxture. Te designed welded steel structure had a Ushaped confguration, which consists of part1 constrained by attachments in the x and y directions and by pressure force in the z-direction. Part1 and attachments constrained the y and z directions of part2 and part3. Teir x-direction was constrained by pressure force. Terefore, part 2 and part 3 were used in measuring the distortions along the x-axis. For each experiment, two specimens were welded; one was cooled on the fxture and the other outside the fxture. Results showed that the distortion increased with an increase in heat input. Also, minimum distortions were developed while cooling was on the fxture, compared to maximum distortions if the cooling was outside the fxture. In a laser welding process, Liu et al. [156] performed a measurement of clamping force variation. Tey used preset clamping forces in diferent experiments on diferent workpieces with thickness, namely, 1 mm, 1.2 mm, and 1.5 mm. Te measurement of clamping forces revealed that the local material expansion resulted in variations in the clamping forces during the welding process. It was found that the established HAZ (heat afected zone) induced a thermal expansion, therefore increasing the clamping forces [144]. It was also proposed that the weld breaking load and weld strength can be improved by choosing optimal clamping forces. Te efect of the fxture in the various welding process is depicted in Table 3.

Fixture Development for FSW of Pipes.
In the FSW of pipes, the fxture plays a very crucial role. Te fxture development for FSW of pipes will require two approaches, as presented in Figure 15. In the frst approach, the pipes are mounted onto a rotary fxture and rotate during welding. Tis will constitute the welding speed parameter. Te second approach being the pipes can be held stationary and the entire tool head can be rotated. Te former is the type that all the laboratory scale studies follow, as this facilitation is easy to establish. All the researches discussed in Section 3 holds the same state as the frst approach. Whereas the latter is a difcult one, as it requires a complex machine setup for performing FSW [160]. In the case of pipe rotation, the setup intends to aid the welding speed (rotational speed of pipe) that is independent of the machine, which controls the tool rotational speed and tool plunge depth. Hence, this particular parameter becomes unanimous during the welding process. When the entire tool head rotates, it will also make the spindle head to which the tool is fxed also to rotate, which also requires a guiding rail. Also, the axial load also needs to be enforced along the tool rotating axis, which is continuously moving. As many factors are included, this will lead to a decrease in the quality of the weldment produced.

Fixture Desirability for FSW of Pipes.
Te fxture development is an integral part of the FSW of pipes, and the frst approach remains the common choice due to its relative simplicity compared to the second approach. However, the developed fxture using the frst approach has to satisfy the following requisites for successful facilitation of the pipe welding process for Figure 16 an industrial application.
(i) Holding of the pipe frmly without any slippage: Te pipes are to be clamped frmly to the rotation fxture (along the mandrel) so that zero slippage is ensured throughout the welding process [161]. Also, care is to be taken that the clamping forces are not afecting the weld property. (ii) Withstanding the axial load exerted by the tool: During the welding process, the tool is plunged into the weld zone with an axial load calculated according to the material mounted. Tis axial load may be low for AAs, whereas high for steel. Te action of the tool load will make the fxture appear Advances in Materials Science and Engineering to be in a fxed beam condition with a point load at the centre, which is a perceived diagram of the pipe fxture (D-pipe diameter, 1/2 L-pipe specimen length). Tis will induce the bending moments to be generated at the clamped points (fxed locations) as well as bending at the weld zone (centre of the beam) as per the bending in the following equations: Bending moment at end supports, M � F y L Deflection at the weld point Where F y is the tool load acting in a downward direction (kN), E is Young's modulus of the pipe material (MPa), and I is the area moment of inertia (mm 4 ). Tis efect may cause the pipes to bend at the joints, which will result in poor cause root defects. As per (4), this bending will increase as the length of the pipe between the end-supports the increase. In contrast, the bending decreases when thickness (T) and diameter (D) increase proportionately (since I � TD 3 /12). Tis factor if underestimated may result in root defects, which is a result of poor integrity at the toot side. Studies have also been reported where these root defects are addressed by novel hybrid welding processes [117].
(iii) Support to hold the integrity of the weld zone viz., mandrel: Backing the weld zone is a major facilitation for holding its integrity. In the case of plate welding, backing plates are used [162]. For pipes, a mandrel is used as a backing structure. Tis mandrel is to possess favorable thermal properties like a high melting point and low thermal expansion when compared to that of the pipe material considered. Te mounting and removal of pipe on the mandrel must be made accessible. Expanding mandrels would be the apt option for this case [114]. Mostly, mandrels are made for a specifc diameter of pipe. Hence, each diameter of pipe will require a separate mandrel to be fabricated and employed. (iv) Rotary motion, with variable speed options: Tis is one of the basic requisites of any pipe fxture for delivering rotary motion of pipe, which is parametrized as weld/traverse speed. Tis was achieved using an external motor with a suitable gearbox. Te motor's torque is to have enough capability to overcome the frictional shear forces at the tool-pipe junction [121].
Te second approach will have more complexity as the entire tool head has to rotate, ensuring the uniform tool load. Technological advancements may pave the way for the selection of approaches based on the application requirements. Figure 17 shows the facilitation chart for FSW of pipes for obtaining sound pipe joints.

Advances in Materials Science and Engineering
Tus, the impact of fxtures on the quality of the weld is very high when it pertains to weld geometries other than fat. Te circular geometries like pipes have critical compensations from the fxture and required perfect facilitations in order to achieve sound weld joints.

Conclusions
Te FSW of pipes has started attracting researchers and the industrial community in reaching industrial standard applications. All the studies that had been reported so far in FSW of pipes have their own fxture being developed by the frst approach. But the fxture setup will vary from one study to another. Tus, FSW of pipes would require a thorough study of the fxtures to obtain sound weld joints. As the process parameters are dependent on the fxture, the optimization of the fxture elements also needs to be incorporated. One more challenge for FSW to hit the market is the unavailability of dedicated machines. Since the rotary pipe fxture is used as an external unit, the control mechanisms do not become interdependent. Also, the controls of fxturing elements were not mechanized. Te optimization works on the fxturing elements that can provide relationship characteristics to help creating controller mechanisms for the rotary pipe fxtures. For this purpose, dedicated machines can be designed and developed for FSW of pipes in future studies. Tis will pave the way for computer-aided control over the process and fxture parameters, thus increasing the efciency of the pipe welding process. Concerning the material choice, Al pipes are mostly exploited by the FSW process due to its economic tool selection options. Te FSW tool is still considered to be costly compared to conventional Al welding tools. But the economical joining of steel pipes using FSW will be challenging unless the development of a tool for joining steel is made comparatively cheaper.

Future Research Directions.
Te studies related to the development of fxtures for FSW are only count to be few. Te research directions that the future work can focus will be: (i) Joining of thermoplastic pipes using FSW is one of the areas which researchers can explore for all possible outcomes. (ii) Development of a generic fxture that can hold to perform FSW on a variety of joints can be attempted. (iii) Development of intelligent control systems is another area to automate the FSW for diferent materials and diferent geometrical requirements with good physical characteristics. (iv) Future studies can focus on integrating diferent conventional and nonconventional welding processes with FSW to eliminate the defects and extensive requirement of rigid fxturing elements. Tis can extend the choice of materials.

Data Availability
Te data supporting the fndings of the study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.  Advances in Materials Science and Engineering 19