Parametric based optimization of friction stir welded wrought AZ80A Mg alloy employing response surface methodology

An endeavour was put forward to friction stir weld flat plates of AZ80A Mg alloy by employing the central composite based design technique of response surface methodology, taking into consideration the parameters namely tool rotational speed, traverse speed, tool’s tilt angle, and pin geometry of the employed tool. Numerical model was formulated to correlate the relationship amidst the employed parameters of FSW process and the responses, namely tensile strength and elongation percentage of the AZ80A Mg alloy joints. The formulated model was also optimized to attain AZ80A Mg alloy joints possessing highest value of tensile strength. Competency of the formulated numerical model was validated employing the analysis of variance and the observations of the affirmation experiments plotted in the form of scatter diagrams revealed an appreciable agreement with the values of the anticipated models. Response and contour plots generated from the established numerical model was employed to understand the interactive impacts of the parameters of FSW process on the variables of the response. AZ80A mg alloy joints fabricated during the employment of tool possessing straight cylindrical geometry at atilt angle of 0.630, rotational speed of 962.077 rpm, tool traverse speed of 2.105 mm sec−1 possessed the highest tensile strength of 195.299 MPa and was proven to be free from flaws.


Introduction
AZ80A is a superior strength wrought alloy of magnesium (Mg), frequently being employed for extrusions and forgings demanding extreme fatigue strength and excellent resistance against creep.This AZ80A Mg alloy possesses a fairly larger values of tensile strength amidst the Mg alloys.It has a reasonable low temperature of melting and a low electrical based conductivity [1,2].The rate of machinability exhibited by AZ80A Mg alloy is excellent and can be heat treated or extruded readily.Owing to these reasons, AZ80A Mg alloy is frequently employed both in the aerospace and automotive sectors for fabricating superior strength components, parts and accessories including gear boxes & rotor hubs for helicopters, inter stage fairings & frames of missiles, gear struts for landing of aircrafts, brake housings for automobiles, frames for bicycles, supercharger and screw machine accessories etc [3,4].
Joining of these AZ80A Mg alloys cannot be achieved by employing traditional welding processes, as the alloys of Mg gets oxidized very easily and is prone to catch fire easily during the employment of arc & gas welding processes.Moreover, the easy and ready tendency to corrosion towards the electrodes (made out of copper) also hinders the employment of these conventional welding processes [5].In addition to this, joining of AZ80A Mg alloy using other fusion category welding processes (including metal inert gas welding, tungsten inert gas welding) will result in the generation of coarse grains, partially melted zones and cracks in the zone of joint owing to its larger thermal conductivity and coefficient of linear based expansion.Apart from this, during the employment of resistance spot welding to join AZ80A Mg alloy, Mg reacts at a faster rate with the atmosphere leading to the formation of layers of oxide on its surface, very rapidly, thereby deteriorating the quality of the fabricated joint [6][7][8].
Some of the researchers have made experimental attempts to weld together the alloys of Mg (especially AZ80A Mg alloy) by employing solid state welding methodologies including laser beam welding, electron beam welding [9,10].But results of these experimental attempts were not satisfactory.For instance, the efficiency of the laser beam welding process is highly dependent on the physical attributes of the metal to be joined together.In this regard, as the alloys of Mg (including AZ80A) possesses certain inherent features including as inferior degree of absorptivity of beams of laser, easy tendency to get oxidized and larger solubility for hydrogen in the state of liquid etc These physical attributes of the Mg alloys when being welded by employing the laser beam welding processes results in excessive formation of large sized pores in the joint regions, loss of various elemental constituents and formations of porous oxide type inclusions, thereby leading to the generation of inferior quality laser beam welded joints [11].
Likewise, experimental research works put forward with the objective of joining together alloys of Mg employing the electron beam welding process have also reported the occurrence of unstable joint pool, massive spatter, undercuts etc, in the welded Mg joints [12].Moreover, employing electron beam based welding process consumes more time to produce welds and there prevails a need for high degree of vaccum (in the order of 10-3 to 10-6 Torr).Another drawback with respect to electron beam based welding process is that the capabilities of the employed vacuum chamber limits the size of the workpieces to be welded and due to these reasons, electron beam welding processes is also not widely employed for joining together Mg alloys [5].
At the same time, friction stir welding (i.e., FSW) process categorized as a unique solid state welding process was proven to be very much effective in joining together alloys of Mg.This was mainly due to the fact that during the employment of the FSW process, formation of the joints takes place at temperatures lower than the point of melting of the parent metals [13,14].Another attractive feature of this FSW process is the refinement in the micro-structure.In conventional fusion based welding processes, the employment of those welding techniques results in the formation of a joint zone possessing large sized, unrefined grains in their micro-structure.At the same time, as during the process of FSW melting of the parent metals does not occur, the friction based heat (almost 75%-88% of the temperature of melting of parent metal) plasticizes the metal along the line of joint, thereby resulting in the generation of finely refined and homogenously distributed grain structures in the zone of nugget [15].Thus, FSW process can be regarded as an innovative, entirely hazard free, solid phase welding process which is capable enough to fabricate flaw free superior quality longitudinal weldments, especially butt, for a wider range of metals of different thickness [16,17].
Despite these merits, attainment of flaw free Mg alloy joints possessing appreciable tensile strength using the process of FSW is a challenging endeavour, as the process based parameters combination (including distinctive tool pin profiles and other parameter combinations), employed during the FSW process plays an inevitable role in ascertaining the quality of the Mg alloy welded plates [18,19].Some of the process based parameters of the FSW process which are significant w.r.t the joint quality includes rotational speed of the tool, its speed of traverse, load being exerted axially downwards, geometry of the tool pin, diameter of the tool shoulder, distance of tool offset, angle of tilt of the tool etc Selecting the proper and ideal combination of these significant process based parameters of the FSW process is really a very tedious and time consuming task.Traditional ways of selection of suitable combination of these parameters, so as to attain good quality joints were time consuming [18].Furthermore, these traditional combination selection methodologies relied on trial and error based techniques, which were not very much efficient.Alternative to these conventional parameter selection techniques, there exists several design strategies to anticipate the outcome for a provided batch of process based parameters, more effectively and in shorter duration [20].Methodology of Taguchi is one of these alternative methods for selecting suitable parameter combinations, in which an orthogonal based array is employed to reduce the total number of experimental runs.Major limitation of this Taguchi based methodology is its attained results are very much relative and does not exactly predict which parameter has a high impact on the performance based characteristics.Moreover, it does not analyze all the possible combinations of variables that impact the outcomes of the process, particularly when the process is complex, non-linear or dynamic [21,22].As a result, in the recent years, several researchers have preferred response surface based design methodology for their experimental works with respect to the FSW process and have proved that the employment of this RSM ( i.e., response surface methodology) is effective in formulating the empirical relationship amidst several parameters of the FSW process and in anticipating the responses [23][24][25][26].As a result, in this experimental investigation, an attempt was put forward to attain AZ80A Mg alloy joints possessing highest value of tensile strength through the employment of central composite based design technique of RSM.Numerical models were formulated to anticipate the outcomes of the FSW of AZ80A Mg alloys and are compared with that of the examination reports (attained from SEM, Macro and Microstructural Examination) of the friction stir welded joints.All the AZ80A Mg alloy flat plates were butt welded using the technique of FSW by employing a semiautomatic category of FSW machine housed with a 416 ×815 mm work table and this table had the capability to traverse in multiple directions including a 410 mm distance of travel both vertically and horizontally, a 525 mm distance of travel longitudinally and the photograph of this employed machine can be seen in the figure 1(a).

Schematics of experimental investigation
During this experimental investigation, FSW of the AZ80A Mg alloy was carried out by holding together strongly these plates using uniquely designed and fabricated metallic clamps from both the sides of the distinctly designed fixture and jig assembly.A tool fabricated out of M42 grade high speed steel possessing five distinct pin geometries was used for FSW the flat plates of AZ80A Mg alloy plates and the photograph illustrating the distinctive tool pin geometries employed in this experimental investigation can be seen in figure 1(b).

Parameters of FSW process
Based on the literature survey [26][27][28], the most dominant parameters which were proven to play a vital role in impacting the properties of the AZ80A Mg alloy joints to be attained by employing the FSW process are listed out in table 2, along with their respective distinctive levels, notation and units.Due to the broad spectrum of the dominating parameters of the FSW process, it was proposed to choose a central composite category response surface based matrix of design possessing 4 dominant factors (parameters of FSW process), 5 distinct levels having 31 runs.
Due to the broad spectrum of the dominating parameters of the FSW process, it was proposed to choose a central composite category response surface based matrix of design possessing 4 dominant factors (parameters of FSW process), 5 distinct levels having 31 runs.Many trial experiments were performed to ascertain the range of working of all the dominant FSW parameters considered in this work.The upper and lower limit of these considered dominant factors (i.e., parameters) were coded as +2 and -2 respectively.The middle values were calculated using the below mentioned equation:-  where X i is the mandatory coded value of the variable X, X is any value of the variable ranging from X max to X min .Xmax and Xmin are the highest and lowest level of the variable respectively.The 31 sets of the coded scenarios comprise a half replication of 24 = 16 factorial based design along with 7 points of center and 8 start (i.e., axial) points.All the parameters of the FSW process at the middle level (i.e., 0) constitute points of center and at the same time, the combinations of each parameter of FSW at its higher value (+2) or lower value (-2) with the other 4 parameters at the middle level constitute the star based points.As a result, the 31 experiment based runs permitted the assessment of the two way interactive, quadratic and linear impacts of the parameters of the FSW process on the ultimate tensile strength of the AZ80A Mg alloy joints.As recommended by the matrix of the design, a total of 31 AZ80A Mg alloy joints were friction stir welded and photographs of the friction stir welded AZ80A Mg alloy joints are displayed in the figures 2(a) and (b).

Formulated numerical model
A total of 31 experimental runs as described in table 3 were employed as the input data in formulating the numerical equation by the employment of experimental design methodology, employing RSM.The functions of response namely highest tensile strength (HTS) and elongation percentage (EP) of the joints were considered as the functions of the geometry of the tool pin (G), tilt angle of the tool (A), tool's traverse speed (T) and rotational speed (R).It is being expressed as: The 2nd order polynomial type regressive equation employed to describe the response of surface was given by,  The preferred polynomial for 4 factors was expressed as,   The larger value of R 2 towards 1 and the lower value of standard error announces that the numerical relationships are quite competent and can be employed to anticipate the responses without reasonable error.Larger value of adjusted R 2 escalates the variation and expresses more beneficial fluctuating elements in the formulated model.The statistics based result describes a larger R 2 value of 0.9797 and 0.9557 and values of adjusted R 2 of 0.9620 and 0.9169 for the highest tensile strength and percentage of elongation respectively announcing that a very large degree of a match prevails amidst the values of anticipated numerical relationship and experimental investigation, thereby validating the values of experimental investigation.Competency of the formulated numerical model was assessed by employing the ANOVA (i.e., analysis of variance) technique and is described in table 5.It was a proven fact w.r.t ANOVA that if the computed F-ratio value of the numerical model exceeds the F-ratio tabulation value at a level of confidence of 95%, then the formulated numerical model can be regarded as satisfactory.From table 5, it can be apprehended that the F-ratio calculated value of the assessed model is very larger than the F-ratio tabulated value at a level of confidence of 95% demonstrating that the formulated numerical models are satisfactory.

Validation of experimental result
Validity of the experimental relationships was tested by plotting the scatter diagram taking the investigation run value as the x-axis and the anticipated value as the y-axis and the plotted scatter diagrams for the highest tensile strength and elongation percentage are illustrated in figure 3. It can be visualized from these scatter diagrams that the plots are scattered very nearer to the line of 45 degrees, thereby announcing the perfect competency of the formulated numerical and experimental relationships.Investigational realistic runs were performed to verify the validity of the formulated experimental and numerical relationships.A total of 5 AZ80A Mg alloy joints were friction stir welded employing distinctive values of factors apart from those employed in the matrix of design and their highest tensile strength and elongation percentage were estimated.The attained results are tabulated in table 6.From this table 6, it can be apprehended that the anticipated values are perfectly acceptable without any observable error.

Macro and microstructural analysis
Optical macrograph of the cross section (being crosswise to the direction of travel of tool) of the friction stir welded AZ80A Mg alloy joint possessing highest tensile strength (fabricated during trial run no.3 at a 950 rpm, 2 mm sec −1 tool. 1 degree tool tilt angle and by a straight cylindrical tool pin geometry) is illustrated in figure 4. In this macrograph, we cannot observe any sort of welding flaws including cracks, tunnels, voids etc, in and around the nugget zone (i.e., NZ).
Optical microstructures of the cross section (being crosswise to the direction of travel of tool) of the parent metal (i.e., AZ80A Mg alloy) and the heat affected zone (i.e., HAZ), the thermo mechanically impacted zone (i.e., TMIZ) and the nugget zone (i.e., NZ) of the friction stir welded AZ80A Mg alloy joint possessing highest tensile strength (fabricated during trial run no.3 at a 950 rpm, 2 mm sec −1 tool. 1 degree tool tilt angle and by a straight cylindrical tool pin geometry) is portrayed in figures 5(a)-(d) respectively.
The microstructural images were attained with the help of Dewinter type inverted trinocular metallurgical microscope and the software incorporated with this optical microscope was employed to enhance and analyze  the images.Polished samples were etched with an acetopicral solution (0.4 g picric acid, 13 ml ethanol, 3 ml glacier acetic acid and 3 ml boiled water) to reveal the microstructure of the weld, base metal (BM) and HAZ regions.By examining figure 5(a), it can be observed that the metal of investigation (i.e., AZ80A Mg alloy) possesses large sized, irregular grain structures and the grains are unevenly scattered, non-uniformly distributed.The parent metal inherits a dendritic network of solid solution of Mg together with huge sized precipitates of Mg 17 Al 12 precipitates at the boundaries of the grains.
Figure 5(b) illustrates the heat affected zone of the friction stir welded AZ80A Mg alloy joint and it can be seen that in this zone, the impact of the temperature (during the FSW process) have fused the grains of this zone and the particles of fragmentation in this zone have experienced initial recrystallization.By observing the figure 5(c), which illustrates the thermo mechanically impacted zone (TMIZ) of the AZ80A Mg alloy, it can be understood that the grains in this zone have undergone thermo mechanical transformation due to the combined impact of the peak temperature generated by the rotational speed of the tool and the stress caused by its shoulder diameter.
The nugget zone of the friction stir welded AZ80A Mg alloy joint is shown in figure 5(d) and it can be visualized that the grains in this zone have undergone an appreciable transformation.For instance, the large sized sporadic grain structures of the parent metal have been transformed into finely refined small sized grain structures which are equally and homogeneously distributed in the nugget zone.This announces that the nugget zone have underwent severe plastic deformation due to the impact of the mechanical based stirring action of the employed tool's rotating probe during the process of FSW.This refinement in the grain structures has taken place owing to dynamic recrystallization [29].

Inferences from tensile tests
Figure 6 portrays the photographs of the tensile test specimens of the friction stir welded AZ80A Mg alloy joints were prepared as per the ASTM: B557M-10 standards.Tensile related testing of these AZ80A Mg alloy joints has revealed that the AZ80A Mg alloy joint fabricated during the employment of a straight cylindrical pin profiled tool at a 950 rpm rotational speed, along with a speed of tool traverse of 2 mm sec −1 , with the tool employed at a  Joints possessing reasonable tensile strength have undergone fracture at the side of their retraction, owing to the attainment of superior bonding at the nugget zone due to the experience of optimal generated heat and rate of strain.It was observed that the generation of heat and consolidation of the plasticized metals were low at the side of retraction when compared with that of the side of advancement, which have contributed for the fracture of the AZ80A Mg alloy joints on their retracting side.
The joint fabricated during the employment of the taper threaded cylindrical pin profiled tool have exhibited lowest value of tensile strength (trial run no.10).These low quality joints have experienced fracture at their nugget zone, owing to the generation of low volume of frictional heat [30].The results of the tensile tests have revealed that the tool pin profile have played a vital role in deciding the tensile strength of the friction stir welded AZ80A Mg alloy joints.Relevant stress strain curve for these tensile test specimens attained during the trial runs mentioned in the table 3 is illustrated as figure 7.

Investigation of contour and response surface plots
3 Dimensional response surface based plots for the outcome (namely highest tensile strength) attained from the formulated model of regression are illustrated in the figures 8(a)-(f).In these 3 Dimensional plots, the optimized value of tensile strength is portrayed by the apex of the surfaces of response.
The interactive impacts of the parameters of FSW on the outcomes (namely tensile strength and elongation) can be apprehended in an easier manner by examining of the contour plots being illustrated in figures 9(a)-(f).Figure 9(a) demonstrates a circular shaped contour announcing the autonomy of impacts of the FSW parameters namely tool's rotational speed and its pin geometry at consistent center points of 2 mm sec −1 traverse speed and 1 degree tilt angle of the employed tool.By analyzing the figure 9(b) it can be understood that the modifications in the tool's pin geometry is marginally more responsive to the change in the tensile strength of the fabricated AZ80A Mg alloy joint, when compared with that of the modifications in the tool's traverse speed at    consistent center points of 950 rpm rotational speed and 1 degree tilt angle of the employed tool.Likewise, from the figure 9(c) it can be observed that the modifications in both the tool pin geometry and its angle of tilt have impacted the joint's tensile strength at consistent center points of 950 rpm rotational speed and 2 mm sec −1 traverse speed of the employed tool.
Figure 9(d) reveals that the modifications in the tool's rotational speed is marginally more responsive to the change in the tensile strength of the AZ80A Mg alloy joint when compared with that of the modifications in the tool's traverse speed at consistent center points of straight cylindrical pin profiled tool and 2 mm sec −1 tool's traverse speed.Similarly, figure 9(e) demonstrates that the modifications in both the tool tilt angle and tool's rotational speed have impacted the joint's tensile strength at consistent center points of straight cylindrical pin profiled tool and 2 mm sec −1 tool traverse speed.From the figure 9(f), it is visible that the modifications in the tool tilt angle is highly responsive to the modifications in the joint's tensile strength than the modifications in the traverse speed at consistent center points of straight cylindrical pin profiled tool and 950 rpm rotational speed.Moreover, from these 3 dimensional response surface based plots, it can be visualized that in majority of the conditions at the highest tensile strength the tool's rotational speed was observed at the center point of 950 rpm.

Discussions on direct impacts of FSW parameters
The direct impacts of the FSW parameters on the tensile strength of the friction stir welded AZ80A Mg alloy joints are illustrated in the figures 10(a)-(d).From these figures 10(a)-(d), it can be apprehended that the tensile strength of the joints escalates directly with the simultaneous escalation in the tool's traverse speed, rotational speed and its angle of tilt, then declines after attaining its highest value.Graphs, i.e., figures 10(a) and (b) exhibit that the tool's pin geometry and its angle of tilt plays a vital part in ascertaining the joint's tensile strength.The decline in the tool's tilt angle have resulted in poor binding owing to the generation of surplus volume of frictional heat and at the same time, the escalation in the tool's tilt angle have also resulted in poor binding owing to the generation of lower volume of frictional heat [31,32].AZ80A Mg alloy joints fabricated using the straight cylindrical pin profiled tool have exhibited highest value of tensile strength and the joints fabricated using the taper threaded cylindrical pin profiled tool have exhibited lowest value of tensile strength.This was mainly due to the proven fact that the modifications in the pin geometry of the employed tool alters the shear stress and will eventually lead to the variation in the development of friction, thereby reducing the volume of heat being generated.This insufficient generation of heat impacts and hinders the layer by layer unification of plasticized metal in the nugget zone, thereby declining the strength of the AZ80A Mg alloy joints [33].
From the graph illustrated in figure 10(c), it can be visualized that both the larger and slower rotational speeds of the tool have resulted in improper binding of the plasticized metal owing to the generation of frictional heat in surplus and insufficient volumes.This was mainly due to the recorded fact that the lower tool rotational speed leads to reduced heat input and lower values of peak temperature.This low temperature results in an inadequate reaction of plasticized metal, thereby hindering ideal plastic deformation of material in the nugget zone [20].As a result, macro sized cracks and tunnel flaws will occur in the fabricated joints, thereby diminishing the quality of the joints.On the other hand, larger tool rotational speeds will cause exorbitant stirring by the pin of the employed tool, leading to detachment of huge sized Al 3 Mg 2 intermetallic aggregates from the parent metal surfaces and these detached aggregates are very much large so as to get distributed in a uniform manner in the nugget zone, thereby leading to improper binding and generating flaws including cracks, voids etc, in the fabricated joints [17,34].
The direct impact of the traverse speed of the employed tool is graphically illustrated in the figure 10(d) and from this graph, it can be understood that both the larger and lower traverse speeds have reduced the tensile strength of the fabricated AZ80A Mg alloy joints.This was due to the fact that the input of generated heat is inversely proportional to the traverse speed of the employed tool [14,35].For instance, in this experimental investigation, the employment of larger tool traverse speeds of the tool have led to the generation of insufficient heat and this insufficient heat generation have contributed for the development of incomplete joint interfaces.This insufficient heat generated at larger tool traverse speeds hinders proper mixing of the plasticized metal from both the sides and leading to the formation of voids in the nugget zone.At the same time, the employment of lower tool traverse speeds escalates the heat input in the nugget zone, thereby generating intermetallic structures in larger volumes and the development of these intermetallic structures and their uneven distribution in the nugget zone reduces the tensile strength of the fabricated AZ80A Mg joints, as these intermetallic structures are prone to formation of cracks in an easier manner [24,36].
From these graphs illustrating the direct impacts of the FSW parameters, it can be visualized that the employment of parameters of FSW process at optimized values have resulted in the attainment of AZ80A Mg alloy joints possessing higher values of tensile strength and these optimized parameters have contributed for the ideal flow of plasticized metal from both sides of the weld surfaces and the uniform diffusion of these plasticized materials across the weld surfaces have contributed for the refinement of grain structures in a fine manner and for the distribution of these refined, fine sized grains in a uniform manner in the nugget zone [37].

Discussions on interactive impacts of FSW parameters
Figure 11(a) illustrates the interactive impacts of the employed tool's pin geometry and its tilt angle on the tensile strength of the fabricated AZ80A Mg alloy joints.From this graph, it can be apprehended that for all the employed tool pin profiles (except for the taper threaded cylindrical pin geometry) the escalation in the tilt angle of the tool initially escalates the joint's tensile strength and then declines after attaining its highest value.
For the taper threaded cylindrical pin geometry, the escalation in the tilt angle initially contributes to the escalation in the tensile strength and during majority of the employed tilt angles, the joints fabricated by employing taper threaded cylindrical pin geometry exhibits a lower value of tensile strength when compared with the other employed pin geometries.This is mainly because of the generation of frictional heat in improper volumes during the employment of this pin geometry [22,38].
The interactive impacts of the traverse speed of the tool and its tilt angle on the tensile strength of the fabricated AZ80A Mg alloy joints is graphically portrayed in the figure 11(b).
From this graph, it can be visualized that for all the tool traverse speed, the escalation in the tilt angle of the tool initially escalates the tensile strength and then declines after reaching its highest value.It can also be apprehended that the tool's traverse speed is inversely proportionally to the tensile strength of the tool tilt angle at an angle of 0 degree.This proportionality gets modified as the angle of tilt escalates and the highest value of tensile strength is attained during the employment of tool traverse speed of 2 mm sec −1 along with a tool tilt angle of 1 degree.It can also be understood that the employment of larger tool traverse speeds at lower tool tilt angle reduces the tensile strength of the joints [39,40].

Optimization of FSW parameters
With the objective of maximizing the tensile strength of the fabricated AZ80A Mg alloy joints, the employed parameters of the FSW process was optimized by the employment of desirability function which was formulated using the Design Expert Software.Regression based equations attained from the formulated numerical modelling was employed as a function of objective.Constraints being fixed as the limits of the parameter's values were employed in this established function.
The optimized values of parameters of FSW process for attaining AZ80A Mg alloy joints possessing highest value of tensile strength of 195.299MPa were tool traverse speed of 2.105 mm sec −1 , rotational speed of 962.077 rpm, 0.63°tool tilt angle and a straight cylindrical pin profiled tool.This combination of optimized parameters were attained based on the several number of iterations carried out using the methodology of hill climbing.Images of parent metal and various zones of this fabricated flaw free sound quality AZ80A Mg alloy joint obtained during Scanning electron microscope (i.e., SEM) based examinations are displayed in the figures 12(a)-(d).By comparing the grain structures in the parent metal and nugget zone of the flaw free joint in the figures 12(a) and (c), it can be understood that the employed tool's straight cylindrical pin geometry have played a significant part in the attainment of joint's highest tensile strength [41].This was mainly due to the impact of the larger pulsating impact generated by this straight cylindrical pin profile, which in turn have contributed for a smoother flow of plasticized metal, thereby leading to the transition of large sized, unevenly scattered grain structures into finely refined, small sized homogenously distributed grains [26,42].

Conclusions
In this experimental investigation, an endeavour was put forward to fabricate AZ80A Mg alloy joints (employing FSW process), possessing highest tensile strength by employing central composite based design technique of RSM.Numerical and experimental relationships were established to ascertain the response namely highest value of tensile strength and elongation percentage friction stir welded AZ80A Mg alloy joints.The conclusion are summarized below:-  • ANOVA based analysis revealed that the formulated numerical model was effective in forecasting the responses of the friction stir welded AZ80A Mg alloy joints at a 95% level of confidence.
• Impacts of the process parameters of FSW process namely rotational speed, traverse speed, pin geometry of the employed tool and its tilt angle on the tensile strength and elongation percentage of the fabricated AZ80A Mg alloy joints were analysed, based on the models of regression.
• AZ80A Mg alloy joints attained during the employment of tool possessing straight cylindrical pin geometry at a rotational speed of 962.077 rpm, tool traverse speed of 2.105 mm sec −1 and the tool being tilted at an angle of 0.630, exhibited highest value of tensile strength of 195.299MPa.
• The employed tool's straight cylindrical pin geometry have played a significant part in the attainment of joint's highest tensile strength.It was also proven that the employment of larger tool traverse speeds at lower tool tilt angle reduced the tensile strength of the joints.
• The tensile test obervations announced that the tilt angle of the tool had a important role in ascertaining the flaw free quality joints.The escalation in the tool's tilt angle have contributed for the enhanced distribution of the plasticized metal under the shoudler's surface and in turn have improved the mechanical properties.
• Macro and Micro structural examinations proved that the employment of optimized parameters of FSW process have contributed for the fabrication of flaw free AZ80A Mg alloy joints and the alrge sized unevenly scattered grain structures have been transformed into finely refined, small sized homogenously distributed grains due to the impact of dynamic recrystallization.
• From the response surface plots and interactive impact graphs, it was visualized that the tool's traverse speed is inversely proportional to the tensile strength of the tool tilt angle at an angle of 00.Highest tensile strength was attained at a tool traverse speed of 2 mm sec −1 along with a tool tilt angle of 10.
• Contour plots revealed that the employment of larger tool traverse speeds at lower tool tilt angle reduced the tensile strength of the joints.

2. 1 .
Machine and material 6 mm thick flat plates of AZ80A Mg alloy possessing a length of 110 mm and width of 55 mm were taken as the material of investigation in this work.Several chemical based constituents of this Mg alloy are mentioned in table 1.The flat plates of the AZ80A Mg alloy being investigated in this work exhibited a tensile strength of 287 MPa together with a yield strength of 198 MPa and an elongation percentage of 5.9% respectively.

Figure 1 .
Figure 1.Photograph of (a) FSW machine (b) distinctive tool pin geometries employed in this experimental investigation.

Figure 2 .
Figure 2. (a) & (b) Photographs of friction stir welded AZ80A Mg alloy joints using distinctive combinations of parameters of FSW process.
Where b 0 is the mean of responses, the coefficients b 4 , b 3 , b 2 and b 1 are the linear terms, the coefficients b 44 , b 33 , b 22 and b 11 are the quadratic terms and the coefficients b 12 , b 13 , b 14 , b 23 , b 24 and b 34 are the interactive terms.All the coefficients were assessed and tested for their influence at a level of confidence of 95%.The finalized numerical model being formulated to anticipate the highest tensile strength and the elongation of the friction stir welded AZ80A Mg alloy joints are mentioned below:-

Figure 3 .
Figure 3. Investigational versus anticipated values of responses (a) Highest tensile strength and (b) Elongation percentage.

Figure 4 .
Figure 4. Cross sectional macrograph of the friction stir welded AZ80A Mg alloy joint.

Figure 5 .
Figure 5. Optical micrographs of the (a) Parent metal (i.e., AZ80A Mg alloy), (b) the heat affected zone (c) thermo mechanically impacted zone and (d) nugget zone of the friction stir welded AZ80A Mg alloy joint possessing highest tensile strength.

Figure 6 .
Figure 6.Photographs of the tensile test specimen extracted from the friction stir welded AZ80A Mg alloy joints attained during the trial runs.

Figure 7 .
Figure 7. Graphical illustration of the stress strain curve for the tensile test specimen of the friction stir welded joints attained during the trial runs.

Figure 8 . 3
Figure 8. 3 Dimensional response surface based plots illustrating the interactive impacts (a) tool pin geometry and tool rotational speed (b) tool pin geometry and tool traverse speed (c) tool pin geometry and tool tilt angle (d) tool traverse speed and tool rotational speed (e) tool tilt angle and tool rotational speed and (f) tool tilt angle and tool traverse speed on the tensile strength of the friction stir welded AZ80A Mg alloy joints.

Figure 9 .
Figure 9. Contour plots illustrating the interactive impacts (a) tool pin geometry and tool rotational speed (b) tool pin geometry and tool traverse speed (c) tool pin geometry and tool tilt angle (d) tool traverse speed and tool rotational speed (e) tool tilt angle and tool rotational speed and (f) tool tilt angle and tool traverse speed on the tensile strength of the friction stir welded AZ80A Mg alloy joints.

Figure 10 .
Figure 10.Direct impacts of FSW parameters namely (a) Tool pin geometry (b) Tool tilt angle (c) Tool rotational speed and (d) Tool traverse speed on the tensile strength of the friction stir welded AZ80A Mg alloy joints.

Figure 11 .
Figure 11.Interactive impacts of (a) Tool pin geometry and Tool tilt angle (b) Tool traverse speed and Tool tilt angle on the tensile strength of the friction stir welded AZ80A Mg alloy joints.

Figure 12 .
Figure 12.SEM images of the (a) Parent Metal-AZ80A Mg alloy (b) Thermo mechanically impacted zone and (c) Nugget zone of the friction stir welded AZ80A Mg alloy joints.

Table 1 .
Several chemical based constituents of the AZ80A Mg alloy.

Table 2 .
Parameters considered in this investigation and their respective levels.

Table 3 .
Matrix of design and runs of experimental investigation.
Competency of the formulated numerical modelStatistics based outcomes of the formulated numerical relationship are described in table 4. When the value of R 2 was 1, the anticipated numerical relationship value absolutely coincides with that of the value of the experiment.

Table 4 .
Summary of test results of ANOVA.

Table 6 .
Results of confirmatory experimental investigations.