Mechanical and Microstructural Properties of B 4 C/W Reinforced Copper Matrix Composite Using a Friction Stir-Welding Process

.


INTRODUCTION
Extensive research attempt has been conducted regarding the FSW process due to its enormous advantages over traditional welding techniques.The FSW process is the most significant and desirable technique in solid-state welding because it is crackfree, fume-free, has less shrinkage and excellent mechanical strength, etc.Also, FSW finds a broad range of utilization in various sectors, such as construction, marine, aircraft, trains, automobile industries and fuel tanks to obtain quality welding of various parts.Although FSW has many reliable features, due to technological growth, it creates new challenges and problems, such as microbores, nonhomogeneous welding surfaces, high heat-affected zones, and less mechanical strength of welding.Nevertheless, the needs for CMC in various applications are increasing day by day in the manufacturing sector.Therefore, in view of improving the welding quality of the composite materials, various research attempts have been made in the last decade worldwide.
Zamani et al. [1] carried out experiments with the FSW process on aluminium silicon composite work material and optimized the process parameters using an RSM technique.They planned the design of experiment including the transverse speed of tool and suggested that the optimal process parameter combination was 1300 rpm rotational speed and 70 m/min transverse speed.Moreover, the effect of coupling process parameters increases the machining rate significantly.Argesi et al. [2] attempted to join a pure copper and aluminium alloy with SiC particles in the FSW process.They noted that 50 m/min of welding speed and 1000 rpm of rotational speed produced higher tensile strength.The application of SiC particle increases the hardness of welding strength from 160 HV to 320 HV.Sudhagar and Gopal [3] carried out experiments with FSW to fabricate the surface copper composite using Si 3 N 4 -reinforced particles in various concentration levels.They analysed the mechanical and microstructure properties of the weld and also observed that while rotating the tool work piece reached recrystallization temperature, which significantly hinders the micro-hardness of welding.The wear property increases with increased reinforcement particles over the surface copper composite by about 15 %.Thapliyal and Mishra [4] utilized the machine-learning concept to weld the copper work material in the FSW process.The most influential parameters (i.e., welding speed, rotational speed and axial forces) are considered for the experiment.They noted a 94 % improved welding strength using this optimal combination among 119 experiments.Harisha et al. [5] welded the copper and aluminium 6083 using the FSW process to study the optimal process parameter combination.They noted a 70 % improved tensile strength in the parameter combination, such as a rotational speed of 1000 rpm and a welding speed of 50 m/min.Nagesh et al. [6] welded dissimilar materials, such as copper and brass, using the FSW process.They compared the results of FSW, such as tensile, hardness, and the impact strength of dissimilar welding strength with parent metal welding, to explore the detailed nature of welding.Senthil et al. [7] studied the FSW process parameters for aluminium 6063 composites using the response surface methodology (RSM) method.They noted the tensile and yield strength of 167 MPa and 145 MPa, respectively, in the parameter value of 1986 rpm of tool speed.Uniform microstructure properties were also noted with the first optimal parameter combination.Karrar et al. [8] studied the effect of tools' rotational and transverse speeds on dissimilar welding in the FSW process.They welded the aluminium alloy 5083 and copper as work material and observed intricate microstructure in the heataffected zone.Many intermetallic mixers were seen in the welding zone, which leads to nonhomogeneous micro-hardness.The maximum tensile strength was obtained in 1400 rpm tool speed and 120 mm/min.Kolnes et al. [9] prepared a tool for machining materials such as aluminium, copper, and stainless steel, using TiC, and WC-Co-based ceramic composite in the FSW process.The composites are fabricated with a powder metallurgy technique, which prevents the damage caused by the temperature developed through friction.They noted that 80 % of the TiCmixed composite electrode produces a fine machining surface in the aluminium work material.Sahu et al. [10] investigated the FSW welding characteristics for magnesium work materials.During the welding process, Zn materials are added to improve the welding quality.Also, the addition of Zn with work materials forms an MgZn alloy, which produces high wear strength.Jimenez-Mena et al. [11] attempted a dissimilar weld with materials (e.g., aluminium, steel) using the FSW process.They reported that interlayering of the work material increases when increasing the load application on the work material.Also, the presence of Co in steel increased the toughness of weld and propagated cracks in the aluminium plate.Garg and Bhattacharya [12] carried the welding on aluminium alloys such as 6061 and 7075 using the FSW process.They investigated the mechanical characteristics such as tensile, flexural, and fracture strength of the welding.Copper particles are used during the welding as interference, and 2 mm tool pin is employed in the tool shoulder.Souza et al. [13] carried out experiments in Al-Ce-Si-Mg aluminium alloy using the FSW process.The maximum tensile stress (102 MPa) noted with lesser tool rotational speed using a triangular pin profile surface and higher micro-hardness found at higher rotational speed.Shettigar et al. [14] conducted the experiments in rutile-reinforced aluminium alloy 6061 composite using the FSW process.They planned the experiments with tool, welding speed, and various tool profiles shapes in order to obtain better machining quality.They conducted the experiment so that rotational, transverse speed contributed more to welding accuracy and homogeneous reinforcement distribution among the welding surface.Khojastehnezhad and Pourasl [15] welded the aluminium alloy 6061 with copper through FSW process along with copper particles.They noted defect-less welding in the rotational speed of 950 rpm and 50 mm/min welding speed.The observation shows the welded composite at the edge has higher hardness due to the aluminum and copper bonding.Babu et al. [16] investigated the optimal process parameter for aluminium alloy 2219 with the FSW process to obtain a defect-free machining surface.They used a genetic algorithm to find the possible parameter combination and suggested that a 1005 rpm tool speed at 3º tool angle produces excellent machining performance.This optimization technique diminishes the machining cost significantly due the shorter machining time.Peddavarapu et al. [17] prepared the composite using Al-4.5Cu based alloy and reinforced it with TiB2 particles; this was employed as a work material to study the performance of the FSW process.Normal treaded profile tool was used as tool.They mentioned that uneven material flow, which produces an uneven circumference in the welding.Herbert et al. [18] conducted the experiments in a Cu-and SiC-reinforced aluminium metal matrix composite, prepared with the stir-casting method and employed in the FSW process to find the optimal welding range of parameters.They used the combined tool profiles (e.g.square and treaded) on the welding and analysed the mechanical and microstructure properties of welding.The hardness of the welding is higher than that of the base metal.Sahu et al. [19] optimized the process parameters of FSW using a fuzzy logic grey analysis method for aluminium and copper welding.They considered the major process parameters, such as tool, welding speed, depth of weld and shape of tool pin.The microstructures of both sides of the metals have fewer defects due to the optimal solution of welding.Ahmadkhaniha et al. [20] carried out the experiments in the FSW process with pure magnesium to find the optimal process parameter combination.They applied the L9 OA and analysed the effect of all parameter combination using the Taguchi method.The harness of magnesium increases with increased tool speed.Shirazi et al. [21] examined the influences of tool and welding speed on the welding strength of aluminium alloy 5456 via the through FSW process.They reported the findings, which showed that, in addition to the optimum parameter, the remaining combinations have a substantial impact on the welding quality.Kumar et al. [22] experimented with aluminium and copper-based composite on the FSW process under various machining speeds and welding speeds.They noted fewer defects with the welding speed in the range of 70 mm/min to 100 mm/min.They also suggest that the microstructure of welding was significantly influenced by the welding and rotational speeds.Suresh et al. [23] studied the influences of various process parameters on the mechanical and microstructure properties of FSW on aluminium alloy 2219 work material using an RSM technique, which was employed to obtain maximum mechanical strength.They found the optimal process parameter combination of1627 rpm rotation speed, 083 mm/s welding speed, and 12.2 kN axial force.
The aforementioned sources were used to explore the optimal solution of FSW process parameters for various aspects, such as aluminium-based composites, copper-reinforced work surfaces, varying concentrations reinforcing particles, and the joining of copper to other metals, such as aluminium, stainless steel, titanium, etc.However, experiments with the copper-based composites and their optimization in the literature are rare [24] to [26] and there is no evidence of hybrid CMC in FSW process.The optimization of the process parameters for every machine is done carefully to achieve the right output without taking major effort.Additionally, optimised values generate high accuracy and reduce unneeded time and cost for quality.In line with that, the TOPSIS and GRA methods are very prominent techniques that are successfully employed in other manufacturing sectors to reveal the optimal solution [27] to [30].The reinforcement materials, such as W and B 4 C particles, provides excellent mechanical strength, thermal stability, and high affinity nature with other materials.Therefore, in this experiment work, material is fabricated with the stir-casting method using commercially available pure copper (Cu) as the matrix and W and B 4 C particles as reinforcements in various concentration levels.Based on L18 OA and major influencing parameters, such as the percentage of reinforcements, rotational speed, welding speed, and axial force are used to conduct experiments.The process parameters of FSW are also optimized using simple and prominent techniques, such as TOPSIS and GRA.Furthermore, SEM analyses are carried out on the welding surface to give better understanding of microstructure.

EXPERIMENTAL WORK
Commercially available pure copper is considered as the base material, W (30µm), B 4 C (50 µm) are employed as reinforcements, and three different CMC materials are fabricated using a stir-casting furnace as shown in Fig. 1.An indigenously created bottompouring and electric-stir casting furnace is employed to develop the composite.The Cu rods (25 mm diameter) are cleaned thoroughly and cut into small pieces 10 mm thick.The rods are loaded into a crucible cylinder that is coated with stainless steel and the temperature is maintained at 1200 ºC during the melting process.The molten Cu is stirred well using a stir setup that enhances the distribution of reinforcements uniformly throughout the composite.Stainless steel coating is carried out in the stirrer and crucible, which protects the unwanted material amalgamation with CMC.The preset volume of preheated W and B 4 C particles is mixed with the molten copper to enhance the wet ability and mechanical strength with copper material.The weight percentages of reinforcement and composition of composite are displayed in Table 1.The well-mixed molten metal is poured into a die (100 mm × 100 mm × 6 mm) to obtain the hybrid composite.
The fabricated CMC plates are employed to study the weldability using the FSW process.The CMC work material in the plate is fastened with a customized vertical milling machine for the FSW process.The work plates are arranged in a butt joint design, and the tool is fixed with the tool holder against the workpiece, as shown in Fig. 1.The non-consumable tool used in the experiment is made of Tungsten carbide with a shoulder diameter of 18 mm.The tool is designed with a hexagon-shape pined with length and diameter of 6 mm and 4 mm, respectively, for the purpose of stirring the work material, as presented in Fig. 2. The rotating tool is plunged over the work material and allowed to rest for 20 seconds to retain its normal temperature.Afterwards, the tool moves over the work material along its length to the weld.Due to the high friction between the tool and electrode, heat is induced, which makes the work materials soften.The shoulder pin of the tool blends the softened work material by moving it from retreating portion to other side which produces a sound weld joint.In this technique, work materials welded using both heat and mechanical energy but in traditional fusion welding only heat energy is employed to weld the metals.The percentage of reinforcements, tool rational speed, welding speed and axial forces influence the welding nature of FSW process.Therefore, these parameters are considered to be input controlling factors.L 18 OA is employed to investigate the process parameter over the output responses.The excellence of welding is evaluated by performing different mechanical and micro-structural testing, including for tensile, impact and hardness.The work materials are sliced into the standard dimensions according to ASTM standard through wire cut EDM.Based on the ASTM E8 standard, test samples are fabricated.Two specimens are cut from every welding specimen and two output responses obtained, which are averaged and considered in order to assess the welding quality.Tables 2 and 3 present the initial parameter levels and design of experiments with output responses respectively, which are considered based on the literature [3].The micro-hardness of welding is measured using a Vickers hardness testing machine (TE-JINANWDW100, Jinan Test Machine Co. Ltd., China) at different places of welding crosssections, and average hardness values are considered for evolution.ASTM E23-16a standards are followed to evaluate the impact strength of welding, and a V-notch has been created in welding specimen at right angles to the welding joint.1.1 Multi Objective Optimization Techniques TOPSIS: TOPSIS is a popular and very powerful technique to separate the correct parametric mixture from the restricted investigated combination.The steps for this technique are scheduled below.[27] Step 1: The conclusion matrix having 'n' characteristics and 'm' option and it is characterized in Eq. (1).
where J ij is the output of i th option with relevance to the j th characteristic.
Step 2: The normalization of the matrix is carried out with Eq. ( 2).
Step 3: For each output, responses have been assigned with equal weights to be Wt j (j = 1, 2, …, n).The standardized weighted choice matrix M = [m ij ] attained using the Eq. ( 3). where Step 4: The suitable best solution is assessed using Eq. ( 4) and the worst solution is attained using Eq.(5). ^`, Step 5: The distributions among every option are intended from the best solutions are obtained using Eq.(6).
The division of option from the worst solution is attained using Eq.(7).
Step 6: The relations closeness of the dissimilar options for the solutions are obtained using Eq. ( 8).Step 7: The k i standards values are graded in downward order to discover the optimal parameters mixture.

Grey Relational Analysis Technique
In GRA, output reactions of different elements should be reformed into the dimensionless values.Therefore, those values are standardized to the variety of zero to one using Eqs.( 9) to ( 12) [28].The tensile, impact, and micro-hardness values must be higher, which considered better and intended using the Eq. ( 9); for lower, the superior is desirable, which is intended using Eq.(10).
Here, ' oi S divergence series is attained from the orientation series O 0 * S and comparability series O i S * .The variety 0 ≤ £ ≤ 1 comprised for the distinctive coefficient: Grey relational grade ( ¥ i ) is added, summing up of grey relational coefficients which corresponds in Eq. (12). is assisted to discover the connection of situation and comparability ideals.

Effect of Reinforcements on the Microstructure of Welding
SEM is employed to investigate the microstructural properties of the welding joints.The macro-size welding images of various reinforced specimens under different machining conditions are analysed.The microstructure of the welding zone and the parent metal of the welding using various reinforced work materials (e.g., 90 % CMC, 85 % CMC and 80 % CMC) are presented in Figs. 3 to 5. The consistency of welding joint on all sides of the parent metal is found to be better in 80 % CMC materials than other reinforced work materials and lesser regularity found in 90 % CMC material.The microstructures of welding zone for all type of reinforced materials are analysed using SEM images.The SEM analysis shows that the dissemination of grains due to high temperatures enlarges the welding zone on the work material, which could be observed as a bright exterior next to the parent metal.The microstructures of various welding surfaces show that associated coarse grains with consistent grain outlines in a welding zone is attained by means of the elevated heat of the FSW process [20].Fig. 3 presents the microstructure of welding zone and its field emission scanning electron microscopy (FESEM) image for 90 % CMC work material.In this, 90 % CMC material creates minor non-homogeneous welding on the parent metal; a small number of micro-voids were produced by over welding.Also, the distribution of welding is found to be higher on this CMC work material than others.Welding through 85 % CMC work material creates micro-cracks in the welds and a few granular fractures were observed, which is presented in Fig. 4. The effective diffusion of B 4 C, W. and Cu produces extended dendrites in the welding zone which lead to micro-crack structures on the welding.Fig. 5 shows 80 % CMC work material, which provides superior consistent welding zone and a very minor amount of slip microstructure obtained across the welding surface [16].White impulsive particles are found above the welding surface, which is due to the dispersion of Cu and W materials into the matrix exterior layer of work material.The elevated carbide content in the reinforcement material indicates the prominent oxidation at elevated temperature, which is leads to the development of molten metal on welding zone.

Micro-hardness
A Vickers micro-hardness test was conducted for various reinforced copper work materials on welding regions; the results of the welding presented in Fig. 6.The graph reveals that the hardness of the 90 % CMC work material is the lowest.The micro-hardness of welding is generally based on the microstructure variations.It is clear from Fig. 6 that the standard hardness of welding formed by 80 % CMC work material is 3.42 % higher than that of welding by 90 % CMC work material.Also, the hardness for 80 % CMC work material at parent material is shown to be higher compared to the welding using 90 % CMC work material, due to the enhancement of carbide and W in the parent metal [20].The standard hardness of 80 % CMC work material is 2.8 % more than that of 90 % CMC work material work material.The welding and parent metal hardness are 108 HV and 113 HV, respectively, when 80 % CMC work material.The

Tensile Properties
The average tensile strength of welding by different CMC materials is shown in Fig. 7.It is observed that the 90 % CMC materials creates 20 N/mm² tensile strength.The tensile strength of 85 % CMC and 80 % CMC materials welding produce 68 N/mm² and 119.2 N/mm² respectively, which is superior to the 90 % CMC materials.It is observed that all wedding samples create the higher tensile strength than 90 % CMC in all factors of welding parameters due to the increasing of reinforcements [31].The cross-sections of welding for different CMC materials are presented in Fig. 8. addition of 15 % B 4 C reinforcements on 90 % CMC metal requires more time to melt than 85 % CMC metals that make the microstructure alteration.It is implicit that standard hardness when using of 85 % CMC creates 2.56 % higher hardness compared to 80 % CMC materials.The alteration of cooling rate in welding zone modifies the crystal metal microstructure headed for an austenitic structure [21].Also, the crystal microstructure is inflated with the increase in temperature, which becomes softer when welding.Hence, the standard hardness of welding was found to be less important than the welding and parent metal excluding 90 % CMC metals welding.

TOPSIS
The outcome responses of FSW process (e.g., hardness, tensile and impact strength) for different CMC work materials are optimized through the TOPSIS method.The preference values of outcome responses are calculated using Eq. ( 1) to (8).The weights for the responses are assigned equally under ideal conditions.The ranking of experimentation is presented in Table 4.The outcome responses of the experimentation are transformed from multi-attribute optimization to single objective optimization using Taguchi's and TOPSIS method combination.The uppermost preference value (k i ) is considered the optimal parameter combination, and the highest rank is termed "first optimal solution".Consequently, it is observed that the 7 th experimental run (0.8376 k i value) is selected as the optimal parameter combination for the finest performance of FSW process.The experimental runs 2 nd and 15 th presents the next suitable optimal parameter solution.Therefore, the suitable optimal parameter combination is found to be that 15 % B 4 C, 1200 rpm rotational speed, 9 mm/min welding speed and 9 kN axial force using TOPSIS.conditions are mathematically studied using ANOVA method, and all individual parameter solution over the outcome responses are investigated.Furthermore, the F-test results are employed to find the most significant parameter to obtain the good performance measure.Table 5 makes apparent that composite work materials, i.e., the percentage of reinforcements with copper material contributes more impact on welding characteristics by around 43.89 %.Subsequently, the significant factor is axial force that produces 32.68 % contribution on the FSW process.

GRA Method
The GRA method is a prominent technique employed to optimize the process parameters of the FSW process, such as micro-hardness, and tensile and impact strength.The GRC and GRG values are derived from Eqs. ( 11) and ( 12) for all experiments.
The weights for the outcome responses are assigned equally (i.e., 1) and GRG values are presented in Table 6.The highest GRG value is selected as optimal parameter solution.Hence, the table suggest that the 7 th experimental run (0.8465) holds the first rank and is termed as the optimal solution for better performance.The experimental combinations 2 nd (0.8461) and 10 th (0.8226) have the next best optimal parameter combination using the GRA method.Therefore, the suitable and optimal parameter combination found to be that 15 % B 4 C, 1200 rpm rotational speed, 9 mm/min welding speed and 9 KN axial force using GRA.

ANOVA for GRG
The results obtained from the GRG for the various parameter combinations are analysed mathematically using ANOVA, which is presented in Table 7.The GRA approach is used to optimize the obtained results for different CMC materials.According to this, the percentage of reinforcements in work materials affects the welding performances significantly at about 37.85 %.The axial force of welding tool creates impact on the welding performance around 33.22 %, which is the next significant factor in the FSW process.

CONCLUSIONS
This paper attempted to express the advantages and performance study of FSW process with copper composite material in reinforcement addition.Three CMC materials are prepared using pure copper (Cu) as the matrix; tungsten (W) and boron carbide (B 4 C) particles are reinforcements for various concentrations.The multi-objective decision-making methods, such as TOPSIS and GRA methods are used to find the optimal parameter combination, and experiments are planned according to the L 18 orthogonal array (OA), using the most influential parameters, such as the percentage of boron carbide reinforcements, tool rotational speed, welding speed, and axial force.
• Based on the optimization results, 15 % of B 4 C, 900 rpm rotational speed, 15 mm/min welding speed and 6 kN axial forces produces the better mechanical strength on the welding using both TOPSIS and GRA techniques.• The microstructure of welding reveals that consistency of welding joint on all sides of the parent metal is found to be better in 15 % B 4 C reinforced CMC materials than other reinforced work materials, and lesser regularity found in 5 % B 4 C reinforced CMC material.• 80 % CMC and 85 % CMC reinforced copper work metal produces 68 % and 10 % higher impact toughness respectively than that using of welding on 90%CMC work metal.• Based on the ANOVA table of TOPSIS, composite work materials, the percentage of reinforcements with copper material contributes more impact on welding characteristics of around 43.89 %.Subsequently, the next significant factor is axial force, which produces a 32.68 % contribution on the FSW process.• Therefore, the mechanical strength of welding with FSW process increases with an increase of the percentage of reinforcement in the copper composite material.Also, these types of materials could be used for the applications for which high mechanical strength is required.• Furthermore, experiments with hybrid composite with pure copper can be conducted, and assistance with FSW process such as heat energy can be experimentally tested with the FSW process to enhance the welding quality.

Fig. 2 .
Fig. 2. Friction stir welding; a) experimental setup, b) tool, and c) work piece i = 1, ,2, …, m, s = 1, 2, …, n, where m is the sum of experiment, n is the sum of the observed data.smaller the better.The standardized values are added in Eq.(11), which is employed to compute the grey relational coefficient (GRC).

Table 1 .
Composition of CMC

Table 2 .
Input parameter and its levels

Table 3 .
Design of experiment and output responses

Table 7 .
ANOVA table for GRG