Abstract
Industry 4.0 industries are relying on automation processes using robots. Robots are multifunctional reprogrammable machines tuned to be used in any process with high accuracy and repeatability. The advent of intelligent technology allows better precision in the welding process. One such robotic welding process is the robot microfriction welding process which is slowly replacing conventional ones. Robot microfriction stir welding (RMFSW) process mainly relays on friction between the tool and material making it more suitable for joining metals in the industries such as automotive and aerospace applications. The robot-assisted MFSW process can provide better joints without human intervention. In this work, a shrinking flange has been designed and fabricated to grip the end effector with the robot. A microfriction stir welding tool was manufactured using EN24 and used to weld 1 mm thin aluminium 1100 sheets. Both the flange and MFSW tools were designed in SOLIDWORKS software. Taguchi L9 was designed with three factors such as motor speed, traverse speed, and plunging depth. The robot microfriction stirs welded samples had controlled bead width and depth of penetration. Mechanical results show an improvement in hardness after the welding process. TOPSIS optimization technique was carried out. The motor speed of 20,000 rpm, traverse speed of 2 mm/sec, and plunge depth of 0.7 mm were found to be the best-optimized parameters.
1. Introduction
One of the unique challenges in producing compact products is the making of small and microcomponents. Especially, joining materials of 1000 μm or below it is a difficult process. This microjoining process can be achieved through highly accurate and precise automated robots. Micromanufacturing using industrial robots is possible by designing a special-purpose gripper tool for a particular process. One such microjoining process is microfriction stir welding (MFSW), which was developed in the welding institute by Wayne Thomas. It is an eco-friendly and solid-state welding process [1]. Solid-state welding is a process of joining two metals or alloys by producing friction between the tool and the base materials. The same method is followed in microfriction stir welding where the dimensions are reduced to a micron level. A tool made of a solid rod tougher than the base material is used in microfriction stir welding. The hard MFSW tool is made of a pin and shoulder that is to be rotated at greater velocity and simultaneously moved through metals to join it by creating moderate heat and friction [2, 3]. Since MFSW is a solid-state process based on friction it produces low-level heat compared to other welding processes. It produces good-appearance welds with low distortion, at a relatively low cost. The only disadvantage of this technique has been its exit keyhole. MFSW is very useful in applications such as welding of complicated products like thin-walled structures, battery covers, and electrical and electronic components [4–6]. The μFSW process has four phases which is conducted in the order mentioned in Figure 1. Some of the principal process parameters of microfriction stir welding are the rotational speed of the tool, forward velocity of the tool and plunge depth. The torque or rotational force of the tool decides the frictional force between the tool and the material which transports the metal around the tool in the plastic zone. This parameter is important to produce the friction at the shoulder of the tool which gives the welding finish [7–9]. Traverse movement of the tool is essential to make the material move around the tool and move it forward. Plunge depth is decided by the thickness of the material. For example, for a sheet with a 1 mm thickness, the maximum plunge depth can be up to 0.9 mm [10]. It is very difficult to carry out fusion welding in a 1 mm sheet due to metallurgical problems and distortion. MFSW is utilized to carry out welding without distortion [11–13]. The disadvantage of the MFSW process is that it has a keyhole at the last stage, this technique is very scale-sensitive so high precision and accuracy are needed. The major advantage in the MFSW process compared to microresistance welding is that the joints are made at lower temperatures without any need of fluxes, gas, or consumable electrodes automatically reducing the risk of contamination or cleaning delicate components. This avoids the need of postwelding treatments which consequently reduces the welding cost. Mahidhar et al. has done TOPSIS optimization for laser-welded haste alloys [14].
The objective of this work is to carry out RMFSW process in aluminum 1100 using ABB IRB 1410 robot. In order to achieve the process a motor with high torque capacity is used as an end effector for the robot. A microfriction stir welding tool has been designed and manufactured in EN24 material. In this project, we are designing and manufacturing a special-purpose shrinking flange to grip the Makita RT0700C alternating current motor to the ABB IRB 1410 Project plan is shown in Figure 1. The aluminum 1100 is welded using RMFSW. The optimization of process parameters of the RMFSW was done using TOPSIS optimization.
2. Experimental Details
Aluminum 1100 alloy of 1 mm thin sheets with a dimension of mm was used throughout this study. The composition of Aluminium 1100 alloy is shown in Table 1.
2.1. Design of the Robotic Gripper
A special type of shrinking flange has been designed in SOLIDWORKS software and manufactured in H30 material to withhold the Makita RT0700C motor. Microfriction stir welding tool has been designed in SOLIDWORKS software and manufactured using EN24 material.
2.2. Design of Backing Plate
The backing plate was designed in SOLIDWORKS to use in any type of welding process. It is manufactured in mild steel material and the whole setup is coated with blackening powder to prevent rust. The model is designed to act as a backing block and also as a shielding gas provider below the sheets. Turning it upside down it also can be used for RFSW process. The SOLIDWORKS design of the Backing plate is presented and the manufacturing block is presented in Figure 2.
(a)
(b)
2.3. Design of Microfriction Stir Welding Tool
The tool was designed in SOLIDWORKS software and the tool material was decided by considering workpiece material. In this work, aluminium 1100 alloy was chosen as a work piece material which has a hardness value of 144 HB. MFSW tool was selected based on a higher hardness value than the base material since our aluminum is a soft material. EN24 was chosen as MFSW tool with a hardness value of 248 HB. The tool was designed by considering the bending moment in the account so to reduce it a step-by-step cut design is used (Figure 3).
(a)
(b)
ABB IRB 1410 robot was used in this project to do the microfriction stir welding process with aid of Makita RT0700C AC motor and specially designed microfriction stir welding tool. The welding setup is shown in Figure 4.
The robot MFSW was done based on the L9 Taguchi array by considering three levels of factors named as M = Motor speed (RPM), T = Traverse speed (mm/Sec) and = Plunging depth (mm). Table 2 shows three levels of parameters with three factors.
The L9 Taguchi design of the experiments table with both input and output parameters is mentioned in Table 3, where DOP, is the depth of penetration, and BW is bead width.
The robot microfriction stir welded (RMFSW) 1 mm thin sheets of aluminium 1100 alloy sheets are shown in Figure 5.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
After the RMFSW process, the welded specimen were cut down into 10 × 10 mm samples using the wire-EDM method and then cold-mounted. The cold-mounted samples were polished using emery sheets and the disc was polished using diamond paste. Further, the samples were etched using Keller’s reagent [150 ml H2O, 3 ml HNO3, and 6 ml HF] for 20 seconds. Macrostructure was taken using a machine vision system and microhardness testing was carried out in a micro-Vickers hardness tester with the model number of HMV-G made by Shimadzu.
3. Results and Discussion
3.1. Macrostructure
The macrostructure of the welded samples is taken from the machine vision system. BW and DOP of the samples are measured using VMS 3.1 software. Measured values are tabulated below. All the samples have achieved a bead width of 3.5 mm approximately and the samples have achieved the applied depth of penetration with a variation of ±0.02 mm. Sample images are revealed in Figure 6 and (see Tables 4 and 5).
(a)
(b)
3.2. Microhardness
Microhardness test was carried out in Vickers hardness tester. The values are taken in the interval of 10 mm on either side of the weld zone and on the weld zone. The average microhardness values of the welded samples are tabulated below. Samples 7, 8, and 9 had the highest hardness of 43 HV.
3.3. Optimization (TOPSIS)
TOPSIS stands to be a technique for order preference by similarity to the ideal solution which is a common technique that has been used in automobile industries for decision making. This technique has six stages as follows: Stage I: Determination of a Performance Matrix Stage II: Calculation of Normalized Decision Matrix Stage III: Calculate Weighted Normalized Decision Matrix Stage IV: Determination of the Ideal (Di+) and (Di−) Stage V: Calculate the Euclidian Distance between Ideal (Di+) and Ideal (Di−) Stage VI: Determine the Relative Performance Closeness to the Ideal Soln.
In this project, depth of penetration and hardness is considered to be larger the better criteria whereas bead width is considered to be smaller the better criteria. Stage I: Determination of a Performance Matrix Calculate the sum of squares for DOP, BW, and microhardness to calculate the performance matrix. The performance matrix of the above-mentioned parameters is shown in Table 6. Stage II: Calculation of Normalized Decision Matrix Equation (1) is used to calculate Normalized Decision Matrix. Table 7 shows the corresponding values. Stage III: Calculate Weighted Normalized Decision Matrix Weights are allocated to output parameters based on optimization techniques. Weights are distributed equally in percentage to the input parameters. So, 0.35, 0.3, and 0.35 has been selected as equal percentage weight for the input parameters. Corresponding values are mentioned in Table 8 and the matrix is shown in Table 9. We can calculate the weighted normalized matrix, Stage IV: Determination of the Ideal (Di+) and (Di−) Table 10 reveals ideal worst and best numbers. Ideal Positive Soln. = (min or max). Ideal Negative Soln. = (min or max). Stage V: Calculate the Euclidian Distance between Ideal (Di+) and Ideal (Di−) Euclidian Distance values calculated using equations (3) and (4) are shown in Table 11. Stage VI: Determine the Relative Performance Closeness to the Ideal Soln.
Table 12 reveals relative performance closeness and its rank.
Figure 7 shows graphical representation of Pi (closeness value) vs. L9 samples.
3.3.1. Response Value Table Using Closeness Value
A3 B2 C3 was found to be the best optimal parameter in the Taguchi L9. Motor speed had the highest influence on input parameters followed by plunge depth and traverse speed, respectively. Figure 8 shows the best optimal level of the parameter (see Table 13).
The mean line was found around 15.5 above this point all the input good results but our optimal level was found to be out of L9 so a confirmatory test was done.
3.3.2. Confirmation Test
In order to find the optimized parameter set Pi value was predicted using the following equation:where ℎm average of Pi values, ℎ is the mean of Pi values at the optimum level. The optimized parameters are presented in Table 14.
The experiment was done again using predicted values and the optimal parameter combination is found with DOP = 0.7 mm; BW = 3.5 mm; and hardness = 43.3 HV. Figure 9 shows the microstructure of the optimized parameter.
4. Conclusion
(1)Robot microfriction stir welding (RMFSW) of 1 mm thin aluminum 1100 sheets was successfully done using ABB IRB 1410 robot.(2)L9 Taguchi was designed with three factors and three levels.(3)In microstructure analysis, the weld zone had a fine grain structure, which increases the hardness value of the material.(4)Microhardness seemed to be higher after the welding process which has increased up to 43.3 HV.(5)In TOPSIS sample number 7 had the highest rank compared to all other samples. Sample 7 has input parameters such as motor speed = 20000 rpm; traverse speed = 1 mm/sec; and depth of penetration = 0.7 mm. It had a dop of 0.7 mm; a bead width of 3.45; and a hardness of 43 HV.(6)A confirmation test was carried out and A3B2C3 was found to be the best optimal input parameter.(7)Motor speed is found to be the most influencing parameter among all three, succeeded by plunge depth and traverse speed, respectively.Data Availability
The data used to support the findings of this study are made available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.