52 CHAPTER 4 OPTIMIZATION OF PROCESS PARAMETER OF FSW ON AA 5083 USING TAGUCHI METHOD (SINGLE OBJECTIVE) 4.1 INTRODUCTION Welding is one of the vital and widely used manufacturing processes in any manufacturing industries. The welding technology has grown substantially in aerospace and shipping industries with a common goal of achieving higher strength and weld joint efficiency. Selection of optimal welding conditions is a key factor in achieving this condition. In order to obtain a high quality weld the manufacturer has to set the input controllable factors at their optimum levels, with the minimum effect of uncontrollable or noise variables on the levels and the variability of the responses. This chapter presents the Taguchi approach to optimize the process parameters in Friction Stir welding of Aluminium Alloy AA 5083.The Rotational speed of the tool, Axial load and the Transverse speed are the most significant parameters taken into consideration. The optimum levels of the process parameters are determined. The predicted optimal value of tensile strength is confirmed by conducting the experiment using optimum parameters 4.2 TAGUCHI METHOD Taguchi, a Japanese quality engineer, widely recognized as the Father of Quality Engineering (Taguchi 1996) addresses quality in two main
53 areas: off-line and on-line quality control. Both of these areas are very cost sensitive in the decisions that are made with respect to the activities in each. Offline quality control refers to the improvement in quality in the product and process development stages. On-line quality control refers to the monitoring of current manufacturing processes to verify the quality levels produced (Ross 1996). The most important difference between a classical experimental design and a Taguchi-method-based robust design technique is that the former tends to focus solely on the mean of the quality characteristic while the later considers the minimization of the variance of the characteristic of interest. Although Taguchi method has drawn much criticism due to several major limitations, it has been able to solve single response problems effectively. Taguchi method attempts to optimize a process or product design and is based on three stages: Concept Design or System Design Parameter Design Tolerance Design. The following are the steps to be followed for process parameter optimization (Datta et al. 2008): Step 1: Determine the quality characteristic to be optimized Step 2: Identify the noise factors and test conditions Step 3: Identify the control factors and their alternative levels Step 4: Design the matrix experiment and define the data analysis procedure. Step 5: Conduct the matrix experiment Step 6: Analyze the data and determine optimum levels for control factors. Step 7: Predict the performance at these levels.
54 4.2.1 Selection of orthogonal array (OA) The selection of orthogonal array to use predominately depends on these items in order of priority: The number of factors and interactions of interest The number of levels for the factors of interest The desired experimental resolution or cost limitations As three levels and three factors are taken into consideration, L9 OA is used in this investigation. Only the main factor effects are taken into consideration and not the interactions. The degrees of freedom (DoF) for each factor is 2 (No. of levels -1, i.e. 3-1=2) and therefore the total degrees of freedom will be 3 2=6. Generally the DoF of the OA should be greater than the total DoF of the factors. As the DoF of L9 is 8, it can be suitable for the study. 4.2.2 Levels of process parameters It has been clearly shown in the literature (Lakshminarayanan et al. 2008, Fujii et al. 2006, Peel et al. 2003, Hirata et al. 2007 and Cui et al. 2008) that FSW process parameters such as rotational speed, transverse speed and axial load significantly influence the process and play a major role in deciding the quality of the weld. When the rotational speed is decreased to 600 rpm all the tools (triangular prism, cylindrical with and without threads) produce high tensile strength, and the dependence of the tool shape is not significant (Fujii et al. 2006). A tapered cylindrical column tool without threads will significantly affect the mechanical properties of the joints at higher rotational speed. Therefore, the rotation speed should be decreased (600 rpm) so that plastic deformation of the material could be transported from the front to the
55 back of the tool effectively (Fujii et al. 2006). A high stress intensity can be developed at the pin shoulder corner, when the transverse speed and axial load increase. As the tool pins are heat treated to VHN 50, the tools are comparatively harder and brittle, resulting in fracture of the tool pin (Chen et al. 2008).The axial load of 8.240 kn has been reported to be the optimal value for some of the aluminium alloys (Leal and Loureiro 2008). When the welding speed is 80mm/min the tensile strength of AA 5083 is considerably low (Ancona et al. 2007). Selective vaporization of volatile constituents is probably responsible for this behaviour, since lower the welding speed, more the magnesium loss from the weld zone (Zhao et al. 1999). In this investigation the three parameters such as rotational speed, transverse speed and axial load are taken into consideration, the values and their levels are listed in Table 4.1. Table 4.1 Process parameters values and their levels Level Rotational speed Transverse speed Axial load (rpm) (mm/min) (kn) Level 1 500 115 9 Level 2 650 135 13 Level 3 800 155 17 4.2.3 Experimental procedure The experiments and tensile testing of the specimens were carried out as per the sessions 3.3 and 3.4 mentioned in the 3 rd chapter. The experiments were conducted according to the designed L9 Orthogonal Array and the values are listed in Table 4.2.
56 Table 4.2 Experimental layout using L9 orthogonal array for tensile strength Sl.No Rotational speed Transverse Speed Axial load Tensile Strength Mpa 1 1 1 1 191 186 188 2 1 2 2 171 175 169 3 1 3 3 147 154 146 4 2 1 2 287 280 279 5 2 2 3 295 280 284 6 2 3 1 228 231 225 7 3 1 3 157 156 156 8 3 2 1 127 129 131 9 3 3 2 130 135 136 4.2.4 Results and discussion 4.2.4.1 Signal to noise ratio The S/N Ratio is calculated based on the quality of the characteristics intended. The objective function described in this investigation is maximization of the tensile strength so, the larger the best S/N ratio is to be calculated. The formula used for calculating the S/N ratio is given below Larger the Best -10 log 10 1 n 1 2 n i 1 yi (4.1) where n - number of replications y i - observed response value
57 The tensile strength of the Friction Stir welding joints values is analyzed to study the effects of the FSW process parameters. The experimental data are converted into mean and S/N ratio. The calculated mean and S/N ratio values are tabulated in Table 4.3. The main effects, average mean and S/N ratio values of all levels are calculated and listed in Tables 4.4 to 4.6. Irrespective of the objective function whether maximization or minimization the larger S/N Ratio corresponds to the better quality characteristics. Based on both mean and S/N ratio values the optimal level setting is RS 2 TS 1 AL 3 i.e., the rotational speed is to set at 650 rpm, the transverse speed has to be set at 115mm/min and the axial load is to be 17 kn based on the experimental results. Table 4.3 Mean value and S/N ratio for tensile strength Sl. No. Input Parameter Response Mean RS TS AL T1 T2 T3 value S/N Ratio 1 500 115 9 191 186 188 188.3 45.4970 2 500 135 13 171 175 169 171.667 44.6910 3 500 155 17 147 154 146 149.0 43.4564 4 650 115 13 287 280 279 282.0 49.0029 5 650 135 17 295 280 284 286.333 49.1311 6 650 155 9 228 231 225 228.0 47.1572 7 800 115 17 157 156 156 156.333 43.8809 8 800 135 9 127 129 131 129.0 42.2097 9 800 155 13 130 135 136 133.667 42.5154
58 Table 4.4 Main effects of the process parameters on tensile strength Process Parameter Average value Main effects Level Mean S/N Ratio A(RS) B(TS) C(AL) A(RS) B(TS) C(AL) L1 169.7 211.7 182.0 44.54 46.23 44.95 L2 270.0 197.7 196.0 48.57 45.38 45.37 L3 138.0 168.3 199.7 42.76 44.26 45.55 L2-L1 100.3-14 14 4.03-0.85 0.42 L3-L2 132-29.4 3.7-5.81-1.12 0.18 Table 4.5 Response table for tensile strength signal to noise ratio Level RS TS AL 1 44.54 46.23 44.95 2 48.57 45.38 45.37 3 42.76 44.26 45.55 Delta 5.81 1.97 0.60 Rank 1 2 3 The optimal setting is RS 2 TS 1 AL 3 based on S/N ratio. Table 4.6 Response table for tensile strength mean Level RS TS AL 1 169.7 211.7 182.0 2 270.0 197.7 196.0 3 138.0 168.3 199.7 Delta 132 43.3 17.7 Rank 1 2 3 The optimal setting is RS 2 TS 1 AL 3 based on Mean.
59 4.2.4.2 Analysis of variance (ANOVA) The purpose of ANOVA is to find the significant factor statistically. It gives a clear picture of how far the process parameter affects the response and the level of significance of the factor considered. The ANOVA table for both mean and S/N ratios are calculated and listed in Tables 4.7 and 4.8. The main effects for mean and S/N ratio are plotted in Figures 4.1 and 4.2. The F test is being carried out to study the significances of the process parameter. The high F value indicates that the factor is highly significant in affecting the response of the process. In our investigation, for the material AA 5083 the rotational speed is a highly significant factor and plays a major role in affecting the tensile strength of the weld. The effect of axial force does not make any impact in the responses. Table 4.7 Analysis of variance for tensile strength means Source DoF Seq SS Adj SS Adj MS F % Contribution RS 2 25893.4 25893.4 12946.7 45.78 88.6363 TS 2 2317.4 2317.4 1158.7 4.10 7.9327 AL 2 436.6 436.6 218.3 0.77 1.4945 Residual Error 2 565.7 565.7 282.8 1.9365 Total 8 29213.1 29213.1 100 DoF - Degrees of freedom, Seq SS- Sequential sum of squares, Adj SS- Adjusted sum of square, Adj MS- Adjusted mean square, F- Fisher ratio.
60 Table 4.8 Analysis of variance for tensile strength S/N ratio Source DoF Seq SS Adj SS Adj MS F % Contribution RS 2 14.584 14.584 7.292 1.83 45.67777 TS 2 4.848 4.848 2.424 0.61 15.18416 AL 2 4.544 4.544 2.272 0.57 14.23202 Residual Error 2 7.952 7.952 3.976 24.90604 Total 8 31.928 31.928 100 DoF - Degrees of freedom, Seq SS- Sequential sum of squares, Adj SS- Adjusted sum of square, Adj MS- Adjusted mean square, F- Fisher ratio. Main Effects Plot (data means) for Means RT TS 250 225 200 175 150 500 650 AF 800 115 135 155 250 225 200 175 150 9 13 17 Figure 4.1 Main effects plot for tensile strength mean
61 Main Effects Plot (data means) for SN ratios RT TS 48 46 44 500 650 AF 800 115 135 155 48 46 44 9 13 Signal-to-noise: Larger is better 17 Figure 4.2 Main effects plot for tensile strength S/N ratio 4.2.4.3 Predicted value of tensile strength Based on the experiments, the optimum level setting is RT 2 TS 1 AL 3. The additive model to evaluate the predicted tensile strength is taken from the literature (Lakshminarayanan and Balasubramanian 2008). The average values of the factors at their levels are taken from Table 7.5 and the predicted value of the response is given below: Tensile strength (predicted) = RT 2+TS1+AL3-2T = 270+211.7+199.7-2(191.5889) = 298.222 MPa (4.1) where RT2 - Average mean value of rotational speed at 2 nd level TS1 - Average mean value of transverse speed at 1 st level AL3 - Average mean value of axial force at 3 rd level T - Overall mean
62 4.3 CONFIRMATION RUN The confirmation experiments were carried out by setting the process parameter at optimum levels. The rotational speed, transverse speed, and axial load were set at 650 rpm, 115 mm/min and 17 kn respectively. Three tensile specimens were subjected to tensile test and the average value of the Friction Stir welded AA 5083 was 301Mpa. 4.4 SUMMARY In this chapter, the process parameter of Friction Stir welding is optimized using Taguchi method. From this investigation, following important conclusions are drawn. The L9 Taguchi orthogonal designed experiments of Friction Stir Welding on aluminium alloy AA 5083 were successfully conducted. The FSW process parameters were optimized to maximize the tensile strength of the joint. The optimum levels of the rotational speed, transverse speed, and axial load were found to be 650 rpm, 115 mm/min and 17 kn, respectively, for cylindrical tapered column tool without threads. The rotational speed plays a vital role, and contributes 88.64% to the overall response. The axial load does not affect the response significantly.