4.TAGUCHI ANALYSIS FOR THE EXPERIMENTAL RESULTS To examine the 100Cr6 bearing steel through DCT, and to identify the optimum parameter condition of the DCT process, the prediction of the Taguchi analysis is used. The Taguchi analysis and ANOVA, gives the optimized parameter setting and the most influential factor of the DCT process respectively.
The quality performance of the deep cryo-treated 100Cr6 steel samples based on the L9 OA was evaluated by conducting the reciprocatory wear tests. Three replications were taken in each test.
4.1 Signal-to-Noise Ratio: Taguchi recommends the use of the S/N ratio to measure the quality characteristic deviating from the desired values. The ratio measures the level of performance against the level of noise factor on performance. The method of modeling the S/N ratio depends on whether the quality characteristic is smaller-the-better, larger-the-better, or nominal-the-best. The response values from the three replicates were transformed into S/N ratios. The quality characteristics selected in this study for better performance of the 100Cr6 bearing steel is wear (should be the lower-the-better). The expressions to calculate the S/N ratio the lower-the-better characteristic is a non-negative measurable characteristic that has an ideal state value of zero. The wear loss values obtained for each consecutive experiment is used for S/N ratio calculation. Minimization of the quality characteristics (the wear loss) can be expressed as S/N ratio=-10 log_10〖[1/r〗 ∑_(i=1)^r▒y_ij^2 ] (4.1) Where r is the number of replications and yij is observed response value where i=1, 2….n, j=1, 2….k; k is the number of experiments. The computed S/N ratios using the above equations for each quality characteristic of the DCT and measured wear weight loss are furnished in Table 4. Table 4. Measured Value of Wear Weight Loss and S/N Ratio Exp. No Wear Weight Loss (mg) S/N Ratio Y1 Y2 Y3 AVG Y 1 1.21 1.34 1.44 1.33 -2.498 2 1.05 1.41 1.19 1.22 -1.767 3 1.35 1.32 1.26 1.31 -2.348 4 1.06 1.12 1.05 1.08 -0.645 5 0.88 0.94 0.92 0.91 0.784 6 1.58 1.6 1.73 1.64 -4.286 7 0.93 1.01 1.08 1.01 -0.073 8 1.01 1.45 1.34 1.27 -2.146 9 1.35 1.21 1.42 1.33 -2.474 Fig 2 Average Wear Loss for each Experiment Figure 1 shows the average wear loss for the 9 experiments for L9 (34) OA .The minimum wear weight loss occurs at fifth experiment. This briefs that 5th experiment is the optimal set of parameters for DCT process. Response value for each parameter is calculated by taking average of S/N ratio values for levels in each column. The response value of level 1 in the first column of OA should be averaged (i.e., experiments 1-3), then response value of level 2 in the first column of OA should be averaged (i.e., experiments 4-6) and the response value of level 3 in the first column of OA should be averaged (i.e., experiments 7-9). The above computed response values are for process parameter cooling …show more content…
For the cooling rate level 2 has the highest value therefore A2, similarly for soaking temperature B2, soaking period C3 and tempering temperature D1 are having the highest response values. Rank for each process parameters is applied accordance higher values is better response characteristic of the DCT level in Table 5. Rank is given to each process parameter to find the most influencing factor among the four parameters. Hence soaking period has the highest response value of -0.564 compared other response values. Likewise the response values are compared with other parameters and ranks are …show more content…
The peak values in the graph are considered as the optimal values. Rank 1 shows that soaking period is the most influencing factor.
4.2 Anova The goal of the ANOVA is to estimate and test the effect of different treatments on the response variables. In the present work, the ANOVA is used to investigate, which DCT process parameters significantly influence the performance characteristics, among the factors, namely, the cooling rate, soaking temperature, soaking time and tempering temperature, for the multi response values of wear resistance. Using the ANOVA, the influence of the DCT parameters on the quality targets can be examined. The percentage contribution of the variance can be calculated using the following equation. Correction factor CF=1/m [∑_(i=1)^m▒y_i^2 ]2
The quality performance of the deep cryo-treated 100Cr6 steel samples based on the L9 OA was evaluated by conducting the reciprocatory wear tests. Three replications were taken in each test.
4.1 Signal-to-Noise Ratio: Taguchi recommends the use of the S/N ratio to measure the quality characteristic deviating from the desired values. The ratio measures the level of performance against the level of noise factor on performance. The method of modeling the S/N ratio depends on whether the quality characteristic is smaller-the-better, larger-the-better, or nominal-the-best. The response values from the three replicates were transformed into S/N ratios. The quality characteristics selected in this study for better performance of the 100Cr6 bearing steel is wear (should be the lower-the-better). The expressions to calculate the S/N ratio the lower-the-better characteristic is a non-negative measurable characteristic that has an ideal state value of zero. The wear loss values obtained for each consecutive experiment is used for S/N ratio calculation. Minimization of the quality characteristics (the wear loss) can be expressed as S/N ratio=-10 log_10〖[1/r〗 ∑_(i=1)^r▒y_ij^2 ] (4.1) Where r is the number of replications and yij is observed response value where i=1, 2….n, j=1, 2….k; k is the number of experiments. The computed S/N ratios using the above equations for each quality characteristic of the DCT and measured wear weight loss are furnished in Table 4. Table 4. Measured Value of Wear Weight Loss and S/N Ratio Exp. No Wear Weight Loss (mg) S/N Ratio Y1 Y2 Y3 AVG Y 1 1.21 1.34 1.44 1.33 -2.498 2 1.05 1.41 1.19 1.22 -1.767 3 1.35 1.32 1.26 1.31 -2.348 4 1.06 1.12 1.05 1.08 -0.645 5 0.88 0.94 0.92 0.91 0.784 6 1.58 1.6 1.73 1.64 -4.286 7 0.93 1.01 1.08 1.01 -0.073 8 1.01 1.45 1.34 1.27 -2.146 9 1.35 1.21 1.42 1.33 -2.474 Fig 2 Average Wear Loss for each Experiment Figure 1 shows the average wear loss for the 9 experiments for L9 (34) OA .The minimum wear weight loss occurs at fifth experiment. This briefs that 5th experiment is the optimal set of parameters for DCT process. Response value for each parameter is calculated by taking average of S/N ratio values for levels in each column. The response value of level 1 in the first column of OA should be averaged (i.e., experiments 1-3), then response value of level 2 in the first column of OA should be averaged (i.e., experiments 4-6) and the response value of level 3 in the first column of OA should be averaged (i.e., experiments 7-9). The above computed response values are for process parameter cooling …show more content…
For the cooling rate level 2 has the highest value therefore A2, similarly for soaking temperature B2, soaking period C3 and tempering temperature D1 are having the highest response values. Rank for each process parameters is applied accordance higher values is better response characteristic of the DCT level in Table 5. Rank is given to each process parameter to find the most influencing factor among the four parameters. Hence soaking period has the highest response value of -0.564 compared other response values. Likewise the response values are compared with other parameters and ranks are …show more content…
The peak values in the graph are considered as the optimal values. Rank 1 shows that soaking period is the most influencing factor.
4.2 Anova The goal of the ANOVA is to estimate and test the effect of different treatments on the response variables. In the present work, the ANOVA is used to investigate, which DCT process parameters significantly influence the performance characteristics, among the factors, namely, the cooling rate, soaking temperature, soaking time and tempering temperature, for the multi response values of wear resistance. Using the ANOVA, the influence of the DCT parameters on the quality targets can be examined. The percentage contribution of the variance can be calculated using the following equation. Correction factor CF=1/m [∑_(i=1)^m▒y_i^2 ]2