OBRABOTKAMETALLOV technology Vol. 27 No. 2 2025 “Larger-is-Better” LB S N 2 1 1 1 ( / ) 10 log R j R y (3) “Smaller-is-Better” SB S N 2 1 1 ( / ) 10 log R j y R (4) “Nominal-is-Best” NB S N 2 0 1 1 ( / ) 10 log ( ) R j j y y R (5) where yi is the the value of the parameter obtained in the i-th trial; R is the the number of repetitions of the trial; μ is the mean value of the data; σ is the standard deviation of the data. In the Taguchi experimental design, an L18 orthogonal array was used, selected based on the number of process parameters and their defined levels. The design included five parameters: workpiece electrical conductivity, gap current, gap voltage, pulse-on time, and pulse-off time. One variable (workpiece electrical conductivity) was varied at six levels, and the remaining four were varied at three levels. These parameters are designated as A, B, C, D, and E. Table 3 presents the process parameters and corresponding levels used in the experiments. The Taguchi method requires the calculation of degrees of freedom (DoF) to select a suitable orthogonal array for design of experiments. The workpiece material electrical conductivity, having six measurement levels, determines five degrees of freedom. Each of the remaining four parameters (gap current, gap voltage, pulse-on time, and pulse-off time), varied at three levels, has two degrees of freedom per variable. Therefore, the total number of DoF is 13. Based on this, a mixed orthogonal array L18 (61 × 34) was chosen as satisfying the criterion of possessing seventeen degrees of freedom. The structure of the L18 array is presented in Table 4. The experiments were conducted in accordance with the Taguchi L18 orthogonal array methodology. Two key principles of design of experiments (DoE) were implemented in this study. First, to enhance the statistical reliability of the results, the principle of replication was used, involving conducting multiple repeated measurements for each parameter set. This allows for improved accuracy in the estimation of main effects and their interactions, as well as a proper assessment of experimental error. In this study, three repeated measurements were conducted for each parameter combination. Second, data were collected for each experimental condition. Based on the obtained data, the signal-to-noise (S/N) ratio was calculated for each experimental condition using equations (3)–(5), according to the selected quality characteristics (MRR, TWR, and SR). Analysis of variance (ANOVA) was used to determine the significance of the influence of various EDM process Ta b l e 3 Process parameters and its levels Parameters Code Levels Electrical conductivity of workpiece (S/m) A NiTi NiCu BeCu 3268 (untreated) 4219 (treated) 5515 (untreated) 5625 (treated) 5645 (untreated) 5902 (treated) Gap current (A) B 8 12 16 – – – Gap voltage (V) C 40 55 70 – – – Pulse on time (µs) D 13 26 38 – – – Pulse off time (µs) E 5 7 9 – – –
RkJQdWJsaXNoZXIy MTk0ODM1