In this study, we conducted a turning experiment and defined the input machine variables such as vc, f, and ap, as well as output criteria such as surface roughness (SR), vibration acceleration (VA), and tool life (TL), in order to optimise the process parameter. We employed three MCDM models, TOPSIS, CODAS, and EDAS, to rank the options with regard to the given criteria in this study in order to find the best process parameter. As previously stated, the weights of the criterion are determined using the Entropy technique. We utilise the decision matrix described in the Entropy technique and then normalise decision matrix (rij) it is using Eq. (1). We next use Eq. (2) to determine the entropy of each criterion, which is represented by Ej. Using Eq. (3), we derive the weights of each criterion, indicated by Wj. The weightage of each criterion Wj for SR, VA and TL is 0.154, 0.557 and 0.289 respectively.
The normalised matrix values for different TOPSIS process parameters were determined using Eq. (9), and the normalised values are shown in Table 3. By considering performance aspects: Lower the better (LB) is preferable for SR and VA, higher is better for TL. We may generate a weighted normalised matrix by multiplying the entropy weighting value with Eq. 9. We get weighted normalised matrix (\({V_{ij}}\)).
Table 3
Closeness coefficient values of TOPSIS analysis
Exp.No | Normalized Decision Matrix | Weighted Normalized Decision Matrix | Di+ | Di− | Ci | Rank |
| SR | VA | TL | SR | VA | TL | | | | |
| mm | m/s2 | min | mm | m/s2 | min | | | | |
1 | 0.176 | 0.123 | 0.222 | 0.027 | 0.068 | 0.064 | 0.037 | 0.153 | 0.807 | 5 |
2 | 0.224 | 0.202 | 0.196 | 0.035 | 0.113 | 0.057 | 0.072 | 0.108 | 0.599 | 9 |
3 | 0.333 | 0.370 | 0.157 | 0.051 | 0.206 | 0.045 | 0.162 | 0.013 | 0.072 | 17 |
4 | 0.186 | 0.148 | 0.157 | 0.029 | 0.082 | 0.045 | 0.060 | 0.138 | 0.698 | 7 |
5 | 0.237 | 0.243 | 0.183 | 0.037 | 0.135 | 0.053 | 0.093 | 0.085 | 0.477 | 12 |
6 | 0.333 | 0.391 | 0.157 | 0.051 | 0.218 | 0.045 | 0.174 | 0.004 | 0.021 | 18 |
7 | 0.231 | 0.225 | 0.157 | 0.036 | 0.125 | 0.045 | 0.089 | 0.094 | 0.514 | 11 |
8 | 0.261 | 0.343 | 0.157 | 0.040 | 0.191 | 0.045 | 0.147 | 0.029 | 0.167 | 16 |
9 | 0.286 | 0.256 | 0.144 | 0.044 | 0.143 | 0.041 | 0.106 | 0.076 | 0.416 | 15 |
10 | 0.223 | 0.217 | 0.339 | 0.034 | 0.121 | 0.098 | 0.067 | 0.114 | 0.630 | 8 |
11 | 0.160 | 0.098 | 0.326 | 0.025 | 0.055 | 0.094 | 0.004 | 0.174 | 0.979 | 1 |
12 | 0.220 | 0.167 | 0.274 | 0.034 | 0.093 | 0.079 | 0.044 | 0.131 | 0.750 | 6 |
13 | 0.180 | 0.127 | 0.261 | 0.028 | 0.071 | 0.075 | 0.028 | 0.153 | 0.845 | 4 |
14 | 0.252 | 0.284 | 0.300 | 0.039 | 0.158 | 0.087 | 0.105 | 0.076 | 0.419 | 14 |
15 | 0.192 | 0.132 | 0.274 | 0.030 | 0.074 | 0.079 | 0.027 | 0.151 | 0.847 | 2 |
16 | 0.241 | 0.274 | 0.287 | 0.037 | 0.153 | 0.083 | 0.100 | 0.079 | 0.440 | 13 |
17 | 0.184 | 0.126 | 0.261 | 0.028 | 0.070 | 0.075 | 0.028 | 0.153 | 0.847 | 3 |
18 | 0.235 | 0.223 | 0.235 | 0.036 | 0.124 | 0.068 | 0.077 | 0.098 | 0.562 | 10 |
Each alternative experiment's the positive and negative ideal distance solution values were computed using Eq. (7). Using the \(D_{{ij}}^{+}\), \(D_{{ij}}^{ - }\) values, the closeness coefficient index (Ci) was determined for each alternative using Eq. (8). The values of the proximity coefficients for each experimental run utilising the L18 orthogonal array are shown in Table 3. Because it indicates the highest proximity coefficient; the values recorded in each trial, from the 18 closeness coefficients showed that experiment 11 had the best multi-response characteristics. Experiment no. 11 yielded optimal process settings in the MCDM methods for the output responses this will conformed with higher Ci values indicates the best setting for the Inconel 718 alloy machining.
The CODAS technique uses a two-step rating procedure to assess the attractiveness of particular traits. Calculate ei and ti distances using the NS [32–34, 39]. The normalised and weightage normalised performance values for each criterion are calculated using Eqs. 9–11. Eq. 12 is then used to determine the NS. Based on the NS, Eq. 13 offers two options: ei and ti distances. As indicated in Table 4, Eqs. 14–15 were used to generate the relative assessment matrix (Hi) and ranking scores. Because it represents the highest relative assessment core; the values recorded in each trial, highest relative assessment score value of (7.166) is consider as the best parameter settings and corresponding experiment no. 11.
Table 4
Relative assessment matrix of CODAS method
Exp.No | Normalized Decision Matrix | Weighted Normalized Decision Matrix | Negative- ideal solution | Relative Assessment matrix (Hi ) | Rank |
Ei | Ti |
SR | VB | TL | SR | VA | TL | | |
| mm | m/s2 | min | mm | m/s2 | min | | | | |
1 | 0.909 | 0.798 | 0.654 | 0.140 | 0.445 | 0.189 | 0.319 | 0.437 | 3.291 | 2 |
2 | 0.716 | 0.485 | 0.577 | 0.110 | 0.270 | 0.167 | 0.142 | 0.211 | -0.295 | 10 |
3 | 0.483 | 0.265 | 0.462 | 0.074 | 0.148 | 0.133 | 0.014 | 0.019 | -1.916 | 17 |
4 | 0.863 | 0.664 | 0.462 | 0.133 | 0.370 | 0.133 | 0.238 | 0.300 | 1.433 | 6 |
5 | 0.678 | 0.404 | 0.538 | 0.105 | 0.225 | 0.155 | 0.097 | 0.149 | -0.982 | 14 |
6 | 0.483 | 0.251 | 0.462 | 0.074 | 0.140 | 0.133 | 0.011 | 0.011 | -1.927 | 18 |
7 | 0.694 | 0.436 | 0.462 | 0.107 | 0.243 | 0.133 | 0.109 | 0.147 | -0.786 | 13 |
8 | 0.616 | 0.286 | 0.462 | 0.095 | 0.159 | 0.133 | 0.031 | 0.051 | -1.774 | 16 |
9 | 0.561 | 0.383 | 0.423 | 0.087 | 0.213 | 0.122 | 0.075 | 0.086 | -1.211 | 15 |
10 | 0.719 | 0.452 | 1.000 | 0.111 | 0.252 | 0.289 | 0.204 | 0.315 | 0.806 | 8 |
11 | 1.000 | 1.000 | 0.962 | 0.154 | 0.557 | 0.278 | 0.452 | 0.653 | 7.166 | 1 |
12 | 0.730 | 0.586 | 0.808 | 0.113 | 0.327 | 0.233 | 0.221 | 0.336 | 1.137 | 7 |
13 | 0.893 | 0.770 | 0.769 | 0.138 | 0.429 | 0.222 | 0.312 | 0.452 | 3.185 | 4 |
14 | 0.637 | 0.345 | 0.885 | 0.098 | 0.192 | 0.255 | 0.145 | 0.210 | -0.244 | 9 |
15 | 0.836 | 0.742 | 0.808 | 0.129 | 0.413 | 0.233 | 0.300 | 0.439 | 2.899 | 5 |
16 | 0.665 | 0.358 | 0.846 | 0.103 | 0.199 | 0.244 | 0.139 | 0.210 | -0.353 | 11 |
17 | 0.874 | 0.779 | 0.769 | 0.135 | 0.434 | 0.222 | 0.316 | 0.454 | 3.275 | 3 |
18 | 0.682 | 0.440 | 0.692 | 0.105 | 0.245 | 0.200 | 0.134 | 0.214 | -0.431 | 12 |
A novel and efficient MCDM technique is (EDAS). In this strategy, the attractiveness of alternatives is measured by their distance from an average answer [27, 36, 40]. The EDAS approach may be efficient for solving the turning of Inconel 718 alloy using graphite mixed lubricating oil and process parameters were adjusted since the average solution is determined by an arithmetic mean.
Eqs. 1 and 2 are used to obtain the starting matrix and average solution (16–17). Based on the criteria type, Eqs. (18) and (19) were utilised to determine positive and negative distances (PDA, NDA). The second step was to calculate the weighted total of the positive and negative distances from the average response (SPi, SNi). These figures were calculated using Eq. (20). Eq. 21 utilises the final evaluation score (ASi) to rank the optimal process parameter based on the values of SPi and SNi, as shown in Table 5. Because it represents the highest relative assessment core; the values recorded in each trial, experiment no. 11 demonstrated the best multi-response characteristics among the values of the L18 the appraisal score (ASi). surprise, Experiment no. 11 shows the best parameter settings, with a larger Appraisal score value indicating that the related experiment is closer to the normalized values of SP and SN value. In all three MCDM method, It has been discovered that maintaining the best parameter configuration of environment (2 i.e. ASCF), at Vc, f and ap were 50 m/min, 0.15 mm, and 0.4 mm respectively results in greater values of proximity coefficient, decreasing the SR, VA, and maximizing the TL.
Table 5
Evaluation based on Distance from average solution EDAS
Exp.No | Positive distance from average PDA | Negative distance from average NDA | NSP | NSN | Appraisal Score | Rank |
SR | VA | TL | SR | VA | TL | | | | |
| mm | m/s2 | min | mm | m/s2 | min | | | | |
1 | 0.235 | 0.440 | 0.000 | 0.000 | 0.000 | 0.022 | 0.521 | 0.985 | 0.753 | 5 |
2 | 0.029 | 0.078 | 0.000 | 0.000 | 0.000 | 0.137 | 0.082 | 0.910 | 0.496 | 12 |
3 | 0.000 | 0.000 | 0.000 | 0.441 | 0.685 | 0.310 | 0.000 | 0.064 | 0.032 | 17 |
4 | 0.194 | 0.327 | 0.000 | 0.000 | 0.000 | 0.310 | 0.402 | 0.798 | 0.600 | 8 |
5 | 0.000 | 0.000 | 0.000 | 0.026 | 0.106 | 0.195 | 0.000 | 0.787 | 0.394 | 13 |
6 | 0.000 | 0.000 | 0.000 | 0.441 | 0.783 | 0.310 | 0.000 | 0.000 | 0.000 | 18 |
7 | 0.000 | 0.000 | 0.000 | 0.002 | 0.026 | 0.310 | 0.000 | 0.780 | 0.390 | 14 |
8 | 0.000 | 0.000 | 0.000 | 0.130 | 0.562 | 0.310 | 0.000 | 0.347 | 0.173 | 16 |
9 | 0.000 | 0.000 | 0.000 | 0.239 | 0.167 | 0.367 | 0.000 | 0.496 | 0.248 | 15 |
10 | 0.032 | 0.012 | 0.495 | 0.000 | 0.000 | 0.000 | 0.416 | 1.000 | 0.708 | 6 |
11 | 0.305 | 0.553 | 0.438 | 0.000 | 0.000 | 0.000 | 1.000 | 1.000 | 1.000 | 1 |
12 | 0.048 | 0.238 | 0.208 | 0.000 | 0.000 | 0.000 | 0.381 | 1.000 | 0.690 | 7 |
13 | 0.221 | 0.419 | 0.150 | 0.000 | 0.000 | 0.000 | 0.610 | 1.000 | 0.805 | 2 |
14 | 0.000 | 0.000 | 0.323 | 0.092 | 0.295 | 0.000 | 0.249 | 0.748 | 0.499 | 11 |
15 | 0.168 | 0.398 | 0.208 | 0.000 | 0.000 | 0.000 | 0.597 | 1.000 | 0.799 | 4 |
16 | 0.000 | 0.000 | 0.265 | 0.046 | 0.249 | 0.000 | 0.205 | 0.808 | 0.506 | 9 |
17 | 0.204 | 0.426 | 0.150 | 0.000 | 0.000 | 0.000 | 0.602 | 1.000 | 0.801 | 3 |
18 | 0.000 | 0.000 | 0.035 | 0.019 | 0.016 | 0.000 | 0.027 | 0.977 | 0.502 | 10 |
4.1 Surface roughness (Ra)
Dry machining roughness ranges from 0.685 to 1.291 µm, whereas ASCF machining roughness ranges from 0.698 to 0.978 µm. Increasing the f causes the Ra value to rise throughout the machining process. At a Vc of 50 m/min, f of 0.15 mm, and ap of 0.4 mm, ASCF machining lowers roughness by 19 to 33 percent compared to dry machining. This is due to the presence of solid lubricants in the tool tip contact, which minimises the amplitude of waviness on the machined surface while maintaining a uniform profile surface as viewed in 2D. [41] Fig. 2 shows the presence of feed marks and scratches on the surface morphology. Figure 2 depicts an optical microscopic examination of a machined surface at the ideal conditions Vc of 50 m/min, f of 0.15 mm/min, and ap of 0.4 mm in both dry and ASCF conditions. When compared to ASCF machining, dry machining produced more grove markings and burned marks. This is due to the presence of solid lubrication, which improves heat absorption and lowers build-up edge (BUE)development.[42]
When compared to the ASCF state, the machined surface has a non-uniform surface profile with a greater value of Ra. Surface roughness was a smooth, uniform surface profile without any appearance of waviness at the beginning stage of machining in both the dry and ASCF machining environment. During dry machining, chipping and BUE cause a greater value of surface roughness on the machined surface. Dry machining creates more feed marks on the workpiece surface than ASCF machining because of the lower Vc and f. This is due to BUE being formed as a result of the long engagement of chip with cutting inserts relatively large cutting temperature generated near the cutting edge at the higher feed rate [43, 44]. Low groove markings and improved Ra on the machined surface are achieved with constructive chip breakability and decreased BUE during ASCF machining. Figure 3 depicts a 2D surface analysis of a machined surface at the ideal conditions Vc of 50 m/min, f of 0.15 mm/min, and ap of 0.4 mm in both dry and ASCF conditions. The figure shows that in the dry state, there were more feed marks with non-uniform surface profiles, which is attributable to vibration build-up during the machining process. Surface roughness has a direct link with vibration build-up. The presence of solid lubricants droplets on the tooltip interface reduces the amplitude of waviness on the machined surface with a uniform profile surface as seen in the 2D surface at the optimum condition Vc of 50 m/min, f of 0.15 mm/min, and ap of 0.4 mm at both dry and ASCF condition, as seen in the images.[45].
4.2 Tool life
In dry machining, higher Vc causes greater plastic deformation at the tool-workpiece contact, resulting in shorter tool life. The presence of small spray droplets of solid lubricants during machining in ASCF minimises the coefficient of friction along the workpiece and the cutting edge, reducing cutting edge sharpness during machining at high temperatures. In contrast to dry machining conditions, ASCF machining extends tool life by up to 40%-60%. Figure 4 illustrates the SEM analysis of tool wear at the optimal conditions of 50 m/min cutting speed, 0.15 mm/min feed rate, and 0.4 mm depth of cut in both dry and ASCF conditions. Craters and chipping were evident on the tool surface due to tool wear in both dry and ASCF machining environments. In the dry condition, more craters and chipping were found on the surface due to the tool and work interface at higher speeds, whereas in the ASCF condition, fewer craters and chipping were found on the surface due to the presence of solid lubricants during the machining process, resulting in a lower tool wear rate. During the turning of the workpiece in both dry and ASCF conditions, the adhesive chipping layer was visible on the rake surface of the tool.[46, 47] When opposed to dry machining, ASCF displays less crater wear and chippings due to the presence of high thermal stress and high strain hardening rate during the machining process.[42, 45, 48, 49]. The finding is also in line with that of An et al.[50] and pal et al.[51] Solid lubricants provide a cooling effect, which lowers the cutting tool temperature at the interface zone and increases tool life. ASCF machining demonstrates less substantial fracture and diffusion area, and the damage is significantly smaller owing to solid lubrication, which minimises the heat created during the machining process and results in longer tool life.