OBRABOTKAMETALLOV TECHNOLOGY Vol. 26 No. 4 2024 fi rst order cutting speed V (contribution of about 47.77 %) and second order cutting speed V2 (contribution of about 12.67 %) have the greatest infl uence on the microhardness. Table 7 shows the ANOVA results for the F-values of the roundness error for roller burnishing under dry and NFMQL cutting conditions. Under dry conditions, the roundness error is signifi cantly aff ected by the higher order of feed (contribution of about 32.5 %), cutting speed (contribution of about 25.89 %), number of passes (about 15.47 %), as well as the eff ect of interaction between the cutting speed and the number of passes (contribution of about 18.46 %). On the other hand, the roundness error for NFMQL condition is observed to be highly aff ected by the feed (contribution of about 67.36 %), number of passes (contribution of about 11.81 %), and higher order of cutting speed (contribution of about 9.57 %). It is obvious that the feed under NFMQL cutting conditions and the number of passes under dry conditions had a signifi cant eff ect on the surface roughness. Cutting speed seems to have the greatest eff ect on microhardness, and feed rate and number of passes come in second and third place. On the other hand, this eff ect seems to be more pronounced under NFMQL cutting conditions. Under dry conditions, cutting speed has a signifi cant eff ect on roundness error; under NFMQL cutting conditions, feed has a signifi cant eff ect. It is evident from Figs. 2-4 and Tables 5-7 that the process parameters are in contradiction with the benefi cial responses. In addition, multi-objective optimization of these competing parameters is necessary to obtain the desired results. In the current work, the desirability function method is used to optimize the parameters of the roller burnishing process under NFMQL conditions to achieve the minimum roundness error, maximum microhardness and minimum surface roughness. Using this method, the optimization of several response variables becomes the optimization of a single desire function, and each response variable is converted into a desirability function [20–23]. Table 8 lists the range of the response function and the process variables. Table 8 illustrates the minimum and maximum limits of surface roughness, microhardness and roundness error based on the experimental results. A one-way transformation is used to transform each response into its corresponding desirability function [20–23]. Using the Design-Expert® software optimization module, a multi-objective optimization of roller burnishing was carried out in this study. The desirability of surface roughness, microhardness and roundness error was evaluated for each level of independent factors. The desirability of minimum surface roughness, maximum microhardness and minimum roundness error was then computed into one desirability function. The optimal process parameters for the smallest surface roughness, maximum microhardness and minimum roundness error under NFMQL conditions are shown in Table 9. The current study reveals that the ideal parameters for roller burnishing of Al6061-T6 alloy are a cutting speed of 357 rpm, a feed of 0.17 mm/rev and four passes. These results give a minimum surface roughness of 0.64 μm, a maximum microhardness of 130.19 HV and a minimum roundness error of 3.514 μm. However, it was found that the ideal parameters for roller burnishing of Al6061-T6 alloy under dry conditions are a cutting speed of 344 rpm, a feed of 0.25 mm/rev and four passes. This gives a minimum surface roughness of 0.807 μm, a maximum microhardness of 119.2 HV and a minimum roundness error of 4.282 μm. Ta b l e 8 Constraints for optimization of process parameters for NFMQL cutting conditions Parameters Goal Min.limit Max.limit Cutting speed, V (rpm) Is in range 100 500 Feed, f (mm/rev) Is in range 0.1 0.2 Number of passes, N (mm) Is in range 1 5 Surface roughness (Ra) (μm) Minimize 0.61 0.83 Microhardness (HV) Minimize 108 136 Roundness error (Re) (μm) Minimize 2.8 9.2
RkJQdWJsaXNoZXIy MTk0ODM1