Datasets describing optimization of the cutting regime in the turning of AISI 316L steel for biomedical purposes based on the NSGA-II and NSGA-III multi-criteria algorithms

There are several methods of analysis used in the metalworking industry for dry machining processes and with Minimum Quantity Lubrication (MQL). Evolutionary methods [1] have been used in the decision-making process in the machining process to select the optimal data and to analyze the behavior of variables such as cutting speed (V), feed rate (f) and cutting depth (ap). This work addresses the use of evolutionary algorithms of low dominance class II and III (NSGA-II and NSGA-III) to analyze from the multicriteria approach the initial wear of the cutting tool (VB), the energy consumption (E) and the machining time (t) in the turning process of the AISI 316L steel workpiece for biomedical purposes. As input variables to the algorithm with 54 records, there are: cutting speed (V: 200, 300, 400 m/min) and feed rate (f: 0.1, 0.15, 0.2 mm/rev). The experiment was developed for a dry (1) turning operation and with the use of MQL (-1). For the MQL lubrication regime, a TRI-COOL MD-1 lubricant was employed, a vegetable type used in ferrous and non-ferrous metal cutting operations. A BIDEMICS JX1 ceramic cutting tool was used.


a b s t r a c t
There are several methods of analysis used in the metalworking industry for dry machining processes and with Minimum Quantity Lubrication (MQL).Evolutionary methods [1] have been used in the decision-making process in the machining process to select the optimal data and to analyze the behavior of variables such as cutting speed (V), feed rate (f) and cutting depth (a p ).This work addresses the use of evolutionary algorithms of low dominance class II and III (NSGA-II and NSGA-III) to analyze from the multicriteria approach the initial wear of the cutting tool (VB), the energy consumption (E) and the machining time (t) in the turning process of the AISI 316L steel workpiece for biomedical purposes.As input variables to the algorithm with 54 records, there are: cutting speed (V: 20 0, 30 0, 40 0 m/min) and feed rate (f: 0.1, 0.15, 0.2 mm/rev).The experiment was developed for a dry (1) turning operation and with the use of MQL (-1).For the MQL lubrication regime, a TRI-COOL MD-1 lubricant was employed, a vegetable type used in ferrous and non-ferrous metal cutting operations.A BIDEMICS JX1 ceramic cutting tool was used.
© The experiment was developed for a dry (1) turning operation and with the use of MQL (-1).The cutting speed (V) in three levels (200, 300 and 400 m / min; 1, 0, -1) and the feed rate (f) in three levels (0.1, 0.15 and 0.2 mm / rev; 1, 0, -1) were taken as independent variables, three replications were made of each run (Table 1).For the MQL lubrication regime, the TRI-COOL MD-1 lubricant of vegetable type was used for cutting operations.In the output variables, the main cutting force (Fz), cutting tool wear

Value of the Data
• These data are important, since they facilitate in the technological process of machining AISI 316L steel for biomedical purposes, with MQL and dry, to predict the initial wear of the cutting tool through the multicriteria method with artificial intelligence.The algorithms used facilitate the determination of the hierarchy of the best alternative in a very short time.• The data can be used in research related to decision-making in the AISI 316L machining process for biomedical purposes, evaluating the best results in terms of production quality and low costs, for workshops or companies in the metalworking sector.
• The mathematical optimization model used, with multicriteria analysis, quickly evaluates the best values of the dependent variables in the two environments (dry and MQL) analyzed.

Objective
The optimization of manufacturing processes by material removal is a very active area of study in the scientific literature.By applying optimization tools, the management of operations in machining workshops is improved.With the optimal data, it is possible to increase production, guaranteeing greater durability of the cutting tools.With the interpretation of the data from the research "Optimization of the Cutting Regime in the Turning of the AISI 316L Steel for Biomedical Purposes Based on the Initial Progression of Tool Wear" [1] , it was performing a multicriteria analysis of the dry machining operation and MQL of 316L steel for biomedical purposes.By applying the NSAG-III algorithm as a new generation of the non-dominant evolutionary genetic algorithm, and with the input data of the cutting speed (V) and the feed rate (f), it is possible to establish the hierarchy of solution variants with the recommended parameters.

Data Description
The data set from the results ( Table 1 ) of the experiment was used, when machining AISI 316 steel specimens for biomedical purposes with a hardness of 148 HRB on a CNC lathe.A BIDEMICS-type ceramic tool was used as a cutting tool.A HAAS ST10 CNC machining center was used as a machine tool.
For the analysis, we started from a population of 54 variants shown in Table 1 .The results of convergence of the Pareto front when arriving at the variants, in the NSGA-II and NSGA-III algorithms that are used, are presented in Tables 2 and 3 respectively.The optimal selection of values for the best approximation of the Pareto front is shown in Tables 4 and 5 .The non-uniform distribution of the Pareto front points for the NSGA-II algorithm is presented in Fig. 1 .In the case of the NSGA-III algorithm, it is shown in Fig. 2 , where a better approximation between the points is observed in the graph.Its uniform distribution is due to the incorporation of reference points.Fig. 1 shows the behavior for cutting speed V = 100 m/min, feed rate 0.1 < f < 0.2 mm/rev and cutting depth at 1.3 < ap < 1.8 mm.In this data set, the analysis of the objective functions showed that the energy consumption (0.013 ≤ E ≤ 0.018 kW.H/dm 3 ) is minimized when performing the roughing operation, to reach the estimated production levels in a machining time t ≥ 0.5 mm; and initial tool wear behaves in the range of 0.007 ≤ VB ≤ 0.009 mm.

Table 3
The NSGA-III non-dominated ordering algorithm shows the results for the restrictions presented in 32 variants.The constant cutting speed with Vc = 100 m/min, the feed rate behaves 0.1 < fn < 0.2 and the depth of cut at 0.5 < Ap < 1.2.In this state, in relation to the mentioned restrictions, the results of the studied objective functions showed fundamentally: it is achieved to decrease the energy consumption in the values of 0.015 ≤ E ≤ 0.017 for the material removal to reach the production levels in a time t ≥ 0.5, and with the tool wear in the values of 0.005 ≤ VB ≤ 0.009.
In the Table 4 show the analyzing the approximation with the pareto front, the behavior of the tool wear during the turning operation, the value of time (t) in the interval of (0.90-0.99) minutes per cubic decimeter has been taken as reference, in which the best behavior is appreciated, with the lowest and most stable values of feed rate (0.005) millimeters per minute.Meanwhile, the energy consumption is set at (0.002-0.08) kilo what hour per cubic decimeter, for a volume of material removed of 5-17 millimeters.
2023 Published by Elsevier Inc.This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

Table 1
Data of the experimental values of the machining of AISI 316L steel for biomedical purposes.It includes 2 columns and shows the 54 measurements made to the specimens in the dry turning and MQL operation.As input variables we have cutting speed (V), feed rate (f) and lubrication regime (LR).As output variables we have wear of the cutting tool (VB), surface roughness (Ra) and cutting force (Fz).

Table 2
The NSGA-II non-dominated classification algorithm shows in 20 variants, the results for the restrictions (t: machining time; E: energy consumption; MRR: Material Removal Rate).

Table 4
Result of the best solutions for the NSGA-II algorithm.

Table 5
Result of the best solutions for the NSGA-III algorithm.

Table 6
Comparison of the results of the statistical analysis between the algorithms.