Abstract—Optimization of surface Roughness (Ra) and Material Removal Rate (MRR) to achieve higher productivity and improved quality machining is an objective in various machining processes. This study develops predictive models and optimizes the machining performance of SKD11 material during milling based on three cutting parameters: Cutting speed (Vc), Depth of cut (Dc), and Feed rate (Fr). The Taguchi Orthogonal Array (OA) design determines the appropriate number of experimental samples. Moreover, the Group Method of Data Handling (GMDH) method develops a predicting model for Ra. Then, the NSGA-II technique is employed for multi-objective optimization. The results show that the triple model (3 variables) for Ra exhibits the most accurate predictive performance among the nine types of GMDH models, with the highest R2 of 0.981, the lowest Root Mean Square Error (RMSE) of 0.074, and Mean Absolute Percentage of Error (MAPE) of 3.6 in training, and 0.959, 0.119, and 5.864 in validation. Applying NSGA-II for multi-objective optimization generates 70 Pareto solutions, representing the trade-offs among the cutting parameters and the two conflicting objectives of Ra and MRR. Three solutions from the Pareto set are selected and machined to validate the optimal values, revealing deviations of less than 10.3% from the optimal values. This study addresses a critical challenge in milling by developing a method to optimize Ra and MRR, which are often conflicting objectives. Hence, manufacturers can improve production efficiency and product quality simultaneously.

Keywords—optimization, surface roughness, material removal rate, group method of data handling, nondominated sorting genetic algorithm II

Cite: Cong Chi Tran, "Modelling and Optimization of Surface Roughness and Material Removal Rate in Milling SKD11 Using GMDH and NSGA-II," International Journal of Mechanical Engineering and Robotics Research, Vol. 13, No. 6, pp. 618-627, 2024.

Copyright © 2024 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.