Scholarlink

This research aims to predict and create a model of cutting force as a response to multiple parameters in the face milling process of JIS SKD 11 tool steel material. The face milling parameters are coolant flow rate (Q), cutting speed (Vc), feeding speed (Vf), and axial depth of cut (Aa). The flow rate (Q) for cryogenic cooling has two levels. Cutting speed (Vc), feeding speed (Vf), and axial depth of cut (Aa) each have three levels. Using the Taguchi method, it has an L18 mixed-orthogonal array as the experiment design. Using ANOVA, the parameters are qualified to only those significant toward cutting force. The prediction uses Sugeno type-1 and Mamdani type-1 fuzzy inference systems (FIS). The tuning of FIS is conducted using genetic algorithm. The tuned parameters were rules-only and rules-and-output. The membership functions are set to be Gaussian The results are shown in surface plots and root mean square error (RMSE). The result shows errors are less than 5%, which tells the cutting force has been successfully predicted.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of the Cutting Force in the Face Milling Process on JIS SKD 11 Tool Steel Material with Cryogenic Cooling Using Fuzzy Inference System

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Chezta Ahmad Muzakky,
  • Arif Wahjudi,
  • B. O. P. Soepangkat

摘要

This research aims to predict and create a model of cutting force as a response to multiple parameters in the face milling process of JIS SKD 11 tool steel material. The face milling parameters are coolant flow rate (Q), cutting speed (Vc), feeding speed (Vf), and axial depth of cut (Aa). The flow rate (Q) for cryogenic cooling has two levels. Cutting speed (Vc), feeding speed (Vf), and axial depth of cut (Aa) each have three levels. Using the Taguchi method, it has an L18 mixed-orthogonal array as the experiment design. Using ANOVA, the parameters are qualified to only those significant toward cutting force. The prediction uses Sugeno type-1 and Mamdani type-1 fuzzy inference systems (FIS). The tuning of FIS is conducted using genetic algorithm. The tuned parameters were rules-only and rules-and-output. The membership functions are set to be Gaussian The results are shown in surface plots and root mean square error (RMSE). The result shows errors are less than 5%, which tells the cutting force has been successfully predicted.