<p>Ceramic matrix composites, particularly C/SiC are known for their outstanding characteristics such as high toughness, excellent specific strength and superior resistance to high temperatures and wear. Due to their unique properties, these materials are increasingly utilized in aerospace, defense and aviation industries. However, their machining presents challenges because of their complex nature including anisotropy, heterogeneity and changing thermal behavior. In machining processes, cutting force is the key factor / index which needs to be minimized up to acceptable limit. In this paper, a Hybrid Physics-ML model has been introduced for prediction of axial cutting forces in rotary ultrasonic profile milling (RUPM) of such materials. The model follows the energy conservation theorem, indentation fracture theory and penetration trajectory. The model has been validated by experimental rotary ultrasonic profile milling of C/SiC, for which the simulated and measured valued of cutting forces found nearly matched (equal) for most of the experimental groups. The mean absolute percentage error (MAPE) and standard deviation was calculated as 3.25% and 2.92, respectively. The slight variation in few experimental groups was caused by anisotropy, heterogeneity, inhomogeneity and complex nature of such composites. The study revealed that cutting forces decreased by increasing spindle speed, but started to increase with the increased in cutting depth and feed rate. The proposed Hybrid Physics-ML model is novel, robust and can be applied to optimize machining parameters and saving (from rejection) of such expensive materials at industry level.</p>

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

A Hybrid Physics-ML model for axial cutting force prediction in rotary ultrasonic milling of ceramic matrix composites

  • Muhammad Amin,
  • Obaid Muhammad,
  • Muhammad Faisal Rathore,
  • Waqas Saleem,
  • Songmei Yuan

摘要

Ceramic matrix composites, particularly C/SiC are known for their outstanding characteristics such as high toughness, excellent specific strength and superior resistance to high temperatures and wear. Due to their unique properties, these materials are increasingly utilized in aerospace, defense and aviation industries. However, their machining presents challenges because of their complex nature including anisotropy, heterogeneity and changing thermal behavior. In machining processes, cutting force is the key factor / index which needs to be minimized up to acceptable limit. In this paper, a Hybrid Physics-ML model has been introduced for prediction of axial cutting forces in rotary ultrasonic profile milling (RUPM) of such materials. The model follows the energy conservation theorem, indentation fracture theory and penetration trajectory. The model has been validated by experimental rotary ultrasonic profile milling of C/SiC, for which the simulated and measured valued of cutting forces found nearly matched (equal) for most of the experimental groups. The mean absolute percentage error (MAPE) and standard deviation was calculated as 3.25% and 2.92, respectively. The slight variation in few experimental groups was caused by anisotropy, heterogeneity, inhomogeneity and complex nature of such composites. The study revealed that cutting forces decreased by increasing spindle speed, but started to increase with the increased in cutting depth and feed rate. The proposed Hybrid Physics-ML model is novel, robust and can be applied to optimize machining parameters and saving (from rejection) of such expensive materials at industry level.