Electric discharge drilling (EDD) plays an important role in the advancement of aerospace manufacturing due to its ability to produce complex hole geometries with high precision. In this study, a statistical analysis and data-driven modeling of hole circularity have been proposed during EDD of aerospace alloy. Statistical analysis-based prediction model is developed where the experimental results of hole circularity have been considered. These are the functions of four EDD process variables (i.e., Discharge current, Pulse on time, Pulse off time, and Flushing pressure). Through training and validation, the statistically learned model exhibits the complex relationships between input EDD process parameters and output geometry, enabling it to accurately predict the hole circularity. Reliability and effectiveness of the developed model are represented by R-Square and Adjusted R-Square value which are quite at suitable level of 88.8% and 75.7%, respectively. Furthermore, contour plots of interactive effect of process parameters are developed and influence of EDD interactive process parameters on hole circularity is discussed. Finally, optimal range of process parameters was found as: Discharge Current: 8–9 A, Pulse on time: 7.3–8 µs, Pulse off time: 2–2.5 µs, Flushing pressure: 75–80 bar. The proposed study demonstrates the precise prediction and optimization of drilled geometry.

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

Statistical Analysis and Data-Driven Modeling of Hole Circularity in EDD of Aerospace Material

  • Kedari Lal Dhaker,
  • Amit Sharma,
  • Tasnim Arif,
  • Gaurav Shukla,
  • Akshat Shukla,
  • Shwetank Avikal

摘要

Electric discharge drilling (EDD) plays an important role in the advancement of aerospace manufacturing due to its ability to produce complex hole geometries with high precision. In this study, a statistical analysis and data-driven modeling of hole circularity have been proposed during EDD of aerospace alloy. Statistical analysis-based prediction model is developed where the experimental results of hole circularity have been considered. These are the functions of four EDD process variables (i.e., Discharge current, Pulse on time, Pulse off time, and Flushing pressure). Through training and validation, the statistically learned model exhibits the complex relationships between input EDD process parameters and output geometry, enabling it to accurately predict the hole circularity. Reliability and effectiveness of the developed model are represented by R-Square and Adjusted R-Square value which are quite at suitable level of 88.8% and 75.7%, respectively. Furthermore, contour plots of interactive effect of process parameters are developed and influence of EDD interactive process parameters on hole circularity is discussed. Finally, optimal range of process parameters was found as: Discharge Current: 8–9 A, Pulse on time: 7.3–8 µs, Pulse off time: 2–2.5 µs, Flushing pressure: 75–80 bar. The proposed study demonstrates the precise prediction and optimization of drilled geometry.