Scholarlink

In today’s industries, reliability prediction emerges as a crucial method for accurately estimating and predicting the lifespan of components utilized in daily applications. This paper explores process parameter optimization to identify the most optimal welding parameters for achieving maximum reliability. In this work, AA-7075 material is used. Aluminum alloy work pieces are welded based on three varied parameters: weld travel speed, wire speed, and current, as per the Taguchi method of L27 Design of experiment array. Subsequently, these work pieces are subjected to tensile testing to record failure values such as tensile strength and time to failure. Soft computing methods, specifically artificial neural network and fuzzy logic, are employed to predict reliability, failure time, and optimize process parameters. The study compares these results with conventional industry methods, aiming to offer a more efficient and accurate solution to existing reliability concerns. It is found out that the predictive ANN model is capable of giving a better reliability prediction of TIG welded aluminum joint than fuzzy logic.

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

Predicting the Reliability of TIG Weld Joint of Aluminum Alloy Using Soft Computing Approach

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • S. Kapilash,
  • S. Pradeesh Dharshan,
  • S. Sai Prasath,
  • M. Shiva,
  • J. Gokulachandran

摘要

In today’s industries, reliability prediction emerges as a crucial method for accurately estimating and predicting the lifespan of components utilized in daily applications. This paper explores process parameter optimization to identify the most optimal welding parameters for achieving maximum reliability. In this work, AA-7075 material is used. Aluminum alloy work pieces are welded based on three varied parameters: weld travel speed, wire speed, and current, as per the Taguchi method of L27 Design of experiment array. Subsequently, these work pieces are subjected to tensile testing to record failure values such as tensile strength and time to failure. Soft computing methods, specifically artificial neural network and fuzzy logic, are employed to predict reliability, failure time, and optimize process parameters. The study compares these results with conventional industry methods, aiming to offer a more efficient and accurate solution to existing reliability concerns. It is found out that the predictive ANN model is capable of giving a better reliability prediction of TIG welded aluminum joint than fuzzy logic.