Scholarlink

<p>The current work explores the deep drawability of 17-4 PH stainless steel sheets using a hybrid model combining finite element (FE) simulations and artificial neural network (ANN). The drawability was assessed under various heat-treated conditions, viz., as-received (AR), solution heat-treated (SHT), and age-hardened (AH). FE simulations of deep drawing were performed using DEFORM 3D software. Four different hardening laws, namely Hollomon, Voce, Ludwik, and Swift, were incorporated with the Hill 1948 yield function and the Cockcroft and Latham (C-L) fracture model to predict fracture behavior and evaluate the effectiveness of each hardening law. The deep drawing experiments revealed variations in austenite content due to the transformation-induced plasticity (TRIP) effect, which enhanced drawability in AH conditions. The wall region exhibited the lowest austenite percentage. The Hollomon law, combined with the Hill 1948 yield function and the C-L fracture criterion, effectively fitted the experimental data, and this combination is used to generate data for training the ANN. The ANN model developed in this work accurately predicted the draw depth and thickness strain at various temperatures and times for both SHT and AH conditions. The results demonstrate the potential of the ANN model as a tool for predicting the drawability of 17-4 PH stainless steel sheets.</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

Prediction of Deep Drawability of Heat-Treated 17-4 Precipitation Hardening Stainless Steel Sheets by Numerical Simulation and Neural Network

  • 期刊论文

  • 发表日期：

    2025年10月17日

查看全文
  • Rishabh Saxena,
  • R. Ganesh Narayanan,
  • P. S Robi

摘要

The current work explores the deep drawability of 17-4 PH stainless steel sheets using a hybrid model combining finite element (FE) simulations and artificial neural network (ANN). The drawability was assessed under various heat-treated conditions, viz., as-received (AR), solution heat-treated (SHT), and age-hardened (AH). FE simulations of deep drawing were performed using DEFORM 3D software. Four different hardening laws, namely Hollomon, Voce, Ludwik, and Swift, were incorporated with the Hill 1948 yield function and the Cockcroft and Latham (C-L) fracture model to predict fracture behavior and evaluate the effectiveness of each hardening law. The deep drawing experiments revealed variations in austenite content due to the transformation-induced plasticity (TRIP) effect, which enhanced drawability in AH conditions. The wall region exhibited the lowest austenite percentage. The Hollomon law, combined with the Hill 1948 yield function and the C-L fracture criterion, effectively fitted the experimental data, and this combination is used to generate data for training the ANN. The ANN model developed in this work accurately predicted the draw depth and thickness strain at various temperatures and times for both SHT and AH conditions. The results demonstrate the potential of the ANN model as a tool for predicting the drawability of 17-4 PH stainless steel sheets.