Prediction of the flow stress of a high alloyed austenitic stainless steel using artificial neural network

Zhaohui Zhang*, Dongna Yan, Jiantao Ju, Ying Han

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Citations (Scopus)

Abstract

The high temperature flow behavior of as-cast 904L austenitic stainless steel was studied using artificial neural network (ANN). Isothermal compression tests were carried out at the temperature range of 1000°C to 1200°C and strain rate range of 0.01 to 10s-1. Based on the experimental flow stress data, an ANN model for the constitutive relationship between flow stress and strain, strain rate and deformation temperature was constructed by back-propagation (BP) method. Three layer structured network with one hidden layer and nine hidden neurons was trained and the normalization method was employed in training process to avoid over fitting. Modeling results show that the developed ANN model exhibits good performance for predicting the flow stresses of the 904L steel. Therefore, it can be used to reflect the hot deformation behavior in a wide working window.

Original languageEnglish
Title of host publicationEco-Materials Processing and Design XIII
PublisherTrans Tech Publications Ltd.
Pages351-354
Number of pages4
ISBN (Print)9783037854396
DOIs
Publication statusPublished - 2012
Externally publishedYes
Event13th International Symposium on Eco-Materials Processing and Design, ISEPD 2012 - Guilin, China
Duration: 7 Jan 201210 Jan 2012

Publication series

NameMaterials Science Forum
Volume724
ISSN (Print)0255-5476
ISSN (Electronic)1662-9752

Conference

Conference13th International Symposium on Eco-Materials Processing and Design, ISEPD 2012
Country/TerritoryChina
CityGuilin
Period7/01/1210/01/12

Keywords

  • Artificial neural network
  • Austenitic stainless steel
  • Back-propagation neural network
  • Hot deformation

Fingerprint

Dive into the research topics of 'Prediction of the flow stress of a high alloyed austenitic stainless steel using artificial neural network'. Together they form a unique fingerprint.

Cite this