Open AccessDetermination of GTN Model Parameters Based on Artificial Neutral Network for a Ductile Failure
Author(s) -
Yassine Chahboub,
Szabolcs Szávai
Publication year - 2021
Publication title -
journal of mechanical, civil and industrial engineering
Language(s) - English
Resource type - Journals
ISSN - 2710-1436
DOI - 10.32996/jmcie.2021.2.1.1
Subject(s) - artificial neural network , enhanced data rates for gsm evolution , ultimate tensile strength , materials science , tension (geology) , process (computing) , structural engineering , identification (biology) , experimental data , artificial intelligence , computer science , composite material , engineering , mathematics , statistics , operating system , botany , biology
The Gurson – Tvergaard – Needleman (GTN) mechanical model is widely used to predict the failure of materials based on laboratory specimens, direct identification of Gurson – Tvergaard – Needleman parameters is not easy and time-consuming, and the most used method to determine them is the combination between the experimental results and those of the finite elements, the process consists of repeating the simulations several times until the simulation data matches the experimental data obtained at the specimen level.This article aims to find GTN parameters for the Compact Tension (CT) and Single Edge Tensile Test (SENT) specimen based on the Notch Specimen (NT) using the Artificial Neural Network (ANN) approach. . This work presents how the ANN could help us determine the parameters of GTN in a very short period of time. The results obtained show that ANN is an excellent tool for determining GTN parameters.