dc.description.abstract |
Material behaviour modelling involves the development &mathematical models based on experimental data, experts' observations and reasoning. Against the rigorous iterative exercise of developing mathematical models, machine learning (ML) model neural networks (NN) offer a fundamentally different and appealing approach to the derivation and representation of material behaviour relationships. Such networks would contain sufficient information about the material behaviour complexities, non-linear characteristics, stress strain behaviour, material properties etc. Further, these networks could be used effectively as material model to reproduce the trained experimental data and untrained experimental data. This paper addresses identification of comprehensive data set end developing a systematic approach for material model using NN Demonstration examples of this study are taken up using the experimental |
en_US |