Parameter identification of an elasto-plastic behaviour using artificial neural networks–genetic algorithm method

Crystal plasticity simulation is a widely used technique for studying the deformation processing of polycrystalline materials. However, inclusion of crystal plasticity simulation into design paradigms such as integrated computational materials engineering (ICME) is hindered by the computational cost of large-scale simulations. In this work, we present a machine learning (ML) framework using the material information platform, Open Citrination, to develop and calibrate a reduced order crystal plasticity model for face-centered cubic (FCC) polycrystalline materials, which can be both rapidly exercised and easily inverted. The reduced order model takes crystallographic texture, constitutive model parameters, and loading condition as inputs and returns the stress-strain curve and final texture. The model can also be inverted and take a stress-strain curve, loading condition, and final texture as inputs and return the initial texture and constitutive model parameters as outputs. Principal component analysis (PCA) is used to develop an efficient description of the crystallographic texture. A viscoplastic self-consistent (VPSC) crystal plasticity solver is used to create the training data by modeling the stress-strain behavior and evolution of texture during deformation processing.

/pdf/machine-learning-based-reduce-order-crystal-plasticity-3a0tpmw2er.pdf

Machine Learning–Based Reduce Order Crystal Plasticity Modeling for ICME Applications

The main purpose of the present work is to measure the stress–strain behavior under warm conditions (about 100 °C) of advanced high strength steel (AHSS) sheets up to large strains compared to uniaxial tension. The test equipment consists of two main parts, i.e., a hydraulic bulge tester and a heating device. A mechanical system is attached to the test equipment for measuring the membrane stress and thickness strain at the bulge pole. The stress–strain curves were measured for three kinds of AHSS sheets with the proposed test method for various initial temperatures (10, 50 and 100 °C). The proposed method does not provide isothermal stress–strain curves because the specimen temperature increases during the test due to the effect of deformation-induced heating. A numerical scheme using thermo-mechanical finite element (FE) simulations is suggested to deconvolute the isothermal stress–strain curves.

Balanced Biaxial Testing of Advanced High Strength Steels in Warm Conditions

The sustainability of neural network applications within finite element analysis in sheet metal forming: A review

This paper presents an approach, based on machine learning techniques, to predict the occurrence of defects in sheet metal forming processes, exposed to sources of scatter in the material properties and process parameters. An empirical analysis of performance of ML techniques is presented, considering both single learning and ensemble models. These are trained using data sets populated with numerical simulation results of two sheet metal forming processes: U-Channel and Square Cup. Data sets were built for three distinct steel sheets. A total of eleven input features, related to the mechanical properties, sheet thickness and process parameters, were considered; also, two types of defects (outputs) were analysed for each process. The sampling data were generated, assuming that the variability of each input feature is described by a normal distribution. For a given type of defect, most single classifiers show similar performances, regardless of the material. When comparing single learning and ensemble models, the latter can provide an efficient alternative. The fact that ensemble predictive models present relatively high performances, combined with the possibility of reconciling model bias and variance, offer a promising direction for its application in industrial environment.

/pdf/single-and-ensemble-classifiers-for-defect-prediction-in-1dr37b6fep.pdf

Single and ensemble classifiers for defect prediction in sheet metal forming under variability

This paper deals with the identification of material parameters for an elastoplastic behaviour model with isotropic hardening using several experimental tests at the same time. But, these tests are generally inhomogeneous and finite element simulations are necessary for their analysis. Therefore an inverse analysis is carried out and the identification problem is converted into a multi-objective optimization where prohibitive computing time is required. We propose in this work a hybrid approach where Artificial Neural Networks (ANN) are trained by finite element results. Then, the multi objective procedure calls the ANN function in place of the finite element code. The proposed approach is exemplified on the identification of non-associative Hill’48 criterion and Voce parameters model of the Stainless Steel AISI 304.

R. Hambli

Papers

Identification of Constitutive Parameters using Hybrid ANN multi-objective optimization procedure