Post - Knowmats

A Framework for the Calibration and Validation of Multiscale Material Models

October 01, 2018 | by DatapointLabs | views 8956

Multiscale material models are being increasingly applied for high-level simulation of complex materials, such as continuous reinforced material products (unidirectional and woven product forms). These multiscale material models require input data from a minimum of experimental tests, which are then used to characterize a multiscale material model that can be used in structural simulations within a variety of commercial finite element solvers, including OptiStruct, RADIOSS, Abaqus, and LS-Dyna. Using these models, it is possible is to predict the performance of layups from single layer properties, as well as performance of these composites under complex loadings. We present a framework where the required experimental data are collected, including a process for maintaining traceability and consistency of the experimental data using the Matereality software. Experimental test data are transmitted to the HyperWorks Multiscale Designer software for development of an appropriate multiscale material model. The resulting multiscale material model data is stored within Matereality linked to the source experimental data. Different manufactured layups are tested and compared to simulation in a validation step which provides a measure of the solution accuracy.

By Jeffrey Wollschlager, Altair Engineering; and Megan Lobdell and Hubert Lobo, DatapointLabs

Presented by Megan Lobdell at the 2018 Global Altair Technology Conference, Paris, France