Menu
Search Contact

Surrogate Structures

Highly detailed and nonlinear models within 1 Element

  • 1D Element outside
  • A I stiffness formulation inside
  • example: Neural model mimics an analytic spotweld model

  train A I structural models from experimental data or detailed models

Surrogate Materials

  train A I constitutive models from experimental data or micromechanical models

handle increasing complexity and diversity

  • composite materials
  • functionally graded materials
  • multi-scale materials

 

References

Due to the increasing usage of complex materials in the development of proper material models for the prediction within Finite Element simulations has become an extensive task. Soft Computing and Machine Learning methods can be very beneficial for getting the complexity under Control.

  • A. Kuhn, T. Palau, G. Schlager, H.J. Böhm, S. Nogales, V. Oancea, R. Roy, A. Rauh, J. Lescheticky: A comprehensive Approach to the modelling of complex materials with machine learning models within finite element simulations, 10th international conference on modeling and applied simulations, Rome, Italy, 12-14 Sept. 2011
  • T. Palau, A. Kuhn, S. Nogales, H.J. Böhm, A. Rauh: A Neural Network based elasto-plasticity material model, European Congress on Computational Methods in Applied Sciences and Engineering (ECCOMAS 2012), Vienna, Austria, September 10-14, 2012
  • T. Hinterdorfer, A. Kuhn: Modelling Mechanical Joints with an Artificial Neural Network Based User Element in Abaqus/Explicit, SIMULIA Technology Days, 14.11.2017, Linz