Self-learning FEM/NMM approach to identification of equivalent material models for plane stress problem

• Autorzy: Pabisek E.
• Języki publikacji: EN

Abstrakty

EN

The autoprogressive and cumulative algorithms, basing on `on line' formulation of patterns and the training of NMM (Neural Material Model), are evaluated to be comparable in case of uniaxial stress state problems. It is shown in the paper that for the plane stress boundary value problems the autoprogressive algorithm, in which NMM is trained for each load increment, is superior to the cumulative algorithm. In order to formulate a small NMM and accelerate the convergence of the iteration of computed equilibrium paths to the monitored paths, a smaller number of inputs NMM is discussed and a modified selection of the training patterns is applied. A new approach is proposed with respect to the designing of NMMs, combining the `on line' and `off line' training of neural networks. The discussed problems are illustrated with two study cases. They are related to the formulation of NMMs for the identification of equivalent materials in plane trusses made of the Ramberg–Osgood material and for elasto-plastic plane stress boundary value problems.

Słowa kluczowe

EN

neural networks; constitutive modelling; material model; FEM

PL

sieć neuronowa; metoda elementów skończonych

• Wydawca: Instytut Podstawowych Problemów Techniki PAN
• Czasopismo: Computer Assisted Mechanics and Engineering Sciences
• Rocznik: 2008
• Tom: Vol. 15, No. 1
• Strony: 67--78
• Opis fizyczny: Bibliogr. 24 poz., rys., wykr.

Twórcy

autor

Pabisek E.

• Cracow University of Technology, Institute for Computational Civil Engineering, ul. Warszawska 24, 31-155 Cracow, Poland

Bibliografia

• [1] T. Furukawa, G. Yagawa. Implicit constitutive modelling for viscoplasticity using neural networks. Int. J. Num.Meth. Eng., 43(2): 195-219, 1998.
• [2] J. Ghaboussi, J.H. Garrett, X. Wu. Knowledge-based modeling of material behavior with neural networks. Journal of Engineering Mechanics, 117(1): 132-153, 1991.
• [3] J. Ghaboussi, D.A. Pecknold, M. Zhang, R.M. Haj-Ali. Autoprogressive training of neural network constitutive models. Int. J. Num. Meth. Eng., 42(1): 105-126, 1998.
• [4] J. Ghaboussi, D.E. Sidarta. New nested adaptive neural networks (NANN) for constitutive modelling. Computer and Geotechnics, 22(1): 29-52, 1998.
• [5] R. Haj-Ali, D. Pecknold, J. Ghaboussi, G. Voyiadjis. Simulated micromechanical models using artificial neural networks. Journal of Engineering Mechanics, 127(7): 730-738, 2001.
• [6] Y.M. Hashash, S. Jung, J. Ghaboussi. Numerical implementation of a neural network based material model infinite element analysis. Int. J. Num. Meth. Eng., 59(7): 989-1005, 2004.
• [7] A. A. Javadi, T.P. Tan, M. Zhang. Neural network for constitutive modelling in finite element analysis. GAMES, 10: 523-529, 2003.
• [8] S. Jung, J. Ghaboussi. Characterizing rate-dependent material behaviors in self-learning simulation. Comput. Methods Appl. Mech. Eng., 196(1-3): 608-619, 2006.
• [9] S. Jung, J. Ghaboussi. Neural network constitutive model for rate-dependent materials. Comput. Struct., 84(15-16): 955-963, 2006.
• [10] L. Kaczmarczyk, Z. Waszczyszyn. Neural procedures for the hybrid FEM/NN analysis of elastoplastic plates. GAMES, 12: 379-391, 2005.
• [11] M. Lefik, M. Wojciechowski. Artificial Neural Network as a numerical form of effective constitutive law for composites with parametrized and hierarchical microstructurc. CAMES, 12: 183-194, 2005.
• [12] M Lefik, B.A. Schrefler. Artificial neural network for parameter identifications for an elasto-plastic model of superconducting cable under cyclic loading. Comput. Struct., 80: 1699-1713, 2002. [13] M Lefik, B.A. Schrefler. Artificial neural network as a incremental non-linear constitutive model for a finite element code. Comput. Methods Appl. Mech. Eng., 192: 3265-3283, 2003.
• [14] E. Pabisek. On algorithms for identification of a neural network based model of equivalent material in real structures. Archives of Civ. Eng., 2007. (accepted for publication).
• [15] H.S. Shin, G.N. Pande. On self-learning finite element codes based on monitored response of structure. Computer and Geotechnics, 27: 161-178, 2000.
• [16] H.S. Shin, G.N. Pande. Identification of elastic constants for orthotropic materials from a structural test. Computer and Geotechnics, 30: 571-577, 2003.
• [17] D.E. Sidarta, J. Ghaboussi. Modelling constitutive behavior of materials from non-uniform material tests. Computer and Geotechnics, 22(1): 53-71, 1998.
• [18] J.C. Simo, R.L. Taylor. A return mapping algorithm for plane stress elastoplasticity. Int. J. Num. Meth. Eng., 22(3): 649-670, 1986.
• [19] T. Strzelecki et al. Mechanics of Nonhomogenous Continua. Homogenisation Theory (in Polish). Wydawnictwo Naukowe, 1996.
• [20] Z. Waszczyszyn. Artificial neural networks in civil and structural engineering: Ten years of research in Poland. CAMES, 13(4): 489-512, 2006.
• [21] Z. Waszczyszyn, E. Pabisek. Hybrid NN/FEM analysis of the elastoplastic plane stress problem. CAMES, 6: 177-188, 1999.
• [22] Z. Waszczyszyn, E. Pabisek. Neural network supported FEM analysis of elastoplastic plate bending. In BUTE Research News, pages 12-19, Budapest, Hungary, 2000.
• [23] Z. Waszczyszyn, L. Ziemiański. Neurocomputing in the analysis of selected inverse problems of mechanics of structures and materials. CAMES, 13: 125-159, 2006.
• [24] G. Yagawa, H. Okuda. Neural networks in computational mechanics. Archives of Comp. Methods in Eng., 4: 435-512, 1996.