Neural network assisted crack and flaw identification in transient dynamics

• Autorzy: Stavroulakis G. E. , Engelhardt M. , Likas A. , Gallego R. , Antes H.

Warianty tytułu

PL

Identyfikacja pęknięć i wad strukturalnych w stanach nieutalonych za pomocą sieci neuronowych

• Języki publikacji: EN

Abstrakty

EN

Crack and flaw identification problems in two-dimensional elastomechanics are numerically studied in this paper. The mechanical modelling is based on boundary element techiques, with special care of hypersingular issues for the cracks. The possibility of partially or totally closed cracks (unilateral contact effects) is taken into account by linear complementarity techniques. Backpropagation neural networks are used for the solution of the inverse problems. For dynamical problems, a suitable preprocessing of the input data enhances the effectiveness of the procedure. For the two-dimensional examples presented here, the proposed method has similar performance for classical crack and flaw identification problems. The identification of unilateral cracks is a considerably more difficult task, which nevertheless, can also be solved by the same method, provided that a suitable dynamical test loading is applied.

PL

W pracy zajęto się problemem identyfikacji pęknięć i innych uszkodzeń strukturalnych w dwuwymiarowym stanie odkszałceń sprężystych za pomocą metod numerycznych. Modelowanie mechaniczne oparto na metodzie elementów brzegowych ze szczególnym uwzględnieniem kwestii osobliwości pęknięć. Możliwość powstawania częściowo lub całkowicie zamkniętych pęknięć wprowadzono poprzez wykorzystanie liniowej metody komplementarności. Dla rozwiązania zagadnień odwrotnych użyto sieci neuronowych ze wsteczną propagacją. W zgadnieniach dynamicznych efektywność zaproponowanej procedury zwiększono poprzez odpowiednią wstępną obróbkę danych wejściowych. Na przykładzie dwuwymiarowych modeli opisywanych w pracy stwierdzono podobną skuteczność metody, jak w przypadku klasycznego zagadnienia identyfikacji wad strukturalnych. Wykazano, że identyfikacja jednostronnych pęknięć, która jest znacznie trudniejszym zadaniem, jest możliwa za pomocą zaprezentowanej metody, jeśli do analizy modelu przyjąć odpowiednio dobrane obciążenia testowe.

Słowa kluczowe

EN

crack identification; neural networks; unilateral contact; wave propagation; nondestructive evaluation

• Wydawca: Polskie Towarzystwo Mechaniki Teoretycznej i Stosowanej
• Czasopismo: Journal of Theoretical and Applied Mechanics
• Rocznik: 2004
• Tom: Vol. 42 nr 3
• Strony: 629--649
• Opis fizyczny: Bibliogr. 27 poz., rys.

Twórcy

autor

Stavroulakis G. E.

• Department of Applied Mathematics and Mechanics, University of Ioannina, Ioannina, Greece

autor

Engelhardt M.

• Institute of Applied Mechanics, Carolo Wilhelmina Technical University, Braunschweig, Germany

autor

Likas A.

• Department of Computer Science, University of Ioannina, Ioannina, Greece

autor

Gallego R.

• Department of Structural Mechanics, University of Granada, Granada, Spain

autor

Antes H.

• Institute of Applied Mechanics, Carolo Wilhelmina Technical University, Braunschweig, Germany

Bibliografia

• 1. Alessandri C., Mallardo V., 1999, Crack identification in two-dimensional unilateral contact mechanics with the boundary element method, Computational Mechanics, 24, 100-109
• 2. Antes H., Panagiotopoulos P.D., 1992, The Boundary Integral Approach to Static and Dynamic Contact Problems. Equality and Inequality Methods, Birkh¨auser, Basel-Boston-Berlin
• 3. Dom´ınguez J., 1993, Boundary Elements in Dynamics, Computational Mechanics Publications, Southampton and Elsevier Applied Science, London
• 4. Engelhardt M., 2004, PHD Thesis, Technical University of Braunschweig, Germany (in preparation)
• 5. Fedelinski P., Aliabadi M.H., Rooke D.P., 1994, Dynamic stress intensity factors in mixed mode, in Boundary Elements XVI, Brebbia C.A. edit., Comp. Mech. Publications, Southampton, 513-520
• 6. Gallego R., Domınguez J., 1995, HBEM applied to Transient Dynamic Fracture, Proc. ICES’95, Hawaii, USA, Edit. S.N. Atluri
• 7. Gallego R., Dom´ınguez J., 1996, Hypersingular BEM for transient elastodynamics, International Journal for Numerical Methods in Engineering, 39, 10, 1681-1705
• 8. Gallego R., Dom´ınguez J., 1997, Solving transient dynamic crack problems by the hypersingular boundary element method, Fatigue and Fracture of Engineering Materials and Structures, 20, 5, 799-812
• 9. Guiggiani M., 1992, Direct evaluation of hypersingular integrals in 2D BEM, Notes on Numerical Fluid Mechanics, 33, W. Hackbusch, edit., Vieweg, Braunschweig, 23-34
• 10. Guz A.N., Zozulya V.V., 2002, Elastodynamic unilateral contact problems with friction for bodies with cracks, International Applied Mechanics, 38, 8, 895-932
• 11. Liang Y.C., Hwu C., 2001, On-line identification of holes/cracks in composite structures, Smart Materials and Structures, 10, 4, 599-609
• 12. Likas A., Karras D., Lagaris I.E., 1998, Neural network training and simulation using a multidimensional optimization system, Int. J. of Computer Mathematics, 67, 33-46
• 13. Mansur W.J., 1983, A time-stepping technique to solve wave propagation problems using the Boundary Element Method, Ph.D. Thesis, Universiy of Southampton, U.K.
• 14. Oishi A., Yamada K., Yoshimura A., Yagawa G., 1995, Quantitative nondestructive evaluation with ultrasonic method using neural networks and computational mechanics, Computational Mechanics, 15, 521-533
• 15. Portela A., Aliabadi M.H., Rooke D.P., 1992, The dual Boundary Element Method: effective implementation for crack problems, Int. J. Numer. Meth. Engineering, 33, 6, 1269-1287
• 16. Rus G., Carlborg, 2001, Numerical methods for nondestructive identification of defects, Doctoral Thesis, Departamento de Mecanica de Estructuras, Universidad de Granada, Spain
• 17. Rus G., Gallego R., 2002, Optimization algorithms for identification of inverse problems with the boundary element method, Engineering Analysis with Boundary Elements, 26, 4, 315-327
• 18. Saez ´ A., Gallego R., Dom´ınguez J., 1995, Hypersingular quarter point boundary elements for crack problems, Int. J. Numer. Methods Engineering, 38, 1681-1701
• 19. Stavroulakis G.E., 1999, Impact-echo from a unilateral interlayer crack. LCP-BEM modelling and neural identification, Engineering Fracture Mechanics, 62, 2-3, 165-184
• 20. Stavroulakis G.E., 2000, Inverse and Crack Identification Problems in Engineering Mechanics, Kluwer Academic Publishers, Dordrecht, and Habilitation Thesis, Technical University of Braunschweig, Germany
• 21. Stavroulakis G.E., Antes H., 1997, Nondestructive elastostatic identification of unilateral cracks through BEM and neural networks, Computational Mechanics, 20, 5, 439-451
• 22. Stavroulakis G.E., Antes H., 1998, Neural crack identification in steady state elastodynamics, Computer Methods in Applied Mechanics and Engineering, 165, 1/4, 129-146
• 23. Yagawa G., Okuda H., 1996, Neural networks in computational mechanics, Archives of Computational Methods in Engineering, 3, 4, 435-512
• 24. Yusa N., Cheng W., Chen Z., Miya K., 2002, Generalized neural network approach to eddy current inversion, NCT and E. International, 35, 609-614
• 25. Zeng P., 1998, Neural computing in mechanics, ASME Applied Mechanics Reviews, 51, 2, 173-197
• 26. Zgonc K., Achenbach J.D., 1996, A neural network for crack sizing trained by finite element calculations, NDT and E. International, 29, 3, 147-155 2
• 27. Ziemianski L., Piatkowski G., 2000, Use of neural networks for damage detection in structural elements using wave propagation, In: Computational Engineering using Metaphors from Nature, Edit. B.H.V. Topping, Civil-Comp Press, Edinburgh, U.K.