Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Proceedings
ISSN 1759-3433 CCP: 79
PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY Edited by: B.H.V. Topping and C.A. Mota Soares
Paper 262
Solid-Shell Finite Elements in Shell Structural Problems: A Comparative Analysis M.P.L. Parente+, A. Roque+, R.A. Fontes Valente*, R.M. Natal Jorge+, R.J. Alves de Sousa*, J.M.A. César de Sá+ and J.J. Grácio*
+IDMEC - Pólo FEUP, Faculty of Engineering, University of Porto, Portugal
, "Solid-Shell Finite Elements in Shell Structural Problems: A Comparative Analysis", in B.H.V. Topping, C.A. Mota Soares, (Editors), "Proceedings of the Seventh International Conference on Computational Structures Technology", Civil-Comp Press, Stirlingshire, UK, Paper 262, 2004. doi:10.4203/ccp.79.262
Keywords: finite elements, solid-shell, enhanced assumed strains, shell structures, transverse shear locking, volumetric locking.
Summary
Reliable numerical simulation of sheet metal forming processes is based on
the correct reproduction of the high stress and strain gradients usually involved in
those industrial processes. A crucial point in the computational modeling of
elasto-plastic behavior, on the other hand, relies on the proper choice of the finite element
formulation to be employed. During the last decades, shell elements
represented the priority choice among end-users, mainly driven by the options
offered by explicit finite element codes. Within this aspect, adoption of elements
based on reduced integration was the trivial option, in order to fully benefit from the
small CPU time. Nevertheless, as higher computational power has became
increasingly cheaper and accessible, researchers' attention turned back to full
integration-based elements, within a sounder theoretical framework and naturally
free of hourglass instabilities, common in reduced schemes.
In this sense, the present paper aims to show one of the current research lines of the authors, focusing on the development of reliable and efficient shell and solid-shell elements. What distinguishes the present formulation from the majority of those previously presented in the literature is the fact that solely the Enhanced Assumed Strain (EAS) method is employed in order to improve the performance of conventional displacement-based elements. The EAS approach is responsible for the elimination of the transverse shear strain locking pathology, dominant in shell elements as thickness values tend to zero. In relation to solid-shell formulations, the EAS method provides an unified treatment of not only the transverse shear locking but also the volumetric locking, arising from the incompressibility constraints directly coming from the constant volume requirements within the plastic range, in elasto-plastic simulations. In both cases, the conventional formulation is enriched by a set of internal, element-wise, variables, consistent with the integration rules employed. These variables do not affect the overall dimensions of the equation system to be solved, once they are completely eliminated, at the element level, by static condensation procedures. In the end of the paper, it is shown a numerical benchmark involving a small thickness shell structure, and including large displacements, large rotations and non-linear material constitutive laws. The example is treated by means of a coarse mesh, consisting of both regular and distorted meshes. Allied to the presence of a curved reference surface, elements pose themselves in warped configurations, which frequently introduce convergence and accuracy difficulties to EAS-based finite elements formulations. However, this does not seem to be the case with the present elements, which led to reliable results for all the situations analysed. The results obtained are compared with solutions coming from reduced numerical integration procedures, with the obtained results being equivalent. Nevertheless, the formulation presented appears to be more general, namely in the treatment of a general type constitutive laws and in the avoidance of parasitic (hourglass) numerical modes, usual in reduced integration schemes. The present work is grounded on the work of the authors in the field, represented by the following references, which also contains the details about the theoretical and algorithmic aspects of the formulation, as well as an extensive set of linear and nonlinear benchmark tests. References
purchase the full-text of this paper (price £20)
go to the previous paper |
|