Computational & Technology Resources
an online resource for computational,
engineering & technology publications
PROCEEDINGS OF THE FIFTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY
Edited by: B.H.V. Topping, G. Montero and R. Montenegro
A New Adaptive Mesh Generation Strategy for Structural Shape Optimization Problems Using Evolutionary Algorithms
G. Bugeda1, J.J. Ródenas2, E. Pahl3 and E. Oñate3
1Universitary School of Industrial Technical Engineering of Barcelona, EUETIB-UPC, Spain
G. Bugeda, J.J. Ródenas, E. Pahl, E. Oñate, "A New Adaptive Mesh Generation Strategy for Structural Shape Optimization Problems Using Evolutionary Algorithms", in B.H.V. Topping, G. Montero, R. Montenegro, (Editors), "Proceedings of the Fifth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 219, 2006. doi:10.4203/ccp.84.219
Keywords: structural shape optimization, adaptive remeshing, sensitivity analysis, evolutionary algorithms, mesh sensitivity, differential evolution.
It is well accepted that evolutionary methods are a very powerful and robust tool for the solution of general optimization problems with the ability of not getting trapped in local minima, as in the case of deterministic methods. Nevertheless, the use of these methods requires the analysis of an important number of different designs. In structural shape optimization problems, the computational cost and the quality of the solutions are very much dependent on the quality of the finite element meshes used for the analysis. One important ingredient of the numerical analysis is the strategy for the generation of a proper mesh for each design. Here we can see two types of strategies:
Two optimization problems have been used to test the behavior of the proposed methodology. The first one, with known analytical solution, consists of finding the best shape for the external boundary of a pipe subjected to a uniform pressure applied over the internal circular boundary. The numerical analysis clearly shows that the results converge to the analytical solution. The second problem consist of finding the optimum shape of a fly-wheel subjected to centrifugal and tangential loads. In this case the proposed methodology has been used to optimize a preliminary layout optimization obtained by topology optimization. The results of this problem have been shown to be similar to previous results obtained by a classical deterministic shape optimization strategy based on shape sensitivity analysis .
The examples show that the integration of the adaptive remeshing strategy into the evolutionary algorithm does not affect the convergence of the optimization process and ensures a good evaluation of the objective function and the constraints for each different design.
The proposed strategy provides a control on the quality of the analysis of each design in the least expensive way because only one single analysis is performed for each different individual.
purchase the full-text of this paper (price £20)