@inproceedings {pub3148,
	title = {Evolutionary Crashworthiness Topology Optimization of Thin-Walled Structures},
	author = {Mariusz Bujny AND Nikola Aulig AND Markus Olhofer AND Fabian Duddeck},
	year = {2016},
	month = {July},
	abstract = {As in many other disciplines, also in crashworthiness, the extensive growth of computers{\textquoteright} power led to the development of techniques for numerical simulations. In
particular, this allows to use numerical optimization methods to develop better structures and shorten the vehicle design cycle, what is a must in case of the hard
competition on the car market. A basis for most of the state-of-the art methods for crashworthiness topology optimization, form so-called voxel elements, being threedimensional, regular brick finite elements. The basic idea in such cases is stated as follows: Firstly, define the design domain in the space that is not occupied with nonstructural vehicle elements such as wheels, engine, etc.; Secondly, fill the volume of the design space with voxels; Thirdly, eliminate redundant voxels according to the optimization method. This results in creation of so-called zigzag structures that can be used as a reference for positioning of the structural beams. Such a design is assumed to be optimal with respect to the given objectives (e.g. energy absorption, plastic deformation, etc.). On the other hand, main car structural components are made of thinwalled panels and beams. In such a case plastic buckling is the principal phenomenon that influences the total energy absorption. In the optimization process based on voxel elements, structures made of thin metal sheets cannot be obtained, which leads to
completely different phenomena in energy absorption, which do not correspond to the buckling of thin-walled structures. As a result, the use of an optimized design obtained from any voxel-based optimization method as an inspiration for final thin-walled structure is questionable and alternative methods have to be developed. We propose a novel approach using evolutionary algorithms for optimization of thin-walled structures. For evaluation of the method, a 2D transverse bending of a rib-reinforced thin-walled structure is considered. Parameterization of the design is realized through defining the position, orientation, length and thickness of each reinforcing rib. The ribs can cross each other and join if they are sufficiently close to each other. As an optimization method both standard Evolution Strategy (ES) and the state-of-the-art Covariance Matrix Adaptation Evolution Strategy (CMA-ES) are used and their performance is compared. The results show that evolutionary optimization algorithms can be efficiently used for crashworthiness topology optimization of thin-walled structures. },
	publisher = {University of Leeds},
	booktitle = {Proceedings of the 11th ASMO UK/ISSMO Conference}
}