Density Filter Control of Thickness-to-Length Change of Composite Structures
2013 (English)Conference paper, Presentation (Other academic)
The homogenized material optimization (HMO) problem is a novel structural optimization problem that we have developed for optimization of fiber reinforced composite structures. In the HMO problem we apply a smeared-out approach to model the material properties of fiber reinforced composite materials. The objective of the HMO problem is to maximize the stiffness of a composite structure by means of finding the optimal distribution of composite material, belonging to a fixed set of fiber orientations, across the design domain. In order to obtain manufacturable solutions, we have introduced a linear density filter as a restriction method to control the thickness variation across the design domain. To examine the effect of the density filter on the thickness variation and the objective function value of composite structures, obtained in the HMO problem, we have performed numerical tests for different load cases, mesh densities and range of the filter radius.
It is observed that for the present problem the thickness variation was mesh-independent. Both the thickness variation and objective function value depend on the load case used in the HMO problem. For all load cases the thickness variations exhibits an approximately piece-wise linear behaviour for increased filter radius. Furthermore, it was observed that an increase of filter radius would result in an moderate increase in objective function value for the solutions obtained from the HMO problem. From these results we conclude that by using a density filter, the HMO problem can be used to obtain manufacturable designs for composite structures.
Place, publisher, year, edition, pages
Homogenized material optimization, composite structures, density filter, structural optimization
Engineering and Technology
IdentifiersURN: urn:nbn:se:liu:diva-102314OAI: oai:DiVA.org:liu-102314DiVA: diva2:676159
10th World Congress on Structural and Multidisciplinary Optimization, May 19-24, 2013, Orlando, Florida, USA