This presentation was made at the NAFEMS European Conference on Simulation-Based Optimisation held on the 15th of October in London.
Optimisation has become a key ingredient in many engineering disciplines and has experienced rapid growth in recent years due to innovations in optimisation algorithms and techniques, coupled with developments in computer hardware and software capabilities. The growing popularity of optimisation in engineering applications is driven by ever-increasing competition pressure, where optimised products and processes can offer improved performance and cost-effectiveness which would not be possible using traditional design approaches. However, there are still many hurdles to be overcome before optimisation is used routinely for engineering applications.
The NAFEMS European Conference on Simulation-Based Optimisation brings together practitioners and academics from all relevant disciplines to share their knowledge and experience, and discuss problems and challenges, in order to facilitate further improvements in optimisation techniques.
In the constant effort of optimizing structures in terms of load-balanced design and extremely lightweight components, composite structures are a key player, offering the possibility of tailoring the material to the application. To be able to fully exploit the potential of using innovative materials it is essential to couple structural and material optimization which is contrary to the present design practice. The introduction of composite materials as part of the design formulation for structural optimization is both to determine the optimal spatial distribution as well as the optimal use of the material, i.e. the orientation and anisotropy of the local material tensor which is controlled by the composite microstructure. In practice, the local control over the microstructure is rather limited. The only eligible design factors are: the volume ratio of the matrix and the reinforcement, the orientation of a given mi-
crostructure and the topology of the microstructure, all within a xed set of design variables.
As an engineering approach, which so far has been focused on the optimization of laminated composites, the design problem can therefore be reduced to finding the optimal distribution of orientation angles of a material reinforcement in order to satisfy a global structure design objective. A lot of eort has been put on developing a general continuum based structural optimization method, especially for the simultaneous structural and material optimization, in such a way that the non-convex nature of the problem, resulting in the difficulty of avoiding local optima, is overcome. Existing approaches for material optimization focus on directly finding a physically meaningful solution based on a limited number of predefined candidate angles (Discrete Material Optimization), with the risk of local optimum solutions. However, other approaches which avoid the local optimum problem by relaxation of design space (Free Material Optimization) face the problem that the optimization may yield a theoretically optimal structure but not always a physically feasible structure, especially for more complex structures or loading scenarios.
In the current work a new approach for the simultaneous optimization of structural layout and material, dealing with structures subjected to multiple loads, is proposed. It is
based on the "soft-kill" BESO (Bi-Evolutionary Structural Optimization) and adresses two aspects of structural optimization: the design of the global geometry (topology) and the more or less detailed design of the material itself in terms of orientation and anisotropy of the local material tensor. The new method opens up the design space as the material can be directly optimized for the functional needs at the structural scale. It yields physically realistic material congurations and is based on a reasonable amount of design variables but without adding unnecessary restrictions to the design space. The concept is illustrated by some example problems for laminated composites. It should be noted, however, that the concept can be generalized to 3-dimensional composite topologies.
|Author||Christa Lang. M|
|Date||15th October 2019|