Computational Design Methods for the Optimization of the Density Distribution in Graded Concrete Components under static and dynamic loading

点击翻译按钮获取中文摘要

This research aims to develop a numerical design tool for determining the optimal density distribution within functionally graded structural concrete components subject to static and dynamic actions. This tool should be used to design functionally graded frame components for dynamic loading, which are then to be produced at ILEK and tested in Berkeley. Functionally graded structural components are characterized by location-dependent, continuous changes in their properties. Density gradation of concretes (i.e. a continuous change in the porosity within the component) serves to accurately adjust the properties of the material to the actions it is subjected to. High-impact zones are made from high-performance concrete whereas low-impact zones consist of porous concrete. This approach makes it possible to save up to 60% of the mass of the material, such as in floor structures subject to bending. Initial tests showed that this new lightweight construction method was associated with a high energy absorption capacity.Only after developing numerical design methods will it become possible to determine continuous, three-dimensional property gradation patterns. The optimization steps implemented in the preliminary research were limited on account of the simplified geometry, the linearly elastic material assumption, and the restriction to optimizing component rigidity. Methods were limited to the algorithms, design variables, and target functions provided in the commercially available optimization tool (Abaqus Atom). No alteration or adjustment of these parameters was possible. The design tool to be developed within the scope of the applied-for research grant should also enable thermal conductivity optimization whilst considering static restrictions and make it possible to design a density layout subject to dynamic actions. The design tool should then be made available free of charge to make a contribution to disseminating the new approach. Some of the optimization algorithms and simulation methods, such as the Concrete Damaged Plasticity Model to describe the non-linear material behavior of concrete, were developed at the University of California at Berkeley. Prof. Filippou has been conducting research on describing concrete structures subject to static and dynamic actions since 1983. The team of scientists tasked with the related activities and the existing infrastructure at the host university thus provide the perfect setting to successfully complete the envisaged research project. Once completed, this project will provide design methods to determine three-dimensional property gradations for structural concrete components subject to static, thermal, and dynamic actions. Besides the automation of manufacturing processes currently under development at ILEK, this design tool is the missing link to reliably produce structural concrete components with free, three-dimensional density gradation.