High performance topology optimization computing platform

2020 
Abstract One of the most challenging tasks in the construction industry nowadays, is to reduce the material demands and distribute, in the same time, the material among the structural system in the best possible way. Topology optimization is a design procedure that is increasingly used, to generate optimized forms of structures in several engineering fields. The current paper presents the Topology Optimization (TO) module of the High-Performance Optimization Computing Platform (HP-OCP) which focuses on civil engineering problems. More specifically the SIMP method [1] is implemented and the topology optimization problem is solved by using the OC algorithm. The HP-OCP is a platform which evaluates several objective functions, such as the volume of the structure, the compliance etc. and can solve constrained or unconstrained structural optimization problems. The above libraries are developed in C#. The core of the platform is created in such way that it can be integrated with any CAE program that has OAPI, XML or any other type of data exchange format. In the proposed work the structural analysis and design software SAP2000 is used. Theoretical aspects are discussed in order to implement the mathematical formulation in a commercial software. Basic and specific features are applied and representative examples are performed. One of the highlights of the proposed work is that the above module can be used for all kind of finite elements. Benchmark tests are presented with structures that are simulated by 2D plane-stress elements, 3D-solid elements and shell elements. Furthermore, it is independent of the type of the mesh, structured or unstructured, so both examples are presented. In the proposed work a powerful tool for both architects and civil engineers is introduced. The analysis and design of the structures are performed in SAP2000 software, in order to achieve a realistic result that could be a solution for a real-world structure.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    6
    References
    3
    Citations
    NaN
    KQI
    []