Topology optimization is an established tool for finding the optimal geometries for a specific initial design space, boundary conditions and constraints. When using topology optimization for problems where the design space is not exactly defined but given as a range in which values of individual dimensions are allowed to be, we do not know which design space is optimal for the best final solution. It is necessary to search the entire available design space to find the optimal solution. This process is very time-consuming since each combination of the design space requires its own topology optimization. We tested the use of design of experiments in combination with topology optimization to determine the region of the optimal design space.
The method was discussed and verified on the example of a beam, where certain parameters of the design space and boundary conditions were not exactly determined. We performed several topology optimizations with design spaces of different dimensions in the Ansys 2024 R1 software. We found that changing values of different dimensions has a different effect on the final mass of the solution. For the case under consideration, we identified the optimal design space and found that the design spaces with the largest or smallest volume do not give the best results.
|
