1.3 Hotspots in topology optimization for AM
1.3.1 Optimization design of multiscale structures
By means of inverse homogenization and other modeling approaches (geometry modeling, parametric design, artificial intelligence), a variety of microstructures have been designed. As microstructures show great superiority in lightweight and multifunctional properties, the design of multi-scale structures that utilize microstructures as basic material units has been increasingly focused. The advanced additive manufacturing technology facilitates the physical prototyping that also promotes the exploration of multi-scale structures. The existing strategies for multi-scale structure optimization include:
● Lattice structure topology optimization.
● Concurrent optimization of freeform macrostructures and microstructures.
The main difference between these two strategies lies in the different generation strategies of microscale structures. The first method uses existing parameterized microstructures. This method is efficient in multi-scale structure design since in-process homogenization is not required, but to avoid the connectivity problem, only a limited types of microstructures are adopted, indicating a restricted optimization space. The other method generates microstructures via freeform topology optimization. This method could fully explore the optimization space, but the computation burden is heavy.