Stress constrained topology optimization with free-form design domains

Shape optimization of Dirichlet boundaries based on weighted B-spline finite cell method and level-set function

“…Clearly, the present solution is less effective. We can resort to the level set based topology optimization method [42][43][44][45][46]. …”

Heaviside projection–based aggregation in stress‐constrained topology optimization

“…In order to show the validity and advantages of our approach to stress‐constrained problem, we compare it with the KS and p‐norm aggregations. In our implementation of the KS and p‐norm approaches, the active‐set–based strategy introduced in the work of Cai and Zhang is employed to aggregate local stresses, and the adaptive normalization scheme proposed in the work of Le et al is utilized to narrow the gap between the maximum stress and the stress limit.…”

“…Since the KS and p‐norm aggregations are the most commonly used approaches to deal with local stresses, this comparison would help to further illustrate the efficiency of the HPI approach. Based on the results in other works and our own numerical tests, the local stress constraints are subdivided into aggregation groups. The aggregation parameters for the KS and p‐norm functions are set to η =10 and P =8, respectively.…”

“…As the single aggregated global constraint cannot effectively control the local stress level, multiple global constraints are employed in the works of Le et al, Holmberg et al, and Deaton and Grandhi and different grouping strategies are also studied. To further improve the effectiveness of global constraints, the active‐set strategy can also be applied to reduce the number of stresses in each group . Recently, a p‐norm correction method is proposed to make the aggregated constraints work efficiently .…”

“…To further improve the effectiveness of global constraints, the active-set strategy can also be applied to reduce the number of stresses in each group. 14 Recently, a p-norm correction method is proposed to make the aggregated constraints work efficiently. 15 Evaluation of different aggregation approaches can be found in the work of Payten and Law, 16 in which the effect of the number of constraints and the aggregation parameters are also investigated.…”