Shape Transformation Approaches for Fluid Dynamic Optimization

The contribution is devoted to combined shape- and mesh-update strategies for parameter-free (CAD-free) shape optimization methods. Three different translate the sensitivities computed by adjoint procedures into simultaneous updates of both discretized domain are employed in combination with a mesh-morphing strategy. Considered methods involve linear Steklov–Poincaré (Hilbert space) approach, recently suggested highly non-linear p-Laplace (Banach method, hybrid variant which Hilbert space. scrutinized optimizing power loss two-dimensional bent duct flow using an unstructured, locally refined grid that initially displays favorable properties. Optimization results compared respect convergence, computational effort, preservation mesh quality during sequence. Results indicate all reach, approximately, same converged optimal solution, reduces objective function about 18% this classical benchmark example. However, as regards quality, more advanced Banach space advantageous comparison even when update performed save costs. In specific, while cost method 3.5 2.5 times pure respectively, metrics 2 1.7 improved respectively.