Shape optimization using structural adjoint and RBF mesh morphing

Adjoint solvers are acquiring nowadays a growing importance in shape optimization especially when dealing with fluid dynamic applications; their use for structural optimization is however still limited. In this work an optimization workflow based on the synergic use of a structural continuum-discrete adjoint variable solver and the commercial morpher RBF Morph™ is presented. Shape sensitivity information with respect to the objective function is exported as deformation maps on the interested geometry and transferred to the morpher that, after a proper filtering and setup, allows to update automatically the numerical grid. By employing a gradient based logic it is finally possible to achieve an evolutionary optimization. The proposed method effectiveness is shown with two examples: a cantilever beam and a structural bracket.