Multi-disciplinar topology optimization via PaLS (Parametrized LevelSet)

Goal Implement a fully functional proof-of-concept workflow for topology optimization.

Tasks

[ ] Level set parametrization of initial design (user-supplied geometry) up to a specified tolerance via a optimization procedure. Ideally the nr. of RBF functions can be increased in a outer loop (macro-iteration) if the parametrization cannot fullfill the user-request tolerance for a specified nr. of shape function (note: add a user-defined "budget" to contain the nr. of shape functions).
[ ] Test different class of shape functions (NURBS, ...) to find the most appropriate class to represent complex geometries (with/without sharp features) with the least nr. of shape functions.
[ ] Test artificially augmentation of the nr. of RBF parameters (dilation factor, center of the RBF, etc...) to increase the nr. of dofs while retaining the minimal nr. of shape functions.
[ ] Workflow for topology optimization: Interface to solver: CFD (Gloria), thermal (?), structural(?)
[ ] Workflow for topology optmization: interface to modefrontier.
[ ] Define relevant test case for the following applications: CFD+structural (which specific application), CFD+thermal (heat exchanger) -> definition of the objective function, etc.

optimad / mimmo