Model Order Reduction for FSI problems: POD-based partitioned and monolithic approaches
Please login to view abstract download link
The numerical simulation of a solution of a Fluid-Structure Interaction problem is very expensive, in terms of computational time and of computer memory. This is even more evident if we take into account a parameter dependence (e.g. the Lamè constants of the solid). In the framework of making the numerical simulation more feasible, Model Order Reduction (MOR) represents a promising direction. In this talk I will present two algorithms, one of which is based on a monolithic approach and one which is based on a partitioned approach. Both these algorithms are designed within the MOR framework, and they rely on a Proper Orthogonal Decomposition technique to generate a suitable linear space of approximation. I will show some numerical results for different toy problems: trying to highlight the pros and cons of the two approaches.
