Development of efficient algorithms for fluid -structure interaction framework and its applications

Computer-based simulation tools are becoming integrated to solve multiphysics and interactions between different disciplines. One of such problems that is of importance is fluid structure interaction (FSI). The goal of this study is to develop efficient and robust algorithms for a FSI framework. Due to the advantages of reusability of well-validated simulation codes for fluid and structural analysis, loosely-coupled methods are now dominant over directly-coupled methods. In loosely-coupled methods, it is important to transfer data efficiently and accurately between inherently unmatched grids used in different disciplines. In addition, efficient moving grid algorithms and time mapping techniques are essential. A new hybrid interpolation method for deformation mapping is suggested. The method provides efficient, accurate, and smooth interpolations in comparison to traditional methods. A partitioned volume grid movement algorithm is developed in which a local reference system is used for the viscous layer, and existent linear and semi-torsional spring analogy is used for the inviscid region. The algorithm reduces computational time significantly, and it prevents vertex-to-edge or vertex-to-face interpenetrations incurred in a large deformation by traditional spring analogy methods. Numerical results with test functions and an aircraft wing deformation illustrate the efficiency and accuracy of the suggested deformation mapping and volume grid movement methods. The validated algorithms are integrated into a FSI framework. Three FSI applications are demonstrated: pulsatile blood flow in an artificial straight/curved stenotic artery with considering vessel compliance; pulsatile blood flow in in-vivo patient-specific femoral artery bifurcation; and an aircraft wing flutter.