Verification and validation of a finite volume immersed boundary method for the simulation of static and moving geometries

This study reports the verification and first experimental validation of an advanced cut-cell finite volume immersed boundary method (IBM). This novel IBM approach was developed by Jasak at the University of Zagreb and released under the name Immersed Boundary Surface Method (IBS). It is implemented and publicly available within the open-source computational fluid dynamics framework foam-extend. In IBS, a temporarily immersed boundary patch is introduced into the cut-cell, resulting in a grid that is virtually boundary-fitted. This allows the simulation of any complex static or moving geometry on an unstructured (collocated) computational mesh without interpolation or simplification in the vicinity of the immersed boundary. In this study, two benchmark simulations were performed to assess the accuracy of the method and validate its implementation for laminar flows of both, Newtonian and non-Newtonian fluids. On the one hand, an extrusion die was used to compare the results of IBS with those of boundary-fitted (BF) simulations with respect to different viscosity models. On the other hand, IBS was validated experimentally with a modified rheometer set-up. Compared to BF simulations, the results of IBS are in very good agreement for both Newtonian fluids and non-Newtonian fluids. Furthermore, with respect to experimental measurements, IBS reproduces the measured torques very well and is therefore able to reproduce physical problems realistically. With these results IBS is verified and also experimentally validated for the first time.