6. Summary of results

This test case meets its primary objective: to test functionality HYDR_VGM. In addition, it makes it possible to have a classical modeling problem \(«\mathrm{BO}-\mathrm{BG}»\) with steep front treatment. We do not have reference solutions to compare ourselves with, however the results obtained have the classic appearance of this type of resolution. So we make it a case of non-regression.

2 numerical schemes are tested here: finite elements and decentralized finite volumes (edges). If the overall appearance of the results is the same for the 2 diagrams, it is clear that the finite volumes greatly reduce the gas peak observed at the interface between the materials. While this gas peak has a physical reality, its magnitude is considerably amplified by finite element schemes.

Decentralized edge finite volume schemes are recommended for this type of problem (less expensive than off-center meshes).