Gas Concentration errors with SORSCA and FLUINJ

Emmanuel Resch

Hi, I am having some confusing problems with mass removal and redistribution. The physical setup is a 2d structured mesh, quasi-rectangular domain (one large rectangle for channel flow, with a smaller rectangle on top which represents a porous ceramic), with helium as a background fluid and hydrogen and water as scalars. The cells along the upper boundary of the smaller rectangle have a mass removal condition (using FLUINJ) and scalar redistribution (using SORSCA) such that hydrogen is consumed and water is produced. The larger rectangle has standard inlet and outlet conditions. There are symmetry conditions on the sides and bottom, so that 2-D axisymmetric flow is simulated. I have density set as ideal gas as a funtion of T and p.

My UDFs seem to be working fine. The problem is that the overall mass balance doesn't make any sense. It seems that the outlet density and inlet density are somehow fixed to not vary, although there is variation within the domain. This is the same for the mass fractions and concentrations for inlet and outlet for both scalars remain the same. What should be happening, from inlet to outlet, is that hydrogen is to be consumed along the length, water should be produced along the length, and the fluid density should thus increase along the length.

Does anyone have any suggestions?