Discussion Forum

Hi, I'm simulatin air through a narrow rectanguclar conduct. Before the conduct i added an inlet region so the constant velocity inlet condition is far from the wall condition. The walls of the entrance region have symmetry condition so fluid won't have any resitance in this region, and the wall of the condurc are tangent to those symetry BC. The problem is that around the transition from simmetry to no slip condition. i have massflow oscilation up to 2% of the total mass flow. I've diminish this efect by reducing the mesh elments size, but is still far from mass conservation.

i'm working with incompresible flow, the mass flow for any given x is estimated as the integration of rhou (u is x component in velocity and rho the density wich is constant) in the transversal sección (yz planes) i used 2 nested general projetions to reduce the 3d data to 1D, and so i ploted (VinletAinlet - intY(intZ(u))*100, so it should be 0 all the way. I checked with post processing derived values integrals so i'm sure is not a problem with the general projections. The problem is not the Global mass conservation but the mass conservation inside the domain. The starting peak is where the wall change from symmetry BC to no slip wall

The images show the sideview of the domain and the mass flow socilation % relative to the inlet mass flow, so is 0 at inlet The conduct starts at x=0 and ends at x=0.5. The flow direcction is X, Reynolds is 400 and the width of the canal is about 0.006.

I'm thinking that these may be a finite element problem and i should try with a finite volume solver.

Any thoghts are welcome and thanked