[mireis]

I’m a student and I would want to analyze a part of a pressure regulation chain. To do this I consider just a plenum with a pipe for the inlet and a pipe for the outlet. The latter narrows at the end to impose a mass flow rate of 5 mg/s at the outlet whereas the mass flow rate at the inlet is around 3000 mg/s. What I want to obtain is the difference in pressure between a point near the outlet and a point in the middle of the plenum.
I use mass-flow-inlet and pressure-outlet as conditions. In the latter I impose a target-mass flow rate of 5mg/s. With a net flow rate like this one (2995 mg/s) the logical thing would be that the pressure at the plenum and its mass would increase almost in a linear way. Nonetheless what I get from the analysis is that the pressure rises in the first 100ms and then it converges to a constant value. Moreover if I calculate the mass of the model at different instants of time (using the average density that Fluent gives) the mass increases rather less than it was expected (net mass flow rate multiplied by time). It is as if there was a drain inside the plenum!
Is it a problem of convergence? I use a time step of 0.0005s and 100 iterations per time step at the beginning, then it says it converges after 20 iterations. I’m completely lost but… could it be related to the fact that Fluent evaluates convergence comparing the mass flow rate that enter and leave each cell (continuity equation)? In a tutorial about a stationary case I read that a way to check convergence was to report the difference between the flux at inlet and the flux at outlet... but in this case, where the fluid is compressible there is an accumulation of mass inside the plenum!!

Thanks!

[mireis]

Do you think it could be a problem related to the quality of the grid? what is the maxim skewness I should have in a transient simulation for a compressible flow?