[lavdwall]

Hi everyone,

I am facing a problem with the reactingTwoPhaseEulerFoam solver (OF-4.1). For some reason, the total volume fraction of particles in my computational domain keeps increasing. As the density of my particles is constant, this means that mass isn't conserved.

I first thought it was a problem with my solver (I made some adjustments to the original reactingTwoPhaseEulerFoam solver), but then I did the fluidized bed tutorial (RANS) again with the original solver and I noticed the same phenomenon. OK, I know for the tutorial, it can be related to very basic (read: less accurate) mesh, discretization schemes etc., but in that case it is still weird that volume fraction is mainly increasing, and never decreasing, it should be more or less constant anyhow... See figure in attachment for the tutorial case (I basically plotted every value that comes after 'alpha.particles volume fraction = ' in the logfile).

I thought it could be related to boundary conditions. In my case, at the inlet, I only feed gas (U and alpha are fixed at 0 for particles). At the outlet, I use the inletOutlet boundary condition for alpha (inletValue 0 for particles, 1 for gas) and inletOutlet boundary condition for U (inletValue 0 for both phases). That should keep my particle mass constant, or even decreasing as particles can/may flow out.

I have tried with firstOrder and secondOrder divSchemes, Euler and backward ddtSchemes, uncorrected/limited/corrected snGradSchemes, switched from smoothSolver to GAMG for every variable, but in none of those cases the particle volume fraction remains constant. As in the attached figure for the tutorial case, the increase in particle volume fraction is only small, but after a lot of iterations/time, it is significant. For example, in my case, at some point I have 2.2grams of particles in my reactor, while I am only feeding 2grams initially...

Has anyone experienced the same problem? Is it a bug? Or how can I solve it?

Greets,
Laurien

[ssss]

Well it seems that you have an increase of the volume fraction of less than a 0.05%, I wouldn't call that a problem. If you think there is an issue, you should report your bug to OpenFOAM's MANTIS bug report system

[lavdwall]

OK, I agree that 0.05% is not much, but if I let that case run for a longer time, it keeps increasing so in the end it is not negligible anymore...
As I said in my previous post, for my actual problem, I need to simulate lots and lots more of iterations and after a day of simulation, my particle mass/volume fraction has increased by over 10% of the initial value, which is quite significant I think.

[yfjok22]

Originally Posted by lavdwall

OK, I agree that 0.05% is not much, but if I let that case run for a longer time, it keeps increasing so in the end it is not negligible anymore...
As I said in my previous post, for my actual problem, I need to simulate lots and lots more of iterations and after a day of simulation, my particle mass/volume fraction has increased by over 10% of the initial value, which is quite significant I think.

Hi,buddy

My guess is the evaporation and condensation in the process make the density changes but not directly related to alpha. Because based on reading of the code, I somehow feel in this solver, volume fraction alpha and mass fraction Y are not coupled with each other. So it may make alpha non-conservative somehow
I don't know if it's right, welcome for discussion of it

[saddy]

hey i was discussing with my colleaguesand the conclusion was:
MASS IS NOT CONSERVED HERE IF SYSTEM IS CLOSED, SO if use it make sure your system is open