[waq]

Hi everybody,

I am confused these days with the pressure convergence when using K-Omega SST model and rhoSimpleFoam solver.I have to do two series of simulations for internal pipeflow and then compare them with eachother. The cases are very simple, the mass flow increases at inlet of the pipe from 30g/s to 70g/s with a increasing step 20g/s and at the outlet is the static pressure of 1 bar specified. The diameters of the pipes are 30mm and 50mm. Their length is the same 4 meters.
In the simulations of 50mm pipe everything goes well. But when the pipe changes to 30mm diameter, start the pressure residual to diverge. I have tested many ways to fix this problem, included change the y+ value(0.15, 1, 10 tested), change the boundary conditons to wallfunction and no wallfunction, change the initial conditions to with massFlowRate and velocity, change the solver of p to PCG and a combination of PCG and GAMG, change the relaxationsfactor of p to 0.5 and 0.7, change the thermophysicalProperties to sutherland Transport, hconst , econst and janaf. But all the measures failed....
At last I have specified both the inlet and outlet pressure and give zeroGradient settings for their velocity, hope to see, until which pressure difference will diverge the pressure residual. The result indicated that, until about 15400 Pa pressure difference start the residual to diverge.

Thank you very much for your help.

[waq]

This is the screenshot for the Mesh.