rhoPimpleFoam with fixedFluxPressure + flowRateInletVelocity --> p diverges

tatu · February 24, 2013, 14:20

Hello!

I'm trying to specify mass flow at compressible subsonic inlets, and I have understood that this can be done using fixedFluxPressure and flowRateInletVelocity. This seems to work fine for rhoCentralFoam, but not for rhoPimpleFoam; unless I assign a very low mass flow (flowRate), p diverges (to negative and positive value) usually on the first time step. I'm using OF 2.1.1.

Any ideas how to overcome the p divergence? For fixedFluxPressure I need to store the matrix rAU = 1/UEqn.A(). I guess I should calculate rAU for fixedFluxPressure after the U is completely solved for that time step, i.e. after line 85 in pEqn.H, right?

By the way, in 2.1.x there is a new feature in fixedFluxPressure: pressure diffusivity. I guess this is meant to "smear" the pressure gradient, or is it? Could this be of help for my case? Also, instead of dividing the phi with density (in compressible solvers), the solver seems to use the field phi*UEqn.H()/UEqn.A() for adjusting the pressure gradient.

Tatu

styleworker · March 21, 2013, 15:10

In alphaContactAngleFvPatchScalarField.H is a note:

"Note that if any of the first three options are used the boundary conditionon p_rgh must set to guarantee that the flux is corrected to be zero at the wall" (fixedFluxPressure)

Does it mean the flux is corrected to be zero across the wall, or along the wall?

February 24, 2013, 14:20	rhoPimpleFoam with fixedFluxPressure + flowRateInletVelocity --> p diverges	#1
tatu New Member Tatu Pinomaa Join Date: Oct 2012 Location: Finland Posts: 16 Rep Power: 13	Hello! I'm trying to specify mass flow at compressible subsonic inlets, and I have understood that this can be done using fixedFluxPressure and flowRateInletVelocity. This seems to work fine for rhoCentralFoam, but not for rhoPimpleFoam; unless I assign a very low mass flow (flowRate), p diverges (to negative and positive value) usually on the first time step. I'm using OF 2.1.1. Any ideas how to overcome the p divergence? For fixedFluxPressure I need to store the matrix rAU = 1/UEqn.A(). I guess I should calculate rAU for fixedFluxPressure after the U is completely solved for that time step, i.e. after line 85 in pEqn.H, right? By the way, in 2.1.x there is a new feature in fixedFluxPressure: pressure diffusivity. I guess this is meant to "smear" the pressure gradient, or is it? Could this be of help for my case? Also, instead of dividing the phi with density (in compressible solvers), the solver seems to use the field phi*UEqn.H()/UEqn.A() for adjusting the pressure gradient. Tatu

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
fixedFluxPressure in viscoelasticFluidFoam	Joana	OpenFOAM Running, Solving & CFD	0	November 9, 2012 12:03
Questions about buoyantPimpleFoam and rhoPimpleFoam	Mojtaba.a	OpenFOAM Running, Solving & CFD	6	August 1, 2012 04:50
1.7.x -> buoyantPimpleFoam -> hRhoThermo -> incompressible and icoPoly3ThermoPhysics?	will.logie	OpenFOAM Programming & Development	1	February 16, 2011 20:52
1.7.x -> buoyantPimpleFoam -> hRhoThermo -> incompressible and icoPoly3ThermoPhysics?	will.logie	OpenFOAM	0	December 16, 2010 07:08
CAD -> gMsh -> enGrid -> OpenFOAM Problem	AlGates	OpenFOAM	7	August 6, 2010 12:46

March 21, 2013, 15:10		#2
styleworker Member Join Date: May 2012 Posts: 55 Rep Power: 14	In alphaContactAngleFvPatchScalarField.H is a note: "Note that if any of the first three options are used the boundary conditionon p_rgh must set to guarantee that the flux is corrected to be zero at the wall" (fixedFluxPressure) Does it mean the flux is corrected to be zero across the wall, or along the wall?