[pauljwells]

I have a tet mesh - 800,000 cells - that I want to run steady-state compressible. The mesh is too coarse to capture any boundary layer effects, but in Fluent I have run it viscous as it seems to help with convergence.

I have got as far as converting the mesh and playing with boundary conditions in FoamX, but not really understanding what I'm doing even after trying the tutorials.

The mesh is a jet aircraft engine intake in a freestream. In Fluent I would set pressure inlets and outlets for the farfield, and a pressure outlet for the engine face. I would calculate the static pressures from compressible flow functions.

As a complete newbie I'm at a bit of a loss where to start with Foam, even down to not being sure which solver to use. Any patient hints would be gratefully received!

Once I have a solution I then want to perform a lagrangian calculation of water droplets impinging on the inlet, so anything I can do to that would make this easier (or possible) woud also beof interest.

Many thanks in advance

[grtabor]

Well, as to what to use, rhoTurbFoam would seem to be appropriate, unless you have transonic flow, in which case sonicTurbFoam. Look in applications/solvers/compressible. These are transient codes, so you would need to run them to steady-state or modify them to use SIMPLE rather than PISO.

Pressure outlets would certainly be possible; including internal calculation of the values if necessary. I've only ever used velocity inlets with Foam, but it wouldn't be too difficult to modify one to base the velocity on the pressure, I wouldn't have thought.

Gavin

[pauljwells]

I've read in my mesh, which looks ok, and I've input values for all the parameters that appear to be needed. Not seeing any other way to do it I hve input pressures as SI, so for example my outlet pressure boundary has p=101325...

As soon as I try to run the case I get this, and have no idea what it means

CaseServer running.....
debug::switchSet(const char*, dictionary*):
Cannot find DimensionedConstants in dictionary /home/paulwells/OpenFOAM/OpenFOAM-1.3/.OpenFOAM-1.3/controlDict

[gschaider]

Have you checked the file /home/paulwells/OpenFOAM/OpenFOAM-1.3/.OpenFOAM-1.3/controlDict ?

In my installation (which is the way it comes with the standard distribution) it's approximately 604 lines long and near the end there is a list with the DimensionedConstants. Maybe your controlDict got damaged? (Don't know how. I don't suppose you edited it yourself. From the "CaseServer running....." I deduce that you're using FoamX. Havn't done that in a long time but I faintly recall (but I might be wrong) that there was some menu point where you implicitly manipulated that file. But I don't know which and blaming FoamX is very uncool so I'll shut up). Just extract that file from the distribution-tars and replace it in your installation.

[pauljwells]

Yes, I'm using FoamX. and you are right about the controlDict file - I found another post that had the right data to add to the end of the file and that worked, thanks!

I got a little further and am now searching for info on:

Starting time loop

Mean and max Courant Numbers = 0.623619 104.133
Time = 0.001

diagonal: Solving for rho, Initial residual = 0, Final residual = 0, No Iterati ons 0


--> FOAM FATAL IO ERROR : Unknown discretisation scheme <>

Valid schemes are :

28
(
skewLinear
Phi
localBlended
QUICK
vanLeerV
upwind
limitedCubicV
limitedLinearV
vanLeer
limitedCubic
SFCD
midPoint
SuperBee
linearUpwind
SFCDV
MUSCL
Gamma
GammaV
downwind
localMin
localMax
limitedLinear
linear
UMIST
reverseLinear
weighted
blended
cubic
)