CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   WSGG-Model / CFX 11.0 / "Unavailable Variables" (

danbir November 12, 2009 11:55

WSGG-Model / CFX 11.0 / "Unavailable Variables"

i am simulating a rocket combustion chamber using ANSYS CFX 11.0 including radiative effects. Therefore a WSGGM approach is used as spectral model to calculate the influence of radiation. The WSGG model worked without any problems in the previous computations but unfortunately the absorption coefficients and weights i used were not appropriate for the conditions in the combustion chamber. Actually they were only valid for a pressure of 1atm but the chamber pressure is about 200atm.

I identified a much better WSGG approach called Spectral Line WSGGM (Denison) which is valid for my case and implemented it to CFX using CEL expressions. After that i validated the model.

At the moment i am trying to carry out a computation using my new epxressions but CFX-Pre 11.0 always prompts the following error message:

"Parameter 'Weight' in object '/FLOW/DOMAIN:Default Domain/FLUID MODELS/THERMAL RADIATION MODEL/SPECTRAL MODEL/GRAY GAS:Gray Gas 1' has been assigned an expression that references the following unavailable variables: molf"

This problem was already discussed in this forum a few times but no solution was posted by now. I found out that the weight of a WSGG model is only allowed to be a function of Temperature T in CFX (see the file RULES in cfx/etc folder).
I tried to modify the line "Dependency List = T" in the file RULES to "Dependency List = molf, T" or "Depedency List = molf, T, pabs" (I need another WSGGM approach which calculates the weights using the absolute pressure, too.) but this was not successful.

Has anybody got an idea how to solve this problem?

Thanks in adavance.

stumpy November 12, 2009 17:49

Although modifying the Dependency List has been mentioned before on this forum, those dependency lists are usually there for a good reason. While sometimes you can get away with changes, in general you can't!
So, having said this, you can try the following hack. Create an Additional Variable equal to molf, write your weights in term of the AV rather than molf, then change the Dependency List to ANY (because I don't know how to allow dependency on just AV's). I'm pretty sure this won't work - the radiation solver is a separate solver within the main solver, so it doesn't have access to the usual variables that the main solver does - but it's all I could think of. Assuming it fails, then I would go to ANSYS for support.

All times are GMT -4. The time now is 15:42.