[turbo-cfder]

In cfx, there is an inherent homogeneous classical condensation equation. I can use it normally following the instructions in the tutorial. But I am preparing to proceed with the next step by writing a new condensation equation. So I first tried to write an equation that is the same as the built-in equation of CFX for simulation. This way, I can know if the equation I wrote myself is correct. Unfortunately, the equation I wrote did not calculate the correct result.The correct result is approximately 1e26, while my result is only around 1e6. I carefully checked my equations but didn't find any problems. May I ask if anyone knows where I went wrong? If the information I have submitted is not enough, please let me know and I will add some more. Thank you very much.

[ghorrocks]

I suspect the built-in equations are linearised and have other numerical treatments so they converge better. You do not get that when you write your own equation in CEL. So it is possible the difference is just numerical.

Have you tried things which will give you better numerical accuracy, like:
* tighter convergence
* finer mesh
* double precision numerics
* More accurate discretisation schemes (both space and time - if relevant)

[turbo-cfder]

Originally Posted by ghorrocks

I suspect the built-in equations are linearised and have other numerical treatments so they converge better. You do not get that when you write your own equation in CEL. So it is possible the difference is just numerical.

Have you tried things which will give you better numerical accuracy, like:
* tighter convergence
* finer mesh
* double precision numerics
* More accurate discretisation schemes (both space and time - if relevant)

Thank you very much for your reply. In fact，I can converge during the simulation using both equations (RMS 1E-5). In addition, the results calculated by the CFX built-in equation are the same as those in the literature, so I believe the mesh meets the requirements. I don't know if there was an error in the CEL nucleation rate equation I wrote or if other operations were missing, which is currently my biggest confusion.

[ghorrocks]

If you are matching the literature results with the CFX model that sounds like your model is accurate which is good.

The equation you show on your first post - where did you get that from? Is it from the CFX documentation? If so, which equation number?

Also, which version of CFX are you using?

[Opaque]

Originally Posted by turbo-cfder

Thank you very much for your reply. In fact，I can converge during the simulation using both equations (RMS 1E-5). In addition, the results calculated by the CFX built-in equation are the same as those in the literature, so I believe the mesh meets the requirements. I don't know if there was an error in the CEL nucleation rate equation I wrote or if other operations were missing, which is currently my biggest confusion.

If you already have meaningful results, you should be able to evaluate your expression in CFD-Post using the existing fields until you find what is missing from the equation.

Typically, unit consistency or scaling of a variable (normalized?)

[turbo-cfder]

Originally Posted by Opaque

If you already have meaningful results, you should be able to evaluate your expression in CFD-Post using the existing fields until you find what is missing from the equation.

Typically, unit consistency or scaling of a variable (normalized?)

This sounds like a very effective suggestion, I will give it a try. Thank you for your reply

[turbo-cfder]

Originally Posted by ghorrocks

If you are matching the literature results with the CFX model that sounds like your model is accurate which is good.

The equation you show on your first post - where did you get that from? Is it from the CFX documentation? If so, which equation number?

Also, which version of CFX are you using?

actually，only the result with CFX built in model is basically matching with the literature。.When I use the nucleation model I defined, the results are quite biased, even though I follow the same equation.The equation is from the literature which is "Modelling of condensing steam flows in Laval nozzles with ANSYS CFX".CFX documentation mentioned it in “Ansys CFX-Solver Theory Guide” which equation number is 5.280.these two equations look a bit different,but i think they are the same.The CFX vision is 2022R1.I expect you can give me some advise.
best wishes

[ghorrocks]

The equation you quoted in post #1 looks quite different to eqn 5-280 in the theory manual. Eqn 5-280 is J = A.exp(-(delta)G*/k.Tg).

I do not have time to check your equation in detail, but are you sure the leading terms in your equation are equivalent to A in 5-280? And (delta)G* is in eqn 5-272 and does not look like your exponential function. Are you sure the exponential term is correct?