How to clip the critical radius in non-equilibrium condensation model with a CEL?

Saeef · October 25, 2018, 10:15

I am troubled in the simulation of supercritical CO2 non-equilibrium condensation.
I have tried my best to make it converge. However, it doesn't work.
I have reduced the time step to 10^-8s, adopted the double precision calculation, and used the single-phase converged result as the initial field.

An article mentioned a trick to clip the critical radius of droplet.
I'm wondering how to clip the critical radius with a CEL?

It's an output parameter, so I'm confused.

ghorrocks · October 25, 2018, 16:43

First of all look at the FAQ on floating point error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F

Assuming the simulation is set up correctly the best way to increase your numerical stability is to improve your mesh quality. This can make a huge difference in difficult material models like yours.

I would put off fiddling with factors like the critical radius for now - your problem is likely to be more fundamental than that.

alirezame · October 26, 2018, 01:11

Hello,

I am the corresponding author of that paper you have mentioned. Yes, we have to write couple of CEL functios to force the solver to change the critical radius limits. But in your case, the divergence is not becase of the critical radius. As far as I remember, we could get converged results, but the droplet size was non-realistic. can you explain more abour your case, boundary consitions and other stuff. In parallel I try to find the CFDX file for that simulation to give you the CEL functions which we wrote. Good luck.

Saeef · October 26, 2018, 07:49

Quote:

Originally Posted by ghorrocks

First of all look at the FAQ on floating point error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F

Assuming the simulation is set up correctly the best way to increase your numerical stability is to improve your mesh quality. This can make a huge difference in difficult material models like yours.

I would put off fiddling with factors like the critical radius for now - your problem is likely to be more fundamental than that.

Hi,ghorrocks
Since I have obtained converged result with single-phase simulation, I think the mesh quality should be sufficient.
Thank you for your advice.
Best regards,
Saeef

Saeef · October 26, 2018, 08:34

Quote:

Originally Posted by alirezame

Hello,

I am the corresponding author of that paper you have mentioned. Yes, we have to write couple of CEL functios to force the solver to change the critical radius limits. But in your case, the divergence is not becase of the critical radius. As far as I remember, we could get converged results, but the droplet size was non-realistic. can you explain more abour your case, boundary consitions and other stuff. In parallel I try to find the CFDX file for that simulation to give you the CEL functions which we wrote. Good luck.

Hi,alirezame
That's really a coincidence. I didn't expect you were in this forum.

I have just started the research of SCO2 non-equilibrium condensation.
The first study case is the Laval nozzle which you have calculated.
The detail of my settings in CFX-Pre was attached here and the .rgp files were generated by my own matlab codes.
It referred to the steam turbine case in CFX tutorial.
My email address: liuzhiyuan@iet.cn
Appreciate your help.

Best regards,
Saeef

ghorrocks · October 26, 2018, 19:42

Quote:

Since I have obtained converged result with single-phase simulation, I think the mesh quality should be sufficient.

Incorrect. Multiphase models are much more intolerant of poor meshes than single phase models. The fact that a single phase model converges OK does not imply the mesh is OK for a multiphase model.

When you are having troubles with convergence improving mesh quality ALWAYS helps, and sometimes completely fixes the problem. Mesh quality is often underestimated by inexperienced CFD people.

mdshujanali1 · July 9, 2021, 23:50

Quote:

Originally Posted by alirezame

Hello,

I am the corresponding author of that paper you have mentioned. Yes, we have to write couple of CEL functios to force the solver to change the critical radius limits. But in your case, the divergence is not becase of the critical radius. As far as I remember, we could get converged results, but the droplet size was non-realistic. can you explain more abour your case, boundary consitions and other stuff. In parallel I try to find the CFDX file for that simulation to give you the CEL functions which we wrote. Good luck.

Hi Alirezame, I am MD Shujan Ali, a PhD candidate at Texas A&M University. I am reading your paper on sCO2 condensation in de Laval nozzle. I am having difficulty using this model. Please would be be able to send your CFX file or CCL file to me. It would be a great favor if you can share your files. My email md_shujan@tamu.edu

mdshujanali1 · July 9, 2021, 23:53

Quote:

Originally Posted by Saeef

Hi,alirezame
That's really a coincidence. I didn't expect you were in this forum.

I have just started the research of SCO2 non-equilibrium condensation.
The first study case is the Laval nozzle which you have calculated.
The detail of my settings in CFX-Pre was attached here and the .rgp files were generated by my own matlab codes.
It referred to the steam turbine case in CFX tutorial.
My email address: liuzhiyuan@iet.cn
Appreciate your help.

Best regards,
Saeef

Hi Saeef,

I am also working with sCO2 condensation inside de Laval nozzle. Please can you send me your CFX File or CCL File? It would be a great favor if you can share your files. My email md_shujan@tamu.edu

Best regards,
MD Shujan Ali

xiangpica1 · September 27, 2021, 22:39

Quote:

Originally Posted by Saeef

Hi,alirezame
That's really a coincidence. I didn't expect you were in this forum.

I have just started the research of SCO2 non-equilibrium condensation.
The first study case is the Laval nozzle which you have calculated.
The detail of my settings in CFX-Pre was attached here and the .rgp files were generated by my own matlab codes.
It referred to the steam turbine case in CFX tutorial.
My email address: liuzhiyuan@iet.cn
Appreciate your help.

Best regards,
Saeef

Hi, Saeef
Have you solved the critical radius clip problem? May you share us how to write the CEL to clip the critical radius?
My email address is xupengcheng@buaa.edu.cn
Many thanks!
Best regards

October 25, 2018, 16:43		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	First of all look at the FAQ on floating point error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F Assuming the simulation is set up correctly the best way to increase your numerical stability is to improve your mesh quality. This can make a huge difference in difficult material models like yours. I would put off fiddling with factors like the critical radius for now - your problem is likely to be more fundamental than that. __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Droplet condensation model - What is it for?	Piti	CFX	0	February 25, 2009 12:29
Thermal phase change condensation model	sudhakar	CFX	0	June 6, 2008 05:21
New turbulence model - CEL or Junction Box?	flga	CFX	3	October 31, 2006 03:04
Water vapour condensation in CFX-5.7.1	hdj	CFX	1	November 27, 2005 07:15
Condensation model	Matthew Okasinski	FLUENT	0	May 12, 2000 15:07

October 26, 2018, 01:11		#3
alirezame Member Join Date: Nov 2013 Posts: 57 Rep Power: 12	Hello, I am the corresponding author of that paper you have mentioned. Yes, we have to write couple of CEL functios to force the solver to change the critical radius limits. But in your case, the divergence is not becase of the critical radius. As far as I remember, we could get converged results, but the droplet size was non-realistic. can you explain more abour your case, boundary consitions and other stuff. In parallel I try to find the CFDX file for that simulation to give you the CEL functions which we wrote. Good luck.