Stage Frame Change + Porous Domain

CFX_Beginner · July 30, 2015, 10:59

Hello,

i have a problem with the CFX-Solver being unable to fully calculate my project. I think the error is caused by a rotating Porous Domain that i have in my setup.

When i replace the 2 Porous Domains with Fluid Domains and leave the rest of the settings untouched the solver will run

When i keep the Porous Domains and change the Frame Change Option in the Domain Interfaces between the stationary and the rotating Domains of the flow channel from Stage to Frozen Rotor and leave the rest untouched the solver will run too.

With the settings i want to have (Porous Domains AND the Stage option as Frame Change in the Domain Interfaces between the stationary and rotating Domains) the solver will return an error after ~ 13 iterations.

So i think there is a problem with the Stage option in the Frame Change option of a Domain Interface and the Porous Domains itself. Cause when i leave one out the solver will finish the calculation.

Does anyone of you know more about this or is this not enough information to say that the problem is the interaction between Porous Domains and the Stage Frame Change option?

ghorrocks · July 30, 2015, 19:31

Have you tried all the normal techniques for tricky convergence - improve mesh quality, smaller time steps, double precision numerics?

If that does not work have a look at the expert parameters available for porous regions, and the options available to change porous region numerics. Some of them affect convergence so they may be relevant.

CFX_Beginner · July 31, 2015, 07:25

Hi ghorrocks,

thank you for your answer.
Do you know which of the expert parameters and options of the Porous Domain have an impacton the convergence?

Can explain to me how i can realize those two techniques you mentioned (smaller time steps, double precision numerics).

ghorrocks · July 31, 2015, 09:31

Have a look at the available options in CFX-Pre. Look at the expert parameters, the porous material section and the solver settings tab.

Smaller time steps - change the physical time scale.
Double precision - in solver manager when you start the simulation activate the double precision option.

CFX_Beginner · August 3, 2015, 12:37

Hi ghorrocks,

sadly my attempts with the smaller time steps and double precision failed.

I attached a picture of my parameters of the porous domain, i think you mean this by expert parameters, right?

ghorrocks · August 3, 2015, 19:39

No, the expert parameters are in a separate section. Also read the documentation about porous regions as it says a lot about how to use the model.

CFX_Beginner · August 4, 2015, 05:00

I think i read all the articles that the Ansys Help provide about Porous Domains, but in my oppinion they are not explaining it very well.

Can you link me the article you mean, maybe i left that one out?

ghorrocks · August 4, 2015, 19:41

I cannot guess what section of the documentation you have not read.....

Have a look in the theory manual and the Modeling Guide. They both have useful information on this topic.

CFX_Beginner · August 5, 2015, 16:23

My problem is that there is only the Chapter 1.8.1, which is only describing the loss model of porous domain.

But that is not enough information to fully understand what causes my problem.

I attached a picture of my search results, maybe i am missing some chapters?!?

ghorrocks · August 5, 2015, 20:20

That sounds unlikely.

Not all the options are described in the documentation anyway. I would try all available options yourself with a small test case and see what they do in your case. Then you can work out yourself what is important and what is not. This is what I do when looking at problems with a new physical model.

Sanyo · August 10, 2015, 05:28

Can you please describe your problem setup? What is the porosity of your porous domain? What is the fluid involved? Sometimes this problem can come due to wrong case setup.

July 30, 2015, 10:59	Stage Frame Change + Porous Domain	#1
CFX_Beginner New Member Phil Join Date: Jul 2015 Posts: 19 Rep Power: 11	Hello, i have a problem with the CFX-Solver being unable to fully calculate my project. I think the error is caused by a rotating Porous Domain that i have in my setup. When i replace the 2 Porous Domains with Fluid Domains and leave the rest of the settings untouched the solver will run When i keep the Porous Domains and change the Frame Change Option in the Domain Interfaces between the stationary and the rotating Domains of the flow channel from Stage to Frozen Rotor and leave the rest untouched the solver will run too. With the settings i want to have (Porous Domains AND the Stage option as Frame Change in the Domain Interfaces between the stationary and rotating Domains) the solver will return an error after ~ 13 iterations. So i think there is a problem with the Stage option in the Frame Change option of a Domain Interface and the Porous Domains itself. Cause when i leave one out the solver will finish the calculation. Does anyone of you know more about this or is this not enough information to say that the problem is the interaction between Porous Domains and the Stage Frame Change option?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
fluid porous domain interface	lauda	Main CFD Forum	0	July 31, 2014 10:29
Error finding variable "THERMX"	sunilpatil	CFX	8	April 26, 2013 08:00
No turbulence generation in a porous domain	siw	CFX	3	September 25, 2012 09:10
CFX domain comparison	Kiat110616	CFX	4	April 3, 2011 23:43
Heat transfer in porous domain CFX V12	Hashmi	CFX	0	August 15, 2009 13:05

July 30, 2015, 19:31		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Have you tried all the normal techniques for tricky convergence - improve mesh quality, smaller time steps, double precision numerics? If that does not work have a look at the expert parameters available for porous regions, and the options available to change porous region numerics. Some of them affect convergence so they may be relevant.

July 31, 2015, 07:25		#3
CFX_Beginner New Member Phil Join Date: Jul 2015 Posts: 19 Rep Power: 11	Hi ghorrocks, thank you for your answer. Do you know which of the expert parameters and options of the Porous Domain have an impacton the convergence? Can explain to me how i can realize those two techniques you mentioned (smaller time steps, double precision numerics).

July 31, 2015, 09:31		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Have a look at the available options in CFX-Pre. Look at the expert parameters, the porous material section and the solver settings tab. Smaller time steps - change the physical time scale. Double precision - in solver manager when you start the simulation activate the double precision option.

August 3, 2015, 19:39		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	No, the expert parameters are in a separate section. Also read the documentation about porous regions as it says a lot about how to use the model.

August 4, 2015, 05:00		#7
CFX_Beginner New Member Phil Join Date: Jul 2015 Posts: 19 Rep Power: 11	I think i read all the articles that the Ansys Help provide about Porous Domains, but in my oppinion they are not explaining it very well. Can you link me the article you mean, maybe i left that one out?

August 4, 2015, 19:41		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	I cannot guess what section of the documentation you have not read..... Have a look in the theory manual and the Modeling Guide. They both have useful information on this topic.

August 5, 2015, 20:20		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	That sounds unlikely. Not all the options are described in the documentation anyway. I would try all available options yourself with a small test case and see what they do in your case. Then you can work out yourself what is important and what is not. This is what I do when looking at problems with a new physical model.

August 10, 2015, 05:28		#11
Sanyo Member Sanyo Join Date: Apr 2009 Location: India Posts: 62 Rep Power: 17	Can you please describe your problem setup? What is the porosity of your porous domain? What is the fluid involved? Sometimes this problem can come due to wrong case setup.