Two-way coupling of Lagrangian particle

Questions and remarks about code_saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Post Reply
lehongduc
Posts: 23
Joined: Thu Aug 20, 2015 3:05 pm

Two-way coupling of Lagrangian particle

Post by lehongduc » Fri Dec 02, 2016 10:16 am

Hello all,

I would like to use two-way coupling of Lagrangian particle in Code_Saturne 4.0.3 but I have an unrealistic value of my results (about fluid velocity). I'm looking for some information about two way coupling of CS and I found a topic about it a year ago:

http://code-saturne.org/forum/viewtopic.php?f=2&t=1787.

And eventually, there has been no validation of two-way coupling yet.

My question is : are there any improvement or validation of two-way coupling since last year?

For my case calculation, it's quite confidential, I cannot upload it in the forum.

Thanks in advance.
Best regard,
LE Hong Duc

Yvan Fournier
Posts: 3084
Joined: Mon Feb 20, 2012 3:25 pm

Re: Two-way coupling of Lagrangian particle

Post by Yvan Fournier » Fri Dec 02, 2016 5:10 pm

Hello,

Actually we have one validation test case which does 2-way coupling, but only tests some combinations.

Are you using specific physics such as combustion, or just basic Lagrangian options ? Which turbulence model are you using ?

There have been some general improvements in version 4.3, but it has only undergone simple validation (the full validation will come this winter for an April release of 5.0).

Regards,

Yvan

lehongduc
Posts: 23
Joined: Thu Aug 20, 2015 3:05 pm

Re: Two-way coupling of Lagrangian particle

Post by lehongduc » Mon Dec 05, 2016 10:45 am

Hello Yvan,

Thanks for the reply. I'm using only basic Lagrangian options. Turbulence model is "k-epsilon". The options of 2-way coupling is: nstits = 1, ltsdyn = 1.

I tested the integration order of SDE (nordre = 1 or 2), and also the activation or not the dispersion model (idistu = 0 or 1), those options doesn't fix my problem.

Could I have access to the validation test case of 2-way coupling that you mentionned?

Thanks in advance,
Best regard,
LE Hong Duc

Yvan Fournier
Posts: 3084
Joined: Mon Feb 20, 2012 3:25 pm

Re: Two-way coupling of Lagrangian particle

Post by Yvan Fournier » Tue Dec 06, 2016 1:24 am

Hello,

For which version of the code would you need the validation test case ?

Regards,

Yvan

lehongduc
Posts: 23
Joined: Thu Aug 20, 2015 3:05 pm

Re: Two-way coupling of Lagrangian particle

Post by lehongduc » Tue Dec 06, 2016 9:04 am

Hello Yvan,

I'm using CSv4.0.3. It would be nice if you have the validation test case of this version. If not, the validation of another version closest to v4.0.3 could be fine, I will try the same validation test case on v4.0.3.

Thanks in advance,
Best regard,
LE Hong Duc

Yvan Fournier
Posts: 3084
Joined: Mon Feb 20, 2012 3:25 pm

Re: Two-way coupling of Lagrangian particle

Post by Yvan Fournier » Tue Dec 06, 2016 12:40 pm

Hello,

Here is a case:

Regards,

Yvan
Attachments
jet.tar.gz
(121.81 KiB) Downloaded 96 times

lehongduc
Posts: 23
Joined: Thu Aug 20, 2015 3:05 pm

Re: Two-way coupling of Lagrangian particle

Post by lehongduc » Tue Dec 06, 2016 1:12 pm

Re,

Thanks very much! I'll take a look at it.

Regard,
LE Hong Duc

lehongduc
Posts: 23
Joined: Thu Aug 20, 2015 3:05 pm

Re: Two-way coupling of Lagrangian particle

Post by lehongduc » Fri Dec 09, 2016 1:42 pm

Hello Yvan,

I found my "bug". I have to set first order of numerical scheme of particle dispersion model for 2-way coupling. (nordre = 1 in uslag1.f90). I didn't look closer to see what is the probleme with nordre = 2 (maybe some probleme in predvv.f90 for source terme (ligne 16510))... but my case works now.

Thanks again,
Best regards,
LE Hong Duc

Yvan Fournier
Posts: 3084
Joined: Mon Feb 20, 2012 3:25 pm

Re: Two-way coupling of Lagrangian particle

Post by Yvan Fournier » Tue Dec 13, 2016 5:45 pm

Hello,

Yes, when updating the algorithm for V5.0 (in progress), our Lagrangian expert detected an issue in the 2nd order terms. As the fix was quite intrusive, it was done only in V5.0 (current development trunk). We should force 1st order for future V4.0 bugfix releases to avoid this issue.

Best regards,

Yvan Fournier

Post Reply