CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Trouble with big domain of study, (http://www.cfd-online.com/Forums/fluent/117811-trouble-big-domain-study.html)

johnnyp May 15, 2013 11:40

Trouble with big domain of study,
 
Good afternoon,

I have a big trouble with a big model. So, I decide to decompose the model in three sub-models.

My question is how I can run the study only one time, but using different models in the next manner:

-on the first model I will apply velocity_inlet and pressure_outlet
-on the second model I will apply as input instead of velocity, the pressure_outlet from previous model and as output pressure_outlet2
-on the third model I will apply as input instead of velocity, the pressure_outlet2 from previous model and as output pressure_outlet3
:confused:
If somebody have other idea or suggestion how to do that or equivalent solution, please help me.

Thanks in advance to all community for support.

Best regards.

P.S. I used ANSYS 14.5 with corresponding Fluent.

Liufeng_ustb May 15, 2013 20:35

Hope this can help you
 
Quote:

Originally Posted by johnnyp (Post 427790)
Good afternoon,

I have a big trouble with a big model. So, I decide to decompose the model in three sub-models.

My question is how I can run the study only one time, but using different models in the next manner:

-on the first model I will apply velocity_inlet and pressure_outlet
-on the second model I will apply as input instead of velocity, the pressure_outlet from previous model and as output pressure_outlet2
-on the third model I will apply as input instead of velocity, the pressure_outlet2 from previous model and as output pressure_outlet3
:confused:
If somebody have other idea or suggestion how to do that or equivalent solution, please help me.

Thanks in advance to all community for support.

Best regards.

P.S. I used ANSYS 14.5 with corresponding Fluent.

Hi, Johnny. I think your meaning is how you can revise slightly with your case and accomplish your simulation.

I think you can initialize. That is to say, when you finish the calculation of your first model, you do not need to reread your mesh or case, you can directly modify your model in Boundary Conditions, and initialize.

Recently, I am doing my undergraduate thesis. Because it always is of divergence, so I need to continually revise some options if it happens this situation. For every try, it is impossible for me to start a new setting for the whole process. So I use initilize when I fail every time.

I hope this can help you.

All the best

blackmask May 15, 2013 21:28

You did not get it. What Johnny wants is to decompose a large problems into several uncoupled or loosely coupled problems.

This however, depends on the dominant feature of the flow. For parabolic dominant flow like boundary layer it is feasible to cut the computational domain into several pieces then solve them sequentially. For other kind of problems I would suggest you looking for more powerful computational resources. You would work around your problem with the technique describe in this thread (http://www.cfd-online.com/Forums/flu...-boundary.html) but I prefer the hard(ware) way.

johnnyp May 15, 2013 23:19

Good morning,

Thanks for yours advice and sugestion. This is what Im looking for (thanks for link blackmask). I will try to implement the routine, and we keep in touch.

Thanks in advance for support to all comunity. If somebody have other ideea or sugestion beside of recommandation made by blackmask, I waiting.

Best regards.


All times are GMT -4. The time now is 03:03.