CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Which Viscous model to use to simulate airflow over obstruction? (https://www.cfd-online.com/Forums/fluent/209683-viscous-model-use-simulate-airflow-over-obstruction.html)

amitshah3333 October 22, 2018 15:21

Which Viscous model to use to simulate airflow over obstruction?
 
4 Attachment(s)
Hi,
I am a beginner to Fluent. I am trying to model airflow over obstruction to get air streamline. I tried K-Epsilon (Real), K-Omega (SST) and LES model in 2D.
Distance between inlet and obstruction is 1.5m and between output and inlet is 2m. Air velocity at inlet is defined @ 10m/s. Setup model is shown in file. Output (image) of velocity streamline are attached. From Velocity profile, it is seen that flow is not reaching obstruction and starts spreading before that.
Please guide me where I am making mistake.


Thankyou,

LuckyTran October 22, 2018 16:19

There isn't a mistake per se so much as it is simply a visualization problem. You are seeding the start of your streamlines using your inlet. Try inserting a plane that slices through the domain near the obstacle. Have you tried just plotting raw velocity vectors?

amitshah3333 October 23, 2018 02:51

5 Attachment(s)
Thankyou for your time and guidance. I increased time of study and captured images.
KE and KW seems to give almost same results but LES gives total different result in terms of air flow. How to make out which model to use for simulation?
Also velocity vector of LES is attached in next post.

Thanks a lot,

amitshah3333 October 23, 2018 02:52

Velocity vector of LES
 
1 Attachment(s)
Please find image of velocity vector of LES.

LuckyTran October 23, 2018 09:54

So you can see with vectors there is a lot of flow over the osbtacle which is not captured by streamlines simply because of the way streamlines are generated.


The velocity in LES doesn't have the same meaning as velocity in RANS. In RANS you solve for the time-averaged velocity (call it an apple). In LES, you solve for the time-accurate velocity (call this an orange). To even compare LES to RANS, you need to take the time-average of the LES results (to compare apples to apples). You can do this by checking the box data-sampling for time statistics and then letting your LES run for a very very very long long long time.


Then you'll still find some differences probably. Which to choose? Whichever you like.

amitshah3333 October 23, 2018 10:20

Thanks a lot for your timely response.
How to do statistical study of data sampling?Where is this option in Fluent?


All times are GMT -4. The time now is 19:27.