Learning Forum

[abhinavh]

Dear All Members,

I have been trying a lot to obtain Time-averaged plots from LES simulations in stirred vessels with multiple impellers through ANSYS Fluent 19.2.

The problem is that the contours for the mean velocity flow field look strange for the inner rotating zones and fine for the static zones for the stirred vessel as shown in the attached snapshot. The instantaneous absolute and relative plots are fine without any issue. I have double-checked the boundary conditions and also tried the new version R20 but the problem still exists.

Did anyone come across this problem in Fluent? I have done everything as per the instructions in manual using unsteady statistics and data sampling method.

Looking forward to hearing from you at your earliest convenience. Please refer to the snapshot of the mean velocity plot I need your help to fix it.

Kind Regards,
Abhinav Hazarika

[Rob]

What does the instantaneous image look like?

[abhinavh]

 

The instantaneous relative velocity profile looks like the attached snapshot below

.

Looking forward to your feedback .

[Rob]

Looks like the RMS is relative rather than absolute velocity (or the reverse). Check what you're monitoring. 

[abhinavh]

The snapshot which I posted in my first post earlier is of the mean velocity and not RMS.  I am sorry but I could not understand your feedback clearly. Could you kindly elaborate?

It would be nice to have a solution for the strange mean velocity flow profile in the inner rotation zone. I hope there is no unfixed bug in Fluent?

 

[Rob]

What do you mean by the "mean velocity"? Time averaged? 

Fluent has the ability to report relative and absolute velocity. As you're rotating a cell zone if you report relative velocity you may get some odd results. 

[abhinavh]

Yes, Time - averaged. 

Please find attached the picture to make it more clear what I am reporting.

Looking forward to the solution.

 

[Rob]

How long has the model run for, how did you initialise & when did you start data sampling?

[abhinavh]

The model ran for 50 seconds in total. The total time sampled was 30 seconds so I started data sampling after 20 seconds of my simulation time. 

Initialization was standard initialization based on the absolute reference frame.

 

[Rob]

OK. You should have a sensible result. I assume you're solving absolute and haven't been messing with the settings?

[abhinavh]

Yes, I am solving absolute and did not mess the settings. What do you think would be the issue then? 

[Rob]

Not sure, I'll check this end: that may take a few days. 

[abhinavh]

Ok, thank you. Looking forward.

[Rob]

Looks like the velocity averaging isn't working as expected.  Have a look at custom field functions and how to add data into the time statistics. 

[abhinavh]

Ok, thanks for the update. I will have a look.

[abhinavh]

How much time will ANSYS need to fix the velocity averaging problem in the GUI? I have not yet tried custom field function for time statistics.

Best Regards,

Abhinav

[Rob]

It depends on the cause, and isn't something I can comment on.