-
-
June 28, 2025 at 10:56 am
scabo
SubscriberHi I am simulating pipe flow using RSM BSL model and getting good validation for streamwise velocity with experiments. But the Reynolds stresses at the wall are going to zero. THere is a thingg that because of a wall, the Reynolds normal stresses are damped and reddistributed along other 2 directions. THis wall reflection is missing in RSM BSL. How can i incorporate this in Fluent in the BSL model? any idea is helpful, thanks
-
June 30, 2025 at 7:40 am
Rob
Forum ModeratorYou can rewrite the wall functions using UDFs, that's non trivial and help is limited to pointing you at the UDF manual. Why are you wanting to rewrite the wall functions?
-
June 30, 2025 at 2:18 pm
scabo
SubscriberHi-I see that the Reynolds stress terms graphs along a line do not compare well to the experiments when using the RSM BSL. But when i am using the RSM linear pressure strain-it matches very well for the RE stresses but the primary velocity does not compare well but i am getting the Re stresses more accurate. Now therefore i was wanting to create a turb model including the positives from both BSL and Linear Press Strain (LPR). Does that make sense? If you can give me some idea it wll he helpful.
Â
So in short: Primary velocity agree more for BSL and Re stresses agree more for LPR.
-
June 30, 2025 at 2:51 pm
Rob
Forum ModeratorI'd be looking into the velocity mismatch, along with checking the experimental data too. If you want to do some tinkering you're either rewriting the wall function (if you can for RSM) or possibly messing with the stress source terms.Â
-
June 30, 2025 at 2:54 pm
scabo
SubscriberI did not fully get what you meant. The velocity magnitudes and Re stresses are coming different for RSM BSL and RSM LPR. And then what do you finally suggest?
-
June 30, 2025 at 3:11 pm
Rob
Forum ModeratorI don't know what you're comparing what with, nor how well refined the mesh, timescale etc are so if you want to mess with the models I have suggested two options.Â
-
June 30, 2025 at 3:19 pm
scabo
SubscriberOh-no the system is fully mesh independent and it is a steady state i have modelled. But including the wall reflection is not trivial? Can you just let me know how should it be done?
-
June 30, 2025 at 3:26 pm
Rob
Forum ModeratorYou want to alter a built in model. I've given you two possible options, there may be more. Both will require UDF coding. That's as far as I can cover on the Forum.Â
-
June 30, 2025 at 3:38 pm
scabo
SubscriberOkay thanks
-
- You must be logged in to reply to this topic.
-
3407
-
1057
-
1051
-
896
-
887
© 2025 Copyright ANSYS, Inc. All rights reserved.