-
-
October 5, 2024 at 9:58 ammarco.adornoSubscriber
Hi,
I'm using Fluent to solve a transpiration cooling problem. Basically, I have two domains (see figure for clarity):
1) Hot gas domain: channel flow with hot air
2) Porous domain: rectangular domain in which a coolant is made flow through a porous media and then injected in the main flow
I need to couple the interface porous_domain - hot_gas in order to pass the temperature (of the solid part of the porous material) and fluid velocity from domain 2) to domain 1) and heat flux + pressure from domain 1) to domain 2). (Also here, see figure 2 for clarity)
I'm trying to do this by using the "System Coupling" (SC) tool offered by Ansys, but I'm having some troubles:
1) SC does not support two Fluent participants in the workbench, so I'm using it from the GUI
2) SC does not support (probably) passing velocity and pressure as coupling variables, so now I'm running a simplified version of the project in which I exchange only the thermal quantities
3) Even after these considerations, after I create the cases, make the .scp files, save them in the respective folders (afaik I need to work on a folder in which I create subdirectories where I save the .cas.h5 and the .scp files of every participant), once I run the solve command the tool is stuck in "Awaiting connections from coupling participants..." status. I've added the participants as the manuals explain and created the coupling interfaces with the corresponding variables.
By looking online, I've seen that a problem on 2023 ansys versions (I'm using ANSYS 2023 R1) is that SC doesn't start if IPV6 is enabled in your network, so I did it, but it didn't solve the problem. It may be a license problem, but I didn't arrive at a conclusion by looking online from this point of view.
Since this is not working, my backup plan is to couple the simulations via a Python script using pyFluent, but SC would have been easier and more straightforward to implement. But I'm also open to any suggestion concerning the coupling of the variables
I hope that the explanation is clear, I'll appreciate a lot every piece of advice you could give me.
Thank you!Â
Marco
-
October 7, 2024 at 12:02 pmRobForum Moderator
Why aren't you just solving in one Fluent case?Â
-
October 7, 2024 at 2:27 pmmarco.adornoSubscriber
Hi Rob, thank you for your answer.Â
I'm not doing it because by using a single domain to simulate both hot gas and porous material we have two option:
1) Use the LTE (Local Thermal Equilibrium), so assuming that fluid and solid (inside porous domain) are always locally at the same T. This is not valid generally, and additionally, this doesn't allow me to model the interface as a wall, meaning that the interface gradients will not be the correct ones.Â
2) Use the LTNE (Local Thermal Non Equilibrium), so solving two coincident domains in the porous material: one fluid domain, in which we solve continuity momentum and energy eq, and one solid domain in which we solve only the energy equation. This is closer to reality, but Fluent doesn't allow me to thermally couple the solid interface with the hot gas domain.Â
This is why I need to couple them externally.
-
October 7, 2024 at 3:15 pmRobForum Moderator
Complicated. In theory you may be able to use sink/source terms and User Defined Memory locations but it's very much not simple. I suspect it's also going to be beyond what we can cover on the Forum as it's not fully documented.Â
-
October 7, 2024 at 3:30 pmmarco.adornoSubscriber
Can you describe in more detail this approach?Â
Btw, I managed to couple the domain in the way I want by using a python script with pyfluent and coupling the simulations every few iterations.
-
October 7, 2024 at 4:07 pmRobForum Moderator
Unfortunately not: we're very limited on the Forum and can only use "public" information. The PyFluent approach may be a better option, especially if you've already got it working.Â
-
- You must be logged in to reply to this topic.
-
801
-
427
-
265
-
220
-
201
© 2024 Copyright ANSYS, Inc. All rights reserved.