I'm running an uncoupled CFD/thermal analysis to determine the effect of flowing water on the heat distribution.
I've created a very simple proof-of-concept model to verify the results, and I'm seeing some strangeness.
First, I created the thermal model and meshed it appropriately. Then I copied the scenario and created a CFD model by suppressing the 3 non-fluid parts, applied the appropriate CFD boundary conditions, ran it, and got steady state results.
Then, back in the thermal model, I have applied the Fluid Convection results using the other scenario and have run the model. It appears as if it's not applying the velocities right, as can be seen in the outputs below.
Here is the water slice mesh:
Here are the results of the water flowing in steady state:
There is a fixed velocity on the left, which is why it's all green. And here's a view of the output face:
So that all looks right to me.
Here's the thermal model:
You can see the water layer inside of it. There is a surface heat source between the bottom (white) and middle (pink) sections. The top and bottom surfaces of the column are held at 0 degrees.
With no water flowing, the results look like this:
That's as I would expect.
However, with flow, the results look like this:
Any idea what could be going on here?? I can upload the model if the answer isn't obvious from the images above.
I'm stuck. I have no idea what's going on. I can't tell if this is a bug in Multiphysics or if I'm doing something wrong.
Right now, my conclusion is that uncoupled Multiphysics results cannot be trusted. This simple model demonstrates a fundamental flaw.
Perhaps a coupled simulation would give accurate results? I'm not sure, because the simulation won't finish...
I've attached the model, if anyone wants to jumps in. (Need to remove .zip from the end of the .z01, .z02, and .z03 pieces before unzipping.)
Here's the last part of the zip file. Remember to use the .zip from the .z03 file before unzipping.