[GulhermeSantos]

Hello everyone,

I'm currently studying an intercooler and for the finned part I'm using porous media to replace the actual fin geomtry to save computational resources.
To validate the flow around fin and see if the porous media represents with precision the fin, i'm simulating a porous region inside a channel before i import this porous region into the real geometry of the intercooler. Just some pre-validation.

I'm simulating this with 1 mass flow inlet and 1 outlet ( because the pressure outlet was diverging) 1 wall with constant temperature and 3 symmetry planes ( I'm just using half a channel in height and 1/10 in width of thereal channel). The porous resistance coeficients and solid resistance are from a real fin. I'm simulating this with equilibrium and non-equilibrium. In non-equi. i set the boundary conditions of the solid phase as adiabatic in inlet and outlet and constant temperature in the wall, and 1 phase intersection with the htc that i discovered with an correlation.

Don't know why in either models ( equilibrium and non-equilibrium) the amount of heat transfered to the wall is smaller than the heat lost by the water inside the channel. The heat lost by the water I used a field function that calculates the energy that comes in and comes out of the channel (${AxialVelocity}*${Density}*${SpecificHeat}*${Tem perature}*$${Normal}[2]) and compared with the heat transfer report of the wall. It seems in the non equilibrium that increasing the htc between the solid phase and the fluid phase that the heat transfered to the wall decreases.

What am i doing wrong? Is there something that i need to do to connect the 'fin' to the wall? In the equilibrium model there shouldn't be this problem since the fluid and the solid phase are conectec by just 1 equation.

If you have any questions pls feel free to ask.

waiting for your reply.

guilherme

[fluid23]

Are you using a porous baffle or porous region?

[GulhermeSantos]

Hi thanks for replying,

I'm using porous region fot both models (equilibrium and non-equilibrium).
If you wanr i can upload the simulation files for each model.

[fluid23]

What version are you running? I probably cannot open it. Pictures are usually sufficient.

I am still a little fuzzy on what your issue is exactly, but I would caution you to double check the porosity value as this will affect heat transfer rates.

[GulhermeSantos]

I'm modeling a real intercooler so all the parameters including the porosity are real.
The main problem that I found was that when I do an energy balance to the half channel that I modeled to simplify a real channel of the real geometry, the energy balance doesn't match. I calculate the energy entering and exiting the system with the flow with mcpT and the heat transfer report to know the energy that is transfered to the wall. This balance does not match because the energy lost by the flow is greater than the energy transfered to the wall, and since all the other boundaries are symmetry I don't know where is the energy going.

[fluid23]

Have you checked mesh quality? Is your energy residual converging? It sounds like you have some numerical diffusion perhaps. I use porous regions a lot, but not for thermal analyses so I am stabbing in the dark here a bit.

I often run in to continuity issues with porous regions and its almost always a mesh issue.

Also, how much energy is being lost? 1%? 10%? 50% A little perspective might help.

[GulhermeSantos]

In the attachments go an image of the values. You can see that is almost 50% difference between the heat lost and the heat transfered in the wall