[anengineer]

Hello,

I am trying to simulate an Stirling engine in 2D axisymmetric.

In the simulation, the engine has 2 chambers, one hot (left side and where the fluid, nitrogen, is expanded), which has some of the walls at 550 [ºC], and one cold (right side and where the nitrogen is compressed), which has some of the walls at 70 [ºC].

I model the movement of the pistons with an UDF file.

The problem comes with the regenerator, that is a stainless steel mesh situated at the middle of the transfer space.

As far as I know, I need 4 parameters to model the regenerator: POROSITY, Inverse of permeability (1/alpha), inertial resistance (C2) and the relative viscosity.

For the first approximations, I am calculating with {porosity 1,1/alpha=0,C2=0,relative viscosity=1.992 (calculated with Fluent's Manual equation)}, which means that the whole region is fluid, and I obtain the same results as without the regenerator.

But when I try with the real {porosity (0.7), 1/alpha = 0, C2 = 0, relative viscosity=1.922}, the results I get are crazy: the cold side (that is the compression chamber) suddenly becomes the hot side, because the temperature increases a lot (1200 [ºC]) and the temperature of the hot side decreases (200 [ºC]). I suppose that the fluid is not accurately being transferred from one side to the other because of the regenerator model.

I am completely lost and I do not know how to continue with this... If anyone has already done something with porous zones please, I really need help.

The model:


https://ibb.co/MZ4T5WQ
https://ibb.co/MZ4T5WQ
The simulation with Porosity=1, 1/alpha=0, C2=0


https://ibb.co/YyWDr1n
https://ibb.co/YyWDr1n

The simulation with the real porosity, Porosity=0.7, 1/alpha=0, C2=0


https://ibb.co/L6my42V
https://ibb.co/L6my42V