[aerosjc]

Hi,

For a channel flow, in numerical simulation, I use periodic boundary condition. Thus, the driving force cannot be pressure gradient in the flow field. I add source term to the NS eq. It's reasonable to add source term to both momentum eq and energy eq. However, I find the source term in energy eq causes pressure history going up. I mean the pressure history after the flow reaches statistically steady state.

Source term I use in momentum eq and energy eq is



The source term in energy eq seems not exactly correct, because it uses the bulk velocity. But that doesn't affect the results. Even corrected s_e causes pressure going up.

With source term in energy eq, pressure history is as follows,



Without source term in energy eq, pressure history is as follows,



Pressure history fluctuating is correct. However, I can't understand why I apply force on the flow while not work on the flow. And I know, it's related to the adiabatic boundary condition I use at the wall.

Anyone can explain to me why only force is applied on the channel flow?

[FMDenaro]

I never worked on the system coupled with the energy equation using mass forcing, many year ago I worked with my collegues on the pressure driven condition and the code worked.

I used in my code the constant pressure gradient as forcing in the momentum equation. This term would lead to a work in the form v.Grad p to be taken into account in the kinetic energy equation. Since you are using the compressible form, this term remains in this form and cannot be transformed only in Div (vp). Concerning the internal energy, to tell the true I don't know if such work (that is reversible and is not associated to entropy production) produces anything else in the equation that is shown here
https://www.researchgate.net/publica..._with_OpenFOAM