[not_fluent_in_fluent]

Hello,

I've been having an issue that hopefully you all can help me solve. I'm fairly new to using fluent and so I wanted to start with a simple, heated vertical channel with flow driven by natural convection. I am not using the Boussinesq approximation because I am using the "incompressible-ideal-gas" option for density (which was described to me as the one to use if I don't want the pressure affecting the density but do care about the density being a function of temperature).

For my boundary conditions, I have a pressure inlet at the bottom and a pressure outlet at the top (both with 0 gage pressure). Along the vertical sides, I have a heated wall with no slip.

My issue is that I am getting positive flow up along the walls in the boundary layers, but then towards the center of the channel, the flow reverses and I am getting a negative velocity in the y-direction.

I feel like this (the negative y-velocity) shouldn't be happening assuming that the air is quiescent outside of the boundary layer. Do you guys have any suggestions? I've already looked at a bunch of the forum posts here, in addition to different videos online. I haven’t been able to correct this, and so I’m at a loss. If you guys could point me in the right direction.

Sidenote: It feel like Fluent is trying to drive the average velocity through the channel to zero. The picture of the Matlab code is me calculating the average velocity over the top, middle, and bottom of the channel.

I'll include some pictures that hopefully help show what I described earlier.

https://docs.google.com/document/d/1...it?usp=sharing

I appreciate all your help!

[not_fluent_in_fluent]

I figured out my issue...

In order to use the "incompressible ideal gas," I needed to specify the operating density as the density of the fluid at the inlet. This results in positive velocities all the way around. This solution worked for a heated channel as well as a simpler heated vertical plate.