[KyleA&M]

Hello all. So I'm trying to run natural convection on starccm+ with water as the medium. I'm running with a constant density for EoS, and am using the boussinesq model and gravity enabled. I have a heat flux input on some pipes, and a heat flux removal on a different part to drive the flow.

Now I'm trying so simulate as if the inlet and outlet pipes were connected to a reservoir, so I thought I would use a stagnation inlet and pressure outlet. It wasn't until I was getting reversed flow that i realized that the stagnation inlet was total pressure while the outlet was just pressure, I'm assuming gauge.

As far as I know P_total=P_gauge+density*(velocity^2)/2 . As I don't know the velocity I don't know what to use for total pressure, I can guess a velocity and run it but that wouldn't be right. How else is there to model buoyancy driven flows?

Also I get crappy convergence, seems starccm+ isn't that great for this...

[cwl]

Why do you think it is exactly Absolute (P + r*v*v/2) Pressure? - It is a computational result, but not a common boundary-value-set variable;
I'd suggest that it is P_total = P_reference + P_Inlet.
Two Pressure Inlets might work as well, with Flow-Split Outlet maybe.

Also I get crappy convergence, seems starccm+ isn't that great for this...

In such simple problem - user should be considered guilty first, not the program ;>

[KyleA&M]

Thanks for the reply. So when I set the boundary stagnation inlet it wants me to specify total pressure. If P_total=P_ref+P_inlet, than I would have to specify a P_inlet that I don't know? Thats kind of why I assumed it was how I specified before.