[cbooks77]

Hi all,

I'm modeling generator exhaust dispersion using singlephase with species transport. I have specified the atmospheric boundary layer logarithmic velocity and turbulence boundary conditions via UDF according to literature for neutral atmospheric stability.

Boundary conditions are velocity inlet, pressure outlet, symmetry on sides and top, and wall for ground. For incompressible flow these simulations work great and a homogeneous boundary layer is present in the domain.

My problem is that the generator exhaust is very high temperature (~900F) and I want to include buoyancy by using the incompressible ideal gas law to capture temperature effects leaving the exhaust pipe. I set operating density to zero to include hydrostatic pressure field. Obviously the pressure outlet cannot be assigned zero gauge or you will have inconsistent pressure gradients at the boundary and unrealistic reverse flow. My first thought was to specify the outlet profile via a UDF which did reduce some reverse flow. But there is still a small discontinuity in the pressure gradient at the boundary causing reverse flow. Extrapolation of the interior pressure field to the outlet would be pretty much identical to an outflow condition so thats off the table.

Outflow cannot be used if you are modeling unsteady flows with varying density, even if the flow is incompressible per Ansys documentation.

Is it possible to set up a mass flow inlet with a boundary layer profile via UDF? Any advice would be greatly appreciated !

[LuckyTran]

Actually the pressure outlet can be assigned zero because the pressure that you specify boundaries should not include the hydrostatic head. That is, you do not need a profile/UDF to specify the rho*g*z part of the pressure. You just need to worry about the flow part.


If you get reverse flow despite this, then you have a problem with reverse flow not because you didn't account for the hydrostatic pressure.

[cbooks77]

Hi LuckyTran,

Please see the attached images. The pressure contour picture shows the plot of static pressure when I use my UDF pressure profile. The second picture shows the static pressure when set to zero gauge. I have set operating density to zero to include the hydrostatic head.

Edit: The issue is the incorrect pressure profile ruins the homogeneity of velocity gradients in the flow direction. Also the flow is moving in the negative y direction.