I have set up a model of an inline fan using the boundary conditions of an opening and a mass flow outlet, as advised by engineers at CFX. For some reason I seem to be getting negative pressure readings at a point downstream of the fan. I have used 1atm reference pressure and 0Pa relative pressure on the opening boundary and have my fan spinning in the correct direction. I am using GGIs between the fan inlet and opening and fan outlet and the main outlet test pipe and the frozen rotor approach.

I have tested different mass flow values and have tried different meshes and turbulence models (SST and k-e), but have the same problem, so it must be something basic I am missing in the setup.

Would anyone have any suggestions as to what I may be missing? I have put the question to staff at CFX over the weekend but as yet haven't had a reply, so would really appreciate any comments people may have.

Thanks in advance for your help.

Jenny

Pressure is only a nominal value (relative value). Talking about pressure increase or decrease makes more sense than the single point value.

If you set every boundary conditions as test setup (For example, 0 Total Pressure at the entrance of inelt side duct and mass flow at exit of outlet duct), you should get correct pressure. In this case, the static pressure at duct inlet and wheel inlet is negative. The wheel makes the work and raises the pressure.

If you set your Reference Pressure (Pref) to zero you may run into round-off issues (because you end up taking the difference between two large numbers). It's better to always set the Reference Pressure to be in the range of your calculation. A simple way to do this is to set Pref to the value at one of your boundaries and set that boundary condition to a relative pressure of zero, as Jenny has done.

The "Pressure" variable is then a Guage Pressure, relative to your domain reference pressure. A value of Pressure less than zero simply means it is lower than Pref. Note that "Total Pressure" is also reported relative to Pref. To see the Absolute Pressure, just plot or calculate using the "Absolute Pressure" variable or add your reference pressure in.

That said, it is still possible to get a negative Absolute Pressure if your fluid properties are not a function of Pressure. If the density is constant, the pressure level really has no meaning and it is only the relative pressures that matter. Physically this makes no sense (although it is possible to have a liquid in tension), but fluids which don't change with pressure don't exist!! So just take care when interpreting your results.

-CycLone

Thank you for your replies. I've double checked that I wasn't getting the pressures mixed up and I'm looking at the absolute pressure still being a negative value of 10Pa, where in reality my experimental value was 30Pa, so somehow 40Pa has got lost somewhere. I am currently trying to refine the mesh, but having some difficulties with resolving some complicated parts of the geometry (see new question submitted on Sunday).

Thanks again for your help.

Best wishes, Jenny