I use Floworks to simulate a radial fan, using a local rotational region, a flow rate as Inlet and a static pressure as outlet BC. I tried several mesh settings and different flow rates but the calculated pressure drop is always smaller than the experimental value. In some cases there also occurs a big pressure loss at the outer region of the disabled volume, which is used as rotational region.

Has someone tried to use EFD for such a kind of analysis or have any experience about it ? I guess the pressure correction term at the boundary between rotational and standrad mesh is wrong ?

The rotation region and the surrounding geometry must be absolutely axisymmtrical relative the rotation axis. Make sure that this requirement is satisfied - any non-axisymmetrical geometry or flow disturbance can have a huge impact on the solution accuracy. What version of FloWorks do you use?

I used Floworks 2009 Sp0.0. The rotating region and everything inside was axisymmetrical, but the outer geometry wasn't.

The whole model don't need to be axisymmetrical, but it is highly recommended for the geometry adjacent to the rotating region. In no case the rotating region boundaries should intersect a non-axisymmetrical geometry. It is also recommended to have the axisymmetrical flow field at the rotating region boundaries. You may also try to adjust the shape of rotating region to make the flow velocity vectors as much close to normal to the rotating region boundaries as possible.

If the radial fan is rotating at particular RPM then why do you want to specify inlet flow rate as BC?

I specify an inlet flow rate to see, which pressure drop belongs to this flow rate. If you don't specify any flow rate, you only get one special operating point but no pressure-flow rate-curve for this fan, as you need it.

Hi,

you may try to solve your problem using "turbulent only" option. In some case the default turbulent and laminar option dampen the pressure drop in fan. I don't know exactly why but I think it is a boundary layer resolution in strongly curvature flow issue.

Regards

Thanks,

It's a good idea. I will try this option.

Best regards