[hwsv07]

Hi

I have a problem which I can't figure why.

I have created a case of a thin actuator disc in a gas fluid domain (2 Regions).

I have created interface boundaries between the inlet/outlet/sides of the actuator disc and the fluid.

If I have understood the manual, interface boundary take precedence over the the boundary type, so the type of boundary (e.g. pressure outlet/wall etc) would not matter. Therefore, I just left it as default (wall in this case).

I have prescribed a specified Momentum Source with a constant value in the flow direction and plotted the streamwise velocity.

Streamwise Velocity.jpg

The disc is centered along the origin axis. Radius is 0.3m.
Flow is along y-axis.
Momentum source is prescribed at the Actuator Disc region, which I believe would apply a uniform momentum drop across the disc.

I can't understand why the flow is so non-uniform along the radius of the disc. I think it should be more or less uniform (at any distance downstream of the disc) since the momentum source is uniform.

Can someone please help? Thank you.

[fluid23]

Star-CCM+ has lots of options for this and from your choice of words it isn't clear what you are doing.

Virtual Disk: Creates a momentum source that is intended to model rotor geometry. This means it is typically spinning. There are 3 options, body force propeller method and blade element momentum method both account for radial variations so you should not see a uniform momentum change. The third option is the 1-D momentum method and this I believe will give uniform loading. I have never used it.


Fan Interface: Creates an interface between two regions that applies a momentum change between faces 1 and 2. This is not a 3D region like a momentum source, but does include radial variation and rotation. However, this won't work well with a disk that experiences a pressure drop (i.e. momentum extraction vs addition). The swirl velocity on face 2 is driven by the pressure change which must be positive for rotation to develop.

Momentum Source: This is the classic old school definition for momentum sources. You specify a 3D region and give it a step change in momentum in cartesian coordiates (probably others too). Unlike the previous options, this is not dependent upon inflow conditions. (i.e. not dependent upon flow rate, velocity, etc...)

What exactly is your disk modeling? For something like an airplane propeller or wind turbine I would use blade element virtual disk. For a marine propeller body force prop method (so sayeth the help doc's). For simple fans, the fan interface is fast and easy. I would try to avoid the momentum source if I can, at least for a classical disk problem. The other tools are likely more applicable to your situation.

[joona]

What would be the best way to model a high bypass turbofan?

[Tellur]

Originally Posted by MBdonCFD

Fan Interface: Creates an interface between two regions that applies a momentum change between faces 1 and 2. This is not a 3D region like a momentum source, but does include radial variation and rotation. However, this won't work well with a disk that experiences a pressure drop (i.e. momentum extraction vs addition). The swirl velocity on face 2 is driven by the pressure change which must be positive for rotation to develop.

Hi Matt,

Sorry to dig this up. Was wondering if you could advise on the Fan Interface. Is it generally possible to apply such an interface without knowing the performance characteristics, perhaps pressure drop, from the actual fan? Also, would this method be more accurate than the typical MRF option?

Ehm hope this makes sense (physics not my strong suit, I know ironic).

Kind regards,
Theodore.

[fluid23]

No, the fan interface cannot be employed without knowledge of fan performance. It is only meant to approximate the effect of a fan between one interface and another. With MRF you are actually modeling it.

[fluid23]

To clarify, it takes the inflow condition at the upstream interface then applies a pressure rise and swirl to the flow at the downstream face based on the performance data you specify. Everything in between is 'magic'.

[Kiran Peter]

Hi Matt,

Could you please explain how to apply the momentum source technique properly on the 1 m thick actuator disc ?

Thank you

Originally Posted by fluid23

Star-CCM+ has lots of options for this and from your choice of words it isn't clear what you are doing.

Virtual Disk: Creates a momentum source that is intended to model rotor geometry. This means it is typically spinning. There are 3 options, body force propeller method and blade element momentum method both account for radial variations so you should not see a uniform momentum change. The third option is the 1-D momentum method and this I believe will give uniform loading. I have never used it.


Fan Interface: Creates an interface between two regions that applies a momentum change between faces 1 and 2. This is not a 3D region like a momentum source, but does include radial variation and rotation. However, this won't work well with a disk that experiences a pressure drop (i.e. momentum extraction vs addition). The swirl velocity on face 2 is driven by the pressure change which must be positive for rotation to develop.

Momentum Source: This is the classic old school definition for momentum sources. You specify a 3D region and give it a step change in momentum in cartesian coordiates (probably others too). Unlike the previous options, this is not dependent upon inflow conditions. (i.e. not dependent upon flow rate, velocity, etc...)

What exactly is your disk modeling? For something like an airplane propeller or wind turbine I would use blade element virtual disk. For a marine propeller body force prop method (so sayeth the help doc's). For simple fans, the fan interface is fast and easy. I would try to avoid the momentum source if I can, at least for a classical disk problem. The other tools are likely more applicable to your situation.