[liladhar]

yes this is a free surface model...with well defined interface between two phases(air and water)...
the disc is absolutely below the waterline...and as the disc is below the waterline streamline pattern for air superficial velocity...cannot be calculated...

[ghorrocks]

I do not know why it is not working for multiphase. You are going to have to fiddle with it and see if you can get it to work. Alternately contact CFX support.

[liladhar]

thanks glenn for your kind help.
i sent the querry to ANSYS support...waiting for their reply.
Is there anybody in this forum who can simulate propeller behind the ship hull using actuator disc?
Please guys help me....I would be very thankfull to you....
thanks in advance.

[Esan]

Hi, I'm working on a project that requires to replicate a propeller using a momentum source. I have the propeller specifications but I'm still unsure on how to define the expressions that is required for axial, radial and theta components.

I would really appreciate if anyone is able to provide some guidance.

[ghorrocks]

Which bit do you not know? How to set the momentum source (this is covered in the CFX documentation), or how to specify it in ax, rad, cyl components (google cylindrical coords or something like that)?

[Esan]

Originally Posted by ghorrocks

Which bit do you not know? How to set the momentum source (this is covered in the CFX documentation), or how to specify it in ax, rad, cyl components (google cylindrical coords or something like that)?

Ghorrocks thanks for your quick reply.
I've already generated the momentum source and tried the cartesian option, but because the flow was not realistic, I have to try the cylindrical option for which I'm not sure how to specify the three components.

I suppose that each component is equal to density times its own velocity component divided by timestep. However, the question is that how to specify these three velocities as expressions or functions as they are not uniform.

The dimensions of the propeller, the RPM and the maximum velocity to be produced are known.

[ghorrocks]

Most people do not bother with the radial or theta components. Most people just model the axial component. That will make things much simpler. You can easily add a turbulence source as well if that helps.

Generally fans are implemented as momentum sources by making the exit axial velocity of the fan equal to the fan performance curve at that pressure drop.

[Esan]

Originally Posted by ghorrocks

Most people do not bother with the radial or theta components. Most people just model the axial component. That will make things much simpler. You can easily add a turbulence source as well if that helps.

Generally fans are implemented as momentum sources by making the exit axial velocity of the fan equal to the fan performance curve at that pressure drop.

Thanks for your time and consideration. I need to produce the swirl as well because the dispersion of the flow is important for my application.

Sorry for asking too many questions, I'm new to the software and unfortunately started with a difficult project. Would you please confirm if F=den*velocity/timeStep is correct? and if yes, how the velocity components should be expressed?

How does the turbulence source work?

[ghorrocks]

If dispersion of the fan's outflow is important then it is probably more important to get the turbulence right rather than include the swirl component.

Please post your CEL so I can understand what you have done in context.

Also please let us know what you know about the fan you are modelling. Most fans have a flow rate versus pressure difference curve. You can implement this performance curve in the source term. But if you know some other performance parameter for the fan you will have to do something different.

The easiest way to do your components would be to define a coordinate frame at the centroid of the fan and define the source term components using that.

[Esan]

Originally Posted by ghorrocks

If dispersion of the fan's outflow is important then it is probably more important to get the turbulence right rather than include the swirl component.

Please post your CEL so I can understand what you have done in context.

Also please let us know what you know about the fan you are modelling. Most fans have a flow rate versus pressure difference curve. You can implement this performance curve in the source term. But if you know some other performance parameter for the fan you will have to do something different.

The easiest way to do your components would be to define a coordinate frame at the centroid of the fan and define the source term components using that.

The model is a propeller of a marine vessel with the following specifications (model scale):
Diameter : 0.18m
Hub diameter : 0.0547m
Number of blades : 4
RPM : 660
100% Pitch
The maximum velocity (calculated by Dutch method) : 1.89m/s


I've not specified any expressions yet to be shown in CEL. I was trying to play around with numbers in the three components. Here's the resulted streamline and velocity contour. The flow profile is what I'm looking for however, I need to verify it with my prop. http://www.4shared.com/download/hY7UNqobba/3CompStr-Zoom.png?lgfp=3000

[singer1812]

In past for swirl on fans I have done it as Glenn said.

Subdomain of Fan Flow area (hub portion is removed).

Axial flow is as per fan curve from manufacturer.

I usually ignore the radial component.

The theta component I just use based on a CEL expression involving fan subdomain location distance from hub centerline*Fan RPM

[Esan]

Originally Posted by singer1812

In past for swirl on fans I have done it as Glenn said.

Subdomain of Fan Flow area (hub portion is removed).

Axial flow is as per fan curve from manufacturer.

I usually ignore the radial component.

The theta component I just use based on a CEL expression involving fan subdomain location distance from hub centerline*Fan RPM

Thanks,

I am getting not realistic flow profile when I include the theta component when specified an expression for theta as follows:

theta component= (den*(radius*RPM))/timestep
But I've put the radius as constant, how is it?

if not how should I specify the theta component?

[singer1812]

Radius isnt constant. Theta flow velocity will be higher on outside of fan than it will be near hub.

You need to use the actual node distances from the hub centerline.

How are you verifying "not realistic"? Axial velocity (as a function of radius) can be easily checked against your radius*radians/sec (I hope you are not keeping the number in RPMs).