[cosmictrex]

Hi all,


I'm working on modeling propeller slipstream with the actuator disk model. OpenFOAM offers this in fvOptions and the tutorials include a turbine siting case.


What I need is a velocity source from this disk which would flow over a lifting surface, simulating the wing-in-slipstream effect. I assumed that actuationDiskSource allows me to do this. However, the induced velocity from the propeller is a function of the propeller rotational speed, and there is no input parameter of actuationDiskSourcee related to the propeller rotational speed. So I'm not quite sure if this is possible in OpenFOAM.


My question is: how do I model actuator disk as a propeller to create a slipstream source in OpenFOAM?

[hirota]

Originally Posted by cosmictrex

Hi all,


I'm working on modeling propeller slipstream with the actuator disk model. OpenFOAM offers this in fvOptions and the tutorials include a turbine siting case.


What I need is a velocity source from this disk which would flow over a lifting surface, simulating the wing-in-slipstream effect. I assumed that actuationDiskSource allows me to do this. However, the induced velocity from the propeller is a function of the propeller rotational speed, and there is no input parameter of actuationDiskSourcee related to the propeller rotational speed. So I'm not quite sure if this is possible in OpenFOAM.


My question is: how do I model actuator disk as a propeller to create a slipstream source in OpenFOAM?

Hi,
I'm facing the same problem.
Did you solve the problem here?

[cosmictrex]

Originally Posted by hirota

Hi,
I'm facing the same problem.
Did you solve the problem here?

I have not solved the problem of propeller-as-actuator disk formulation in OpenFOAM. However, I have found several alternatives:

• Dynamic mesh (AMI) + RANS -- This is simple to setup and actually rotates the mesh to simulate propeller rotation at the cost of computational power. This is because UAV propellers spin at relatively high speed compared to the free-stream flow and the unsteady solver needs small time-step to be stable. I successfully completed a case with a Courant number of 2 (PIMPLE). For my application, this requires too much computational time and data storage.
• Steady MRF -- although I have not been successful in getting the solution to converge.
• VLM -- This is what I'm working with right now. VSPAERO comes with propeller models (actuator disk and rotating blades) implemented in vortex lattice formulation. This is the least expensive method I have found but it does not capture all the physics (limitation of VLM). I'm working on accuracy assessment of this method for my application. It seems to yield a somewhat large amount of wing-in-propeller moment prediction error compared to RANS solver. For more information check out this recently released paper.

Cheers

[hirota]

Originally Posted by cosmictrex

I have not solved the problem of propeller-as-actuator disk formulation in OpenFOAM. However, I have found several alternatives:

• Dynamic mesh (AMI) + RANS -- This is simple to setup and actually rotates the mesh to simulate propeller rotation at the cost of computational power. This is because UAV propellers spin at relatively high speed compared to the free-stream flow and the unsteady solver needs small time-step to be stable. I successfully completed a case with a Courant number of 2 (PIMPLE). For my application, this requires too much computational time and data storage.
• Steady MRF -- although I have not been successful in getting the solution to converge.
• VLM -- This is what I'm working with right now. VSPAERO comes with propeller models (actuator disk and rotating blades) implemented in vortex lattice formulation. This is the least expensive method I have found but it does not capture all the physics (limitation of VLM). I'm working on accuracy assessment of this method for my application. It seems to yield a somewhat large amount of wing-in-propeller moment prediction error compared to RANS solver. For more information check out this recently released paper.

Cheers

Thanks for the reply.

I also did this method "Dynamic mesh (AMI) + RANS". I had the same impression as you. The computational cost is very high... It takes too much time...

The MRF calculation for the propeller alone was successful for the time being, but I am still considering whether it will be effective for me as I have to do the calculation including the hull.

I have never tried VLM, but I am very interested in it.
Does this VLM require any rewriting of the code?
In short, what are the steps to represent VLM?

[cosmictrex]

Originally Posted by hirota

Thanks for the reply.

I also did this method "Dynamic mesh (AMI) + RANS". I had the same impression as you. The computational cost is very high... It takes too much time...

The MRF calculation for the propeller alone was successful for the time being, but I am still considering whether it will be effective for me as I have to do the calculation including the hull.

I have never tried VLM, but I am very interested in it.
Does this VLM require any rewriting of the code?
In short, what are the steps to represent VLM?

I would really appreciate it if you could share your MRF case. I could not get the solution to converge even for a very simple case and still wondering what I did wrong.


The VLM route does not require any additional code writing. OpenVSP (open-source) comes with GUI which helps you model your vehicle including the propeller very effectively. There might be some learning curve in how to use the software but overall it is relatively easy to set up + much cheaper computationally compared to FVM. The geometry and solver setup are done within the same program. You can learn more from this wonderful presentation + demo of VSPAERO.


Good luck

[hirota]

Originally Posted by cosmictrex

I would really appreciate it if you could share your MRF case. I could not get the solution to converge even for a very simple case and still wondering what I did wrong.


The VLM route does not require any additional code writing. OpenVSP (open-source) comes with GUI which helps you model your vehicle including the propeller very effectively. There might be some learning curve in how to use the software but overall it is relatively easy to set up + much cheaper computationally compared to FVM. The geometry and solver setup are done within the same program. You can learn more from this wonderful presentation + demo of VSPAERO.


Good luck

Sorry for the delay in getting back to you.

In my case, the propeller is placed inside the analysis region of a rectangular body, and the rotation of the propeller is taken into account in the MRF.
In this case, the calculation converged and I was somewhat satisfied.

However, now I have extended this case and the calculation is not converging.
In this case, the propeller is mounted behind the hull, and the rotation of the propeller is taken into account in the MRF.
The only difference between this case and the previous one is that the analysis area is extended and the hull is placed in front. I may have made a simple mistake, but the cause is unknown.

If you want to know more details, please let me know.


I'm a bit hesitant to start a new OpenVSP project, as the time left is very limited for me.

[Sachin m]

Originally Posted by hirota

Sorry for the delay in getting back to you.

In my case, the propeller is placed inside the analysis region of a rectangular body, and the rotation of the propeller is taken into account in the MRF.
In this case, the calculation converged and I was somewhat satisfied.

However, now I have extended this case and the calculation is not converging.
In this case, the propeller is mounted behind the hull, and the rotation of the propeller is taken into account in the MRF.
The only difference between this case and the previous one is that the analysis area is extended and the hull is placed in front. I may have made a simple mistake, but the cause is unknown.

If you want to know more details, please let me know.


I'm a bit hesitant to start a new OpenVSP project, as the time left is very limited for me.

Hi hirota,

Were you able to solve the issue.
If possible could you share your case set up with me

Thanks

Best regards
Sachin