Propeller simulation gone wrong

[Bruno Marques]

Dear all,

I am currently doing a VSPAERO simulation for a wing with 2 propellers (one on each side of the wing) for 3 different combinations of advance ratios J and thrust coefficients Tc:

Case 1 - J=1.11, Tc=0.024

Case 2 - J=0.95, Tc=0.101

Case 3 - J=0.81, Tc=0.2

I have done the simulations with VLM analysis, Mach 0.147, alpha 0º-10º with increments of 1º, 11 wake iterations, leading edge suction/vortex lift analysis, Vinf=50, Rho=1.225 and the remaining rotor parameters were calculated using the expressions of J and Tc.

However, the alpha-CL curves for the 3 cases that resulted from these simulations are exactly the same. Am I doing anything wrong?

Kind regards,

Bruno 

[Rob McDonald]

What do the lift distributions look like? The Cp plots in viewer? Do
the other results match your intuition for these cases and variations?

Rob

> --
> You received this message because you are subscribed to the Google Groups
> "OpenVSP" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to openvsp+u...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Bruno Marques]

Hi Rob,

The lift distributions are basically the same for the 3 cases so something strange is happening or maybe I am doing some error in the simulation. Attached to this email you can find the results that were exported for the 3 cases in excel and corresponding csv files.

Kind regards,

Bruno

[Jasper]

Hi Bruno,

They shouldn't be the same. Can you share your .vsp3 or .VSPAERO files?

Maybe the propeller calculations did not work; did put in your power coefficient as well, and how did you calculate it? Normally you would need additional info besides the advance ratio and thrust coefficient to determine the power coefficient. 

[Bruno Marques]

Dear Jasper,

I am comparing the results of this simulation with the experimental results obtained in the pdf attached (page 107). Basically, this paper says the following about the experiment:

- Propeller diameter of 0.236 m.

- Dynamic pressure of 1500 Pa.

- The advance ratios specified before.

- Shaft power of 3.6 kW.

- The thrust coefficient associated with each advance ratio.

Assuming a air density of 1.225 Kg/m^3 and with the formula of the dynamic pressure I calculated the air velocity of the experiment (about 50 m/s). With this velocity and with the formula of the advance ratio I calculated the number of rpm of associated with each advance ratios. Than with this and by knowing the shaft power I calculated the power coefficient (P/(rho*n^3*D^5)). Than I had all the variables to make the simulations.

Attached to this document you can find the .vsp3 files of each simulation and a excel file with all the variables used. You can find all the attached files here: https://drive.google.com/open?id=1v88j-TLQf7jN6bUOl3ZbFIFv_6pYWvUU

Thank you in advance for your feedback.

Kind regards,

Bruno Marques

[corp...@gmail.com]

Hi Bruno,

I don't understand if you used one configuration ( I mean different geometry )  for all three cases, or one config. for each case.

Corrado

[corp...@gmail.com]

Sorry, wrong sentence..

....( I mean different geometry ) is ( I mean the same geometry)

[Bruno Marques]

Hi,

Sorry it was my mistake. I used the same geometry for all the cases, which was the one attached to this message.

Kind regards,

Bruno

[corp...@gmail.com]

Did you use the three thrust coefficients listed in the document, keeping the power constant at 3.6 kW in all three cases ?

[Bruno Marques]

Yes exactly. I did that

[corp...@gmail.com]

I have enclosed the curves obtained considering different CP for each J and CT , rather than a constant CP as in your case. This is based on the assumption that CT is a function of the thrust, the thrust is function of the power, then you cannot have different CT and J and only one CP.

As you can see this approach follows the  trend shown in the document. I used your model, slightly modified in the region of the propeller’s wake, where a denser tessellation is required to get a more defined wake. Also the tesselation of the disk is increased. Propeller rotation is inboard.

I hope this can help and any comment is welcomed.

Corrado

[Bruno Marques]

Hi Corrado,

Thank you very much for all the help. This is a very important part of my thesis and was stuck on it for a few weeks. Can I ask you what CP values did you use for each J and how did you calculated them?

Kind regards,

Bruno Marques

[corp...@gmail.com]

Here the main passages:

J=V/(n*D) -> n= V/(J*D)   [1]

CP= P/(rho*n^3*D^5)      [2]

Thrust= prop_efficiency*Power/airspeed   [3]

CT= Thrust / (rho*n^2*D^4)= prop_efficiency*P/(airspeed*rho*n^2*D^4)    [4]

( Use prop_efficiency=1 , since different values don’t affect the trend you are searching for.)

Since you have three values of CT, for each CT find the value of P that satisfy eq. [4], put the found P value in [2] to obtain CP. Now you have ,for each J, the couple (CT,CP) that, with D, n are the necessary input for VSPAERO.

I hope there are no errors above, but check it as well and good work for your thesis.

Corrado

[Bruno Marques]

Hi Corrado,

Thank you for all the help. Just one question, what do you mean by "propeller rotation is inboard"? Is it inboard up or inboard down? Is that a positive value for the RPM in VSPAERO or a negative one? Or is the propeller rpm positive for one disk and negative for the other?

Kind regards,

Bruno Marques

[Bruno Marques]

Besides, I would also like to know, in the advance case setup option of VSPAERO, which method did you use? Was it "Batch calculation", "Jacobi Preconditioner" or "Leading edge suction"?

Attached to this message you can find the updated version of my excel with the new CP values and RPM values. Can you tell me if they are the same of yours?

[Rob McDonald]

Those options shouldn't influence your answer.

Batch calculation -- only determines whether alpha/beta sweeps are
handled inside or outside of VSPAERO itself. No change in results
should occur.
Jacobi Preconditioner -- determines details of how matrix solver
works. Should not influence flow solution in any significant way.
Typically don't need to change this setting unless you experience
stability regressions from an earlier version.
Leading edge suction -- model should only kick in with significant
leading edge sweep.

Rob

[Bruno Marques]

Ok. Thanks for the information Rob.

[corp...@gmail.com]

I left “batch calculation” on only. I used the file I attached last time, but I didn’t save each calculation separately, just record each result and then changing the rotors parameters for the new analysis in the same model. Therefore the curves are derived by such records, taken at three a.o.a. for each case (0,5,10 deg), to speed up the test.

Propellers are rotating inboard down, but it was a random choice.

Corrado

[Bruno Marques]

Thank you very much for the support. It worked perfectly. I have another question. I now need to make two plots of the spanwise distribution of the tangencial force and normal force. Which are these variables in VSPAERO? Are threy cy and cx?

[Rob McDonald]

CFx,CFy,CFz are the force coefficients in body-axes -- x, y, z as
expected. For small AOA, these will basically be CD,CL, CS (side) in
wind axes.

I'm not sure what you mean by tangential in this case, but hopefully
the above definitions are clear.

Rob

[Bruno Marques]

I am sorry. I meant the tangential and normal force coefficients. I wanted to compare the results of VSPAERO with the ones that are on the pages 114 and 115 of the pdf file attached here: https://drive.google.com/drive/folders/1v88j-TLQf7jN6bUOl3ZbFIFv_6pYWvUU

[Rob McDonald]

On a 3D wing, the word 'tangential' does not uniquely specify the
direction of a force.

Reading the text of the document you link to, they indicate that they
are integrating an airfoil section to result in the forces they are
considering. From that context, it is clear that they are referring
to cx as the tangential force.

Rob

On Mon, Feb 12, 2018 at 7:37 PM, Bruno Marques

[Bruno Marques]

Ok. Thank you for the support Rob.

Kind regards,

Bruno