Difficulty in Obtaining Aerodynamic Coefficients for Lenticular Geometry

[Kartikey Karnwal]

Respected members,

I am currently working on analyzing a lenticular-shaped geometry in OpenVSP, and I would like to confirm whether it is possible to calculate aerodynamic coefficients (such as CDtot and CDi) for this type of configuration.

To provide details, I created the geometry starting from an ellipsoidal shape with the following parameters:

• X radius = 2 m

• Y radius = 2 m

• Z radius = 0.5 m

For the analysis, I set:

• Uw = 40

• Um = 60

All the input values and parameters were calculated and set correctly. However, the issue I am facing is that the tool does not generate the expected coefficient values. In fact, the value of Cdtot on the graph consistently comes out as 0, even after proper setup.

This raises a question: Is it possible to calculate aerodynamic coefficients in OpenVSP for geometries that do not have a conventional wing-like configuration, such as a lenticular body? If yes, could you please suggest the correct approach or adjustments required to obtain meaningful results?

I have attached the .vsp3 geometry file and relevant screenshots of my setup for reference. Any help, suggestions, or references would be greatly appreciated.

Thank you for your support.

[Brandon Litherland]

For non-lifting bodies, such as fuselages, pods, tanks, etc., the recommended approach would be to estimate the parasitic drag using the Parasite Drag tool.  The inviscid, potential flow based, VSPAERO solver will not accurately model this type of problem.  In this case, there should be some drag equations out there to estimate the drag on such a body. I'm not thinking of an exact formula off the top of my head but you could likely find something useful in Raymer's Aircraft Design or Hoerner's Fluid Dynamic Drag.  Many of the VSP drag formulas in the PD tool are from Raymer or Hoerner.  Hoerner will likely be your best bet.

To restate, VSPAERO does not solve for separation, stall, or viscous flows.  Any components of the inviscid solution are estimates and corrections rather than computed solutions, per se.  For something like this, you would need Navier-Stokes type solvers like OVERFLOW, StarCCM, PowerFlow, OpenFOAM (free), and the like.

[skyc...@gmail.com]

A follow on question to this, just confirming that any geometry in the thick set do not contribute to lift either, correct?

[Rob McDonald]

Any component -- thick or thin -- that has a wake attached will contribute to forces and moments.

Any component -- thick or thin -- that does not have a wake will only contribute to moments.

OpenVSP attaches wake lines to 'wing' type components -- this is typically wings, propeller blades, and sometimes bodies of revolution.

When the OpenVSP GUI prepares a model for running in VSPAERO, it draws the identified wake lines on the screen.

In the supplied file, the geometry was modeled with an ellipsoid -- which does not get a wake attached, so it will never support a force.

I made a quick approximation of the shape using a Wing component with Ellipse airfoils.  I used blending to make it reasonable, but the wing tips aren't great.

An approach like this will support forces.

However, you must then think about the physics.

This model is specifying that the flow separates from the aft point of an ellipse.

In reality, predicting the aft separation point for blunt body flow is an extremely tricky problem.  The flow will curve around the trailing edge to some extent, reducing the force on the body, approaching the wake-free solution.

I have run a smooth blunt body 'potato' shape in Euler CFD before.  This body had no symmetry in any axis or direction.  There were no sharp edges anywhere to force separation.  Run at any angle of attack, the CFD solution 'found' the separation point that results in zero forces on the body -- i.e. the wake-free solution that VSPAERO would give.

A RANS code would not be able to find this force-free solution -- its flow would separate somewhere else, causing there to be forces on the body.  However, who knows if that separation location is correct.

Try mounting a blunt body in the wind tunnel -- if you want to put a sting out the back, then that sting will influence the separation location on the aft body.

Trying to get lift out of a blunt shape is hard.

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.
To view this discussion visit https://groups.google.com/d/msgid/openvsp/9f1228fc-9102-405a-9210-39db892cff1fn%40googlegroups.com.

[Kartikey Karnwal]

Dear Sir,
Thank you very much for your prompt response. I will certainly review all the parameters you provided.

To view this discussion visit https://groups.google.com/d/msgid/openvsp/CAEppYpHoB-09jZWUUQ%3Dr-o0_ugkdukfXK_dmR7xU5hRvEm6NXQ%40mail.gmail.com.