About accuracy of drag/lift-to-drag ratio

[Eric Cheng]

Hi all,

When I used VSPAERO to calculate plane or wing drag coefficient, I suddenly thought of a serious problem. 

Owing to the Vortex-Lattice Method or the Panel Method does not consider the viscosity, so its drag coefficient is only the induced drag coefficient, not including the frictional resistance. 

So if the calculated CL is accurate, but the total drag coefficient may be inaccurate because the friction drag cannot be captured. If so, is the lift-to-drag ratio calculated using VSPAERO unreliable?

So will the lift-to-drag ratio or the total drag coefficient calculated by VSPAERO have a big error with the true value?

Has the stickiness（viscosity）correction been considered in the VSPAERO source code? For example, these correction methods of wall function.

I would be very grateful for any reply，Thanks in advance.

Eric

[Rob McDonald]

VSPAERO has a parasite drag estimate -- but it will certainly underpredict the drag of your aircraft.

It may do OK for the components it knows about, but real airplanes have various sources of drag that are left out of a VLM or Panel type model.  These can't possibly be captured by VSPAERO.

So yes, the L/D calculated by VSPAERO will be optimistic.

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 on the web visit https://groups.google.com/d/msgid/openvsp/cc9eacc6-14c0-4a95-8eb1-63a2ad271f03n%40googlegroups.com.

[Eric Cheng]

Okay, thank you very much, Rob.

Is there a way in VSPAERO to automatically superimpose the results of the calculation of parasite drag and induced drag for visualization?

 Because in a low-speed general aviation aircraft, the total drag of the aircraft is mainly composed of induced drag and frictional drag. If the result of the superposition of these two is as accurate as possible, then maybe the lift-to-drag ratio result will not be too bad?

Eric

[Brandon Litherland]

The CD0 from VSPAERO is a good first estimate of the solution but you'll notice that CD0 is actually lift-dependent meaning that Dave Kinney is approximating an empirical fit of the NACA 0012 viscous drag to the model.  A good way to capture or break apart (and then sum) all the sources is to run VSPAERO and subtract the minimum reported CD0.  This would be the pure parasite (non lift-dependent) contribution from VSPAERO based on a NACA 0012.  Keep the remaining CD0 difference as CDi_visc.  Then run a quality parasite drag build using the Parasite Drag tool.  See Raymer or other texts for good values to use here and what to apply for extra excrescence drag.  The sum of CDi + CDi_visc + CD_parasite will be pretty darn close to expected performance.  Any addition accuracy needed is tuning your model to known configurations and then checking that the results are repeatable.  This takes some effort but is well worth the trouble.

[Rob McDonald]

We've thought about adding something like that, but have not done so yet.

That said, the visualization of VSPAERO results in OpenVSP (in the Results Manager) is not really meant to be the end of your visualization journey.  This is not a complete plotting tool that allows you to adjust formatting, size, fonts, etc.  It can not be used to plot arbitrary stuff.  It is not intended to produce publication ready plots (or even boss ready plots).

The results visualization built-in to OpenVSP is meant to be a quick-check diagnostic tool to help you develop confidence in your solution.  In my experience, the best way to diagnose many things is through visualization.  Also in my experience, people do not perform visual checks nearly as often as they should.  So, we developed this tool to make it trivially easy and to force the user to look at the result.  

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/d803e284-5fdf-441d-8760-01c75690b1e3n%40googlegroups.com.