VSPAERO on STL Model?

[Luke Bergeron]

I have a simple airframe STL (attached) that I'd like to do VSPAERO analysis on. It loads in OpenVSP without error, but when I run VSPAERO, nothing happens apart from a few errors (also attached).

What am I doing wrong/how can I do analysis on my STL? The wiki doesn't have any info on this.

Thank you!

[Cibin Joseph]

Quick answer:

VSPAERO only works on geometry generated in OpenVSP. An STL file is just a collection of triangles/points with normal information. You'll need to fit a surface to it or create wings and fuselage that match the dimensions to perform analysis using VSPAERO.

Cibin

--
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/816350eb-fb37-4e71-8bc4-7bf9bd769810n%40googlegroups.com.

[Rob McDonald]

VSPAERO may actually run on an imported geometry like this.  However, there will be some challenges.  First, you must make sure that the imported geometry is watertight.

Second, you will need to run VSPAERO in Panel method -- and I would recommend running with the 'alternate file format' selected.

Finally -- and this is the big problem.  The mesh you have shown in the screenshots is inappropriate for a solution.  The triangles shown are not sufficient to represent a wing -- you need better spanwise resolution for a good aerodynamic solution.  Most CAD programs only export STL files suitable for 3D printing -- the geometry (straight lines) can be represented with a small number of triangles.  However, the aerodynamics can not be represented with such a coarse model.  If you can change settings in your CAD program, you might be able to decrease the maximum edge length to achieve a suitable mesh.

Otherwise, I would encourage you to re-create your geometry in OpenVSP.  It looks like a very simple geometry that should not take long to recreate.

Rob

--

[Cibin Joseph]

@Rob McDonald didn't know VSPAERO could run the Panel method on STL geometry. That's interesting.

Luke, here's an eyeballed starting point [VSP file] if you want to play around with VSPAERO. You'll have to adjust the dimensions and airfoils to get an exact match though.

Cibin

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/CAEppYpHsjG%2BYHNNACuU1C%2B_8Hur4pehmomiZxpXq7grZVDNaQQ%40mail.gmail.com.

[Luke Bergeron]

I tried to model the plane in OpenVSP using the native toolset and got something close, but the airfoil seems to be a bit tricky. I make my own .dat file since my airfoil can't be made with a NACA 4 series AFAIK (I'm an EE not an Aero E, so I have limited experience). When I used the custom .dat file to form the main wing section, I see odd black wireframe traces on the wing. What's going on here and will it affect the accuracy of my VSPAERO simulation?

I've attached the .dat file and screenshot of the wing section with simulation results. CG was auto-calculated to (2.55, 0, 0.271).

Also, do the simulation results look right?

Thanks!

[Rob McDonald]

Looking at the wing tip wireframe, I can see that your airfoil file does not have information in the right format for OpenVSP.  Where did you get the point file?  What airfoil are you trying to use?

Because of the strange way the wing is being modeled, I would not yet trust the VSPAERO results.

I also would not use the auto-calculated CG.  That CG works -- but it assumes you've put in detailed density information for all of the parts -- and you've identified them as either a thin skin, or a solid part.

Instead, for most airplanes, you have some freedom to move internal parts around to get the CG where it needs to be aerodynamically.  Therefore, for initial calculations, I would either place the CG at the nose (typically 0,0,0) , or the 1/4 chord point of the wing.

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/c104a4ca-973a-40b3-9ab6-0f39fb0ba27cn%40googlegroups.com.

[Luke Bergeron]

I made the .dat more or less by hand based on a previous CAD file of a custom wing. How can I fix it? Do you know of any NACA airfoils that would closely match it?

Yeah, I figured the CG was not in the right spot, but OpenVSP has a horrendous UI, so I didn't want to mess with CG too much, but I also really didn't want it at (0, 0, 0).

[Rob McDonald]

I'm sorry that you feel that OpenVSP has a horrendous UI.  Do you have any constructive suggestions to improve it?  Most OpenVSP users much prefer it to CAD or other ways of modeling an aircraft.

Many people asking for help wouldn't start by trashing the tool to the primary developer.  You do you though.

Was the airfoil in the CAD anything in particular?  Aerospace engineers don't generally just draw horrendous airfoils in CAD and then treat them as gospel.  We also usually stay away from horrendous airfoils on horrendous airplanes.

You will find a collection of thousands of airfoils on the UIUC Airfoil Data Site.  OpenVSP should read any of the files on that site.  There are three formats typically used -- they usually represent both the top and bottom surface of an airfoil with about the same number of points -- and typically similar spacing.  Many of the formats start at the trailing edge lower surface, wrap around the leading edge and then finish at the trailing edge upper surface -- but other formats do it differently.

Your airfoil looks to be about 16% thick and it has a flat bottom -- which is easy to build, or to model in horrendous CAD programs.  But horrendous airfoils drawn in horrendous CAD usually exhibit horrendous aerodynamic performance.  The Clark Y is an old and popular airfoil that has a flat bottom, you might consider it.  The NACA 4412 is a fairly close match to a Clark Y.

Your airfoil seems to be rotated such that the bottom surface is 'horizontal'.  This can be accomplished with incidence or by modifying the airfoil.  But it was probably just another arbitrary choice by a lazy CAD operator who didn't have a good reason to set the wing at any other angle.  Usually, an airfoil's points are specified such that its 'natural' position has specific aerodynamic meaning -- though a very long time ago, airfoils would have been drawn with the flat bottom 'flat'.  A wing's incidence is usually set based on the trim condition for the airplane -- the relative angle between the wing and the horizontal tail is called decalage, it matters.

You might try approximating your airfoil as a NACA 4416 to start.  You could try the 4-digit modified or other airfoil types to get some more freedom to approximate your horrendous shape.

Running an aerodynamic analysis with the CG at 0,0,0 is OK.  The CG is only used as a reference point for the moments.  After the aerodynamic analysis is finished, you can easily re-compute the moments about any other point using statics.  One common procedure is to run first with 0,0,0 and calculate the position of the neutral point.  Then, you might set the CG about 10% of the chord in front of that location.

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/6a95b5f3-7579-4198-be4f-895e47d44cb2n%40googlegroups.com.