Problem with .step format import


I'm pretty new to this software, and I currently have the 32 bit version of the PANEL software. I'm using the programme to carry out my 4th year university project which is looking at sailplane aerodynamics. I have been using OpenVSP for creating my geometries and Rhino3d to shift these into either .step or .igs formats.

I have, however, never been able to run a proper analysis in it as the solver fails. I predict that it is a meshing problem and I have tried to implement the solutions in the FAQ section to no avail. Is there any way in which I could possibly upload a geometry for some feedback on where I am going wrong?

Any help would be greatly appreciated as I feel like by now I have exhausted all options:P


Skin friction and facets

Take a step back. What is your main objective? I'd guess to design a high L/D airfoil. In which case you will need to model skin friction. The panel method does not model skin friction, so I don't see how this path you've chosen is going to help you.

On a more practical note, OpenVSP exports facets, so trying to convert them to NURBS in Rhino is pointless - you'll end up with an unmanageable number of planar surfaces - one for each facet. Also there is a good chance that the wings pass directly through the fuselage, with no intersection. For simulations you need water tight geometry.

I'd suggest you use a true NURB modeler such as Rhino or a CAD system to generate your geometry.

Sorry I'm late to this

Sorry I'm late to this discussion, but I wanted to set the record straight for posterity.

While OpenVSP can generate a mesh, it also does output smooth NURBS surfaces. The *.3dm Rhino export the user was most likely using are smooth NURBS surfaces. Rhino can write those out as STEP or IGES files for just about any CAD software to use.

The upcoming OpenVSP v3 will drop support for Rhino files, but will instead write STEP files directly. Since Rhino can read STEP files, we aren't losing anything.

More information

The aim of the project is to consider the design of a lower winglet on the sailplane and see if it provides performance benefits. The skin friction will be hand calculated, so using the panel method was to find the induced drag.

Vsp has a mesh function built in, with a subsequent possibility to export to CART3D in *.tri format. This unifies the geometry into a watertight mesh and, by reading about, it seems that it operates very well in the CART3D capacity. I was going to try this and see if any progress could be made from there.

Otherwise, would you say then that trying to use OpenVSP for Caedium geometries is a useless idea?

Thanks for your help, it's greatly appreciated. I've been fretting over this quite a bit haha! (as deadline day looms in a couple of months:P)

Try OBJ format

Thanks for the clarification.

Given you are using the old 32-bit version of Caedium, you are limited to using Alias/Wavefront (.obj) as the geometry format. If you can somehow massage the facets from VSP to be water tight in OBJ then you might be able to make progress in Caedium. You will need to have a geometry feature line along the wing trailing edges to attach your wake.

You will still likely face problems around high curvature regions if the facets are too coarse. For more see "Geometry Exchange".

I still recommend that you try a different geometry route using a NURB modeler.

Still not working

I started again from scratch creating the models in Rhino 3D within the university. However, on inputting the files it still does not work and as a results I've had to go about my analysis by different means. I think a large part of the problem must be down to having a 32-bit Windows system. Although, when I had the trial period and was using simpler geometries, things were working good ( minus a few graphic troubles).

Is it possible to cancel the license? Unfortunately the software has been of no use. Although i would definately reconsider trying it at another date on a 64-bit machine once one had became available to myself.

Thanks for your help.

Be specific

Can you be more specific about what isn't working?

  • Is your geometry water tight?
  • Did you try using the Accuracy tool to improve the resolution of the leading edges?
  • Did you attach wakes to trailing edges and make sure the wakes don't intersect geometry downstream?

Panel methods work best when you simplify your geometry as much as possible - any sharp edge will cause a problem unless it's a trailing edge with a wake attached.

Please share your geometry (STEP file) so I can I take a look? I recommend using Dropbox and our contact form.

There is no fundamental difference between the 32-bit and 64-bit Panel methods.

Once a license is activated time starts counting down and it can not be canceled.