Understanding surface import modifications

[JA_10341562]

Hello community.

Looking for someone who's walked my particular path before. I'm deep in the docs and forum posts, and I'm looking for a friend to lead me back out of the forest. 🙂

I work in airplane design. I am investigating options for a workflow where data about the final skin of an aerodynamic surface (i.e. wing, aileron, canard, etc) is generated and analyzed outside of Creo. This happens in an open source tool called OpenVSP. The aero surfaces can be exported in STEP and IEGS. I prefer IEGS because I can put them in a text editor and read them to be sure of exactly what I'm getting. For example, for the fuselage, I have 8 IEGS entities of type 128, "Rational B-Spline Surface". They are arranged around the four quarters of the forward half and aft half of the teardrop-shaped fuselage.

When I import into Creo 10 with "default" import settings, things "seem" to work fine. But our aero designer is suspicious of the various "fixes", "reparameterization", and "simplifying" algorithms that Creo turns on by default. He would like know that the aero surfaces are coming into Creo exactly as they were generated and analyzed in our upstream tools.

As I turn off various options in the IEGS and STEP import dialog box, the model often has problems with importing, which would require intensive work in the IDD to fix. I don't think these STEP and IEGS files have problems; they import easily into other tools.

Does anyone have a document or tutorial that explains how to see and analyze the differences in a surface introduced by Creo's algorithms? For example, if we can prove to ourselves that the end result imported into Creo varies by less than the manufacturing tolerance, then of course the differences are immaterial. If the modifications Creo is doing have consequences on the aerodynamic performance of the finished surface, then we have got to turn them off, and deal with the consequences.

Thanks for ideas.

[Chris3]

There is someone here that probably has a better answer, but can OpenVSP export points on the surfaces? If so then you could quickly dump all of the points to a point table, export it back out and and verify that the points have been imported 1:1. For a sanity check you could also measure from the point locations to the surface to verify that the surface geometry is following those points.

[JA_10341562]

Yes, I like the idea of comparing the points.

The goal is to avoid rasterizing the mathematical shapes from the b=spline surface, so comparing points is not quite what I want. I guess what I'd want to do is compare the control points coming from OpenVSP's IEGS file (which I can write programs to parse) with the control points that end up in Creo. But it's unclear how to analyze differences I find there; is a small difference within manufacturing tolerances, or does it have bigger consequences further out down the spline?

[Chris3]

Why can't you have both? If OpenVSP won't let you export out the points without eliminating the mathematical splines and surfaces can you export out two models and then just overlay them on top of each other? The more points that you have the less concern you will have about deviations in areas where the point could isn't as dense.

[kdirth]

The best tip I have received is to match the export accuracy.  Start a new part and set the model accuracy to the accuracy the file was created at.  Then import the model setting model accuracy to external.  This helped me a lot with some surfaces sent from a customer in STEP format, I expect it will also make a difference with IGES.

[tbraxton · Accepted solution]

I know that in the compressor blade design space the use of .ibl files to generate curves are used to directly replicate the aero surfaces from the aero-thermodynamic design department. These curves are then used as construction features in Creo to build the surfaces using Creo core surface features (not Style surfaces). The input to Creo design department is a series of .ibl curve files to define the surfaces required to create the solid geometry.

You can use this approach as well for the outer skin surfaces of an aircraft.

How to import .ibl files:

https://www.ptc.com/en/support/article/cs107337