I have an analytical process that generates points which describe a surface.
This data is well ordered, ijk, xyz
Simplest format for me was STL.
All analylitical point data explicitly preserved.
I have now tried two ways to process this in NX so that I can get a set of sheets that can be integrated in to downstream design process.
Method 1, use the rapid surfacing tool.
When this works, it works great, but in cases where the STL data encompasses higher curvature, it becomes impossible to keep the edge reasonably congruent with the initial STL points, and as adjacent STL zones are no longer sewable.
Modified my analytcal tool to also outpout the edge points used for each STL file.
Read these in as seperate groups of points
Made each a fit curve.
Now the four fit curves can be used with fill surface to create a sheet.
But... the "Fit to Facet Body" does not let me pick a my original STL convergent body.
Does that STL data need to be converted to some other facet format?
Any help is appreciated
So I answered my own question regarding the fill surface, fit to facet.
I changed the convergent facet surface to an NX facet surface and the tool works
Poorly, but it works.
So far my sheet surface is as far away as .4 inches from my initial STL points.
This is not very good.
I will play with this method a bit and see if I can get it closer.
I think it needs a patch count control on the filled surface.
Any thoughts on how to turn an STL into a sheet?
Think of this as a wing, I have say 250 points around the wing at 75 slices.
I need the large number of slices to capture change along the wing, and the large number of points to capture some changes in local curvature (its not realy a wing)
The through pts tool only allows a 24 by 24 matrix of points.
Not big enough.
I could read in 250 sets of points 75 times and make splines.
And then re-read in the sets of points in the cross direction as 75 points 250 times
And then use those to create a surface.
But thats a lot of work, and the scripting process seems like a lot of work.
I've only used the through points command a few times (on small numbers of points), so I don't know all the limitations. The help file makes it sound like 25 is the maximum, even for a multiple patch surface; however, I don't see why this would be a limitation. If you create a degree 3 spline through points, you can specify as many points as you like. The resulting spline may have many knot points (or patches if we extend this to a B-surface), but there is no inherent limitation to the number of points you can specify. I'm curious if this is an actual limitation of the command or just poor wording in the help files.
I tried the "through points" surface tool on a file that contains 105 rows of data with 40+ points per row, set to create a degree 3 surface. It worked; the input dat file can be found here if you want to try it yourself. The thread that the link leads to also contains a journal file to help automate the creation of the surface. In my test, I didn't use the journal; I just started the through points command and gave it the .dat file.