For Car modelling you need often complex patches to create clean surfaces.

but.....



The Moi network runs only with 4 splines perfectly.... you get a clean patch (mesh) and structure.



In Inventor its possible to get a clean patch with more than 4 splines, it seem to be a kind of "N-Network"



first back to Moi it looks clean..........



...... the mesh is crazy, but it works.

Inventor seems to create a "basesurface" (a bit like flow in moi) in the right "formwave" and cut it automaticly to get a patch......is this possible in Moi.

Here are the Splines, maybe some has an workflow, how it is possible in moi

Hi zarkow,
Here's a way that might help to jump through that hoop.



I wasnt really happy in that I did the final network using a custom tolerance set very low to get a nice looking surface, but this could probably be avoided by creating the projection curves with more care to create a better flow for the network generation. You can tell in the end where I had to extend the wheelwell cutting curves out to get a good trim. Iif you can trim the wheelwell with the original curves, then you havnt deviated much from the intent. A little more care in that earlier part may get better results. But really, this is just doing what we see Michael repeat alot about making a larger surface to trim down. creating the surface in this way from the start would get even better results. Here I just worked with the curves you gave.

Thanks BurrMan,
that is a good idea, maybe for a script too.

A surface rebuild or anything like that moi needs .

I have seen a couple other apps with "extend surface".. That would be cool to have.

Thanks Burr !

Hi Zarkow,

> Inventor seems to create a "basesurface" (a bit like flow in moi) in the right "formwave"
> and cut it automaticly to get a patch......is this possible in Moi.

Currently MoI does not have a mechanism for fitting one big surface through an odd outline structure like that. In the future I do want to add a mode for the new N-sided command to be able to choose between the "star" method that it currently does or that single surface fit type method.

But it's difficult to implement a well behaved single surface fit, the fitting mechanisms that do this tend to be prone to making oscillations and lumps in the surface result, and it also tends to be difficult to get enough accuracy out of the fit so that the generated surface actually touches all the curves closely enough to be trimmed directly by the 3D curves.


Often times it's better to do what Burr shows and make an initially 4 sided more simple outline and then produce the final outline by trimming that, rather than trying to directly build one surface to an odd shaped boundary.

- Michael