Matching Start and End of MeshSweep Section Curve

Hello,

I created an interpolate curve through random points. This curve is a period curve and is used as the rail for Mesh Sweep.

I want to fix a minor problem where once the section curve sweeps around the rail and comes back to where it started it does not match with its initial state; its rotated as it finishes the rail.

I want to make it so that the whole form is a continous loop; in other words one does not understand where it started from or ended at. Yet now the disconnection makes it obvious.

I would really appreciate if someone can help me on that.

Files are attached,

Kind Regards,

Levent

Attachments:

Replies to This Discussion

Permalink Reply by Joseph Oster on August 15, 2015 at 5:49am

There is no trick required, as far as I can see... But I don't have the 'MeshSweep/Mesh Pipe' plugin.

It looks like you went to considerable effort to create multiple cross-sections on 'pFrames'? But you only need one, unless you want to change the shape along the way?

Permalink Reply by Levent Ozruh on August 15, 2015 at 9:11am

The Mesh Sweep gives a result of higher resolution so preferably I would stick to that one. The multiplication was for something else, it seems I forgot to change it but as I do what you say the result remains the same...

Permalink Reply by Joseph Oster on August 15, 2015 at 9:18am

"higher resolution"? That doesn't sound likely to me, since NURBs have no resolution until you bake them. The GH preview may be lo-res but bake it and it should be perfect.

Permalink Reply by Levent Ozruh on August 15, 2015 at 9:27am

The baked NURBs happen to have unsmooth stuff like the one I attached.

Permalink Reply by Joseph Oster on August 15, 2015 at 9:41am

That's because the radius of the curve is too small at that point for the size of the cross-section; a completely different issue than mis-alignment of sections at beginning and end of the sweep, as depicted in your original post.

Permalink Reply by Levent Ozruh on August 15, 2015 at 9:55am

Do you know any way to solve that without going to much manually? Since I will try this grasshopper sketch with different seeds for the random component in the very beginning, a general solution could be of better use.

Thanks,

Permalink Reply by Joseph Oster on August 15, 2015 at 10:25am

Well... The first thing that occurs to me is scaling the section based on the radius of the curve. Hook up the following to see how it changes as you move the 't' slider (note that the curve is "reparameterized" to work with t=0..1):

You could divide the curve (the more divisions, the more accurate), get the radius at each point, and then scale the section down if the radius is less than "N". Might take a little experimentation with parameters to get the right domain, but based on what I see looking at the radius output, I would start with something like: