Sweep Mystery

Next
 From:  TMeeks
9829.1 
I created a profile and a simple sweep rail. I then resized the sweep rail and only slightly changed its path for different sizes.

In some sizes the sweep is capped. With other sizes, the sweep is NOT capped. All sizes have the same number and type of control points. The size was altered by moving control points.

I have attached a small .3DM to demonstrate what we are seeing.

EDITED: 27 Apr 2023 by TMEEKS

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
9829.2 In reply to 9829.1 
Hi Tom, it looks like the sweep result has unjoined ("naked") edges inside it.

I believe this is caused by some segments being just barely not tangent to each other, like just a little more than 1 degree off from being tangent.

This makes the sweeper try to do mitered corners which involves extending surfaces and intersecting them. Intersection calculations can tend to be difficult on things that cross each other at a very shallow angle like that because they kind of have something more like a "zone of intersection" rather than a really well defined intersection.

The cure will probably be to tune up control points to make everything that should be tangent to be directly in line with each other.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
9829.3 In reply to 9829.1 
Hi Tom, try this attached version. This one has the sweep rail tuned up so that all segments are exactly tangent to each other.

Now it should not be trying to do any extensions and intersections for corner mitering.

Things that are really close to being tangent but just barely not tend to complicate intersections.

- Michael
Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  TMeeks
9829.4 In reply to 9829.2 
Based on your explanation I found and fixed the issue. The solution will be invaluable to us!

I had widened the Sweep Rail by pivoting on a Corner Point. Based on your help, I pivoted on the standard point above it and it worked.

THANKS!

I love this product.

By the way, Version 4 is a bit more forgiving than Versions 3!

I will download your fix. Thanks, again.

EDITED: 27 Apr 2023 by TMEEKS

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Phiro
9829.5 
If it could help, to explore this mystery, when you move a point which is not a sweet point, the sweep is capped.

The path is made with hard and soft points with lines and curves joined.
If you use a rebuild fuction, the problem is resolved.
If you use a ReconstructCurve, it's solved too.
If you use the script delcorners, it's solved too.

Michael, is it possible to view the different types of points in MOI ?
Soft, Hard and those are semisoft/semihard ?
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  TMeeks
9829.6 In reply to 9829.5 
"Michael, is it possible to view the different types of points in MOI ?
Soft, Hard and those are semisoft/semihard ?"

Great suggestion! I, too, would find this very helpful... as I did your accompanying suggestions. In fact, I will incorporate those ideas into our training to reduce the frustration of a sweep no acting as we expected.

Thank you so much!
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Phiro
9829.7 
For example, Inkscape do it with squares and diamonds
squares are soft points, diamonds are hard points.

Perhaps a script could change type of a point (or joint)
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  Michael Gibson
9829.8 In reply to 9829.5 
Hi Phiro,

> Michael, is it possible to view the different types of points in MOI ?
> Soft, Hard and those are semisoft/semihard ?

There isn't anything for that currently but it's something I would like to add in the future. It will probably be something that shows up in Join initially, like some kind of optional report after doing a Join that tells you if segment ends are tangent to each other or not.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All