Sweep along a 3d polyline path without tilting

Sweep along a 3d polyline path without tilting

joel.moeller
Contributor Contributor
1,749 Views
8 Replies
Message 1 of 9

Sweep along a 3d polyline path without tilting

joel.moeller
Contributor
Contributor

I have a square profile that I'd like sweep along a 3D-polyline. It's important that the profile stays perpendicular to the XY-plane, but that's not happening:

joelmoeller_0-1736926342096.png

The top object is the 3D-polyline. The segments are tilted, with the 'sides' showing (particularly on the second and last segments).

 

For the bottom objects, I exploded the polyline to individual line segments and swept them. This result is more what I'm looking for, with a 'flat' top and bottom of the profile that only tilts with the Z-change of the line but not sideways. However, the disjointed corners are going to be hard working with.

 

I've tried to make sense of the Alignment and Twist (and Bank) options of Sweep but without success sofar.

0 Likes
1,750 Views
8 Replies
Replies (8)
Message 2 of 9

imadHabash
Mentor
Mentor

Hi,

You may need to try SWEEP command options ( Alignment, Twist ) that may help you.

 

Imad Habash

EESignature

0 Likes
Message 3 of 9

joel.moeller
Contributor
Contributor

I have experimented with both Alignment and Twist without getting what I want. I do want Alignment set to 'Yes' so the profile is perpendicular to the start of the sweep guide, though I might be able to work with aligning the profile my self. Twist I've tried to have Bank on, having Twist at 0 and tried various other numbers for twist but nothing appears to make what I want.

0 Likes
Message 4 of 9

leeminardi
Mentor
Mentor

Try the following:

  1. Create a block that is a rectangle 0.5 x 0.5 with its base point in the center.
  2. Use UCS ZA to define a UCS with its origin at the start vertex of the 3dpoly and its 2nd Z-axis point at the 2nd vertex.
  3. Explode the block.  Note, the distance command with show that the z coordinates of the lower two vertices of the polylinne square is zero (with UCS World active) indicating that the square is aligned to the world XY plane. 
  4. Copy the polyline to the 2nd vertext and use LOFT to make a solid.
  5. Repeat the process for successive segments of he 3dpoly.

The following compares the results from your model.

leeminardi_0-1736961714159.png

 

The next task is  how do you want to deal with the transition from one solid to the next?

 

leeminardi_1-1736962153249.png

You can union the two solids together and then, to fill the gap, define a series of 3DFACES and then convert the 3dfaces to regions which can be used by SURFSCULPT to create the solid that fills the gap.

leeminardi_3-1736962387797.pngleeminardi_4-1736962558297.png

 

 

lee.minardi
0 Likes
Message 5 of 9

Washingtonn
Collaborator
Collaborator

As mentioned by Lee, you will need to deal with the transitions. 

 

You could apply the cross section along the 3D Polyline and use the loft command.  This would require multiple placements  prior to and after each change in directions and possibly midpoints as well to minimize distortions and force the majority of the segments to be straight. See example below.

Washingtonn_0-1736975671368.png

Washingtonn_1-1736975681958.png

 

 

0 Likes
Message 6 of 9

leeminardi
Mentor
Mentor

 In addition to making multiple sections s washingto suggested, you could just reposition the sections away from the vertex then make copies of the section and loft.

For example, in the image below (starting from the bottom) we see the sections correctly positioned at teh 3dpoy vertices. The sections at the iteror vertices ar moved way from the vertex using osnap nearest. Copies of the interior sections are made then used to make solids with loft.

leeminardi_0-1736978745230.png

 

lee.minardi
0 Likes
Message 7 of 9

Washingtonn
Collaborator
Collaborator

Lee, the OP had specific requirements:

"...This result is more what I'm looking for, with a 'flat' top and bottom of the profile that only tilts with the Z-change of the line but not sideways. However, the disjointed corners are going to be hard working with."

 

Here is a plan view (showing only Z-changes) - I drew short lines from each of the vertices and then offset the lines to each side.  This gave me the bottom corners of the rectangles.  Vertical lines were drawn at each vertex which provided the third point for alignment and insured the sides stayed vertical. As you mentioned, the change in direction caused some issues (deformities) but adding more cross sections could smooth it more.

Washingtonn_0-1736983424713.png

 

0 Likes
Message 8 of 9

joel.moeller
Contributor
Contributor

This is intresting, thanks Washingtonn! I think I understand what you're doing but I'd appreciate if you could share your file.

0 Likes
Message 9 of 9

Washingtonn
Collaborator
Collaborator

Just follow the procedure outlined above.  I have already trashed the file.

0 Likes