Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Swept: twisted and curved .... is causing distortions

Status
Not open for further replies.

kherman

Military
Jan 21, 2014
20
I am on NX 10.

I am trying to make a swept feature that can twist and is curved.

The problem:
The shape as it is swept is not maintaining a rectangular cross section perpendicular to the guide's tangent. My worse case so far is that the "cross sections" can be a trapezoid with corners at 135 degrees. This issue manifest most notably at the center of the sweep.

What I am doing
There are two sections that are both rectangular. The rectangles can have different shapes. One could be 2x20 and the other 4x6.

I have a guide that is curved. It's start and end points are the centers of each rectangle. The start vectors are normal to the rectangle face. (a long winded way of saying that it is a centerline).

Default settings for swept except for what follows:
Interpolation is Cubic.
Alignment set to pattern
orientation is fixed
scaling is area law
law type is cubic
start cross section area is the area of the start rectangle
end cross section area is the area of the end rectangle


Solution?
I should be able to define a spine that should resolve this. Spine is greyed out tough. I read that I need two guides to do that. So I added a straight line to the start and end of the guide and tried that. The guide selection was still greyed out.

Spine selection:
Lets you select a spine curve.

Use a spine curve to control the orientation of the section strings and to avoid distortions caused by the uneven distribution of the parameters on the guides. A spine string works best when it is normal to the section strings.

More solutions?
I can not fathom that this is a unique problem that I am having. There must be a way to handle this. I just can not find the missing detail that I need.

Note:
For various reasons, I can not upload a prt file. I can include screen shots if needed.

UPDATE 1:
An image showing what is going on.
 
Replies continue below

Recommended for you

Can you provide at least a picture of what it is that you're attempting to model?

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Digital Factory
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.
 
I am having some success by changing the Orientation Method::Orientation to Vector Direction and selectingthe Z-axis. I have to go through the process of testing the results ...... might get to that in a few hours.

I don't know why the Z-axis is works though. Choosing X or Y mangles the shape. If this is the fix, I'd like to better understand the rational as to why it works. I read the NX documentation for this setting and I don't understand it.
 
Could you provide the part itself as it's hard to see what sort a path the center-line follows?

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Digital Factory
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.
 
For reasons I can't get into, I can't upload the prt file.

The centerline is a studio spline. I create a studio spline with default settings then specify a start point and the normal of the start rectangle as the orientation. I do the same for the end point and end orientation. Then NX comes up with a solution to the edge curve that is "nice" according to the customer.
 
Changing orientation to vector axis of Z improved things.

Went from 45-135 degree corner angles to an 80-100 degree range. I have no idea why this is still happening. Tolerances?
 
I think techinically, what I need to do is have NX provide Uniform isoparametric curves with one guide
 
kherman said:
Default settings for swept except for what follows:
Interpolation is Cubic.
Alignment set to [highlight #FCE94F]pattern[/highlight]

Pattern? I'm guessing that you mean "parameter". Try turning on the "preserve shape" option, it should help to keep the corners square.

On second thought, "preserve shape" probably wont' help in your situation. It will help to keep the corners sharp (two faces meeting at an edge rather than a sharp corner approximated by a spline); however it may not help to keep the corners "square".

www.nxjournaling.com
 
"Preserve Shape" isn't doing it. It does seem add extra edges to the sweep though (unverified). Regardless, I was using preserve shape all along and agree that it is correct to have it selected.
 
PrintScaffold,

I appreciate the idea. I think at this point a different solution may be needed. I'll try to give it a try soon.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor