tubular...

[PAULKORENKIEWIC]

guru's....

I'm trying to model two tubes that are a constant distance away. They travel in one direction then turn 90º and twist about before making another 90º and head off at 15º. They "sort of" look like this:


and from the side:


I'm being forced to model each leg individually and then there is that "overlap". I know I could change my sweep to a swept blend and add more "control" sections, but seems like there should be a way to do this more elegantly. Any thoughts?
thanks in advance...

Paul Korenkiewicz
FEV , Inc.
4554 Glenmeade
Auburn Hills, MI, 48326-1766


This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.

[rapidoconsultin]

Paul

Have you tried making two curve sketches with the second a discrete distance from the other. Next you should be able to select which curve you want your sweep to reference. I'm sure this is a series of sketches or series of points with the curve running through points.

Regards

Eric Mills

Design and Project manager

http://www.rapidoconsulting.com

eric.mills@rapidoconsulting.com

[CBenner]

<u></u>

[CBenner]

Ok.... no idea what just happened. My reply WAS:

Do you have Pro Piping? This can be done as a flexible tube in Non Spec Driven piping. We use it for things such as air hoses, etc.

Chris Benner
Designer II
Powell Fabrication & Manufacturing, Inc.
<u>www.powellfab.com</u>
P: 989-681-2158
F: 989-681-5013
Confidential
The information contained in this communication is confidential and may also be subject tolegal privilege(s). This information is only for the limited use of the individual to whom it is intended. If you are the intended recipientyour use is strictly limited to your knowledge and use of the information only and you may not forward, distribute, disseminate, copy or publish this information in writing or verbally to otherswithout the express written consent of the sender and/or creator of the information. If you are notthe intended recipient or the agent or employee responsible for delivering this e-mail to the intended recipient, youruse, dissemination, distribution, copying, forwarding, or publishing ofthis information, in anymannerwhatsoever, isstrictly prohibited. If you have received this e-mail in error, please destroy it immediately and notify the sender by telephone at (989) 681-2158 or by return e-mail. Thank you.
<u></u>

[PAULKORENKIEWIC]

Let me try to clarify things more...

There is a set trajectory for the "center" of the two pipes. It is a straight vertical section followed by radius curve to a straight horizontal section. The pipes come down the vertical section side-by-side. Thru the radius curve and a short section of the horizontal section they "twist" 90º ending up going out the rest of the horizontal section one on top of the other... Thru all of this the distance between the pipes is constant.

I managed to get a variable section sweep surface whose edges I thought I could use as trajectories for the pipes as they twist:


BUT when I put in swept blends, I don't have control over the normal direction so the ends don't blend correctly with the constant section pipes... hmmm....


I don't have pro/pipe.

Offset points won't do it because of the twist... or it would be fairly "cludgey" at best.

thanks...


Paul Korenkiewicz
FEV , Inc.
4554 Glenmeade
Auburn Hills, MI, 48326-1766

[BrentDrysdale]

Hi Paul,,
The swept surface ribbon and use edge is exactly the technique I would use. Without knowing the detail of your sweep I suspect the problem is that you are getting non tangent transitions at the problem points you show. If I am doing this sort of thing I pay particular attention to these transitions; often I end up adding small straight sections to enforce this. The other thing that might be a consideration is using a sweep control other than the default "Normal to Trajectory", an example is "Normal to Projection" which we use quite a bit.

Looks like you are very close so good luck.

Regards, Brent Drysdale
Regards, Brent Drysdale
Mechanical Designer
Tait Electronics Ltd (www.taitworld.com)
New Zealand
Ph. +64 3 358 1093

Korenkiewicz, Paul wrote:Let me try to clarify things more... There is a set trajectory for the "center" of the two pipes. It is a straight vertical section followed by radius curve to a straight horizontal section. The pipes come down the vertical section side-by-side. Thru the radius curve and a short section of the horizontal section they "twist" 90ºending up going out the rest of the horizontal section one on top of the other... Thru all of this the distance betw een the pipes is constant. I managed to get a variable section sweep surface whose edges I thought I could use as trajectories for the pipes as they twist: BUT when I put in swept blends, I don't have control over the normal direction so the ends don't blend correctly with the constant section pipes... hmmm.... I don't have pro/pipe. Offset points won't do it because of the twist... or it would be fairly "cludgey" at best. thanks... Paul Korenkiewicz
FEV , Inc.4554 Glenmeade
Auburn Hills, MI, 48326-1766

[gkbeer]

I think I'd try modeling a curve that represents a centroid between the
pipes, then use that as a basis for a variable section sweep to create a
ribbon whose edges represent the tube center lines. The item to vary would
be the angle about the centroid. That angle could be managed by a relation
driven by the trajpar and a graph feature.

Once you have that ribbon surface, use it's edges to drive the sweep of the
tubes. Hide the ribbon surface on a layer, or better yet, build the ribbon
feature in the assembly's skeleton, and copy geom the relevant geometery
into the tube parts.

[PAULKORENKIEWIC]

Thanks for the additional suggestions... they pretty much centered on what follows...

I wasn't able to eliminate the end issues with a VSS... though, I will admit to always being a bit confused over all the projection references...

I was able to put some points on the edge of my "ribbon", then create a new curve thru those point and set the tangencies at the ends appropriately. Using those curves for my sweep cleaned up the end issues BUT the distance between the curve that was supposed to be constant, wasn't and was a bit lower... so far the best "work around".

Another thought I have been trying to work out is controlling the "ribbon". It's just a VSS of a line with the angle defined in a relation as 90*trajpar. Trajpar goes from 0 to 1, so this relation sets the angle as a linear progression from 0 to 90. BUT what I really need is to force the ends to be tangent. So I thought I could use a sine function to do this. I got as far as converting trajpar to an angle from 0º to 180º, then the sine of that goes from 0 to 1 back down to 0. I thought I could use that multiplied by the linear slope and get just what I wanted. I have been playing around in excel trying to get it to work, but seem to be missing something... What I have is show below with the "RED" sketch what I really need... Any suggestions?


Paul Korenkiewicz
FEV , Inc.
4554 Glenmeade
Auburn Hills, MI, 48326-1766

[CM10]

Paul,shoot me an e-mail with your part.

F

In Reply to Paul Korenkiewicz:

Thanks for the additional suggestions... they pretty much centered on what follows...

I wasn't able to eliminate the end issues with a VSS... though, I will admit to always being a bit confused over all the projection references...

I was able to put some points on the edge of my "ribbon", then create a new curve thru those point and set the tangencies at the ends appropriately. Using those curves for my sweep cleaned up the end issues BUT the distance between the curve that was supposed to be constant, wasn't and was a bit lower... so far the best "work around".

Another thought I have been trying to work out is controlling the "ribbon". It's just a VSS of a line with the angle defined in a relation as 90*trajpar. Trajpar goes from 0 to 1, so this relation sets the angle as a linear progression from 0 to 90. BUT what I really need is to force the ends to be tangent. So I thought I could use a sine function to do this. I got as far as converting trajpar to an angle from 0º to 180º, then the sine of that goes from 0 to 1 back down to 0. I thought I could use that multiplied by the linear slope and get just what I wanted. I have been playing around in excel trying to get it to work, but seem to be missing something... What I have is show below with the "RED" sketch what I really need... Any suggestions?


Paul Korenkiewicz
FEV , Inc.
4554 Glenmeade
Auburn Hills, MI, 48326-1766

[PAULKORENKIEWIC]

Now this is cool... Tim uses evalgraph to just totally avoid the "math" issue with getting the ends tangent... Very nice...

Paul Korenkiewicz
FEV , Inc.
4554 Glenmeade
Auburn Hills, MI, 48326-1766