Pipe Dreams

[Bill Tillman]

I have this project which contains some 10" SCH 40 pipe (10 3/4" O.D.). The model was provided for me from a Revit model converted to DWG. I can make out the shape of the tube but need some help to recreate it. The tube is rolled to a varying radius and it's compound.

 

Okay, so I got the DWG file, the tube had been converted to a block inside of a block. No big deal. I'm needing to trace the center of this tube so I can extrude some other shapes along the same path. I'm drawing line from points on the circle that form the O.D. of the tube and passing them through the center, or at least that's what I'm trying to do. What I'm finding is that when I pick a point on the outside wall and draw a line from there which passes through the center of the tube and I end up with a line that is exactly 10 3/4" long. But I'm finding that when I move the end point to one of the other intersections along the wall of the tube I can also end up with 10 3/4" length. Which is not a good thing because that means the line doesn't pass through the center.

 

Here's a couple of questions I'm curious about

1. Is there a way to break up the block so that I can easily draw these lines which pass through the center?
2. What I'd really like is to get rid of the lines which make up the wall and just be left with the circle paths which make up the tube.

 

Any advice would be appreciated.

 

3DPipe.dwg

[lrm]

It appears that the tube lies on a plane so the first step is to create a UCS with an XY plane that passes through the center of the tube.  I used three points located at the center of the tube at its ends and approximate midspan.

 

I tried fitting an arc to the curve by finding the intersection of of the perpendicular bisectors for the lines connecting the end points to the midpoint but determined that the shape could not be approximated with an arc so decided to create a spline for the tube's centerline.

I like creating splines using Control Vertices  as it gives me better control.   I started by creating a spline with 7 CVs making sure it started and ended at the precise end points of the tube. I wasn't too careful about the location of the other 5 CVs as I know they would be moved.  It is important to keep in mind that the second CV and the next to the last CV control the tangent direction and radius of curvature of the spline at its beginning and end.  Once I had the 7 CVs and with all object snaps off I moved the 2nd CV relative to the end making sure to keep the tangent vector correct but observing how the radius of curvature was changng as it met the tube's curvature.  Once I was happy with the beginning and the end shape I moved on to moving CVs 3, 4 and 5.   The results are not perfect but should be sufficient.  It is best to use as few CVs as possible. Many people think more is better but more CVs (or fit points as well) can actually produce worse results as their is an increased likelihood  of getting ripples in the curve when using too many defining point (CV or Fit).

[SEANT]

One issue not immediately apparent is that the two ends of the Revit geometry (Pipe) are cut at an angle not square to the prescribed arc.  Hence, the inner and outer sections are elliptical, not circular. 
 
Using the sectional geometry near - but not at - the ends, circles could be recreated via 3PtUCS then 3PtCirc (Radii 5.40 and 5.03).  An additional application of that procedure near the pipe’s midpoint gives three circles to work with.  Like the above sequence, the three circle centers can be used with 3PtUCS then 3PtArc.
  
A circle can be swept along that arc, and the two ends can be SolidEdit – Offset to extend beyond the Revit geometry, and 3Pt Slice can match the odd angles.

3DPipe_2016.dwg

[Bill Tillman]

Thanks all. Yes, this one is indeed a tricky one. This 10 Sch 40 pipe trims out the end of a larger steel structure. The tube is basically rolled with several different radii and then tiled about 26°. It's made up of several pieces and once it's together it spans about 80 feet. Tomorrow morning we're making one of the trim to the trim pieces and will place it on one of the tubes to see how it fits.

 

I took the tube from the 3D model and isolated it to work with it. It wasn't too hard to figure out the tilt angle and once I had that I set the UCS so I would be looking orthographically at the thing. Then I carefully and painstakingly drew lines along what will be the top of the tube once it's in place. These lines I then joined and AutoCAD turned it into a 3D Polyline which presented problems. So I exploded it and then ran the FLATTEN command to get everything in the same plane. That allowed me to offset the lines which I then filleted together. Took a while and I gotta believe there is a better and more efficient way of doing this. I'll know tomorrow morning if it worked as planned when we cut a template on the burn table.

Edited May 18, 2022 by Bill Tillman

[SEANT]

9 hours ago, Bill Tillman said:

The tube is basically rolled with several different radii

Is that true?

[Bill Tillman]

SEANT, yes, even though it's not clearly visible. This total assembly is made up of 3 pieces which spans over 82 feet. And the radii change several times between the bottom and the centerline of it. It's symmetrical so that means I only need to get one side of it and then mirror it. The layout did work when I put it on top of the rolled tube. So we're proceeding with fabrication.

Edited May 26, 2022 by Bill Tillman

[SEANT]

Very good.  Sounds like you are off and running.  Post photos of the completed project if at all possible.