Elevations on a curved 3D polyline

[Breten]

So I am working on laying out data for some survey crews along the centerline of the road.  I have a BC and EC elevation, but the vertices along the CL are at different elevations, which were the original elevation of the poly line before being transformed into a 3D polyline.  Is there an easy way to make the full line follow the slope from BC to EC other than manually shifting the vertices?

[TerryDotson]

If you don't find a better solution, MapWorks Base contains a 3dPoly Proportional Slant tool that lets you pick a starting/ending point and it will set all vertices between to the grade defined by those points.

[troma]

So you have a 3D polyline with many chords forming a curve, and the start and end of it are at the correct elevations, but the middle points need to be adjusted, is that it?
I would advise you to create a feature line from your polyline. It should inherit the same elevations. Then you can do an elevation edit and 'flatten' the grades between these two points, so that it all forms one straight grade.

[ksorsby]

The Grading tool 'Set Gradient/Slope Between Points' works with 3D polylines. Selecting the start and end points, the intermediate vertices are interpolated between those levels. You can also pick the start point and gradient and it will calculate levels based on that too.

Be careful with what you say is the start point though as line directions seem to play a big part as to whether it works correctly or not. Another undocumented Autodesk bug feature.

Kevin

[AllenJessup]

You've gotten some good suggestions. My choice would be to use a Featureline. However the question I have is; Is the grade between the BC and EC regular. If the horizontal curve falls fully or partially along a vertical curve. That will complicate matters some.

Usually by the time you're to the point of staking the road. You'd have an alignment defined, a design profile established and a corridor built. Then you'd be able to get your CL elevations from the corridor surface.

Allen Jessup

[jmayo-EE]

The OP has some good recommendations here but I think Allen hit an important point. If this file has a lot of these 3d plines and they are used in a surface it will have a large overhead and therefore a performance hit from thousands of chord vertices. Feature lines will allow the surface to be defined with fewer points and performance will be optimized. Now the fact that you can have real curves instead of chorded representation of those curves is another big advantage.

[ksorsby]

In addition I would say that using feature lines as curves in a surface is, in my experience anyway, not always great. The surface always triangulates to from the ends to the curve midpoint like a large chord, and I end up having to manually add in loads of intermediate level points on the feature line anyway to get it to triangulate as close to the curvature as possible, so you end up with the same problem. Multiply this process by 100's of feature lines....

Or is there a setting I've missed or a better way to get a surface to triangulate exactly to a curve?

Kevin

[jmayo-EE]

In the attachment the pool and wall model fine with curves or varing radii. We don't have issues with Fline curves in surfaces. Perhaps your mid-ordinate distance is too high for the curves? If I have small radii, like 2 units or less I use a mid ord of .01, if the radii is between 2 and say 20 I'll use .1, if the radii are 20 or more I go to a mid ord of 1. If we have radii much larger you can get away with 5 or 10.

[Neilw_05]

I think Allen's suggestion to use an alignment is a better solution. Most survey controlers can upload alignments which allows the surveyor to perform calculations along the curves, etc.

I would consider making the polyline into an alignment and profile.

[ksorsby]

John,

Yes, you're right. I'd never played with that variable, the default is 1m but it does work much better with lower numbers like you say.

Cheers,

Kevin

[jmayo-EE]

I would also use a corridor for all but the smallest proposed road projects.

[jmayo-EE]

HTH Kevin.

[troma]

---

@Anonymous wrote:

...

If I have small radii, like 2 units or less I use a mid ord of .01,

if the radii is between 2 and say 20 I'll use .1,

if the radii are 20 or more I go to a mid ord of 1.

If we have radii much larger you can get away with 5 or 10.

---

A little off topic, but wouldn't it be great if you could set this up ahead of time so you didn't have to mess with it?  Maybe if you could set it by degrees of radius turned, rather than mid-ordinate distance.  So you could set it at 15 degrees—you would end up with 24 points on a full circle, or 6 for every right-angle bend.  Maybe 20 or 30 degrees would be better?  But either way, you could set it once and leave it, as it would apply just as well to all curves.

[Breten]

Some great suggestions thanks guys.  The majority of my experience is with pipenetworks so I'm catching on.  Sadly as I am construction mnagement I am working from drawings provided by the client.  The drawings are stripped down so there's no civil 3D data, however the linework is all 3d.  I've hd o create my own alignments for the survey crews to stake out.

[Neilw_05]

It may not be a lot of work to create an alignment using the best fit tool.

Once you get the alignment you can project the 3D poly to the profile and trace it.