Extrude Along Path & DTM

[Frank O'Kobrick]

Is there any way to use the NURBS curve created by the Send To Surface feature as the path for a Extrude Along Path object? I'm trying to make a more realistic curb & gutter on a 3d model using a DTM model created from a survey.

Any other ideas to achieve the same effect would also be appreciated. I have client that may need a lot of this kind of thing for some photo realistic site models.

[michaelk]

All EAPs use a NURBS curve as the path. Even if you create the EAP with 2D geometry as the path the EAP command converts the path to a NURBS curve.

In 2010 (and I'm assuming 2009) drawing a poly line and running the Send to Surface command generates a 3D poly. That won't work for an EAP path. So you will need to convert it to a NURBS curve first.

But it should work. Depending on the length of the path and your site model you may want to knock down the number of vertices in the NURBS curve to keep the redraw time down.

hth

michaelk

[Frank O'Kobrick]

That's what I thought but it does not seem to work very well. Maybe my paths are to long or complicated. I might try breaking them up & connecting them back after the extrude.

Thanks

[GWS]

I find complicated paths when using the EAP command often cause a 'failure'. Splitting the path is a real pain but can often help.

I was hoping parasolids would help this, and it has, but only a bit.

[michaelk]

I think the workflow that might work for you is this:

Draw the polyline over the DTM

Send to surface

Ungroup.

Convert to NURBS.

This will give you a nurbs curve that could be the path of an EAP. But it could have so many vertices that the EAP fails. I'm not sure if this is because it's too complicated or it's because the vertices are so close together that the EAP "folds into itself".

I just tried it on a simple DTM and got 2500 vertices. 50 is probably enough. 100 is certainly enough.

So select the NURBS curve. Model > 3D Power Pack > Rebuild NURBS... experiment w/ the number of points that will work for you. See if 50 works.

Now the EAP is much more likely to not fail and produce the solid you're looking for.

hth

michaelk

[Pat Stanford]

I don't think they fail often due to too many point, but rather due self-intersecting surfaces.

By default the profile extrudes along its geometric center. If the path has sharp curves, it is likely that the profile will interesect with itself on the inside of the curves.

If you can draw the path and profile so they will extrude along the "outside" (ie the path runs along the smallest part of the extrude), you can often get things to work. But you will have to manually edit the Profile to move the point you want to run along the extrude to the 0,0 point of the edit window.

[michaelk]

I think you're right about the self-intersecting surfaces problem.

In my experimenting, reducing the number of vertices helps. Maybe fewer vertices = fewer folds?

[Assemblage]

Just a minor comment: Pat's comment about editing the profile relationship to its locus is very useful. I've been working for years on the assumption that VW only EAP through the profile's geometric centre, which often has meant awkward manual calculations for the path dimensions to achieve a desired overall size for the extruded object. There's no way of knowing any different from the tool/help notes etc - why don't they mention this??