Polyline weirdness

[neilyj666]

I have received a dwg (derived from another modelling system I suspect) that has polylines that show elevations of  e.g. 623258 in Quick Properties but when I hover over a vertex the hotgrip turns red and the elevation shows as the expected value (elevation correct.jpg).

LISTing the polyline also confirms the elevations are as expected so why does quick properties show the elevation so incorrectly??

[AllenJessup]

I don't see anything right off. What does the model look like in 3D? What elevations do you get if you list the polyline? Can you post a drawing with one or two of the plines?

Allen

[neilyj666]

It's the green line on Layer verges - looks fine when viewed in 3D as does the full drawing. LIST shows elevations of the correct order of magnitude

[Jeff_M]

LIST shows that the polyline is not in the world UCS: 

Extrusion direction relative to UCS:
X= -0.4858 Y= 0.8741 Z= 0.0000

Also, inspect each vertex. many are at z=78.8500, the first two & last two are not.

[AllenJessup]

What Jeff said! Workaround - Explode line, use AeccCreateFeatureLines to convert all segments to Featurelines, use AeccJoinFeatures to join all segments in to one FL. There is at least one other line that isn't planar to the WCS. You can select them using QSelect and an elevation range.

HTH

Allen

[Jeff_M]

No need to explode/convert/join. Use the Convert 2d polyline to 3d command, then back to 2d. The resultant LWPoly will have the elevation of the original first vertex.

[AllenJessup · Accepted solution]

---

@Jeff_M wrote:

No need to explode/convert/join. Use the Convert 2d polyline to 3d command, then back to 2d. The resultant LWPoly will have the elevation of the original first vertex.

---

Glad you noticed that Jeff. But I don't think that "back to 2d" is appropriate if the elevations are needed. I tried it on another non planar line in the drawing which was visibly stepped in 3d. Converting it to a 3d polyline gave good results with the vertices at the original elevations. Converting it back to 2d gave a flat polyline that only matches the first vertex. But using 2d to 3d is much better than what I came up with.

Allen

[Jeff_M]

---

@AllenJessup wrote:

---

Converting it back to 2d gave a flat polyline that only matches the first vertex. 

Allen

---

Since the original was a LWPolyline I made the assumption that it was supposed to be at one elevation.

[BrianHailey]

<edit>

Wow, I feel like a loser taking so long to create this post and seeing all the posts that have been made since I started working on it.

</edit>

First thing I noticed is it didn't have the typical grips a polyline would have (mid grips are square instead of rectangles). Listing the polyline, it lists as a LWPOLYLINE but the vertices have different elevations and that shouldn't happen: http://screencast.com/t/DHh6UJFFIaB

Try converting the polyline to a 3D polyline and then convert it back to a 2D polyline.

http://screencast.com/t/jIa3zgk4fhhe

[gunnarl]

I've noticed the elevation difference when listing vs. properties box in polylines extracted from a Civl 3d surface before. I would try the workaround stated above, convert to 2d polyline command. If i remember right that is how i fixed it before. I would think the explode and re-join should work as well. Weeding the polylines may help too. I haven't tested any of this in your drawing, i'm away from Autocad at the moment, but i'm pretty sure one of these workarounds will work.

[neilyj666]

Thanks for all the replies guys.

The drawing was supplied after a request to the client to supply a 3D drawing for quantification purposes

To put the previously supplied drawing extract into context, I attach a 3d orbit view of the dwg as I received it - the coloured lines look to be at the correct level (site is 1200m x 800m), the mesh effect is section lines at level 0 and the red dots are objects at level -999. I just isolated all lines with elevation > 1.0m and all 3Dpolys and built a surface which looked ok.

I need to maintain the level so the convert to 3D polyline seemed to do the trick but what effect (if any) would the "...extrusion direction relative to UCS.." that Jeff identified have on the integrity of the model?? It doesn't look like it matters when I check the surface.

I have had the "...extrusion direction..." problem before when trying to use lineworkshrinkwrap around what were ostensibly simple 2D polylines. I think this problem derives from the conversion proces from whatever system the drawing was produced in originally to the dwg file.

[AllenJessup]

Once you've converted those polylines to 3D polylines, you shouldn't have any problem. The extrusion data on the 2D polyline means that they are not in a plane that is parallel to the WCS. The conversion removes that component.

I can't tell how the lines you're working with got that way. But you could duplicate that by using the Align command and aligning a 2D polyline to 3D points.

Allen

[neilyj666]

---

@AllenJessup wrote:

Once you've converted those polylines to 3D polylines, you shouldn't have any problem. The extrusion data on the 2D polyline means that they are not in a plane that is parallel to the WCS. The conversion removes that component.

---

---

Thanks Allen - that was the answer I was hoping for.

---

I can't tell how the lines you're working with got that way. But you could duplicate that by using the Align command and aligning a 2D polyline to 3D points.

---

I suspect that this extrusion was introduced when the lines were export from their originating software package (I have had this extrusion before when trying to join polylines)