Importing 3D dwg terrain, getting 3D lines only

[Ignacio Azpiazu]

A dwg import of a huge terrain model as a library part is resulting in a 3D script with a zillion LIN_ commands --all 3D lines, no faces.

Does that sound like a typical result which you avoid using some typical procedure or toggle switch that I am missing?

[Not sure if the original dwg does have the faces but I would assume it does. In a worst case scenario it does seem that scripting a MASS out of the zillion coordinates would be pretty straightforward and fast.]

[Eduardo Rolon]

Usually DWG terrain models are a gazillion of lines since the command that is used is 3dface which is not a solid element (the points collected on site are connected by the software using 3dlines). Try either openning the site in SketchUp and do a dwg export from there or get a dxf or 3ds out from autocad. All this takes in consideration that the dwg model was not exploded in AutoCAD which has the result of "erasing" all the 3dfaces and leaving only lines floating in space.

[Anonymous]

Ignacio

I don't know if your model is just a terrain model, or if it has other 3d data (buildings, walls).

IMHE, it is never a good idea to import a terrain calculated on another software.
That other software did some nasty things to the surveyor data (like creating totaly fictitious 3d data), and probably you have some guy also fiddling around with point coordinates that just where not taken correctly from site.

Best you can do is try to detect where that surveyor data is on your file (usually on a separate hidden layer), or, if not present, ask for it from your client.

Then just import it via ArchiTerra, and make your own 3d model.

[Ignacio Azpiazu]

Thanks Eduardo and Krippahl. Both very good points. I imported the 3D data into Excel with the idea of trimming it to the point xyz coordinates, eliminating duplicates and using that to create a MASS but I had 464,000 LIN_ commands. For bringing it into Excel I need to chop it off into 7 parts, because Excel will only take some 64 K rows per sheet. And then most of that data must not be actual survey data, so my whole approach doesn't make sense. I'll look around for the survey.