Fit walls to objects - walls wrongly fit to ground plane

[lupin]

Using fit walls to objects, some (but not all) of my walls are fitting down to the ground plane (z=0) instead of up to my roof layer objects.

I'm drawing up existing conditions from a land survey, so my model is around 35m above the ground plane.

There are no objects (loci etc) on the roof layer that I can find at z = 0.

Sorry not to post a screenshot, but I'm not sure how to upload one to the techboard.

Any help much appreciated!

[lupin]

actually they're fitting to points near z=0 (e.g. z=-278.576mm) But I'm still not sure why. There's nothing there to fit to.

[Jim Smith]

Is your roof in a separate layer?

[bcd]

It can be very useful to create a 'boundary' Design Layer where you place a single copy of your roof &/or lower boundary. This way stray framing, loci, etc will not hamper the Fit command.

[lupin]

thanks for asking Jim, and sorry for the repeat post.

yes my roof is in separate layer.

My site model is also in a separate layer.

It seems to matter which order I do "fit walls to objects".

If I do the top of the wall first, and then the bottom separately, it's more reliable.

Also, strangely, it seems to work better if I make a more complex extrude (for a fancy parapet wall) than a simple one. I can then edit the extrude to simplify it and redo the "fit walls to objects" to get the correct parapet wall shape.

I'm wondering if VW doesn't like that my site model layer is set to z=0, whereas my building model layers are more like z=37000 and z=40000.

(I'm working in mm).

Finding these work arounds is time consuming - I can't work out the logic of them.

[lupin]

Thanks bcd for that tip. This model is pretty simple so far, I don't think there's stray geometry in my roof layer, but that method would be a good way to go for complex models.

[bcd]

Give it a go, you might be surprised that it works.