Pads that ignore grade limits...?

[DaveNorton]

Hi, 

 

I'm putting together some site models that are simple by comparison to the majority of site models I've worked on over the past few years, but I'm banging my head against a wall with the simplest ever thing here and need to ask for some help...

 

 

Model data is derived from EA LiDAR data, converted to 3d contours (QGIS). I do this process all the time for modelling and have never had this problem. 

 

I've rebuilt the models several times, validating 3d data (no problems)etc and reducing their size to make as workable as possible.

 

Essentially, I'm just trying to create a berm, with approximately 1:3 side slopes, sat above an existing terrain. On one side of the model (West), this has worked, with 1:3 side slopes to the grade limits. So that's good: it works sometimes!

 

On the other side (berm to the east) the same process delivers only a vertical edge to the pad, no side slopes, and no amount of offsetting of the grade modified/redrawing/re defining its options will change this. Please see the attached model. I've had this problem dog me for two days straight now, both with trenches, and berms. It's never a problem before - and I'm lost for solutions!

 

Any thoughts? Am I just being eye-wateringly stupid about something? It wouldn't be the first time...

 

Thanks for looking

 

Dave

 

 

 

Berm Dave Norton - Help.vwx

[Jeff Prince]

@DaveNorton I have an explanation and solution for this, but I would like to see if one of the experts here or from Vectorworks can get you sorted before I share it.

 

@Tom W. @Katarina Ollikainen

[Tom W.]

Patently not an expert but be glad to understand what's going on. I think the elevations are wrong. Is the terrain really 2m below OD? When you turn the skirt on you can see the model is 'upside down'. If I raise the DTM + lower the min elev it works. I also took the source data, simplified the polys, converted to 3D Loci + created new DTM in new file + everything works as it should there. 

[DaveNorton]

Tom and Adios amigos - thank you. Yes, the actual site is below sea level, in the East of England. Interesting. 

 

I will try the poly simplification, and conversion to 3D Loci (I haven't done that before). In this case Tom, what factor of simplification did you apply? And did you keep the elevations as per the original 3d contours (EA data?)

 

Thanks so much. Already feeling less insane with it!

[Tom W.]

2 hours ago, DaveNorton said:

Tom and Adios amigos - thank you. Yes, the actual site is below sea level, in the East of England. Interesting. 

 

I will try the poly simplification, and conversion to 3D Loci (I haven't done that before). In this case Tom, what factor of simplification did you apply? And did you keep the elevations as per the original 3d contours (EA data?)

 

Thanks so much. Already feeling less insane with it!

 

I'm going to look at it again... Perhaps the elevation thing was a red herring. (I am in East Anglia too but in the mountainous part compared to your elevations 🙂)

[Tom W.]

Ok forget what I said about the elevations, skirt, etc - that was rubbish. What I have found though is that if you make the Grade Limits around the east berm 1.1m bigger the pad develops sloping sides. Not sure why this would be... Hopefully @Adios amigos hasn't hit the road just yet + can enlighten us!

[Tom W.]

Ok + when I take the source data, simplify the polys (simp tolerance 1m) + create a new Site Model, the original size Grade Limits works (applies a slope to the east berm). So the issue is to do with the density of the source data relative to the size of the slope...???

[DaveNorton]

Thanks Tom

 

Yes, when I create a large offset to the grade limits, that works fine, creating continuous falls from pad to ground, as expected.

 

I rebuilt the model with simplified 3d polys, to 3d loci, and initially, the berm worked. Then I tried to join the two parts of the pad, and grade limits together to make one longer 'L' shape. That reverted to having vertical sides, and wouldn't go back to sloped sides. 

 

I am guessing that yes, it is about the density of the source data, and working with very small offsets to grade limits. This was never an issue in previous VW editions or service packs though (this anomaly has only happened, in my experience, since moving to SP 6, VW 2023 last week)....

 

Thanks again!

[Jeff Prince]

You guys are flirting with the answer and in the right neighborhood.  No need to rebuild the site model.

 

Rare problem, but easy to fix.

Essentially, your modifier does not have enough points along its perimeter for Vectorworks to triangulate the surface to your liking.

The fix is to add vertices to the modifier along its perimeter.

 

When you see weird triangulation like this:

 

 

It's a harbinger of doom when it comes to site models and usually means you need to add some vertices to assist the triangulator and/or revisit your site model and hunt for overlapping vertices on 3D polys.

Edited August 10, 2023 by Adios amigos

[Tom W.]

13 minutes ago, Adios amigos said:

You guys are flirting with the answer and in the right neighborhood.  No need to rebuild the site model.

 

Rare problem, but easy to fix.

Essentially, your modifier does not have enough points along its perimeter for Vectorworks to triangulate the surface to your liking.

The fix is to add vertices to the modifier along its perimeter.

 

When you see weird triangulation like this:

 

 

It's a harbinger of doom when it comes to site models and usually means you need to add some vertices to assist the triangulator and/or revisit your site model and hunt for overlapping vertices on 3D polys.

 

I don't think I've EVER heard anyone say this before + I'd never have thought of it myself so very good tip Jeff thank you. But makes perfect sense when you point it out. It's going in the manual... 🙂

[Jeff Prince]

2 minutes ago, Tom W. said:

 

I don't think I've EVER heard anyone say this before + I'd never have thought of it myself so very good tip Jeff thank you. But makes perfect sense when you point it out. It's going in the manual... 🙂

 

Let me know if you need someone to write the foreword for your manual 🙂

 

[DaveNorton]

Adios amigos & Tom you are both brilliant. Thanks for giving up your time to help, you genuinely both saved my week 🙂 I don't draw straight features like this generally, they tend to be more shaped....and so it makes perfect sense...

 

Have great weekends both!