Trimble Business Center

I'm not getting a clean subgrade surface after applying site improvements and building takeoff surfaces. 

Please advise on what additional info is needed to resolve. 

Finished design:

FD with subgrades adjusted (one example of several) The slots are roughly 0.85' x 0.67':

Hi Richard,

where does the FSL surface comes from. Is it clean, no error flags? Does it have a vertical spike in that vicinity?

My FSL from the Designer had a spike between curb and pavement. I had to explode the surface first, delete that line work and re-triangulate.

Since that seems to be a curve in your example, do you use polylines or a curve to separate the upper and lower level of that step.

If it's a polyline make sure it's either absolutely identical or offset far enough from each other.

A few weeks back I was about to go mad with instructions "leaking" into each other, since the pavement type outlines from the designer weren't perfect. Tiny little overlaps caused havoc. TBC also didn't like near zero segments length within the line.

And if you drape the line work onto the FSL before adding it to it make sure it's far enough from vertical edges. Especially if it is a polyline it might go up and down along that edge. Site improvements don't like those near vertical segments either.

Examples from an early version of my problem model.

Hello Ronny thanks for your reply,

(I'm going to assume FSL = Finished Design)

I create the FG via linestrings.

Below are the same clips with the controlling linestrings and triangles turned on.

As you can see, no flags or spikes and no apparent reason as to why the vertical downward spikes would be created by the subgrade routine. 

Finished Design:

FD with subgrades adjusted:

And here's a shot of the flood in that area:

Richard

I would have to get the data to investigate - the challenges with these automated subgrades is where you have 2 different subgrade thicknesses that butt up against each other along a curved line that is chorded in the surface model and then we trim out those vertical walls that we used to have in the subgrade model - my guess is those spikes are where it lost track of the two depths somehow and pushed one of the subgrade down at those spots by the full depth not the partial depth - if we can get the specific data then we can likely track the issue and improve the algorithm so it happens less or not at all - since all data sets are different it may or may not be easy to reproduce the same issues that you are seeing - it could also be that the FG linework has some very small segments in the source lines that fail on our point tolerance somehow and that can also cause these types of issues to show up.

What are the two depths of subgrade being applied either side of that line - is one of them the 0.83' that you are seeing in the downspike?

Alan 

The HD paving is 1.25'. The LD paving is 0.67'. I actually got the dimensions flipped in the clip so the downspike = the LD SI. 

File shared offline.