Comparing cogo points to a surface

[Anonymous]

Hi,

My client has asked for data to be presented in a few different ways.

We are looking at erosion to embankments.

I have completed several aerial surveys and presented the data as a tin volume with cut/fill elevation banding.

Now they would like to look at the verification GPS survey points that we collected in a table form.

For my own purposes, I have been using a LISP function that I found on this forum, but it presents the vertical difference between point and surface.

The client would like to see the perpendicular difference between point and surface.

Given the steepness of the embankments, this makes sense.

Ideally, I would like to hand the data over as a spreadsheet, along with the drawing set.

Any tips on how to analyse the points like this?

I am using C3D2015

Regards

[neilyj666]

No sure there is an OOTB solution for this as all reporting is done vertically.

[LeafRiders]

If the embankment has a constant side slope there should be a mathematically way to get the perpendicular distiance in excel. I would need to dust off my Trig textbooks to break it down (not today). You would assume 90deg from embankment sideslope, calculate angle between vertical line (knowing that depth) and the side slope. I don't know a method in Civil 3D to get you what you're looking for. Good Luck.

[tcorey]

Considering you would need to analyze each point to each face on a face-by-face basis, a point could be perpendicular to any number of faces. What would you do in that case?

Tim

[autoMick]

Challenging!

How many survey points are there? What's the shape of the embankment? Concave, convex or even grade? Are the survey points taken as cross-sections or are they just random points anywhere on the embankment?

If the embankment is roughly in a straight line (not going around a steep river bend for instance), then an approach I have taken before is to extract the triangles from the original surface, group with the points and rotate the whole mass in 3D space so that the bank surface is basically flat. Now recreate a surface from the triangles and then you can use your LISP function to document the now vertical differences from your points.

This obviously doesn't give perpendicular offsets, but it will be closer in that it will reflect the average slope of the embankment. Clearly, all the points surface location will not be at real-world coordinates, but you will be able to get the measurements.

Assuming a more complex verison, I'm not sure of any way in Civil3D to do it but I remember using some tools in Sketchup that would allow you to select a face or group of faces, then select a point in 3D space, then the software would draw face-normal lines bewteen the faces and the point. Getting a surface and the survey points into sketchup is easy enough, but even once you have these face-normal lines, you would still have to get the length of each, find the minimum one and record it. There's no automated way I know of.

I'll be interested to see if there's some more elegant solutions out there!

Cheers

- Mick

[AllenJessup]

You might be able to do something with the NOR function in the Geometry Calculator. But it would take a bit of lisp to automate it. At one time I tried to use it to create a Normal offset of a Surface. But I didn't have enough time and skill.

nor(p1,p2,p3)

Determines the 3D unit normal vector to a plane defined by the three points p1, p2, and p3. The orientation of the normal vector is such that the given points go counterclockwise with respect to the normal.

The following illustrations show how normal vectors are calculated:

[Anonymous]

Please tell me how this can work:

If we assume that the slope under the point is constant through the point where it is perpendicular, can we use an expression to multiply the elevation difference of the two surfaces times the cosine of the inverse tangent of the slope?  That mathematically should be the perpendicular distance.

[Anonymous]

It appears expressions can only refer to one surface, and you can't combine expressions of two types to make this work, but you can end up with a label that gives you the elevation of the volume surface (difference between the two surfaces) and references the lower surface to give you decimal to multiply the elevation to get the perpendicular elevation you need.

Here is the reference text:

COS(ATAN({Surface Slope}))

Here is what I get on a fill around a 3:1 slope:

6.90 (fill)

0.94

So if I multiply the two together that is reasonably close to the perpendicular fill.

[Anonymous]

Sorry for taking so long to get back to this, I've done a lot of travelling in the last few days.

I have been reading the post replies via email.

The slope is 2:1 typically, however, some of the points are recorded above the top of slope, next to the rail itself, so simply using a trig function in excel isn't going to work entirely.

I have pointed out to the engineers that if the points are below the design surface, it is a fail regardless of if you are measuring perpendicular to the surface or not.

As there is no elegant way to do this, I will just resort to adding pages of cross sections with each report. Engineers love cross sections so it isn't a total loss.

Thanks for all the suggestions, I'm still learning about LISP, so I will be coming back to this to try again in the future.

Regards

[Anonymous]

I did not do what I did in excel, this was in "expressions" in the surface labels for slope.  I then created a label using the expression that gave the elevation of the volume surface (the cut or fill between the two surfaces) and the factor to multiply by to get the perpendicular to the surface immediately below the point.  This is as close to automated that I can think of doing it.  As someone mentioned above, each point in Civil 3D is perpendicular to more than one surface triangle, in fact there is a perpendicular distance to every single trinagle in a surface model.  Having Civil3D determine which triangle has the perpendicular line contained within itself sounds like a major undertaking.

In the attached file I created an existing surface, a field shot surface and a volume surface.  I have the expression and the label style with several points labeled.  Take a look and even if this doesn't wotk for Ben, somebody tell me it could be useful.