Help with creation of volume boundaries

[eh8MUHT]

I'm exporting a geological model (which would be a boundary of a volumetric unit) and trying to generate the surface of that model on Civi3D. The boundary is represented by 3D faces.

Steps taken: 

1 - open the file on Civ3D (.dxf file, also tried with landxml (.xml))
2 - create a surface
3 - go on the definition, drawing objects. select object type as  3D faces and select the entire model. After the generated surface is done, I use the alignment tools to generate a profile view, but the surface is kinda spiky because we have different volumes (the geological unit is not single unit, but different boundaries of the same material) it generates only one surface, not the boundary.

   

Is there a way to generate the boundary of that volumetric unit, what we would expect would be as below:

Where rather than the profile view "bouncing" from each node of the top and bottom we rather see a true slice.

See the attached file to play with!

 

[camden.n.dean]

Maybe it's because I am not a geologist, but I'm having a hard time understanding your question.

It seems like you want the envelope of the surface profile. I don't understand the physical meaning of this, but it might be a good idea to use Best Fit Profiles. If you want an upper and lower envelope, then you'll need to get points at each vertex of the profile; then, fit a Best Fit Profile to the high ones and a second Best Fit Profile to the low ones.

At some point though, it might be just as quick to draw it by hand.

[ferodriguesMUDP7]

Hey @camden.n.dean, thanks for your answer.

The problem itself is a step before the profile creation. We have a 3D faces representing the volume(s) of material and we want to generate a boundary (surface?) of it to see them in profile view. The way of generating surfaces described at the beginning is not enough to do it (so it shouldn't be a surface?) because Civil3D is bouncing between the top and bottom when we do a profile view, since it generates only one surface. Is there a way to represent the enclosed volume with that 3D faces so we can see it in profile view / do volume calculations?

[camden.n.dean]

Again, I'm not sure that I totally understand so feel free to continue to correct me.

You have 3D faces that are defining your surface. Your surface is representing a stratigraphic boundary, say, bottom of sand layer. I assume you have at least one more surface to define another stratigraphic boundary, say top of sand layer.

If this is the case, then to find the volume you should use a volume surface comparing those two surfaces.

You also might be interested in the geotech module for C3D. More info on it here and here.

Let me know if this is helpful, or if I am way off base.

[autoMick]

If I understand correctly, you have 3D geological features and you wish to determine the volumes of the separate strata. Civil3d can't actually deal with 3D surfaces, which is ironic given its name. So instead of using the XYZ coordinates for the whole 'bubble' containing the material you want the volume for, you need to separate the upper and lower surfaces, - which is what a Civil3d volume surface will be able to compare to do its calculations. So typically you would have a surface for:

- (0) ground surface

- (1) top of material 1

- (2) top of material 2,

- (3) top of material 3, etc.

Then if you have a volume surface which compares (1) with (2) you will have the volume of material 1, etc.

Am I interpreting your issue correctly?

[fcernst]

Wrong workflow..

You need the boring hole data .. not DXF.. and take that to the Geotechnical Modeler .. the GM takes the boring hole data in several formats..

[autoMick]

I'm assuming the faces in the DXF include both the upper and lower bounds?

I had a quick look a sub-set of your DXF and whilst there's some possibility of determining the 3d envelope, there isn't a clear separation between the top and bottom, so any manual categorisation would be problematic at best.

What I think you need to do is to go back to the source data and export the interfaces between the different materials as separate DXFs (or XYZ coordinates), so you can then create unique surfaces to do the comparisons as explained above.

If you post an example of separated data, then the volume calculation will be easy to demonstrate.

[eh8MUHT]

Cheers Mick, sounds like you are understanding what we are after. Just to help, here's a rough diagram of what we're describing. In the top half, is what we have, units of material isovolume surfaces. The example file uploaded would be one of those materials. What we're trying to do is to is use an alignment and then a profile view to see what intersects the alignment (Second half of image), but what we get is the first image in my original post, where the profile view is "bouncing" between the top and bottom surfaces of the material volumes.

We of course would also like to do a volume calculation.

I'll try and get you some example files of the surfaces of each unit to see what you can do that we are missing, thanks for the help!

[autoMick]

Yes the concept is clear now. I misunderstood in that I was thinking you were looking at a open cut mine or similar. If you are looking just for a road/rail cut through a hill then that might make life easier.

The geotechnical module for Civil3d can help if you have borehole data (i.e. lat/long point on the ground surface and depth of each material boundary below that point) - as it will create the surfaces by interpolating material boundaries between boreholes. If you are dealing with seismic or other data then it gets trickier. 

From what I could see from your drawing you have randomly distributed points, so the challenge is to work out what is actually the top of the material or the bottom (just because a point is higher than a neighbour doesn't mean its automatically the top). If that is the case then I think some sort of gridding approach is probably the best

- it would help if you provided some raw ascii data. 

In the snip below I took some of your dxf and gridded min and max values, created 2 surfaces corresponding to those values and then it's easy to convert to a profile, or calculate volumes. I added a random third lower boundary just for demonstration.

You can cut a design surface (i.e. your road) through these and extract volumes etc. or just visualise the materials exposed in plan view. 

[eh8MUHT]

Mick, I've attached a small area of 4 different materials as xyz points. I wonder if this is something you could demo how to work with? You'll see clearly in these the issue you described, some XY locations have a clear top and bottom point, but others have multiple points.

Side note: We wouldn't / won't be going via the geotechnical modeller workflow.. as these are results of a different geotechnical modelling workflow.

Looking forward to your reply!

[autoMick]

I had a go with the data using the 1st 3 materials as an example. Method I used was to grid the data by 5 unit (feet? metres?) cells - once extracting the minimum and again to get the maximum. Then I created 6 surfaces

0_min

0_max

1_min

1_max

2_min

2_max

from those extracted points (I've attached an example min and max file in case its any use). Then just plotted the profiles.

As you can see from the snip, the strata seem to overlap unacceptably. It is probably to do with the grid cell size, but who knows? Obviously the larger the grid, the smoother it will be but the less resolution and more error you will have. The other issue is that there's no way of knowing whether there's numerous contacts within the same material body or if you're intersecting multiple lens of a material embedded with another material. Maybe this accounts for all the overlaps?

You need to have some method to reliably know what is the top of each mass of one material and what is the bottom.

[ferodriguesMUDP7]

hey @autoMick , could you tell me how did you get those min and max values to generate the top and bottom surfaces?