Problems With Cross Sections

[nigelro]

I am evaluating volumes of concrete removed from a surface and the volumes of concrete poured to replace the demolised concrete.

I have created surfaces for each stage of the process in individual drawings.

I have created a drawing for the section views and successfully created cross-sections showing all the surfaces.

When I compute materials, determining cut material between 2 of the surfaces, the end area is not always determined at every station.

Has anybody got an idea why this might be happening?

I have previously noticed end area "shapes" with breaks in them. Each time, this was because there was  a void in the surface triangulation.

Is there any way that I can interrogate a surface to check whether the triangulation is incomplete, with voids in it?

I posted last week and included the attached screen print showing incomplete triangulation which I only noticed by turning on the "Shades of Gray" visual style.

There must be some way to check surfaces without having to use the memory gobbling visual style.

I am running Civil 3D 2011

[mathewkol]

If you just want to fill in all of the holes, create a "Show" boundary around your entire surface.  This should work.

[nigelro]

Thanks Matt,

That does seem to help.

I am still unable to figure out why the cross-section areas are not evaluating properly.

Since posting, I have created a completely new set of sections with a new data set.

I am still having the same problems.

In one instance, the area was partially determined. I attach screen shots of a complete section and a partially evaluated section.

Can anybody think of anything that I should check?

[mathewkol]

Tough to say unless you attach your data set.

[machadt]

Hi,

what is the procedure that you are using to create this surfaces?

Thanks,

[nigelro]

I build the surfaces in the following way

1.) Import cogo points.

2.) Create new surface and add all points to surface definition.

3.) Create feature lines along the edge of the area covered by the points on the surface

4.) Convert the feature lines to breaklines

5.) Make the outside bounday into a "SHOW" boundary

I attach Part I of a sample data set here and Part II in my next post

Copy and paste the files to a folder "W3_TEST"

The 2 surfaces are W3S_DEMO250 and W3S_DEMO288

The cross-sections are in W3S_TEST_SECTIONS

The isections at stations:

0.40, 7.60, 9.20, 9.60, 10.00

have not computed correctly

[nigelro]

Part II of Data Set

You will have to create data references between this drawing and those in Part I

[andrewpuller3811]

We generally change the surface style to show the slopes.

Standard settings for our template are red, so we get red triangles where the triangles exist and black holes where they don't triangulate. Only takes a quick visual scan to see if surface fully triangulated.

Not automatic but is quick to implement and easy for people to understand.

[nigelro]

Tatiana

I am still having the problems with the cross-sections.

Earlier I attached a sample data set but I have received no further suggestions from the forum.

I think that it is now a matter for Product Support.

I really need to get this matter sorted out. The problem is occuring with every sample group that I work with.

Thanks

Nigel

[dnl999]

did you find the solution ?  If you did can you post the solution ?

I have the same problem.

[mathewkol]

Nigel, we are here to help as long as it doesn't take too much time out of our own day.  When you say things like "...

You will have to create data references between this drawing and those in Part I.." This means that we have to do extra work in order to BEGIN  helping.  Many times this kind of iunstruction turns us off.  It's better to give us a drawing we can ver yquickly open and look at instead of several tasks we need to accomplish before we can assist.

[nigelro]

Matt

I completely understand.

Unfortunately, I am unable to load all the files together due to size constraints and have had to split them up and hopefully give some guidance on how to get the sample working in the same way I have been using it. I do not see any other way to deal with the data.

Nigel

[machadt]

Nigel,

I being working on this.

Do you have to create sample lines every .2 mts?... I am just wandering...

Thank you for your patience.

Regards,

[nigelro]

Tatiana

Yes I do and sometimes even closer together.

[mathewkol · Accepted solution]

My previous comment notwithstanding, I had some time today.  I moved one of the surfaces up by 0.1mm and it fixed the gaps I could see.  Sometimes when your surfaces exactly meet, there are problems.  By having them overlap every so slightly often fixes these problems.

[nigelro]

Thanks Matt

I shall try that in my project and see what happens.

[nigelro]

Matt,

I have tried what you have suggested on one project which has 5 surfaces. 

I tried various permutations of shifts.

It did alter the outcome for a couple of the unevaluated sections but did not solve the problem.

It strikes me as a "duct tape" fix-up. There is obviously a problem with the alogrithm that Autodesk needs to address.

Quite frankly, I find it scary that that such a fundamental operation can  produce at random, spurious results,

Nigel

[mathewkol · Accepted solution]

I agree with your duct tape analogy.  This problem doesn't often appear for me bu I as well as many others have seen it before.  Currently, I con't know of any solid fix.

[machadt]

I was able to confirm this behavior, and I will log it with our Dev team.

But thank you Matt, it seems that he has a good workaround for it.

[nigelro]

Tatiana

I am still having problems with the cross-sections and am tired of trying to make patch up solutions.

Matt's resposes indicate that the problem is a software problem that has been around for a while. This really worries me as I work with cross-sections a lot and cannot afford to worry about weather the AutoCAD output is correct or not and have to check each and every cross-section each time I run a volume report.

I am suprised that neither you or your supervisors have asked me for a complete data set so that you can try and analyse the problem and hopefully find the reason for the errors.

FYI, my company is shortly to embark on a major highway construction project. My role among other things will to supervise the tracking of earthwork and grading quantities through the project. The way things are, I cannot see  that we can use Civil3D for this, considering that there is such a fundamental problem with it.

Meanwhile, I am manually editing the Autodesk xml volume reports to give me the correct output, after manually computing the cross-section areas that Civil3D has failed to automatically compute.

More expensive time wasted.