Vertical Surface Comparsion?!? Is it even possible? You tell me..

[Mallen1987]

So I have two 3D scans of a retaining wall (one from 2013 and another from 2014) and I need to compare the two and see if it has moved. I am wondering if anyone in the world knows how to do so. Since the surfaces are vertical, and not horizontal, this has become a real 'head-scratcher' for me. I've tried to use the 'X' values as elevations, but since the retaining wall is not due east-west in the 'real-world' it isn't along the 'X' and 'Y' in AutoCAD, therefore when I tried to compare the surfaces, I get a gradual increased elevation difference from South-North due to the rotation from North (in the real-world) and the coordinates produced... SOMEONE PLEASE HELP, I CAN PAY!

[BIGAL]

Use rotate3d and turn the wall to be on the X Y plane make a surface then you can compare two surfaces.

[Tiger]

Isn't it enough to flip the USC so that the back of the walls are the X-Y-plane? Otherwise, Bigal's suggestion should work?

 

Are the walls exactly vertical btw? Didn't think that was possible... if they are not exactly vertical, wouldn't it work with a diff-surface anyway? It would be interesting to try it out if you can share the file with us.

[eldon]

Instead of comparing surfaces, I would take the conventional approach of drawing section lines, and then you can easily compare readings from both surfaces on the same section.

 

When I was monitoring a retaining wall, each section was marked, so that successive readings could be directly compared.

 

I am afraid that it is the way of modern surveying to do a blanket survey, and then try and pick out the meaningful results.

[eldon]

Here is one that I did earlier. You probably have all the right data, but you have to get the presentation of the facts in a readily understood format.

[BIGAL]

Ret walls are often a problem when surveying the top can overhang the bottom and when contouring goes back to front and sometimes requires a little nudge of the points to correct.

 

The problem I see is if you use something like CIv3d and the vertical surface wobbles in and out then the surface in the true vertical plane will be all over the place as the algorithm used to make the triangles is based around XY position. Like you say Eldon if the wall has a lean then you can probably do very easy one profile across base and lots of cross sections with a exagerated vertical scale..

[Mallen1987]

Thank you guys for the replies, I haven't had a chance to try your suggestions yet, though I will soon. As for you Tiger, I should be allowed to share it, but remember its scan data, so its big, I'm pretty sure the surface XML's are 25 mb. Let me know if that format is would suffice for you. Talk soon.

[Tiger]

It would be awesome to try it out, XML is fine for me. You can use any number of filesharing sites (dropbox, Sprend.com or GDrive are the ones I use, depending on the occasion) and send a link to thorssonc (a) gmail.com for me.