TIN model - time spent

[Openantics]

I am just beginning to create TIN models, and was curious how much time people spend actually perusing the triangles for anomalies or mistakes or even how efficient they were created?

 

Mine are not that big, and in a couple of hours i can look at each TIN for efficiency but is that a poor use of time?

 

I am not sure how well to trust the TIN model!

 

TIA,

 

Darren

[Organic]

What program are you using (I assume you are not just using basic AustoCad)?

[Openantics]

Civil 3D

 

Actually, i am learning what TIN modeling is in a different program but will migrate to Civil 3D shortly.

 

thanks

[Tyke]

If you have sufficient well placed break lines most of your TIN triangles will be pretty well formed.

 

After creating the TIN I change the visual style to see the surface and not particularly the TIN and then by using ORBIT I can quickly look over the surface and any anomalies are quickly found. The TIN can then be edited to give a better model. For a surface of around 15000 m² it takes typically 10 minutes.

[Openantics]

Thank you all for the replies: can someone offer links to a mini-study into the world of TIN models? I will be taking a software course for this topic shortly, but would like to do a little homework first and was hoping someone could offer any info on where i can learn more about the workings TINs, and how to edit, create, and work with.

 

always, TIA!

[eldon]

I would think that most folk here just use TINs without knowing the mathematics or philosophy behind them. I think that you would find more technical information about them with a search on the web.

[SLW210]

Try searching Triangulated Irregular Network. How much accuracy are you expecting?

[Openantics]

Hi SLW210 - sorry to respond so late as i was not emailed of new reply!!

 

accuracy is important in that the resulting contour maps look correct. For example, we survey a lot of small creeks, and if each TOB shot isn't properly connected it is glaringly obvious.

 

questions:

1. what does one do if bad data is found AFTER the tin model is created?

 

i spent much time realigning many triangles to connect similar featured topo shots resulting in a very good looking contour map. Only then to find a few clusters of shots with bad numbers which couldnt be 'fixed', and i needed to delete. i guess this is a software specific problem, although the small data loss did not affect the surface - I lost all the effort i made re-aligning the triangles when i had to re-run the model WO the bad points!!

 

TIA, Darren

[Organic]

Try searching Triangulated Irregular Network. How much accuracy are you expecting?

 

The mathematics behind surfaces and contouring is quite interesting. Most civil engineering uni courses teach this as part of their math components.

[eldon]

Speaking from personal experience, and using Land Desktop (from a few years ago), when the contours are formed and there are obvious glitches, I examine and correct the points and breaklines forming the ground model, and redo the TIN modelling process. It takes but a minute or two, and I really could not do with reforming the triangles (although there is a triangle editing feature).

 

Bad data can come from the surveying process, which must be edited/deleted if necessary. You cannot just lump all survey points into an automatic process, and come out with a perfect contour map every time.

 

Incidentally, when the contours are curve fit, if one uses the Fit option in Pedit, then sometimes the contour takes obviously wrong routes, especially where contours are at close intervals. I set the SplineType to 5 and Spline fit. I know that this is technically wrong in that it pulls the line away from the node points, but it does give a much more pleasing picture. Contours are only a way of representing the land features, and you really can't say that they are that precise!