DTM inaccurate

[islandmon]

Welcome to DTM v11.5 ; )

Apparently some of these niggling issues were worked out in v12.

The shifting of contours from your origin is a known 'bug' ( although the symbol contours are accurate to the symbol origin ).

Regardless, DTM is both an art & science. The best success results when the Existing DTM grid of 3D loci is equal to the XY grid of the contour separations, eg, 10x10 , and the Z values fall within the declared elevation ranges.

[Lynda]

Thanks for your reply. Not sure whether to be glad it's not just something I'm doing wrong or infuriated 'cos it's created more work!

I don't totally undestand your refernce to grid separations - I wasn't using a grid knowingly but my z elevations are definately within the declared ranges.

Can't ugrade an educational version to 12 unfortunately and can't afford to buy fresh as I'm having to do ACAD too. be intersting to find out how the opposition software works out when I finally get around to learning it!

Thanks anyway. lynda

[Thom]

Actually I find this is still a problem in v12. I have forwarded a copy of examples to tech support. I usually work from 3d poly's. It seems that the dtm would rather take the easy route vs. following the contour. Road crowns completely disappear. Ditches make for weird lines. Swales are all over the place. Hopefully they figure out away around this. I would rather have better accuracy than speed.

[ 01-04-2006, 05:01 PM: Message edited by: tvetter ]

[Lynda]

Yes, I'm getting the same problem with ditches and swales too. Both on the existing DTM and any new ones ce=reated for the proposed model. Makes a bit of a mockery of cut and fill calcs.

Infuriating and looks like it could be back to the drawing board!

[Robert Anderson]

Lynda, I'd like to see your data set. Please send it to randerson@nemetschek.net

[Lynda]

I have created a DTM from 3D loci. These match the original contours exactly.

When I create an existing and proposed DTm some of the contours created no longer match the existing positions. e.g. my 54m contour has been consistently moved east by 19m.

Other contours have some curves squared off (quite drastically). This is without any modification.

When I update my proposed model with site modifiers this is exagerated in some areas. Outside the fenced areas contours are altered when there is nothing happening there. The more I do the stranger some of the contours get. I have tried using elements in a new file from scratch but have just achieved the same result of crossing contours. One area looks like a childs scribble! There are also some very strange V shapes appearing.

Am I doing something wrong with the fences? How are you supposed to set the elevation of these - by sending them to the surface, choosing a height (which is difficult on a slope), or leave it at 0 regardless of the pad or road height as it is a 2d object?

I was doing well until the site became more complicated yet I can't even seem to backtrack the proposed model by eliminating items. Ironic as I'd almost finished the remodelling!

Help gratefully accepted.

11.5 on Windows 2000.

Lynda

[islandmon]

Lynda, the 'Grid' is the XY Grid of the page [ Page>Set Grid... ] which can be set to the grid of the major or minor contour spacing. Placing your 3d Loci ( when possible ) at the X-Y intersections creates a less random and a more uniform contour plan.