Graded region tips/tricks?

[patricks]

Anybody have any tips for getting the topo lines of a graded region to do what I want? It's not so bad when the contour lines are relatively smooth, but when I have to start creating swales (v-shapes in the topo lines for water drainage) or when the topo line comes to a 6" curb (topo lines jogs down halfway towards the next topo line, then continues on the other side of the curb), it just seems VERY VERY difficult to get these things to behave.

For example, I place spot points in a v-shaped or U-shaped pattern to create a swale, but instead of connecting the points with the topo line, it keeps wanting to just run a straight line across the top of the U or V, and not connect the rest of the points. Or when trying to show a topo line jogging down along a curb, I will end up with 3 topo segments coming out of a single spot point. So frustrating! It makes me just want to print out the site plan w/ existing (demolished) topo and hand-draw the new lines. :roll:

[patricks]

arghhhh I really can't stand how toposurface creation behaves... It's almost as if I have to place my points where I want the topo lines to be, and then FORCE the lines to connect the points by placing extra points that are like an inch higher or lower, on both sides of where I want the line to go! :banghead:

[Merlin]

...Wow....and actually changing points doesn't change the contours?

[iru69]

The Revit topographic tools are extremely frustrating to use because the developers appear to have invented their own completely illogical way of generating topo lines (similar to the problem of splines in Revit). This needs to be fixed.

[Justin Marchiel]

i agree. it would also be nice to have a tool that allowed you to add points based on a surveyors relative points. too often i receive a survey drawing that has a reference point that they picked as 100.00m. now this is always different then the 100.00m (or 100'-00") that i set my main floor to be.

it would be nice if somewhere you could set your floor plan to be a geodetic elevations and then set topo points based on the geodetic elevation. i have to take the points they give me, convert them based on my building floor height and then insert. too much work for a tool that doesn't work great in the first place.

Did i miss something, or how are other peopl handling this?

Justin

[Tobie]

I know lots of people are going to respond, object to my suggestion, but it works for our small scale projects. I use split surface. I do suggest using detail lines to define the area you are going to split. When you split the topo it gives you points on the edge. I use these on both new sections to align the heights. The amount of cut and fill does not behave correctly with this method. It will however give you the edge of the batter on elevation and site plan. You might have to do the swale in 2 sections.

[patricks]

Let me illustrate...

in the first image you can see 3 topo lines to the right and above the property line, and their points. Each contour is supposed to run up to the left, then cut back to the right to create the swale, and then when it gets to the curb it's supposed to run back to the left halfway to the adjacent contour and then continue up across the driveway, but you can see how the line cuts across (straight up) instead of following the points in the curve near the bottom.

In the second pic you can see the heavy line at the right which is elevation 100' and how it does the same thing at the bottom. Then move over to the left to the 104' contour, and it's a real mess. Not only does it follow the points in the curved part near the bottom, you can also see all the extra points I had to stick in the grass and the parking lot, just to make it follow the other points along the curb. Same thing up near the top to the right of the parking spaces, where it follows the curving curb. But then you can see where Revit feels the need to connect lines all the way across the freakin' parking lot for no apparent reason! So I'll have to go and stick another point in between contours to tell it NO you're not supposed to do that!

It would be nice if it could follow the rules of topography drawings, i.e. countour lines can never start and stop inside the surface, topo lines can never cross, and can never split into 2 or more lines.

[Merlin]

mmmm....I see what you mean.....as we say in Oz......."oooo! Bugger!"
John Mc

[gordolake]

Hi everyone,

I have similar issues wit h my topo a simple DWG import of polylines with elevation and it appears to me that revit will not interpolate correctly. It doesn't matter if I add or remove points revit will not draw the intermediate contours as expected.

Regards

Steve.

[Tobie]

I think Revit will "understand" Topo a lot better if you could specify the ridge and valley lines.

[patricks]

Hi everyone,

I have similar issues wit h my topo a simple DWG import of polylines with elevation and it appears to me that revit will not interpolate correctly. It doesn't matter if I add or remove points revit will not draw the intermediate contours as expected.

Regards

Steve.

In your situation it looks like you might need more contour lines in the CAD file.

I have found that if I import a DWG that has polylines for every foot of topo, Revit seems to make the surface just fine (the topo lines on the Revit surface follow the DWG import pretty well). But trying to make the topos go where I want them to is just beyond difficult.

[Merlin]

I think Revit will "understand" Topo a lot better if you could specify the ridge and valley lines.

hmmm.......Yes....there are times when it is crucial we have spot levels and contours in agreement.....maybe a "Wishlist" request would be to have a spot level "override" where we determine some levels are crucial at those points then overides/adapts the contours to suit/incorporate those points...this could be VERY useful...any thoughts?

John Mc

[Mike Hardy-Brown]

I have adjusted my site settings, it helps a bit, but they never are spot on either. Play around in your intervals. I usually set mine to 500(mm). It gets as close as anything to the dwg.

[Gadget Man]

hmmm.......Yes....there are times when it is crucial we have spot levels and contours in agreement... They are times???? :shock: Gosh, I thought that especially the Spot Levels should be followed to the milimetre! After all, they are the ONLY relatively accurate points in whole contour survey, aren't they? Everything else is just much less accurate approximation based on these points! So, while the actual contour lines usually don't pass through these points (obviously), they (Spot Levels) should take an absolute precedence in calculations of the contour lines! So if a user inserts a Spot Level it should force-alter contours accordingly - without simplifying them!

Alternatively, let us choose the level of simplification, according to our current job's needs! (Percentage scale would do the trick nicely... using, for example, already existing "Simplify Surface" function).

Yes, I have experienced this strange and frustrating behaviour many times too, but with so many other things in REVIT not working as they should, this was just a drop in the ocean (well, all right, maybe just a bucket... ;) )

[beegee]

>> they (Spot Levels) should take an absolute precedence in calculations of the contour lines! So if a user inserts a Spot Level it should force-alter contours accordingly - without simplifying them!
>>)I suspect it isn’t that simple.



My understanding of the problem is that the Triangulated Irregular Network ( TIN ) calculation must first determine which points to connect as triangles. Consequently, some implementations of TIN represent the surface in each triangle using a mathematical function chosen to ensure that slope changes continuously, not abruptly, at the edges of the triangle

There are many algorithms that perform this and at least 3 major ones –



Fowler and Little.





VIP

Drop Heuristic.

Having chosen one of the available algorithms, there are several ways to connect the vertices into triangles.



"Fat" triangles with angles close to 60 degrees are preferred since this ensures that any point on the surface is as close as possible to a vertex

This is important because the surface representation is likely most accurate at the vertices

The two main methods are




Distance Ordering and Delaunay Triangulation.




Delaunay is more common. 3 points form a Delaunay triangle if and only if the circle which passes through them contains no other point


When the TIN yields a flat triangle ( one having no defined aspect since its points occur on the same contour ) , methods must be employed to determine its slope by comparison with other triangular faces.

In a nutshell , there are a lot of complex calculations being done by Revit as you add topo points to a surface.


I’m just surprised that the contours bear any relation at all to that calculated by the Surveyors software, given that it is likely using an entirely different mathematical model.

[blads]

Makes a lot of sense when you put into perspective like that, thanks BG...

[Gadget Man]

I suspect it isn’t that simple...
Thank you BeeGee for your in-depth explanation. Wow!

However, I do not intend to write a program to do it - I just want to use one, which will reproduce (model) the most accurate real life situation. And I still say, that Spot Levels are the ONLY really solid pieces of information taken in the field, that are easy to input in the computer, so, regardless how difficult it is to do it, they should be followed faithfully.

Because if they aren't and the computer representation is only vaguely similar to real life don't bother at all and still do it in 2D...

[beegee]

Spot levels are the basic data that is used to generate the topo and contours, but the results may vary from what you might expect when you place a point, due to the factors I mentioned.
Changing the algorithm to force that point to occur on a contour may well have other unexpected and unwanted results for the remaining topology.
To some extent, it depends on the density of data that you are dealing with, as well as the continuity of the slope. Dedicated surveying software can deal with these sorts of situations, but I don't think we can expect to see those programmes incorporated into Revit.

[iru69]

Sorry beegee... maybe I'm missing your point - all the algorithm stuff goes right over my head - but I'm not looking for anything super fancy. It's pretty straight forward that when you place a spot elevation point that match the contour settings, e.g. the example Patricks posted, the contour lines should follow. Unfortunately, they too often do not, and therefor Revit *needs* to be fixed.

Maybe there needs to be a way to indicate which points are contour points and which points are extraneous spot elevations.

[beegee]

My point is that its anything but straightforward. It depends which points are chosen by the algorithm and then how the slope calculations are done between faces. I once did write a very simple programme to do these sorts of calculations, ( much more basic than what is in Revit :) ), but I'm not an expert in this field. I just know enough to point out that its not a simple fix.

It would be good if the topo could be divorced from the point data, once it was generated, and the user could then adjust contour lines, as intelligent objects, to regenerate the surface.


Sorry beegee... maybe I'm missing your point - all the algorithm stuff goes right over my head - but I'm not looking for anything super fancy. It's pretty straight forward that when you place a spot elevation point that match the contour settings, e.g. the example Patricks posted, the contour lines should follow. Unfortunately, they too often do not, and therefor Revit *needs* to be fixed.

Maybe there needs to be a way to indicate which points are contour points and which points are extraneous spot elevations.