Tying Feature Lines to Corridor

[d_reno]

Here is the deal. I have to grade the infield, turfed area, between a proposed airport taxiway and an existing runway. The taxiway is being designed by others and we have to match into it. I have taken the other designer's data and created a corridor. So far, so good. At various spots I drew feature lines that go from a feature line representing the existing edge of the existing runway to the dynamic taxiway, snapping to the top layer of the corridor. I have additional feature lines snapped to the first set of feature lines. 

I want to create a surface bounded by the runway and the taxiway. Since I can't control the taxiway design, I have to adapt my infield design to it. So if the other firm comes back and tells me the taxiway has been moved, I want all of my feature lines to move when I update the corridor.

Since I do not know how to add a corridor to a surface, I created a dynamically linked feature line along the edge of the taxiway where I have to tie in. As a test to see which objects would react to which, I changed the profile of the corridor after having done all of this. The feature lines I drew from the existing runway edge to the corridor by snapping to it remain totally unchanged. The dynamically linked to the corridor feature line stayed with the revised corridor, but sent a bunch of spikes down to the unchanged fetaure lines I drew sanpped to the corridor.

I want this to all update automatically - the corridor controls, the feture lines that are tied to the corridor should follow it, and the feature lines that are tied to those feature lines should follow them after they've followed the corridor.

Any tips on how to achieve this?

thanks.

[dibr0804]

Im not real sure, but I think the only way you could get feature lines to move along with a corridor/profile is to use assemblies that tell it to do that.

[d_reno]

That sounds like it would be a feature line that runs parallel with the roadway, right? these features lines run across the space in between two separate roadways. so when I raise/lower one of them I would want the end of the FL to raise/lower, with the other end still pinned to roadway no. 2.

[dibr0804]

So would there be a way to create this assembly and then target the surface of the corridor (roadway 2). I am not very good with assemblies, I like 3d polys and points.

[jmayo-EE]

If you have two corridors, extract the feature line you want to target from one corridor and the other corridor can target the feature line.

[d_reno]

Yeah, John, that should work. Kind of. I have a bunch of grading to do in between the two roads, but if I create regions I can do it that way. Do you know of an easy way to incorporate a corridor into a surface? Or must they be separate?

[jmayo-EE]

Typically one would paste a corridor surface into another surface. Then add feature lines, gradings and/or more surfaces.

[d_reno]

Thanks. You'd think there'd be a way to tie an entire site with more than more corridor and independant grading objects and feature lines into one site-wide surface. But I guess not. I am not a huge fan of pasting just because of the way it "fudges" where the two surfaces meet. If you've ever done DTM modeling for GPS dozers, it pretty much has to be exact, everywhere, or it gets built wrong.

Thanks for your input!

[Joe-Bouza]

Fudges? I thought it replaces, no? At least that's what help says. Plus are they building a surface or your design? I would think you'd want to give the dozer design lines and grades

[Joe-Bouza]

Wouldn't pasting the two corridor surfaces and adding a show boundary between them be dynamic? (devils advocate here)

[jmayo-EE]

Yes but I was just helping with the target question.

Waitin' for a steak dinner... Allen said he's is buying. 🙂

Oh and Happy PI Day Eve. Warmest wishes to all of course.

[Joe-Bouza]

Oh sure, I was laying out more Ideas for the OP.

Make mine medium. Fort Lee?, Bronix?

[d_reno]

From what I understand when C3D pastes a surface, and how LDT did in the past, is that it goes out to the first whole 3dFace outside the boundary of the pasted surface when deciding what to keep of the original surface. So if your 3dfaces are about 50 feet on all sides, it might interpolate 45 of those feet when yuou paste something else in. It depends on the surface to be pasted into, and the outer boundary of the surface you are psting in.

Because of this, I again don't like to paste unless I have complete control over how it interpolates between the two surfaces - that is I know how it will interpolate because I left it no margin to anything but what I want. If I changed a corridor I would presumably have to go in and look at this area where the interpolation takes place to be sure it was right - like if a daylight line got kicked out 10 feet because I raised the CL profile or something.

When I provide this stuff for usde with dozers, it has to be an exacty model of what is being built, curb & gutter, inlets graded right, intersections, supers, anything you can possibliyt think of. Or they will build it wrong. They don't stop to check the model - they assume what I give them is right. If it's off 20 feet they may notice it, but not if it's off a half a foot. The machine is doing the thinking, not the operator. And on a site - like a school I modeled that had three different kinds of curb, five different kind pf playing fields, different sidewalks and pavement structures, all of it - you may have any number of different cuts from finish to subgade, so I *always* give them finish and let them figure it out. Anything made of concrete or bit too for that matter, has to be right on to 0.01 of a foot - exactly the same as if you were staking it convetionally. Except that when staking I can stake a curve with three points - PC, mid, PT. a GPS dozer needs the entire curve, reduced to segments so that with enough segments to define an arc, or it will build it wrong. Then my phone starts ringing and my day is ruined. This goes for vertical curves as well. about the only time I relax is if it is dirt, then nearest tenth is OK, because who cares as long as it drains properly.

Sn you can see that one surface would be ideal. I did a mile long airport taxiway and apron, with ponds, storm sewer & ditches, connecting taxiways, parking lots, fuleing stations, etc. Out C3D guys here modeled it to spit out plan/pro and XS sheets (the construction documents) using about 15 or 20 centerline alignments, and it was up to me to create a model to build it from. It worked GREAT, it is amazing what the stuff can do when it is used right. but that moedl probably took me two starights weeks *after* the corridors and everything were modeled. The only thing we staked on the whole job were things like storm sewer and light pole bases.

[d_reno]

sorry for the typos, I am going as fast as I can since I am supposed to be working!

[Joe-Bouza]

---

<snip> The machine is doing the thinking, not the operator. And on a site - like a school I modeled that had three different kinds of curb, five different kind pf playing fields, <snip>

I'm sorry, but that sounds like a real problem, to me.

So these machines (pardon my ignorance) they're like 3d-printers grinding through the site and a finished product comes out the back end. All the while NO ONE is checking things out? The IDR's must be fascinating to read. Could you post a copy of one?

[d_reno]

It's exactly like a 3d printer. these printers weigh 80,000 pounds is all. each corner of the blade is good to sub-centimeter accuracies.

and yes, there's usually someobody checking. our DOT specs call for two checks where they (gently) drop the blade of the dozer onto a mark so they know the system is working. and then there are a certain amount of randomly generated checks up and down the grade that are also specced.

but this is for real. and surveyors have kind of inherited the making of these models since we stake jobs and know about building and think in three dimensions. and the model you give them absolutely has to be right or they are going to come after you to pay for moving dirt twice.

I will see if I can post the airport job I mentioned. the school with all that crazy stuff going on was built about ten years ago, so I no longer have access to those files (my first model was a hospital campus in 1999). the end result out of Cvil 3D, though, is a 3dface representation of what is being built. you *can* use alignments and templates, but we all know that there aren't many roads that don't have intersections and turn lanes and variable ditchs, etc.

[Joe-Bouza]

I'd like to see one of those in action.

So the surveyor does the model for the GPD dozer - so thats the stakeout, right?

[sboon]

I'm trying to figure out exactly what you're trying to do here.  You have a known boundary at the edge of the existing taxiway, and a second known - but changing boundary at the edge of the designed taxiway.  Are you trying to create a straight grade from one to the other, or some sort of swale in the infield space - either one of these could be done easily enough with a corridor.

If you want to design something with more freeform shapes then you could use a grading group.  You already have a featureline along the existing edge and you can extract another one from the corridor.  If you draw two more featurelines from the corners of the corridor to the existing edge, and create a closed area then you can add an infill object.  This will give you a flat surface to start from.  Any featurelines or grading objects that you add within the space will add more data to the grading group surface until you get what you want.  If the corridor changes then you will probably break the connections at the outside corners and the infill object will disappear but it's easy enough to reconnect them and add another infill.  The surface will automatically regenerate.

A sketch showing a typical section, or somthing to show what this surface might look like would be useful.

[neilyj666]

We use GPS units mounted on dozers and graders and have got excellent results.

The main problem is that models are generally supplied to the finished ground levels whereas we need to do the bulk earthworks to the formation levels (which is a PITA to do in C3D - Autodesk please create a tool to accomplish this easily - other software manages it...!!!). This is the time consuming part but when its done, it eliminates the need for most of the "peg bashing"

[jmayo-EE]

IMO i would not let a paste operation transition or interpolate any grading ever. Every time I paste surfaces there is a common 'paste' line that has the same x,y,z, values on both surfaces at every vertex along any edge where they meet. I design and define the transitions. In a site plan sense the border of my FG lies entirely on the EG. With this I never see pasting errors.

I use multiple surfaces to build a composite FG. I have some projects with up to 10 base surfaces for basins, swales, demolition areas, project pahses, offsite work, etc. When the plans are put together I am working with one compostite surface.