When topo'ing a curb island like the one pictured, where the PC and PT points are difficult to pinpoint, one should err towards the curve, lest one map a rather phallic looking form. That is, if you can't quite discern where the straight ends and the curve begins, cheat to the curve side.
I swear there's something goofy with the way cad draws these because I've been mapping an island like this and it seems very, very clear where the PC and PT are, but cad draws it like this anyway. But yeah I eventually figured it out the hard way that going a few tenths into the curve seems to solve this issue.
LisCAD has a collection routine for mapping tangent curves. You shoot 2 points on the tangent in, a point on the curve and 2 points on the tangent out. The software then calculates the PC & PT. It works great. I wish Leica would incorporate it into Captivate, the have slowly but surely been adding other LisCAD routines.
@john-putnam That'll make a pretty picture in cad, but I'd bet money the PC and PT of the linework won't match the PC and PT on the ground.
That sounds great for planimetric mapping but not that great for surfacing where the grades are not uniform. Also, there is a manner of coding F2f in C3d that draws just as you describe as well. Cases where I am mapping for planimetrics alone are so rare that I just always assume that I'm surfacing and proceed accordingly.
LisCAD calculates a 3D position for the PC/PT so if your shots are close to the actual points you still get a good DTM. When I was still using LisCAD, I would still take my shots at what I thought was the PC/PT, this routine just forced them to be tangent. It is the rare occasion that I'm providing existing condition mapping for my engineering clients so I'm guessing 98% of my mapping is 3D with surfaces.
I'm guessing it is better than you think, unless you are using a 6 foot straight edge to figure out where the PC/PT actually is. Guessing in the field usually gets me within a several of tenths of the calculated points. The resulting DTM is going to be more accurate than the obviously incorrect bulb return. If I'm that worried about things then it is time to break out the scanner.
I was out a couple weeks ago wrapping up a boundary and was locating some gravel roads and driveways. I didn’t have my glasses on and was not paying attention and had something like this pop up but I had started another curve in wrong direction so it looked like a big bubble under neath the gravel cup deep sac where I showed the driveway heading to the old house. When I did my field to finish in office I was well that ain’t right. A lot of those like you have shown here curb islands have like a .45 ft flat tangent piece not always at the nose either it’s where the desires to break the water they can be a bigger bear.
Speaking of surveying kerbs
Did you know you can use horizontal and vertical offset control codes in Access to generate the whole structure from one measurement.
For example, measuring from the back of a kerb
BC H 0.1 will generate a second line to represent the front of the kerb
BC H 0.1 H 0.11 V -0.1 gives the back, the top and the gutter line
BC H 0.1 H 0.11 V -0.1 H 0.25 V -0.05 gives the whole thing including a lip of channel.
It does help that you can save the whole code string in your library, bring it up and adjust on site to suit the actual numbers.
I also find it useful to have separate code strings saved for mountable kerbs and for vehicle crossings
(I work in metric, sorry if the numbers above make no sense - just adjust to suit your location)
Did you know you can use horizontal and vertical offset control codes in Access
This is also possible in C3d's F2f, as it was in LDT's. It would work great if you had a mass of uniform curb and gutter (or anything similar) to tie. In practice there are always driveway and wheelchair letdowns that would require constant modifications to the complicated coding. So opportunities to use this manner of coding are too limited to allow for me or my people to become proficient, IMO.
I use 3 codes one for gutter lip, gutter flow line, and tfc. I use bc and ec with civil3d and usually by taking points for the ramp and other pieces I get a pretty good curve. Never used POC function part of it. I have done reverse curves using that way as well.
@jimcox Yeah, those offset commands are nice in theory and I've used them occasionally, but I usually don't trust the feature I'm mapping to be uniformly built or to have settled evenly.
In my area we usually shoot the lip, flow line, and top back of curb. In cad we'll offset the top back of curb 0.5' towards the flow line so the drawing looks nice, but that line won't be used in the surface. To date I've never had anyone ask what the face of curb elevation is.
We use the POC code, see example below. It makes nice curves and if you have your figures constructed in 3d, you can get the elevation along any point (including the non-shot PC)
Here is another option with more complicate reverse curbs. For that one you use a begin curve action code, PC for ours, a the beginning of the first curve then take a shot with no action code at the PRC of the curves then take a shot coded “PT” for end curve at the end of the second curve with as many points as needed along the curve but this reverse curve fit doesn’t use the “POC” action code for point on curve.
@wa-id-surveyor I didn't run into reverse curves too often but when I did I remember trying to do something like:
EA PC
EA
EA PT PC
EA
EA PT
The software didn't recognize what I was trying to do by stacking up the control codes like that but putting a 0.10' straight segment at the PRC seemed to work. Looks like the software is smart enough to detect a reverse curve on its own; I assumed it would try to shoehorn a curve through all of the points based on the first point after the PC being left or right of it.
The F2f in C3d will draw that reversing curve very nicely without any mention of the PRC point at all.
@john-putnam in carlson, use the "afit" field code and it does the same thing. Derives a fitted curve based on incoming and outgoing tangents
@bstrand the prc point should be EA PT EA PC. Then it will draw correctly
@va-ls-2867 I thought I tried doing that too and it didn't work, but I can't remember for sure. Adding begins and ends to each string may have worked though, but I don't remember that either. In any case if a simple EA shot there will make it draw right then that's probably the quickest and simplest way to go.