Last week at Autodesk University I was part of a discussion with some of the senior people involved in the development of Civil 3D. The topic was railway alignments, and how this might be added to the software. One of the questions they asked was about railway curves and railway spiral definitions. They are already aware of chord based curves and clothoid spirals but not the lesser known spiral definitions used in railway design. I know that some of the people on this site need to recreate old rail alignments and have a lot of knowledge on this topic.
The focus of Civil 3D is design, but it seems to me that the ability to recreate old alignments with various types of spiral would be a function useful for surveyors and designers. If anyone can post links similar to the one below then you might be able to help develop a tool that will be useful for the surveying community.
It would be interesting to see how they deal with the approximation built into the railroad formulas when fractional-station chords occur.
As I think I understand it, a railroad curve with all 100-ft station chords is the same as the highway curve of the same radius and delta, but just described by a different degree parameter and stationing "length". There are no approximations to cause discrepancies at whatever precision you care to work.
The fun begins when you have a fractional-station chord on the end, particularly a 50-ft one on the end you start the layout from. Then to computer precision it isn't a tangent circular curve any more. The layout from one end won't match a layout from the other end to computer precision. If you change the station of the PC to xx+00 then you get a slightly different layout than if it was xx+50. The difference on the ground wasn't too significant for a gentle curve, but would be easily measurable for a sharp one, and would be quite obvious at computer precision.
I think my approach would be to convert the railroad description to a highway description internally, label it in whichever style is chosen, and ignore the layout problems. If people want to see the effects of layout on it, defining the software functions gets messy.
I also grabbed the C3D Product Manager at AU and talked to him about Parcels with Spirals, so we can create Parcels that abut highway or railway curves.
It might be if you used 5730' for the radius of a 1° curve for each, but Highway Degree of Curve and Railroad Degree of Curve have different Radii Definitions.
I have always believed that you cannot create a parallel spiraled ROW that is truly parallel to a C/L spiral. That being the reason C/L spirals are used with an offset distance to the ROW when describing a line along the ROW line that is parallel to a spiraled C/L. Yes, my software will come close as most will, but is it something we should be using, or should we stick with C/L spirals and offsets when describing these types of ROW lines?
jud
My point of view is that a circular curve has one radius and delta. You can describe it by different definitions of "degree" and "length". It's not the radius that's different, it's the kind of "degree of curvature"
Making the railroad degree equal to the highway degree results in two different curves, but there is no reason to match those numbers. Fahrenheit and Celsius - equal numbers describe different temperatures, but there is only one temperature in the room so use the different numbers. Apples and Oranges.
Here is my version of the formulas (MS Word)
http://dl.dropbox.com/u/25124076/CurveFormulas.doc
In our little Port Authority subway system we use Allen Spirals;
His book was published just before we started tunneling and there may be a link between the Chief Engineer, Assistant Engineer and Allen.
Do you have an example of one of these? Are the ROW spirals geometrically "true" or are they approximated, similar to the way that Civil3D handles offset alignments through curves?
> I have always believed that you cannot create a parallel spiraled ROW that is truly parallel to a C/L spiral. That being the reason C/L spirals are used with an offset distance to the ROW when describing a line along the ROW line that is parallel to a spiraled C/L. Yes, my software will come close as most will, but is it something we should be using, or should we stick with C/L spirals and offsets when describing these types of ROW lines?
> jud
True, let's not forget that spirals are engineering curves. They are designed for engineering purposes, that is transitioning the curve. But it has no use in Right-of-Way. I had an old boss who started surveying in 1926 and was one of the most respected boundary surveyors in the state and he was respected for his math skills (he was one of the first to use coordinate geometry before there were computers). He told me when you come across a spiral highway or railway make the ROW a simple curve. He suggested a compound curve on the right of way as the spiral was too complicated for right of way purposes. I still do this and no the ROW is not perfectly parallel but so what.
Although rather RARE (at least out here), I encounter “spiral” Railroad Right-of-Ways from time to time.
These are Warranty Deed conveyances which CLEARLY and unequivocally recite Tangent, Spiral (Searles), Simple Curve, Spiral (Searles), Tangent, Courses ALONG the ROW LINES, or (more commonly) as being parallel to a recited centerline alignment (with Spirals).
Whenever possible (or practical) “they” used a ROW specific [single] Simple Curve (different than the Centerline curve w/ spirals) for the ROW, which embraced the area required for the Railroad construction and operation, but in some cases this was simply not possible (or practical I guess).
In every case that I have seen so far, these ROW Spirals have been Searles Spirals, so this isn't a big deal, and the ROW is in fact parallel to the centerline spiral (the Searles Spiral being a compounding series of simple curves).
I have several examples of these Deeds in my files, but they are in Storage down in Utah.
Where things get a little tricky, is in “blanket” Right-of-Way deeds, that say... “a strip of land 100 feet wide running parallel to the centerline as staked on the ground through the NW1/4 of Section X.”
Unless and until you get out there on the ground, and tie in all of the EVIDENCE, you don't really know what to expect, or how to proceed. Many of the Local Railroads actually went out and set formal Monuments (often, but not always, recited in the deeds) at the PS, PCC, PCS, PT, and the existing Fences conform to these monuments. In many cases, the ONLY way to “FIT” this physical/spatial evidence, is to solve the Parallel Searles Spirals (not a big deal).
Your Mileage may of course vary, and the contrary may be shown.
Loyal
