I’m in the process of rebuilding some track at a model railroad club. One of the problems I’m trying to correct is that some of the curves have very irregular radius, probably because they were originally laid out by eye by someone who was vision impaired. I’m very familiar with the techniques of laying out curves by measured length of string (or straight edge), or by the use of templates. But neither will work well here because the scenary (which I don’t want to remove) will get in the way. It occurs to me that an alternate way to both measure the curvature and to correct the alignment is by measuring the deviation from a straight line over a given distance. In other words, if you put, say, a 1-foot ruler so that one end is at a right angle to the track center line, you should be able to determine the curvature from the number of inches the other end of the ruler deviates from the track center line. Trouble is that I’m not an engineer or a draftsman, so I don’t know how to easily make the conversion from inches of deviation to radius. I could, I suppose, figure it out by laying a big sheet of paper out on my garage floor, drawing various curves and measuring the deviation. But it occurs to me that someone may have done all of this work before, perhaps in constructing a club layout with very broad curves that can’t readily be located with meauring strings or templates. Any takers?
Consider picking up a few track gauges from Ribbonrail. They have both 5 and 10-inch lengths in HO scale. They also have N-scale gauges. Each gauge is a shaped curve which exactly fits between the rails. The make a spread of gauges of various curve radii.
If the curve isn’t quite the desired radius, you can tell whether it’s too large or too small based on where the gaps are when you try to fit the gauge between the rails.
I think the only time you really need to worry about a curves radius is when it is sharp, or minimum radius. Even if it is an advantage to real railroads to have the same radius throughout the curve, I’m sure there were exceptions due to some geographical feature.
What I would try to do is relay the curve with flex track and make it as smooth as possible on the roadbed that is already there. Plus, if you aren’t willing to remove any of the existing scenery, you won’t be able to change the curve very much anyway.
Is there any chance the discrepancy in curve radius is because the person laying the track incorporated easements?? If so, the curve would start on a very gradual radius, then become progressively sharper, then after a distance begin to become more gradual again when transitioning back to straight track.
I suspect the curves were laid out that way on purpose. The reason a lot of layouts have a nice even, almost “stamped out” 180 degree look is because many of us have to have the most curve we can fit on our limited spaced home layouts.
Years ago I visited a layout that had large sweeping curves and none of them had the 180 degree turns that most of us are use to seeing. By that, I mean a curve would start out with the minimum set radius for the layout, that would get larger as you went around the curve. Or the reverse would happen, where a larger curve would tighten down to the minimum set radius for the layout. This removes some of the toy like train set look that a lot of our home layouts have.
I cut out cardboard templates of various radius to make sure I was never below my minimums on the different parts of my layout. That being said I laid my track by eye and that makes my track routing flow better in my opinion. As I add in the surrounding landscape it appears the route was chosen to pick its way through the mountains much like the prototype would. A stamped out 180 degree turn looks like the track was put down before the mountains were there.
If this is a large layout the need for 180 degree set radius turns may not be necessary and that may be the reason curves with a changing radius exist. I wish I had that luxury on my home layout. I have one that goes from a 32" and opens up to a 36" as it goes around and even with that slight difference I think that 180 looks much better. Sure I gave up a 34" radius curve all the way around, but it was a tradeoff worth making for appearance.
Brent[C):-)]
I’m going to break this down into two sub-sections, one of which will answer your original question:
- To determine the radius of a smooth model railroad curve:
- Pick a point somewhere near the center of the curve.
- Using a ruler, measure a ten inch chord in each direction from your base point.
- Determine the angle between the two chords. It will exceed one hundred degrees.
- Subtract from 180. Then divide by 2.
- Look up the sine of the angle you just found.
- Divide 5 by the sine. this is the curve radius.
And now you know why prototype curves are stated in degrees…
- Why are you re-laying the track?
If the problem is a series of abrupt radius changes (the square wheel effect) you can correct that ONLY by lifting and re-laying. But if the `problem’ is that the curve radius is not a constant, more parabola than circle, it shouldn’t need work.
In my own modeling, I have a bunch of spiral easements connected by short tangents and arcs of curves. The curvature obviously isn’t uniform, but cars on those rails move smoothly along. None of that, `Student driver taking a corner,’ lurching common to un-eased minimum radius curves. That is the result you should be striving for.
Chuck (Modeling Central Japan in September, 1964 - on trackwork as bulletproof as I can make it)
We have used the ribbonrail gages and they are great. We have also cut out of masonite curve templates. Both are great methods. The longer ribbonrail are more effective to me. This is just one place to get them.
http://www.toytrainheaven.com/.sc/ms/cat/HO%20Track--Ribbonrail
I second the poster who mentioned easements and would add that having compound radii is very prototypical. Think about it like this. If you can have 26" radius for part of the curve but could later open it up to 32" radius, why wouldn’t you.
Richard
Just to clarify so we’re on the same page, if you think of it in terms of the radius of the curve, you would start from straight track and go into a short section of say 60"R curve, then 50"R, then 40" down to say a minimum radius of 30". Then coming out of the curve, it would be the opposite, 30"R - 40"R - 50"R - 60"R then straight again. That’s a curve with “easements” so you can “ease” into and out of the curve…except you wouldn’t normally do it in distinct segments, you’d start with the largest radius and the track would spiral in until it reached the tightest radius of the curve.
Falcon
I would suggest this method of measuring curves: Determine your largest and smallest radii. scribe this onto a piece of cardboad (cardboard like Big Screen TVs come in). Cut out a semi-circle using that radius. Now you know that any curve larger that that will work.
Since real railroads do not conform to perfect circles you can use the cardboad as a guide to check your curves but not necessarily to draw the curve.
Jim
The formula you are looking for is the tangent offset. I will post it below. As a quick and easy alternative consider the bent stick method of laying out a curve with an easement, except in this case pin it in the center-line of your existing track and draw a center line along the bent stick. You will likely need to put a pin near the center of the curve to stop the stick from trying to straighten out the corner.
HORIZONTAL CURVES
That is the link to the pdf document that covers the formula you are looking for.
Sound like since it’s already sceniced, and the turns have real bad whips and kinks, you will just have to relay close ot the origional position, but this time “Using a real good Eye”. If you could cut out a cardbord template to approximate the radius (using one rail), you might be able to use this as a guide.
Nobody who knew what they were doing would have laid out the curves on this layout on purpose. I certainly agree that it’s not necessary (or prototypical) to have a perfectly even radius throughout a curve.
Interesting problem and interesting discussion.
If I didn’t know better, I would think that this thread is about my layout.
When I first laid track on my current layout back in 2005, I used the pencil and string method to trace the curves on the bare plywood surface. I then laid flex track to form the curves. I did not solder the rail connections which resulted in kinks. While I tried to maintain smooth curves, there were variations in the radius.
The result was a series of continuing derailments, especially with steam engines.
My solution has been to completely remove the curved sections leaving straight track at either end. I then replace the removed track with soldered sections of flex track to assure that there are no kinks at the track connections. I use the 10" Ribbon Rail Curved Track Alignment Gauges to maintain a perfect radius. Since the original track layout was intended to be a 32" radius curve, I rely on the 32" radius Ribbon Rail gauge, but I also keep on hand a whole series of Ribbon Rail Track Alignment Gauges ranging from 28" radius to 40" radius. These additional gauges are helpful to assess the true radius of the relaid curves. In order to rejoin the curved sections to the straight sections of track, some easements naturally occur.
Rich
hi Falcon,
the ribbon rail way is easy to get a good idea where you will have to relay your tracks when horizontal curves are involved. Of course you could make your own templates. Hopefully it is possible to maintain the wished minimum radius throughout the layout, without having to change the subroadbed.
When vertical issues have to be solved too, i am afraid a major rebuild might be needed. When you know all the trouble spots you will be able to decide which parts can be corrected without removing any scenery.
Paul
On a real railroad, the scenery often dictates the curves, but on a model railroad, the curves should dictate the scenery. In other words, if reliable operation is important, don’t let the scenery prevent you from correcting this problem properly.
Some photos of the trouble spots would be helpful and some information on the layout construction would help too. Hopefully, somewhere beyond the ends of this troublesome curve is track which is acceptable. Determine what’s needed to get between those two points, ignoring the scenery issue completely, then remove whatever is in the way of the needed replacement. I suspect that you’ll discover that those “acceptable” areas of track cannot be connected by an acceptable curve within the area of the original curve, and that your repair will extend farther than you may think.
Any stop-gap type of repairs which you make now are ones which you (or someone else) will need to make again and again until it’s done properly. The alternative is to make this area of the layout into a static diorama.
Wayne
I know it’s not what anyone wants to hear, but I think that the only way you’re going to make this curve really work the way you want it to is to completely remove all the track and most of the roadbed, and re-build it from the ground up. I had a bad curve which I kept trying to fix bit by bit, but I finally had to start all over to get it right.
As others have said, if this is the case, tearing it all up and relaying carefully sounds like the only viable alternative.
I was involved with a similar project on a client’s railroad where the original crew did a quick-and-dirty job of tracklaying. After months of messing around with spot fixes, he eventually tore nearly all of it up and relaid it. The improvement was like night and day.
Best of luck.
I haven’t measured a radii in the last 25 years of layout building. I use an old method that’s called the Bent Stick or something like that. You take a stick (1/4"x1/2"x 8’ or 10’) anchor it at one end (a person holding it in place or using screws) on the centerline, then bend it around until the curve suits you and trace along the stick to make the centerline, then you just move the stick to the end of the center line and repeat. To meet up with straight sections, mark that first and then use the stick, this gives you nice curves with natural easements. To do a double track main, just measure the distance you want between centerlines and mark that every 8-12 inches and then line the stick up to those marks and connect them with a marker. You don’t have to worry about radii or easements, it’s all automatic. You can find the sticks in the crownmolding/trim section of your lumber store. A wood stick can be bent down to about 24"r, a polyurathane trim stick can probably get down to 20"r or so. This method eliminates the need for all templates, strings and compasses.
note: if you do this solo, you can anchor the stick at the ends by screwing 2 3"-sheetrock screws about 4" apart on one side of the stick and one screw in between the two screws on the other side of the stick, this will also work on foam.
I use this method lay out the outline of the plywood subroadbed first, once that’s cut out and mounted, I then use the stick to lay out the actual centerlines to lay the cork roadbed to. If you’re using WS foam bed, instead of marking the centerline, mark for the outer edge of the foam bed as it’s hard to follow a centerline precisely with the foam.
Without tearing up any scenery your right of way might look a little tight. I think without tearing it all up you might be able to loosen up the middle of the curve and leave the ends of the radius tied down. With the ends tied down you might be able to bring in the pinched portion of the curve. this will flex the ends out wider towards your fixed ends. Depending on how severe your radius is you might be able to work this. This will give you a more uniform radius throughout the curve. if you run into clearance with the scenery you may have to add or remove some track to either end of the curve. Or move the angle point of the curve towards either end to fit. This has worked for me in the past. Good luck…
I don’t know what sort of track you are dealing with but with springy flex track sometimes the problem will fix itself. I once purchased someone elses old layout for my son. There was one place with a kink that would derail anything with a six axle truck. I fretted over how to fix it it as you are, but then it hit me. It didn’t matter what the radius really was as long as the kink was gone. It was Atlas flex track, so I simply pulled the “stressing” nails around the kink and let it take its natural form. The curve went wide and I just pulled some of the nails off the straight and let it shift to one side of the roadbed. It worked great. Later I used this at my club’s layout where one track had a pretty severe reverse curve in it. I just loosened it up (it was glued) and let the track make its own natural transistion and then reglued. Once again it worked great.
Of course if you have the Shinohara type flex track that is not springy or hand laid rail with individual ties this is not even an option to consider.