Prototypical Curve Radii

[Anonymous]

What are some prototypical curve radii? Some of my engines jump the track because it's really tight and I thought that I'd go ahead and relay the track to match real life.

Can anybody out there help me? I have a 2-8-0, SD40, and a F7.

Regards,
Rich

[Anonymous]

You'd better have a lot of room! Prototypical curves had radii equivalents (because that is not the way the construct them) of thousands of feet! The Credit Valley Railroad, built in the late 1800s north of Toronto had radius on the mainline of 1910 feet, and 850 on the branches. That would be just under 22 feet and 9'9" respectively in HO. Most radii in HO range from 18-30 inches on HO layouts.

How tight are your curves? Are there other problems you have with the engines? Do any of your rolling stock jump the curves as well? If you can provide some more info, maybe we (collectively on the forum) can get to the bottom of what must be an annoying problem!

Andrew

[ndbprr]

Real railroads don't use the radius of a curve for a description. They use degrees of curvature. There is a formula but I forget it right now (along with a lot of other things I am supposed to remember). As I remember in HO a 30" radius curve would have a 5mph speed limit on it in real life. EMD engineer manuals list how tight a curve an engine can take both coupled and uncoupled by the way.

[Anonymous]

Andrew:

The most practical way to go is to contact the manufacturers of the locomotives in question and ask them for minimum radii. Most of the current makers have websites you can consult.

As an alternative go to the NMRA website ( www.NMRA.org ) and have a look at the Standards and Practices tab which covers locomotives and rollingstock.

GoodLuck

Randy

[rogerhensley]

What you want is NMRA RP-11 Curvature and Rolling Stock . It gives you the definition of train and rolling stock sizes be seperating them into classifications and then offers a chart showing curves for several scales according to size classification with P being the very largest steamer and passenger car consist.

Go to http://www.nmra.com/standards/consist.html and take the RP-11 link.

[Anonymous]

Thanks Randy, but it was Rich's question...! [;)]

The 2-8-0 (assume IHC or Bachmann) should be good on 18" radius curves. The SD40-2 is a 6-axel diesel, so may need wider curves - maybe 22"+, and the F7 is a 4-axel, so may again be ok on 18" radius.

Anyway, Rich - if you can provide more info, maybe we can help! And Randy has a good suggestion about contacting the manufacturer.

Andrew

[Anonymous]

My understanding of the way curves are measured in terms of degrees is this:

The degrees measurement refers to the degree of bend in the track within 100 feet of run. This is how it would get built, since you have surveyed out a straight (tangent) track centerline, then to begin your curve you string a line out 100 feet, swivel the surveyor's eyepiece to the correct degree, and mark the spot. Move the equipment to the new spot, align it based on the chord (straight-line segment between points on a circle) you just laid out, and then swivel a few degrees again and mark your spot 100 feet out.

Model railroad curvature usually comes in around 100' to 300' in radius, in scale feet. If you measure a 100' long chord across your curve, you'll see the degree of bend is probably over 45 degrees. For real railroads, curvature approaching 10 degrees was considered severe!

[Dallas Morlan]

QUOTE: Originally posted by Avondaleguy

My understanding of the way curves are measured in terms of degrees is this:

The degrees measurement refers to the degree of bend in the track within 100 feet of run. This is how it would get built, since you have surveyed out a straight (tangent) track centerline, then to begin your curve you string a line out 100 feet, swivel the surveyor's eyepiece to the correct degree, and mark the spot. Move the equipment to the new spot, align it based on the chord (straight-line segment between points on a circle) you just laid out, and then swivel a few degrees again and mark your spot 100 feet out.

Model railroad curvature usually comes in around 100' to 300' in radius, in scale feet. If you measure a 100' long chord across your curve, you'll see the degree of bend is probably over 45 degrees. For real railroads, curvature approaching 10 degrees was considered severe!

Almost right. [:)] The stakeout or "deflection" angle turned (what you referred to as swiveling) from the tangent line is one-half of the angle at the radius point (center) of the circle. The railroad degree of curvature is defined by the angle at the radius point for a 100' chord (Straight-line between the points on a circle). Making the curve even more severe! I teach this stuff for a living[;)]

[Anonymous]

I think we're talking about the same thing, despite my disuse of the proper jargon! If I understand you correctly, if you strike a 100 foot long chord across the curve, then the tangents at those two points will diverge at the degree of curvature. You would have to use 1/2 of the curvature degree in order to lay out the chord itself, as opposed to placing a stake in the ground several hundred feet away...

This is what I was getting at, if not very clearly. Thanks for the clarity!

[Dallas Morlan]

Avondaleguy, I believe you've got it[8D]
The tangent lines into and out of the curve could be extended to intersection. The angle at the intersection (Point of Intersection or P.I.) is the difference in direction of the tangents. The angle, measured from a point on the first (entry) tangent at a given distance from the P.I. to a point on the second (exit tangent) the same distance from the P.I. defines the location of the curve. The line between these two points is the chord of the curve. This makes the angle from the first tangent (deflection in my above post) half the intersection angle and the angle at the second tangent line the other half of the total change in direction.

[|)] At this point the class is normally nodding and blinking[|)]

Only when the chord length is exactly 100 feet does the intersection angle define the degree of curvature. The field layout procedure simply strings together a series of these small curves until the required amount of change in direction is achieved. One curve on a section of the Old Penn Central mainline near Dennison, Ohio enters each end of the main curve with a series of three easement curves. (Going from memory here but will define general application) First easement 1 degree curve, radius 5,729.65 ft. for 100 ft. chord; second easement 2 degree curve, radius 2,864.93 ft for 100 ft chord; third easement 3 degree curve, radius 1,910.78 ft. for 100 ft. chord; main curve was 5 degree with radius of 1,146.28 ft and running for approximately one mile. At the exit from the main curve the sequence of the easement curves was reversed.

[zzz][zzz]BY THIS POINT THE CLASS IS USUALLY SNORING LOUDLY[zzz][zzz]

That five degree radius would be very large for a model layout.
For HO = 13.1756ft or 13ft 2 3/32 inch. For N = 7.1642ft or 7ft 1 13/32 inch

[ndbprr]

I understand what you are saying but I don't understand how you determine the center of the curve for the intersection of the two radii unless they are at 90 degrees to the 100' chord. Is that assumed?

[Dallas Morlan]

The radial lines are 90 degrees to the tangent lines at the beginning of the curve, first point on both the arc of the curve and the chord line, and at the end of the curve, last point on both the arc and chord line. By definition a tangent line touches a curve at only one point and a line through that point and the center of the circle is at right angles (90°) to the tangent line.

[ndbprr]

Got it! but that seems to not require the 100' chord segments then.

[Dallas Morlan]

The entire concept is using geometry to work around obstacles. The initial route location surveys were the straight lines only. Curves were then computed to fit in the office (field office in many cases) for later field layout. In the real world the radius lines are of little use. Think of trying to swing an arc with a 1,000 ft. long rope in the woods or mountains[banghead]

The 100 ft. long chords stayed within the area of the roadbed and could be used for construction through the steps of clearing of trees, grading of alignment, placement of ballast and track.

[BR60103]

I just bought a reprint of a book from the end of the steam era which gave some interesting curve data.
The normal minimum radius for all locomotives was 1000'.
It also tabulated the minimum radius with .75" gauge widening for a number of locos. This usually came out to a figure around 300'. This was intended for shop or roundhouse operation, not mainlines. I expect even the 1000' radius would be dead slow operations.
(I can copy the figures if anyone's interested. Many of the articulateds would go around sharper curves than rigid frame locos.)
1000' is 138" in HO.
300' is 41" in HO.

[johncolley]

As a point of interest, I have the specs for an industrial spur off the Indiana Belt line in which it says the absolute minimum radius for such is 459' (or 12 1/2 degrees) This scales out to a little over 63 inches in HO. So for our purposes 36 might be good for industrial or incidental spurs, 48" is a good place to start for mainlines, and from experience, 72" looks great with 85' passenger equipment. Another thing to check out is the "equivalent radius" of turnouts, # 8 and above are in the 110" range! That's a BIG curve!

[BR60103]

That 1 degree curve in Dallas's easement was over a mile radius. That's over 60 feet in HO.
Anyone for buying the houses next door?

[johncolley]

How about finding an empty K-mart? There would be room for a GREAT Free-mo setup! In the interest of realism, it's Big Curves, and big turnout frog numbers I have some 1920's official drawings of the original yards either end of the Cascade tunnel that GN built. From the mainline, primary turnouts were #11 and secondary to spurs, etc. were #9. Thankfully we can approximate this in code 83 with Walther's #10 and #8's. Anything bigger calls for handmade. A lot of modern high speed mainlines have #16 and even some places are #20. So, Atlas, et al, get with the program!

[pkeppers]

The discussion on curve degree sounds pretty good. It should be pointed out that the degree definition on highway curves is slightly differnt in that the degree is the central angle subtended by 100' of arc length, not cord length which railroads typically use. Where I have to disagree is inthe discussion of easement curves, also called spiral curves. I'm not familiar with the specific location noted above that had 3 constant radius curves of progressivly sharper curves leading into the main curve but typical railroad practice is to use what is called a spiral curve, a curve of changing radius to go from tangent to curve. Also, it's interesting to note that highway engineers are now using radius length in place of degree in most instances. It's a result of being forced to do highway plans in metric byt hte federal governmtn in the mid 90's. Degree didnt mean much in metric so we used radius and now that we are back to english radius seems to have stayed.

[Jetrock]

All this discussion of prototyle curve radii makes those of us with shelf layouts breathe easier, since curve radius is a less critical issue...Of course, we could get into the subject of frog numbers to point out that #4 and #6 frogs are almost as unprototypically sharp as those 18" curves...

and then there are trolley curves, which can be as sharp as 40 feet in radius--that's a touch under 6 inch radius, in HO, with plenty of prototype examples to back us up, and single-point switches of equally sharp radius (the frogs are curved but are about the equivalent of a #1.5-#2 frog!)

Of course, the non-trolley crowd don't have to hang trolley wire, and can typically buy motive power and rolling stock without having to haunt eBay and the expensive brass parts of local hobby shops.

The bottom line, though, is that even with the most accurate and detailed model, we are still simulating and tightly compressing the spaces we are modeling, rather than providing a totally accurate representation. Give your curves as much radius as you can--but don't sweat too much if you can't fit a layout with 1-degree curves in your basement!

[NEALNP]

If one were modeling the Tehachapi Loop in N scale, and was wanting scale curvature, the loop would be approximately 14' across! But wouldn't the trains look great!

[Dallas Morlan]

QUOTE: Originally posted by pkeppers

The discussion on curve degree sounds pretty good. It should be pointed out that the degree definition on highway curves is slightly differnt in that the degree is the central angle subtended by 100' of arc length, not cord length which railroads typically use. Where I have to disagree is inthe discussion of easement curves, also called spiral curves. I'm not familiar with the specific location noted above that had 3 constant radius curves of progressivly sharper curves leading into the main curve but typical railroad practice is to use what is called a spiral curve, a curve of changing radius to go from tangent to curve. Also, it's interesting to note that highway engineers are now using radius length in place of degree in most instances. It's a result of being forced to do highway plans in metric byt hte federal governmtn in the mid 90's. Degree didnt mean much in metric so we used radius and now that we are back to english radius seems to have stayed.

I agree with all of your post. The current practice for both railroads and highways is spiral curves. The location I described above was documented on a railroad valuation plat dating from the 1920s. The plat showed the mainline track and right-of-way that may have been in this location prior to the Civil War (Lincoln's funeral train was reported to have used the route). The use of compound curves predates spirals. Use of compound curves on railroads results in track alignment that approximates a sprial. The use of spirals increased as mechanical calculators were developed in the late 1800s and really became common with the use of computers (1950s) for alignment calculations.

There are two points I should have made clear above. First point, having access to the radius point of a curve is not necessary to layout the curve. The proceedure that has been used in real world since the beginning is the deflection and chord distance mentioned above. Second point, using this combined with compound curves and flex track should work well for approximating a spiral. Setting "controlling points" at intervals of 100 feet (13.79 Inches in HO) or 50 feet and allowing the track to adjust in between these points. Haven't applied it to a model railroad yet but this is what happened in prototype operations many years ago.

[johncolley]

If you want another way to form an easement other than the batten try this on butcher paper:
1. decide on the overall degree of curve from tangent to tangent.
2. using an appropriate( HO, N, etc.) scale and a protractor, layout the first segment 100 ft. at 1 degree, double that on the3 second segment 100 ft. at 2 degrees.
3. continue doubling the angle for every 100 feet to the middle of the curve, then start cutting it in half until you are at the final tangent.
4. lay some flex track on this centerline and try your longest rolling stock on it to see how it will look on the layout.
5. If you like the results, draw lines 1 1/2" either side of the centerline and use this as a pattern for your sub-base and roadbed. I find it is a lot easier to use the batten method,but, hey, whatever works best for you. Enjoy trying out different methods and learning, that's what model railroading is all about, enjoyment!

[Dallas Morlan]

John Colley, thanks for the reply. Very similar to what I had in mind. However, spirals have defined relationship to the circular curves. What I have in mind is using the deflection layout method to define 50 ft. control points. Starting with a radius 4 or 5 times the final circular curve. Changing to a curve radius 3 times the final circular curve. Then to twice the radius of the final circular curve and allowing the flex track to make a 150 ft sprial. This would be about half the length of spiral normally used in highways and the limited amount railroad work I have seen. Just more of the selective compression necessary for modeling. For N Scale the spiral would be about 11 inches long.

[johncolley]

As I said above, the batten method is the easiest for me. I use generous easements roughly based on the three following spec's.( for HO): up to 36" radius - Offset = 1/2", L/2 = 9", 42" radius - O= 9/16", L/2 = 9 3/4", 48" radius - O= 5/8", L/2 = 10 1/2", 54" radius - O = 11/16", L/2 = 11 1/4", and 72" and larger radius - O= 12", L/2 = 12". And don't forget to leave 100 feet between reverse curves.

[Anonymous]

To Dallas Morlan, I can only say that I'm glad that all it takes me to lay out track is to drill a couple holes in a stick and to swing an arc on some plywood. Thanks for the great expertise. Bob T

[ddechamp71]

QUOTE: Originally posted by masonjar

You'd better have a lot of room! Prototypical curves had radii equivalents (because that is not the way the construct them) of thousands of feet! The Credit Valley Railroad, built in the late 1800s north of Toronto had radius on the mainline of 1910 feet, and 850 on the branches. That would be just under 22 feet and 9'9" respectively in HO. Most radii in HO range from 18-30 inches on HO layouts.

Use Z scale.......In Z, a 10° curve has a 80cm / 32 inches radius.....
Dominique

[Anonymous]

QUOTE: Originally posted by rt2907

To Dallas Morlan, I can only say that I'm glad that all it takes me to lay out track is to drill a couple holes in a stick and to swing an arc on some plywood. Thanks for the great expertise. Bob T

[Anonymous]

The following data came from a Fairbanks-Morse brochure on its "Train Master" locomotive. Brochure was entitled, "The Lackawanna Story."

Minimum radius of curvature = 191 feet (30 degrees)
Minimum radius of curvature with locomotive coupled to AAR 40 ft. 6 in. freight car = 212 feet (27 degrees).

212 feet in HO would be about a 29 inch radius