OK, here's something to get you guys thinking. I am doing a study for our group here in Niagara to see what type of coach can go around what degree of curvature.

Take for instance, a 4 axle coach. What is the tightest curve that say 2 or 3 of them together can handle. In the same respect, lets also look at a 6 axle coach, under the same circumstances. Now, I'm sure that the different manufacturers had different specs on thier cars. Also, for instance, would a commuter car, say a Long Island 4 axle commuter car, be able to handle something better than a 4 axle Pullman. The coach lengths should all be figured in around the 80ft area.

Now, lets apply the same theories to locomotives. I've put a hudson around an 18degree curve, and that was pretty tight. I figure that is about the tightest you'd want to put a steam locomotive with 6 drivers around.
So, there you have it. Those are my questions, so lets see what kind of answers we can come up with. At this time I don't have the actual degree of curvature, so just go with what info I've posted above.

Ken Jones Jr.
Niagara Railway Museum
http://www.nfrm.ca

Someone with a grasp of calculus will have to answer this one, that is to say, someone other than me. But I can tell you it doesn't have anything to do with the number of axles. Three-axle trucks have plenty of lateral motion, especially if they're still solid-bearing.

What counts is these dimensions: car length over the buffers and diaphragm hangers, distance between truck centers (and the resulting overhang between the truck centerpin and the buffers), coupler length, and coupler swing. You will also need to know not only the tightest radius of your curve, but the rates at which the curvature changes, as actually measured. Was your curve laid out by an engineer with parabolic transitions, or a track gang with a bulldozer? Taken together, these dimensions will tell you whether the couplers will run out of swing in their hangers, or the buffers or diaphragm hangers will go bang as the carbody corners come together and slide past each other going into and out of the curve.

Remember that pulling or pushing makes a difference as the slack runs in in the couplers and draft gear. Using interlocking (Types "H" or "F") couplers makes a big difference, too, as these drawheads form a single drawbar when coupled together. Type "E's" will give you a lot more freedom due to the movement between the couplers. The results will be different, too when cars of different dimensions are coupled together in varying combinations. Some passenger cars have very limited coupler swing, and baggage-car trucks are closer to the ends of the car than on coaches, and their couplers are shorter.

When we installed Type "F" couplers on some 80-foot CN coaches, we didn't know how to calculate any of this, so we just gave the new 100,000-lb. AAR-standard coupler hanger as much swing as the centersill of the car would permit, and then watched and prayed as the cars negotiated the tightest curve we knew they would operate over. (It worked, but only so long as the draft gear was in tension. Roughly speaking, a 2-8-4 and 80-foot coaches with Tightlock couplers both seem to bind up at about the same point.)

Roughly I remember hearing that lightweight (streamlined) 4 axle cars would go around an 18 degree curve, so amazing you should choose that as your example. But that was a 'design' figure with everything shiny new, no wear, and well lubed. I would derate that for worn or sharp flanges, less than perfect track, old rusty coupler hangers with lots of drag and friction, etc. Lets do this example for level track also, as any vertical easement or kinks with a load under tension, might lift one car or another.

Also, if you really "HAD" to move a car around a tight curve, I would consider using it as a single car move, perhaps pulled by a critter at less than slow speed, and with a wreck knuckle adapter to provide additional swing.

Bob Kutella

The limit on car curving is usually coupler swing. Truck rotation is pretty high for most cars and with the elimination of underslung generators most trucks on current cars will rotate pretty far.

That being said most passenger cars today are designed to negotiate a 250' (23°) single curve coupled. Reverse curves are usually worse and the the worst case for a passenger car is a back to back curve with little or no tangent in between. The sharpest I've seen for this condition is facing #7 turnouts presenting a s curve. These were common in some terminals LAUPT and Grand Central come to mind. Most Railroads limited this to #8 turnouts which are common in most all terminals.

The year was 1995 behind the diesel shop at CN's MacMillan Yard north of Toronto. That curve was so tight the middle driver of ex CNR Hudson #5700 was right off the rail. And 'no', it was not under steam :)

The design calculations for railroad cars coupled on curves are found in the AAR Mechanical Division "Manual of Standards and Recommended Pratices Section C Part II", which is the Specification for Design, Fabrication and Construction of Freight Cars. The AAR standard uses a "base" car for the design calcualtions. The worst case is a short car coupled to a long car. Both curves and crossovers are analyzed. This standard does not address steam locomotives.

As others have said, coupled coaches present the biggest challenge on sharp curves. The draft gear is part of the problem. The buffers also can play havoc with each other and what they are coupled too. I've seen a coach lifted off the rails because the buffer rode up over a coupler pocket on very rough track. The vertical alignment is as crtitcal as the horizontal, IMHO. Over verticals tend to make the cars hop around. Rusted up draft gear is another problem, as is worn out coupler shanks.

Steam locomotives negotiating curves is another matter. I know of a local shortline with a large 4-8-4 that has a wooten firebox. The track leading to the former enginehouse was layed out to a 19° curve, as was stated in the old employee timetables under the maximum curve this class of locos could negotiate. It "looked" tight but the loco regularly ran over that curve. Passenger cars went over it, coupled together, regularly.

Another RR has a #6 turnout and has moved single passenger cars through it for some time now. Their track still needs surfacing but is fully functional. The rusty rail does create problems from time to time but that is self-cleaning with use.

Many books exist on this topic. Consult a qualified railway engineer or surveyor for further help.

Rob