I doubt it's ever been much of a consideration. Back in the twenties, when 33' rails were still common, the PRR designed a boxcar with 32'-3" truck centers. When 39' rails were becoming common in the WWII era, so were 50' boxcars with truck centers around 41'.

If the Mechanical Dept. ever gave rocking cars any thought at all, they'd just grumble that the Engineering Dept should fix the track.

"close" only counts in horseshoes... when dealing with harmonic resonances, a bit of difference makes a lot of difference. 32'3" is sufficiently different than 33 to do the job, if that's an issue. I've scanned halftone work at 300dpi and gotten horrible interference patterns. I scan at 305 dpi, and they go away.

The speed where this became critical was around twenty mph. The Rio Grande had restrictions on train operations which said if a train could not go at least 25 mph, it was to run at less than 15 mph. Rock off derailments were a problem at one time, but welded rail has put an end to this problem.

At one time I was called by an engineer from the C and O who had several derailments they were trying to figure out. He had noticed new WP cars of the same series involved in several derailments. He was correct. The cars had Ride Control trucks from a foreign foundry where the snubbing was not working correctly. These were IPD cars on a short lease so we just pulled them out of service. Probably bankrupted the car owner. Without snubbing these cars would rock off the tracks under the wrong conditions. Some of thes same cars had derailed on the Rio Grande.

I don't think the problem is related to welded vs. jointed rail. It appears to have more to do with the overall condition of the tracks, and sharpness of curves. CSX used to have a rule - Restricted Equipment Rule 7 (RE-7), that a car that came under this rule, in territory designated with that restriction, a train that could not maintain 21 mph or faster, had to drop below 14 mph. My understanding has always been that 15-20 mph was the range where a car on sketchy track was most like to rock-off. In Baltimore that rule caused some grief, because the Beltline - the single-track part of the Terminal from the Bailey Wye to Clifton Park, through the Howard St. tunnel, was designated as RE-7 territory, and Eastbound it's a .8% uncompensated grade, with the steepest part being around the Baltimore Streetcar Museum, oh, and the speed limit is 25 mph to begin with. Needless to say, getting a run at the grade, the bottom of which is in the middle of the tunnel, is the goal of every engineer with a train that is close on maximum tonnage rating for the locomotives on the train. With a RE-7 car in a train, you knew you were going to end up below 14 mph, and probably below 10 mph, with a distinct chance of stalling before Huntington Av., with a train that was maxed out with a car with that restriction. CSX eventually did away with that restriction on the Beltline, and it looks to be completely gone from the rule book presently.

If you're talking about some sort of vibrations caused by the click of the wheel on the rail joint, yes, you are correct. Everyone else is talking about the car rocking, and all that takes is for both trucks to be in the dip caused by a low joint at the same time. A foot or two is less than the wheelbase of the truck; also less than the length of the depression caused by the low joint, so yes, it matters. But North American railroads have always used a variety of car lengths, with a variety of truck centers, so it is impossible to pick a rail length that avoids them all. The better solution was to eliminate the low rail joints, which is what welded rail does.

There's an Urban Legend that English railways have rail joints opposite each other because they use 60 foot long rails and 60 foot long passenger coaching stock. A sequel to it claims that if American railroads ever adopt double length 78 foot long rails, they will have a problem with staggered joints and 85 foot long passenger cars!

Opposed joints result in bounce. Staggered joints result in rocking. There was opposed jointed track in the US, but when the knuckle couplers were adopted bounce would cause them to come uncoupled. Many of the opposed jointed lines were converted to staggered joints to avoid this problem. On passenger cars the six wheel trucks reduced the low joint problem and rocking.

Having worked for CNW since 1975 I'd regularly come across 31' 33' and 36' light rail on the Webber line and Troy Grove sub.

36-foot rail is merely used 39-foot rail that was relaid after first having 18-inches cropped off of each end to eliminate end batter, worn and cracked bolt holes and any bonding cables. 31-foot rail sounds like 33-foot rail that was cropped 12 inches on each end.

After each cropping, the rail is typically relaid on a less important branchline. I once inspected a former PRR branchline in Ohio that was laid with 130PS rail in 54-foot lengths. On closer inspection I realized that each was actually two 27-foot rails that had been thermite welded together. The thrifty PRR had taken 39-foot rails through three cycles of cropping - 39' to 36', 36' to 33' and 33' to 30' - before doing it one more time and then welding them, resulting in the 54-foot lengths.

/s/ Larry
Lawrence G. Lovejoy, P.E.

I've enjoyed this thread; it's very interesting.

On the former Southern Railway rails of The Bluegrass Railroad Museum, much of it is 75lb 33 and 30 feet long lengths. We'd thought for a while that it was 85lb rail, but each rail that we replaced with that size, it turned out to be too tall.

We also have a pile of 60lb rail in 33 and 30 feet long lengths, off of a nearby Army Base.

What methods would have been used to crop the rail? In the manner of the previous post by Mr. Lovejoy? Also, what mthod would have been used to cut rail, down the track, by a section crew? In the pre-gasoline engine era?

Thank you,

James

Most rail trimming or cropping would have been done back at a shop with a large power saw. The roundhouse facility in Martinsburg, W.Va., was used for switch fabrication after it was replaced as a division point facility around 1895 or so; it continued in this function and others (such as rebuilding track scales and fabricating bridge components) until its closure in the 1990s. The power saw for this was a large electric motor affair in an iron shed behind the roundhouse, and fed with a roller system similar to that of a sawmill. That saw made the loudest noise you ever heard.

This facility has been the object of preservation efforts over a number of years now. Unfortunately, CSX took out almost all the tools that were part of the shop; some were real antiques, including some large planes that had the bed move under the blades, rather than moving the blade, as you do in a hand plane. There were also punch and shear sets as well. All that's left today are a forge or two and a steam hammer in the blacksmith shop, the latter being run for demonstration a few times with compressed air.

Prior to that, rail (in the much smaller sections used in 1900 or so) was cut by hand. You can imagine what the guys who did that all day must have looked like; I don't think you would want to get into an argument with them or anyone else from that time. I would imagine a track crew also would cut rail by hand.

I seem to recall the Arden Trolley Museum (now the Pennsylvania Trolley Museum) had some powered rail saws (reciprocating) and rail drills that ran on compressed air from a locomotive or a trolley car. These devices were slow and cumbersome compared to modern equipment. Perhaps someone here can provide photos or other illustrations of this equipment.

In the summer of 1968, I saw PennCentral rail gangs breaking 155 pound Pennsylvania System rail to make odd lengths to close a gap at the end of a welded rail string. They used a rail chisel to make a notch in the base of the rail and then pounded the rail with a hand hammer about 20 times until it broke cleanly.
Sample purchase order contracts for rail from the mills required the buying railroad to accept up to a certain quantity of rails shorter than 39 feet long in the order, by full feet, but no shorter than 27 feet in length. This way the mill didn't get stuck with the short ones, especially if it was a section that only the PRR would likely order. Short sections could be used fin switches, crossings, and guard rails, or for welding into extra long lengths. Those 54 foot long rails might have been made from 2 of the short new ones.

The ex-PRR line from Louisville to Indianapolis has rail that was cropped, some once, others twice. As a result, it was one reason for the speed restrictions on this line in the PC and Conrail eras for Amtrak trains. The line was once host to PRR T1s, 2-10-4s and high speed running. But, by the Amtrak era, the Floridian had several derailments on the line.

Thank you for the history.

I'd heard of the chisel and hammer method. I'd like to see that done, at some point. It sounds like it might make a good demonstration (as would many old fashioned work methods), but whaling on it for the 20 licks, hoping that it would work, would be a little nerve wracking.

James

Keep in mind the ever changing metallurgy of the steel. From control cooling to vacuum induction to the fancy new Japanese rail steels (to Reardon metal)... that may have obsoleted the chisel and hammer method.

If a few well placed chisel strikes could break a rail or weaken it to where traffic would break it, you'd think it would be a favored method by the OSS and the British Dept. of Dirty Tricks. Lot easier to sneak a chisel behind German lines than explosives.