Can anyone explain to me why a soft or uneven track will cause a derailment? There was a derailment not long ago in my area due to track being rough (crews reported a bumpy ride in that spot).
So what is going here? Do wheels actually lift off the railhead in these bad track areas? I guess they’d have to… but it’s hard for me to imagine such a heavy thing as a railcar getting airborne on the track. Any insights on this?
Thanks.
I’m no expert, but just from casual observation, have you ever noticed the way cars rock from side to side when transitioning form a level spot to an uneven one?
Once they get to swaying, it’s pretty eriee standing very close to such and area.
I would imagine that such movement , over time, causes spikes to loosen, from the extra lateral force on them.
Just a guess though.
Low rail joints are mostly to blame for the rocking action of which you speak. At certain speeds, which varies depending on the distance between opposing rail joints, a harmonic rocking motion sets up. Each low joint increases the amplitude of the sway. Eventually the speed becomes “just right” to amplify the rocking, causing the car’s wheels to actually lift off the track, and then eventually either the car tips over, or the rail under the non-lifted side tips and the car falls to the ground.
In years past, when most rail was laid with 39-foot sticks, operating crews were instructed to avoid the “critical speed” that caused most derailments of the sort we are discussing here. The “critical speed” was between 12 and 22 mph.
On the CNW, we had rules that prohibited us from operating in the critical speed zone; unless we were accelerating or decelerating, we had to go either 10 mph or 25 mph, not in between.
The place where noticed it the most was at a local bridge, right where the track transitions onto the bridge.
If you looked at the track up close, you really couldn’t see anything unusual, but if you were 1/8 mile down the track you could see a slight dip in one rail just before the bridge.
And, especially the hi cube box cars would get to rocking, as they passed from low spot to bridge.
After they re did all the ballast, the problem seemed to go away.
Keep an eye on that spot. Over time (depending on how deep the resurfacing was done), that low spot will redevelop. The track guys on the CNW called the phenomenon “memory” in that the track and ballast somehow retain a ‘memory’ of the pre-existing condition, and the flaw will redevelop.
I’m sure the mudchicken can give us a much more detailed (and possibly corrected) assessment of the situation.
Harmonic rocking is what you are thinking of.
At the speeds Big Z posted, the cars can create a wave effect, and if you mix a empty or two in with a block of loads, or vice versa, you can tip it over.
The weak link effect, so to speak.
Ed
Yes, harmonic rocking speed, between 14 and 22 MPH, by our old ETT instructions, on everything but solid empty trains, so it was either run real slow or violate by about 3 MPH, in a couple of locations where we had 20 MPH restrictions account non-electric lock switches.
Back when there was a lot of jointed rail, you can guess what most hoggers did.
It was a looser time!
Soft track induces wheel lift, partially from the harmonic rocking, partially from the friction on the wheel flange on the gauge side ball of the rail which dynamically allows the wheel to climb. Once the bottom end of the wheel flange gets on top of the rail - it’s all over and the wheel can go over to the field side and the derailment begins.
If track dynamics weren’t so foregiving, this would be a greater problem than it is.
Soft ballast, low joints, worn rail,damaged (“burned” rails are the most common) rail, damaged (worn) switch frogs, points (that have been “picked”), all contribute to the potential for a derailment. All of the above are the results of diferred maintence,or track that is in an area that is poorly drained, or unstable. Most employees are sharp enough to know where the bad spots are and they excercise caution.
Mudchicken:
Is the bit I wrote about the “memory” correct, or were the CNW guys a bit off?
With jointed rail or poorly drained ballast pockets (which tend to be off the ends of newer larger bridges with headwalls or dump planks) - YES
Mudchiken:
I saw reference on a MOW form to a “badly pumping joint.” I’m assuming that refers to a joint that has lots of up-and-down action when the train passes over it. Correct?
Cornmaze–
Don’t know where the ferrous fowl is right now, but the answer to your question is yes. And, if there is no geotextile filter under the track, the up and down action will literally pump mud from beneath and in the subgrade up into the ballast. Typically you will see the effect in what looks like (and is) very muddy ballast.
Hope that answers your question.
The mudholes pump so badly that you get pseudo-crawdad holes, complete with the little mound/towers.
drephe- iron feathers has spent his weekend working on a proposal put out by the regulators in cshave’s backyard that tells me that state is in big trouble with people in charge that shouldn’t be. The lack of RR basics & survey basics understanding is absolutely frightening.
Uh-oh!
The memory effect – which is quite real – is usually one of two things: this may be a soft spot at some considerable depth in the subgrade or it may be due to differential compaction. If track is raised or surfaced, it is almost impossible to get the same compaction in the new raise of ballast as you had in the old after years or decades of pounding (this does NOT apply to mudpumping or just plain soft areas!). With time, the spots that were raised more will, paradoxically, settle more. After a few decades the effect will disappear… This is also why when track is surfaced, whether it’s raised or not, there will usually be a slow order until a certain number of tons have passed over the area.
MC – you shouldn’t be surprised at the lack of RR basics & survey basics understanding – but it is frightening (if not downright terrifying!)
I teach workshops on track inspection and maintenance and I find many railroaders know many of the basics but not the whys. The issue of car rocking, track surface memory, etc., is actually rather complicated. It gets measured several ways, such as runoff, deviation from uniform profile, deviation from zero crosslevel, warp, etc. There are also FRA regs for such things as repetitive low joint combinations. There is a surprising amount of math if you look at the issue correctly, far more than most track employees are taught and probbaly more than most supervisors completely understand.
The issue of track memory is a serious one. The track fill and subgrade have impact from passing trains passed into it, thus fill material can be crushed or decomposed from the pressures. This means that certain spots have inferior subgrade conditions which lead to spots returning over time. Welded rail will often show signs of where the old joints were for just this reason.
Transition areas between track structure types is also a problem area. Track typically has some flexing in it as opposed to bridges. This means a very short area has to handle the change in design, which often leads to track degradation at that point.
This topic can and is many days of lectures and discussions just to get the basics of the field’s knowledge.
Bart Jennings
Bart – thanks! Wish I were younger – I’d come listen any time.
The problem of transitions is really vicious. As you note, it is because of the flexing (a pretty necessary part of track, considered as a complete system subgrade to rail head) in normal track sections vs. very little – if any – on bridges. Track degradation? Oh yes. This is one reason why I would specify ballast deck bridges any time I could (can’t always use them). I would note, though, that the problem isn’t just bridges. Highway crossings usually have enough of a different modulus, too, to cause problems, and so does the change from cut to fill (which is why cuts are taken deep, then backfilled).
Bart - Pay a visit a little more often. Would have been nice to have you around a few weeks ago to deal with the fella who could not understand where something “looked bad” and was still within FRA tolerance. Going to Louisville this fall?
Mud
A local branch in my area has a good number of mud-pumping joints. Ballast is a mix of gravel, crushed rock, pea gravel, and mud.
For the past 20 years that I’ve noticed, the railroad dumps new ballast between the rails after tie change-outs. Then machines come in to groom the ballast. After all is said and done, the new ballast is shoved off onto the shoulders of the track where it doesn’t seem to do anything, and the muddy, pumping joints remain as before. (?!!)
I don’t get it.