Now y’all can make your high-school kid expertise observations again, but this report specifically mentions worn constant-contact side bearings as a major contributing cause of this incident, and describes that it is difficult to detect the condition with visual (and, by extension perhaps, usual ‘remote sensing’ at conventional detector locations) methods. The number of cars potentially affected by this condition is, to me, staggering and significant.
What is the current wisdom regarding ‘automatic’ detection and remediation of this problem? What added inspection procedures or tests ought to be developed or “canned” to address the situation?
Thanks! Sounds like a problem with an expensive remedy.
"3.0 Findings
3.1 Findings as to causes and contributing factors
The accident occurred when the A-end truck of the 6th car from the tail-end of Canadian National (CN) freight train M36831-01 (CN 368) derailed at Mile 153.92 of the CN Kingston Subdivision. When the derailed truck contacted the main track crossover at Mile 151.45, the 5 trailing cars derailed and separated from CN 368.
With VIA Rail Canada Inc. (VIA) passenger train No. 47 (VIA 47) slowed to 51 mph and Canadian National freight train M36831-01 slowed to 11 mph, the trailing A-end of car GTW 623373 struck the front of the VIA 47 locomotive and scraped along the north side of the locomotive and the 5 VIA coaches.
The excessive truck hunting on car GTW 623373 was influenced by the type of car, the speed of the train, and the worn condition of the A-end truck, which further reduced the warp resistance in a truck type that was known to be susceptible to excessive truck hunting.
An opportunity to restore the constant contact side bearing resilient members was missed when they were not replaced during a related repair at another railway. Allowed to continue in service, the resilient members likely deteriorated to the point where they no longer provided effective damping.
3.2 Findings as to risk
If inspection programs do not identify constant contact side bearings with inadequate preload force and ineffective damping, trucks that have the potential for excessive hunting will remain in service, increasing the risk of derailments due to wheel climb or wheel lift events.
3.3 Other findings
The dragging equipment detector and related alerting systems and procedures, which were designed to inform train crews of the derailed condition of a train, worked as intended.
The train crews of both Canadian National freight train M36831-01 (CN 368) and VIA Rail Canada Inc. (VIA) passenger train
I would wonder if there are established parameters for the parts in question. Did the unidentified repair facility look at them and decide they were fine? Did they even look at them?
This looks to be a part that would be virtually impossible to check in the field. Many of us have “checked” the shocks in our car by bouncing on the hood or trunk. That would be a little difficult on a railroad car.
It was noted that the bearing can be in contact with the car and have no effect due to weakening of the resilient member. That means a visual inspection is virtually worthless, unless a gap can be seen.
That leave us with the need to verify proper operation any time the truck is separated from the car, and with an accepted method (as well as specifications) for doing the testing.
Simply renewing the parts any time the car is off the trucks could be somewhat expensive, depending on parts cost. And unless the cars are thusly disassembled on a fairly regular basis, it still wouldn’t cure the problem.
It seems that the design for those constant contact side bearings is being asked to thread a needle by sharing the loading with the center bearing to thus carry just the right amount of weight; not too little, and not too much.
But the inevitable wear changes the loading that they carry. On top of that, there is no practical way to thoroughly inspect them.
Overall, it seems like a bad design with too many Band-Aids subsequently added in an attempt to make it reliable.
The basic problem is that there is a center bearing intended to allow the truck to pivot easily when under load so it can go around curves; and then a friction brake is added to the bearing in order to inhibit its ability to pivot so it does not hunt. It is trying to have it both ways.
Optimally, the truck has to pivot just easily enough to go around a curve, but no more easily than that. The requirement to thread that needle seems too refined for the general design of freight car trucks. A better approach would be to cure the causes for truck hunting instead of trying to inhibit it after it starts.
Somebody once told me that there have been truck pivot problems due to a little too much anti-hunting friction being applied. The result was that empty cars would climb the rail in the relatively sharp switch curves because the friction of the constant contact side bearings was too great to allow the trucks to pivot for the curves.
I advise that you look at this more carefully, with a better engineering perspective.
The constant-contact side bearings are NOT, to my knowledge, designed to optimize truck yaw or hunting behavior. This is a secondary consequence of their being correctly installed and aligned/calibrated to the right lateral support contact force. As you can readily imagine, there are all sorts of considerations involved with the frictional contact between the bearing and the surface it bears upon, which are completely incidental to the CCSB ‘doing its primary job’ but which are of great interest and, now that the TSB has brought its importance to the fore, great importance.
There really isn’t a ‘better’ way to control truck rotation than by using something out at the radius the CCSBs act at. My father tinkered for years with hydraulic rotating stops, adjustable progressive resistance, emergency truck centering and so forth acting around the center pivot – there’s just not enough structure and too much potential force for those sorts of approach to work well. So mechanisms and devices out at the CCSB are the ‘right’ approach, since you’re already providing much of the needed structure to implement them in having functional CCSBs in the first place.
One thing of importance is to develop the right kinds of progressive resistance to truck swing. Something with ‘restoring spring’ action is valuable, but not if it’s going to put progressive resistance on the bolster of a 3=piece truck; not only will it predispose you to derailing on sharp curves but I suspect to things like lozenging and skew if other parts of the truck structure are worn or misconfigured. Therefore I would keep the ‘general’ restoring force relatively light.
Some version of the approach used on steam-locomotive trailing trucks might be used here, where the CCSB contact surface is slightly ‘ramped’ progressively away from
“Since the 1990s, to address the tendency for trucks to hunt excessively at high speeds, railway cars have been built with CCSBs installed.”
There are two issues:
Truck pivot
Truck hunting
Constant contact side bearings arrest #2 while allowing #1.
I understand your points about restorative action it but I am not sure where it fits into the above purpose. You certainly don’t want a spring-like restorative force on the truck pivot. The pivotal truck alignment is restored by the track resuming tangent after coming out of a curve.
And it is not really restorative action that you wan
If you look carefully at the report, you will see that the problem with the CCSB loading is related to wear in other (unspecified, but critical) parts of the truck in order for the dangerous effects to occur. I wish they had made this a bit clearer in their prose, just as I wish they hadn’t left the unwary with the somewhat post hoc ergo propter hoc idea that CCSBs are intended primarily as an anti-hunting device by restricting truck yaw…
Note, again as I understand it, that truck yaw (i.e. ‘nosing’) is NOT hunting, which is a combination of nosing and rolling. The CCSB acts in part on this by preventing or controlling the roll to where it will not ‘couple’ into the truck dynamics. When the truck can ‘lozenge’ there are additional degrees of freedom to produce what amount to complex resonances driven (primarily) by tread and flange force cycling. I trust buslist, Paul Milenkovic, PDN and others who are knowledgeable in this will correct the picture as appropriate.
Fixing the roll instability is, I thought, what adjusting the CCSBs to ‘spec’ primarily does, but there is just as much value in addressing the (uncontrolled, with good center pivot lubrication) yaw. Whether or not this is best done via friction vs. making a consistant effort to eliminate play in the sideframes, etc. is an interesting question. I would note that the idea of providing skew bracing (via, for example, the little X-frame that attaches to the special brackets on a couple of truck types) has not seemed to catch on at all in this country, although extensively tried elsewhere (M636C, what was the use and acceptance in Australia?), which is an indication that a properly-maintained truck will not lozenge even at comparatively high axle load and fuel-efficient track speeds.
A big piece of the puzzle is that a three-piece truck is inherently possessed of a relatively low polar moment of inertia given the lever arm of flange and tread f
Wizlish,
Here is a quote from a patent on Constant Contact Side Bearings (my emphasis in red):
“The art has also contemplated constant contact side bearings for railcars. Constant contact railcar side bearings not only support a railcar body with respect to the bolster during relative rotational movements there between [as do the “gap style” side bearings that preceded the development of CCSBs], but additionally serve to dissipate energy through frictional engagement between the underside of the railcar body and a bearing element thereby limiting destructive truck hunting movements.”
Here is a quote from product information by Miner (my emphasis in red):
I was coming around to agree with you about the anti-yaw functionality of CCSBs but was too lazy to edit the prior posts. I subsequently got something of an earful on the subject. Let me just say that the design of CCSBs could be ‘adjusted’ to make them definitively better in that part of their role comprising truck-hunting reduction.
The thing from White is from volume 2 on the American passenger car, I think shortly after he was discussing the Winans antifriction wheel (which also precluded any effective foundation brake short perhaps of the sort of arrangement used on the German high=speed locomotives). If I remember correctly the ‘flexible’ truck he was describing was just what you were saying; the whole ‘sideframe’ of the truck was spring, with the ends resting on very small axlebox bearings (in those days they apparently thought that railroad bearings worked like watch bearings, the smaller the journal area the lower the friction…). This was if I recall correctly in the 1830s with some mention it was used later for things that did not need a brake … tenders being an example of that. I will find a page for you if I can locate my copy.
Several of these engines have Fox trucks on the tenders, but what I am referring to are the tender trucks on locomotive #936 shown in the 5th picture from the top. They also show up on a few other engines such as #1310 and #1353 a little further down on the page.
I have always wondered what these types of trucks were called. They were widely used. The famous #999 has them on the tender. The latest examples that I know of were used on CMStP&O ten-wheelers built around 1900.
This type of truck has a center bearing that is non-load bearing. All
Wonder if the loco truck problems that Amtrak is having is related to this problem ? Maybe Amtrak’s failure to maintain proper procedures is of this nature ? Know Amtrak cannot keep enough trucks in stock.
I don’t know about the Amtrak problems, but the only time I have seen any truck hunting was the lead truck of a Great Northern FP7 A-unit while driving alongside of it at 80 mph. It was cycling about 4-6 times per second with a total swing of maybe 2-4”
I have seen a lot of trains go by, and never have noticed any truck hunting, assuming that it would be visually obvious.
You literally beat me to it by 10 seconds and a slip of the finger!
“Making the Case for Long-Travel Constant-Contact Side Bearings”
Now, it does have to be noted that the Miner .pdf article had an apparent date of mid-2005 on it in buslist’s reference, so it might have been an earlier version. I tried correcting the ‘internface’ in the reference, but that alone didn’t fix things.
The interesting thing about these trucks, if I am reading the design right in your description, is that they have none of the cross-articulation flexibility of the three-piece (or in a cruder sense, the Fox, and a more refined sense, the Taylor) trucks. The rigid frame may be to keep the bearing brasses strictly aligned and the axles parallel, with any cross-level accommodation sort of approximated by different deflection of the long leaf springs. That picture is so unusual that I have to think I’m missing something. (Note the emphasis on cross-level accommodation by the long-travel CCSBs in the O’Donnell article, and the mention of the issue that was brought up in the earlier “TSB derailment” thread about carbody stiffness influencing derailment propensity on spirals or going into and out of curves…and how to compensate effectively for it…)
Am I correct in presuming that the springs are pinned to the axleboxes and their top ‘saddles’ slide on the tender bolster? And the telescoping center pivot has no ‘play’ in it to let the H-frame do anything but rise and fall and pivot to follow curves?
I do note that the polar moment of inertia of the H-frame in rotation promises to be very small, smaller even than a weak archbar truck, if the mass of the spring is in fact not pinned so it moves with the frame. That would indicate to me that you have accurately identified the thing as an analogue of CCSBs restraining a low-polar-moment frame in yaw (and by extension, in hunting), but with a larger moment arm (instead of being out on the bolster, it is out o