RWM:
Got it. Never took into consideration the pickup/setout aspect of the industry. The CSX line in question doesnt seem to do any block swapping or industrial set outs on this line, so that never entered my mind.
Ed
Falcon:
That is great stuff. How wide spread are these new generation of detectors?
Are each trains identification coded, forgive my basic non technical discussion, but similar to bar coding to account for each train’s reading as it is read by subsequent scanners? Otherwise, how would the system account for trains overtaking other trains, etc. (You gotta realize that i am pretty slow on the technical side here).
This sure is quite a jump from the old days at Porter Tower, when the HBD east of town would print out on a graph the temps of the bearings and the operator would visually interprete the info.
What is next? How far is the industry from having on line sensors on the bearings. Retailers are doing something similar to this determining the distribution of product into geographic areas. Granted the technology for reading temperature and transmitting info is considerably more advanced…but is that on the drawing board, or rather the screen at this time?
Ed
Ed, the “change” is 99% about organizational methods and IT support, rather than field-based. It’s all in how the railroad deals with information flowing from the field into a central location. You can let the information flow by you without collecting it or organizing it and just look for the outliers, or you can do something with it and use it to discern more subtle trends and relationships.
RWM
HBD’s have a heck of a time with steam excursions…
I was under the impression that detectors measured the defect from the last axle, not the first.
As for “bar coding” the trains - ACI died a long time ago (those red, white, blue, and black bar codes on the cars and locomotives), but I’m thinking that most power, and definitely most, if not all cars, have RFI (radio frequency identification) tags that can be read by a trackside sensor.
It’s not a reach to tie a detector report to a lead locomotive, assuming the infrastructure is there to transmit both to some receiving entity. It might even be possible to tie such a report to a specific car in the train, which would certainly simply locating the problem.
In addition to hotboxes and dragging equipment, I seem to recall that there are some “wheel impact detectors,” whose function it is to find flat wheels.
WILD or Wheel Impact Load Detectors as well as High/Wide load (Clearance) Detectors. Are there any other types than these? HB, DE, WILD, Clearance.
Ed, your mention of the graphic printout at Porter reminded me of a couple of things: I remember seeing one in the C&O dispatcher’s office in Grand Rapids, and watched it chart a train passing over the detector site (near Hartford, Michigan, if I remember correctly–this would have been the only detector at that time, in the late 1960s or so).
And the CNW’s detectors had to be similar. There was a special signal associated with them–lunar white lights fitted into a PRR-style position-light signal head. The signal would display three horizontal lights to stop the trains at a designated point (clear of crossings, etc.) unless the dispatcher had inspected the printout and cleared your train (it would then show three vertical lights). If he stopped you, he would tell you how many axles (or cars–they weren’t hard to figure from the printout) from one end or the other the blip showed.
Larry, some detectors did tell you how many axles from the rear. Those were the type that had a readout (three-digit axle count) that was visible to crews in the caboose. The probably didn’t react at all until a defect was noted, so with an unknown number of axles having gone by ahead of the defect, they’d count the axles that were left.
All cars have AEI tags now, Larry, you’re correct. Don’t know why they don’t give you the associated car if there is a problem–except that I’ve heard variances between detectors recording the axle count on the same train, or a train going through a detector with an odd number of axles (uh-uh!), so errors could creep in.
And yes, I also mentioned the wheel impact detector, sort of. Not only would they detect flat wheels, but also those that are out of round, for whatever reason. Flat spots can be detected by visual inspection, bit an axle that is no longer centered would be a tad trickier. That would, however, produce an undesired impact on the rails.
And as for steam locomotives, I still remember hearing when UP 3985 hitting the detector between Winfield and
That’s the IT function I referred to earlier – the information from the HBD associates with the OS information from the dispatching computer and the AEI information from the AEI reader. You don’t have to do anything out in the field to make this happen other than put a data radio onto the HBD that transmits its output into a file that can be stored on a server at the central office - and most of them have this already. The hard part is taking all of the data flows from all the different sources in the central office and properly and accurately associating them with each other on an automatic basis, and then building history files for each car, and each truck and each wheelset on each car.
RWM
For example - from the NS Northern Region - Pittsburgh Division Employee Timetable No. 1 In Effect August 4, 2008 - GENERAL INFORMATION - B. EXPLANATION OF CHARACTERS on page 1 [‘’(5 of 158)‘’ of the ‘PDF’ version at - http://blet73.org/Pittsburgh_Division_Timetable.pdf ]:
Train Inspection Detectors:
DED — Dragging Equipment Detector
HCD — High Car Detector (includes Excessive Height Detectors)
HBD — Hot Box Detector (includes TSA, SAD and HBD detectors)
WCD — Wide Car Detector
SSD — Stress State Detector
SWD — Sliding Wheel Detector
WID — Wheel Impact Detector
All train inspection detectors are listed on the station pages according to milepost location. Unless otherwise indicated, train inspection detectors are Radio Alarm and operate in both directions on single or multiple track. Detectors on Single Track — Track will not be shown.
Note that the 2 that I have ‘bolded’ are additional types - also, there are 2 different kinds of clearance detectors - High Car and Wide Car. There may well be others . . . [:-^]
The 'SSD — Stress State Detector ’ is apparentl
Most of the hot box detectors across Iowa on the UP only give temperature when giving the no defects or integrity faiure. Some don’t even give that. Some you can slow down or even stop on, but most if your speed drops under 10 mph you can almost expect an integrity failure. (Once the train I was on got 3 no defect messages on the same detector)
Most of the new wayside signals that were installed when the CTC was put in have dragging equipment detectors that only talk on defect. The one that hit us today gave a total axle count. (The defect was near axle one. While bucking snow drifts the air hose came lose and struck the detector and immediately gave a defect tone.)
The detectors on the Spine do give axle counts.
The detectors across Nebraska only talk on defect. Even then sometimes they don’t give a tone or talk when they find a defect. They are tied into the block system and will hold a signal against the train that the dispatcher must manually clear.
The UP has what they call the Hot Bearing Desk. When they call they generally give a car’s initial and number and which axle to check.
The last time I was stopped by a hot box detector, the heat generated was from a tie-down chain on a flat that had come loose and wrapped itself around the axle.
Jeff
Would the purpose of stating the temperature be for operating the train due to extreme high and low temperatures? Such as below 10 degrees (due to air brake restrictions) and above 90 degrees due to possible rail integrity issues?
Ed
I don’t know how close to any drawing board, but one of those things the RRs would like to do. The hurdles were always, “how do you power the system?” and “how do you transmit the information?”. These hurdles get flattened when ECP braking is implemented. There may be enough “smarts” in the ECP braking processor to handle other tasks w/o adding anything more than the sensors. There are other freight car health monitors that would be useful as well: Wheel impact detection, stuck truck detection and truck hunting detection, off the top of my head.
The term “truck hunting” has appeared a couple of times.
Definition please.
ed
The St. Louis paper today reported a freight train received a dragging equipment defect and stopped. They found a corpse tangled in the couplers between cars. They believe he was snared while trying to climb through the train when it was stopped. He was dragged for six miles.
One of our supervisors seriously suggested that we could train dogs to snoop out hot boxes with their noses. I suppose the dog could give the axle count by stamping his paw on the ground.
Although we don’t have any yet, the latest thing is the audio detectors, that listen for the “growling” of failing wheel bearings.
Nick
From the Salient Systems webpage for its Hunting Truck Detector at: http://www.salientsystems.com/prod_hunting.html [emphasis added - PDN]
Hunting trucks underneath rail cars can violently oscillate from one rail to the other as they traverse along tangent track, inducing excessive lateral forces that significantly contribute to the rapid wear of rail and rail cars in a relatively short time. This particular type of degraded vehicle performance is a leading cause of damage to delicate lading. At a minimum, hunting trucks cause increased fuel consumption. Unchecked, severe damage to truck components and derailment can result.
There’s something like 1’’ or 1-1/2’’ of ‘play’ / tolerance/ ‘slop’ between even a properly gauged wheelset*, and the nominal tangent track gauge of 4’-8-1/2’‘, as measured at the gauge line at 5/8’’ below the tops of the rails. While that works great to provide flexibility in curves to keep the trucks from binding, on tangent track it can be too much of a good thing. That’s one reason the PRR went to a slightly narrower ‘standard’ gauge of 4’-8-1/4’’ on tangents for a while . . .
- Paul North.
*Can’t remember exactly what that dimension should be. If I don’t find it, someone please post it. Thanks.
EDIT - See also the lengthy and learned discussion in the Wikipedia article on this, at -
Ideas I’ve heard about in the past (don’t know if they were ever given serious trials):
Treating one of the bolts at the end of each bearing so that it would show a distinctively different color when the bearing overheated.
Making one of the bearing bolts spring-loaded so that a pup-up temperature sensor (think Butterball here) would close the circuit on a battery-powered device that would transmit an audio and/or radio-frequency signal (radio-frequency to notify the crew, audio for an on-ground inspection once they stopped).
Perhaps these ideas were abandoned when it was realized that these bolts are actually needed to hold the bearing together!
Well…if the bolt isn’t holding the bearing together…that’s one way to find the offending car/bearing…[:-^]
Given the advances in RFI, taking the ideas Carl mentions a step or two further might involve a small RFI transponder on each wheel. The transponder would include a sensor for heat, which would set an alarm condition on that device. Next time a trackside monitor interrogated the transponder it would show the alarm, which could then be handled much as they are now.
The downsides are the sheer number of transponders needed and the fact that it would not give real-time status - only when the car passed a sensor.
IIRC, roller bearings do/did include a cartridge that would go off when an overheat condition occurred. I’m pretty sure it stunk (as did friction bearing hot boxes), and may have smoked. Of course, now there’s no one in a caboose to smell it or see the smoke…
In the January-February 1962 Rock Island employee’s magazine is an obituary for “Queenie.” She was a collie owned by an assistant trainmaster at St. Joseph, Mo that had learned to sniff out hotboxes.
They even had a picture of her at work, walking along side a cut of cars.
Jeff