The local CSX hot box scanner provides considerable information, including track #, length of train, train speed, and number of axles. Great info.
How do these scanners work? I can definately see how the number of axles are determined, but how does the scanner determine the train speed and the length of the train? Is the train speed based on a reading based on the average speed of the train (which can change if the train is braking or accelerating)? If the speed is used for determining the length of the train, then would the change of speed also affect the length reading?
Basic physics tells me that if you know the speed of an object, and the elapsed time (during the passing), one can easily determine the length of the object. For instance, if a train is going 60 mph and it takes 70 seconds to pass, then the length would be around 6156 feet in length.
So, how do these scanners determine the speed and the length of the passing trains? What about when a train stops during the passage of the scanner? Would that reading be null and void?
I cannot give you an answer, but I can share some observations.
Where I live, I can hear three separate defect detectors for the CP (MP 37.8, 51.6, and 72.7). These detectors broadcast “temperature”, “number of axles” and “train length”, and a “safety message”, but not train speed. And as there are 14 Amtrak Hiawatha’s going by every day, I hear the detectors frequently.
I have noticed that these detectors all give a different length for the same train. This error is up to 10’ which might not sound like much, but the trains are only 6 coaches and 2 locomotives long (IIRC about 640’). This error percentage, extrapolated out for a 125 car freight train, would seem to indicate a potential error of over 100’ per train. Not so good if you’re trying to determine if your train will fit into a tight siding or a certain track in the yard.
I can’t give you a technical anwer but what I’ve seen/heard. I was waiting for a particular train to reach the HBDE (Hot Box/Dragging Equipment) detector on the CN at MP181.1. The train ahead of the one I was waiting to photograph had to stop with half of the train on the detector. The detector (I assume) didn’t see any more axles passing so it gave its report, albeit with a shorter train length than expected. Once the train began to move again and eventually cleared the detector it gave another report, again, with shorter train length. A conductor I spoke with at a later time informed me that should they stop over a detector they can add up the axles to get counts and lengths as well. I’d think that would make finding a problem slightly more complicated should one arise (especially in the 2nd part of the train). You’d have to figure out how many axles were in the “first” report and then go from there.
While there is the possibility of radar and lasers (LIDAR), I suspect that speed is determined by measuring how long it takes for a wheel to cross two points, which really don’t have to be very far apart. Trending can help deal with any variations, so once the train clears the HBD/DED, the detector can announce the speed, along with everything else.
The detectors I hear most often (CSX) generally report location, track (if more than one), defects if any, train speed, and axle count.
Every now and then I’ll hear a bogus axle count - at least I’d consider an odd number to be bogus…
If a train stops (or even slows down too much) over one of our detectors, we get an “integrity failure” message. But our detectors don’t give speed and train length, or even axle count (IIRC–it’s been a while).
Somebody who’s familiar with the CSX detectors might be able to enlighten you as to how it’s done. We have pages of instructions concerning our various detector types, and what to do in the event of certain readings. Perhaps some information about how they operate can be deduced from CSX’s instructions.
I realize the detectors aren’t perfect and that hopefully they’d catch whatever would be wrong before you see the car from the tower-say one isn’t caught? Would the crew on the ground be looking for dragging equipment prior to humping or are they looking for reporting marks for the next ‘kick’? Would the car get pulled out to a RIP track (depending on what was wrong) or something else?
The trains are inspected in the receiving yard after they arrive and before they are humped. The types of defects usually caught by the detectors aren’t the same as are likely to be caught in these inspections (save for dragging equipment). We just hump the cars to one of two bad-order tracks (one for light repairs, the other for more extensive stuff, wheel changeouts, etc.). Nowadays, we get bad orders caught by some of the wheel defect detectors–stuff that the car-knockers couldn’t catch, such as wheels out of round. These detectors can also catch shelling, but that’s something that should be caught in the inbound inspection as well.
Our hump crews should be on the lookout for brakes applied (either hand brakes or air that wasn’t completely bled off during the inspection). If there’s something more obvious (dragging chains or those horrendous metal straps) they’ll do what they can to fix them. There are times when I report what I see that is either a defect or an accident waiting to happen–such as open plug doors (we’ve talked about this one before), or damaged lading (not as common any more, but we used to see cracked windows back when auto racks were open).
One time, a few years back, we had a monster load of some sort on a heavy-duty flat car that we were putting onto its track. I noticed something that didn’t look right, and a closer look suggested that the load had been damaged by putting it somewhere that it hadn’t cleared. After I reported the damage, the car was inspected. Turns out (as I was told later) that I saved the UP a six-figure damage claim, because the damage could be traced to the delivering railroad (which I won’t name, but its reporting mark has three letters).
RWM will probably chime in with a better answer but here goes…
Shelling is the process by which portions of the wheel tread melt, displace on the tread and reharden. This causes a build-up on the wheel that makes it go kur-KLUNK, kur-KLUNK. kur-KLUNK, leading to rail damage, further wheel damage, damage to lading, derailment, etc.
Photo of ‘shelled’ car wheel, from a Transportation Safety Board of Canada report on - “Railway Investigation Report - Main-Track Derailment - Canadian National Train Q-120-31-12 - Mile 114.8, Montmagny Subdivision - Lévis, Quebec - 12 November 2004 - Report Number R04Q0047 - RAIL REPORTS - 2004 - R04Q0047” at -
"The running surface of the failed wheel exhibited shelling spots adjacent to the wheel flange, one with a diameter of 1.7 inches and the other with a diameter of 2.4 inches (seePhoto 2). The 2.4-inch shelling spot was not continuous, but the 1.7-inch shelling spot exceeded the limit prescribed by the Association of American Railroads (AAR), which requires that wheels with continuous shelling spots on the running surface measuring 1 inc
Thanks, Paul, I did not think of braking heat. [D)] I can see the heat of braking being held because of the hot brake shoe being held against the hot wheel tread.
HBDs detect speed of the train using the same technology that a grade-crossing predictor (GCP), aka constant-warning time device, does for a grade-crossing approach circuit. It works like this. The GCP places a frequency onto one of the rails at the grade crossing. The frequency travels outward in both directions a measured distance (that depends upon the maximum possible speed of the approaching train. At the end of that distance a “narrow-band shunt” connects the two rails. The shunt consists of a wire that passes through a filter which allows an electrical circuit through the device only within a narrow band of frequencies that happens to correspond to the frequency that the GCP is putting out. This allows the shunt to reject the d.c. track circuit voltage, electrocode, etc., that might be present. The freq
How important is it for the crew to have length of train info, when their wheel report should also state that info. Is it just another precaution? The train length, while critical for importance of operations, is something of a head scratcher for me. Train speed - sure, not bad to check vs internal instrumentation.
BTW, CSX HBD activated this evening…first time I have heard it. Interestingly, the crew did not report to the dispatcher and continued moving. It was an unusual symbol W003, with only 221 feet of train (operating at 56mph).
Much like the coal power generating station thread, this is one which needs to be read a couple of times…appreciate the explanation.
I’m certainly no expert on detector technology. But one important development in this area which hasn’t been mentioned yet in this thread is the “networking” of detectors. Historically, each detector operated as an isolated island. Today, however, detectors may be linked to a cetnral office. This allows a railroad’s computers to see not only the condition of a train as a passes a single detector, but changes from detector to detector. That can be important. For example, if the temperaturet of a bearing has increased significant from one detector to the next, that can be a warning of an imminent failure even if the 2nd detector’s reading, by itself, wouldn’t be of concern. Another important development in this area has been the deplyment of detectors that sense things beyond the traditional hot bearing or dragging equipment conditions. For example, there are detectors that actually listen to bearings and can recognize the accoustic signature of a failing bearing. There are also detectors that measure wheel impacts. It’s a fascinating, and still developing, technology.
How important is length! From a dispatcher’s and a conductor’s point of view, a whole lot more important than speed …
With apologies to the genius of Harry J. and Rita Fink, screenwriters:
“I know what you’re thinking. “Does this train have six thousand feet or only five?” Well, to tell you the truth, in all this excitement after those last three pickups and set-outs and the lousy paperwork from customer service, I kind of lost track myself. But being as this next siding will only fit five, and we’ll stab three Z trains and a business-car special if it doesn’t fit, you’ve got to ask yourself one question: Do I feel lucky? Well, do ya, punk?”
RWM: Around here we have a HBD that is located inside the island circuit but not the actual crossing of the track over a street (HBD 200ft north of the crossing gate).(Crossing is a constant time circuit) Also the signal and switch for a TCS controlled siding is 400 ft south of the same crossing. Is this hard to engineer or not?
Not hard to engineer, at least after someone else has already done it and you can pull the sheets out of the drawer and see how they did it. Just a lot of extra work to engineer, and a lot of details to get right in the circuit drawings, checks, instrument house wiring, field work, construction, cut-over, and maintenance.
As far as hotbox and dragging equipment detectors are concerned, knowing train length may not be much of an issue. The hotbox and dragging equipment detectors with which I am familiar identify the location of each problem they find by stating how many axles it is from the head end.
Some installations require that the train pass with a minimum speed in order for the hotbox detection to be effective. Other, more sophisticated ones do not have this requirement and will even allow a train to stop on the detector without reporting an integrity failure.
Sometimes a detector will broadcast a false reading possibly due to a heavy brake application that the system misinterprets as either a hot axle or a hot wheel. After the train is stopped and the conductor inspects the problem(s), if no exceptions are taken, rules require that a further inspection must be made of the 5-cars (or 20-axles) either side of the alleged defect on both sides of the train.
On some railroads if two consecutive scanners detect the same problem with the same mechanical reefer, the railroad requires it to be setout - no matter what!
