Since Dave Klepper has now made details of his approach for a ‘safe oil train’ fully disclosed for any patentable purpose, perhaps it’s now appropriate to take up the discussion on a public forum.
Dave argues that it makes sense to ‘scale up’ an electromagnetic track brake, as used in transit applications, and apply it to individual cars of freight consists. I will leave it to him to describe his approach in more detail. It would be applicable to a wider range of applications than just for oil trains, although the politics associated with that particular application might drive the funding and effort needed to produce a workable system in minimum time.
A great potential advantage of the electromagnetic brake is that it works entirely in parallel with ‘normal’ tread air brakes, and at least theoretically can be operated with cost-effective energy storage (in one version, supercapacitors) via a distributed wireless control system. Discussions of the physics (in the ‘Concept for a Safe Oil Train’ thread) have been interesting.
I see more potential for this system than just as a form of emergency brake. If the application of the electromagnetic brake system can be modulated, much of the need for ECP air braking can be reduced. In ‘runaway’ situations where train speed exceeds braking ability, the track brake may serve effectively as a means of reducing train speed (and energy) to where more conventional methods of braking can be re-established. The system might also, in some cases, assist in keeping derailed cars aligned with the track, and prevent some tendency for cars to jackknife following a derailment.
On the other hand, the system must be designed and implemented in a way that does not produce more problems than it purports to solve, or introduce new failure modes that would constitute grounds for new liability. The fact that no electromagnetic brake system has been actively developed for mainlin
Even on interurban and street railway and ligh railway systems, operators are encouraged to use the magnetic track brake only for emergency stops. In all cases where it has been used, includidng violation of this specific instruction, I know of no damage to the equipment or problems with the brake. However, repeated use in regular service (fast braking gave more leeway on operators making schedules, perhaps while extending their coffee brakes at the end of the line) on one Brooklyn streetcar line lead to corregated track at regular stops, which then required rail-grinding. I suggest a test train with magentic track brakes as close as possible to what is in use on light rail cars, modified only to the extent required to mout them securely iinside standard freightcar truck sideframes, and the train equipped with the other important ideas for uniform braking controlled elecdtrically, derailment detection from adapted whieel-slip control technolgy, monitoring by the entineer with overeride control. Of course any break-in-to should produce he same braking reaction as a derailment detecdtion. When the train is tested with simulated dertailmlents, we will learn whether the track brake in its present form contributes to faster stops, is free from problems, etc., and is capable of being scaled up without damaginng equipmenet. A prototype car and pair of cars is the very first step, after detainled engineering.
And comparison of wheel rotation is the most effective, quickest, and easiest to implememt way to recognize a derailment, because the equipment to do so is avialable, and can also charge the batteries (or supercaps). If a car is upended and wheels are spinning freely, the derailment was noticed considerably ealrier by the break-in-two. The air hose and electrical connection to that car were broken earlier, and almost certainly, before ovrturning wheels one side of one or both of its trucks hit the ground, at different times.
Once you have a power source and some smarts on the cars, you can do lots of things! Just doing braking proportional to load w/wheelslide, you should be able to get stops in the area of .25-30gs, which is uncomfortably fast. If you add on an eddy current or friction track brake, you could really tie things down in a hurry.
I think the improvement to “smart” brakes w/o any track brake would be pretty impressive - not to mention all the other safety improvements you could get with a “smart” freight car:
Full time bearing monitoring
“Stuck truck” detection
broken spring detection
hot/broken/flat wheel detection
near-real time track defect detection
Higher “restricting” speeds
lading monitoring
truck hunting detection and correction
I think the timeline on universal application of stuff like this is a couple decades…
I agree that this will take some time. I am not sure universal application of ECP brakes will ever happen. However, a radical improvement for specialized application to oil trains may be possible to begin within a couple years. In the meantime, it would not hurt to tell the Senators about at the hearing in a couple weeks from now.
I agree, but I want to see the simpler technology that I propose installed in a two or three year time frame. It can be a step to the kind of smart car you envision, especially once we have communication throughout the train and a power source.
I’d be concerned about the increased concentration of braking force into the track structure. For example, presently a full-legal load 263,000 lb. tank car about 65 ft. long has an average weight (vertical) of about 4,000 lbs. per foot. At full emergency braking today, when the wheels slide the rate of deceleration (same as the coefficient of friction per Don Oltmann above) would transmit a horizontal braking force of from 1,000 to 1,200 lbs. (horizontal) per foot into the track structure. In comparison, most Light Rail Vehicles are considerably - well, lighter, on the order of only 1,000 -1,100 lbs. per foot of length.* Thus, even a very high LRV braking rate of 0.5 g (16.1 ft./ sec.^^2, or about 11 MPH per second - same as 0 to 60 in 5.5 seconds, only the other way) would still produce only about half as much horizontal load on the track structure as today’s maximum conventional freight train braking.
I am concerned that to increase the latter substantially with a magnetic track brake could lead to damage of the track structure, anything from tearing fastenings (rail clips, etc.) loose to causing the track to slide and buckle, and thereby causing a derailment anyway !
The horizontal load on the track structure has me a bit spooked as well, which is why I picked 0.25g for a target retardation rate for a hypothetical magnetic track brake for freight cars. It would seem natural to assume that 25% coefficient of friction between steel wheel and steel rail would also apply to the steel rail and tie plate. Going much above that would be placing a lot of trust in the track hardware’s ability to hold the rail, and there’s reason to doubt if the rail could handle 0.25g braking by heavy freights.
OTOH, the braking rate can’t be too low as 0.25g results in a 500’ stopping distance from 60 MPH - that rate would bring a train to a stop before more than 10 cars got piled up - assuming the track held.
The lateral overturning moment on the rail was one of the principal things I was watching, back in those days before Pandrol clips, concrete ties, and very long span CWR became accepted practice. Much of the vertical moment is overcome by the near-adjacency of the underailed wheels of the train; and to an extent the inertia of the rail; while the deflection (upward) may be substantial, it ought to be well within the elastic limit of the rail even if the vertical section isn’t as amenable to bending as the horizontal.
Now, with Pandrol clips, I might have to wonder about whether the rail might tend to arch under the train, and just how far the whole works might rise up out of the ballast if the clips hold. Once the clamping force on the rail is neutralized, or Lord help us the ties come out of the ballast, it’s Katie bar the door resisting the longitudinal moment.
The longitudinal force causing the rail to move is, of course, substantial, and I feel reasonably assured that beyond some (perhaps calculable) braking rate the amount of creep will become nonlinear. I started by looking at better rail anchoring, including how the anchors were faced against the ties and how far down on the ties they extended (to prevent the surge moment from overturning the ties rather than moving them in plane. I no longer have the calculations I made of the resistance between anchor and rail, but it was as I recall sufficiently high that slippage there wasn’t a limiting factor.
There’s a tradeoff between the desire to have very rapid deceleration and the need to rebuild the track geometry following an ‘anomalous event’ (e.g. crash braking in 10 seconds with the railhead perhaps nearly glowing with the generated heat…) I was proceeding on the assumption that the former was an operative condition (driven in part by plaintiff’s bar o
Mgnetic track brakes have been in use on interurban and street railway and light rail lines for over a century with none of the specific track problems that seem to worry these posters. Only problem has been repeated use as a service brake at car stops causing corrigation solved by grinding. I am not at the prresent time proposing increasing the force of the brakes but simply to expereimentally determine whether the existing brakes make a useful contribution. From that we can discard, keep, or scale upward as experiments determine. The track structure of most interurban lines was far lighter that just about any frieght rail line. I am repeating a previous answer. This has been discussed already.
And when large amounts of Bakken crude and HAL traffic is routed over interurban, street, and light rail lines, I’m sure such an observation will become relevant.
I suppose you could have a separate air reservoir on each car that would spin a generator to engage track brake. Energy is integral PdV between the high and low pressure?
This is part of what was covered in the Euclid ‘safe oil train’ thread, and I encourage Dave to synopsize all that here, or copy and paste from the posts in that other thread. Erikem went to some pains to provide numbers to substantiate how the brake could work as Dave designed it.
Basically, the alternators charge up supercapacitors (over a comparatively long period of time; presumably this would be varied with demand so that minimum fuel use would be involved to achieve the charge). Other methods of energy harvesting are of course possible, but the high current at relatively low voltage that best characterizes supercaps producing magnetism is best provided via an electromagnetic ‘transducer’ like an alternator (or the field-modulated permag generator I was discussing). I am of the opinion that some additional battery or ‘wire’ charging is highly desirable, for a number of more or less expedient operating reasons, but that is a detail-design matter and not critical to Dave’s explanation of how his system works.
Dave, I think, has a small expert-system computer trigger the system via analysis of the differential rotation of his axle-mounted alternators. There will have to be special cases to handle wheelslide, flats, etc., and I remain unconvinced that the actual change in rotation during the early stages of a derailment will actually be proportionate to flange depth (ever see what a derailed wheel does to wood ties?), but the dependence on differential rotation is not a major concern if it proves unworkable. There are other combinations of inputs that when detected trigger one type of alarm, and if analyzed may trigger a different type. Remains to be determined what constitutes an alarm serious enough to … we ne
Others have already mentioned the possibility of the lateral stresses on the rails possibly causing the track to buckle, and this would be a particular concern on sharper curves. The same forces may also cause problems if the train should happen to be crossing a bridge, especially a truss bridge or a movable span.
The forces will be repetitive as each car passes over a particular spot while the train brakes to a stop, and repetitive stress will take advantage of any weakness it finds!
The track braking forces generated by a 40 to 60 ton Interurban or street car are a far cry from the forces that would be generated from a 100 car unit train with each car weighing in at 143 tons. One or two cars exerting braking force on the track structure is totally different than a full train doing it.
Additionally having each car equipped with electrical gear having sufficient electrical power to operate the brakes and HAZMAT in the near proximity of high powered electrical gear that when it malfunctions is able to produce sparks that could ignite product that otherwise would not have achieved ignition. Sounds like checking the contents of a gas tank with a match…
Again, I am not initially suggesting increasing the power of each track brake to compensate for the added wieght of loaded tankcar. Thus, I do not expedt the track brakes on a freight train to stop a train as quicly as on an interuban car or PCC streetcar, as quickly as an auto on dry pavement. But if the stopping distance can be cut to a half or third, the applicaton may be valuable. Since the forces are distributed along the 100 cars, again I see no problem, indeed the forces at each car should be less than those under locomotives holding speeds low on a steep downgrade using only dynamic brakes.
The concept would certainly be tested before applied in revenue service, and strains and movement in track stucture certainly will be ovserved and recorded.
And again, bridge structures on interurban lines were far lighter than those on freight railroad.
My original proposal had a battery for each car, but supercaps may be better. The battery woud be charged by the same alternators that compare wheel rotaton, and aslo by train line from the locomotive when available.
The only sparks associated with this sort of equipment was will trolley poles and trolley wire and relays and contactors, all of which are NOT associated with this sort of equipment, all solid state switching and contdrol. I agree that safeguads against false triggering of an emergency stop are necessary, and hopefully can be implemented. Note, have false break-in-two’s ever have triggered emergency stops?
I was not referring to the spelling error. I just wondered what you meant by the word “false” when referring to a break-in-two. And I wondered about your questioning whether it would cause a brake application.