This thread: http://cs.trains.com/trn/f/111/t/283377.aspx sent me down a rabbit hole on the internet, to read more about a train derailment where an out of control train went flying down a hill but not around the corner at the bottom. In one of the sources, it explains that part of the problem was that the helper unit was using dynamic brakes. When the helper went into emergancy braking, it automatically turned off the dynamic brakes and things went downhill from there- literally. The article went on to say that SP later changed things so that going into emergency braking didn't automatically turn off the dynamic braking. Huh? Why was that ever allowed to begin with?
Thanks to Chris / CopCarSS for my avatar.
Succinctly, because leaving dynamics on while applying brakes slides the very expensive cast wheels on at least some axles on locomotives very promptly beyond repair. The whole wonderful world of blended brakes deals with this sort of situation, but ask any railroader how to make blended brakes work on a random one-pipe-braked train on real-world track in weather and you may hear interesting language (that I could not transcribe here).
Meanwhile there is the issue that getting a train with composition brake shoes over a certain speed-- about 23mph on the B&O around Sand Patch, for a famous set of cases in point -- will result in a runaway no matter what if the dynamics kick out... and if you were controlling the train with dynamics at around 20mph and they cut out for any reason you can easily accelerate past the equivalent of Vne in seconds. Which is why current rules hold you well shy of that speed when controlling the train with dynamics, and that in turn can greatly impair downhill timekeeping.
Overmod Meanwhile there is the issue that getting a train with composition brake shoes over a certain speed-- about 23mph on the B&O around Sand Patch, for a famous set of cases in point -- will result in a runaway no matter what if the dynamics kick out... and if you were controlling the train with dynamics at around 20mph and they cut out for any reason you can easily accelerate past the equivalent of Vne in seconds. Which is why current rules hold you well shy of that speed when controlling the train with dynamics, and that in turn can greatly impair downhill timekeeping.
Having read that accident report, I suspect the "23mph" Vne depends on the grade and weight per operative brake shoe - or more succintly, braking horsepower per shoe. IIRC, both Sand Patch and the Palmdale cutoff line down Cajon Pass are 2.2% grades. Vne for the same train on a 1% downgrade might be 40 to 50mph.
Overmod- Your explanation is going over my head. As I understand it, dynamic braking would slow the axles down, the train brakes would slow the wheels down? Where does a sliding wheel fit into this picture?
Murphy SidingOvermod- Your explanation is going over my head. As I understand it, dynamic braking would slow the axles down, the train brakes would slow the wheels down? Where does a sliding wheel fit into this picture?
Axles and wheels are mechanically one and the same, wheels are interference fit to the axles and they all act as a unit - If a rolling engine wheel set contains 100K pounds of roling energy and dynamic brakes apply 75K pounds of retardation on that 100K pounds and then the emergency brake application applies an additional 50K pounds of retardation to the wheel set we no have 125K pounds of retardation applied to a wheelset that only has 100K pounds of energy to be disappated - result locked wheel sliding on the rail.
Note - all numbers and measures have been pulled out of the air - I am not an engineer and am not in possession of the proper engineering terms for the actions that take place.
Never too old to have a happy childhood!
Just as with the old adage 'brakes don't stop the car' neither dynamics nor brake shoes actually slow the train -- the adhesion at the contact patch between wheel tread and railhead does that. Now if you have full dynamic, there is force going through the patch to, essentially, force the wheels around. The braking is electromagnetic but the contact patch doesn't care. Now we throw on emergency air brakes, which are somewhat carefully designed not to lock up an idling wheelset when fully applied. However they are now applying against a wheel already being braked, so the result goes more positively to stopping wheel rotation than the poor contact patch can take. It skids just like the Cadillac back wheel... and remember that on the railroad the coefficient of sliding friction can be dramatically less than static friction. If in fact the sliding friction is less than the effective braking torque from the dynamic, the wheel would continue to slide even if the air brake were released were the dynamic not dropped out.
When the automatic air brake is set, whether from a service application by the engineer or an emergency application from anywhere in the train, the locomotive's independent brake will also apply. The engineer actuates or "bails off" the independent to keep or moderate the engine brakes from applying and possibly sliding the engine's wheels.
They started equipping engines with a dynamic brake interlock valve that would automatically keep the engine brakes from applying when an automatic brake application was made while in dynamics. Even with it, engineers are still supposed to bail off the independent duting an automatic application. We're not supposed to rely on it workong. I think this feature may have made them comfortable with allowing the dynamics to remain on during an emergency application.
Now with PTC, the DBI feature is being removed. For short trains or light engine moves made in PTC territory the only brakes PTC affects are air brakes. PTC needs the braking effort from the locomotives in those cases, whether the engines are in power or dynamics.
Jeff
When using dynamic braking and an air brake application is made the locomotive engineer is supposed to bail off (release) the air brakes on the locomotive to prevent the sliding of the wheels. There are also automic systems that will reduce the dynamic braking force if the wheel do slide, whether because of an air brake application or slippery rail conditions.
When I was in engineer's school in Cumberland, Md. it seemed wrong to me to shut off dynamic braking in a run away situation. I asked the instructor and he pretty much said "Well that's the way it works". I doubt that I had any influence, but some time later Chessie set up the dynamics to stay on in an emergency.
In order to keep Lithonia... well, that was embarrassing, Murphy... from becoming still more concerned, we should remind him at this point that there are actually three systems of braking involved here: the automatic Westinghouse brake on the train, the straight locomotive 'independent' brake, and the dynamic brake or AC equivalent. The 'bailing' feature only releases the independent, which is arranged to apply together with the automatic train brake "for safety and convenience".
In theory since the independent is a straight brake, you could then modulate it (apply and 'graduated release', as if it were a passenger brake) rather than waiting for accelerated release and recharge. Figuring out how much straight air won't quite slide the first susceptible axle is not something I'd expect even the best engine man to sense on trailing units; modern blended-brake controllers have means of sensing wheelslide and adjusting the 'combined' braking effort for maximum deceleration -- in some modern locomotives, right down to a couple of mph.
This can be facilitated greatly on AC synthesized drive locomotives, where the induction motor is better 'motored in reverse' via the rotating fields to produce braking. As with an aircraft with beta thrust or reversers this can work right down to zero road speed without losing efficiency, whereas traditional DC dynamic drops off substantially at low speed. (I will let Jeff and others who have used them take up the implementation of extended-range dynamic braking...)
It is possible to recover the dynamic brake after an emergency application on engines that weren't equipped to keep them working. I was told this by an instructor who worked grade territory. I've never had, and don't expect to, the chance to try it out.
When the brakes go into emergency, or the engineer's valve is used, the emergency application will open up the PCS which cuts off the power or dynamics. The automatic handle needs to be placed in the emergency position, if it isn't already. It needs to be left there at least two minutes. The rules require recovering the air brake only being done after stopping. This means moving the brake valve handle to release. It's possible to recover the PCS without completely recovering the air. Instead of moving the handle to release, move it to "handle off" position. It won't release the emergency application, but it will allow the PCS to reset allowing the dynamics to work again.
So I've been told.
jeffhergert It is possible to recover the dynamic brake after an emergency application on engines that weren't equipped to keep them working. I was told this by an instructor who worked grade territory. I've never had, and don't expect to, the chance to try it out. When the brakes go into emergency, or the engineer's valve is used, the emergency application will open up the PCS which cuts off the power or dynamics. The automatic handle needs to be placed in the emergency position, if it isn't already. It needs to be left there at least two minutes. The rules require recovering the air brake only being done after stopping. This means moving the brake valve handle to release. It's possible to recover the PCS without completely recovering the air. Instead of moving the handle to release, move it to "handle off" position. It won't release the emergency application, but it will allow the PCS to reset allowing the dynamics to work again. So I've been told. Jeff
Also, I taught my students that if they have an undesired emergency application, the first thing you don't do is put the brake handle in the "emergency" position or the engines will lose either the power to run away from the rear of the train or the dynamic brake to hold the train back!
.
Big JIm is right , it's true.
Overmod In order to keep Lithonia from becoming still more concerned, we should remind him at this point that there are actually three systems of braking involved here: the automatic Westinghouse brake on the train, the straight locomotive 'independent' brake, and the dynamic brake or AC equivalent. The 'bailing' feature only releases the independent, which is arranged to apply together with the automatic train brake "for safety and convenience". In theory since the independent is a straight brake, you could then modulate it (apply and 'graduated release', as if it were a passenger brake) rather than waiting for accelerated release and recharge. Figuring out how much straight air won't quite slide the first susceptible axle is not something I'd expect even the best engine man to sense on trailing units; modern blended-brake controllers have means of sensing wheelslide and adjusting the 'combined' braking effort for maximum deceleration -- in some modern locomotives, right down to a couple of mph. This can be facilitated greatly on AC synthesized drive locomotives, where the induction motor is better 'motored in reverse' via the rotating fields to produce braking. As with an aircraft with beta thrust or reversers this can work right down to zero road speed without losing efficiency, whereas traditional DC dynamic drops off substantially at low speed. (I will let Jeff and others who have used them take up the implementation of extended-range dynamic braking...)
In order to keep Lithonia from becoming still more concerned, we should remind him at this point that there are actually three systems of braking involved here: the automatic Westinghouse brake on the train, the straight locomotive 'independent' brake, and the dynamic brake or AC equivalent. The 'bailing' feature only releases the independent, which is arranged to apply together with the automatic train brake "for safety and convenience".
Thanks all for a really excellent discussion, a model for what this Forum does best.
Murphy SidingI'll let Lithonia know, so he won't be concerned.
Overmod Murphy Siding I'll let Lithonia know, so he won't be concerned. Thanks, I appreciate it; helps me from having to stammer excuses.
Murphy Siding I'll let Lithonia know, so he won't be concerned.
Thanks, I appreciate it; helps me from having to stammer excuses.
Part of the theory of having Dynamics shut down when a Undesired Emergency Application hasppens is to have the head end of the train 'run apart' from the rear end of the train so as not to have the rear end of the train impact the rear of the head end at speed and causing a a 'high' speed impact derailment when a train breaks in two account a broken knuckle or a pulled out drawbar.
With each end of the broken in two train having several thousand tons of mass the kinetic energy of the collision will be tremendous.
A few things.
Additional factors in the San Bernardino wreck were that the weight of the train was improperly calculated, and the DB was not working on at least one of the trailing locomotives, the Engineer being unaware of this. So this particular train's 'maximum safe speed' for the grade was lower than normal, essentially it was set up to fail.
The automatic bail off while in DB feature is a customer option, CN gets it but those blue GECX ET44AC demonstrators were not so equipped when CN first started leasing them (later buying them, most have since been repainted and modified to CN specs).
Some older units automatically deactivate DB if the independent brake is applied even a bit. This will happen even if you are not in emergency.
Freight locomotive and car air brakes do not have any automatic anti-lock feature.
If you get wheelslide while in DB the computer will attempt to regulate braking effort to avoid skidding, just as it attempts to control wheelslip while in power. Newer units, especially AC's are better at this, older DC units tend to drop their load significantly or completely in both power and DB.
Greetings from Alberta
-an Articulate Malcontent
Overmod In theory since the independent is a straight brake, you could then modulate it (apply and 'graduated release', as if it were a passenger brake) rather than waiting for accelerated release and recharge. Figuring out how much straight air won't quite slide the first susceptible axle is not something I'd expect even the best engine man to sense on trailing units;
In theory since the independent is a straight brake, you could then modulate it (apply and 'graduated release', as if it were a passenger brake) rather than waiting for accelerated release and recharge. Figuring out how much straight air won't quite slide the first susceptible axle is not something I'd expect even the best engine man to sense on trailing units;
It is tough, especially if the brake pistons are not all adjusted equally. I find that more than 40-50 PSI in the brake cylinder will cause skidding at low speeds, but this is different for each individual unit and if the brakes are badly worn or out of adjustment a full 72-80 PSI independent application will barely stop the unit.
I'm speaking of our large 6-axle road power with composite brake shoes here, yard engines with double clasp brakes or older power with cast iron brake shoes will have different (lower) maximum brake cylinder pressures.
Overmod This can be facilitated greatly on AC synthesized drive locomotives, where the induction motor is better 'motored in reverse' via the rotating fields to produce braking. As with an aircraft with beta thrust or reversers this can work right down to zero road speed without losing efficiency, whereas traditional DC dynamic drops off substantially at low speed. (I will let Jeff and others who have used them take up the implementation of extended-range dynamic braking...)
Extended range DC DB works pretty good until about 5-8 mph, then starts to drop off. I've never run anything with the old 'standard range' DB but I believe it would start to lose effect somewhere in the 15-20 mph range.
AC DB is absolutely amazing, you can bring a train to a stop and hold it stopped on a gentle grade with just the DB.
BaltACD Part of the theory of having Dynamics shut down when a Undesired Emergency Application hasppens is to have the head end of the train 'run apart' from the rear end of the train so as not to have the rear end of the train impact the rear of the head end at speed and causing a a 'high' speed impact derailment when a train breaks in two account a broken knuckle or a pulled out drawbar. With each end of the broken in two train having several thousand tons of mass the kinetic energy of the collision will be tremendous.
That is counter-intuitive to the fact that you are trying to hold the train back in dynamic.
The big problem with the dynamic dropping out when the train went into emergency was that it would actually "time-out" before the train came to a stop. I can guarantee you that if the train happened to have a large consist of locomotives and/or the head end of the train was much heavier than the rear, you were going to break a knuckle before you could get enough braking effort on the units just before you stopped. In such a case, there just wasn't enough braking effort in the units to keep the head end from running out. Remember, there was a limit to how many units could have the independent brakes in operation.
I suspect we all have seen the UP Tornado derailment from a number of years ago. While I suspect the train was not braking at the time the tornado derailed the cars in the train - the video shows the cars behind the engines derailing with the head end continuing to move and slowly comes to a stop or near stop - meanwhile the momentum of the balance of the tonnage of the train keeps it moving (derailed and all) until it impacts the engines.
https://www.youtube.com/watch?v=VakBHlqXi1g
Our community is FREE to join. To participate you must either login or register for an account.