Ryan
Court Jester
Yeah, it's pretty much impossible for the train to stop quickly enough to "launch" anyone*.
* (unless it hits something or derails)
* (unless it hits something or derails)
Another Grade Crossing Fatality
- Thread starter Ryan
- Start date
Help Support Amtrak Unlimited Discussion Forum:
- Status
If 40 mph (~ 60 fps), 10 85' superliner cars, stopping in 850 ft (its length): that's approx. -0.085g... a little less than the type of force one would experience in a car braking heavily; so, nobodies going to get physically launched, but if they're standing and not braced, will probably end up on their butts. ... I would guess more in the 30, sub-30mph range.Since you are the one that thinks they'd get launched, what is your guess?
(null)
But likewise I'd defer to the train buffs here: given a 500-600 ton train, and only small friction/contact patches on the rails, I would expect lots of sparks and noise with that type of deceleration, if it were even possible.... but begin human, could very well be wrong.
Time for the experienced to chime in here...
[of note: in a quick web search: seems light-rail type cars can decelerate at 0.15g in full-on emergency; passenger trains with passenger type loads, around 0.03g - would would mean a closer to a sub-30 mph number... but again: time for those with multi-thousand posts and millions miles of travel experience to chime in: reality always trumps google.]
My rough calculation is for contact velocity, not initial when engineer sees car on tracks and tries to stop the train velocity... suspect there's a couple train's lengths there before contact.
Last edited by a moderator:
Ryan
Court Jester
Those numbers look somewhat reasonable.
You haven't met Jishnu yet, but he's one of the uber-experts.
Just using your numbers, you can get the .085G down to .03G by tripling the stopping distance - a.k.a the engineer dumps the air 2 train lengths before impact, which sounds reasonable to me.
You haven't met Jishnu yet, but he's one of the uber-experts.
Just using your numbers, you can get the .085G down to .03G by tripling the stopping distance - a.k.a the engineer dumps the air 2 train lengths before impact, which sounds reasonable to me.
Sounds almost like closure then...
though one has to remember: in most 30 mph collisions, most of the energy gets turned to heat very quickly, all parties come to a standstill and it's over quickly (1/10th sec) - in the case of train vs car: suspect the car slows the train down very little, ie, the 40-45 fps train velocity continues right through the car, applying that "force" to everything in its path - in one side and out the other: 1 ton car vs 500 ton train - outcome is painfully terribly predicable.
though one has to remember: in most 30 mph collisions, most of the energy gets turned to heat very quickly, all parties come to a standstill and it's over quickly (1/10th sec) - in the case of train vs car: suspect the car slows the train down very little, ie, the 40-45 fps train velocity continues right through the car, applying that "force" to everything in its path - in one side and out the other: 1 ton car vs 500 ton train - outcome is painfully terribly predicable.
This video is very instructive on many levels. Trains looks to be moving at least 50 or 60mph when it hits the auto, yet by the time the last car exits the video frame, it has slowed considerably, maybe 25mph or less... all in just the length of the train.
Last edited by a moderator:
One can use a superliner car, 85 ft, as a yardstick: 1 car length per second = 58 mph; 1 car length in 2 seconds = 29 mph.... time to get the stopwatches out
I see 2 P-40s, a baggage car and 8 sl cars. P-40 is 69', baggage car about same (can't find anything on web for it), and SL cars at 85'. After playing about 20 times and trying to single frame it and use clock in upper left: initial speed 55mph, last car speed 40mph, ie, 15mph deceleration in 900 ft... again same 0.03g number. Maybe/hopefully others have better players and can interpolate the clock and car boundaries better (and surely have younger eyes).
Last edited by a moderator:
NW cannonball
Conductor
Consider - Amtrak trains have better brakes, higher working brake-pipe pressure, better couplings. Can stop (relatively) fast -- the braking formulas and track adhesion and all that been known for many decades.
Get in front of a freight train - (you 100 kg - SUV 2000kg - Amtrak train a few thousand tons - freight train (with worse brakes) maybe 10 thousand tons -- aircraft carrier may 100 thousand tons) -- you just go splat.
So why don't people get it?
Everybody knows that, if you collide at top speed with an NFL linebacker who outweighs you by 50% -- you will be dumped on your butt.
Every pedestrian knows that you do not step in front of a speeding SUV or a bus, and on the road you do not cut off a 50-ton rig (not more than once)
Maybe there is some limit to how threatening an artifact might feel? Maybe 1000 times just doesn't register?
Better crossing gates -- sure.
Don't know what more to say.
Get in front of a freight train - (you 100 kg - SUV 2000kg - Amtrak train a few thousand tons - freight train (with worse brakes) maybe 10 thousand tons -- aircraft carrier may 100 thousand tons) -- you just go splat.
So why don't people get it?
Everybody knows that, if you collide at top speed with an NFL linebacker who outweighs you by 50% -- you will be dumped on your butt.
Every pedestrian knows that you do not step in front of a speeding SUV or a bus, and on the road you do not cut off a 50-ton rig (not more than once)
Maybe there is some limit to how threatening an artifact might feel? Maybe 1000 times just doesn't register?
Better crossing gates -- sure.
Don't know what more to say.
Snarky answer: in competition for Darwin Award.
More real answer: not appreciating 100 fps trains... by the time you see it, AT MOST you have 5 seconds to not be in its path, or you will die. This sad lady that died in this crash: the crossing gates came down on her car - I suspect she thought the light they were waiting for would change before the train arrived, instead of immediately getting out of the car and running for her/their lives.
I assume an engineer is governed by the FRA regs in terms of when they blow the horn the first time... is that also when the crossing guards first sense the train? 500 ft out? 500 ft at 100 fps = 5 seconds - not a lot of time to ponder and wonder what to do, ie, hear horn, and in wrong place, time for action is instantly now.
VentureForth
Engineer
How would the engineer know to start stopping? The only way is if the car was stopped/stuck on the tracks. If that's the cast, the likelyhood of a fatality would be much less. If the car went around the crossing, or got stuck at the last second, the loco engineer would have no preminition to start slowing down before impact.
In addition to the crew and victim, I also feel really bad for the car(s) stopped behind her who had to witness this.
I cross an Amtrak crossing in my neighborhood on my way to work every morning. About a half-block beyond the crossing is a traffic light. I'm pleased to say that I have never, ever seen anyone stop on the tracks in the 3+ years I've been living here. Everyone leaves space, even when the light turns green and everyone starts moving. I don't cross until the vehicle ahead of the tracks has moved forward enough for me to have room to get over the tracks without stopping, and no one has ever honked.
If I did see someone stop on the tracks, I would be incredibly worried until they were clear.
I cross an Amtrak crossing in my neighborhood on my way to work every morning. About a half-block beyond the crossing is a traffic light. I'm pleased to say that I have never, ever seen anyone stop on the tracks in the 3+ years I've been living here. Everyone leaves space, even when the light turns green and everyone starts moving. I don't cross until the vehicle ahead of the tracks has moved forward enough for me to have room to get over the tracks without stopping, and no one has ever honked.
If I did see someone stop on the tracks, I would be incredibly worried until they were clear.
In this case, my impression is that there was a line of cars waiting for a light and the line backed up across the tracks - so the engineer had at least a chance of seeing something blocking the tracks ahead... but if he was doing something else in the cab and didn't have his eyes glued to the forward window, then those precious seconds would have been lost - don't believe an engineer can keep his eyes out the forward window 100.00% of the time, and do his job... could be wrong.... actually might be an opportunity for computer assistance: a forward camera and software that recognizes the forward rails [not hard to id in an image] and look for obstructions - might have to be disabled in snow covered rails, but for 90%+ of the time, the rails should be visible and obvious [even at night given the front lights on the loco].
We're on the main UP line here in RDD and have roads that follow and cross them. Too many times I've seen people waiting on the tracks, luckily not waiting for a light, but simply a stop before making a right turn [south Bonnyview at 273 comes immediately to mind]... and I always cringe... wonder if the pause before the tracks and wasting the extra 2 seconds isn't worth it, ie, per Sorcha - don't cross the tracks until there is a spot on the other side.In addition to the crew and victim, I also feel really bad for the car(s) stopped behind her who had to witness this.
I cross an Amtrak crossing in my neighborhood on my way to work every morning. About a half-block beyond the crossing is a traffic light. I'm pleased to say that I have never, ever seen anyone stop on the tracks in the 3+ years I've been living here. Everyone leaves space, even when the light turns green and everyone starts moving. I don't cross until the vehicle ahead of the tracks has moved forward enough for me to have room to get over the tracks without stopping, and no one has ever honked.
If I did see someone stop on the tracks, I would be incredibly worried until they were clear.
Very very true... but being hit at 60 mph vs 30, or 20, or 10, might make a significant difference in terms of outcome...
and if one had a reliable camera early warning system, per the math above, given a 1/2 mile (+/-) of warning one could actually have the train stopped.
[though one has to wonder about the safety of such, ie, I understand there is only one train per section, but if a heavy freight is one section behind, if it would be able to deal with a stopped passenger train, ie, the passenger train doesn't hit the car and the one driver, but then 40 people on the passenger train get killed because a freight overtakes it from behind... sounds like there needs to be more than an engineer warning system, but a system wide warning.]
If that's directed at me: not in the least... just suggesting a methodology whereby harm is minimized.
VentureForth
Engineer
Regardless, the fact that the last car of the train is next to the wreckage indicates that the train hit the car at a relatively slow speed, regardless whether he braked 1/4 mile before impact just before impact, or at impact, the impact speed was relatively slow.
And it was enough to be fatal, and enough to be tragic.
And it was enough to be fatal, and enough to be tragic.
Ryan
Court Jester
I'm pretty sure we're not.
Not sure how you can come to that conclusion.
Not sure how you can come to that conclusion.
I kind of thought the same thing. It seemed to be directed at gmushial's hypotheses about early warning systems, the physics of it, etc. He was talking about minimizing harm, like putting back-up cameras on cars, but I can see how it appears that he's saying the engineer should have done more.
Mumble... going back and rereading my words: yes I can see how someone might interpret them that way, unfortunately. If communications is the art of the sender having the receiver hear the intended message - looks like I did a less than perfect job ;( But, no, I don't fault the engineer at all, zero, zilch. I assume that the engineers job is to when possible, to keep an eye out the front window, but, there are other things that he has to be doing, that are very much part of his job, that can keep him from being there 110% of the time. I guess I operate on the theory that his job primarily is to operate the engine; and that it is the job of the track layout engineers to give him a safe place to run his train. Given the braking capability of a passenger train, one simply can't have a system that is absolutely reliant on an engineer making sure that all 10,000 crossings that he has to traverse are clear, each and every time, etc - if that were case, then trains would take weeks to get across country. But on the otherhand, is it his responsibility to add that extra margin of safety when possible: absolutely. My comments w/re a automated obstruction checker, which is no different from having a second set of eyeballs overseeing a delicate operation - they're not absolutely required, but in general the outcome will be better with them. Even in my software world, on difficult problems, though in theory I should be able to solve/resolve all problems by myself, sometimes I call in another person to look over my shoulder to see if they see something that I'm overlooking.
Having said such: the more I think about an automatic obstruction checker, the more I like the idea... not so much just for saving lives because of people sometimes do stupid, really stupid, things... but also as a cost savings to the RR companies: I suspect there are lots of things that end up on the tracks, which do damage to the loco, that potentially cause a train to be late (no, I'm sure that would never happen ;-) ), and cost monies to repair. A couple hundred dollar system, that just might save a life or two, but probably more likely, save many times that in repair bills and lost time.
Hopefully I'm more successful in my attempt to communicate this time ;-)
That's the painful part of the internet - it's hard to convey meaning and tone sometimes, even if someone is an accomplished writer. It's easy to misinterpret things. Sometimes I wonder if I use too many smileys, but I'd rather do that than take a chance I'm coming off as mean or sarcastic or bored.
Sometimes I wonder if I use too many smileys, but I'd rather do that than take a chance I'm coming off as mean or sarcastic or bored.A scanning algorithm for the video camera to check for visual obstructions would be a major technical challenge, but in the age of self-driving prototype cars, it might be feasible. However, unless the camera system also includes a night vision or IR camera, it would be limited to bright daytime use. Of course, it would also be of limited use on curves or obstructed views or in complex locations with multiple roads over or near the tracks. The false alarm rate would have to be very, very low.
The better solution is to upgrade the crossing with vehicle detection sensors that detect when a vehicle or something large is sitting on the tracks when the gates go down. There are such sensor systems which are being installed on the Chicago to St. Louis and Michigan 110 mph corridors. But they cost money to maintain, likely occasionally generate false alarms, and will not be fool proof. There has been considerable research and efforts put into reducing grade crossing accidents which has succeeded in reducing the annual number of fatal collisions. You may want to check the FRA website for technical reports. Or dig up the NC DOT report on grade crossing design safety experiments and trials they did some year back and was in the NC DOT website as I recall.
That is, unless the irresistable force (train) hits an immovable object (another stopped train, or brick wall, or ...?). Then, the passengers on the train might be launched. I suppose passengers might be launched, without striking said immovable object, if the rate of deceleration approached the gravitation force of the earth (9.8 meters per second per second), but that can't be achieved with existing braking equipment on Amtrak passenger cars.
Ryan
Court Jester
Yeah, I covered that.
- Status
