Trains.com

Metro North, 6 dead

20463 views
372 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    December 2007
  • From: Southeast Michigan
  • 2,983 posts
Posted by Norm48327 on Sunday, February 8, 2015 2:27 PM

Euclid
 
We also know the public reported problems with the crossing protection system earlier the day of the crash.  That does not prove that a system malfunction caused the crash, but it is quite a coincidence. 

 

 

Are you aware that witness reports after the fact are the least reliable source of information? Each is likely to see, and tell, the story differenly leaving the authorities to determine the true facts. People see things from their own perspective. Reality can be vastly different.

Norm


  • Member since
    December 2001
  • From: Denver / La Junta
  • 10,820 posts
Posted by mudchicken on Sunday, February 8, 2015 2:28 PM

BroadwayLion

Signs will not be read. A new crtossing protection is required. The kind used to protect Federal Buildings, it comes up out of the ground and can stop a tank.

If a car is on top of it when it comes up, the car will just stay there until it goes back down again. Naturally it has to be much further away from the tracks so that if it does lift a car it will not be half on the tracks. It will need good warning lihgts to let people know to STOP AND WAIT RIGHT THERE!

Oh well, ideas of LION might not be practicle, but what the heck do you expect from a LION.

 

ROAR

 

Tried and failed. (McLean, IL on SPCSL racetrack) If man designs and builds it, man can manage to subvert it.

Mudchicken Nothing is worth taking the risk of losing a life over. Come home tonight in the same condition that you left home this morning in. Safety begins with ME.... cinscocom-west
  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Sunday, February 8, 2015 2:41 PM
Yes Norm.  That explains why you never hear of a witness being used in a trial. 
  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Sunday, February 8, 2015 3:30 PM

 

Euclid
 
We also know the public reported problems with the crossing protection system earlier the day of the crash.  That does not prove that a system malfunction caused the crash, but it is quite a coincidence. 

 

Observation from 25+ years of handling reports of 'malfunctioning' crossings where the OSP says the crossing didn't give them enough time to react.  Their thought processes were somewhere other than operating their vehicle when the protections began to operate - after a period of time they noticed them and were 'surprised'.

On my carrier, whenever this is reported, a maintainer is dispatched.  Findings upon arrival - protection is functioning as intended, review of crossing equipments log reveals no exceptions to operation in any period preceding the reported 'defect'.

My carrier takes the proper functioning of crossing protection seriously.  I suspect all carriers do, including Metro North.

Never too old to have a happy childhood!

              

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Sunday, February 8, 2015 3:53 PM
We know the gates were apparently working because one came down on the victim’s vehicle.  I am not sure if we know whether the red flashing lights activated when they were supposed to.  The second driver might have confirmed that they did. 
However, if they activated after the gate lowered, then it is perfectly understandable how the driver would have been struck by the lowering gate.  The lights are supposed to warn drivers before the gate comes down so they have time to stop for it.  Without the flashers warning of the gate descent, it would be like having a traffic light go from green to red. 
We also know the public reported problems with the crossing protection system earlier the day of the crash.  That does not prove that a system malfunction caused the crash, but it is quite a coincidence. 
We also know that traffic engineers recognized the crossing as having a driver visibility problem due to the curves in the road. 
And then there is this little nugget of news from this link:
Quote from link:
“All railroad grade crossings have gate arms that are designed to lift automatically if they strike something like a car on the way down, railroad safety consultant Grady Cothen said.”
 
We know that did not happen in this accident. 
  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Sunday, February 8, 2015 4:19 PM

^

Why the double posting?  This was posted on page 3!

Never too old to have a happy childhood!

              

  • Member since
    September 2010
  • 2,515 posts
Posted by Electroliner 1935 on Sunday, February 8, 2015 4:41 PM

When I was working on the PRR in the 50's, Gates were released to lower by gravity and the descent was controlled by the motor acting as a generator into a resistance. Otherwise it would freefall and could do damage. No sensor was used to cause the gate to rise if it struck smething, it would just remain on whatever it landed on until the track circuits came clear (no train on track) and the motor circuits were energized to raise the gate to the upright and latched. I doubt that that design has changed. I know nothing is totally perfect but railroad signalling is designed to be fail-safe. I remember a crossing where a washout had caused the signal case to fall over but the crossing signal had one red light lit in both directions. Not flashing but lit. for the crossing signals to "clear" ie, gates rise and flashers go dark, the approach track section and the section between the approach sections had to be clear. Energized and no shunt. Today, they use audio tone overlay electonics but they provide the same functions. The poor lady failed to heed the sign "DO NOT STOP ON THE TRACKS" and paid a high price. Habit and highway design traps people into bad choices. Whether she thought she could get clear, why she did not back up or for whatever reason she failed to avoid being on the track when the train arrived. And then the train pushed the car into the energized third rail which being anchored to its supports was pushed up and into the lead car spearing it and setting it on fire due to the high fault currents arcing to the inside of the car. Such is how this tragedy occured. I suspect the five passengers who died were on the third rail side of the car. Hell of a way to go.

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Sunday, February 8, 2015 5:25 PM

I am not very familiar with this third rail system, but this article describes what happened with it.  As I understand it, the third rail runs along the track just a bit raised off of the ground.  Then it is interrupted at a grade crossing, so the train loses contact with it as it passes through the crossing. So as the train passes through the crossing, it approaches the stub end of the third rail on the other side of the gap.

It sounds like the train hit the vehicle, and sort of "snowplowed" it under the blunt end of the facing third rail, and lifted the third rail off of its supports as train carried the impaled vehicle down the track.  And the dislodging rail pierced the vehicle and followed a path through it that led into the first car of the train. 

As it entered the train, the third rail broke up into 39 ft. lengths.  So the passengers were faced with 39 ft. javelins of rail coming through the car at high speed.  Of course, the rail was not traveling, but the train was swallowing it at 58 mph.  Apparently this rail remained energized as this process played out.  Here is an article that talks about it:

http://www.lohud.com/story/news/transit/2015/02/08/metro-north-train-suv-snagged-third-rail-valhalla/23070285/

The article’s expert says that the rail design was not to blame for the crash.  I wonder about that.  I would not say that the third rail caused the accident.  However, if the third rail had stayed intact, it is perhaps likely that nobody on the train would have been killed.

There must be a lot of grade crossing collisions within this third rail system.  I would think that, in the case of a collision, trains shoving an impaled vehicle into the butt end of the third rail would be highly likely. A lot of cars that get hit by trains stick to the front of the engine and are carried a long distance.   

 

  • Member since
    October 2006
  • From: Allentown, PA
  • 9,810 posts
Posted by Paul_D_North_Jr on Sunday, February 8, 2015 5:39 PM

Worthwhile column on the problems with a grade crossing warning system forcing trains to stop, etc.:

http://ten90solutions.com/confessions_of_a_heartsick_man 

"This Fascinating Railroad Business" (title of 1943 book by Robert Selph Henry of the AAR)
  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Sunday, February 8, 2015 6:02 PM

Paul_D_North_Jr

Worthwhile column on the problems with a grade crossing warning system forcing trains to stop, etc.:

http://ten90solutions.com/confessions_of_a_heartsick_man 

 

Interesting and informative!

Never too old to have a happy childhood!

              

  • Member since
    November 2005
  • 4,190 posts
Posted by wanswheel on Sunday, February 8, 2015 9:36 PM
“Do not stop on tracks” is good, but it needs help. I’d like to see a wordless picture sign, showing a side view of a car stopped on the track with a train coming, within a red slashed circle. That, plus a plain old-fashioned stop sign on this side of the tracks, so that everyone has to stop before the tracks, whether there’s a train coming or not, and take the moment to see what space there is on the far side.
  • Member since
    December 2001
  • From: Northern New York
  • 25,022 posts
Posted by tree68 on Sunday, February 8, 2015 9:49 PM

Maybe this would help:

LarryWhistling
Resident Microferroequinologist (at least at my house) 
Everyone goes home; Safety begins with you
My Opinion. Standard Disclaimers Apply. No Expiration Date
Come ride the rails with me!
There's one thing about humility - the moment you think you've got it, you've lost it...

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Sunday, February 8, 2015 10:21 PM

Never too old to have a happy childhood!

              

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 7:58 AM
It seems to me that this inverted third rail design would produce exactly what happened in the Valhalla crash in at least 50% of crossing collisions with vehicles.  It is a predictable death trap for train passengers riding near the head end
So I wondered how this obvious danger could be acceptable, considering that the danger would be multiplied by a presumably large number of grade crossings in this third rail system.
However, now I find that there are only two lines with this type of third rail, and only a “few dozen” crossings.  So I conclude that this death trap for train riders is only limited by the small number of crossings improving the odds for passengers.
So now, there are calls to close those crossings.  It is amazing that it took this horrific accident to see the obvious hazard. 
  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 10:20 AM
It is true that a PTC-based obstacle detection system can only detect vehicles trapped by the gates within the time interval after they lower.  Nevertheless, it appears that that would have been sufficient in this case to prevent the collision. 
The crossing activated 39 seconds before the crash.  Since the victim’s vehicle was fouling the crossing at the time of activation, it would have been detected as an obstacle at that time.
In 39 seconds, a train traveling at 58 mph would travel 3317 feet.  So had there been a vehicle detection system, it would have signaled the train to stop when it was 3317 feet away.   After putting this train into emergency braking, it traveled 950 feet, so we know its stopping distance.  Therefore, at 3317 feet away, with a stopping distance of 950 feet, it would have stopped 2367 feet short of the crossing and six people would not have been killed.   
  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Monday, February 9, 2015 11:54 AM

Read -

http://ten90solutions.com/confessions_of_a_heartsick_man

and revise your calculations.  While the lights began at 39 seconds prior to impact - the gates began their decent some, unspecified, seconds after the lights began operating and then would have contacted the rear of the vehicle at some seconds after the gates begand their decent.  I am going to guess, that the gates began their decent 8 seconds after the lights started flashing and would have come into contact with the vehicle 2 to 3 seconds after the decent started, so the gate would have contacted the rear of the vehicle approximately 10 seconds after the lights started flashing or 29 seconds before impact (and it could very well have been less than 29 seconds as I believe the FRA minimum requirement is 20 seconds). 

Emergency is not the desired braking method for trains as it can, for a variety of reasons, cause catastrophic accidents on it's own.

Never too old to have a happy childhood!

              

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 12:28 PM
 
BaltACD
...and revise your calculations.  While the lights began at 39 seconds prior to impact - the gates began their decent some, unspecified, seconds after the lights began operating and then would have contacted the rear of the vehicle at some seconds after the gates begand their decent.  I am going to guess, that the gates began their decent 8 seconds after the lights started flashing and would have come into contact with the vehicle 2 to 3 seconds after the decent started, so the gate would have contacted the rear of the vehicle approximately 10 seconds after the lights started flashing or 29 seconds before impact (and it could very well have been less than 29 seconds as I believe the FRA minimum requirement is 20 seconds). 

Okay, say the vehicle became a fouling obstacle when the train was 20 seconds away.  At 58 mph, that would mean that the train was 1701 ft. away.  It required 950 ft. to stop, so if it was warned by a PTC-based obstacle detection system, the train would have stopped 751 ft. short of the crossing and six people would not have been killed. 
  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Monday, February 9, 2015 1:12 PM

Euclid
 
 
BaltACD
...and revise your calculations.  While the lights began at 39 seconds prior to impact - the gates began their decent some, unspecified, seconds after the lights began operating and then would have contacted the rear of the vehicle at some seconds after the gates begand their decent.  I am going to guess, that the gates began their decent 8 seconds after the lights started flashing and would have come into contact with the vehicle 2 to 3 seconds after the decent started, so the gate would have contacted the rear of the vehicle approximately 10 seconds after the lights started flashing or 29 seconds before impact (and it could very well have been less than 29 seconds as I believe the FRA minimum requirement is 20 seconds). 

 

Okay, say the vehicle became a fouling obstacle when the train was 20 seconds away.  At 58 mph, that would mean that the train was 1701 ft. away.  It required 950 ft. to stop, so if it was warned by a PTC-based obstacle detection system, the train would have stopped 751 ft. short of the crossing and six people would not have been killed. 
 

Train was in emergency prior to striking the vehicle - how long before hasn't been discolosed.  The train stopped 950 feet after striking the vehicle while the train brakes were in emergency, undoubtly getting additional retardation from pushing the vehicle along the right of way as well as having the 3rd rail entering the rail car and providing additional retardation.  The 'unaided' stopping distance of the train from 58 MPH has not been disclosed in either normal service braking or emergency braking conditions.

Never too old to have a happy childhood!

              

  • Member since
    July 2006
  • From: North Dakota
  • 9,592 posts
Posted by BroadwayLion on Monday, February 9, 2015 1:44 PM

The Sprauge type rail (under-running as on MNRC) is wrapped in ain insulated cocoon with only the bottome of the rail exposed. Brackets told it from the top.

The angled end of the rail permits the shoe (and aught else for that matter) to slide under the rail and perhaps lift it. The insulated cocoon and the brackets holds the rail together as it is pushed up.

The type of rail otherwise used (I am told that Culver was the first to put a protective board over it) slopes down at the end to lift the shoe up and onto the rail. Now the Brentwood accident on the LIRR demonstrates that things can still push under the rail or make contact with the rail, but the rail is not held by brackets as the Sprague rail. It is basically just sitting on the insulator seats with very little other than gravity to keep it there. In the event of an impact against the rail, it is usually just pushed off of its seats and is shoved to the ground.

Whatever. That is the LION's Story and him sticks to it

 

ROAR

The Route of the Broadway Lion The Largest Subway Layout in North Dakota.

Here there be cats.                                LIONS with CAMERAS

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 1:55 PM

BaltACD
Train was in emergency prior to striking the vehicle - how long before hasn't been discolosed.  The train stopped 950 feet after striking the vehicle while the train brakes were in emergency, undoubtly getting additional retardation from pushing the vehicle along the right of way as well as having the 3rd rail entering the rail car and providing additional retardation.  The 'unaided' stopping distance of the train from 58 MPH has not been disclosed in either normal service braking or emergency braking condition.

I got the stopping distance from the link you provided in which it says this:The eight-car train stopped 950 feet after the brake was applied, Sumwalt said.”  It says “after the brake was applied.”  It does not say “after hitting the vehicle.”  I agree that tearing up the third rail and the resistance of shoving the vehicle would have added some deceleration.  But there is still a fair margin in the numbers.  So I conclude that there was at least a good chance that the PTC-based obstacle detection system would have prevented the crash.
I did read that other link, Confessions of a Heartsick Man, but I have a hard time following his description because it is so incredibly wordy.       

 

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 4:07 PM
 
Lion,
Your insight on these lines is most welcome.  I was thinking that you would be familiar with them.  Is the “rail” of the third rail composed of actual railroad running rail section, or is it just some sort of bar stock?
  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Monday, February 9, 2015 4:43 PM

Euclid
 
 
BaltACD
...and revise your calculations.  While the lights began at 39 seconds prior to impact - the gates began their decent some, unspecified, seconds after the lights began operating and then would have contacted the rear of the vehicle at some seconds after the gates begand their decent.  I am going to guess, that the gates began their decent 8 seconds after the lights started flashing and would have come into contact with the vehicle 2 to 3 seconds after the decent started, so the gate would have contacted the rear of the vehicle approximately 10 seconds after the lights started flashing or 29 seconds before impact (and it could very well have been less than 29 seconds as I believe the FRA minimum requirement is 20 seconds). 

 

Okay, say the vehicle became a fouling obstacle when the train was 20 seconds away.  At 58 mph, that would mean that the train was 1701 ft. away.  It required 950 ft. to stop, so if it was warned by a PTC-based obstacle detection system, the train would have stopped 751 ft. short of the crossing and six people would not have been killed. 
 

Not quite correct, it stopped in 950 feet while it was shoving a 2 ton Mercedes SUV and plowing up a 3rd rail....the additional mass/weight and resistance would affect the distance.

23 17 46 11

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 4:58 PM

Ed, I mentioned that in a following comment.  I said: 

"I agree that tearing up the third rail and the resistance of shoving the vehicle would have added some deceleration. But there is still a fair margin in the numbers. So I conclude that there was at least a good chance that the PTC-based obstacle detection system would have prevented the crash."

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Monday, February 9, 2015 5:02 PM

Darwin candidates 'outwit' any system designed to prevent their candidacy.

Never too old to have a happy childhood!

              

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 5:05 PM

BaltACD

Darwin candidates 'outwit' any system designed to prevent their candidacy.

 

That false belief is the entire problem that is holding up progress in making things safer.

  • Member since
    May 2012
  • 5,017 posts
Posted by rcdrye on Monday, February 9, 2015 5:09 PM

MNRC third rail is shaped like a sideways H with a top and bottom shaped more or less like an ordinary rail head.  The top is held by the insulation, in turn held by the brackets.  The actual rail is smaller in section than an overrunning third rail. Brackets are screwed to extra-length ties about every 15-20 feet.  The rails' breakaway characteristics compare favorably to overrunning third rail.

What almost everyone is overlooking is that this is more or less the first accident in which the design of the third rail was really involved, and then only with a scoop in the form of an SUV shoved into the end.  For something that's been around over 100 years, the Sprage-Wilgus underrunning third rail has a fantastic safety record.

  • Member since
    December 2001
  • From: Northern New York
  • 25,022 posts
Posted by tree68 on Monday, February 9, 2015 6:00 PM

Euclid

Ed, I mentioned that in a following comment.  I said: 

"I agree that tearing up the third rail and the resistance of shoving the vehicle would have added some deceleration. But there is still a fair margin in the numbers. So I conclude that there was at least a good chance that the PTC-based obstacle detection system would have prevented the crash."

Note, however, that the engineer made an emergency application.  I would presume that a PTC system would initially make a service application (or a penalty application), which would not slow the train anywhere near as quickly.  That 950' would have been well over 1,000 feet.

Said application would merely serve as notice to the engineer that there was an issue - which he (or she) may already know.  It would then be up to the engineer to decide to override the service/penalty application with an emergency application.

LarryWhistling
Resident Microferroequinologist (at least at my house) 
Everyone goes home; Safety begins with you
My Opinion. Standard Disclaimers Apply. No Expiration Date
Come ride the rails with me!
There's one thing about humility - the moment you think you've got it, you've lost it...

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 6:15 PM

.

 

 

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Monday, February 9, 2015 6:24 PM

tree68
 
Euclid

Ed, I mentioned that in a following comment.  I said: 

"I agree that tearing up the third rail and the resistance of shoving the vehicle would have added some deceleration. But there is still a fair margin in the numbers. So I conclude that there was at least a good chance that the PTC-based obstacle detection system would have prevented the crash."

 

Note, however, that the engineer made an emergency application.  I would presume that a PTC system would initially make a service application (or a penalty application), which would not slow the train anywhere near as quickly.  That 950' would have been well over 1,000 feet.

Said application would merely serve as notice to the engineer that there was an issue - which he (or she) may already know.  It would then be up to the engineer to decide to override the service/penalty application with an emergency application.

 

If the obstruction developed when the train was close, I would not presume that the system would start with a service application.

Generally, I think the whole point would be to react to an emergency obstruction, and therefore apply the maximum stoppering effort as early as possible. This overrides the engineer because it can see farther.

However, if the obstruction were detected say 40 miles ahead, then I can see it starting with a service application.

If this system were in place at the Valhalla crossing, it would have put the train into emergency at the first detection of the obstruction.

 

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Monday, February 9, 2015 6:33 PM

Maybe the crossing protection should take control of automobiles in the area of the crossing and apply their brakes?

https://www.youtube.com/watch?v=NqrS-wMGKYM

Never too old to have a happy childhood!

              

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy