5 Killed: Michigan Collision with Wolverine (2009)

[jis]

i feel sorry for those on board all those pax stuck on the train for 4 hours. did amtrak bus them to chicago after the train was towed back.

According to reports, the passengers were bussed to Ann Arbor and then put on a train from there to Chicago.

 

[caravanman]

Interesting. I had assumed that the wheel flats were caused either by a parking brake not being released, or faulty wheel cylinders not releasing when they should.

To my way of thinking, if it is possible, as you say it is, for the engineer to derail the Amtrak train by stopping suddenly, then it might be a lot safer to govern the amount of force that can be applied to the braking process?

Ed

 

[henryj]

I was astonished to learn in a post above that it is possible to derail an Amtrak train by the engineer applying full emergency brakes. This seems like a major design flaw?
Ed

Lock up the brakes on your car at 100mph and see what happens.

I think a more apt, though perhaps less relevant analogy is - lock up the brakes of a tractor-trailer at 100 MPH and see what happens. The risk of derailment stems from the fact that when the engineer dumps the brake pressure, the coaches slam forward into a declerating locomotive.

But wait, you say, the coaches have brakes too, and they are connected to the locomotive. Yes, but it takes a certain amount of time for the decreasing pressure to travel all the way back to the last coach. So the front cars start slowing before the back cars do. Now this all happens very quickly, but the resulting forces in the train can cause it to jump the tracks.

The solution to this is to use electrically controlled pneumatic brakes, in which the signal to brake travels via an electrical wire that spans the length of the train. Since the electrical signal transmits nearly instantly, all the brakes apply uniformly. But ECP brakes are very new technology and only a handful of trains in the US have them - each car needs to be equipped with ECP technology for the system to work. The Amfleets, and all of Amtrak's fleet, were created long before it.

Now it's worth noting that the above described effects are much more pronounced with freight trains, I'm not sure how much an Amtrak train would be affected by them.

It has always been my understanding that passenger trains DID have electrically controlled pneumatic brakes and have had for decades, well before Amtrak. Did I miss something? Did they take them off all Amtrak trains somewhere in the past? Somehow I don't think so.

Also, the danger of a derailment is comes from the vehicle damaging the track as it is dragged along or some of the debris getting under the engine's wheels.......not from the braking. That would only apply to long freight trains that might double up on a curve or something.

 

[Green Maned Lion]

Amtrak trains, and all FRA-compliant trains running in this country, use a variation of the automatic air brake invented by George Westinghouse. Basically, the locomotive produces pressure and charges the brake line. The pressure in the brake line holds the brakes in the car in the off position. When the engineer applies the brakes normally, the air pressure is reduced causing the brakes to apply. When the engineer puts the train into emergancy from the engine, all air is released from the system through the engine. In trains with EOT/FRED mechanisms are put into emergency, "Wilma" tells "FRED" to also dump the air pressure out of the rear of the train. When a car's emergancy brake is pulled, air is dumped from the entire train through the valve in that particular car.

Lastly, if anything were to disrupt the brake line in any way, the air would explode out as in an emergancy braking application, and the train would come to a halt.

Assuming you don't close the cocks on the system (which can be safely demonstrated not to be the case with a brake test) the system is practically fail-safe- because its default condition, unlike, say, automobiles brakes, is fully applied.

 

[caravanman]

That is indeed pretty much my understanding of how the air brakes work. I am still struggling to comprehend that a simple emergency brake applicartion at speed could derail the train, which was the original statement that attracted my attention.

Ed

 

[jis]

Doesn't Amtrak already use a form of ECP (Electronically Controlled Pneumatic) brakes on all its trains except the Auto Train? I have read that Amtrak is planning to equip the Auto Train too with a form of ECP.

 

[amtrakwolverine]

watching the news on this just now a couple pulled up to the tracks after the crash and jumped out the truck and ran to the wreckage to see if they could help they were banging on the car windows trying to see if anyone was alive. they said the conductor stuck his head out the window yelling get out of here leave and they said no we're trying to help.

 

[ALC Rail Writer]

Foolhardy move. The tracks were not clear, they could have added to the casualty list.

 

[oldtimer]

I would like to correct a couple of things:

1: All Amtrak cars and locomotives are equipped with a wheel slide sensing device that is designed to prevent wheels from sliding regardless of speed. This modulates the brake cylinder pressure preventing any axle from stopping before another axle on that specific car or locomotive. This is the equivalent of the Anti Lock Brake System in your auto.

2: As the GML stated "When the engineer applies the brakes normally, the air pressure is reduced causing the brakes to apply. When the engineer puts the train into emergency from the engine, all air is released from the system through the engine." This is partially true. The normal operation is correct, but when an emergency application is made the rapid reduction of pressure in any car will cause will cause a valve in that car to go into emergency and dump more brake pipe air. This causes the rapid reduction to the next car which is also equipped with that same valve to apply the emergency brakes in that car and so on and so on. Passenger cars very rarely run into the locomotive or each other, An emergency stop often goes unnoticed by the average passenger if no object is hit.

I would also like to add my condolences to the engineer,crew,survivors of the victims and also to those in the white SUV that had to witness this tragic event.

I would also like to point out that the original French built Turboliners, RTG's, based in Chicago from 1973 to 1981 were equipped with ECP braking and had some fantastic stopping distances. On clean dry rail from 79 mph stops of under 1,000 feet were measured, and from 50 mph stops of 500 feet were measured.

 

[ALC Rail Writer]

And thus the turboliners leak into this thread!

 

[Longford]

Here's a link to a Chicago Tribune news article about the crash, posted a short time ago:

http://www.chicagotribune.com/news/chi-ap-...0,7449430.story

 

[AlanB]

I wonder how cameras would work at a crossing. They look for motion after the light has changed, or in this case, the gate has gone down. But there will always be movement because the train goes through! I'm sure there is a technical solution for that though.

Most red light cameras that I know of are not motion sensative. If they were, they'd be taking pictures of the cross traffic in motion.

Most red light cameras are activated by your car passing over a wire loop in the roadway right after the stop line. That coupled with checking to see if the light is red via the control box, is what triggers the camera.

 

[DET63]

Here's a link to a Chicago Tribune news article about the crash, posted a short time ago:
http://www.chicagotribune.com/news/chi-ap-...0,7449430.story

The crash occurred around 12:30 p.m. in Canton Township, about 20 miles west of Detroit, police Sgt. Craig Wilsher said. He said the vehicle was heading north when it crossed the train tracks and was struck. The train typically travels about 67 miles per hour at the site of the crash, [police spokesman Sgt. Mark] Gajeski said.
...

Gajeski said the car was pushed about a mile from one road crossing to another.

If the cop was right, and the train was going 67 mph (not 79 mph or whatever others have speculated), how long should it have taken the train to stop? A mile, as the cop suggested?

 

[ALC Rail Writer]

If the cop was right, and the train was going 67 mph (not 79 mph or whatever others have speculated), how long should it have taken the train to stop? A mile, as the cop suggested?

Many people are trained to think a train takes a mile to stop because that's a figure that is thrown around to prove a point. Remember that the media will think anything over a half-mile is equal to "nearly a mile".

According to the account a truck pulled over and people went over to offer assistance. If that is the case, they were probably only a few hundred yards at most from the crossing.

 

[ALC Rail Writer]

http://news.yahoo.com/s/ap/20090709/ap_on_..._train_hits_car

The car was broadsided and pushed down the tracks. It was crumpled underneath the front of the train and ended up right-side up, its roof and front crushed.Gajeski said the car was pushed about a mile from one road crossing to another.

James Reese, 59, of Royal Oak, who was taking the train to an Ann Arbor museum with his wife and grandson and was riding in the second car, said he felt a brief "surge" of the brakes but "no impact and no sound."

That does sound weird.

 

[DET63]

If the cop was right, and the train was going 67 mph (not 79 mph or whatever others have speculated), how long should it have taken the train to stop? A mile, as the cop suggested?

Many people are trained to think a train takes a mile to stop because that's a figure that is thrown around to prove a point. Remember that the media will think anything over a half-mile is equal to "nearly a mile".

According to the account a truck pulled over and people went over to offer assistance. If that is the case, they were probably only a few hundred yards at most from the crossing.

I would hope that a police officer, who may have to testify in court in matters where precise and accurate information is expected (and often required), would be trained to judge what a mile is, irrespective of what news media flunkies may say. IOW, if he says it was a mile, it should be a mile or close to it. Nonetheless, I wouldn't be surprised if the train stopped in a considerably shorter distance than that, especially considering that it was an Amtrak train, not a heavily loaded mile-long freighter, that was involved in the accident.

 

[ALC Rail Writer]

If the cop was right, and the train was going 67 mph (not 79 mph or whatever others have speculated), how long should it have taken the train to stop? A mile, as the cop suggested?

Many people are trained to think a train takes a mile to stop because that's a figure that is thrown around to prove a point. Remember that the media will think anything over a half-mile is equal to "nearly a mile".

According to the account a truck pulled over and people went over to offer assistance. If that is the case, they were probably only a few hundred yards at most from the crossing.

I would hope that a police officer, who may have to testify in court in matters where precise and accurate information is expected (and often required), would be trained to judge what a mile is, irrespective of what news media flunkies may say. IOW, if he says it was a mile, it should be a mile or close to it. Nonetheless, I wouldn't be surprised if the train stopped in a considerably shorter distance than that, especially considering that it was an Amtrak train, not a heavily loaded mile-long freighter, that was involved in the accident.

That AP story seems to confirm it was a mile. "Crossing to crossing"

 

[jackal]

I always imagined that the flanges on the wheels would guide the train, as there is no element of steering within the engineers controls. Is it a fact that a full brake application causes the train wheels to all "lock".. My thought is that despite the efforts of the brakes to stop the wheels rotation, the weight and momentum of the train would keep all the wheels turning for some considerable time?
Ed

Actually, it's very common for the wheels to lock up. Keep in mind the surface that a wheel touches about a rail is about the size of a dime. When you combine very heavy braking pressure with high speeds and large amounts of weight, the wheels do many times (if not usually) lock up. This actually causes flat spots on the bottom of the wheels and if this is bad enough the cars need to be taken out of service till the flat spots can be machined out or the wheels changed.

Keep in mind that an emergency braking application only produces about 20-30% more braking effort than a full-service application. At a full-service application, the air pressure in the 2500cc service portion of the dual-compartment reservoir is pretty much equalized with that in the brake cylinder at somewhere between 50 and 60 psi. Putting the train into emergency adds the 3500ccs of air in the emergency portion of the dual compartment reservoir into the brake cylinder, which sounds like a lot, but it really only increases the pressure in the brake cylinder to about 70-75psi. (These numbers are from memory of brakeman training in spring of 2006. I did jot the calculations down in my notebook--I should go back and look them up. It'd be interesting to review that stuff anyway.)

Also, one has to take into account the fact that steel-on-steel (especially over a dime-sized contact area) has a much lower coefficient of friction than rubber-on-asphalt or other contact methods. (This is one factor of why rail transport is so much more energy-efficient than road transport--a 100-car freight train has about the same contact with the ground on a square-inch basis as an 18-wheel semi truck.)

An emergency or even full-service application can certainly result in locked/stuck wheels, but usually only if the car is an empty (not loaded) or the traction between the wheels and rail is compromised (e.g. wet rail or debris like leaves on the rail). That is, of course, based on my experience with freight cars--I'm not sure how much of a difference people in a passenger car make to the overall weight of the car and/or how heavy a passenger car is compared to, say, an unloaded hopper ("gon") or tank car ("can"). I do know that some cars--usually, in my experience, articulated TOFC cars--have a weight sensor which (pneumatically--nothing fancy like an electronic scale) controls the amount of air allowed to enter the brake cylinder during a brake application, preventing wheel lock-ups.

Of course, I don't know if the rapid application of a brake at high speed would cause the wheel to lock (and once it's locked, the principles of kinetic vs. static friction dictate that it's not as easy to get it spinning again as if it had never locked up).

Interesting. I had assumed that the wheel flats were caused either by a parking brake not being released, or faulty wheel cylinders not releasing when they should.
To my way of thinking, if it is possible, as you say it is, for the engineer to derail the Amtrak train by stopping suddenly, then it might be a lot safer to govern the amount of force that can be applied to the braking process?

Ed

That is actually the most common cause of flat spots, since mass emergency brake applications aren't super common.

It's not exactly easy to control the amount of force that can be applied when doing an emergency brake application. During a service application, it is, but that service application takes many seconds (even minutes, on a long freight train) to propogate throughout the train. If an engineer saw something on the tracks and did a full-service application at 60mph, the effects of the brake probably wouldn't even begin to take effect until well after the train passes whatever caused the engineer to apply the brakes.

In contrast, an emergency brake application dumps all of the air immediately. The aforementioned Number 8 vent valve helps to quickly vent all of the air in the trainline to atmosphere, thus propagating the emergency application quickly. If you have someone plug a long cut of cars in a railyard and you're standing in the middle of the cut, you can hear the air exhausting out of the Number 8 vent valves in extremely rapid succession--probably no more than 10 seconds for a 100-car cut. (It's actually a really cool stereo effect if you're standing in the middle of the cut of cars.) Once the trainline's air is dumped, the brake piston begins actuating within about 3-5 seconds and is fully applied within about 10 seconds (though you can hear the air pumping through the valve applying the last few PSI for a good 30 seconds or more after the emergency application). Unfortunately, you can't apply less than an emergency application without vastly increasing the amount of time it takes to apply the brakes.

That is indeed pretty much my understanding of how the air brakes work. I am still struggling to comprehend that a simple emergency brake applicartion at speed could derail the train, which was the original statement that attracted my attention.
Ed

Well, as you can see, it's not just "a simple emergency brake application." An emergency application applies maximum force as quickly as possible. There are a number of factors which could lead to derailing--differential forces (especially around curves), wheels riding up and over the rail, whatever. Locked wheels may contribute to that, but actually, a locked wheel has less traction than a spinning-but-braked wheel (think of anti-lock brakes on your car versus locking your wheels, especially on ice). If Amtrak cars are equipped with wheel lock sensors (not something I've heard of before, but I've never worked on passenger equipment), that would substantially help. Regardless, though, when massive forces are applied to objects of great momentum--well, crap happens.

It has always been my understanding that passenger trains DID have electrically controlled pneumatic brakes and have had for decades, well before Amtrak. Did I miss something? Did they take them off all Amtrak trains somewhere in the past? Somehow I don't think so.
Also, the danger of a derailment is comes from the vehicle damaging the track as it is dragged along or some of the debris getting under the engine's wheels.......not from the braking. That would only apply to long freight trains that might double up on a curve or something.

Doesn't Amtrak already use a form of ECP (Electronically Controlled Pneumatic) brakes on all its trains except the Auto Train? I have read that Amtrak is planning to equip the Auto Train too with a form of ECP.

Not as far as I've ever heard. Many (most?) subway and light-rail lines use ECP brakes, but it's a fairly new process to mainline rail. Remember, too, that ALL engines and cars must be equipped for ECP in order to utilize them, which is why freight rail companies are only implementing them (at this time) on unit trains whose cars always pretty much stay together (e.g. coal trains out of the Powder River Basin).

 

[AAARGH!]

That AP story seems to confirm it was a mile. "Crossing to crossing"

I just used an application that measures between two points on a Google Map.

It is only a 1/2 mile between the two crossings and based on the helicopter video of where the train stopped, it is approximately .4 miles (or ~2200 feet) from the initial collision site.

Anyway, nowhere near 1 mile. It is not possible to be over 1/2 mile because the train never crossed the second crossing.

 

[AAARGH!]

The Detroit Free Press has posted the Survailance Video of the intersection at the time of the crash. The actual collision is blocked by trees, so no gore involved.

You can clearly see the Ford Fusion never slowed down and simply passed the stopped SUV. It was then instantly hit by the train. It is unlikely they ever saw the train or the train saw them based on how fast they both were going in my opinion.

The driver's license had been suspended the day before for poor driving.

 

[RailFanLNK]

To this day, when Burlington Northern came to my high school in 1978 with "Operation Lifesaver" my whole thought process of a railroad crossing change....and for life. I cross RR tracks (the line the CZ traverses on) 4 times every day. To work, to my UPS route, back from my UPS route to the center and back across the tracks to go home. I believe I may be one of the few that slows down and looks not once but twice before crossing. It is this same line that not only once but twice in my career at work I saw the crossing arms NOT go down. Too much faith in technology and with this situation, just a horrible split second decision that cost the lives of a carload of friends.

 

[MattW]

About the braking force, I noticed from watching the video the train had 2 P-42s, one on each end. Could the second large mass at the end of the train account for not going into emergency? With the cars bunching up against one locomotive, it's one thing, but with another large mass like that, wouldn't that contribute to the possibility of a derailment after an air-dump?

 

[DET63]

What street was the train crossing when the accident took place? The news reports usually give the location simply as "Canton Township, 20 miles west of Detroit," which doesn't tell very much.

On edit: I've looked at my Google maps, and it appears to have been at Haggerty Road. Haggerty appears to have a cantilevered crossing signal, though it's impossible to be certain.

The next crossing to the west is Lilley Road, which is an industrial driveway or access road. The track runs at a WSW angle, so it's impossible to tell exactly how far apart the crossings are, but it would appear that they are about 2000 feet apart.

 

[AAARGH!]

What street was the train crossing when the accident took place? The news reports usually give the location simply as "Canton Township, 20 miles west of Detroit," which doesn't tell very much.

Hannan Road. Use the links I provided in the 1st post to see it.

 

[AAARGH!]

Using Street View on Google Maps, the car could not see the train approaching at the speed he was going. There are trees all along the track.

What a stupid chance to take. It's like Russian roulette.