Skip to main content

I would guess that the track moved during the accident and that the "after" photos of the track geometry are misleading.  Anything can be moved if enough force is applied.  We will not know if a track problem contributed to the accident until the accident report is released.  From what I have read, real railroads use spiral transition curves.  This means that the track radius transitions from straight to a very broad curve and the radius is gradually decreased until the curve reaches its final radius.  I would expect that the curve involved in this accident was built in this manner.  

This is unlike on most model railroads where we go from straight to the final curve radius with the next track section.  This works on model railroads because both our flanges and track is oversize and the mass of the equipment is so much less in comparison to a real train.  I have seen a few model railroads built to prototype standards.  In O gauge, the rail height is about the size of HO track (maybe smaller?) and locomotive flanges are tiny.  

NH Joe

I don't think that strange straight piece was more than 100 ft. and there was a kink in the one rail.  I have observed a lot of track work and never saw a curve like that. But the excessive speed was probably at fault unless  something happened to a rail.  They were going 80  MPH, in the Bennington curve wreck on the PRR mainline on the "Slope" above Horseshoe Curve the steam engine was estimated to have been going about 65, track was elevated,  and the whole thing crossed 3 more tracks and most of it went down the hill. Still 15 MPH on the Slope  and if they go faster horns blow.

I doubt this was hacking, when I hear claims of hacking when accidents happen usually it is the anti technology crowd looking for some reason to go back to 'the good old days", or trying to excuse what this seems to be likely at this point, human error. If someone was going to hack something for terrorist reasons they would hit something large scale, like hitting the central control system of a CTC system, or even into the PTC system's central control computers, and cause widespread mayhem. One train like this would be a lot of effort for relatively little payoff, causing a bunch of trains to crash at once would do more of what a terrorist would want to do than one train going off the rails. Something like hacking into a specific train would require a lot of work, would likely have to be done close range to the train (from what I know of PTC systems, for example, that can tell an engine to 'stop' or 'slow down', they have transmitters all over the route to relay the stuff to the engine, akin to cell towers, but not an expert on them), it is why a central system would be a lot more of a target, get a heck of a lot more bang for the buck compromising a central system, can be done remotely over the internet from anywhere, doesn't require figuring out the code the engine system would respond to and so forth, hack upstream you do a lot more damage a lot easier....

As far as what happened here, last I read they don't see any signs of brakes being applied, it is almost as if the engineer didn't notice ever after they hit the curve, and didn't have time to hit the brakes, regular or emergency. That part is the one that is going to take time, to get the accounts of those in the engine at the time, the telemetry from the train, to try and figure it out. 

With Bostian, the reason the judge tossed out the criminal complaint was that he felt like Bostian had done nothing criminal, that it simply was an accident, that he wasn't on drugs, wasn't deliberately going too fast and from what the Judge seemed to feel felt that Bostian had been distracted by the reports of the rocks being thrown through the other trains windshields. From the article I read it sounded like the DA wasn't going to prosecute but there is a Pennsylvania law that allows someone to sign a complaint against someone and the DA has to act (or some such, I am not a lawyer). Criminal negligence has a different standard than civil negligence, with criminal negligence I seem to recall you have to show a very high level of indifference, the same way that criminal cases have a much higher level of proof than it does in a civil case. 

In this case, it is going to come from the facts, as to why the train was going too fast and why the engineer allowed it to happen. Right now we simply don't know, and won't know until the investigation is done, all the pieces have been put together, and the full picture is formed. History is full of cases of 'slam dunk' verdicts , snap to judgements that led to things like people being lynched who were innocent, and I reserve judgement on the engineer because even though this seems cut and dried, with what we know, we don't know the whole picture yet. It could be the engineer didn't realize he was in a slow section, got distracted, could be something pulled his attention, we don't know. 

TM Terry posted:

Is there any thought as to when we might actually hear what the engineer of the loco has to say? Or has our legal system and associated liabilities made this unlikely for quite some time?

I'm sure that will all depend on what the NTSB decides to release to the public. Then again, once out of the hospital, the Engineer could conduct an interview with the media, however I seriously doubt THAT.

Dominic Mazoch posted:

Odd.  Freight locomotives on the two track line (Two Main Lines?) have PTC operating, and Amtrak and WashDot's do not?

Correct. Outside of the Northeast Corridor, Amtrak has NO UNITS equipped with functional PTC. Also, since the whole Federally Mandated use of PTC was/is NOT funded by the Federal Government, thus Amtrak doesn't appear to have all the necessary funds to install PTC on all their diesel units. 

Sounder commuter use this line farther north  Does the ban apply to them also?

What "ban"?

 

Last edited by Hot Water
Hot Water posted:
Dominic Mazoch posted:

Odd.  Freight locomotives on the two track line (Two Main Lines?) have PTC operating, and Amtrak and WashDot's do not?

Correct. Outside of the Northeast Corridor, Amtrak has NO UNITS equipped with functional PTC. Also, since the whole Federally Mandated use of PTC was/is NOT funded by the Federal Government, thus Amtrak doesn't appear to have all the necessary funds to install PTC on all their diesel units. 

Sounder commuter use this line farther north  Does the ban apply to them also?

What "ban"?

 

I thought Sounder used this "upgraded" line from the new Tacoma station north to the junction with present line.  So are all passenger trains banned from the entire rebuilt route, those that do not have PTC equipment, or in the section where the wreck happened?

Also I thought Amtrak had a few locomotives equipped with PTC for the special rebuilt lines in MI and for the CHI-STL routes?

GVDobler posted:

So when Congress mandates an increase in the minimum wage and doesn't provide companies with the funds, is that an unfunded mandate?

By god I believe he's GOT IT!

I'm just not clear on how that works. If Congress says you can't operate you railroad in an unsafe manor, the tax payer must fit the bill? Funny rule..

Apples and oranges.

 

All OSHA regs are unfunded by the government. They're always the responsibility of the company. The problem is that Amtrak is a quasi public company so a lot of their funding comes from taxpayers. The FAA mandates stuff all the time. Things like airworthy inspections are the responsibility of the air lines, but updating the ATC computer system's going to be a lot of public input. It's not black and white, nor can it be.

PTC, from what I'm told by two BNSF employees, is hardly the 'end-all' that people would like to think it is. Nothing works 100% of the time and they say it fails with some regularity where it is in place. The big problem right now, they say, is the misperception the public has of it, from the Dupont incident thanks to the media.

One news story, I heard, claimed PC would prevent grade crossing accidents. For crying out loud, it's not like the "stop" button on my DCC system that stops a train instantly on my layout. Even if something would lock the brakes on a 1:1 scale train and throw the power into emergency, we all know the thing's still gonna slide a long way before stopping.

Just because something may not work 100% all the time does not negate the possibility that it may have prevented this accident. It's similar to the gun arguments, also lacking in veracity. Just because a different kind of regulation can't prevent every bad instance, doesn't mean it can't prevent some or many instances. As they say, don't let the perfect be the enemy of the good.

I'm sure when the Westinghouse air brake was first implemented, it didn't prevent every train wreck that was the result of the systems it replaced, but I'll be it improved the system by an order of magnitude. We know that the Westinghouse system, where each car has it's own pressurized reservoir to power its own brakes and all the engine needed to do was reduce the pressure on the train line so the triple valve in each car would then release air from its own reservoir to apply that car's brakes, has its own limitations (sorry for the long independent clause).

On modern freight trains, that can exceed a mile in length, the air reduction signal, traveling at basically the speed of sound, can take over 5 seconds for the tail of the train to know that the front of the train has applied it's brakes and the multi-thousand tons of momentum is still pushing forward while the engineer needs the train to stop. So the answer is electronically controlled brakes where instead of an air signal, the "stop" command is electronic and reaches each cars control valve in effectively the same time (a few nanoseconds perhaps). This adds cost to the system since older cars would need to be retrofitted since the entire train would need this new system, but does that mean you shouldn't do it? You still need the train line to pressurize all the cars reservoirs as they don't have on-board compressors, and the electronics would be an add-on.

Professor Chaos posted:
Farmer_Bill posted:

If Congress has mandated Amtrak have PTC by the end of the year then why hasn't Congress allocated funds for Amtrak to accomplish that goal?  

(rhetorical question)  

 

The less funding you provide for government, the more government must operate by regulation rather than by providing solutions.

Huh?

Rule292 posted:
Professor Chaos posted:
Farmer_Bill posted:

If Congress has mandated Amtrak have PTC by the end of the year then why hasn't Congress allocated funds for Amtrak to accomplish that goal?  

(rhetorical question)  

 

The less funding you provide for government, the more government must operate by regulation rather than by providing solutions.

Huh?

The Professor is saying that if the government doesn't take enough away from everyone to pay for everything it tells you, you have to do, you will have to pay for some of what it tells you, you have to do, yourself.

Doug

To try and bring this back to the topic at hand, someone noted that without redundancy any system is not foolproof if it fails and there is no backup, and that is a fair observation.  Aircraft have such redundancy, both military and civilian, and things like spacecraft were/are much the same way.

PTC in of itself is a kind of redundancy, at least in the situation we are talking about. The prime operator even with a PTC system from what I understand is still the engineer, and he/she is relying on signals and messages sent via radio and even their own line of sight to make judgements on operating the train, as well as rules that determine how they may or may not proceed when for example signals fail or other out of the normal situation happens. PTC comes into this as a backup, that if the engineer is operating in a way that is detected to be wrong, it activates  , whether it slows the train down, stops it, or issues some sort of alarm, it is there for when the engineer is apparently not doing their job correctly.  PTC from what I have read is not automated train control, like systems like the BART and other modern commuter systems often use where the engineer is backup to the automated system, it is the other way around. 

Some posters seem to take the position that those asking why PTC was not put on these tracks before the line was opened or saying "PTC is not foolproof" are assuming those asking these questions are trying to deflect blame from the operator of the train (if the facts indeed point to operator error in the end as they seem to right now), and I don't think anyone is, I think what they are saying is that if the technology exists that could have prevented this tragedy (and it could have in this case, there seems little reason to think otherwise, a train doing 80 in a 30 mph zone is a PTC system's bread and butter) it is a shame that it wasn't used here. No system is 100% foolproof, on interlocking systems like the NYC subway terrible collisions have happened because either the mechanism failed, or the motorman purposely bypassed the safety control for example, but the reality is that compared to not having the system in place, the odds of system failure on top of human failure is a lot higher than human control alone, since human beings are even less 100% reliable than a system is.

I for one am not saying that PTC isn't any good, only that people want to be able to rely on such systems as a 'catch all' that can never fail. I think many of you have read the insane claims in the media on what PTC actually is and what it can do if it is in place.

I've recently heard two TV news people declare that PTC would end grade crossing incidents.

I will try to relate my power plant experience to the PTC conversation:

I see the PTC as a "fail safe" system, i.e., if the primary locomotive control system is the engineer, then the PTC intervenes when the engineer (primary system) operates outside the safe parameters of the situation. 

It would be ideal if the PTC had redundancy within itself. That redundancy could be complex, depending on whether it was a mere duplication of equipment, or a different methodology.

I could relate some parallel like systems of steam generation equipment and the boiler operator (locomotive engineer) as they relate to boiler control systems (PTC), but I won't bore you with the details. Insurance providers, over the years, have required more fail safe systems be added to the boiler control/monitoring systems.

Post
×
×
×
×
Link copied to your clipboard.
×
×