Pondering an idea that might've saved some lives..

[DJ700R4]

Hey guys, so I saw the news and heard about the Amtrak accident that happened the other day (r.i.p to those who died), and it got me to thinking, what if there was a system of powered wheels placed at the curve of the derailment? Sounds confusing but what I mean is a system of wheels placed at a curve designed to prevent the train from *completely* rolling over in the case of a derailment. These said wheels would gradually slow the train down as it rubbed against them. (rolling counter clockwise to the direction of the train at maybe 20mph?) Catch my drift? Now how this would support the weight of the train, I'm not entirely sure (I'm sure a good bit of heavy duty fabrication would be involved to hold everything in place though) but as far what material the wheels would be made out of, I think rubber would be a nice idea. (feel free to chime in if you have better alternatives). Something that wouldn't be too abrasive on the train cars.

This system would extended at least a 1/4 mile after said curve so hopefully it would create a chain reaction of the train gradually being pushed back on the tracks.

If you need me to be more specific about something, please feel free to say so. Also, lets not be rude because I seriously think this would be a nice idea to try and implement. It might sound stupid/weird/etc but it might prevent something like this from happening again.

Thanks!

[Greg Moore]

I understand what you're suggesting.

However, I think it's problematic from a number of points of view.

Firstly: it would be hugely expensive to maintain. And who decides what curves you put them on and where? There's probably 1000s of such curves in the country.
Secondly, where would you put the equipment? There's 4 tracks in that area (not counting the diverging ones, etc) in a tight urban area. You simply wouldn't have room to fit them all and this equipment.
Third: The wear/tear on the cars themselves would be huge.
Fourth: Some percentage of passengers would need new underwear after each trip after the first time they enter a curve and hear the thump and see these huge spinning tires rushing towards them.
Fifth: What's the result if these failure? (Say a tire locks up and doesn't spin). Does it rip off a window or two by mistake?

But finally, I think it's a bad idea because it addresses the problem in exactly the wrong way. It adds whole new levels of complexity in an effort to solve a problem when other solutions exist.

Don't get me wrong, sometimes technology IS the way to go. And I suspect PTC will be at the top of the list.

But first, you have to look at things like procedures and root causes.

What CAUSED this accident? We've had two similar accidents now here in the US in less than 2 years (the second being the Metro-North crash).
We know that one was at least caused in a huge part by the engineer having an un-diagnosed sleep disorder AND a major shift in his work schedule which disrupted his sleep schedule. Now, we know PTC could have most likely prevented that accident, but that doesn't solve the root cause. Experts in psychology and human factors have warned for decades about problems caused by sleep patterns, yet the various transportation industries tend to not treat them very seriously (the Buffalo air crash in 2009 is another example of this. This is NOT unique to the rail industry.)

So while we almost certainly should implement PTC, we should also look at what we're asking of the crews of our planes, trains, busses, tractor trailers.

While most likely not practical for a host of reasons, as one person on the thread about the crash suggested, calling out the signals tends to break people out of any mind-hazes they're in. (this btw is a continual problem with automation. Automation can make routine things much easier, but often can lead to the pilot/engineer/driver/etc not being focused enough to be engaged when they are most needed).

For this crash, it's far to early to really say what the issue was. I have my speculations, but I'm keeping mum for now. And I'm certainly NOT advocating any specific solution without more data.

But let's just say that rarely does an accident have just a single cause. It's usually a confluence of events. Generally it involves a human factor and a mechanical factor. We're all to quick to jump to a mechanical solution when often redressing the human factor is the better choice.

BTW, I recommend (and I may post in the other thread) that folks read Normal Accidents by Charles Perrow to understand how not only sometimes a technical solution won't solve the problem, it can make it worse.

That said, it's always worth considering off the wall ideas. 99% of them will never go anyplace. And perhaps 1/2 the remaining 1% shouldn't but sometimes the best solutions come out of nowhere.

[DJ700R4]

Hmm.
Good point. Very good points.
Makes sense.

[electricron]

In this day of automation and the priority of seeking technical solutions, I wonder maybe the human solution would be cheaper and better? Could returning to past rules of having a qualified second person in the locomotive cab avoid most overspeed accidents, acknowledging that overspeed accidents have always existed? Is finding a perfect 100% technical solution achievable? I believe all of us could agree that achieving a 100% human solution is impossible.

As a passenger riding planes and trains, I would feel safer knowing there were two qualified humans in the cabs and cockpits.

[rohr turbo]

I think your idea is impractical. Remember that the side of a train in a curve is not an arc, but multiple flat sides with angles and gaps between them. Your rubber tires would have to be very close to each other or spring loaded or some other very complex and unwieldy configuration.

The simpler solution (which I still think is impractical) is a railroad guardrail like this: http://upload.wikimedia.org/wikipedia/c ... Loop_1.jpg" onclick="window.open(this.href);return false;


Who cares about scratching up car sides in a twice-a-decade occurance if you can save lives and equipment.

But even that guardrail would likely cost 10s of millions $ and I'd much sooner put the money to acquiring land to straighten the line. Then you solve the safety issue and all trains run faster.

[DJ700R4]

Well I did think about the problem with the gaps in between the different train cars and such.
A spring loaded mechanism was also what I was thinking about.

Thanks for pointing that out too though.

[gprimr1]

Could it withstand a train going that fast though?

I'm curious, if he fell asleep, why didn't the deadman's break kick in?

[george matthews]

Amtrak needs to consult the French (or Germans, and other Europeans) who will advise that high speed rail cannot run on legacy tracks. Possibly there could be automatic speed controls, but rebuilding the track to eliminate historic curves is the safest.

Are the Americans serious about speeding up their trains? If so, they need to eliminate these slow tracks.

[DJ700R4]

Another good point and that's actually precisely why I said there would have to be durable fabrication involved in the whole set up, that and something like the other gentleman said too, also have it spring loaded too.
Combine the two somehow.^^

And George , I see what you're saying.

Thanks for the good input guys, I didn't think I'd get this far conversing about this.

[Greg Moore]

Amtrak needs to consult the French (or Germans, and other Europeans) who will advise that high speed rail cannot run on legacy tracks. Possibly there could be automatic speed controls, but rebuilding the track to eliminate historic curves is the safest.

Are the Americans serious about speeding up their trains? If so, they need to eliminate these slow tracks.

And the first thing the French would say is, "Yes we run the TGV on non-high speed lines, we just don't run them fast". My understanding is that most (all?) of the TGV lines use regular lines in and out of the cities until they reach their dedicated lines.

[chrisf]

The simpler solution (which I still think is impractical) is a railroad guardrail like this: http://upload.wikimedia.org/wikipedia/c ... Loop_1.jpg" onclick="window.open(this.href);return false;

Of course this guardrail, being on the inside of the curve, doesn't protect against overspeed derailments, but rather clotheslining, where a long train gets pulled straight across the chord of a curve. It's difficult to say whether it'd be workable on the outside- even a light train like Amtrak 188 has a lot of energy when it's going 106 mph. It may not be possible to put the guardrail close enough to actually keep the train upright, and train cars are not designed to protect against impacts like these at upper levels. Remember that most of the cars in Amtrak 188 stayed upright and did not make contact with anything.

[chrisf]

also have it spring loaded too.

Springs are likely counterproductive in this application. All that energy getting put into the springs would likely force the train to tip to the inside when the springs rebounded.
Keep in mind that this sort of accident simply doesn't happen if positive train control is in place. That's the only idea that makes sense in preventing accidents like this. A guardrail cannot prevent a derailment.

[DJ700R4]

True

[David Benton]

I would think looking at Catenary design may result from this accident. Would a tubular support tower have done less damage?. Larger diameter to compensate for a thinner wall that would crumple on impact, but maintain holding the gantry up .
I have also noted the general emptiness of the area (apart from Catenary), which helped. this could be designed into outside curves where possible. Highway style Guardrails could be placed around sharp objects on curves, but i don't know how helpful these would be against the energy of a 100 mph train.
But PTC must be at the top of the priorities.

[chrisf]

I would think looking at Catenary design may result from this accident. Would a tubular support tower have done less damage?. Larger diameter to compensate for a thinner wall that would crumple on impact, but maintain holding the gantry up .
I have also noted the general emptiness of the area (apart from Catenary), which helped. this could be designed into outside curves where possible. Highway style Guardrails could be placed around sharp objects on curves, but i don't know how helpful these would be against the energy of a 100 mph train.
But PTC must be at the top of the priorities.

I would think that a breakaway type of catenary support would be somehow feasible, in the style of highway lighting and signage fixtures, but the problem is that if you slow the first car even slightly more rapidly than the following cars, the first car is still going to bear the brunt of the energy stacking up behind it, effectively becoming the "crumple zone" for the following cars. I don't think a guardrail is feasible because of this.