Hacker News new | past | comments | ask | show | jobs | submit login
How the FRA is Regulating Passenger Rail Out of Existence (ebbc.org)
304 points by Symmetry on Oct 1, 2011 | hide | past | web | favorite | 109 comments

This is really sad. I drove a long distance (maybe 60 miles?) for the first time in years yesterday because I'm staying with my mom in the suburbs that have 0 public transportation infrastructure.

First: It's impossible to have a life without a car here when the weather gets bad for biking (I'm outside of Seattle so that's a pretty good % of the year). People without a car are basically prisoners in this town with no culture, where the last bus from town leaves at 6:45. I can't go to work on days I can't bike because I usually work crazy long hours, til 10 or 11pm so I have to work from home if the weather's going to be prohibitive (I didn't bring my waterproof gear with me).

Second: My mind was blown at the mental space driving in a car for a long period of time put me in. Weird stuff like traffic lights gave me this feeling of helplessness. It's a blueprint for a system of arbitrary, total control. The fact that no decisions are really based on the situation at hand but on these lights that mindlessly blink from green to red to green to red and you never interact with anyone or anything except through this sheets of glass. Call me crazy but I really think one of the big influences that's creating the massive societal problems we have in the USA can be traced to the fact that between work, home, school, and whatever destinations we get to we interact with one another in this alienated and antagonistic way.

[edit] I know this rant is a little off topic, but it just highlights to me the need for a coherent public transportation network. It'd be interesting to look at this draconian regulation in relation to what was done to the rail network in the USA in the middle of the 20th century. GM and a bunch of other auto-related corporations formed a coalition, bought up and then dismantled lots of inner-city streetcar networks in order to replace them with buses that they would sell to the cities: http://en.wikipedia.org/wiki/Great_American_streetcar_scanda... [/edit]

"between work, home, school, and whatever destinations we get to we interact with one another in this alienated and antagonistic way."

As opposed to within work, home, or school? Most of our public and private institutions seem to train people to be in abusive relationships.

> As opposed to within work, home, or school? Most of our public and private institutions seem to train people to be in abusive relationships.

These don't have to work on an abuse basis, that's a cultural deficiency rather than a built-in defect. With automobile transportation, the distance and antagonism are pretty much built-in.

> With automobile transportation, the distance and antagonism are pretty much built-in.

Only if you expect them to be. A lot of people enjoy driving.

I don't believe I have ever heard anybody say "The only thing I enjoy more than the 45 minutes I spend commuting to work is the 45 minutes I spend commuting back." I'm sure there's somebody who feels that way, but I don't think it's a lot of people.

I can honestly say this is how I feel commuting on my bicycle. I've always been blessed with beautiful commutes, along the water or across a bridge, and it makes for an excellent way to wake up when going to work and an excellent wedge I could drive between work and not work on the way home. I could gauge the stress of the workday I just finished by how fast I rode home.

You must not know many motorcyclists then. Back when I had a bike I thoroughly and completely enjoyed carving through rush-hour traffic at 20 over the speed limit to arrive at my destination VERY energized and alert. Hands down the best part of my day.

> A lot of people enjoy driving.

"Distance and antagonism" isn't about driving, it's about the relation to other drivers.

Oh, to be sure. Almost everyone would enjoy a nice relaxing drive along a beautiful road. The only problem is for most places in the world, we'd need either a whole lot more beautiful roads, or a whole lot less people.

The problem with government regulatory agencies is that once they solve the problem that instigated their creation, they can't stop. They keep looking for more problems to "fix" in a never ending loop of justifying their existence. Once they become typical bloated behemoth bureaucracies, common sense doesn't work anymore.

1. Once a bureaucracy is created, its 1st act is to perpetuate itself. It has no allowance for self-destruction someday.

2. When one coalition of governments and bureaucracies don't work to one's advantage or agenda, create a new one.

3. If threatened w/ abolition, bureaucracy's main defense is counting d no. of bureaucrats who'll lose job, social tension.

4. Before the main mandate is accomplished, create new functions, prgrams and "needs" to perpetuate itself forever.

5. Create one admin or regulations order, monitor how ineffective it is, create a revised regulations order, monitor...


Once a bureaucracy is created, its 1st act is to perpetuate itself. It has no allowance for self-destruction someday.

Certainly we can all see the results of this sort of behavior, which is a real thing.

You can't extract, from that observation, the general principle you propose, though. The boards, bureaus, and commissions that have allowed themselves to fade away are gone, of course. The ones that act to perpetuate themselves persist. When you observe the behavior of the organizations around you, you're not looking at a fair sample of the organizations that have existed.

And exactly the same is true of private clubs, charities, corporations and companies of every sort. This isn't a characteristic of governmental organizations, it's a property of organizations, period.

Ah, but the crucial difference here is a private organization can no longer exist once it has exhausted its purposes. A privately-funded bureaucracy which simply self-propagates while producing no value goes bankrupt automatically, regardless of what they can spin.

Not really. There are plenty of private organisations that produce no value while consuming funds. There are no major differences between large corporations and the the government in this aspect.

Really? Where are they getting these funds if they aren't providing some service or 'value' in exchange? Maybe you are right but you seem to have stopped mid-argument. Examples might help illustrate your point.

The financial sector or defense? They operate within a framework where they are pretty much guaranteed profits.

That isn't a serious response.

1) Both industries are intimately tied to the government in a variety of ways and so aren't particular good examples of 'private organizations'.

2) The notion that these entire sectors have no value is preposterous.

Debating this sort of thing in the aggregate ('banking sector', 'defense sector') just doesn't make sense. Lots of banks fail and go out of business even outside the fiscal mess of the last few years. Same with defense contractors.

Still waiting for an example of a private firm that provides no value and consumes funds indefinitely.

"Value" is a slippery term, but...

Palotta, Komen, Grassroots, and other professional charity/political fundraisers that sell hope and lies to donors and spend the proceeds on administration.

I have no knowledge of these organizations, but I was assuming we were not discussing fraudulent endeavors.

I guess my larger point is that 'value' is in the eye of the beholder. Someone is giving money to the organizations you listed and must have determined that there was value in that from their perspective.

This goes to the heart of the idea of liberty and freedom of action. Individuals are going to value different things in different amounts and should have the freedom to act upon their own judgement. Someone else may not come to the same conclusion but everyone's circumstances are different and so there isn't going to be any universal agreement on 'value', nor need there be.

To go back to the original point, a private organization can not force individuals to fund their endeavors, while the government can (raise taxes, establish regulations and fees with fines or imprisonment for violation, etc).

Because the government has the unique authority to force compliance we should be extremely circumspect about attempts to expand its scope and power to avoid creating a self-sustaining bureaucracy. For example, the notion that government agencies can't be reduced or eliminated because that would eliminate 'jobs' is a terrible reason to avoid the reduction or elimination. It is an acknowledgment that the government agency doesn't exist for any public purpose other than to cut paychecks. By that rationale we should simply tell the workers to stay at home and at least eliminate the props that support the charade (buildings, electric bills, maintenance expenses, and so on).


A privately-funded bureaucracy which simply self-propagates while producing no value goes bankrupt automatically...

It's a beautiful theory.

The most obvious counter-example would be patent/copyright/trademark trolls.

Another might be financial services companies who sell worthless derivatives as AA grade. Or the analysts who provide those grades. Or the brokers who make the worthless loans the derivatives are based on.

That the US's implementation of this law is heavily lopsided in favor of the patent trolls doesn't change the fact that patents incentivize innovation. (Whether they're effective or not is another debate.)

And does it need pointing out that without the government's intervention, banks that give out irresponsible loans would simply go out of business?

That the US's implementation of this law is heavily lopsided in favor of the patent trolls doesn't change the fact that patents incentivize innovation. (Whether they're effective or not is another debate.)

Whether they're effective or not is another debate? The truth or falsity of your preceding "fact" is unimportant?

IP trolling is the patenting of obvious stuff (or buying out the rights to dumb patents) for the purpose of harassing other parties with legal threats. Whether IP law really does encourage innovation in general or not, that activity clearly does not encourage innovation, and it doesn't "produce value." It's called rent-seeking for a reason. They don't want to get paid for producing value, they want to own things. It's a return to the principles of feudalism.

And does it need pointing out that without the government's intervention, banks that give out irresponsible loans would simply go out of business?

Focusing on the banks alone is missing the forest for the trees. The whole mess involved a stack of businesses from mortgage brokers to banks to investment firms to financial services firms. And there was fraud, dereliction of duty, and malfeasance at every level.

Most of those businesses haven't had any "government interference," and are doing well enough, considering how badly home sales have stalled. Some have absolutely made out like bandits.

> That the US's implementation of this law is heavily lopsided in favor of the patent trolls doesn't change the fact that patents incentivize innovation. (Whether they're effective or not is another debate.)

Moving the goalposts.

FRA as such isn't the problem. See, every European country has its own railway regulatory office, which is perfectly fine, but there is simply so little unification between the different countries, that international railway traffic between various EU countries is so painful, that it simply cannot stand in competition with the road freight. You can easily load your cargo on a truck and move it from Bulgaria to Scotland, but it's incredibly painful to do that on rail.

There are tons of various signalling systems, there are multiple electrification systems (even in a single country), there are different norms everywhere, various load gauges, available railway clearances… it's a HUGE mess. And it's expensive.

Some progress is being made, but of course, it doesn't get so easily through. ETCS (European Train Control System) is being developed and very slowly deployed, but only very slowly, it's expensive to equip a locomotive with it (more so than to put a multiple national signalling systems there) and you can use it on only very few lines. And you will never fix everything. Like, you are not going to rebuild every railway line in the UK to the continental clearance. Won't happen.

So, I don't think there is a problem in having a federal reguatory agency. The problem is that it's dysfunctional and nobody really cares. But not having anything like that is bad as well.

There are even different track widths, with the russian system being wider than the most common standard.

Well, there isn't much of a passenger traffic to the Russia and back, there are facilities to handle the freight, it's an EU border (implying long legal procedures regardless of a transport mode) and the roads there aren't very good (in comparison). Baltic states and Finland lack connections to the rest of EU and again, it's mostly freight only. So it doesn't harm rail transport that much. Less obvious, but much worse problem is that there are incompatible axle loads and clearances, russian carriages are heavier and wider. Too much for the most of standard gauge lines.

Much funnier example of a country with multiple railway gauges is Spain. There are three widely used gauges…

I'd say this is more like government taking a local optimum from one time period, and casting it in stone. You can see the same thing happen with in the nuclear power industry, with light water reactors being given a regulatory blessing that made sense at one point.

Ture and agree, but this is true of any organization. The essence of any organization is to serve its own interests. It can be charitable or ideological or profit-driven, but if it does not serve its interests, it is no longer an organization.

We shouldn’t be surprised that gov’t agencies are similar. The difference is that its funding is decoupled from its benefit to constituents.

Amen. You don't have to go as far as an organization. Get an ad-hoc team together and they'll try perpetuate themselves.

It's normal human behavior, and has nothing to do with the "evils" of bureaucracy.

The problem as far government agencies are concerned is the informal corruption that allows for ever increasing amounts of money to flow into those agencies. This is very hard to fight, since it is almost impossible for even the most dedicated and upstanding elected representative to judge if the funds assigned are justified. Surely there is a need for something like the FRA, and surely there should be rules governing the safety of trains?

The funding isn't decoupled, but it isn't coupled to something as simple and transparent as a free market. There needs to be active democratic oversight, which is way, way harder.

I don't buy that. Every other country with high-speed rail has rail regulations as well.

Exactly. In fact, I'd say that looking at the existing evidence successful rail deployments are entirely uncorrelated with regulatory action. This is just a whipping post for the libertarian nutjobs, and the only notable thing is that it's one that aligns them in goal with the pinko commie rail boosters.

Clearly the determining factor isn't regulation, it's subsidy. Countries that pay for fancy rail systems out of public funds ... have them.

Effective, good-quality regulation is a form of subsidy.

This article is right to look at the costs of not updating regulation to meet reality.

The idea is to regulate business out of existence, then declare capitalism to be a failure.

This is the worst sort of comment -- blanket assertion of a controversial yet vague idea, no actual arguments, no specific statements, nothing to discuss.

It's just as much noise as reddit-style joke threads. Those don't seem to be tolerated here, and neither should comments like this.

Ok, as someone who has long been very close to the rail industry, a few points:

1. "...the Long Island Commuter Railroad (LIRR) in New York City, which has no freight traffic" - Incorrect. The LIRR also is host to the New York & Atlantic railway, a freight line which does indeed operate on LIRR lines daily. Average train gross weight is about 1,200 tons for each freight on that line, not 100 tons.

2. "FRA staffers point out that it is unfair to compare US buffering standards with those in Europe because passenger rail in the US has to contend with more (and heavier) freight traffic." - They are quite right to do so. The US plays host to the highest density of freight rail traffic in the world with most trains exceeding 5,000 tons (some closer to 15,000).

3. "In both Europe and Japan, a competitive business exists in the DMU marketplace. But that market is off limits to US transit agencies because the FRA has effectively created a trade embargo." - Incorrect. In San Diego, the Sprinter lines use Siemens Desiro DMU's, a light design totally unadapted for the US Rail network as far as weight goes. The line also plays host to freight trains at night. How they got around the weight requirement, I do not know. Furthermore, the San Diego Trolley has the line from downtown to El Cajon by way of Lemon Grove. That line also plays host to freight at night - a streetcar line! The market is indeed open, but the trick agencies use to get around the requirement is a bigger question.

4. "The FRA proposed rule would only allow Quiet Zones exemptions at crossings that had been improved with "four-quadrant" gates and curb medians." - for good reason! Many fatalities happen at grade crossings and horns are one of the only really efficient ways to keep people off the tracks. Most other nations have few crossings, preferring grade-separated rights-of-way. However, in the US, drivers are often grossly idiotic and don't pay attention. Not how many grade crossing accidents you see on YouTube...

That said, it is indeed a major issue that the FRA rules apply to any rail line connected to the freight network that spans the nation. It would be far better if the regulations made clear exceptions for trains on passenger-only lines, hours of operation, etc.

For example, let's look at CalTrain. The main line up the peninsula only sees freight traffic at night. By setting operational rules that restrict speeds near freight trains, etc, this would allow much better equipment for CalTrain while continuing to let freight run at specific times.

The FRA very much needs to get with the program and allow better conditional standards.

Rail operators are occasionally able to get waivers from the FRA to run non-compliant rolling stock. Caltrain was awarded such a waiver for a future rolling stock purchase, and the Sprinter was awarded one as well. Typically the FRA requires that there be some sort of temporal separation between the trains, which could mean freight only running at night, or it could mean all trains on the line have some form of electronic PTC (positive train control) that guarantees separation between trains. The freight lines have been fighting PTC for several decades, but they're finally being forced to implement it (of course the FRA hasn't set much of a standard, so there's going to be a dozen different incompatible systems, yay).

The problem is that the waivers are arbitrary and one-offs. There's no standard for light passenger rail in the US, there's just a bunch of hyper-specific waivers for each case and operators must re-apply if their rolling stock changes. So if caltrain wanted to switch from UIC-compliant Siemens Desiro trains to a UIC-compliant hitachi train, they'd have to apply for another waiver.

To top it all off, the FRA's own studies showed that their "buff strength" standards are more dangerous than the more modern energy-absorbing crash structures mandated on European and Japanese lines.

The FRA really does need to be reformed.

Totally agreed.

I'm used to see 10 trains on an electrified railway line an hour in each direction, some of them very light EMUs, longer locomotive-hauled passenger trains and heavy freight (ok, heavy is ~2000 tons). With the last deadly accident with a passenger train hitting the freight (and falling from the bridge after that) in... 1970.

Yep, signaling and operating rules are the biggest concern. If you run a railroad properly, you reduce the risk at far lower cost.

Which country are you in?

Czech republic, but you could, of course, see the same thing in Germany or Poland.

And of course, there are more frequent accidents on the lines that are equipped with only basic signalling. This year, a DMU hit shunting freight and one elderly woman died. But not because the DMU was badly damaged (I even think it's already back in the service), but because she wasn't sitting (the train barely left the station). I don't think that being in a heavier unit would help her. (Now, the line is being equipped with something that looks like ERTMS Regional, but... it's not. Another national standard...)

And yes, there are quite frequent problems with signalling failures crippling whole lines for hours (not causing crashes, at least). Almost like we can't have a thunderstorm without some failure.

Ah, ok. Yeah, signaling in the Czech republic is a bit antiquated last I checked. They've been improving, but dispatching is really suffering as traffic density rises.

Thunderstorms crippling it sound like an issue with proper infrastructure maintenance/siting too.

Well, on the recently upgraded lines, there is an up to date system (bi-directional automatic block with cab signalling, not ETCS, yet, only GSM-R on the main lines), but of course, modernization is very slow indeed. Traffic density is hard to manage not because the signalling is deficient, but because various cost cuts that creates stupid bottlenecks (like not using grade separation at crucial junctions, or, also quite popular, not using switches that allow higher speeds to the diverging track).

(And we even got our own FRA-like nonsense: noise limits that demand to build a kilometre long, three meters high noise barriers for a single farm nearby the line in the forest…)

_delirium pointed out the regulation that allows Sprinter and San Diego Trolley to get around the weight requirement in this comment: http://news.ycombinator.com/item?id=3060835

In doing some more research, Utah's FrontRunner has a completely disconnected network that allows them to get around the weight requirement. Apparently they went with standard US commuter equipment (MPI locomotives and Bombardier coaches) because of grade crossings and collision concerns. No sense in having a super-light train at high speed with grade crossing that carry many trucks.

Another failed government policy regarding rail: the requirement for cleaner locomotives. The vast majority of railroads in the US do not buy new locomotives, and in fact usually use locomotives from the 60s and 70s (bought secondhand). Yet, the federal government is requiring newly bought locomotives to have cleaner emissions. Given that most emissions are from yard switchers (again, old locomotives no longer used in long-haul service), this makes no sense...

As an airliner, lobbying for tightened RR regulation seems like a handsome roi.

/takes off conspiracy hat

I wouldn't be too surprised if that's part of it. In the failed early-90s project to build high-speed rail in Texas, Southwest Airlines spent a lot of money and litigation making sure that all "rail safety" rules would be interpreted as strictly as possible, opposing any streamlining or exemptions granted to the project that would've made it more sensible.

Other main players are the freight railroads and the railroad-workers' unions, neither of which cares too much about passenger rail (freight is much bigger business, on both the corporations' and the unions' side).

Ironic as Southwest had to ground 44 planes because they were flying them without the federally-mandated safety checks. Also, one of them "opened up" in flight.

"All is fair in LUV and war," I guess

Good luck: in the UK the worst train accidents with the largest number of fatalities have been due to poor track maintenance. The accidents have tended to occur on commuter trains with high passenger numbers and many people standing. Not sure what the French have seen as history on their much faster trains.

Track has been an issue in the US, as has losing track of trains and having collisions. The vast majority of fatalities are exactly what these laws won't help with: people swerving the gates, or walking on the tracks.

Or a boat crashing into a rail bridge pier:


(Honestly, I feel like this is one of those "cannot be avoided" situations.)

Or parking your car on the tracks:


There is a link in the article about TGV accidents (although to be fair, it is a website about TGV, so maybe biased toward TGV safety, I dunno). It should be noted though that the French railway system is pretty good from a quality POV, if only because it is used so much (traveling within France by plane rarely makes sense time-wise, and buses are rarely used). The situation may not apply to the US, although it would certainly remain a safe transportation.

But I think it is a bit disingenuous not to mention critical factors like low-cost flight companies, huge facilities in the vicinity of airports and big net of inter-states buses. None of that is really true in France (there are low cost flight companies, but they go to airports which make it a non-starter for business).

Dunno about the UK, but in the Netherlands we get news about major train accidents in surrounding countries and I don't think I've ever (well, maybe once or twice) read about a TGV accident. ICE (German high speed rail) accidents, on the other hand, happen at least once every couple of years.

Which I always find ironic because French sloppiness is as proverbial over here as German precision.

> I don't think I've ever (well, maybe once or twice) read about a TGV accident.

There are accidents pretty regularly (thought not frequently, it's about once every two years). Most accidents are not really mentioned because they're level-crossing (some truck barreled through the level crossing on low-speed regional tracks, got stuck and the train smashed into it at 100~150km/h). The result is usually that the truck or car gets annihilated and the train is due for repairs.

A more interesting incident type is the derailment (especially high-speed), and while that has happened several times the record of the TGV on these has been stellar: it was hinted at in the article, a TGV behaves very much like a solid object, it's a single coherent structure and the cars are very tightly bonded together (that's a major difference between TGV and ICE). As a result a TGV is extremely rigid, and in all derailments so far the derailment was partial (a bogie to a few cars) and the rigidity of the trainset is credited with keeping the train as a whole on the tracks even though some cars were not in contact.

FWIW, TGV has the (dubious) award of highest-speed derailment from 1993: a trainset hit an unsupported rail section (a sinkhole had opened under a track after days of heavy rains resulting in a hole 7 meters long by 4 wide and 1.5m high) at 294km/h (185mph). 4 cars and the rear power unit derailed (a TGV trainset is 2 power units and 8 to 10 cars, this was a "TGV Réseau" trainset so 2/8, meaning half the train was off the rails), the train remained upright and managed to stop (over 2.3km). One passenger injury and one passenger was treated for shock.

Yeah, I meant accidents involving derailment. Sorry if I was unclear. These happen quite frequently with ICE trains.

Actually, there has been exactly one fatal ICE accident. Unfortunately, that was also the worst high speed rail accident worldwide so far:


A design error led to a mechanical failure, which caused a derailment at the worst possible location...

The interesting thing about that disaster is that rail safety folks from both the US and Europe I've spoken to have all said that if it had been the Acela in that accident there would have been far fewer fatalities due to FRA crashworithness requirements.

(I worked on the environmental assessment of the Acela back in the 1990's and worked with testing the first two vehicles off the line in Pueblo and along the NEC NJ "race track in 2000. The firm I was working for back then also did the noise analysis behind the revised FRA horn noise rules)

This article keeps popping up on various geek sites over the years. In addition to the points kposehn brought up, I'll add the following: FRA isn't the reason why HSR sucks. The reason that HSR sucks is because we as a nation don't want to invest in the infrastructure to make a good HSR system. At a minimum that means exclusive ROW (grade-separated crossings) with relatively few stops and long straight sections where it can get up to speed.

As for the FTA buffering standards - it really doesn't matter. Yes, the Acela is heavier than the TGV. That extra weight isn't why Acela service sucks. The Acela is perfectly capable of maintaining 155+ mph speeds for extended periods (I witnessed this myself in Acela enduracing testing). The power cars are more than capable of handling the load - during the first few months of Acela operation, there was a problem with the network connection that linked the front and rear power cars. To get service running until the power could be sorted out, the trainsets were run with only one power car operating. Running with one power vs both power cars (and pulling the dead weight of the second power car) only increased DC to Boston run time by 5 minutes. As for cost, the price of an Acela trainset is within the range of most other popular HSR trainsets (TGV, ICE, Pendolino, etc) albiet at the higher end. The effect of the train weight on track wear is minimal as it's the unsprung mass of the train (essentially the wheels, axles, traction motors and brakes) that is proportional to wear, not the static train weight. And train weight has nothing to do with noise.

kposehn already commented on DMU but I'll add that the biggest impediment I saw to transit agencies adopting DMU's was that since no one else in the USA had them, transit agencies didn't know what to expect in terms of maintenance, operation, and environmental effects. In fact FRA and FTA were essentially begging transit agencies to try them, and it's only been recently that they've been operating in Vermont and other locations.

Finally, regarding FRA horn noise rules: first of all, the preemption of state horn rules originated with Congress who directed FRA to get involved with horn noise (Google "Swift Rail Development Act" for more information). But the reason those rules exist is because whenever there is a grade crossing fatality, inevitably the next of kin sue and all too ofter win in court. As a result, there is a tendency to do anything and everything in the name of "safety" on the part of RR operators, agencies and regulators. As long as this remains true, horns are going to be part of rail travel.

edit: btw we had this discussion at ArsTechnica back in 2008 (I'm Anechoic there as well): http://arstechnica.com/civis/viewtopic.php?p=2671067

> The effect of the train weight on track wear is minimal as it's the unsprung mass of the train (essentially the wheels, axles, traction motors and brakes) that is proportional to wear, not the static train weight.

I find this hard to accept. Just because there's a spring doesn't mean the weight isn't there, bearing down on the wheels, deforming the wheel and the track. Stand next to a train as it goes by, you can feel the ground move. All that bending/unbending of the track and bed contributes to fatigue damage, and fatigue damage goes up as the cube of the load.

> I find this hard to accept.

As did I at first, but it's true. There are several papers that show the math behind this (including one I co-authored called "Performance of ballast mats on passenger railroads: Measurement vs. projections") - unfortunately they are all behind paywalls.

> Just because there's a spring doesn't mean the weight isn't there, bearing down on the wheels, deforming the wheel and the track.

At rest, sure. Once it's in motion, the static load is insignificant (by several orders of magnitude) compared to the dynamic forces, and the dynamic forces are driven by the unsprung mass (and other factors such as the primary suspension frequency).

>Stand next to a train as it goes by, you can feel the ground move.

You're referring to ground-borne vibration (one of my areas of expertise). Again, that is a function of the dynamic forces, not the static forces. The FTA Transit Noise & Vibration Impact Assessment guidance manual (disclosure: I used to work for the firm that wrote the manual and I have a contract with FTA to teach cources based on material in the manual) touches upon this a little bit when it mentions that heavy rail and light rail trains have similar vibration levels (p 7-8, section 7.3, note that "heavy" and "light" in this context refers to train schedule not vehicle weight, but generally speaking heavy rail vehicles tend to be heavier than light rail vehicles) but again, all the good stuff is behind paywalls.

>All that bending/unbending of the track and bed contributes to fatigue damage, and fatigue damage goes up as the cube of the load.

That's a highway stat. Trains are different than roadway vehicles in that there is only one spring separating the vehicle weight from the unsprung mass. In trains, there are two suspension systems, the primary suspension with isolates the wheel/axle from the truck and the secondary suspension that isolates the vehicle body from the truck (and the wheels). When you look at the driving force behind ground vibration, the body weight has little correlation with the forces going into the ground (okay, there is a weak correlation in that heavier trains tend to have heftier axle sets, but the differences here between, say, the Acela and the TGV are minor).

I appreciate the detailed reply.

I am talking about the dynamic load. A moving train generates a standing wave that is a function of the train's weight and its speed. I don't see how adding a spring changes that. This standing wave bends the track and the bed, resulting in fatigue damage.

I know about the standing wave because in the early days of trains, there were problems because the standing wave propagated at speeds the locomotives were able to reach, and this resulted in mysterious destruction of the tracks and a derailment. (The solution was to stiffen the tracks so the wave propagated faster than the locomotive.)

Feeling the ground move down as the wheel trucks pass is dynamic, yes, but it is not vibration, as its "frequency" would be the rate at which the wheels passed by, and its amplitude would be proportional to the static weight.

Furthermore, there will be more localized bending as the wheels and track must deform so the contact area is enlarged to support the entire weight. (Assuming the same elasticity of the track steel.) This local deformation is going to produce cracking and wear of the surface of the rails.

>A moving train generates a standing wave that is a function of the train's weight and its speed.

No, it's really not. It's a function of the unsprung mass, the roughness of the wheel, the roughness of the rails and other factors (fastener type, faster spacing, etc). Train weight never enters in to the equation (literally), only the combined weights of the train components south of the primary suspension.

>Feeling the ground move down as the wheel trucks pass is dynamic, yes, but it is not vibration

??? That is exactly what ground-borne vibration is!

>its "frequency" would be the rate at which the wheels passed by,

Among other factors (wavelength of roughness, spacing of supports, etc), yes, and you can see that in train vibration spectra. You can also see the primary suspension resonance and other components. I'll refer you to the report "Vibration Characteristics of the Transrapid TR08 Maglev System" - (yes, I was a coauthor) were we discussed these issues in the section about the TR08 force density. http://www.fra.dot.gov/downloads/rrdev/fra0206.pdf section 3.3. It's not a perfect analog with steel-wheeled trains but a lot of the same principles apply in terms of the different factors that go into the input force.

You mentioned "the early days of trains" - what era are you talking about? Specifically are you talking about the days before trains had suspensions, or had very stiff suspensions (which was common as late as the 1950's-1970's in transit vehicles until it's effect on vibration was well known - this was a huge issue with BART)? In that case I could very will see that vibration was proportional to gross weight since a stiff suspension is basically a short circuit to ground. That generally isn't true any more, the only exception being some types of freight cars.

edit: a quick Google search provided a paper that supports my thesis: http://library.witpress.com/pages/PaperInfo.asp?PaperID=1957...

I'd really like an explanation of how, if you push down on the track at a point, you are not bending the track down at that point, sprung or not. Move that point along the track, and you create a standing wave, sprung or not. The amplitude of that wave will be proportional to the total weight.

It's not that the train isn't pushing down on the track, it's that the force created by the unsprung mass is several orders of magnitude higher than the static force. The wheelset force is higher than the static force because the primary and secondary suspensions allow the wheels to move independently of the vehicle body.

Imagine you have a 2x4 wood beam suspended across two blocks, with a 200-lb man standing in the middle of the beam - the beam will clearly sag in the middle. But if you place a 50 lb girl jumping up and down on the beam next to the man, the beam will sag significantly more. If you are looking at the effects of forces on that beam so see how it bends or breaks, it's the dynamic forces from the jumping girl that will control things, not from the heavier guy.

At this point I don't know how to explain things without jumping into two degree-of-freedom mechanical models.

But the reason those rules exist is because whenever there is a grade crossing fatality, inevitably the next of kin sue and all too ofter win in court.

It sounds to me like we need a law saying that if a driver drives on to the tracks when a crossing arm is down or the red lights are flashing, that's prima facie evidence that any resulting crash is the fault of the driver. Don't play chicken with a train.

Anything that is not a train that is on railroad tracks is prima facie evidence that fault doesn't lie with the train since trains are the only thing supposed to be on the tracks.

But it doesn't matter. You then get into questions like "were the lights flashing?" "Were the gates down?" "Did the train operator blow his horn? Was it loud enough? Was it long enough?" "Where there sight-line issues between the car and the gates/lights/trains?"

In essence the question is usually "did the railroad do absolutely everything in it's power to prevent the train from hitting something that is trespassing on the tracks" and the mere existence of the grade crossing is evidence that the answer is "no" since the railroad could have spent a million dollars and grade-separated the crossing. And there is your liability.

trains are the only thing supposed to be on the tracks

That's not the case at crossings.

"were the lights flashing?" "Were the gates down?"

Whoever bears legal responsibility for the crossing should be able to show that the systems were in working order.

"Did the train operator blow his horn? Was it loud enough? Was it long enough?"

This shouldn't matter any more than it should matter if I sounded the horn in my car when someone else ran a red light.

"Where there sight-line issues between the car and the gates/lights/trains?"

If you drive faster than you can see and react to hazards, you are negligent and at fault for any resulting collision.

With these things in mind, I'm in favor of a law that automatically holds the driver responsible as long as the flashing lights or gates were working properly. Nobody else should be held responsible when you do stupid things with your car.

The solution to this is not to increase horn noise, but to simply pass a law that states that people can't sue for wrongful death at a train crossing due to lack of safety standards as long as it meets FRA regulations or whatever.

This sort of limiting has been done in other circumstances. You will never solve this problem by changing the regulations, lawyers can always think up new things you should have done when they are paid to :)

Awesome! Thanks for posting, really interesting to see what you have to say about working on the Acela.

Good point as well about unsprung weight - hadn't thought about that. As the Acela is that powerful (it is based on the TGV Atlantique I believe), what in your opinion is the worst issue that set has faced? I really think that if we could buy off the shelf designs (Shinkansen N700 for example) with minimal adaptation, that would solve a large number of problems. (I'm also more a proponent of HSR with EMU's instead of power cars).

I do agree about our nation being unwilling to invest, but I also would add that it is a matter of density. Average travel distance via any HSR line in the US will be much longer than any other nation on the planet, making it much less attractive. There are some corridors which would see benefit, but because of the sprawl that most non-northeast cities have, a train doesn't provide you a realistic way to get from point A to point B outside station-to-station. You still have to get from the platform to your final destination.

> what in your opinion is the worst issue that set has faced?

Well the crappy infrastructure (shared ROW, lots of stops, lots of crossings, 100-year old catanary south of NYC, etc) is by far the most detrimental issue to Acela performance.

With regard to the actual Acela vehicles itself: from what the insiders have told me, the worst issue the set has faced was, frankly, the selection of Bombardier/Alstom in the first place. I was told that when FRA auditioned the various HSR vehicles on the NEC (they actually paid to bring over a bunch of European trainsets to the NEC for testing back in the late 80's or early 90's), Siemens told them that their vehicle (ICE) could meet most of FRA/Amtrak's spec with a little work but not everything. Bombardier/Alstom said they could meet all of the performance requirements with no problems. After getting the vehicle contract, Bombardier/Alstom started to agitate about not being able to meet large parts of the specs which led to compromises in the design and delays. By the time FRA/Amtrak realized they made the wrong choice, they were in too deep.

Again from what I've heard, if Amtrak/FRA had it to do over again, they would have gone with Siemens. It will be interesting to see what happens when it's time to replace the Acela.

Interesting, didn't know about Siemens bid being much closer to the mark. While we can't say that they wouldn't have had problems, I always figured that Bombardier/Alstom really had botched it, especially with the issue where the carbody was too wide for full-tilt.

Do you think the same problems will present themselves on the SEHSR corridor[1]?

[1]: http://www.sehsr.org/

I suspect that the density problem in the US isn't as bad as in Australia.

Very true. As one person I know put it, "Australia is a few cities separated by an ocean of largely useless land."

That Australia has managed to not only grow but thrive despite this is a testament to the people there. One great example is the Pilbara ore mining operations (BHP, Fortesque, Rio Tinto), that run some of the heaviest trains in the world. Even more interesting is that while the national Australian network has a tighter loading gauge and stricter load requirements, the Pilbara railways have the heaviest axle-load in the world at 230+ tons per car. For reference, in the US most trains are limited to 120 tons and in China 80 tons.

Those Pilbara trains are robotically driven, too.

Actually they aren't - they did try though. Currently they use a single-operator ruleset.

Locomotives have a refrigerator, hot plate, coffee maker and stereo with CD player and iPod hookup for the crew member. There's one on YouTube who has done cab rides all the way from the mine to Port Hedland. I think he is rather bored :D

The purpose of government should be to ensure that the minimum expectations of its citizens are met. These minimum expectations are a changing set of things, and I think that better feedback between government agencies and those which are affected by the rulings would be one of the best ways to solve the country's problems.

I think that http://data.gov and http://recovery.gov are a step in the right direction. Experts should analyze the data and blog about it, and the government agencies should be keeping an ear out to what experts are saying. Also the interested public can do the same.

The word "minimum" that I use is not accidental. The problem is that government rarely solves just the minimum set of problems. Once an agency exists, among the new employees there are always those who want to make their mark, and increase the amount of regulation. This is how government grows and grows. It's free for them to regulate but not free for those who have to implement it, and thus they don't feel the right incentives at the time. We need to figure out a way to incentivize government to stick as much as possible to only enforcing minimum regulations. Maybe it can be done by requiring them to get the citizenry to clamor for something before they implement it.

> The purpose of government should be to ensure that the minimum expectations of its citizens are met.

That seems like the purpose of a welfare state.

If you want a government that sticks to the 'minimums' you're asking for, I'd redefine that purpose statement to something like:

The purpose of government is to ensure the natural rights of its citizens, and nothing else.

I've heard that a lot. The problem with this kind of formulation is that "rights" is a loaded term, like "intellectual property". It implies that there are some universally recognized rights that come from nature, and only those. For example, if "intellectual property" was a natural right, then where does the 20 years come from in patents? That number 20, or the length of copyright, show that it's society's conventions and not rights.

Instead, I treat everything as "expectations". Thousands of years ago, most people expected to hunt or farm if they wanted to eat. 500 years ago you didn't expect almost every child to survive past childbirth and the early years. Now we expect all these things and more. Clean water. Food that doesn't contain killer bacteria. Buildings that don't have obvious structural problems.

Obviously we can't ensure things 100% but we as a society can work to set up organizations to regulate and oversee these things. A free market allows anyone to start cutting corners at any time, such as a new and inexperienced company that wants to make things cheaper. For example, a real estate developer can build a building that's not up to code. In theory, we could all walk around with our own subscriptions to private "auditing agencies" like Zagat for buildings, that would inspect buildings as they are built or whatever. But most of us outsource this kind of stuff to a centralized government. Which we vote for. Also, government could tax us to do unprofitable stuff such as the US Postal Service.

However, the problem is that our government lacks incentives to stop growing, and we need to introduce them.

I define "rights" as the freedom to do whatever you want with your body and property as long as you don't infringe on another's freedom to do the same.

But as you say, "rights" and "expectations" are slippery slopes.

> if "intellectual property" was a natural right

I don't believe this is a natural right. If anything, nature shows us that there is no such thing as intellectual property. Once an idea is "out in the open" it's out there for good.

> Also, government could tax us to do unprofitable stuff such as the US Postal Service.

But this is a terrible thing, is it not?

I'd guess at least 85% of mail is junk mail. This is only possible because mail is so cheap and the post office runs at an annual loss in the billions.

Eliminate the post office. Let UPS/FedEx/whoever compete to deliver your mail. Pay $3 to mail a letter. Save trees. Eliminate junk mail.

> However, the problem is that our government lacks incentives to stop growing, and we need to introduce them.

I agree.

I'd start with not allowing the government to print its own money. As long as it can, it will never have an incentive to stop growing.

I am not saying that "expectations" are slippery slopes, it is just the word I use to describe the situation, much like I use "government granted monopoly right" instead of "intellectual property right".

I am just observing what governments do and what their purpose is supposed to be. Once we can agree on what their purpose is, then we can decide on what they should be doing, and perhaps how to make them more efficient at doing it.

People today expect more from their governments than 2000 years ago. If we were talking about "rights", they would be the same throughout time. But it's more mundane than that.

Anyway, "ensuring that the minimum expectations of the citizens are met" is not just for welfare states. It is for ALL governments that serve their people.

If you are advocating increased independence for the Federal Reserve (discussing the U.S. case), are you concerned that it will then grow itself to perpetuate the organization as discussed elsewhere in the comments on this post?

More likely the eradication of the Fed.

The Ron Paul solution? How do you propose to move from the current situation to a non elastic currency backed by a combination of commodities, in a way that doesn't involve hurting a lot of people? And how would it be better?

Life is a natural right. Ensuring that I do not die in a train crash (or because I cannot afford health insurance, or because the factory I work at takes no safety precautions, or because the market choses crash-prone passenger planes, or because some idiot who doesn't believe in spectrum ownership jams my 911 call) all fall within the umbrella of protecting life. Yet these are decried as examples of an overly invasive welfare state.

If Caltrain did a symbolic disconnect, cut off several feet from the track, put up a barrier, a few days work to do and undo. from the national rail network, would they still be subject to the FRA?

In Caltrain's case, they actually got an interesting waiver that allows them to time-segregate the network, where it's part of the national rail network (and can accept freight traffic) at night, but cut off from it (with no freight traffic, and interlocks making sure it's disconnected) during the daytime. In that setup, it would have permission to run non-FRA passenger-rail equipment: http://wilshirevermont.com/2010/05/31/caltrains-amazing-waiv...

However I believe implementation of this is on hold, because they want to coordinate any upgrades with the HSR project.

They've also discussed completely cutting it off from the national rail network at times. Unlike in many places, Caltrain actually owns the rails (rather than a freight network owning them), so that would be possible. There's strong lobbying against it from various landowners along the line, though; even though the line doesn't get a lot of freight usage, the possibility to receive at least an occasional freight shipment increases the value of the land.

Yes Caltrain (actually the Peninsula Corridor Joint Powers Board that owns the track) has a clause in their contract from when they bought the line from Union Pacific that states they can start freight abandonment proceedings if they make a significant change to the line. It's generally believed that was put in there in case they wanted to convert the line to BART, but in theory it could be brought into play for a Caltrain electrification or the HSR project.

Caltrain has so far, however, refused to go down that route. Apparently the one freight train per day is too much to lose.

The article says there's "zero possibility of a passenger train colliding with a freight train" because "there are virtually no freight trains running on the Caltrain line". However, since I had to wait for a freight train last week at the Redwood City Caltrain crossing (and I hardly ever use a grade crossing), I conclude that this article is making stuff up.

Yes, this article is remarkably inaccurate.

Those Trinity Rail Express cars are now used for the Denton-Dallas A-train commuter rail, soon to be replaced with new Swiss-made cars. I like the old ones actually, extremely comfortable and seats like a massive couch.


I can asssure you, the new ones will be quite comfortable as well ;) Granted, they might not have couches as seats but they are very spacious and bright. I like them quite a lot... (They are in use since a few years already here in Switzerland.)

Modern suburban and regional {D,E}MUs are generally becoming more like trams or light rail vehicles in their design and, sadly, even interior. Lighter seats and so on. It's a general trend in Europe as well.

The one issue this article doesn't address is what is being done about the FRA. Are there movements to change the regulations? Overhaul the FRA? Even something as small scale as what, if anything, the author is trying to do about it would have been nice to hear.

I wonder if they have regulatory authority over monorails or maglev?

What this article talks about is only for trains that run on tracks used for freight...so that wouldn't apply to monorails/maglevs...

Monorails and Maglev are remarkably expensive and inefficient, so unfortunately I doubt they'll ever be a viable alternative.

I'll agree with you on maglev at present, but monorail has gotten a bad rap from some false assumptions (e.g. switching) and some early problems.

I think these numerous regulatory agencies are a way of circumventing popular will (read: write mandates handed out by interest groups). Whatever happened to Obama's promise that he would weed out regulations that harm small businesses (I think they would benefit the most from the increase in foot traffic that would result if public rail service were more prevalent).

Can we add "in the U.S." to the title?

The FRA only has jurisdiction in the US.

We have an FRA in Sweden too, except it's our NSA-like thing.

Can we expand the acronym, perhaps?

what strikes me as odd is that one of most cost-effective safety devices, the set built, is unavailable to most passengers on buses or trains.

What is a "set built"? Do you mean "seat belt"? If you do, then you should know that seat belts do nothing on vehicles of that size.

I think I was half asleep when I wrote that :-) wtf

But I'm not sure I understand your point. It seems when large vehicles stop suddenly, unsecured cargo doesn't. Airplanes have them don't they?

I had heard something suggesting seat belts are not cost effective on trains and buses for the amount of lives saved. Perhaps that's what you're thinking of. Still, I would rather have them even if it cost an additional tenth of a cent per mile.

The honking is driving me nuts. They're required to hit the horn four times per crossing. Sometimes it's a series of quick blasts. Other times, they'll lay on that thing for -- I swear -- five full seconds per honk. Midnight, 2am, 3am, they don't care. I'm around a mile away from the track, and I can't imagine what life could possibly be like for those who live closer to the track.

Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact