Why are train tracks flat?

We used to have a tram/streetcar that did that in The Hague but apparently now they use trucks that can go on the rails for that.

We do have an old PCC car that roams the tracks around the city that measures the state of the tracks with lasers, though, the “meettram” (measurement streetcar, blog post in Dutch + photos).

I assume similar questioning led to the discovery of our planet being round…

The original engineering proposal was to put tracks across the Golden Gate Bridge under the roadway, and a great deal of effort was expended in the 1960s and 1970s to prove that the bridge had sufficient extra structural capacity to support the additional loads. The design constraint that determined the track gauge was that a three-car train of cars weighing 50,000 lbs each and traveling at 80 mph be able to withstand a 94 mph crosswind on the bridge. The design engineers at the time determined that the track gauge of 5’6" was required to insure stability and prevent derailment or overturning under these operating conditions. The cars ended up weighing 62,000 lbs, the lightest car per passenger of any transit system in the world.

I’m not sure why you place such extraordinary importance on track gauge. All railcars in the US are custom-built, so the only things BARTD can’t order out of a catalog—that other transit systems can—are frogs and turnouts. Many, many transit systems around the world use gauges other than 1435mm. Streetcar systems in Baltimore, Pennsylvania, New Orleans, and LA used other gauges, as do rapid transit systems in Toronto, Philadelphia, and Tokyo. Around the world, you’ll find lots of transit and suburban operations on meter gauge or 1067mm (Japanese standard), and plenty of wide-gauge operations in Spain, Russia, and India.

I appreciate the additional detail, but really–they were designing for a 94 mph crosswind? Wind speeds have never exceeded 75 mph on the bridge, and that was a freak condition that closed the bridge for several hours. They made a huge sacrifice in the system design to just get one more 9 in the reliability figure. It makes no sense.

I realize that track gauge isn’t entirely standardized. But nobody else in the US uses this gauge. Even Pennsylvania trolley gauge is used in more places.

If gauge were just an irrelevant detail, then the eBART extension would not have switched to standard gauge. But they did so because it’s cheaper. The new trains are used all over the US and Europe and are available in 1435 and 1000 mm gauges.

Although this isn’t strictly related to rail gauge, the all-custom nature of the BART system has meant they’re going to eBay for spare parts. It’s ridiculous.

Apparently the gauge isn’t the only non-standard aspect of the BART. It actually uses cylindrical (non-tapered) wheels which has been causing problems.

It is in fact the exact opposite of a car riding in a V shaped trench. The principal is that used by flat belts running on crowned pulleys.

And yes, train wheels are pressed into the axles, and can’t rotate individually. As mentioned above, this feature is critical to proper centering and cornering without flange contact.

Huh–that’s a new one to me. It certainly explains the banshee screech the trains make. The article says the new replacement cars will have tapered wheels–I wonder if the existing flat track profile will pose any problems there.

An article linked from the one you posted says the replacement cars cost double that of conventional ones. Obviously those costs aren’t all due to gauge or wheel profile (the electricals are wonky as well), but all of those things together indicate an extreme “not invented here” syndrome among the designers.

In general, aren’t we seeing the whole rail-based transportation industry moving towards more uniformity? It used to be that the Dutch railways and the tram companies in the big cities largely designed their own rolling stock, but these days they buy off the shelf with relatively minor customizations.

Just watched. Learned some things, but you’re correct.

It’s been going on for more than a century. Thousands of miles of track were converted from broad gauge to (almost) standard gauge in the US in a 36-hour period starting on May 31, 1886.

It makes sense to have a few different gauges since trolley/logging/etc. tracks have different requirements, but like virtually everything else in a modern economy, standardization has tremendous benefits.

Saying that standardized gauge won the Civil War for the North would be a massive exaggeration, but there’s a bit of truth to it, too.

No, I completely disagree. The principle (nit: not “principal”) of a flat belt on a crowned pulley is that longitudinal and lateral deformations of the tensioned belt tend to make it ride up and center itself at the top of the crown and keep it there. There are no such deformations when tapered steel train wheels roll over rails. What keeps them centered is essentially gravity, because the centered position is the most gravitationally stable lowest-energy state, the same thing that would keep a car on a V-shaped road from “climbing” the sides, and instead staying put at the valley in the center. It’s not a perfect analogy but it’s close.

Ironically, the belt example seems counter-intuitive because the center of a crowned pulley creates greater tension in the belt than if the belt slipped laterally off to the side, hence a higher potential energy state. It’s the deformations that bring it back to the center, and the higher potential energy state is drawn off from whatever is driving the pulleys. So really that’s closer to being the opposite of what keeps train wheelsets on the rails.

No, that’s not true. What keeps them centered is the counter-steering caused by the varying diameter across the width of each wheel.

Gravity could actually have the opposite effect as what you’re describing. If the left wheel has ridden up 1 mm on its bevel, then the right wheel must have ridden down by 1 mm (roughly speaking). So the center point on the axle is the same vertical distance from the earth. However, the train is now tilted, which means that the overall center of gravity (which is above the axle) is now lower, and thus in a lower energy state. In other words, gravitationally speaking, the train is in an unstable equilibrium.

Yes, thank you, that’s an excellent explanation and I humbly accept the correction.

So I guess what Kevbo was getting at with what was really a very rough analogy with the flat belt and crowned pulley was that in both cases there are mechanical forces acting to center the device in question. I accept that, too, but I maintain that the forces are very different and arise from different causes.

Agreed. The resemblance is just that in both cases the object gets “steered” back to the center due to the geometry, but in the case of the belt it’s elastic deformation at work, while in the case of the train wheels it’s the varying wheel radius. As you say, though, it’s a very rough analogy.