This is an excellent science fair project. This is not some hidden solution to derailment. The article is devoid of details other than the student observed great variance in truck spring condition and did extensive hypothetical testing to show it can contribute to derailment.
A contributing factor to both frequency and severity of derailment, sure, but a rare initial cause. “Springs” are not outside the considerations of the FRA. If you ever have trouble sleeping, read an FRA derailment report of some non-catastrophic event. Back in the 90s, they went so far as to identify dampening pads - thin plastic plates in the trucks - as the cause of derailment for new heavier intermodal cars. 1" thick plastic that deformed 1/8" under the 250,000lb+ load hit a harmonic frequency that led the trucks to hunt (wobble left and right but never finding a steady center) until the wheels hopped the tracks when entering a curve.
There is a massive test facility in Pueblo, Colorado that has the smartest engineers on the continent working and testing there. There’s straights, curves, a super loop, road crossings, bridges, and even a concrete wall for crashing. They pull rail cars wired more than your road car all to find something, anything in the data. So can resonant frequency be a big contributor to derailment? Yeah, of course. A push becomes a shove 3 cars later and hits some limit. But there’s absolutely no way to predict or control it because every car, ever load, and every piece of track is going to have some unique critical frequency.
And, for the record, the vast majority of those “1300 derailment per day” are low speed (which the article mentions) inside yards when assembling the full train (which the article does not mention). And, while the individual root cause of other derailments is varied, a majority of them are triggered by hard braking. It takes a few minutes for the brake signal to reach the end of the train - it’s a system operated by air pressure that’s used for both signaling and applying brakes. This means when the locomotive slams the brakes, the back of the train is still trying to shove it ahead at full speed long after the locomotive passes the engineer’s original line of sight. This force can be so great it makes the rail roll out form underneath the train. Add in coupler cushions and the train can shorten itself by a few hundred feet under this pressure, shoving couplers to the sides, taking out all the cushion slack, and adding slack towards the front as the middle and rear cars apply the brakes, creating a slinky longer than a mile weighing 140 tons per link. And yes, they’ve certainly “figured out” that empty cars in front of loaded cars is a huge contributing factor to these forces, there’s nothing easy about staying competitive. This platform shows it’s well aware of how much ground rail is losing to truck transportation. By rate and by severity, trucks vastly outweigh trains in terms of damage. The problem is rail failure is much more catastrophic and gets more news. 1 fiery derailment is news. 100 fiery trucks are a statistic. Same as planes. Way safer per passenger mile, way more deadly in failure.
And why do brakes perform so poorly? Because the benefits that come from “interchange” of rail cars between trains, lines, and companies, also comes with a massive interchangeability headache for any changes. I’m not defending the rail industry as if they’re innocent bystanders, but we should all be able to understand the difficulty in trying to convert 1.4 million freight cars to a new system for the first time in, literally, 100 years. Many of these cars reach 50 years of service before being scrapped. Someone owns the rail, someone owns the locomotive, someone owns the freight car base, someone might own the freight car body, and someone owns the cargo. These are often entirely different entities for each. You can’t really do a soft rollout, it’s all or nothing. So here, we sit, with centurion technology. Ironically, the only type that has the same loco and cars every day is the least progressive load - coal trains. It’s runs from the coal processor to the coal plant and back, nothing else. Every other freight train can be diced and hashed multiple times a day.
It’s not all doom and gloom. The East Palestine derailment put “wayside detectors” into the general public’s lexicon. They were talking about relatively simple infrared sensors placed next to the tracks to look for hot bearings. That’s old tech. There are currently massive sheds being implemented filled with cameras that record terabytes of info for each train passing through. It can identify missing bolts, cracked springs, and other failures on the spot. Control is notified and if it’s urgent, the car can be routed to a repair shop or the train can be stopped. There’s even a type with a trolley that rolls along with each truck to image the entire circumference of the wheels, looking for chips and cracks. A stall in the rolling tech doesn’t mean a stall in the industry. Make no mistake, derailment are expensive and companies and trying damn hard to sell solutions to the rail lines.
Passenger trains have their own headaches. Practically each line has its own design engineers reinventing the wheel. While some lines may buy existing designs (notable, the Amtrak Acela that runs from DC to Boston is a French TGV), they have their own design flares (such as doubling the number of trucks for the Acela). These trains do run as consistent units, so they tend to have intercar communication systems and hydraulic brakes, minimizing the overall braking time. That’s why their derailments usually come from speeding or collisions rather than “random” accidents.
I agree that the title is misleading and I couldn’t believe that nothing was done to analyze the problem but you explained it pretty well.
Thank you for explaining, all this is very interesting!
This is an excellent science fair project. This is not some hidden solution to derailment. The article is devoid of details other than the student observed great variance in truck spring condition and did extensive hypothetical testing to show it can contribute to derailment.
A contributing factor to both frequency and severity of derailment, sure, but a rare initial cause. “Springs” are not outside the considerations of the FRA. If you ever have trouble sleeping, read an FRA derailment report of some non-catastrophic event. Back in the 90s, they went so far as to identify dampening pads - thin plastic plates in the trucks - as the cause of derailment for new heavier intermodal cars. 1" thick plastic that deformed 1/8" under the 250,000lb+ load hit a harmonic frequency that led the trucks to hunt (wobble left and right but never finding a steady center) until the wheels hopped the tracks when entering a curve.
There is a massive test facility in Pueblo, Colorado that has the smartest engineers on the continent working and testing there. There’s straights, curves, a super loop, road crossings, bridges, and even a concrete wall for crashing. They pull rail cars wired more than your road car all to find something, anything in the data. So can resonant frequency be a big contributor to derailment? Yeah, of course. A push becomes a shove 3 cars later and hits some limit. But there’s absolutely no way to predict or control it because every car, ever load, and every piece of track is going to have some unique critical frequency.
And, for the record, the vast majority of those “1300 derailment per day” are low speed (which the article mentions) inside yards when assembling the full train (which the article does not mention). And, while the individual root cause of other derailments is varied, a majority of them are triggered by hard braking. It takes a few minutes for the brake signal to reach the end of the train - it’s a system operated by air pressure that’s used for both signaling and applying brakes. This means when the locomotive slams the brakes, the back of the train is still trying to shove it ahead at full speed long after the locomotive passes the engineer’s original line of sight. This force can be so great it makes the rail roll out form underneath the train. Add in coupler cushions and the train can shorten itself by a few hundred feet under this pressure, shoving couplers to the sides, taking out all the cushion slack, and adding slack towards the front as the middle and rear cars apply the brakes, creating a slinky longer than a mile weighing 140 tons per link. And yes, they’ve certainly “figured out” that empty cars in front of loaded cars is a huge contributing factor to these forces, there’s nothing easy about staying competitive. This platform shows it’s well aware of how much ground rail is losing to truck transportation. By rate and by severity, trucks vastly outweigh trains in terms of damage. The problem is rail failure is much more catastrophic and gets more news. 1 fiery derailment is news. 100 fiery trucks are a statistic. Same as planes. Way safer per passenger mile, way more deadly in failure.
And why do brakes perform so poorly? Because the benefits that come from “interchange” of rail cars between trains, lines, and companies, also comes with a massive interchangeability headache for any changes. I’m not defending the rail industry as if they’re innocent bystanders, but we should all be able to understand the difficulty in trying to convert 1.4 million freight cars to a new system for the first time in, literally, 100 years. Many of these cars reach 50 years of service before being scrapped. Someone owns the rail, someone owns the locomotive, someone owns the freight car base, someone might own the freight car body, and someone owns the cargo. These are often entirely different entities for each. You can’t really do a soft rollout, it’s all or nothing. So here, we sit, with centurion technology. Ironically, the only type that has the same loco and cars every day is the least progressive load - coal trains. It’s runs from the coal processor to the coal plant and back, nothing else. Every other freight train can be diced and hashed multiple times a day.
It’s not all doom and gloom. The East Palestine derailment put “wayside detectors” into the general public’s lexicon. They were talking about relatively simple infrared sensors placed next to the tracks to look for hot bearings. That’s old tech. There are currently massive sheds being implemented filled with cameras that record terabytes of info for each train passing through. It can identify missing bolts, cracked springs, and other failures on the spot. Control is notified and if it’s urgent, the car can be routed to a repair shop or the train can be stopped. There’s even a type with a trolley that rolls along with each truck to image the entire circumference of the wheels, looking for chips and cracks. A stall in the rolling tech doesn’t mean a stall in the industry. Make no mistake, derailment are expensive and companies and trying damn hard to sell solutions to the rail lines.
Passenger trains have their own headaches. Practically each line has its own design engineers reinventing the wheel. While some lines may buy existing designs (notable, the Amtrak Acela that runs from DC to Boston is a French TGV), they have their own design flares (such as doubling the number of trucks for the Acela). These trains do run as consistent units, so they tend to have intercar communication systems and hydraulic brakes, minimizing the overall braking time. That’s why their derailments usually come from speeding or collisions rather than “random” accidents.
I agree that the title is misleading and I couldn’t believe that nothing was done to analyze the problem but you explained it pretty well. Thank you for explaining, all this is very interesting!