Catastrophic bearing failures in trains, on the run.

Bearing failures are almost always caused by lubrication failure, but in any case overloading, dirt or moisture incursion will lead to vibration and overheating, so this is your first line of defense, monitor vibration level and temperature, and you should be alerted to an imminent failure... this allows time to prepare for shutdown and repair, the time you need to prepare, governs the sensitivity of the monitors...ideally

It seems to me that you could buy a really good sensor network for the cost of one train derailment.

Just sayin'

I agree. What has always puzzled me is the insurance carriers. They take the brunt of the derailment expenses most of the time. Most of the derailments aren't as bad as the one in Palestine Ohio. Some money to improve sensors to prevent a derailment in the first place would be in these bean counters' interest.

Insurance companies!! They wont pay for preventive maintenance, in my case taking down a dead tree in my front yard, but they think nothing of paying for the damages to my house if the tree drops on it.

You have to have these sensors (hot bearing detectors, or HBDs) on every line that every train will run on, ensure good communication from the HBD to the control system/train driver, and space them suitably to ensure that they catch every faulty bearing. This is expensive to install (~US$200 000 each on a single track) and to maintain. So I assume most railways try to strike a balance - if it really was that cheap, railways would have done it. They're in the business of making money, and, in addition to the reputational issues resulting from the safety & environmental consequences, derailments are often very disruptive to their business. Having said that, newer developments on wagon-fitted sensors are maturing, and may be prioritised for wagons carrying hazmat payloads, for example.

That is the claim but I find that claim to be disingenuous.

For any bearing to get hot enough to be detected through all of the steel of the bearing housing, the axle has to be spinning for a while. Thus switchyard and side rail track locations are unsuitable for any off-the-train thermal measurement of bearing temperature. While this proposed configuration does limit the number of locations an HBD can be effective (a laudable cost-saving approach) it also guarantees the expensive communication needed for separation between bearing and sensor along with a separation between sensor and operator. Applying a simple nonreversible data logging indicator (color-changing ink or crayon) to each axle end or bearing housing will allow for low-cost imaging to show if the bearing overheated at any time and not just when it rolled past an expensive high-speed thermal imaging camera.

Still, any overheating bearing detection is not preventative. An overheated bearing is a damaged bearing. Attaching an ultrasonic contact microphone to each bearing housing allows for measuring bearing health long before significant damage occurs. That measurement need not be continuous and perpetual but can be part of a six sigma rolling inspection procedure.

That's universal problem in many fields of life, specially where safety is involved,

Other than any type of sensing and early detection, Scheduled inspection and preventive maintenance is best way to avoid catastrophe.

A simple or serious derailment, besides the major causes due to problems with the tracks, the broken axle on the carriage is also vulnerable. Prediction and prevention are ongoing challenges. My question is, if a bearing failed on the run, beyond a manageable extent, what would be possible action? Would ‘no action’ lead to derailment?

If a bearing failure is detected the train idles to a parking area and disconnects the car and leaves it for repair...No action, could start a fire, damage to tracks, or derailment of part or all of the following cars...

Shall I slightly modify your words, "No action, or not much time to act, could start a fire, damage to tracks, or derailment of part or all of the following cars." Like that happened in Ohio. I am imagining a scenario, wherein a passenger train suffered bearing failure, was caught by hot-box detectors and the train stopped. Will its bearing be changed in situ?

That depends on the design of the wheel assembly but I doubt a bearing could be changed in situ and still be affordable. What might be possible is spare rolling supports of the bearing for the axle, a kind of spare tire. This would likely not last very long so the tedious wheel replacement Solareagle posted would still need to be done but getting safely to the next side track would be plausible.

The problem with any temporary fix is the temptation to keep using the cheaper temporary fixes until something really bad happens.

https://www.atsb.gov.au/publications/investigation_reports/2020/rair/ro-2020-017

..."There’s no question that hotbox detectors work. Train accident rates caused by axle and bearing-related factors have dropped 81% since 1980 and 59% since 1990 due to the use of hot bearing detectors, according to the FRA.

But bearings can overheat quickly, the FRA says, and can burn off in as little as 1 to 3 minutes. Hotbox detectors failed to diagnose 124 severely defective bearings in the U.S. and Canada from 2010 to 2018, 117 of which resulted in derailments."...

https://www.trains.com/trn/news-reviews/news-wire/wheel-bearing-expert-to-prevent-derailments-railroads-should-equip-freight-cars-with-sensors/

Causes of train derailments...

https://www.coganpower.com/blog/2016/march/off-the-tracks-the-reasons-trains-derail/

"There’s no question that hotbox detectors work"

That's only true if the people in charge use that information in a timely fashion. I understand that surveillance video shows flames coming from that Ohio train car at a location around 20 miles prior to the actual derailment, and that a hotbox sensor did alert somone aboard that train.

There's also the physics of momentum. With light cars in the middle of the train and heavy cars near the end, when the front of the train slows down rapidly, the heavy cars near the end keep moving, causing the light cars to buckle.

Why not start applying the brakes in a sequential fashion,starting at the rear and going forward?

This would prevent backlash and cascading into the car ahead,would it not?

Machines that make fabrics,and other materials,paper,etc, always braking from the rear,and delayed starting from the rear to prevent wrap up.Some delays on machines are only a few milliseconds apart,but that means a lot in some cases.

I'm pretty far from a train expert, but The last I knew, the brakes on RR cars were released by air pressure, and applied by a lack of air pressure. Every car had a hose connecting its air supply to the adjacent cars.

I assume that the brakes are applied by opening a valve to let the air escape at the locomotive. If that is the case, clearly the cars nearest the locomotive will lose pressure well before the cars at the end of the train. How long does it take for the air in the 150th car to pass through all the other cars and escape at the locomotive?

I'd be very happy to learn that there is some method of electronically controlled braking for trains, so they could apply the brakes at the rear first, but since they eliminated cabooses, I suspect that is not possible.

The braking system used on trains is more complicated and studied than many realize. The compressed air driving the brakes is controlled pneumatically and/or electronically.

These systems are braking of empty or loaded vehicles according to load, anti-blockage braking system to protect wheel slip, monitoring of driver awareness, remote controlled emergency braking, emergency braking systems with passenger control, wheel slide protection devices.

This is far more complex than I presumed. Proper maintenance of such a complex system should be paramont.

"The braking system used on trains is more complicated and studied than many realize. The compressed air driving the brakes is controlled pneumatically and/or electronically."

I didn't realize that the words "braking system" in your post was a link, until I copied and pasted it here, and it turned red.. I'll go back and check it out.

As you indicate, a system that complex will require frequent testing and maintenance; it's probable that such testing and maintenance is far too infrequent.

Perhaps it's nit-picking, but I'm bothered by the phrase "The compressed air driving the brakes..." To me, that phrase implies that compressed air applies the brakes. Surely it is the absence of air that applies the brakes, and air pressure is required to release them! Please correct me if I am mistaken!

You are mistaken.Air must be applied to apply the brakes,unlike on OTR trucks,which are just the opposite.

Google:recent runaway train accident in Canada for a tragic example of air failure.

Contrary to the trucking industry, where air is required to release the brakes,the railroad does the opposite.WHY?It would be much safer to have air pressure release the brakes,would it not?Many runaway train accidents would have been avoided if this logic had been applied in the design of the braking system in the USA.

I am sure other countries may have had better sense when designing their systems by looking at the follies of our systems.

Our system requires the engine to be running and supplying compressed air to hold the brakes closed.If the engine is shut down,or fails,there are no brakes,except for the manually applied ones.

If this happens on even a gradual long downhill slope, it can be disastrous.

I have performed ultrasonic bearing monitoring for preventative maintenance,and ironically a new ball bearing sounds like a train on tracks at first but experience can detect the difference in a bad/ worn/vs good bearing.The process now is totally automated with a computer analysis doing the work, with alarm points programmable.

Perhaps it would help if the RR's had a similar sensor on the tracks. A hot bearing is in the last stages of it's life and imminent failure is close,so why not try to predict and plan ahead? That is the basis of Preventative maintenance.

"ultrasonic bearing monitoring for preventative maintenance, and ironically a new ball bearing sounds like a train on tracks at first"

Can ultrasonic bearing monitoring revel friction bearings as well?

By the way, are friction bearings still being used in new cars? Also, what is the rough ratio anti-friction bearing : friction bearings?

My experience in with ball and roller bearings only.

I do not know about"Friction bearings"Do you mean sleeve bearings, or static bearings,or sliding or rolling,or oilite or sintered bronze,or solid brass?

Yes, lined bearings (Sliding surface bearing with oil lubricated).

While I asked my previous question, in parallel I web-searched and found the following content from a leading bearing manufacturer:

Types of Railway Bearings

There are many different types of railway bearings, each designed for specific applications and operating conditions. Ball bearings are the most common type of railway bearing and are used in a variety of applications, including locomotives, passenger cars, and freight cars. The most common types of railway bearings are:

Roller bearings: these bearings use rolling elements (balls or cylinders) to support the load, reducing friction and allowing smooth operation. They are typically used in high-speed applications such as locomotives. However, they can be noisy and may require more maintenance than other types of railway bearings.

Plain bearings: these bearings rely on a sliding contact between the bearing surface and the journal (the rotating shaft). Plain bearings are often used in lower-speed applications such as passenger cars.

Magnetic bearings: these bearings use magnetic fields to support the load, eliminating friction altogether. They are typically used in high-speed applications such as electric trains.

Ball Bearings: Ball bearings are similar to roller bearings, but they use balls instead of cylinders. Ball bearings are quieter than roller bearings and require less maintenance. However, they can be less durable and may not last as long as roller bearings.

Tapered Roller Bearings: Tapered roller bearings are a type of railway bearing that is designed for use on curved rails. Tapered roller bearings have a tapered shape that helps them grip the rail better, which reduces friction and makes the train run smoother. However, tapered roller bearings can be more difficult to install and may require special tools.

Back to my question, can ultrasonic inspection figure out friction-bearings irregularities?

Another question: Are Magnetic bearings getting popular for railway applications?

With the proper sensors and software,anything is possible.

In an ICE,there is no direct contact between the connecting rod inserts or main bearings and the metal of the crankshaft,there is a pressurized oil film in between all of the bearing surfaces.When the oil fails there is metal to metal contact and imminent failure.A "Rod knocking: is a good example of a detectable sound from a failing bearing.

An engine that is cold can sometimes be heard knocking until the oil pressure builds up.

Some of the older large electric motors had sleeve bearings and an oil reservoir that had to be kept serviced regularly .

If it was kept full, and the proper oil was used,they ran quieter and longer than roller or ball bearings.

Some small motors still use sleeve bearings which are lubricated by a wick-like material saturated with oil from the factory.

Early bearing failure detection by Fast Fourier Transform analysis of the high-frequency sounds produced by the bearing is a long-studied field. The major advantage of this detection technique over thermal "hot box" detectors is the likely early detection, long before complete bearing failure.

This is such an old maintenance standard in many fields of mechanical machines that I am stunned that this is not already regularly performed on train wheel bearings. There might be something inherent in the existing train systems that prevent this test from being routinely performed. I don't want to make up excuses for a critical industry to not fix itself. Hopefully, it's not just greed preventing this from happening.

Sadly,the answer to "Why did they...." or "Why didn't they...." is usually: Money $$$$$.

Maintenance has always been considered the "red headed step child" by big business.

"Don't fix it till it breaks" is the unwritten,and unspoken motto,because down time is lost money.

GM once did a calculation that found that a human life is worth about $200,000 and could calculate the cost of a recall based on that figure.

https://www.beasleyallen.com/article/ivey-memo-how-much-is-your-life-worth/

IMHO:True sociopaths.The same as most,but not all large businesses.

I expect money to be the deciding factor here since everything in business comes down to money. However, future expenses resulting from any present-day financial decision can never be accurately codified. I, therefore, claim the right to hope some additional factor was part of the equation.

I agree,in part with what you said,but if they are estimated,they are always underestimated due to the $$ involved to properly estimate.

It is hard to prove a negative,but foreseeable future events are still controlled by $$.

There are many accidents that are caused by human error,of course,but as in the GM case,it was a purely financial decision to let people burn rather than to fix the problem.See my link above.

IMHO: The person responsible for the final decision to not take remedial action should have been tried and sentenced for the subsequent deaths.

Sociopathic,IMHO.

Sociopath:

What does sociopathic mean?

The term sociopath is used to describe what a mental health professional would diagnose as antisocial personality disorder. Symptoms may include disregard for others, a lack of empathy, and dishonest behavior.

IMHO:A good description of most big businesses.