Trains.com

FRA report on wheel failures and other topics

7025 views
47 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    April 2016
  • 1,447 posts
Posted by Shadow the Cats owner on Thursday, March 11, 2021 7:56 AM

From reading that report it sounds like a triple failure.  Improper metalurgy in the wheels.  To much deferred maintance waiting until they are just at the minimum before doing anything to fix the issue.  And the last one is not enough inspections to catch the problems before they fail.  The railroads have gotten way to reliant on technology to catch problems.  Wheel dynamic impact sensors to catch flat wheels hot box dectection by remote.  So instead of having people look at things with the mark 1 eyeball they go by what the computer says is good.  Doing things like that leads to massive problems down the line that tends to cost millions of dollars when they fail.  We have all heard it that the carmen are under pressure to get the trains out of the yard.  So instead of fixing them properly they say they are good and pray they make it to the next 1000 mile inspection point.  You are just lucky that PSR hasn't caused a major TIH spill in a major city yet.  The odds will catch up to the penny pinchers in the Boardrooms sooner than later and when it does I do not want to be a shareholder in the company it happens to be with.    

 

 

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Thursday, March 11, 2021 6:45 AM

Overmod
The Australians also figured out you could get safer performance and longer life by using a shorter roller bearing... something I did not see coming. It would not surprise me to find they have done their HAL homework and understand the wheel/rail system better. I would not be surprised to find ECP involved somehow.

Well, brake friction heat is said to play a role in the VSR wheel failures.  An air brake application must dissipate a specific amount of heat to the atmosphere.  If all brakes, during an application, set up simultaneously, each wheel should reach the same temperature in the same amount of time.  ECP brakes would set up simultaneously. 

However, conventional airbrakes will set up sequentially, and the ones setting up first, will get hotter during the brake application than the ones setting up later.  So, the earlier/hotter wheels will likely reach a temperature higher than the uniform wheel/braking temperature rise of the wheels braking with ECP.

I wonder if some type of relatively rare anomalies in the practice of setting up brake service applications might explain why these VSR failures occur consistently, but so infrequently compared to the number of wheels in service. 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, March 10, 2021 4:59 PM

The issue is likely not the 'work-hardened zone' in the affected tread, but the material cold-flowing under it.  Think of it as the counterpart to martensite breaking up into platelets on the railhead at 315K axle load... or induction of gauge-corner cracking and then propagation.

Yes, the reprofiling would have to extend out to the face, and yes, that might thin the rim. 

Look up the theory of the 'magic wear rate', which in theory would just get rid of SCC and other micro cracking or platelet deformation before cracks could turn in/down and propagate.  Some of the early rail-grinding discussions involved 'realizing' this rate by organized means.

The Australians also figured out you could get safer performance and longer life by using a shorter roller bearing... something I did not see coming.  It would not surprise me to find they have done their HAL homework and understand the wheel/rail system better.  I would not be surprised to find ECP involved somehow.

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Wednesday, March 10, 2021 3:59 PM

I don’t know if retarders would have much effect.  Some kinds of track impact may cause the actual breaking away of the wheel material, but the initiating cause is the development of the stress crack.

The basic fundamental cause is the development of the zones of tensile stress on each side of the zone of rail contact.   That happens without impact. Eliminating those stress zones seems like it would be a major threading of the needle in the manufacturing process.  As I understand it, the malleability of the steel wheel tread is apparently intended and quantified by specification.  But the tensile stress zones are apparently undesired and unanticipated. 

There is discussion about adopting the practice of periodic wheel truing by reshaping the treads by grinding. The grinding would cut deep enough to take the hollowness out of the tread face.  So it would cut deepest into the outer edges of the zone “compacted” by running under load.  So that would remove the most embedded tensile stress.  Actually, I would think the entire wheel tread, including the flange, would have to be cut down while maintaining the proper form until the hollow zone of the tread is eliminated.  But even doing that will leave the work-hardened loading zone completely intact at the center of the zone, and then diminishing toward its edges where more metal will be removed. 

Interestingly, Rio Tinto in Australia runs higher wheel loads than U.S. practice, and they have no failing wheels.  It is said that they run multiple use wheels and true their treads and flanges on a regular basis.   

The big mystery in this problem in U.S. and Canada is why such a small percentage of wheels are failing.  If you have so many wheels running in nearly identical use patterns, and built to nearly identical specifications; it is hard to understand how so few of those wheels are failing with VSR events.  

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, March 10, 2021 11:38 AM

How might this mechanism be affected by repeated passage through retarders?

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Wednesday, March 10, 2021 10:18 AM

VSR WHEEL FAILURE

 

The focus seems to be on a type of wheel failure called VSR for “Vertical Split Rim.”

The fracture separation plane falls on a vertical reference plane that is perpendicular to the axle of the wheelset and parallel with the rail.  This vertical reference plane falls within an inch or so from either the outside of the wheel or the flange side of it.

This vertical reference plane passes entirely through the wheel tread, although the fracture itself starts at one location on the thread where it intersects the reference plane and progresses around the tread circumference. 

The fracture may ultimately pass through the tread over its entire circumference and thus break off a solid ring as part of the tread from the near outside of the wheel or from near the inside, in which the broken off portion would be essentially the entire flange as a solid ring. 

Also, the fracture may not progress entirely around the wheel circumference and then release as s continuous ring.  Instead it may progress only part way around the circumference, and break off as just an arc shape.  The fracture may continue to propagate and break off additional arc shapes in succession.

This is my understanding of what causes this fracture:  Wheels and rail are malleable and so the pressure contact “compacts” the steel into a higher density near the surface.  It would be like squeezing a steel ball and thus decreasing its diameter as the total weight stays the same. 

The wheel tread is much wider than the rail head.  So the loading of the tread onto the rail is concentrated at a relatively narrow band of the wheel thread.  As the wheel accumulates mileage, the tread is compressed or “cold worked” into a recessed configuration that is referred to as being “hollowed out.”  But this hollowing is not caused by a wearing away of metal.  Instead it is caused by compacting the metal.  So the effect is a trough of compacted steel all the way around the wheel, centered on the wheel tread.

As the metal is compacted into that trough shape, it pulls on the wheel tread metal on each side of the trough where there is no loading of the wheel tread.  You can clearly see this effect if you push your finger into a chunk of bread dough.  The finger makes a dimple that draws the surrounding dough toward the dimple and into it.  So that area around the dimple is being stretched. 

So as the rail compresses the hollowed trough into the wheel tread, it produces a stretch (tensile stress) into the metal on each side of the trough.  Eventually, that stretching of material causes it to pull apart as a fracture, on one side of the trough or the other.  Then depending on which side begins to fracture, it will break off all or a portion of either the wheel flange or of the outside portion of the wheel that generally overhangs the rail.

In the case of breaking at the flange, this not just the standing flange breaking away from the wheel rim beneath the standing flange.  It is instead a breaking away of the entire flange height plus the portion of the wheel rim beneath the flange.  So the breakage is in the wheel rim, and the flange just comes off as a part of the broken off wheel rim. 

 

Article from Trains magazine:

https://www.trains.com/trn/railroads/the-mystery-of-vertical-split-rims/

 

Article from Railway Age:

https://www.railwayage.com/mechanical/freight-cars/wheel-failures-digging-down-to-the-roots/ 

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Friday, March 5, 2021 9:05 AM

The loading of the wheel onto the rail cold forms and changes the profile of both the wheel tread and the rail.  Much of this change is just the plastic flow of the metal being reshaped as though it were dough.  It has been discovered that rail grinding to maintain its correct profile adds life to the rail.  It seems counter intuitive because grinding removes metal, which seems equivalent to wearing the rail out.  But re-truing the rail head profile adds life that exceeds the life lost by wear between re-truing operations.   

When they refer to hollowed wheel threads, I understand that to be a result of the cold reshaping due to metal flow under pressure.  This cold flow not only changes the profiles of wheel and rail, but it also induces stress areas that become prone to breaking out of the wheels.  It also moves metal out of the rail/wheel contact and brings the rail contact deeper into the critical mass fiber of the wheel.  Moving this load into that weaker area of the wheel causes stress cracking to begin.  The cracking then lengthens deeper into the wheel toward the center, and further weakens the wheel.

 

Note:  I have revised my understanding of the fracture process detailed here in green text, and the next post reflects that new understanding. 

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Friday, March 5, 2021 8:46 AM

From the PDF:

As previously discussed, VSRs are extremely rare—if not non-existent—in domestic and international passenger wheels, and international freight wheels, including Australian mining railroads which operate at higher axle loads than the North American freights. The aforementioned railroads generally utilize multi-wear wheels and routinely turn their wheels. However, the North American freight railroads use single wear wheels which typically have zero or one turn before end of life. It appears that the U.S. freight railroads are an outlier in the world for not utilizing multi-wear wheels and routinely turning the wheels.

 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, March 5, 2021 8:44 AM

charlie hebdo
Car loaded weight increases might also be relevant? 

Somewhere there has to be a metallurgical analysis of the increased tendency to failure, and I have no doubt this would assist in finding causes.  By analogy weight increases causing increased railhead problems could be expected to have some effect on the corresponding contact patch.  Shock damage (for example at non-flange-bearing crossings or frogs) might be a factor with some approaches to providing hardened tread.  I believe in this period we've seen much more intensive profile grinding to reduce running shock and noise, which might be conducive in some way to the 'wrong' kind of stress raising.

 

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Friday, March 5, 2021 8:43 AM

charlie hebdo

Car loaded weight increases might also be relevant? 

 

I suspect it is relevant.  Were wheels and axles upgraded specifically for the weight increase, or did the increase fall within the existing wheel and axle capacities? 

  • Member since
    September 2017
  • 5,636 posts
Posted by charlie hebdo on Friday, March 5, 2021 8:18 AM

Car loaded weight increases might also be relevant? 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, March 5, 2021 8:06 AM

zugmann
There has to be a standard somewhere.  Your issue should be with the regulation - not the person holding the gage that is following that regulation. 

I think part of the implication he's making is that the gage designed for one purpose isn't picking up a critical condition affecting safety in another respect.

I had thought that sharp flanges were in themselves so important a safety consideration that they had their own regulation -- so much so that I never looked in the CFR to see exactly how inspection and condemnation standards might have been worded.  What exactly IS present Federal practice regarding sharp flanges?

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Friday, March 5, 2021 7:56 AM

According to this report, the wheel failure problem developed between 1990 and 2000.  Does this coincide with the weight limits increase of rolling stock?  It seems that the wheels in use before this failure phase began must have been just on the safe side of the “razor’s edge” of durability.  Then it crossed over that edge and a vast array of consequent failures have ensued with many possible interrelated causes and effects.

Sorting this out and resolving the problem with just the adequate remedies must be the biggest technical challenge the industry has ever faced.  It involves changing the mass of the various wheel features, changing the metallurgy, changing the manufacturing process, abandoning the single-use wheels, adopting wheel truing equipment and methods, evaluating rail grinding, balancing the wheel truing with rail grinding, evaluating the heat effect of braking, evaluating the effect of ambient temperature, and evaluating how the wheel, axle, and rail loading re-forms metal and adds residual stress.

It seems strange that with this wheel failure issue being a systemic problem with the wheel and rail standards, it is consistent, but the failure rate is extremely low.  And this despite the fact that the conditions are standardized over such a high number of wheels in use.  If wheel durability has reached a tipping point, I would expect the majority of wheels to be failing.  

  • Member since
    January 2002
  • From: Canterlot
  • 9,575 posts
Posted by zugmann on Friday, March 5, 2021 6:30 AM

mudchicken
(*) fun with switch point simple derailments, flanges sharp to the touch / hardly rounded edges and the cop-out is that the gage says it's OK.

There has to be a standard somewhere.  Your issue should be with the regulation - not the person holding the gage that is following that regulation. 

It's been fun.  But it isn't much fun anymore.   Signing off for now. 


  

The opinions expressed here represent my own and not those of my employer, any other railroad, company, or person.t fun any

  • Member since
    January 2002
  • 4,612 posts
Posted by M636C on Friday, March 5, 2021 12:58 AM

mudchicken

Is it thin rim or sharp flanges*? Regardless, overly worn wheelsets that see so very few checks with AAR wheel gage templates... Lot fewer one-spot and car-knocker inspectors out there than there used to be. 

(*) fun with switch point simple derailments, flanges sharp to the touch / hardly rounded edges and the cop-out is that the gage says it's OK. (pegging the needle on the ol' baloney-meter)... If it were not for the appearance of the new generation switch point protectors, we might see a lot more simple yard failures)

 

 

I was an observer to a yard derailment that was associated with a sharp flange and a switch protector, back in 1976.

Thuis was in Mt Newman's port yard at Nelson Point, Port Hedland, in Western Australia.

I was sitting in the trailing cab of a three unit set of locomotives when I heard our train crew call the Yard Tower.  "Has our train been cancelled?" Yard Tower "No, why?" Train Crew "Have you looked at the yard lately/" Yard Tower (expletive deleted).

A loaded train had derailed across the yard ladder taking out four switches, including the one between us and the main line, with 24 loaded 100 ton gondolas leaning at 45 degrees.

They worked out a path for us which involved backing the whole train through the car dumper and bypassing the derailment on the one remaining track. This was on a Friday and my train arrived back on Saturday night.

Since I'd been trained in investigating derailments, I spent my free time on Sunday morning trying to work out what had happened. (It was at least as interesting as anything else happening in town on a Sunday where we had one radio station and one TV station....)

I checked out the leading wheelset on the first derailed wagon and it did indeed have a very sharp flange which shouldn't have been running. I then walked back looking for the site of the derailment.

For those who haven't done this, all our wheels had been turned on Hegenscheidt lathes, and these use a single point tool, So the flanges have a distinctive pattern of parallel lines which under normal running are never affected.

When a wheel climbs on to the rail head, it leaves a distinctive mark with the parallel lines being embossed into the railhead.

I followed this back to the first switch in the yard, where it was clear that the sharp flange had hit the "Mack" switch protector, a casting designed to bump the flange clear of the adjacent switch blade.

There was a curve preceding the switch and the rail head up to the switch only had about half of the head remaining. The switch protector had been almost worn away, but had recently been reversed to provide a new wearing surface.

The sharp flange hit the switch protector and rode up over it, and climbed further on to the rail head. Then we had an amazing stroke of luck. Somehow the wheelset found its way through the switch frog and guard rails without further derailing, leaving the main switch complete and in place. It then dropped off clear of the frog and took out the next four switches.

But we still had two tracks out of the yard while the gang rebuilt the damage.

On Monday I tried to explain all this to the Track Superintendent but I think his eyes glazed over....

Peter

  • Member since
    December 2001
  • From: Denver / La Junta
  • 10,820 posts
Posted by mudchicken on Thursday, March 4, 2021 9:44 PM

Is it thin rim or sharp flanges*? Regardless, overly worn wheelsets that see so very few checks with AAR wheel gage templates... Lot fewer one-spot and car-knocker inspectors out there than there used to be. 

(*) fun with switch point simple derailments, flanges sharp to the touch / hardly rounded edges and the cop-out is that the gage says it's OK. (pegging the needle on the ol' baloney-meter)... If it were not for the appearance of the new generation switch point protectors, we might see a lot more simple yard failures)

Mudchicken Nothing is worth taking the risk of losing a life over. Come home tonight in the same condition that you left home this morning in. Safety begins with ME.... cinscocom-west
  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Thursday, March 4, 2021 8:35 PM

Report seems to indicate that there is more trouble with wheels in the West as opposed to the Eastern carriers and that Covered Hoppers and Gondolas (remember todays 'coal cars' are frequently AAR designated gondolas).  The recurring theme that was mentioned numerous times was 'thin rim'.

Will be interesting if they can pin down some provable reasons.

Never too old to have a happy childhood!

              

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Thursday, March 4, 2021 1:41 PM

That seems like an extremely complex problem.  It even gets into the idea of grinding wheel profiles and reducing the grinding on rail profiles because they affect each other.  

  • Member since
    December 2007
  • From: Georgia USA SW of Atlanta
  • 11,919 posts
FRA report on wheel failures and other topics
Posted by blue streak 1 on Thursday, March 4, 2021 12:21 PM

Report shows pictures of various wheel failures.  One item that is noted that there has been more rim failures since 1990.  Reason not determined yet.

Wheel Failure Investigation Program: Phase I | FRA (dot.gov)

 

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy