One man crews: Spread the enthusiasm

[schlimm]

Five broken knuckles?   The weak link in the chain.  The problem is not hi-tech devices failing.  The problem may be the combination of using a technology over 100 years old and expecting it to function with freight cars now weighing far more along with greatly reduced manpower to inspect and maintain.

[Euclid]

Those are mighty big delays for routine broken knuckles or brake hoses.  Clearly a better way is needed to deal with them.

[dehusman]

schlimm

Five broken knuckles? The weak link in the chain.

Knuckles are designed to be the weak link in the chain.  They are designed to fail before the drawbar or the structure of the car itself.  They are also designed to be be "easy" to replace (as opposed to a drawbar or a car ripped in half). 

They can build (and have built) composite kunckles that weigh 1/2 to 1/3 of a  regular knuckle, but they were actually stronger than the steel couplers and the car/draft gear failed before the knuckle did.

If a chunk of steel about 3-4" thick and 8-10" wide has that many failures, that is why I question those coupling systems that use a single 2" steel rod.

[dehusman]

schlimm

Five broken knuckles? The weak link in the chain. The problem is not hi-tech devices failing. The problem may be the combination of using a technology over 100 years old and expecting it to function with freight cars now weighing far more along with greatly reduced manpower to inspect and maintain.

The knuckles work fine and are strong enough for modern cars.  What is important is figure out WHY the knuckle failed.    Knuckles fail because the in train forces exceed the strength of the knuckle.  That means either the knuckle is weakened (damaged) or something else caused the forces to be higher.  Detecting flaws in a knuckle is tough because its an odd shape and surrounded by other pieces of big metal.  Most efforts focus on proper train handling and elimination of UDE's (undesired emergency brake applications).  In BAltACD's examples the knuckles probably really weren't due to anything wrong with teh design of the knuckle, it was more of in adequate securement  and design flaws/inadequate maintenance of the air hose trolleys and train lines on the cars.

[schlimm]

When one dispatcher on one railroad can note five broken knuckles in his territory in 24 hours, that suggests equipment failure.   The basic design was probably fine even 60 years ago.  But with today's car weights doubling, longer trains, and more HP, the forces experienced are increased and lead to failures, even if train handling is proper.  The railroads do not have the ability to detect flaws in advance consistently.  So perhaps the manner of connecting cars needs to be revisited?  Or weights and train lengths need to be reduced?

[cx500]

As Dave H points out, the knuckle is intentionally the weak link.  If you think about it, that is actually a safety feature. 

A broken knuckle can create tremendous delay with resulting inconvenience to the railroad (and sometimes the public at a grade crossing) but that is generally the extent of the problem.  There are costs involved, for recrewing and the knuckle itself, but of minor magnitude.  On the other hand, if a drawbar breaks or the stub sill breaks off a tank car, a major derailment becomes a real possibility.  That causes far greater delays, damage, and potential risk to human life.

Using distributed power helps manage the train forces and reduce the stress on the couplers.

John

[schlimm]

In Balt's example, I counted at least 11 hours of trains standing still.  Additionally, other trains were delayed or also standing still.   How much per hour (on average) does it cost the railroad to have trains just sitting around?  Multiply that times how number of hours times how many trains per day times times 365.  My hunch is that it is a lot of money.

Dave Husman may think the knuckles are fine but more than one rail engineer has suggested to me quite the contrary.

[Paul of Covington]

   It's a balancing act.   They want to move more tonnage per dollar by putting more tons in each train.   When they see that the problems are costing more than they are trying to save, they will (or should) back off on the loads.

   The alternative would be to re-design and build stronger and heavier sills, draft gear and couplers to maintain the safety factor of keeping the re-designed stronger knuckle the weakest link.   Let them figure the cost of that.

[Euclid]

Paul, I agree with your point that if you make the couplers stronger, they still must remain as the “weak link.”  So that would require strengthening the whole coupler, draft gear, center sill, etc.  That would be very costly due to all the rolling stock, so I don’t see the industry making such a strength upgrade of the rolling stock.

 

However, I think that the real issue is not the fact that knuckles break, but rather, the time it takes to replace the knuckle.  The basic replacement task that could be completed in ten minutes, but a vast amount of time is chewed up in all the logistics of train movement and manual labor to get the replacement knuckle to the site.

 

We are always told that train delays pose a severe cost to the company.  Needing to spend a couple hours replacing a knuckle seems like a problem that needs fixing.   Knuckle replacement needs to be more nimble. Maybe they could give the job to an outside contractor.  It would be the Nimble Knuckle Service.  

[Paul of Covington]

   Euclid, I know your "Nimble Knucle Service" was a bit of tongue-in-cheek, but from what I gather, the replacement of the knuckle is not where the time goes.   It's walking the train to get to it, then after replacing it, coupling back up, pumping up the air and doing a brake check.   No outsider is going to be closer to the action than the crew on the train.

[abdkl]

Presuming the number crunchers include any penalty time for broken knuckles & other UDEs that occur at the rear half of the consist....wouldn't a rear end crew (???!!) with spare knuckles mean less walk time for some train walks?

[BaltACD]

abdkl

Presuming the number crunchers include any penalty time for broken knuckles & other UDEs that occur at the rear half of the consist....wouldn't a rear end crew (???!!) with spare knuckles mean less walk time for some train walks?

If the were still alive.  With the slack in the size trains that are being run, slack action could be a real killer to inhabitants of a caboose - even if they were belted in.

[Euclid]

Paul,

 

I am serious about an independent contract knuckle service.  They would not run the train or pump air.  But they would get the knuckle to the break and replace it.  There would be no railroad crewmen walking the train, carrying knuckles, or moving the train to get the knuckle switched.  The point would be to get the knuckle job down to less than an hour from break to resumption of travel.  That ought to be worth $7500 per incident, thus plenty of incentive for One Hour Knuckle to get the job done.

 

It would be like calling a contract derailment cleanup service like Hulcher.  

[abdkl]

I agree humans in a caboose could be like a golf ball hit with a full swing in a tile bathroom. A "caboose" could still be used as a Rear-End-Device and a traveling knuckle store, without human passengers.

[dehusman]

schlimm

In Balt's example, I counted at least 11 hours of trains standing still.  Additionally, other trains were delayed or also standing still.   How much per hour (on average) does it cost the railroad to have trains just sitting around?  Multiply that times how number of hours times how many trains per day times times 365.  My hunch is that it is a lot of money.

Dave Husman may think the knuckles are fine but more than one rail engineer has suggested to me quite the contrary.

 

One consistent thread on this forum is that whenever there is problem the first response seems to be "lets spend a billion dollars completely redesigning the equipment" instead of fixing the actual cause.  Railroad studies have found that most trains going in emergency are caused by a handle of cars sometimes a specific handful of cars.  One railroad has reduced UDE's with an immediate recovery by keeping track of the cars in trains that suffer them and then seeing which cars are common in multiple trains.  Amazingly enough there are few cars out there with air hose support systems that cause dozens of UDE's (which can cause knuckles)  Fix those few cars and UDE's and knuckles drop.   

 

BaltACD's knuckle on the bridge?  First thing I would do is have the roadmaster out there inspecting the approaches to the bridge to make sure there aren't any dips or jogs in the track or that isn't something sticking up that would hit an air hose.  Next take those cars and don't just replace the knuckles and air hoses, send them to a car shop and have the entire draft gear and air hose support system disassembled and inspected.  Dollars to donuts the air hose trolley or support systems is out of spec.

 

Railroads can track where UDE's and knuckles occur and which engineers have them.  After a while patterns emerge.  High crossings,  certain locations, bad alignments.  There may be engineers that need additional training.  There may be some train make requirements that need to be changed. 

 

In one night there might be a half dozen cars with broken knuckles, but there are about 300,000 that make their trip successfully.  Its probably not the basic design.  Could it be better?  Probably.  Is it worth spending a couple billion dollars to change?  Probably not yet.

[dehusman]

Euclid

I am serious about an independent contract knuckle service.  They would not run the train or pump air.  But they would get the knuckle to the break and replace it.  There would be no railroad crewmen walking the train, carrying knuckles, or moving the train to get the knuckle switched.  The point would be to get the knuckle job down to less than an hour from break to resumption of travel.  That ought to be worth $7500 per incident, thus plenty of incentive for One Hour Knuckle to get the job done.

 

And that would mean they would have to be less than about 15" from the site, that allows 20" to get into the actual location of the knuckle and then 15-20" to fix it.

 

How are you going to station people so they are 15" from any point on any railroad ?

 

How are you going to get people into spots with no roads?

 

 

It would be like calling a contract derailment cleanup service like Hulcher.  

 

And Hulcher can take anywhere from 1-8 hours to get on site, then an hour to be workwise.

[schlimm]

BaltACD

With the slack in the size trains that are being run, slack action could be a real killer to inhabitants of a caboose - even if they were belted in.

Could that also be a major factor in why knuckles break - too much force because of the slack inherent in longer trains?

Another factor could be quality control in the knuckles themselves,  not the design but the materials used in manufacture.  Where are they manufactured in recent years?

[dehusman]

abdkl

Presuming the number crunchers include any penalty time for broken knuckles & other UDEs that occur at the rear half of the consist....wouldn't a rear end crew (???!!) with spare knuckles mean less walk time for some train walks?

Wonderful suggestion.  Unfortunately most knuckle failures are towards the head end of the train, where the in train forces are highest.  If the knuckle is more than a few cars ahead of the caboose, I'd drop a knuckle off the head end, pull the train by, then either fix the knuckle if it was on the trailing end of the car or load the knuckle on the rear car and shove it back to to the rear portion (if it was on the leading end of the car).  Nobody's going to haul a knuckle more than a car length or two.

[dehusman]

schlimm

Could that also be a major factor in why knuckles break - too much force because of the slack inherent in longer trains? Another factor could be quality control in the knuckles themselves, not the design but the materials used in manufacture. Where are they manufactured in recent years?

This of course Assumes that knuckles are breaking more now than they wer in years past.  Do you have any data to support that?

 

(spoiler alert :  overall knuckle failures have decreased over the years)

[Wizlish]

dehusman

This of course Assumes that knuckles are breaking more now than they wer in years past. Do you have any data to support that?

(spoiler alert : overall knuckle failures have decreased over the years)

But hold on a moment: we've been gabbling for weeks about how trains are getting longer and longer recently as profitability in the post-coal post-China-financial-kerfuffle era begins to slide.  The BaltACD story that led to this exchange mentions an untoward number of near-simultaneous knuckle failures, and I for one would be entirely unsurprised to find a spate of these failures as an 'unanticipated consequence' of running longer trains. 

But yes, I fully expect to see this reflected in statistics.  Once a proper base of data on current operations has made its way into them...

[Euclid]

dehusman

 

Euclid

I am serious about an independent contract knuckle service.  They would not run the train or pump air.  But they would get the knuckle to the break and replace it.  There would be no railroad crewmen walking the train, carrying knuckles, or moving the train to get the knuckle switched.  The point would be to get the knuckle job down to less than an hour from break to resumption of travel.  That ought to be worth $7500 per incident, thus plenty of incentive for One Hour Knuckle to get the job done.

 

 

 

 

And that would mean they would have to be less than about 15" from the site, that allows 20" to get into the actual location of the knuckle and then 15-20" to fix it.

 

How are you going to station people so they are 15" from any point on any railroad ?

 

How are you going to get people into spots with no roads?

 

 

 

It would be like calling a contract derailment cleanup service like Hulcher.  

 

 

 

And Hulcher can take anywhere from 1-8 hours to get on site, then an hour to be workwise.

 

Well, on one hand we are told that crew size cannot be reduced because of all the complications with broken knuckles; and then when somebody suggests fixing the knuckle problem, we are told that there is no problem. 

I suggest a dedicated, highly specialized, highly motivated, independent contractor knuckle service, and you say it will not work because it would require staging men with replacement knuckles every fifteen inches along the railroad lines.  

I am not suggesting a need to replace a knuckle in five seconds.  An hour maximum would probably be a vast improvement.  And it would not be a matter of spending a “billion dollars” to redesign the whole railroad simply to fix a tiny problem, as you suggest.  Obviously, it has to be cost effective.  This is not even a public safety issue, so the only point is to increase corporate revenue. 

The only point of my comparison to Hulcher is that they are an independent contract service, and the railroads have found that it saves money to use them.

[Wizlish]

This one-hour knuckle service reminds me of the current trouble finding nurses. 

Can I fugure out a way to deliver even a heavy knuckle to a given site, with reasonable safety, and with any tools and supplies needed if I find (say) that some other part is galled or bent?  Probably.  But is anyone going to want to pay what it would cost me to deliver that level of 'service with a smile'?  Probably not.  And regardless of volume or the costing-down of capital and training, it will never be possible to deliver that level of service at a rate railroads would be willing to pay.  (Even if you carefully explain all the costs and implications of tying up the railroad while your on-the-law crew does manual labor between the cars.)

There may be times when the engineering equivalent of Life-Flight is justified.  There may be times when helicoptering a replacement crew to a train, or driving the special Brandt-style maintenance truck 100 mph to the rescue, would be justified.  But those occasions are likely to be fewer and farther between than pulled knuckles.

I'd be interested to see what Hulcher might put together as a service model for 'outsourcing' this kind of comparatively light emergency repairs.  They have at least the capability to leverage some of the specialized equipment needed, and could find tax offsets to help acquire or maintain it, too.

[BaltACD]

There appear to be a number of posters on recreational pharmacuticals in judging the speed with which a knuckle can be replaced by ANYONE in the size trains that are being operated by today's Class 1 railroads.

There are numerous rule compliance reasons for the length of time taken for the repair as well as the distances that have to be negotiated between crew location and the location of the incident within the train.  Mentioning a broken knuckle is just a easily identifed action - there are any number of other issues that have to be inspected and rectified in the operation of today's trains - hand brakes still applied, sticking air brakes, broken air hoses, leaking trainlines, inspection for defect detector activations etc. etc. etc. 

Talk of 3rd party 'on immediate call' contractors is ludricous.  Who is going to set up a vehicle supplied with 24/7 driver/technician  protection and all the necessary repair items stationed strategically approximately every 50 miles along Main Track routes to provide 'reasonable' response times and charge a price that will return a profit and cover the investments into equipment, manpower, insurance and fringe benefits. 

[Wizlish]

The big problem I see is in the required rapid response.  There is a certain lag in discovering the problem, then another in getting it reported to the point where action is authorized (as opposed to the crew just starting to tie the train down and start walking the consist), and then figuring out where the train is, where the break is, how to get to it, and going there.  Only then can you figure out how long it's going to take to get someone there ... and how long to get them safely ready in place, with tools and PPE, to actually begin work. 

In the absence of practical hypersonics, that means a fixed staff of not inconsiderable size 'on call' at every hour, with Murphy doing his best to ensure it will be cold, dark, and sleeting when the hour arrives. 

Now reverse the process, and the lags, to get the special crew out again, now without any particular incentive to go really fast if there isn't another call waiting for them.

Note how ugly the logistics gets when you have one response crew and five knuckles broken on multiple trains. 

Note that none of this argues there can't be crews that 'help' with road failures -- just that they aren't going to do rapid response that is 'rapid' enough to beat what railroads already pay for with two-man crews.

Frankly, I'm astounded that no one has brought up ECP in this context yet, as it (or something like it) promises to reduce many of the incidents that wind up pulling knuckles or breaking draft gear.  I also have to wonder if there is a reason to provide knuckles on some of the cars in the train, in boxes like the ones on locomotives used for the rerailing frogs.  That would cut the worst part of a crew repair -- the need to lug the heavy piece down the ballast prism and over bridges without walkways, or run the movable piece of the train forward and back to position the heavy piece, before an actual repair can be made.

(Yes, there are problems with the latter, ranging from vandalism to accounting for whose knuckle goes on whose car.  I won't list all the ones I've thought of.  The point I want to make is that net of all problems, putting knuckles on every xth car may still be cheaper and better than a flying-squad repair service model.)

[Euclid]

Wizlish

But is anyone going to want to pay what it would cost me to deliver that level of 'service with a smile'?  Probably not.  And regardless of volume or the costing-down of capital and training, it will never be possible to deliver that level of service at a rate railroads would be willing to pay.   

My premise is based on the assumption that they are already paying for it.  The point is to save money.  If it can’t be done then so be it.  But I will bet the industry said the same thing about derailment services before they saw what an independent contractor with sideboom Cats could do.    

We hear about how much it costs to tie up the mainline.  Valuable merchandise in wrecks is bulldozed aside and destroyed because its loss pales compared to tying up the mainline.  Time is money.  What does it cost to tie up the mainline for three hours to change a knuckle?  It must be at least $50,000.

I too would like to see the Hulcher business model for emergency light repair services.  Are they non-union?  

[Paul of Covington]

   Euclid, I just don't get your logic.   The train crew are there.   They have to walk to the location of the broken knuckle.   After the repair they still have to put the train back together and do a brake test.   The replacement of the knuckle is only one small step.   What would be accomplished by having to wait for someone else to come out and replace the knuckle?

[schlimm]

Wizlish

 

 

dehusman

This of course Assumes that knuckles are breaking more now than they wer in years past. Do you have any data to support that?

(spoiler alert : overall knuckle failures have decreased over the years)

 

 

But hold on a moment: we've been gabbling for weeks about how trains are getting longer and longer recently as profitability in the post-coal post-China-financial-kerfuffle era begins to slide.  The BaltACD story that led to this exchange mentions an untoward number of near-simultaneous knuckle failures, and I for one would be entirely unsurprised to find a spate of these failures as an 'unanticipated consequence' of running longer trains. 

But yes, I fully expect to see this reflected in statistics.  Once a proper base of data on current operations has made its way into them...

 

Does Mr. Husman have a citation for data to support his contention?  Why not share it?  

But that ignores the question of why couplers break.  It also ignores the enormous cost of delays, as well as eventual lost business.

[Euclid]

Paul,

What would be accomplished is the time saved by quicker replacement of the knuckle, including getting the knuckle to the break.  Is that really just a small step in the whole process, as you say?  Removing the broken knuckle and installing the new one is a small step.  Assessing the situation, coming up with a plan to get the knuckle to the break, and executing that plan seems like a much larger step.  But let's consider the example cited by BaltACD.  What is the time breakdown for the details of spending 3 hours fixing a broken knuckle.  

[edblysard]

BaltACD

There appear to be a number of posters on recreational pharmacuticals in judging the speed with which a knuckle can be replaced by ANYONE in the size trains that are being operated by today's Class 1 railroads.

There are numerous rule compliance reasons for the length of time taken for the repair as well as the distances that have to be negotiated between crew location and the location of the incident within the train.  Mentioning a broken knuckle is just a easily identifed action - there are any number of other issues that have to be inspected and rectified in the operation of today's trains - hand brakes still applied, sticking air brakes, broken air hoses, leaking trainlines, inspection for defect detector activations etc. etc. etc. 

Talk of 3rd party 'on immediate call' contractors is ludricous.  Who is going to set up a vehicle supplied with 24/7 driver/technician  protection and all the necessary repair items stationed strategically approximately every 50 miles along Main Track routes to provide 'reasonable' response times and charge a price that will return a profit and cover the investments into equipment, manpower, insurance and fringe benefits. 

 

 And I an sure that on Balt's railroad, just like mine, there are roads that run beside every single track, the entire route, to allow the Fast Action Response Team guys to get right there in minutes.

[BaltACD]