Trains.com

Concept for a Safe Oil Train

23510 views
375 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Sunday, February 16, 2014 6:04 AM

I agree, that that is far-and-away the best approach for the alternator, and allows the wheel sets themselves to remain standard.  Either slipring alternators or rotating magnets could be used.

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Sunday, February 16, 2014 9:31 AM

Euclid

This fireball crisis is more of a marketing problem for the railroads and oil industry than it is an engineering problem. 

It is interesting to note that, although this problem applies to the oil industry as well as the railroads, the main media thrust focuses on the culpability of the railroads for their derailing trains.  The media seems to dismiss the Bakken oil volatility and the mislabeling by the shippers. 

If it were just an engineering problem, the railroads could take comfort because much of that problem is beyond their control.  But since it is a marketing problem, the problem can be anything that the media says it is. 

You can say that it is not a problem because the media does not know what they are talking about—that they exaggerate the danger of oil while ignoring the more dangerous materials that are shipped by rail.  But it does not make any difference.  If the media says the problem is oil-by-rail, then that is the problem.  You can’t end the problem by disproving the media.  You can’t defeat the power of the press by scoffing at it.

A Senate hearing on this rail safety problem had been scheduled for last Thursday, but was cancelled due to snow.  The marketing problem will be on full display once this hearing is under way.  Senators will expect the railroads to ensure the safety of people living along the rail lines because the premise of the problem is that the railroad companies are placing these people at risk for death and injury every day now with the rising oil traffic.  

“Ensure the safety” does not mean that it is enough to create the probability that fewer people will be killed or injured.  There is also the implication that Senators expect a very quick solution to the problem because they refer to the problem as being critical.

Against this backdrop, the industry will tell the Senators that the problem is being exaggerated; that the media is wrong; and that rail is the safest form of transportation.  And then the industry will offer a purely engineering based solution to the problem that will take several years to achieve.  The Senators will want to know if the industry’s solution will ensure the safety of people along the rail lines, and the industry will tell the Senators that “nothing is 100% certain.”   

 

 

 

  •  

                                                              

Posted by Norm48327

onSat,   Feb 15 2014 5:31 PM

                   

  • Ya know Bucky,
  • Your entire post sounds like it comes from a rail hater rather than a railfan.
  • Norm
  • Never trust atoms. They are known to make up everything.

 

****************************************************************************

 

Norm,

You mentioned earlier that you thought I was on the right track with the marketing component of my safe oil train idea.

So I am just putting the problem into perspective by explaining what I mean by a marketing problem.  Just because I am pointing out a peril to the industry does not mean that I hate the industry.  Ignoring the peril will not make it go away.  Pointing it out might be useful if the industry does not see it.  You can’t sidestep a problem if you don’t realize it is there.

There are indeed haters in this equation, but I don’t see them as railroad haters.  They are oil haters, and their greatest current leverage against oil is in impeding of oil shipping by rail.  Public safety is their most potent weapon, and you can bet that their viewpoint will be present at the Senate hearing. 

The hearing will be very interesting to watch.  It is a quest for answers, and answers will be given.  So far, I have not heard any answers other than making tank cars stronger and getting the trains a safer distance from the public.  In my opinion, that solution does not match the problem.  It only very slowly nibbles at the margins of the problem while the Senators and the public expect an end to the problem now.  And if they don’t hear what they want to hear from the industry, they will solve the problem on their own.   

From a marketing perspective, if you can’t solve the problem 100%, and do it now; then make the public and the regulators believe that you will.      

 

 

  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Sunday, February 16, 2014 1:28 PM

Placing the alternator on the end caps of the roller bearing makes sense to me as well, though my preference would still be to use ultracaps to energize the magnet. My concern with derailments is from having a heavy chunk of metal dangling by springs, and what happens if one of the springs break.

Running some numbers, a four inch square with a 1 Tesla flux density will be good for about 10,000 pounds of force, which would give very useful retarding force. With a gap of 1mm between the pole face of the magnet and the rail, this would need one the order of 20,000 amp turns, 100 turns would need 200 amps Noe that 10 awg wire can handle the current for the time involved). 200 amps for 10 seconds would be 2000 coulombs, which could be provided by a 1000F ultracap charged to 2 volts (latest caps are good for 2.5V long term). Number of caps in series would depend on resistance and inductance, along with the change in coil voltage due to the field increasing when the pole faces approach the rails.

Looks doable, though not sure if it make sense.

- Erik

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Sunday, February 16, 2014 2:01 PM

But the alternators also serve as the derailment detector pickups, are existing technoilogy in both applcations.  Remember that it is the difference between the speed of the front and the rear axles that singals a derailment has occured.

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Tuesday, February 18, 2014 5:00 PM

The safe oil train system that I am proposing uses ECP brakes, but I am not calling for the development of ECP braking, since this braking system already exists fully developed.  I am just calling for its application to oil trains.  However, I am proposing a small control modification to the ECP braking system that does not currently exist.  That is the part that changes the braking performance in response to a derailment.

But, aside from what I am proposing, ECP braking alone would offer a substantial benefit over standard pneumatically controlled pneumatic braking of our present universal system.  ECP braking applies more quickly than standard PCP braking.  The quicker application reduces stopping distance by 30-70%. 

Some here have said that braking of the cars behind the derailment cannot possibly keep up with the deceleration of the cars that have derailed.  That may be true in some cases, but generally I think that claims of how fast those derailed cars will stop have been exaggerated.  And I also think that the imagined problem fails to take into account, the enhanced stopping power of ECP brakes working to slow down those trailing cars.

Another benefit of ECP braking is that, not only is application quicker, it is also simultaneous, unlike conventional air brakes which propagate the application sequentially from one car to the next over a period of time.  This propagation lag time creates slack action that can actually cause a derailment.  So ECP brakes actually prevent derailments in some cases by virtue of their simultaneous application. 

Therefore, considering the fundamental advantages of ECP brakes alone, I expect and predict that the railroads will apply ECP brakes to oil trains.  It will make them safer.  It is the most obvious advantageous application because oil trains are said to lacking in safety.  And it will show the regulators and the public that the railroads take the problem seriously and are doing something big about fixing it. 

ECP brakes are perhaps the most high profile advancement in safety and performance that the industry currently has.  Considering the fact that the railroads would like to begin their ECP introduction with specialized service as a way to ease into universal application, putting ECP brakes on oil trains seems like a no brainer.   

 

Here is an informative paper on ECP brakes: http://www.uic.org/cdrom/2001/wcrr2001/pdf/sessions/1_6/465.pdf

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 18, 2014 6:30 PM

One th-th-th-thousand f-f-f-farads on a r-r-railroad car? With wires hanging down the outside of the truck frame?  Nice arc source for the oil if it gets out, perhaps -- or to make a hole in the car, perhaps?

And then there are the fun uses if the cap array is stolen some dark night...

I think I'll pass on ultracaps for this particular application, thank you.  I only hope others do, too.

  • Member since
    December 2001
  • From: Northern New York
  • 25,024 posts
Posted by tree68 on Tuesday, February 18, 2014 8:04 PM

A one farad capacitor will range in size from a tuna can to a 1 liter soda bottle, depending on the voltage involved.

According to How Stuff Works, replacing an AA battery with a capacitor would require a device capable of 11,080 farads...

We'll just have to send out workers with boatloads of penlight batteries to keep those cars ready to go...

LarryWhistling
Resident Microferroequinologist (at least at my house) 
Everyone goes home; Safety begins with you
My Opinion. Standard Disclaimers Apply. No Expiration Date
Come ride the rails with me!
There's one thing about humility - the moment you think you've got it, you've lost it...

  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Tuesday, February 18, 2014 10:09 PM

Overmod

One th-th-th-thousand f-f-f-farads on a r-r-railroad car? With wires hanging down the outside of the truck frame?  Nice arc source for the oil if it gets out, perhaps -- or to make a hole in the car, perhaps?

Sure! A Maxwell  1500F capacitor comes in at 61mm dia and 85mm long. Keep in mind that it is rated for 2.7VDC, and lifetime would be longer if kept at 2.5V.  Five of these puppies in series would give 12.5V, which is WAG for the drive voltage for the track magnet.

For RR service, I would encapsulate the caps in an electronic grade epoxy along with the FET switch to energize the magnet. Charging and voltage sensing would be done through current limited leads. I'd be less scared of dealing with an ultracap than a Li-ion battery with similar short term power ratings.

On a related note, I'd contend that ultracpas would be much nicer for a "Green Goat" switcher than lead-acid batteries.

- Erik

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 18, 2014 10:14 PM

tree68

A one farad capacitor will range in size from a tuna can to a 1 liter soda bottle, depending on the voltage involved.

According to How Stuff Works, replacing an AA battery with a capacitor would require a device capable of 11,080 farads...

We'll just have to send out workers with boatloads of penlight batteries to keep those cars ready to go...

I think you are lamentably behind the times regarding capacitor technology.  If you Google 'ultracapacitor' and look at the size of a multifarad capacitor you will find it considerably smaller -- and far more susceptible to damage by even slight overvoltage -- than the tuna cans and so on you mentioned.

One thing about the 'thousand farad' instantiation is that its output to the brake would almost certainly not be running at the 2V nominal output; the individual supercaps/ultracaps would be in strings.

Another thing that makes ultracaps different from penlight batteries is the very low effective internal resistance.  Short a 3V battery terminal, and the current is limited by the characteristics of the internal chemical reactions.  (But see the YouTube video with the hundreds of 9V radio batteries in series...!) Short a 3V lithium-ion battery and you might get a fire.  Short a 1-farad capacitor even at 2V and expect to see something interesting.  Short 1000F and you will have a memorable experience.

It does look a bit amazing that there are so many electrons stowed up in one little penlight battery (and in the same general range of output voltage as the supercap, too) but remember the definition of a coulomb, and then of a farad.  The difference is what's involved in getting all those electrons to stay inside the can as opposed to inducing them to go from one place to another.  If you look at the history of electrical technology, its early beginnings (in Europe, not in India) are in electrostatics -- but the practical uses develop first with the invention of the 'voltaic pile' and then by the recognition that electromagnetics and induction give far more power, and at higher potential voltage (pun intended) than a standard-potential chemical reaction...

  • Member since
    December 2001
  • From: Northern New York
  • 25,024 posts
Posted by tree68 on Tuesday, February 18, 2014 10:26 PM

Overmod
I think you are lamentably behind the times regarding capacitor technology. 

No question there.  The bulk of my experience is with capacitors in tube and transistor electronic circuits, not high-capacity applications.  I have to go with what I can find on the 'Net. 

I do have to question, however, just how long the capacitors will be able to maintain sufficient current through the brakes.  I don't recall that anyone has said how long it would have to be (unless I skimmed over it), so we have a moving target there.

Short 1000F and you will have a memorable experience.

Been there, done that - maybe not 1000F, but it was the HV for a large CRT display in a piece of weather equipment...  It made quite the snap when we discharged it.

LarryWhistling
Resident Microferroequinologist (at least at my house) 
Everyone goes home; Safety begins with you
My Opinion. Standard Disclaimers Apply. No Expiration Date
Come ride the rails with me!
There's one thing about humility - the moment you think you've got it, you've lost it...

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 18, 2014 10:45 PM

tree68
No question there.  The bulk of my experience is with capacitors in tube and transistor electronic circuits, not high-capacity applications.

Hmmmm...  YOU wouldn't have a copy of the 'bad electron' proposal made for the Navy during WWII, would you?  The one about the 'static' and noise in contemporary AM circuits being due to bad electrons, that were filtered out and forced to charge an 8"-nominal-diameter capacitor, which was then put into one of the ship's guns...  well, drift speed of electrons in salt water is not all that great, so you could work some clear traffic until the bad electrons got back to the hull...

tree68
I do have to question, however, just how long the capacitors will be able to maintain sufficient current through the brakes.  I don't recall that anyone has said how long it would have to be (unless I skimmed over it), so we have a moving target there.

The current would be limited by resistance in the normal manner for these things, if there weren't more explicit provision for 'active' limiting.  I'm surprised there isn't something on the Net somewhere that would explain this. 

I am by no means sure I trust the idea of a very fast capacitor discharge through magnet wire, especially if wound around a formed core, with the idea that the current will be of such short duration that resistance heating will be 'acceptable'.  I don't see the brakes possibly working in the time interval quoted -- perhaps I am simply taking something firmly tongue-in-cheek with too much seriousness.

I do think the setup would be better served by a good modern battery stack with the ultracaps used as 'charge buffers' to limit the rate of change in charge and discharge that the battery chemistry and structure would otherwise suffer.  The idea of using the alternators to power the track brakes directly strikes me as a very expensive, complex attempt to recreate the Loughridge Chain Brake with only some of the historical faults addressed...  ;-}

  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Wednesday, February 19, 2014 1:20 AM

Hmm, time for some numbers...

Fusing current for 10 awg is 500A at 8+ seconds, 6 awg can take 1000A for about 10 seconds. To get 200A at the magnet terminals and with a 12V supply, one would need to keep the wire length to 60' or less with 10awg copper wire and on the order of 150 feet with 6awg wire. 1000F at 2.5V would be 2500 coulombs, so the capacitor would be discharged in 12.5 at a constant 200 amps - possibly marginal with respect to 10awg wire, but no problem for anything much larger. Note that the current will decay as the capacitor discharges, though that may be a feature as opposed to a bug.

A quarter g braking effort will result in a 5.5 MPH/sec deceleration, so 10 seconds of baking at that rate will bring the car to a stop. There isn't much point in going to magnetic track brakes unless the retardation is at least a quarter g.

What I've proposed is probably nowhere near optimal, but close enough to indicate that it is within reason. I've seen electromagnet designs crazier than this (unusual NMR magnets).

- Erik

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Wednesday, February 19, 2014 1:53 AM

Again, there may be more sophisicated systems, but two alternator comparison method is already in use in locomotive slip control and can just as ealisy be applied as derailment detection.   But if you wis h to replace the battery wiht sjuercaps, OK if the supercaps will do the job adequately.  The magnetic track brake is a separate issue.   Electric power is still required on the car.  Head end power can help, but we want the car's electrically controlled brake, and the magnetic track brakes if applied, to work should the car be behind a break-in-two.

  • Member since
    June 2004
  • From: roundhouse
  • 2,747 posts
Posted by Randy Stahl on Wednesday, February 19, 2014 5:23 AM

Decelostats have been around in railroad applications for years, no need to re-invent. Mostly used for disc braking on passenger cars.

 

RSS

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, February 19, 2014 6:05 AM

erikem

Hmm, time for some numbers...

Now we're gettting somewhere.

I presume we are talking about using this system as a one-time emergency device -- would the heat dissipation be sufficient to keep the coil and/or leads from softening and melting after the braking run is finished.  What material are we using to keep the turns insulated from each other that will take the temperature 'long enough' and not crack as the copper expands?  Does it matter in a one-time system if it cracks, so long as it provides the operational insulation (there won't be corona shorting through the cracks at no more than several V)

How do we guarantee that the braking is equally applied to both rails, if the system is applied to the truck?  I would surely wire them in series.   At this deceleration rate, what force would be applied to rotate or skew the truck if it goes over a conventional frog -- the chain system might help this but I think would need a fast-acting tensioner.

I am too lazy to calculate the field decay with dropping voltage, but for 'constant deceleration' you would want the field strength to decay proportional to momentum, no?  So the peak current would only apply at higher road speed, and in any case one of the other uses of a magnetic-brake system is to get the momentum down to where conventional air brakes can work effectively.  I will figure out this curve and the control modality to achieve it later, if erikem doesn't do it first.

Something very significant, perhaps, is that these brakes ARE the practical answer... if modulated short of melting... to the kind of runaway aggravated by brake fade due to composition shoes.  If the brake can generate 5mphps decel in emergencies, it is surely adequate for pulling down speed to where conventional brakes can re-establish authority... the operational difference with dynamic braking being that the 'magnetic' deceleration is no longer dependent on complex locomotive systems and logic, and applied through a relatively small number of relatively expensive wheels.

I do think that it would make sense to run this at a higher nominal voltage than that which is close to the supercap punch-through, and perhaps to use one of the self-healing construction methods.  That might allow use of the system in blended operation at higher speed, and more apropos to the present discussion, allow longer or repeated application at relatively low power for the calculated distributed differential braking that was an essential point of Dave's system for derailments. 

There remain a number of legal, political, and regulatory concerns with the system that we might address.

  • Member since
    September 2003
  • From: Omaha, NE
  • 10,621 posts
Posted by dehusman on Wednesday, February 19, 2014 7:07 AM

There are a couple assumptions in this proposal That I would like to ask about.

Assumption 1:  In normal operation the wheels of a rail car all turn at the same rate.  Do you know this or is this an assumption?  When going around a curve, if the lead truck is rolling on the inside rail and slipping on the outside rail, and the trailing truck is rolling on the outside rail and slipping on the inside rail, the wheels will be turning at incrementally different rates.  If the brakes are applied in normal operation and the brakes apply at different effectiveness on different axles, causing the wheels to turn at different rates, will the system interpret that as a derailment?  What happens if the lead axle has a brand new wheel and the trailing axle has a near condemnable wheel?  Since the wheels are of different diameters, they will turn at different rates.  A 33" wheel has 611 rev/mile, a 31" wheel (1" wear) has 650 rev/mile.  A 36' wheel has 561 rev/mile and a 34" wheel has 588 rev/mile. 

Assumption 2:  If the wheels of a car derail, it will they always turn at different rates.   Do you have any proof of this or is it an assumption?  If you include variance to accommodate wheel wear and braking differentials, do you know that a derailed wheel turns slower than those variances? If normal wear has a differential of 30-40 rev/mile will the differential of a derailed wheel be more than that?  Do you know that or is it an assumption?

Assumption 3:  If a car derails, one of the outboard axles will always be derailed.  What happens if just an inboard axle derails?

Assumption 4:  Both ends of the car can't derail at the same time.  If both ends derail at the same time or near same time (rail rolls over) will derailed wheels turn at rates with a variance greater than the normal variance?

With a locomotive wheel slip system, the most it does is adjust the power to the axle, so if it is "wrong" the consequences are relatively minor.  In this case if there is a false positive, the train catastrophically fails (it stops the whole train).  Big difference in the requirements for accuracy. 

 

Dave H. Painted side goes up. My website : wnbranch.com

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Wednesday, February 19, 2014 9:29 AM

While I welcome a range of discussion on the topic, this thread has divided into a parallel discussion of two completely different brake system concepts.  For anybody looking at it for the first time, it will be impossible to realize that.  It will look like all the technical features run together, as opposed to being two separate proposals.  So, as a result, the thread title will be perceived to offer a solution of enormous technological complexity that is sure to outweigh any advantage.

But to the question at hand:  Is monitoring wheel rotation speed really the simplest way to detect a derailment?

If that method is used, I would think you would need to sense the rotation speed of every axle in the train, and compare that rotation rate to the average rotation rate of all the axles.  Why can’t you just sense the vibration on each truck and compare that to the normal vibration? A change in vibration at the onset of a derailment seems like it would be a much greater change than a change in axle rotation speed.  Thus, the vibration change might be simpler to differentiate from normal, and it has the potential to sense the derailment as early as possible.  Any wheel on the ties will be a perfect vibration producer.

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Wednesday, February 19, 2014 10:01 AM

Euclid:  YOu are making the matter much more compllicated than it is, Euclid.  Any derailmen of any one car will definnitely and in all cases cause a difference in rotation speed between any two of the four axles of the car if both trucks leave the rails and between the axles of one truck that is the only truck t hat leaves the rails and between its axles and the axles of the truck still on the rails.   Thus, with all cars reportiing to the locomotive and to the brake system, the moment any one truck leaaves the rails, the moment any one axle leaves the rails, emergency braking is implemented with the entineer simultaneously notified.  It is a simple system, using existing componants, and it will work.

Using sensors requires a great deal of research, including experimental deraimens, testing to insure that hard coiuplings, low joints, switch points, frogs, insulated joints, don't register false deraklment notices.

And note that your differential braking may create a real hazard in a specific situation because the front of the train won't stop quickly enough where uniform fast braking would have avoided a hit.  Also, magneting track brakes are difficult to apply differentially, and test may prove they do have value in stopping a train uniformly as fast as possible.  Finally, the technology to prevent stringing with differential braking on curves still has to be developed.

DeH:    I have seen movies of derailments and all have the characteristics I've assumed.   The difference in rotational speeds that would trigger the derailment notice would be calculated to be larger than differences in speeds between wheels slipping on the inside rail vs. those slipping on the outside rail.  Actually, although this situation is poasible and thus must be accomodated in the emergency system, in most cases the ouside wheels are running closer to the flange and thus at a larger diameter and circumpherence than the inner wheels (tapered wheel profiles) and slippage is minimized.

I happen to have two MIT engineering degrees, and although my practical onboard railroad expereicne was a long time ago, I did have such experience, plos doing all kinds of work at a trolley museum subsequently.  I think I can make the kind of recommendations I am making with confidence that they will be successful if applied.

The kind of track brakes I'm recommending are the off-the-shelf variety used on light rail cars, modified to fit exactly mounting inside regular freight truck frames and with any lessons as appled to the Indiana High Speeds and C&LE high-speed cars.   And the Inidana cars did operate mu, three in regular service, four in speical occasions.

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Wednesday, February 19, 2014 10:38 AM

Again, I am suggesting application only to uniform well-maintained trains, where braking effort can be uniform, ditto other railcar characteristics.

Wheels running on the ground have have flange diameter rotational speed.   On rails it is tread diameter rotational speed.

Wheels on cars digging into the ground and thus just starting to jacknife, stop turning!

I even once witnessed a slow-speed drailment from trackside.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, February 19, 2014 1:21 PM


The semantic question is that the thread is about 'safe oil trains' now... and not just the original system, at least some details of which have been essentially deprecated (like the bulletproof centersills and massive drawbars doing all the business). 

The 'fork' occurred somewhere in the discussion of differential braking, where Dave decided to run with the idea of bringing the whole train to a halt quickly, and your approach went toward modulating the braking just so that tension across the presumably-derailed cars was optimized.  I'd recommend establishing two new threads, one for magnetic emergency braking and one for differential braking as an adjunct to stopping derailed trains more effectively.

Euclid

But to the question at hand:  Is monitoring wheel rotation speed really the simplest way to detect a derailment?

No, it's not; that's Dave Klepper's opinion. and it only makes sense to detect a derailment if you already need the rotary encoder(s) for another purpose.  The wheels may continue to spin with considerable rotational inertia, or may continue to spin bumping over the ties (with the considerable 'adhesive' weight encouraging the transfer of energy into their rotation).  If you design the system for 'slight' differences, you'll be back into the fun that occurs when wheels wear to different profiles and effective diameters.

If that method is used, I would think you would need to sense the rotation speed of every axle in the train, and compare that rotation rate to the average rotation rate of all the axles.

I wouldn't think so.  Even comparing adjacent trucks would do the job, and that's a simple distributed function for 'smart cars' without their needing to know location in the train (as they would if differential braking were involved).


Why can’t you just sense the vibration on each truck and compare that to the normal vibration?

You can, and I've already said you can, and I've already said that I think this represents a 'better' idea for detecting derailments.  My original device was intended to detect and record periodic vibrations that could be interpreted as flats or other wheel defects, but it is also capable of determining a number of other vibrations and moments that signal different types of failure (e.g. truck skew, bearing failure, and derailment) that require prompt attention.  The device can use a number of modalities to keep itself charged, and to communicate with other equipment both on the train and in wayside or portable detectors.  Etc.

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Wednesday, February 19, 2014 8:26 PM

Overmod
The semantic question is that the thread is about 'safe oil trains' now... and not just the original system, at least some details of which have been essentially deprecated (like the bulletproof centersills and massive drawbars doing all the business). 

Yes, new ideas attract deprecation.  What fun would inventing be without the doubters?  I once did some work for a company where they would assemble us into a big group to brainstorm how to introduce a new product that could be put on the market.  But the human nature of the effort assured that there could be no new product that would work.  Everybody shot down everyone else’s ideas until there were no ideas.  And then we all moved on to the next challenge.

Often the deprecation here has depended on revising what I have said in order to make it less realistic for the purpose of facilitating the deprecation. For instance, I never said that I was proposing “bulletproof centersills and massive drawbars doing all the business,” as you say. 

All I said was “extra strength solid drawbars and continuous extra heavy car center sills.”

It was others who immediately concluded that I was proposing making the drawbars so strong that their weight would increase by over 100 times.  That way, they could say the idea would not work because by adding 100 tons to the drawbars, the car would not be able to carry the weight of the oil.    

I also never said that the extra strength drawbars would be “doing all the business.”  I only said they were one element of a system.  The system relies on pulling into the derailment in order to stretch it out, and a parting of the train defeats that purpose. 

So it makes sense to add some drawbar strength and coupling integrity to help resist the tension inherent with this system as it extends through derailing cars.  But that objective is nothing new.  Drawbars have been upgraded and improved several times for that same objective of making them stronger and less likely to uncouple during derailments.

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Thursday, February 20, 2014 12:20 AM

Euclid, I agree that new ideas should have a forum and be discussed.   I discussed what I see as problems with your idea in previous postings and need not repeat them.  If you still have faith in your concept, by all means contact people in the industry that you think can implement them, as I have tried to do with mine.  I'll post any responses that come my way and you can do the same.  I probably would not have developed my scheme if I had not seen some issues with yours.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Thursday, February 20, 2014 5:23 AM

Euclid
But for the drawbars to play that role, the outer edges of the car ends must be reinforced to withstand the extreme compression that is the reaction force to the drawbar tensile load as the cars try to jackknife.  This is the buffer system referred to in item #3.  It would consist of extra strong corner block structures that would be spaced from one car to the next only far enough apart to allow the cars to traverse curves. 

Stuff appears to be changing here since the version that was originally proposed.  That is a good thing, in an important sense, because it shows that the design is evolving as difficulties are perceived.  The problem is that if the original description is where the changes are made, it creates the impression -- correctly or inappropriately -- that the original inventor has predicted the 'ultimate' form of the system all along, rather than evolved things as concerns arose.  Now I distinctly remember in the early phases of discussion about how the drawbars were going to be made strong enough to resist the jackknifing forces, and how the drawgear was going to have to be optimized to take the substantial forces   The proposal as now written has much more reference to using appropriate devices and methods for the 'anti-jackknifing', and I would like to see some discussion now directed specifically at these aspects.

An important detail of the Miller passenger-car system was the provision of anticlimbers to prevent telescoping of the cars.  This involved the invention of a device which could absorb far more energy than the couplings between cars could themselves sustain, even where there is an absence of flexible and elastic connections needed for trains to traverse normal curves and switches.  A similar system is now proposed for the outer corners of the cars to help prevent jackknifing.

In a very simple form, what is required is a system of 'lateral anticlimbers' as far apart on the faces of the cars as possible, which would lock together if cars were pushed together and prevent further mutual rotation in yaw sufficient to separate couplers (or bend/break drawbar connections).  This must be relatively fast-acting, as otherwise differential momentum effects alone may build up sufficient force to damage any connection between cars, no matter how robust.  One thing to note is that the provision of some forms of longitudinal draft gear, per se, would no longer be anathema, and to an extent the mandatory use of drawbars the full length of the train is no longer critical to the design (the mutual "anti-jackknifing" and anti-roll devices between coupled ends can be different in type, size, and action from those within semipermanently-coupled 'rakes')

One immediate concern is that the mutual 'locking' has to be effective if there is more than just yaw 'misalignment' of the cars.  Should a car roll during derailment, as we have discussed, it may come far enough out of loading gage as to foul an adjacent track.   The system should, if possible, resist this tendency -- the 'catch' being that if the differential-braking method is being used, the compression alignment will not be effective in roll.  There are ways around this -- and I think they should be brainstormed, not just analyzed -- but I do not think very strong (against twist) drawbars are a good solution in and of themselves.

Note that the forces acting to produce jackknifing need to be resisted early.  For those of you that hate physics: think of the action of a hammer.  If even a few inches of travel are required to engage a pair of 'corner anticlimbers', the force exerted with 'lever action' on the center-mounted draft gear will be highly magnified.  This is different from the geometric considerations mentioned in the original description, and in my opinion ought to be noted specifically there.  The use of fins or 'teeth' to interlock the two buffer blocks on adjacent cars, while important, will only slightly ameliorate the tension induced in the drawgear.

I am wondering whether some form of automatic extension of the corner blocks, or taking up of travel between cars, might be appropriate to provide -- without pyrotechnics! -- in the event of a derailment.  In a simple system this might be provided (where ECP braking is being used) by the brake air supply, since pressure drop is no longer being utilized as the control signal.  In the case of the 'buffer' blocks, they might be moved forward by spring to the 'forward' position, with some form of progressive tumbler dropping to provide the required very strong compression resistance without jamming the cars beyond ability to traverse curves... which I think would induce derailments directly if falsely or excessively deployed.

Likewise, something that pulls the cars more closely together when severe yaw moment is detected (in other words, jackknifing needs to be strongly resisted) might be provided.  This wouldn't be a latch, but might involve some sort of positive limit stop that would act as a latch once a permissible amount of mutual yaw had been reached with progressive resistance.  (This suggests another design refinement: that the buffer blocks be designed to provide a restoring moment to the 'upright and locked' alignment when pressed together, rather than just preventing further mutual displacement or roll when engaged.  (I have no doubt that if this was not already in the original plan, it soon will be, to paraphrase Martial... ;-} ).

I think at this point Dave Klepper should start his own 'safe oil train' thread, clearly describing what his system involves and how it handles things like jackknifing and rolled-car clearance issues.  So far, the operative principle appears to be stopping the train in an absolute minimum time while preventing either buff or draft shock to any part of the train, including the derailed cars.  Those discussions are very different in nature from what is involved in the present proposal.  I don't think any of the present posts in this thread regarding Dave's system should be edited (or removed), but going forward there should be a distinction between the two approaches in all appropriate senses.

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Thursday, February 20, 2014 5:53 AM

Euclid, I thak you and appreciate the suggesion.   Again, I probably would not have thought through my ideas on this subject if you had not started the topic.

I thnk I have defined my proposal and its options pretty throoughly in past postings on this thread.  I have presented the basic ideas to both EMD amd GE via email on their commens inpus on their website.  If others have more to say one way or another concerning my damage control system thay can start a new thread and I will participate.

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Thursday, February 20, 2014 9:12 AM

Thanks Dave and Overmod for being very understanding of my concerns.  I agree that there is nothing wrong with anybody posting alternatives here to what I am proposing here. My only concern was about the possible appearance that there was just one proposal being discussed with a doubling of explanation, which would obscure both proposals.

I agree that any ideas such as these have to be shown to people in the industry to have any effect.  And for such a showing, the concept has to be presented in the clearest manner possible.  Communicating the details and purpose of an invention can be an enormous challenge.  Sometimes a prototype or a model is beneficial to the explanation, but in the case of these railroad systems, a prototype is out of the question due to its size and cost.  A physical model might be an option, but I would choose a diagram with illustrations and text callouts showing a sequence of events in the operation of the concept.  Every word needs to be carefully chosen in order to get the most clarity for the fewest words.

Overmod,

Regarding your comments about the buffer system.  I did indeed modify my proposal by removing the buffer system which was part of the original explanation on page one.  In that original intent, the drawbars were made stronger to resist the added tension of pulling on the derailed cars.  Then, I saw that this added tensile strength could be taken advantage of in preventing jackknifing.  That would be accomplished by adding buffers, which would resist jackknifing by going into compression in reaction to the couplers going into tension. 

However, in thinking about it, I realized that the buffers would have to be exceptionally strong to serve the anti-jackknifing function.  And it requires two buffers to serve one drawbar.  So I decided to dispense with the buffers and rely solely on the fundamental anti-jackknifing feature of pulling on the train into the derailing cars.   

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Thursday, February 20, 2014 10:19 AM

Euclid
I realized that the buffers would have to be exceptionally strong to serve the anti-jackknifing function.  And it requires two buffers to serve one drawbar.  So I decided to dispense with the buffers and rely solely on the fundamental anti-jackknifing feature of pulling on the train into the derailing cars.   

I think both features are exceptionally important, and both are warranted far above the idea of trying to put substantial strength into the draft gear and drawbars, etc.

The differential proportional braking is important, and it will work as desired in many cases.  There are many, many other cases, however, in which it won't.  For those situations the 'buffer' anti-overrun devices will be important, particularly when you have enhanced methods of keeping the cars from separating physically.  Even if one or two derailed cars 'get sideways', a method of keeping the cars behind that point physically aligned with the track direction is a positive aspect...

The immediate question is whether adding the appropriate reinforcement behind the buffer anti-overrun faces can be done effectively on a tank car, especially considering the increase in mass required in the tank structure alone due to the double shell.  It is at least technically possible that the force can be carried smoothly into the outer shell with appropriate engineering, and some 'controlled crush' structure might prove practicable without necessarily puncturing the inner shell.  I do think something better than pathetic little after-the-fact shields needs to be worked up...

I do recommend that you retain both design features in the proposals you make.

  • Member since
    June 2002
  • 20,096 posts
Posted by daveklepper on Thursday, February 20, 2014 11:13 AM

But you have to solve the stringing on curve derailments problem and how you can solve that problem is behond my current  comprehension. Also, what if the engineer knows there is worse danger ahead and only the quickest possible stop will avoid disaster?

  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Thursday, February 20, 2014 8:50 PM

daveklepper
But you have to solve the stringing on curve derailments problem and how you can solve that problem is behond my current  comprehension. Also, what if the engineer knows there is worse danger ahead and only the quickest possible stop will avoid disaster?

There will not be any stringlining caused by this system.  A key point that may not be clear is this: 

When I refer to placing the leading cars under tension to pull on the derailing cars, I am not saying that this will be an extraordinary high amount of pulling force.   The tension that I am referring to would only function as a kind of preload tension ahead of the derailed cars to keep them from becoming an obstacle to the trailing cars pushing forward with their kinetic energy. 

In the absence of such tension, if the force of the leading cars and the trailing cars could be kept identical, there would be no tendency to jackknife the derailing cars.  But I see this preloading tension as extra insurance against the possibility that the force of the leading and trailing cars might vary somewhat.  If it varies in a way that the leading cars decelerate slightly quicker than the trailing cars, the system would fail its purpose.  So the preload tension on the leading cars guards against that failure.   

The tendency for stringlining increases with the sharpness of a curve and I don’t anticipate the tension applied by this system during a derailment to be great enough to stringline most mainline curves.  The greater the speed, the lower the amount of extra tension that will be needed.  And the lower the speed, the less need there will be for the anti-jacknifing system.

But beyond the natural unlikelihood of stringlining a curve will be the smart system control itself, which will prevent stringlining. I mentioned that my system would have a controller that would account for factors such as speed and tonnage. One other factor it would account for is location of the train on the line. This would tell it whether the train was rounding a curve or not.  So the system controlling the application of tension would factor in the curve and limit tension if necessary to prevent stringlining.

  • Member since
    September 2003
  • From: Omaha, NE
  • 10,621 posts
Posted by dehusman on Friday, February 21, 2014 6:12 AM

Euclid

When I refer to placing the leading cars under tension to pull on the derailing cars, I am not saying that this will be an extraordinary high amount of pulling force.   The tension that I am referring to would only function as a kind of preload tension ahead of the derailed cars to keep them from becoming an obstacle to the trailing cars pushing forward with their kinetic energy. 

What are going to do going downhill?  The train will be using dynamic brakes which means it will be in compression.  How are you going to create this tension and still stop the train going downhill?

Dave H. Painted side goes up. My website : wnbranch.com

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, February 21, 2014 6:29 AM

dehusman
What are going to do going downhill?  The train will be using dynamic brakes which means it will be in compression.  How are you going to create this tension and still stop the train going downhill?

Well, first by law you're going to be under the speed restrictions applicable if you weren't using dynamic (particularly after the Seventeen Mile Grade problems!), and second, you're going to have the modulated brakes applied to hold the train appropriately.

All the automatic system does, when it detects a derailment (due to changes in vibration, or 'differential rotation' or whatever) is to modulate the braking behind the point or 'car' of detection to a slightly greater rate, enough to produce the differential tension across the coupling behind the derailment. 

Now, the point you raise is nontrivial; when this system actuates, the amount of commanded dynamic becomes, more or less suddenly, highly excessive to 'requirements', and this may result in abrupt slowing down of the consist ahead of the point of derailment.  So the effective amount of proportional braking needs to take this into account in order to keep the (smaller) amount of tension between that forward consist and the leading end of the derailed section. 

I don't think that this is an insoluble, or even a particularly difficult, problem, given the existence of distributed strain gages or similar sensors in the draft gear.  An emergent question is whether there should be some sort of distributed control, or modulation, of the dynamic brake control.  I'm still a bit concerned about active traction control on automobiles, so this isn't as simple as Popular Science might have you believe...  ;-}

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