Trains.com

How to derail a train or stupid railroading

4650 views
75 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    January 2014
  • 8,221 posts
Posted by Euclid on Friday, July 12, 2019 9:33 AM

BaltACD
 
JPS1
 
oltmannd
 Which even my "back of the envelope" calculation showed is what likely happened.

Getting to L/V is tougher to do and you can't reasonably determine it for every car in every location under every possible circumstance.  There are trains built and run that will just plain derail if the go into emergency at the right spot.  But, they don't, so no news is good news...  

How long did it take from the beginning of the railroad era for engineers to work out the mathematics to explain all the forces impacting operations?  Or is it still a work in progress?

 

Still a work in progress as they try to maximize tonnage and minimize costs.

 

I would think they would have a program that would predict the stringline potential of any given train, based on makeup, gradient, power, and curvature.  Maybe it could also take over and prevent the stringline event from happening. 

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Friday, July 12, 2019 9:12 AM

JPS1
 
oltmannd
 Which even my "back of the envelope" calculation showed is what likely happened.

Getting to L/V is tougher to do and you can't reasonably determine it for every car in every location under every possible circumstance.  There are trains built and run that will just plain derail if the go into emergency at the right spot.  But, they don't, so no news is good news...  

How long did it take from the beginning of the railroad era for engineers to work out the mathematics to explain all the forces impacting operations?  Or is it still a work in progress?

Still a work in progress as they try to maximize tonnage and minimize costs.

Never too old to have a happy childhood!

              

  • Member since
    December 2018
  • 865 posts
Posted by JPS1 on Friday, July 12, 2019 8:33 AM

oltmannd
 Which even my "back of the envelope" calculation showed is what likely happened.

Getting to L/V is tougher to do and you can't reasonably determine it for every car in every location under every possible circumstance.  There are trains built and run that will just plain derail if the go into emergency at the right spot.  But, they don't, so no news is good news... 

How long did it take from the beginning of the railroad era for engineers to work out the mathematics to explain all the forces impacting trains operations?  Or is it still a work in progress?

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, July 11, 2019 9:06 PM

petitnj
Stringlining occurs when the pull of the locomotive at the curve angle exceeds the gravity pull. 

Which even my "back of the envelope" calculation showed is what likely happened.

Getting to L/V is tougher to do and you can't reasonably determine it for every car in every location under every possible circumstance.  There are trains built and run that will just plain derail if the go into emergency at the right spot.  But, they don't, so no news is good news...

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    October 2001
  • From: US
  • 591 posts
Posted by petitnj on Thursday, July 11, 2019 7:18 PM

Just to add some physics details to the issue. What keeps trains on the track is the ratio of lateral to vertical force. Lateral is velocity squared/radius toward the outside of the curve and pulling times the sine of the car angles for pull inside the curve. Vertical is gravity. Safe speeds are designed to keep L/V less that 0.5 and some specify less than 0.3.  The accident in Seattle last year exceeded L/V to about 1.0 and the train flew off the track. Stringlining occurs when the pull of the locomotive at the curve angle exceeds the gravity pull. 

 

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, July 11, 2019 6:29 PM

timz

Wonder what the superelevation is -- maybe 2 inches? Think that has much effect?

(The 2008 chart says four inches -- page 34 of the PDF)

http://multimodalways.org/docs/railroads/companies/NS/NS%20Track%20Charts/NS%20Pgh%20Division%20Track%20Chart%202008.pdf

What was the tonnage?

 

Probably still four inches. That'll help things along a bit.  Tonnage around 8000? 140+ cars?  Train resistance balanced out at ~10 mph

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    February 2005
  • 2,366 posts
Posted by timz on Thursday, July 11, 2019 5:56 PM

Wonder what the superelevation is -- maybe 2 inches? Think that has much effect?

(The 2008 chart says four inches -- page 34 of the PDF)

http://multimodalways.org/docs/railroads/companies/NS/NS%20Track%20Charts/NS%20Pgh%20Division%20Track%20Chart%202008.pdf

What was the tonnage?

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, July 11, 2019 5:20 PM

BigJim

 

 
oltmannd

BigJim

So, did you amend the rule book/timetable to limit trailing tonnage on said car?

 

 

 

Yes.  Headed axle limit on trains ascending HSC w/o helpers. Also tonnage limit. 

 

 

 

Was this before or after the fact? And, what particular rule number and tonnage rating?
I ask because as I have said before, in my time there was no restrictions on centerbeam flat cars.

 

 

If there were only two unit on line and the train was pushed from the rear, the derailment doesn't occur.   Car length becomes irrelevant.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    April 2001
  • From: Roanoke, VA
  • 2,019 posts
Posted by BigJim on Thursday, July 11, 2019 1:46 PM

oltmannd

BigJim

So, did you amend the rule book/timetable to limit trailing tonnage on said car?

 

 

 

Yes.  Headed axle limit on trains ascending HSC w/o helpers. Also tonnage limit. 

 

Was this before or after the fact? And, what particular rule number and tonnage rating?
I ask because as I have said before, in my time there was no restrictions on centerbeam flat cars.

.

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Thursday, July 11, 2019 1:10 PM

Excerpt from CSX Baltimore Division Timetable in 2015

CSX Batltimore Division Timetable No. 1 April 1, 2015

4466 PLACING EMPTY CARS IN TRAINS

Empty Car Placement Train Classification Instructions for Manifest Trains:

Empty cars 80 feet and longer (other than a box car) must be placed in the train in such a location that the trailing tonnage behind these empty cars does not exceed the amount listed below. In territory where helper locomotives are used on the rear of the train, their tonnage rating should be subtracted to the trailing tonnage listed below when determining the location for the restricted car(s):

Between          Direction         Tonnage
Hyndman & Sand Patch Westward 3,500
Connellsville & Sand Patch Eastward 5,100
Connellsville & New Castle Eastward & Westward 13,300

Never too old to have a happy childhood!

              

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Thursday, July 11, 2019 12:53 PM

SD70Dude
 
oltmannd
BigJim

So, did you amend the rule book/timetable to limit trailing tonnage on said car?

Yes.  Headed axle limit on trains ascending HSC w/o helpers. Also tonnage limit.  

Looks like NS 'forgot' your instructions.  Oops.

NS forgot their Institutional Knowledge from those they let go putting in PSR.

Never too old to have a happy childhood!

              

  • Member since
    December 2017
  • From: I've been everywhere, man
  • 4,269 posts
Posted by SD70Dude on Thursday, July 11, 2019 12:39 PM

oltmannd
BigJim

So, did you amend the rule book/timetable to limit trailing tonnage on said car?

Yes.  Headed axle limit on trains ascending HSC w/o helpers. Also tonnage limit. 

Looks like NS 'forgot' your instructions.  Oops.

Greetings from Alberta

-an Articulate Malcontent

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, July 11, 2019 5:45 AM

BigJim

So, did you amend the rule book/timetable to limit trailing tonnage on said car?

 

Yes.  Headed axle limit on trains ascending HSC w/o helpers. Also tonnage limit. 

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2019
  • 1,686 posts
Posted by Erik_Mag on Wednesday, July 10, 2019 11:43 PM

oltmannd

All weather adhesion rating for AC units is around 35%.  For DC units with computer based wheel creep control, about 27-28%.  Non-wheel creep (Dash 2, 7 Series), 18-21%.

The 18-21% is in line with the 16% dispatchable factor adhesion given in the 1968 MR article on grades with a bit better control. The 35% available with AC could be done with DC motors with separately excited field windings and pole face windings, but motors with these features would be more like the traction motors in the PRR DD1's.

One other figure from the 60's was a 33% factor of adhesion for the KM desel hydraulics, with the combination of smooth torque and the three axles on a truck running off a common Cardan shaft. Demonstrator testing of the U25B's gave a 25% factor of adhesion on clean dry "western" rail. Durng one dynamometer test, the GE Little Joe demonstrator achieved 35% factor of adhesion - I've seen a few comments about traction current enhancing adhesion by burning away contaminants.

You want to dispatch trains on what you can count on, day in and day out.

 
No argument from me on that - the 1968 article in trains stated that 25% was achievable on clean dry rail, 33% with sanded dry rail and 16% was what you could count on day in and day out. OTOH, the MR figures could be partially available from steam locomotive experience, with the pulsating torque reducing average factor of adhesion. The torque from a properly designed and fed three phase motor can be very smooth, added with a very fast torque fall-off with speed results in optimal conditions for adhesion.
  • Member since
    April 2001
  • From: Roanoke, VA
  • 2,019 posts
Posted by BigJim on Wednesday, July 10, 2019 9:09 PM

oltmannd
You ever had a full tonnage train with three 4000 HP locomotives running in notch 8 at 10 mph on a 9 degree curve with an 80' empty head out? Where and when?


I have not had an 80' flat head out and the cars in question were not "head out".
I have had centerbeam flats near the head end with three 4000 HP engines on line and dragging. And, in my day, centerbeam cars had no restrictions on them. In fact we were glad to see a centerbeam instead of a bulkhead flat as the speed of bulkhead flat were restricted to 45 mph.

.

  • Member since
    April 2001
  • From: Roanoke, VA
  • 2,019 posts
Posted by BigJim on Wednesday, July 10, 2019 8:57 PM

So, did you amend the rule book/timetable to limit trailing tonnage on said car?

.

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 8:39 PM

BigJim
Well, that is your mind talking, not the mechanical dept. Can you prove that even a 73' centerbeam was in the derailment?

17 years in Conrail Mechanical Department - staff engineering position.  Did a lot of locomotive application work, some derailment analysis work and testing.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 8:35 PM

https://www.mrc-rail.com/project/centerbeam/

https://www.bnsf.com/ship-with-bnsf/ways-of-shipping/equipment/pdf/73Centerbeam.pdf

https://www.csx.com/index.cfm/customers/resources/equipment/railroad-equipment/

https://www.up.com/customers/all/equipment/descriptions/centerbeams/index.htm

Standard centerbeam cars are 73' between the bulkheads.  A 73' centerbeam car is 80' 6-1/2" between pulling faces.  There are 61' centerbeam cars still rolling around, but these are not the current standard.

 

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 8:24 PM

BigJim
Has NS said anything about the cause of the derailment or are we to just take your word for it? 

They haven't and probably won't say much.  Your choice if you want to beleive what I've laid out or not.  

You ever had a full tonnage train with three 4000 HP locomotives running in notch 8 at 10 mph on a 9 degree curve with an 80' empty head out? Where and when?

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    December 2001
  • From: Northern New York
  • 25,021 posts
Posted by tree68 on Wednesday, July 10, 2019 7:51 PM

BigJim
Can you prove that even a 73' centerbeam was in the derailment?

They were definitely centerbeams, although the length could be debated.

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
    April 2001
  • From: Roanoke, VA
  • 2,019 posts
Posted by BigJim on Wednesday, July 10, 2019 6:09 PM

oltmannd
The 73' centerbeam I found to use for my estimate is 80' over the pulling faces, which is what counts in this case.

Well, that is your mind talking, not the mechanical dept. Can you prove that even a 73' centerbeam was in the derailment?
The only thing that I have ever seen was that empty 89' flats were limited to trailing tonnage. I've had plenty of center beam flats at or near the head end of full tonnage trains and never had any problem.
Has NS said anything about the cause of the derailment or are we to just take your word for it? 

.

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 4:52 PM

Erik_Mag
The two chief advantages of an AC induction motor is a much more rugged rotor (conductors are uninsulated bars vs insulated wires) and inherent almost constant speed operation (DC series motor speed can increase dramatically with loss of load). The former is good for higher continuous tractive effort and the latter is good for better adhesion.

All weather adhesion rating for AC units is around 35%.  For DC units with computer based wheel creep control, about 27-28%.  Non-wheel creep (Dash 2, 7 Series), 18-21%.

You want to dispatch trains on what you can count on, day in and day out.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2019
  • 1,686 posts
Posted by Erik_Mag on Wednesday, July 10, 2019 2:15 PM

My 2 cents on the AC versus DC.

The two chief advantages of an AC induction motor is a much more rugged rotor (conductors are uninsulated bars vs insulated wires) and inherent almost constant speed operation (DC series motor speed can increase dramatically with loss of load). The former is good for higher continuous tractive effort and the latter is good for better adhesion.

A further advantge of modern AC locomotives is indivdual control of the motors, so power reduction could be limited to the motor that is slipping as opposed to reducing power to the whole locomotive.

Response time of the DC series motor is limited by the inductance of the field winding limiting how fast the the motor current can be varied, a DC shunt motor (e.g. original AEM-7) can respond faster. Ultimate limit for response time for an AC motor is roughly the period of the highest frequency that will provide full power output of the motor.

  • Member since
    July 2010
  • 351 posts
Posted by ns145 on Wednesday, July 10, 2019 11:22 AM

Overmod

 

 
BaltACD
... note CSX derated the 4400 HP the engines were purchased with to 4000 HP for fuel economy savings

 

Interesting that NS seems to have done just the opposite.  As I recall it, their dash-9-40W locomotives were delivered from GE with the 4000hp nominal rating 'for fuel savings', but with a simple override switch that would let them develop the 4400hp on demand; NS subsequently uprated quite a few of them between 2013 and 2015, but  I believe kept the -40 designation on the cab to maintain the class distinction for parts and support.

Did CSX adopt the same electronics for their derating program as GE provided in the NS C40-9Ws, and can the derating be 'overridden' as easily if wanted?

 

 

According to http://www.nsdash9.com/roster.html all of NS' Dash 9's were uprated to 4400 HP and reclassified as D9-44CW's between 10-2013 and 09-2014.  However, not all of the units have had their external model designation lettering updated to reflect the internal changes.  If I recall correctly, this was done during NS' 2014 meltdown in an effort to boost their system velocity numbers.  NS uprated their DC GEVO fleet and reclassified them as ES44DC's at the same time.         

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 10:16 AM

JPS1

 

 
caldreamer
 I ran some calculations for the tractive effort algorythm.  It is pretty close for DC engines, but is way off (too low) for AC engines. 

 

Why would the tractive effort of a DC locomotive be different than that of an AC locomotive?   

 

Well, it is and it isn't.  It depends on how fast you are going, for the most part.

At 20 mph, there wouldn't be any difference.  It's just force x speed = power  since you are nowhere near the adhesion limit and thermal limit (DC only) of the propulsion system.  The AC locomotive might give a bit more since it's electrical transmission system probably has less losses (I think...).

At low speeds, the AC unit can give you full HP down to lower speeds because it's adhesion is better (35% you can count on).  Best you might get from DC is 27%.  AC adjusts the frequency to match speed.  DC has to react to slip by backing down main generator excitation which is slower due to the nature of the machine (inductive windings resist current change)

DC motors also have thermal limits which stop high TE levels from being produced for very long even when available adhesion is good.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 10:06 AM

zugmann

 

 
oltmannd
A three unit consist of SD70ACes operating in notch 8 at 10 mph makes about 400,000# TE. (you can get the speed by timing the Virtual Railroading youtube video and TE = HP x 308/speed)

 

Actually, it was two 70aces and one 70m-2.  Still a lot of TE, but a little less than 3 ACs.

 

I should have caught that.  Actually my 10 mph speed estimate is probably off by more than the TE difference...

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, July 10, 2019 10:04 AM

BigJim

 

 
oltmannd
An 80' centerbeam car

 

Can you prove to me that there was actually an EIGHTY FOOT centerbeam car in the consist? I've never seen one that long and I definately am not seeing one in the published videos.
There was a stringline derailment like this years ago on the Hagerstown Dist. without any long cars involved. If I am not mistaken, it was due to an undesired emergency brake application back near the rear of the train.

 

 

The 73' centerbeam I found to use for my estimate is 80' over the pulling faces, which is what counts in this case.

-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, July 10, 2019 9:41 AM

BaltACD
... note CSX derated the 4400 HP the engines were purchased with to 4000 HP for fuel economy savings

Interesting that NS seems to have done just the opposite.  As I recall it, their dash-9-40W locomotives were delivered from GE with the 4000hp nominal rating 'for fuel savings', but with a simple override switch that would let them develop the 4400hp on demand; NS subsequently uprated quite a few of them between 2013 and 2015, but  I believe kept the -40 designation on the cab to maintain the class distinction for parts and support.

Did CSX adopt the same electronics for their derating program as GE provided in the NS C40-9Ws, and can the derating be 'overridden' as easily if wanted?

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, July 10, 2019 9:31 AM

JPS1
Why would the tractive effort of a DC locomotive be different than that of an AC locomotive? 

Several reasons, the most important of which is that there can be essentially zero derating of the motors both right down to 'locked rotor' zero-speed operation and to avoid wheelslip above the 'microslipping' that gives best creep control.  DC locomotives would require wildly more capable cooling arrangements, probably including sprayed coolant in the blown air, even to approximate what a good induction motor fed a properly-synthesized waveform can do.

The matter of adhesion limit is also a bit different, although I'll leave it up to Erik to make the electrical arguments comprehensible in English.  DC motors typically 'cut power' in some way to accomplish creep control; AC drive can react in different ways that keep full nominal excitation strength active but keep wheelslip from propagating.

We can discuss the pros and cons of larger wheels if those aren't already known to you. 

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Wednesday, July 10, 2019 9:18 AM

JPS1
 
caldreamer
 I ran some calculations for the tractive effort algorythm.  It is pretty close for DC engines, but is way off (too low) for AC engines.  

Why would the tractive effort of a DC locomotive be different than that of an AC locomotive?   

The why's require the expertise of a electrical/mechanical engineer - that I am not.  However, the AC's have a more effective wheelslip control set up than to DC's.

The reality is that DC's have less tractive effort than AC's of the same power and axle configuration. On CSX's Baltimore Belt Line grade (Howard Street Tunnel) a CW44AC is rated for 4900 tons, a CW40DC (EVO's) is rated 3600 tons (note CSX derated the 4400 HP the engines were purchased with to 4000 HP for fuel economy savings). 

Never too old to have a happy childhood!

              

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