CSX/Amtrak crash-un-powered armstrong Mainline switches are supposed to have signal actuators?

2516 views
52 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    April, 2002
  • From: Delmarva Peninsula
  • 100 posts
Posted by SealBook27 on Wednesday, February 07, 2018 1:28 PM
Wasn't there a similar accident in SC a few years ago; a misaligned switch resulting in a collision between two freight trains?
  • Member since
    November, 2013
  • 522 posts
Posted by VOLKER LANDWEHR on Wednesday, February 07, 2018 3:12 PM

BaltACD
I suspect trying to interface relay dependent circuitry with electronic circuitry can be a real adventure.  My guess is that the railroads felt it was more economic and more reliable to do a 'clean installation' of 21st Century signal technology when installing equipment to support PTC requirements.

My question is why designing a PTS system based on GPS and then not use the advantages like e.g.: lineside signals only in specific locations, rolling block?

Is the chosen PTC system in this light really the most cost effective?
Regards, Volker 

  • Member since
    May, 2003
  • From: US
  • 13,202 posts
Posted by BaltACD on Wednesday, February 07, 2018 3:22 PM

VOLKER LANDWEHR
 
BaltACD
I suspect trying to interface relay dependent circuitry with electronic circuitry can be a real adventure.  My guess is that the railroads felt it was more economic and more reliable to do a 'clean installation' of 21st Century signal technology when installing equipment to support PTC requirements. 

My question is why designing a PTS system based on GPS and then not use the advantages like e.g.: lineside signals only in specific locations, rolling block?

Is the chosen PTC system in this light really the most cost effective?
Regards, Volker 

My understanding is that the civilian use of GPS does not have the required precision to be dependable in reliabily differentiating between #1 and #2 track at any given location, the track centers of each track are within the allowable margin of error.

CSX has a number of locomotives that are equipped with lateral accelerometers to find and report 'rough spots' in the track structure.  These locomotives transmit a message to the MofW Desk in Jacksonville with the GPS derived coordinates of the incident that triggered the message.  When communicated to the MofW Supervisor for the particular territory the proviso is that ALL tracks in the area must be inspected for the defect - the GPS precision is not sufficient to indicate the specific track.

The Military used of GPS may be more precise than Civilian - I don't know.

         

Never too old to have a happy childhood!

  • Member since
    January, 2009
  • From: Maryland
  • 6,785 posts
Posted by ATLANTIC CENTRAL on Wednesday, February 07, 2018 3:24 PM

VOLKER LANDWEHR

 

 
BaltACD
I suspect trying to interface relay dependent circuitry with electronic circuitry can be a real adventure.  My guess is that the railroads felt it was more economic and more reliable to do a 'clean installation' of 21st Century signal technology when installing equipment to support PTC requirements.

 

My question is why designing a PTS system based on GPS and then not use the advantages like e.g.: lineside signals only in specific locations, rolling block?

Is the chosen PTC system in this light really the most cost effective?
Regards, Volker 

 

I did lots of control work (not railroad signals) back in relay days and in the early days of solid state - we interfaced the two with no problem.

I'm not really up to date on any of this, but the idea of abandoning line side signals and relying completely on GPS and similar wireless communication seems like a really bad idea.

Rolling blocks beyond the current permissive signals also sounds like a really bad idea.

Equipment does fail, I don't care relay or solid state, radio or hard wire. Redundant with specific absolutes seems like a good idea.

When cab signals came along no body said "lets take down all the signal masts"?

Why should PTC be any different?

As explained over and over by me and others, the old system, had it been working, would have prevented this crash.

Sheldon

    

  • Member since
    December, 2001
  • From: Northern New York
  • 17,840 posts
Posted by tree68 on Wednesday, February 07, 2018 3:27 PM

SealBook27
Wasn't there a similar accident in SC a few years ago; a misaligned switch resulting in a collision between two freight trains?

The incident at Graniteville, SC is the reason we now have things like the SPAF.

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
    November, 2013
  • 522 posts
Posted by VOLKER LANDWEHR on Wednesday, February 07, 2018 4:09 PM

ATLANTIC CENTRAL
As explained over and over by me and others, the old system, had it been working, would have prevented this crash.

Yes, but not others like Amtrak 501.

What Overmod formulated first and I took up in my question is mentioned in the FRA cost-benefit analysis of PTC, IIRC. A number of independent industry experts criticized the railroads for asigning the signal department with the PTC development and thus giving away the named possible advantages. 

Rolling block would rise the track capacity without the need of more track.

Where is the back-up for the signal system? If equipment fails, it is designed failsafe. And the same can be done with PTC.

BTW the European Train Control System (ETCS) will, in its advanced form, remove the need for lineside signals and allow a movable/rolling block.
Regards, Volker

 

  • Member since
    September, 2003
  • 5,086 posts
Posted by Overmod on Wednesday, February 07, 2018 5:04 PM

ATLANTIC CENTRAL
When cab signals came along nobody said "lets take down all the signal masts"?

In what alternate universe do you live, or what is your excuse for not having read contemporary accounts of cab-signal development?  

The very existence of automatic-block signaling in the first place is a compromise and a kludge, properly recognized as such by Frank Sprague before 1911.  As soon as continuous-aspect cab signals became practical there was no need for all the color light aspects to convey block occupancy, and even relatively little for home and distant aspects for fixed hazards.  When continuous-aspect has speed control, even primitive two-speed control, associated with it, it is comparatively simple to incorporate a 'zeroth speed' (throttle modulation to zero, gear to mid or drift, brakes to service) corresponding to passing a wayside point rather than a 'red indication' -- that requiring inductive or ramp infrastructure in addition to the tone modulation for the cab aspects, but not all the complex relays and power circuits needed to work wayside lights for blocks in either direction.

 

  • Member since
    December, 2005
  • From: Cardiff, CA
  • 2,755 posts
Posted by erikem on Wednesday, February 07, 2018 10:08 PM

BaltACD

I suspect trying to interface relay dependent circuitry with electronic circuitry can be a real adventure.  My guess is that the railroads felt it was more economic and more reliable to do a 'clean installation' of 21st Century signal technology when installing equipment to support PTC requirements.

As noted elsewhere in this thread, interfacing relay's and electronics is pretty straightforward (your concern was reasonable). One problem is that the electronics often don't have the ruggedness associated with vital circuit relays used in RR signaling.

A good example of interfacing "relays" with electronics is the audio frequency overlay used in grade crossing signals.

  • Member since
    March, 2003
  • From: Central Iowa
  • 4,092 posts
Posted by jeffhergert on Wednesday, February 07, 2018 11:46 PM

ATLANTIC CENTRAL

 

When cab signals came along no body said "lets take down all the signal masts"?

Sheldon

 

Well, actually the CNW did.  Back in the 1920s when they installed automatic train control with a two aspect cab signal (clear or restricting) on the Chicago-Council Bluffs main line they removed all their intermediate waysides.  Only wayside signals left were at interlockings and the approach signal to them.  (A couple of exceptions were in areas where waysides allowed trains from branch lines to use the main line for short segments without having to have ATC equipment.)

It wasn't until after the UP merged the CNW into it and started installing CTC that waysides started appearing enmasse.  

Jeff

  • Member since
    January, 2002
  • From: On the Ballast.
  • 5,990 posts
Posted by zugmann on Thursday, February 08, 2018 4:43 AM

ATLANTIC CENTRAL
When cab signals came along no body said "lets take down all the signal masts"?

Two railroads around here did just that.  The one just kept the signals at the interlockings, the other kept those plus the one automatic before. All the rest of the automatic signals were sent to the scrap heap.

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

Are we evil or good? Do we walk the fine line - That we'd cross if we could?


  • Member since
    March, 2016
  • 1,256 posts
Posted by CandOforprogress2 on Thursday, February 08, 2018 8:39 AM

Does anyone here have a stock photo of a unpowred railroad relay switch in question? cant find one online

  • Member since
    January, 2009
  • From: Maryland
  • 6,785 posts
Posted by ATLANTIC CENTRAL on Thursday, February 08, 2018 11:29 AM

CandOforprogress2

Does anyone here have a stock photo of a unpowred railroad relay switch in question? cant find one online

 

I don't know how they do that today, but "back in the day", no seperate electrical device was needed. The simple act of lining the switch to the siding would bridge the detection circuit on the main and generate a "red" aspect in both directions.

That's how dectection works, low voltage current flows from one rail to the other through the wheel sets. So connecting them with the diverging route of a turnout does the same thing.

When they don't want that result, they insulate sections of the turnout or rail as needed.

Sheldon

    

  • Member since
    December, 2001
  • From: Northern New York
  • 17,840 posts
Posted by tree68 on Thursday, February 08, 2018 12:49 PM

CandOforprogress2

Does anyone here have a stock photo of a unpowred railroad relay switch in question? cant find one online.

Is this what you're looking for? 

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
  • 5,086 posts
Posted by Overmod on Thursday, February 08, 2018 1:34 PM

ATLANTIC CENTRAL
I don't know how they do that today, but "back in the day", no separate electrical device was needed. The simple act of lining the switch to the siding would bridge the detection circuit on the main and generate a "red" aspect in both directions.

For as long as I can remember, that's been recognized as inadequate and dangerous.  Unlocked points may be in electrical contact sufficient to actuate relays or their more modern equivalents, but still be easy to move under the weight or lateral thrust of a train, and of course by that time the controlling signals will be invisible behind the head end.  What is needed is a separate linkage interlocked with the switch handle so that only when the switch is locked down does any route go away from red (and in proper practice this should also recognize that the shackle of the lock has been inserted to hold the handle in its locked position).

It was my understanding that electric timer locks worked directly on the mechanism to prevent its being unlocked, not just on the points to keep them from moving.  (in my not-so-humble opinion any design that does not assure that is badly flawed.)  Note that this has nothing to do with 'signal suspension' necessarily as the timers can be independently powered (e.g. via batteries recharged with a small solar array) and only run relative to when the switch is first supposed to be moved -- I used to think this idea was a kludge, and in a signal suspension where crews give verbal movement authorization without relining the switch, a fixed unavoidable delay when the crew finds they want to fix such a problem might be extremely counterproductive, but for a range of other problems involving throwing switches close to or under trains it's a good, and signal-problem-proof, kind of solution.

  • Member since
    May, 2003
  • From: US
  • 13,202 posts
Posted by BaltACD on Thursday, February 08, 2018 2:25 PM

ATLANTIC CENTRAL
 
CandOforprogress2

Does anyone here have a stock photo of a unpowred railroad relay switch in question? cant find one online 

I don't know how they do that today, but "back in the day", no seperate electrical device was needed. The simple act of lining the switch to the siding would bridge the detection circuit on the main and generate a "red" aspect in both directions.

That's how dectection works, low voltage current flows from one rail to the other through the wheel sets. So connecting them with the diverging route of a turnout does the same thing.

When they don't want that result, they insulate sections of the turnout or rail as needed.

Sheldon

What you are describing is the normal non-electric locked main track switch in signalled territory.

Electric locked switches have additional locking hardware.  Hardware that detects the immediate presence of a train on the Main Track in order to allow the switch to be unlocked and thrown for a route away from the main track.  Many of these electric locked switches have timers.  The normal switch lock must be unlocked and removed from its hasp to start the timer - after the timer run to completion the switch may be operated to route to the other than Main Track.  Some timers are of a short duration, some are of a very long duration.  Crews often attempt to throw the switch before the timer has run to completion - when this happens the timer normally restarts at the beginning again.  Many train crews give up in disgust and the customer doesn't get service.  Restoring the switch to Main Track alignment does not have any delay.  All delays are built into opening the switch for other than Main Track movement - be that from the Main or from the other track.

Opening non-electric lock switches to the Main Track, by rule crews are to wait 5 minutes before occupying the Main - supposedly this is sufficient time for a train on the Main to either make itslef KNOWN to the crew trying to enter the Main or to have the Main Track track train see a Red Signal in advance of the switch. 

         

Never too old to have a happy childhood!

  • Member since
    March, 2016
  • 1,256 posts
Posted by CandOforprogress2 on Saturday, February 10, 2018 1:44 PM

YES!!!!!

  • Member since
    March, 2016
  • 1,256 posts
Posted by CandOforprogress2 on Saturday, February 10, 2018 1:47 PM

So what are we looking at in the picture?

 

  • Member since
    March, 2003
  • From: Central Iowa
  • 4,092 posts
Posted by jeffhergert on Saturday, February 10, 2018 9:17 PM

tree68

 

 
CandOforprogress2

Does anyone here have a stock photo of a unpowred railroad relay switch in question? cant find one online.

 

Is this what you're looking for? 

 

The switch circuit controller, the silver box on the tie, is the thing that ties the switch position into the signal system.  Notice the rod running parallel to the ties between the two siver boxes.  One end is attached to the far switch point, the other to a lever on the side of the controller box.  ALL hand throw switches, and some derails, not equipped with electric locks in signalled territory have them.  (Depending on the equipment, I've seen pictures of electric lock equipped switches with the controller boxes.)  It is to ensure that the signal system is shunted when the switch is operated.  As Overmod said in a post, it's not safe to depend on the switch itself to shunt the signals.

There are a few different types of Electric Locks.  All keep the switch from being thrown until the system has determined there is no closely approaching movement or a time period has elapsed.  The ones I'm most familiar with locked the handle in place.  As Balt said, you remove the lock and it allows a foot pedal operated blocking device to slightly move, but not enough to release the handle, that starts the process.  If the adjacent block is clear, a light comes on and it will allow the pedal to be depressed and the handle to be moved.  If the block isn't clear or it's not working right, it will run time.  Ours usually were 7 to 11 minutes and the maintainers usually had painted the time period on the mechanism box.  Once the timer had run down, the light was supposed to come on and you could again operate the pedal, release and operate the handle.

Because of potential problems with electric locks, most of our hand throw switches where a train or engine might have to clear the main track in CTC territory don't have them.  They've started using what we call a "leaving signal", one that governs movement over the switch and entrance to the main track.  

Either one, electric lock or the leaving signal (when displaying a proceed indication) waive the need for waiting a prescribed time period from when the switch is first lined until the main track is occupied.

I believe in years past, there were electric locks that had to be released by the dispatcher on the CTC machine.  They could be released in the field in an emergency, but only a signal maintainer could reset them.  We had, and may still have, one like that at one location.  The switches are rarely used, there is usually another route available around them.  It's said operating them without dispatcher consent can lock up the control point they are part of.  Then a signal maintainer has to come out to reset everything.   

Jeff

  • Member since
    September, 2008
  • 1,022 posts
Posted by aegrotatio on Tuesday, February 13, 2018 9:36 PM

I'm having a problem understanding the rationale of running a passenger train at track speed (79 MPH) when the signals are out of service.

 

  • Member since
    March, 2003
  • From: Central Iowa
  • 4,092 posts
Posted by jeffhergert on Tuesday, February 13, 2018 9:50 PM

aegrotatio

I'm having a problem understanding the rationale of running a passenger train at track speed (79 MPH) when the signals are out of service.

 

 

They don't.  Maximum speed without signals is 59 mph.  Freight is 49 mph.

Jeff

  • Member since
    September, 2008
  • 1,022 posts
Posted by aegrotatio on Tuesday, February 13, 2018 10:03 PM

I understand, but the reports state the train was travelling at 80 MPH.

 

  • Member since
    September, 2003
  • 5,086 posts
Posted by Overmod on Wednesday, February 14, 2018 12:12 AM

aegrotatio
I understand, but the reports state the train was travelling at 80 MPH.

Please quote these reports.

If this is regarding the "Cayce" accident, the NTSB has provided a definitive reading off the locomotive EDR, for the timeline in question.  Permissible speed (for TWC during a signal suppression) is 59mph -- that figure representing 'not 60mph' just as 79mph represents 'not 80mph' -- and in any report I have seen based on fact, the locomotive never exceeded that speed in the signal-suppressed zone, particularly from the time the locomotive approached the switch to the point the train went into emergency.

Are you confusing this with the derailment on the speed-restricted bridge in Washington?  That's a whole different set of circumstances.

  • Member since
    May, 2003
  • From: US
  • 13,202 posts
Posted by BaltACD on Wednesday, February 14, 2018 7:36 AM

aegrotatio
I'm having a problem understanding the rationale of running a passenger train at track speed (79 MPH) when the signals are out of service.

I have a problem understanding how some one can conflate two distinctly different incidents that happen on different sides of the country.

         

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.

Newsletter Sign-Up

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

Search the Community