Stay in Line

[m7zanr160s]

What keeps subway switches from changing position while a train is running over them? I heard that a train in a switch block shorts it, making sure it can't change position, is that true? I figure that has to be top priority with an interlocking, because an error there would be disastrous.

[Joe]

A collection of relays are used to ensure safe switch movement. The actual NYCTA signal diagrams are very interesting to read and trace out, mostly in part due to the fact that the circuits are very redundant and go through many series of error-checks.

 

For example, in the process of a signal clearing, the trip has to go down. If the trip doesn't go down, the signal cannot show a clear aspect. However, as a safety measure, the signal system is designed so that the trip must move from the raised to the lowered position. This alerts motormen and signal maintainers to possible mechanical issues with the trip and/or sabotage.

[SubwayGuy]

What keeps subway switches from changing position while a train is running over them? I heard that a train in a switch block shorts it, making sure it can't change position, is that true? I figure that has to be top priority with an interlocking, because an error there would be disastrous.

 

Complicated signal relays prevent not just that from happening, but ANY unsafe move. For instance, signals and switches at opposite ends of track cannot be lined to cause two trains to collide head on. Switches cannot be moved with trains on them, and a timer must run off before they can be moved when trains are immediately approaching them (this is called approach locking of a switch). If an object in the track prevents a switch from throwing completely, the tower operator won't be able to complete the move and then provide a clear signal aspect approaching the switch. All signals protecting a switch must be red before that switch can be moved.

 

It's extremely safe. Even the oldest interlocking machines in the system can do everything I indicated above.

[EE Broadway Local]

The worst accident involving a switch in our subway was the Times Square IRT accident of Friday afternoon August 24, 1928.

 

At 4:55pm, a switch just south of the Times Square station failed. A southbound "gap" train was in the station and, behind it a regular Broadway Express train.

 

The IRT Maintenance Foreman inspected the defective switch and decided it would work if he operated it manually and his assisstant "key-byed" the trains at trackside until repair crews arrived. Warning the Towerman on duty not to touch the lever that operated the switch, they began the procedure of operating the switch manually and keying-by the trains.

 

The gap train was signaled, safely negotiated the defective switch, and moved on to the lay-up track between Times Square and 34th Street-Pennsylvania Station.

 

The regular Broadway Express entered Times Square, made its' stop, and was signaled. As the train approached the defective switch, it was travelling approximately twenty miles per hour. The first eight cars negotiated the switch and the IRT Maintenance Foreman and his assisstant were waiting for the final two cars to pass so they could make another attempt at getting the switch back in its' normal working order. As the first truck of the ninth car negotiated the switch, a faint hissing sound was heard - the unmistakable sound of air escaping from the switch valve. The switch began to curve as the second truck of the ninth car approached.

 

With the switch now curved, the first eight cars, and the first half of the ninth car continued toward 34th Street-Pennsylvania Station while the second half of the ninth car and the tenth car headed for the lay-up track.

 

The result was disastrous and horrifying. The rear truck of the ninth car leapt the defective, now curved switch, and the front quarter of the car was torn from the rest of the car. Amazingly, it briefly remained coupled to the eighth car before hitting the concrete-and-stone partititon immediately beyond the switch. The rear three quarters of the ninth car hit several steel pillars. The tenth car crashed into the ruins of the ninth car and bent like a bow. The third rail was uprooted.

 

16 passengers died and more than 100 injured. To this day, I don't think they ever figured out how the defective switch moved from straightaway to curved position as the Broadway Express was negotiating it.

 

One thing that could've been done - "spiking" the switch so it wouldn't move until repair crews arrived wasn't because of the need to continue switching gap trains which in the afternoon rush hour terminated at Times Square, used the switch to access the layover track and reverse to reenter service northbound.

[Rutgers Tube]

Complicated signal relays prevent not just that from happening, but ANY unsafe move. For instance, signals and switches at opposite ends of track cannot be lined to cause two trains to collide head on. Switches cannot be moved with trains on them, and a timer must run off before they can be moved when trains are immediately approaching them (this is called approach locking of a switch). If an object in the track prevents a switch from throwing completely, the tower operator won't be able to complete the move and then provide a clear signal aspect approaching the switch. All signals protecting a switch must be red before that switch can be moved.

 

It's extremely safe. Even the oldest interlocking machines in the system can do everything I indicated above.

 

So then it's not possible for a motorman to accept a proper line-up only to find a switch set for an incorrect move? For example, an operator punches for the Rutgers Street Tunnel at Broadway-Lafayette southbound, accepts a bottom green aspect, only to proceed a short distance before finding the switch aligned for one of the 2 Avenue pockets. And here's an oddball question, but would a shunt loss, in theory, allow for mistaken switch movement and subsequent fouling if train was passing over, or this isn't possible, or my question just doesn't even make sense? :confused:

[SubwayGuy]

So then it's not possible for a motorman to accept a proper line-up only to find a switch set for an incorrect move? For example, an operator punches for the Rutgers Street Tunnel at Broadway-Lafayette southbound, accepts a bottom green aspect, only to proceed a short distance before finding the switch aligned for one of the 2 Avenue pockets. And here's an oddball question, but would a shunt loss, in theory, allow for mistaken switch movement and subsequent fouling if train was passing over, or this isn't possible, or my question just doesn't even make sense? :confused:

 

Anything is possible. That's why part of a train operator's job is to read the iron in addition to observing the signal aspect. If it's not correct, he is NOT to accept the lineup and must call it in as a signal malfunction. Then it's up to a maintainer to figure out what's up with the relays and why there is a problem. And at that point it quickly gets beyond my understanding.

[m7zanr160s]

Complicated signal relays prevent not just that from happening, but ANY unsafe move. For instance, signals and switches at opposite ends of track cannot be lined to cause two trains to collide head on. Switches cannot be moved with trains on them, and a timer must run off before they can be moved when trains are immediately approaching them (this is called approach locking of a switch). If an object in the track prevents a switch from throwing completely, the tower operator won't be able to complete the move and then provide a clear signal aspect approaching the switch. All signals protecting a switch must be red before that switch can be moved.

 

It's extremely safe. Even the oldest interlocking machines in the system can do everything I indicated above.

 

Cool. Um...checks upon checks.

[Joe]

And here's an oddball question, but would a shunt loss, in theory, allow for mistaken switch movement and subsequent fouling if train was passing over, or this isn't possible, or my question just doesn't even make sense? :confused:

 

By shunt loss, you mean a fail in the track transformer for a given block?

[Rutgers Tube]

By shunt loss, you mean a fail in the track transformer for a given block?

 

As in a weak connection between the train wheels and the running rails which loses the short ("shunt") and consequently results in an open circuit.

 

However, I did some reading last night, and from what I've gathered with regards to automatic block signalling, a rail break or improperly-lined switch will result in an open circuit and a restricted aspect. I was curious as to what might happen if the circuit was opened while the train was already moving through the switch. I'll assume that if the circuit was open, the switch would simply not move at all, and any trailing trains would not be allowed to proceed through the interlocking until the issue was remedied.

[Joe]

You know, I was just discussing this with some people last night, but remembering the actual signal drawings is tough stuff. I'll get back to you later in the week, Rutgers, i'll refer to a signal book.

[Rutgers Tube]

You know, I was just discussing this with some people last night, but remembering the actual signal drawings is tough stuff. I'll get back to you later in the week, Rutgers, i'll refer to a signal book.

 

Thanks, Joe.

 

It's funny, because the concept of Absolute Automatic Block Signalling is very elementary, and even antiquated (there's that word again, dammit!). However, the inner workings of the signalling system are actually quite complex, a great deal more complicated than just "red, yellow and green colors, and sometimes there are two colors on the wall!" There are many, many different conditions which may yield false proceeds (think the 2009 Washington Metro collision), and so much that goes into accident prevention and mitigation of hazardous situations through proactive maintenance and repair. Obviously signal maintainers are required to know a whole lot more than simply how to change a light bulb and and which end of the screwdriver to hold - something that the general riding public takes for granted, and management takes credit for.

 

Here are a few more questions that I've been wondering since I started reading up on this topic:

 

1) Are NYC track circuits wired using one or two relays? Those who are up on parallel and series wiring and relays will know the difference when it comes to detecting broken rails within a turnout.

 

2) Are points detectors used on the switches down there?

 

3) Are scrubber pads used on the car wheels to help clean non-conductive debris, i.e., sand used during a winter storm or tree leaves, etc.?

 

Thanks for the info. Interesting topic!