ATC

shop deals on amazon
Advertise with us
Shop deals on ebay
Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Status
Not open for further replies.
Will PTC be able to stop a train BEFORE it passes a stop signal? I'm thinking of a situation like the wreck of the Colonial where it was lined up with clear signals and moving at normal speed when another train ran through a stop signal (at relatively low speed) and into the path of the Colonial, which hit it.

jb
 
John Bobinyec said:
Will PTC be able to stop a train BEFORE it passes a stop signal? I'm thinking of a situation like the wreck of the Colonial where it was lined up with clear signals and moving at normal speed when another train ran through a stop signal (at relatively low speed) and into the path of the Colonial, which hit it.

jb
Click to expand...
Yes. That is one of the distinguishing features of PTC which makes it complicated. It provides the engine computer with a distance to target and a braking curve to enforce, depending on the characteristics of the train that is involved. The information about the train has to be entered before the start of a run, and updated if things change on the way.
 
jis said:
John Bobinyec said:
Will PTC be able to stop a train BEFORE it passes a stop signal? I'm thinking of a situation like the wreck of the Colonial where it was lined up with clear signals and moving at normal speed when another train ran through a stop signal (at relatively low speed) and into the path of the Colonial, which hit it.

jb
Click to expand...
Yes. That is one of the distinguishing features of PTC which makes it complicated. It provides the engine computer with a distance to target and a braking curve to enforce, depending on the characteristics of the train that is involved. The information about the train has to be entered before the start of a run, and updated if things change on the way.
Click to expand...
That's a point of failure right there. I can see someone entering into the system that the train has 1000 tons rather than 10000 tons. The braking curves will be different.

jb
 
Scenario # 2

Here's the heavy freight at the top of the mountain, ready to descend. The engineer knows that the train can't be going any more than 10 mph when it crests the hill because if it does, it'll be a runaway at the bottom of the mountain. Signals are clear all the way. Normal (maximum timetable speed) is 79 mph.

So what will PTC do? If the train crests the hill at 20 mph, will it attempt to stop it there?

jb
 
I have no idea if anything like this is implemented, but it seems like it should be theoretically possible for a train to "weigh itself" based on how the locomotive is accelerating. If an onboard computer system knows the performance characteristics of a given locomotive or set of locomotives and assumes level ground (or calculates elevation change from watching GPS over time), it could measure the time it takes to arrive at a given speed after a throttle change and get a reasonable estimate for the amount of load attached. It could even take into account barometric pressure and windspeed to calculate air resistance.

That would never be exact without having a way to know the rail friction and the maintenance level of every axle of the train (how smoothly they're turning) and possibly other factors, but it would be close enough to do some sanity checking on that "1000 tons vs 10,000 tons" scenario.
 
John Bobinyec said:
That's a point of failure right there. I can see someone entering into the system that the train has 1000 tons rather than 10000 tons. The braking curves will be different.

jb
Click to expand...
Yup. It is shared with the airline industry. In rail it is braking and accelerating curves, in airline it is the ability to take off, fly and land safely.

For passenger trains this is relatively minor issue. AFAICT they just use weights based on equipment type and number of cars. It is a different ball of wax for freight. But then again the Engineer has to have some idea of the operating characteristic of the train too in order to operate it, and that would involve knowing the power, weight and length I presume. So I guess that the necessary information is not new. It just has to be entered into the system, and apparently this can be done either centrally by whoever creates the information based on lading and marshaling order etc. or may be entered by the Engineer. However, the Engineer is very unlikely to have this info all my him/herself. It has to come from whoever put the train together.

In this the procedure seems to be surprisingly similar to how the airline industry operates their flights with more stringent reporting requirements for weight, fuel and balance, with much greater chance of bad outcomes if done wrong, and yet they seem to be able to operate more safely than the railroads already. Which would in some sense be a matter of concern if this is indicative of a more carefree safety culture among railroaders. I don't know. I hope that is not the case.
 
Last edited by a moderator:
So exactly what was implemented over last weekend at the Frankford Junction curve, real PTC, some kind of ATC or . . .?

There have been some fairly knowledgeable sounding interviews on MSNBC regarding safety systems. One with Secy. Foxx and one with Sen. Thune of the Senate Transportation Committee, MSNBC suggesting it was “ATC” that was implemented so fast in that location. So where else can whatever was implemented in Philly in just two days be implemented next weekend? Who has that list, LOL?

MSNBC suggested to the Senator and Secretary that some system, "ATC" or otherwise, could have been operable, at low cost, a long time ago at Frankford Junction and the 2013 Metro North accident site. That the FRA has a history of responding only to actual accidents not to a potential for accidents. Both the Senator and Secretary deflected any real answers to specific questions. Secy. Foxx stated he “is not at liberty to discuss or disclose” plans for Train and Passenger safety?
 
Frankford Junction and indeed all of NEC has had an ATC systems consisting of Continuous Coded Track Circuit based Cab Signaling System that provides a warning upon detection of failure to comply with signal speed in force in the block, and then applying brakes to bring the train to a stop if the Engineer does not take remedial action within a short time limit. The basics of this system was installed by PRR many moons ago and is considered to be one of the pioneering such installations. It has been enhanced recently to provide for more signal aspects than were available before. Signal aspect "Clear" has no signal speed limit. Signal aspect "Approach Medium" has signal speed limit of 45mph.

The setup approaching Frankford Curve where the derailment occurred is as follows. The curve is less than 0.2 miles from the Shore Interlocking where the Atlantic City trains leave the NEC to head towards Delair Bridge. At Shore Interlocking there is its Home Signal which govern entry into the block that includes the curve in question. About 2 miles to the west of it is the Advanced Signal for the Shore Home Signal.

Prior to the rearrangement of things following the accident, a train heading East which had a clear track through Shore would have seen a Clear indication on both the Advanced Signal and the Home Signals, and it was upto the Engineer to abide by the 80mph speed limit upto Shore and then the 50/60mph limit through the curve. Typically trains would accelerate after passing North Philadelphia station and stick to 80mph or thereabouts until applying breaks just in time to get down to 50/60 to enter the curve. This meant that they were pretty much up at 80 almost all the way to Shore interlocking, and applied breaks just short of Shore to hit the right speed for the curve.

The change that has been put in effect is that the Advanced Signal cannot display "Clear" anymore. When the track is clear through Shore and beyond it displays "Approach Medium" which corresponds to 45mph speed. This enforces a 45mph speed over the two mile block between the Advanced Signal and the Home signal. The Home Signal shows "Clear" releasing the train from the 45mph speed limit. But the train is already slow enough to take the curve safely. The Engineer can accelerate a bit to hit the curve speed limit. The downside of this is that all trains must now run at 45mph between the Advanced and the Home signal, where in the past they used to run close to 80mph.

This is my understanding of what has been done. Those more knowledgeable like Thirdrail may please jump in with any corrections or omissions if needed.

Finally, when ACSES providing full PTC functionality in conjunction with the current ATC system) comes into force, since it is able to enforce speed limits and even enforce braking curves with distance to target information that is available to it from the track mounted transponders, things can be setup so that trains can again run at 80mph almost to Shore and then slow down to 50/60mph for the curve.
 
Last edited by a moderator:
jis said:
Frankford Junction and indeed all of NEC has had an ATC systems consisting of Continuous Coded Track Circuit based Cab Signaling System that provides a warning upon detection of failure to comply with signal speed in force in the block, and then applying brakes to bring the train to a stop if the Engineer does not take remedial action within a short time limit. The basics of this system was installed by PRR many moons ago and is considered to be one of the pioneering such installations. It has been enhanced recently to provide for more signal aspects than were available before. Signal aspect "Clear" has no signal speed limit. Signal aspect "Approach Medium" has signal speed limit of 45mph.

The setup approaching Frankford Curve where the derailment occurred is as follows. The curve is less than 0.2 miles from the Shore Interlocking where the Atlantic City trains leave the NEC to head towards Delair Bridge. At Shore Interlocking there is its Home Signal which govern entry into the block that includes the curve in question. About 2 miles to the west of it is the Advanced Signal for the Shore Home Signal.

Prior to the rearrangement of things following the accident, a train heading west which had a clear track through Shore would have seen a Clear indication on both the Advanced Signal and the Home Signals, and it was upto the Engineer to abide by the 80mph speed limit upto Shore and then the 50/60mph limit through the curve. Typically trains would accelerate after passing North Philadelphia station and stick to 80mph or thereabouts until applying breaks just in time to get down to 50/60 to enter the curve. This meant that they were pretty much up at 80 almost all the way to Shore interlocking, and applied breaks just short of Shore to hit the right speed for the curve.

The change that has been put in effect is that the Advanced Signal cannot display "Clear" anymore. When the track is clear through Shore and beyond it displays "Approach Medium" which corresponds to 45mph speed. This enforces a 45mph speed over the two mile block between the Advanced Signal and the Home signal. The Home Signal shows "Clear" releasing the train from the 45mph speed limit. But the train is already slow enough to take the curve safely. The Engineer can accelerate a bit to hit the curve speed limit. The downside of this is that all trains must now run at 45mph between the Advanced and the Home signal, where in the past they used to run close to 80mph.

This is my understanding of what has been done. Those more knowledgeable like Thirdrail may please jump in with any corrections or omissions if needed.

Finally, when ACSES providing full PTC functionality in conjunction with the current ATC system) comes into force, since it is able to enforce speed limits and even enforce braking curves with distance to target information that is available to it from the track mounted transponders, things can be setup so that trains can again run at 80mph almost to Shore and then slow down to 50/60mph for the curve.
Click to expand...
I think you mean east.

There is also a 65mph curve at MP84 (Second Street), one mile east of North Philadelphia. A train passing through North Philadelphia would go to 80mph (or as close to 80 as possible), slow to 65 for the MP84 curve, accelerate back to 80 for nearly two miles, then slow to 50 (60 for Acela on Tracks 2 or 3) for the Frankford Junction curves.
 
Last edited by a moderator:
jis said:
You're correct. Thanks for catching. Corrected in the original.

Is the explanation clear enough for people to understand? If not, how can it be improved? Any suggestions? Thanks.
Click to expand...
I think it it pretty clear. Basically, ATC is an inelegant solution to enforcing speed through curves (called "civil" enforcement). The speed is controlled through an entire signal block (the length of track between signals). If you are lucky, the signal block and the area requiring speed control align well. If you are not lucky, the block is much longer than the critical area and the train has to creep along (relatively speaking) for a long distance to ensure the speed when it finally gets to the curve is appropriate. That is what is going to happen here, with all trains (including Acela) having to operate at 45mph for two miles prior to even getting to the 50mph curve.
 
Exactly!

And the basic problem there is that the ATC system in place is rather obsolete, based on enforcing current state of the block, which is the best that could be done when it was put in place. More modern ATC systems use distance to target (next signal) and the state of the next signal to decide when to start applying brakes to meet the rule for the next signal (or in case of civil speed enforcement, next enforcement point), like what Engineer do in handling Approach Medium or Limited Clear or such in a non ATC protected railroad in the US.
 
Last edited by a moderator:
Two questions for you Jis. First, the Acelas [used to] take Frankford at 60mph correct?

If that's the case, then why can't the FRA be happy with Cab-60 from the distant signal? I believe it was explained (by you maybe?) that active tilt systems do nothing for safety and are solely for passenger comfort. If that's the case, then isn't Frankford safe at 60mph even if the passengers will be a little unconfortable?
 
AFAIK it is FRA that decreed what was done. It is possible that they wanted to use a code that is understood by equipment that does not know the new codes. Just a random guess. I don't know for sure.

For the Elizabeth curve with a higher speed limit they still went with 45.
 
jis said:
The change that has been put in effect is that the Advanced Signal cannot display "Clear" anymore. When the track is clear through Shore and beyond it displays "Approach Medium" which corresponds to 45mph speed. This enforces a 45mph speed over the two mile block between the Advanced Signal and the Home signal. The Home Signal shows "Clear" releasing the train from the 45mph speed limit. But the train is already slow enough to take the curve safely. The Engineer can accelerate a bit to hit the curve speed limit. The downside of this is that all trains must now run at 45mph between the Advanced and the Home signal, where in the past they used to run close to 80mph.

This is my understanding of what has been done. Those more knowledgeable like Thirdrail may please jump in with any corrections or omissions if needed.
Click to expand...

The only thing I would change is technically, an Approach Medium cab signal will allow the engine to operate up to 45mph, but the indication of Approach Medium calls for you to proceed approaching the next signal at medium speed, which is not exceeding 30mph.

As you get closer to Shore, you could use rule 243 (next governing signal) as long as the home signal did indeed display a clear, but if your cabs signal are cut in, the speed control portion of the system will indeed limit the engine to 45mph. Remember, speed control is a big part of train control.

MattW said:
Two questions for you Jis. First, the Acelas [used to] take Frankford at 60mph correct?

If that's the case, then why can't the FRA be happy with Cab-60 from the distant signal? I believe it was explained (by you maybe?) that active tilt systems do nothing for safety and are solely for passenger comfort. If that's the case, then isn't Frankford safe at 60mph even if the passengers will be a little unconfortable?
Click to expand...

jis said:
AFAIK it is FRA that decreed what was done. It is possible that they wanted to use a code that is understood by equipment that does not know the new codes. Just a random guess. I don't know for sure.

For the Elizabeth curve with a higher speed limit they still went with 45.
Click to expand...
Most equipment understands the cab speed 80/cab speed 60 codes. From what I've been told and seen (I've never been on the new silverliners) , Septa's equipment does not, but the cab signal will default to approach medium. Putting the code in for approach medium was the quickest option since it already existed since that is the crossover speed for Shore. It take time to put in another code point change, and additionally, the 834 automatic probably can not handle Cab Speed Signals. You'd have to change the entire signal bridge since neither arm is equipped to flash.
 
Status
Not open for further replies.

Latest posts

Back
Top