 |
Re: Robust Communications Options for PTC-Why only politicians???
Author: Dr Zarkoff
Date: 09-26-2009 - 12:48
A few comments:
>And, the needs issues I've seen points toward a system that is layered, and highly integrated with additional enhancement of the existing baseline systems.
One thing which bothers me is the phenomena of "over-layering" and the trance-like enchantment with micro-processors (the "Buck Rogers Syndrome"). For example, Knorr came out with its EAB (electronic air brake) about 18 years ago. It "emulates the 26-C system" (quote from the operating manual), except that it doesn't do a very good job of it. Pressure regulation of the brake pipe is twice as sloppy as with the 24 system of the early 1930s, not to mention the 26-C system of the 1950s. This leads to such phenomena as indulging in feed valve braking all on its own. Without going into the details of how this happens, it's been implicated in two coal train derailments: Soldier Summit on the former D&RG about 15-18 years ago and on the CSX at Bloomington WVA about 10 years ago. The symptoms on the event recorders of both are many small (6 psi or less) brake valve reductions with no decrease in train speed. In Soldier Summit, the entire train crew was killed, in Bloomington, the train was going "50 in a 25 zone", derailed, jackknifed, knocked down a house, and crushed a teenager.
Another example is the complexity of BARTD. Even before it started operating, one of the computer whizzes ran a study which predicted that every 20 minutes an event would happen which would bring the whole system to a halt (i.e. the trains would stop). For this he was fired--by Westinghouse, the operating computer and system supplier. It turns out he was wrong; something failed every 10 minutes.
BARTD also had block occupancy detection problems just like WAMTA. It was finally "solved" by installing something called "SORS", meaning Sequential Occupation Release. This meant that a block wouldn't be reported clear until the train went on to the next block. Oddly enough, the SFOBB (Bridge Railway) used something similar to turn off the transmissions of track code into a block once a train had gone onto the next block. Trap circuits on diamonds have used this principle for how long? 90 years?. BARTD's problem was that it tried to get away from impedance bonds and the "old style", low frequency track circuits, using instead AFOs and shunts.
>Several of the posts predict the perils of having politicians involved in the selection process of a PTC system. But aren't the nation's railroads just as susceptible to the panderings of the various manufacturers ???
Not really. The privately-owned rrs have always been extremely conservative when it comes to trying "new" things. For example, non-US made rails and joint bars are still no-nos, and I believe the same may be still true for roller bearings.
>Couldn't XYZ Co. promise the ABC RR a system " with all the bells & whistles " that's cheaper than the competition ???
This would be extremely difficult, to say the least, because of the interchange of locomotives between rrs. It's almost like boxcars these days. The rrs, through the AAR, would require interoperability with systems of other manufacturers. This is a principle which dates to the Burlington Air Brake Tests of 1886 and which has been reinforced periodically ever since. A car with automatic brake system from the late 1880s will function just fine with one which has ABDX, neglecting the length of the train.
I was once told by a signal wacko at a railroad museum that the PRR developed cab signals, and then gave the specifications to US&S and GRS. While I had no way to accredit this particular assertion, just about everything else he expounded on turned out to be correct. A locomotive with US&S cab signal equipment will work just fine picking up codes sent by GRS wayside equipment (and vice versa). However, certain components, particularly with the equipment on the locomotive, aren't interchangeable (for example, to detect the codes, GRS uses an amplifilter while US&S uses a primary relay).
>Supposing XYZ convinces several of the Class 1 's to sign on; wouldn't that system ( for better or worse ) then become the de facto standard ???
It could, and with many things, this is the way it happens. However, with today's liability concerns, most of the new stuff is tested at Pueblo, by the FRA (the Government) which insulates the rrs from these concerns. If the system didn't succeed testing at Pueblo, then no sale (except perhaps for transit agencies).
>Will Class 1's have any influence over the Feds in the selection process ???
You betcha.
>How about propriatary & copyright issues ??? Will a manufacturer bite the bullet & share their technology with the industry, for the mutual benefit of all ???
Over the last 80-90 years, WABCO used to bring out new brake systems and then give the prints to NYAB for the sole purpose of avoiding anti-trust litigation. However, WABCO's and NYAB's current control valves for freight are quite different.
>And, the needs issues I've seen points toward a system that is layered, and highly integrated with additional enhancement of the existing baseline systems.
One thing which bothers me is the phenomena of "over-layering" and the trance-like enchantment with micro-processors (the "Buck Rogers Syndrome"). For example, Knorr came out with its EAB (electronic air brake) about 18 years ago. It "emulates the 26-C system" (quote from the operating manual), except that it doesn't do a very good job of it. Pressure regulation of the brake pipe is twice as sloppy as with the 24 system of the early 1930s, not to mention the 26-C system of the 1950s. This leads to such phenomena as indulging in feed valve braking all on its own. Without going into the details of how this happens, it's been implicated in two coal train derailments: Soldier Summit on the former D&RG about 15-18 years ago and on the CSX at Bloomington WVA about 10 years ago. The symptoms on the event recorders of both are many small (6 psi or less) brake valve reductions with no decrease in train speed. In Soldier Summit, the entire train crew was killed, in Bloomington, the train was going "50 in a 25 zone", derailed, jackknifed, knocked down a house, and crushed a teenager.
Another example is the complexity of BARTD. Even before it started operating, one of the computer whizzes ran a study which predicted that every 20 minutes an event would happen which would bring the whole system to a halt (i.e. the trains would stop). For this he was fired--by Westinghouse, the operating computer and system supplier. It turns out he was wrong; something failed every 10 minutes.
BARTD also had block occupancy detection problems just like WAMTA. It was finally "solved" by installing something called "SORS", meaning Sequential Occupation Release. This meant that a block wouldn't be reported clear until the train went on to the next block. Oddly enough, the SFOBB (Bridge Railway) used something similar to turn off the transmissions of track code into a block once a train had gone onto the next block. Trap circuits on diamonds have used this principle for how long? 90 years?. BARTD's problem was that it tried to get away from impedance bonds and the "old style", low frequency track circuits, using instead AFOs and shunts.
>Several of the posts predict the perils of having politicians involved in the selection process of a PTC system. But aren't the nation's railroads just as susceptible to the panderings of the various manufacturers ???
Not really. The privately-owned rrs have always been extremely conservative when it comes to trying "new" things. For example, non-US made rails and joint bars are still no-nos, and I believe the same may be still true for roller bearings.
>Couldn't XYZ Co. promise the ABC RR a system " with all the bells & whistles " that's cheaper than the competition ???
This would be extremely difficult, to say the least, because of the interchange of locomotives between rrs. It's almost like boxcars these days. The rrs, through the AAR, would require interoperability with systems of other manufacturers. This is a principle which dates to the Burlington Air Brake Tests of 1886 and which has been reinforced periodically ever since. A car with automatic brake system from the late 1880s will function just fine with one which has ABDX, neglecting the length of the train.
I was once told by a signal wacko at a railroad museum that the PRR developed cab signals, and then gave the specifications to US&S and GRS. While I had no way to accredit this particular assertion, just about everything else he expounded on turned out to be correct. A locomotive with US&S cab signal equipment will work just fine picking up codes sent by GRS wayside equipment (and vice versa). However, certain components, particularly with the equipment on the locomotive, aren't interchangeable (for example, to detect the codes, GRS uses an amplifilter while US&S uses a primary relay).
>Supposing XYZ convinces several of the Class 1 's to sign on; wouldn't that system ( for better or worse ) then become the de facto standard ???
It could, and with many things, this is the way it happens. However, with today's liability concerns, most of the new stuff is tested at Pueblo, by the FRA (the Government) which insulates the rrs from these concerns. If the system didn't succeed testing at Pueblo, then no sale (except perhaps for transit agencies).
>Will Class 1's have any influence over the Feds in the selection process ???
You betcha.
>How about propriatary & copyright issues ??? Will a manufacturer bite the bullet & share their technology with the industry, for the mutual benefit of all ???
Over the last 80-90 years, WABCO used to bring out new brake systems and then give the prints to NYAB for the sole purpose of avoiding anti-trust litigation. However, WABCO's and NYAB's current control valves for freight are quite different.
