Trains.com

A Contrarian View of High Speed Rail

19203 views
156 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    July 2004
  • 2,741 posts
Posted by Paul Milenkovic on Tuesday, June 9, 2009 12:44 PM

I don't take personal offense at your questioning me.  I can't for the life of me think what your point is, other than to advance a vision, assert the technical capabilities exist to sustain that vision, and then push back against any information contrary to that assertion.  I have no interest in winning arguments with you, nor do I have any interest in either denying or supporting your vision.   I'm not worried if anyone else reading this forum is swayed or not swayed, nor worried if someone here will start believing the sky is orange and the moon made of cheese unless I type faster and with more use of bold and italics.  If you want to scoff at my knowledge, mox nix to me.

You are NEW around here, aren't you?  Advance a vision, assert the technical capabilities, and then push back against information contrary?  I never saw it expressed so succinctly, but that pretty much defines the passenger train advocacy community let alone the comments section of pretty much any Web page these days.

oltmannd to my understanding actually works for a Class 1 railroad.  Sam1 is an accounting/finance professional who has seen electric power utility deregulation from the inside.  My excuse is that I am a research engineer trying to follow in my father's footsteps, a man who designed the constant-velocity power transmission for the roller test stand at Pueblo, Colorado and who had a stack of notes from his ASME conference abstract and his work at GATX on bogie design that I rescued from the ravages of the red squirrels attracted by his walnut plantings.  I am trying to get up the learning curve of Lagrange-Hamilton and Lyapanov theory to get that work to see the light of day.  Others will have to offer up their excuse. 

Thanks.  It will be interesting.  I hope the "always" and "never" folk don't kill it for the "lets try this" folk.

Don, you lost me there.  I don't know the players without a program and don't know who the "lets try this" folk are.

As to the traffic restricting of PTC, I have to agree with RailwayMan based on what is written in Trains of all places.  The trouble with PTC is that it has to assume the lowest common denominator in braking performance in setting the protection zones.

The narrative for a long time has been "110 MPH Midwest Regional Rail Initiative" to do a kind of near HSR "on the cheap."  I don't think anyone is seriously thinking going faster in the share-the-tracks mode, but RailwayMan is telling us the 90 MPH is stretching it unless there are serious restrictions on freight operations (like not having any).  I don't know where the 110 MPH comes from and whether it was pulled from the air, but it has become a kind of shibboleth in the advocacy community.

All of this is not simply idle jibber-jabber as according to Trains (of course!) the 8 billion in the Stimulus Package is not going to be a downpayment on a single dedicated HSR line but looks to be spread around among as many as 10 different places, pursuing various kinds of improvements to conventional rail shared with freight, and I am sure the 110 MPH thing is going to be a big issue.

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?

  • Member since
    December 2001
  • 8,156 posts
Posted by henry6 on Tuesday, June 9, 2009 1:19 PM

My excuse is that I have been around trains, riding, observing, even organizing and operating special train trips, and special trips (an excercise in utilizing schedules of different systems...rail, light rail, subway, and bus; and different agencies, but  learned a lot about operating practices and equipment usasge), and museum/tourist operations for over 60 years.  I remember the first time I knew I doing 100 mph when returning to NYP from a fan trip to Montauk, LI, NY and those aboard the LIRR MU west of Jamaica stop watched the mile posts annonced a speed a bit over it.  I also heard stories of "hoggers' getting the most out of thier train at the century mark, side by side races of DL&W and Erie west of Elmira, and DL&W passenger trains making up time on the Cut Off at more than 100 more than once.  It happened a lot on all roads.  But I try not to dwell on that, that is nostalgia.  The reality of railroading has changed from the mettle of its operating men to the metals...and wires, and micro proccessors and computers and sattelites and legal limits imposed by insurance companies and lawyers and government agencies.  I like to see new technologies applied to solving the problems and operations of railroads.  Just because it used to be done does not mean it is the way it has to be done, or that it was economical or safe at all to begin with.  I have dealt with railroaders from the lavatory attendents to the president and everyone in between.  I feel I have some experience in railroading in that little respect.  A professional, by no means!  But just enough working with and inside, and learning from experiencing the old and new, coupled with access to the broadest range of employees and practitioners the business had to offer. There is so much to be learned on these pages by those who actually plan railroads and services, operate the trains, operate the business end, et al., that makes these discssions fun and enlightenting.   I know I often throw something in which is above my head or based on what somebody else said or told me. But if I am wrong, I am glad to get the  answers and good conversation. 

RIDEWITHMEHENRY is the name for our almost monthly day of riding trains and transit in either the NYCity or Philadelphia areas including all commuter lines, Amtrak, subways, light rail and trolleys, bus and ferries when warranted. No fees, just let us know you want to join the ride and pay your fares. Ask to be on our email list or find us on FB as RIDEWITHMEHENRY (all caps) to get descriptions of each outing.

  • Member since
    October 2006
  • 1,123 posts
Posted by HarveyK400 on Tuesday, June 9, 2009 1:34 PM

The Amtrak experience was four years ago, and tracks may have been Class VI - certainly rode well if only subjectively.  ATS was FRA-compliant.  Again, this probably was only occasional and not an everyday event.  No exceptional speed was experienced on the return trip.

Being currently involved, you have an insight few of us have access to.  Your wording suggested carrier postures with regard to PTC, which under the circumstances are at least suspect.  I have questions in part because of previous positive postings regarding PTC.  Some of us I think would appreciate an elaboration.

You may be very familiar with current track maintenance equipment; but it is still incredulous to me that manufactures would produce to such individual specifications for track profile measurements.  Most mainline equipment seems to use laser sighting systems; but maybe I'm wrong.

The other thing that makes absolutely no sense is that if track is brought just into compliance to allow trains to operate at a given speed, the least failure or degradation will result in a slow order and need for maintenance relatively quickly.  There is a distinction to be made between maintaining track to a given standard and the work standards for maintenance that would be much finer.  This is why I wonder that machines intended for mainlines can be used, or desired, for only up to Class IV or Class V track.

  • Member since
    November 2007
  • 2,989 posts
Posted by Railway Man on Tuesday, June 9, 2009 2:02 PM

Paul Milenkovic

As to the traffic restricting of PTC, I have to agree with RailwayMan based on what is written in Trains of all places.  The trouble with PTC is that it has to assume the lowest common denominator in braking performance in setting the protection zones.

That's one example of the fundamental change that PTC represents in train control.  Before, reliance was placed upon the good judgement of the train crew to operate in a safe condition.  Because of fatalities caused by single-point failures attributable to humans who did not exercise good judgement, PTC will remove many of those pathways that led to single-point failures.  That sounds excellent in concept, but the problem is that it's not technically or economically feasible to make PTC sufficiently fine-grained to accurately replicate the good judgement of a human throughout the vast range of scenarios that a train confronts in everyday operation in a highly variable world.  Instead, it's only technically feasible to replicate the judgement of an extremely conservative human.  No one wants to design a safety system that doesn't create safety, so when in doubt, everyone will choose the worst-case scenario and design for that.

This creates a significant delta in train performance outcomes.  For example, it's possible to operate a train right at the speed limit and apply brakes at the last possible moment to stop short of the signal, or, it's possible to apply the brakes early just in case the rail might be a bit slippery or the brake shoes a little less effective than ideal.  It's not technically feasible to put the same amount of instanteous experience into the algorithms of a PTC system that a skilled engineer can do after he sees how his train handles and with his detailed knowledge of idiosyncracies of the territory.  Maybe people smarter than I will figure all this out in the next few months -- there is quite a bit of work at TTCI at present into braking algorithms -- but I'm not yet cognizant of any good-news stories.

In other words, we're building a system to eliminate the last 0.00001% of failures, which is an approximatation of the percentage of all train-handling events that could cause a fatality that a human being fails and does cause a fatality.  That's the decision, and we're all on board with it.  I'm faintly amused by the people who are accusing us of dragging our heels, using accusations of the category  "Have you stopped beating your wife yet?" 

RWM


  • Member since
    October 2006
  • 1,123 posts
Posted by HarveyK400 on Tuesday, June 9, 2009 3:22 PM

Railway Man

Paul Milenkovic

As to the traffic restricting of PTC, I have to agree with RailwayMan based on what is written in Trains of all places.  The trouble with PTC is that it has to assume the lowest common denominator in braking performance in setting the protection zones.

That's one example of the fundamental change that PTC represents in train control.  Before, reliance was placed upon the good judgement of the train crew to operate in a safe condition.  Because of fatalities caused by single-point failures attributable to humans who did not exercise good judgement, PTC will remove many of those pathways that led to single-point failures.  That sounds excellent in concept, but the problem is that it's not technically or economically feasible to make PTC sufficiently fine-grained to accurately replicate the good judgement of a human throughout the vast range of scenarios that a train confronts in everyday operation in a highly variable world.  Instead, it's only technically feasible to replicate the judgement of an extremely conservative human.  No one wants to design a safety system that doesn't create safety, so when in doubt, everyone will choose the worst-case scenario and design for that.

This creates a significant delta in train performance outcomes.  For example, it's possible to operate a train right at the speed limit and apply brakes at the last possible moment to stop short of the signal, or, it's possible to apply the brakes early just in case the rail might be a bit slippery or the brake shoes a little less effective than ideal.  It's not technically feasible to put the same amount of instanteous experience into the algorithms of a PTC system that a skilled engineer can do after he sees how his train handles and with his detailed knowledge of idiosyncracies of the territory.  Maybe people smarter than I will figure all this out in the next few months -- there is quite a bit of work at TTCI at present into braking algorithms -- but I'm not yet cognizant of any good-news stories.

In other words, we're building a system to eliminate the last 0.00001% of failures, which is an approximatation of the percentage of all train-handling events that could cause a fatality that a human being fails and does cause a fatality.  That's the decision, and we're all on board with it.  I'm faintly amused by the people who are accusing us of dragging our heels, using accusations of the category  "Have you stopped beating your wife yet?" 

RWM

My understanding was that PTC computed specific train handling from data on cars and lading and monitored performance such as for acceleration and brake response for calibration, taking into account speed, grades, curves, weather, distance to stop or speed restriction, and safety factor.  Weren't these and perhaps other factors used to create algorithms for performance? 

  • Member since
    December 2001
  • 8,156 posts
Posted by henry6 on Tuesday, June 9, 2009 4:01 PM

HarveyK400

My understanding was that PTC computed specific train handling from data on cars and lading and monitored performance such as for acceleration and brake response for calibration, taking into account speed, grades, curves, weather, distance to stop or speed restriction, and safety factor.  Weren't these and perhaps other factors used to create algorithms for performance? 

I get the feeling that the algorithms a seasoned engineer gets through the seat of his pants are finer than any computerized system can produce.  Do you drive a car? Stick or automatic?  Sedan, SUV, or sports car?  Thats the difference.

RIDEWITHMEHENRY is the name for our almost monthly day of riding trains and transit in either the NYCity or Philadelphia areas including all commuter lines, Amtrak, subways, light rail and trolleys, bus and ferries when warranted. No fees, just let us know you want to join the ride and pay your fares. Ask to be on our email list or find us on FB as RIDEWITHMEHENRY (all caps) to get descriptions of each outing.

  • Member since
    November 2007
  • 2,989 posts
Posted by Railway Man on Tuesday, June 9, 2009 4:03 PM

HarveyK400

My understanding was that PTC computed specific train handling from data on cars and lading and monitored performance such as for acceleration and brake response for calibration, taking into account speed, grades, curves, weather, distance to stop or speed restriction, and safety factor.  Weren't these and perhaps other factors used to create algorithms for performance? 

 

Inputs are:

  1. Horizontal and vertical profile (absolute value)
  2. Permanent and temporary speed limits (absolute value)
  3. Tonnage of train (absolute value, or at least what the trainsheet claims; there's considerable discussion on how this will be made more accurate -- I expect a whole lot more AEI readers to be installed)
  4. Braking horsepower of train (estimated value)
Excluded at this time are:
  1. Braking performance history -- no method yet available how to account for fade, ice buildup, localized rail head profile, localized rail head surface conditions
  2. Dynamic brake horsepower -- no method yet available to predict future effectiveness
  3. Weather -- no method yet available to predict local conditions

En route calibration is something of an enigma.  In concept it's a great idea; no one has figured out yet how to do it in a reliable, predictible, and FRA-certifiable fashion.

Algorithms remain under study at TTCI and at the Class 1s.

The FRA Product Safety Plan and Railroad Product Safety Plan requirements under Subpart H and I put the onus on to the vendor and the railway to certify this all works perfectly.

RWM

  • Member since
    October 2008
  • From: Near Burlington, WA
  • 380 posts
Posted by Maglev on Tuesday, June 9, 2009 10:38 PM

In the face of a crumbling airline industry, we need more government investment in railroads:

Airline event reflects industry slump

"Media questions fly at a somber meeting of international carrier executives who are focused on survival. The industry has been rocked by the recession, swine flu outbreak and the Air France crash."
By Peter Pae, Los Angeles Times
June 9, 2009
(excerpts)
"Reporting from Kuala Lumpur, Malaysia -- The telltale signs of an industry in crisis were unavoidable at the biggest airline trade group's yearly gathering..."
"One of the highlights of the event: a panel on whether aviation could ever be economically sustainable. (The consensus: Maybe, but only after further painful consolidation.)"

"'We've never seen anything like this before,' Jong Hee Lee, president of Korean Air Lines Co. and a 40-year veteran of the carrier, said of the travel slump that had hammered the industry. 'Everybody is saying, 'Let's survive.' '"

"The association's 226 member airlines, including the largest U.S. carriers -- Delta, American and United -- account for about 93% of the world's international traffic. No event draws as many airline chief executives, estimated at 150 at this year's meeting."

"In one of the conference's more depressing moments, the group's CEO announced Monday that the industry would lose $9 billion this year, more than double the forecast made just two months ago."

"Revenue is projected to tumble 15% to $448 billion, a drop more than twice as big as the falloff in sales after the Sept. 11, 2001, terrorist attacks."

"And airline executives say there's little relief in sight..."
 
full story at:
http://www.latimes.com/business/la-fi-airlines9-2009jun09,0,3942501.story

"Make no little plans; they have no magic to stir men's blood." Daniel Burnham

  • Member since
    September 2007
  • From: Charlotte, NC
  • 6,099 posts
Posted by Phoebe Vet on Wednesday, June 10, 2009 3:58 AM

I don't know about anyone else, but I have no intention of ever flying again as long as the airport resembles an East German border crossing in the '60s.

Dave

Lackawanna Route of the Phoebe Snow

  • Member since
    March 2004
  • From: Central Valley California
  • 2,841 posts
Posted by passengerfan on Wednesday, June 10, 2009 6:29 AM

Phoebe Vet

I don't know about anyone else, but I have no intention of ever flying again as long as the airport resembles an East German border crossing in the '60s.

You are fortunate I Due to unforseen cirumstances am a frequent but reluctant flyer. There are many times I don't have the time to take Amtrak and many places are not served by Amtrak. If Amtrak provided the service I would be on Amtrak. And lets be honest I can fly far cheaper first class than Amtrak first class. For instance I go from Sacramento to Calgary about 4 times a year and Calgary does not have rail service.  I fly and have a 1 hour change of planes in Seattle and am there in 31/2 hours. I can book my flights with less than two days notice. And yes I tried the train from Sacramento once on this trip taking Via from Vancouver to Edmonton and grabbing a flight from there to Calgary. I must have looked a sight as I could only get coach on Via from Vancouver to Edmonton. The train trip from Stockton to Vancouver on Amtrak was great but it takes two nights. One of those nights in Seattle is in a Hotel.

Al - in - Stockton  

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, June 10, 2009 7:35 AM

Paul Milenkovic

Don, you lost me there.  I don't know the players without a program and don't know who the "lets try this" folk are.

The dogmatic "always" or "never' folk are those who grab a position and then hold on tight.  They include the frt RR that says "never" to any additional passenger service on their routes and advocates who start with the position that HSR must be 180 mph dedicated ROW,etc. etc, and then work backward to the arguements for.

The "lets try this" folk are those who start with a goal or problem statement and then try to find common ground or mutual benefit for the stakeholders.  A good example would be the NCRR program.  NC wanted faster and more frequent service in their state.  NS and NC worked together to improve the NCRR, replacing high maintenance equilaterals with #20s so that the passenger trains don't have to slow down when the switch is lined normal, adding some sections of double track and installing CTC and upgrading track on a more lightly used portion of the RR.  NC gets improved passenger service.  NS gets a more fluid (there's that word, again), easier to operate and maintain railroad.  Of course, in this case, NC had some leverage - they own the RR!

Or, let's say there is some route where 110 mph service is desired and it has several curves that need reallignment.  Rather than just saying "no" to the whole deal, the frt RR sees that the reallignment could benefit them since they could reduce superelevation and all the trouble of maintaining it.  So, they offer up some $$ comensurate with the benefit they'll get (plus a fair return) to help move things along.

Or, they look to see if there is some political quid pro quo for being flexible in their approach toward these things.

Paul Milenkovic
I don't know where the 110 MPH comes from and whether it was pulled from the air, but it has become a kind of shibboleth in the advocacy community

110 is as fast as you can run with conventional equipment on conventional track without having to jump through all kinds of hoops. The only place I know where this is currently being done outside the NEC is CP-169 (west of Schenectady) to Poughkeepsie NY where CSX (and Conrail before them) maintain Class VI track for NY and Amtrak (at their incremental expense). 

From the Conrail days....http://www.railpictures.net/viewphoto.php?id=8020

-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, June 10, 2009 7:44 AM

Railway Man

HarveyK400

My understanding was that PTC computed specific train handling from data on cars and lading and monitored performance such as for acceleration and brake response for calibration, taking into account speed, grades, curves, weather, distance to stop or speed restriction, and safety factor.  Weren't these and perhaps other factors used to create algorithms for performance? 

 

Inputs are:

  1. Horizontal and vertical profile (absolute value)
  2. Permanent and temporary speed limits (absolute value)
  3. Tonnage of train (absolute value, or at least what the trainsheet claims; there's considerable discussion on how this will be made more accurate -- I expect a whole lot more AEI readers to be installed)
  4. Braking horsepower of train (estimated value)

Algorithms remain under study at TTCI and at the Class 1s.

RWM

This produces braking curves that are longer than those currently used to set existing signal spacing and speeds?

Isn't actual speed also an input?

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

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Wednesday, June 10, 2009 7:57 AM

Maglev

In the face of a crumbling airline industry, we need more government investment in railroads:

Airline event reflects industry slump

"Media questions fly at a somber meeting of international carrier executives who are focused on survival. The industry has been rocked by the recession, swine flu outbreak and the Air France crash."
By Peter Pae, Los Angeles Times
June 9, 2009
(excerpts)
"Reporting from Kuala Lumpur, Malaysia -- The telltale signs of an industry in crisis were unavoidable at the biggest airline trade group's yearly gathering..."
"One of the highlights of the event: a panel on whether aviation could ever be economically sustainable. (The consensus: Maybe, but only after further painful consolidation.)"

"'We've never seen anything like this before,' Jong Hee Lee, president of Korean Air Lines Co. and a 40-year veteran of the carrier, said of the travel slump that had hammered the industry. 'Everybody is saying, 'Let's survive.' '"

"The association's 226 member airlines, including the largest U.S. carriers -- Delta, American and United -- account for about 93% of the world's international traffic. No event draws as many airline chief executives, estimated at 150 at this year's meeting."

"In one of the conference's more depressing moments, the group's CEO announced Monday that the industry would lose $9 billion this year, more than double the forecast made just two months ago."

"Revenue is projected to tumble 15% to $448 billion, a drop more than twice as big as the falloff in sales after the Sept. 11, 2001, terrorist attacks."

"And airline executives say there's little relief in sight..."
 
full story at:
http://www.latimes.com/business/la-fi-airlines9-2009jun09,0,3942501.story

Volatility has been a characteristic of the airline business since the get go.  It will always be the case.  Airlines come and go, but to think that airplanes are going to disappear and people are going to flock to trains is unrealistic.

Investment in passenger rail makes sense for relatively high density, short corridors where expanding highways and airways is cost prohibitive.  Thus, a significant per cent of commercial passengers take the train from Philadelphia to New York.  The percentage of commercial passengers taking the train from Charlotte to New York is very low.

  • Member since
    December 2001
  • 8,156 posts
Posted by henry6 on Wednesday, June 10, 2009 8:19 AM
Confusion in this thread is growing.  Can we stop and seperate the two emerging arguements about HSR?  One is the mechanical/technical aspect...(Oltmannd, I think you are missing the point that after a certain level of speed technology and mechanics go in opposite directions for freight and pasenger. This becomes a single issue arguement.)  Second is the aspect of the market place: build it and they will come on one side and scientific research (market studies?) per route and routing coupled with other available means and pricing.  (When do I take a plane? a bus? a train? drive myself?)  Can highways and airways be expanded and at what price relative to use?  In what instances is rail the answer and what instances is rail not the answer?  High speed or otherwise for that matter.  Or can't the two arguements (as I call them) be seperated?

RIDEWITHMEHENRY is the name for our almost monthly day of riding trains and transit in either the NYCity or Philadelphia areas including all commuter lines, Amtrak, subways, light rail and trolleys, bus and ferries when warranted. No fees, just let us know you want to join the ride and pay your fares. Ask to be on our email list or find us on FB as RIDEWITHMEHENRY (all caps) to get descriptions of each outing.

  • Member since
    November 2007
  • 2,989 posts
Posted by Railway Man on Wednesday, June 10, 2009 8:42 AM

 

oltmannd

This produces braking curves that are longer than those currently used to set existing signal spacing and speeds?

Isn't actual speed also an input?

Initial and final speeds are inputs, yes.

The answer on signal spacing is actually that the braking curves are very similar with PTC.  Signal braking distances are derived from a similiar algorithm that uses vertical profile, an assumed maximum possible weight per car (e.g., every car is 158 TPOB), a conservative braking horsepower, and initial and final speeds.  However, signal spacing is only interested in whether a worst-case train that is given an aspect can adhere to that aspect, and leaves it up to the engineer to determine if he is a worst-case train or a best-case train.  Thus signal spacing is almost always much greater than what a train is going to actually do, because no train is ever as bad as the worst case.  Thus in most cases trains are crowding much closer than the signal spacing nominally dictates.  PTC, however, will undo that crowding. 

For example,using one of the rule books, an engineer sees a yellow aspect.  The rule says "Proceed prepared to stop before any part of train or engine passes next signal.  Freight trains exceeding 30 mph must immediately reduce speed to 30 mph."  The "stop at next signal part" is not discretionary, but where the braking from 30 mph to stop occurs and how hard of a braking effort that will be is discretionary, as is the "immediately reduce" part.  What does that mean?  Reduce to 30 the minute you see the signal?  Reduce to 30 after the head end reaches the signal?  How aggressive shall the braking be?   PTC will "see" that signal aspect and make sure the train is at 30 before the head-end reaches the signal, which in many cases means the braking will begin to 30 long before the engineer can visually see the signal, and will follow a worst-case braking scenario to arrive at that target speed.   In a non-PTC system, the engineer of a light train with good dynamic braking on a dry day may run right up to that signal at 45 and then drop the speed in a hurry as he passes it. 

Now think what happens at every permanent speed restriction and station stop.  I've spent enough time in the cab to know that one engineer can drop his train onto the entrance of the speed restriction or platform aggressively, and another engineer will dawdle up to it.  Or put another way, from the viewpoint of a train dispatcher, different engineers can get identical trains across the same territory with remarkably different running times, and both adhere to the rules. 

RWM

  • Member since
    October 2006
  • 1,123 posts
Posted by HarveyK400 on Wednesday, June 10, 2009 10:24 AM

Braking horsepower would seem to have a wide range of values depending on the ratio of loaded and empty cars.  In explanation to other readers, braking force was limited to about 60% of empty car weight to avoid locking the wheels and sliding; but this may have changed more recently time. 

Railroads seem to have gotten away from weighing cars, and the comment on the accuracy of the trainsheet complicates calculating train braking distance.  The railroads' care that the load limit is not exceeded; but seemingly have little interest other than for customer car supply to know how much the lading weighs or if it maxes car volume. 

One thing you didn't mention was the number of cars and the time it takes to set up brakes for the whole train.  It seems newer brakes are faster acting, so that is another variable to deal with.

I don't have anything on fade as a component of braking horsepower in what's left of my library, but this would seem to be a curve related to speed, brake force, and ambient temperature.  Brake force varies for different types of shoes and the individual adjustment of each car.  As a train, one might start with the assumption of an average adjusted for empirical deviations.

With ambient temperatures near or below freezing, I understand how ice buildup is problematic and may need manual input/override if suspected.  The engineer can't really see what's happening behind him, although a passing train might.  Another possibility may be to make running brake applications periodically and prior to the distance required for dynamic braking to bring down speed.  Never having run a train or read an account of this procedure, I have no idea if this is done.

In the same vein, do air brake valves freeze up often from moisture in the line condensing and icing up with falling temperatures?

Dynamic brake horsepower seems to be as readily available as locomotive horsepower and varies with speed, notwithstanding the small variations in actual unit performance on a given train.  Am I missing something?  Who cares what the future may bring?

Weather - temperatures and winds in the area of the train - might be obtained from the Weather Service.  It seems newer locomotives may measure ambient air temperature for more efficient operation.

Even an experienced engineer would have little more than a fuzzy ideal of how the train will handle, taking into account the same but unquantified variables, and act cautiously.  Both engineer and computer can get some idea of braking performance by how long it takes to accelerate the train against the actual tonnage and rolling resistance that allows some recalibration.  I understand this is a combined rolling resistance and drag algorithm and not the whole answer. 

It seems facetious for people to claim that a human can come in hot and the computer not for the very reason of assuring safety.  Sooner or later the hotshot engineer is going to overrun a block and maybe hurt someone.  I am not convinced that PTC would reduce capacity to any measurable degree, even with 50 trains a day. 

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, June 10, 2009 11:11 AM

Railway Man

Initial and final speeds are inputs, yes.

The answer on signal spacing is actually that the braking curves are very similar with PTC.  Signal braking distances are derived from a similiar algorithm that uses vertical profile, an assumed maximum possible weight per car (e.g., every car is 158 TPOB), a conservative braking horsepower, and initial and final speeds.  However, signal spacing is only interested in whether a worst-case train that is given an aspect can adhere to that aspect, and leaves it up to the engineer to determine if he is a worst-case train or a best-case train.  Thus signal spacing is almost always much greater than what a train is going to actually do, because no train is ever as bad as the worst case.  Thus in most cases trains are crowding much closer than the signal spacing nominally dictates.  PTC, however, will undo that crowding. 

For example,using one of the rule books, an engineer sees a yellow aspect.  The rule says "Proceed prepared to stop before any part of train or engine passes next signal.  Freight trains exceeding 30 mph must immediately reduce speed to 30 mph."  The "stop at next signal part" is not discretionary, but where the braking from 30 mph to stop occurs and how hard of a braking effort that will be is discretionary, as is the "immediately reduce" part.  What does that mean?  Reduce to 30 the minute you see the signal?  Reduce to 30 after the head end reaches the signal?  How aggressive shall the braking be?   PTC will "see" that signal aspect and make sure the train is at 30 before the head-end reaches the signal, which in many cases means the braking will begin to 30 long before the engineer can visually see the signal, and will follow a worst-case braking scenario to arrive at that target speed.   In a non-PTC system, the engineer of a light train with good dynamic braking on a dry day may run right up to that signal at 45 and then drop the speed in a hurry as he passes it. 

Now think what happens at every permanent speed restriction and station stop.  I've spent enough time in the cab to know that one engineer can drop his train onto the entrance of the speed restriction or platform aggressively, and another engineer will dawdle up to it.  Or put another way, from the viewpoint of a train dispatcher, different engineers can get identical trains across the same territory with remarkably different running times, and both adhere to the rules. 

Stupid me.  I knew this. I just forgot.  With PTC the engineer has to keep his train "under" the worst case speed/distance curve.  This is exactly how the LSL system works with cab signals, only it has no idea of where the train is on the RR.  It assumes worst case signal spacing on the territory and is very aggressive lots of places it doesn't need to be.

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

  • Member since
    November 2007
  • 2,989 posts
Posted by Railway Man on Wednesday, June 10, 2009 11:22 AM

oltmannd

Stupid me.  I knew this. I just forgot.  With PTC the engineer has to keep his train "under" the worst case speed/distance curve.  This is exactly how the LSL system works with cab signals, only it has no idea of where the train is on the RR.  It assumes worst case signal spacing on the territory and is very aggressive lots of places it doesn't need to be.

 

Like my brain-fade on 539 engines the other day.

There's lots of means by which PTC can be fine-tuned to reduce the capacity loss that is created by making every braking case the worst braking case, which is the safe and simple course.  But the cost of implementation and maintenance of these means, and the cost of the proof to the FRA that safety is not being compromised, become impractical in a hurry.  There's a lot of research money being spent now looking at ways to reduce the capacity impact, but there's not a lot of confidence that by 2015 when in theory PTC is all rolled out, that we won't have taken a 10-20% capacity hit.

By comparison, the other day we got into a question about hump yard capacity.  After talking around with a lot of people who combined had more than 2,000 years of experience designing, managing, and operating hump yards on somewhere around 20 different Class 1s and terminal roads, we concluded that the same hump-yard that had a capacity of 3,000 cars over the hump in 1960 today had maybe 2,000 cars over the hump a day, and the only significant change we could ascribe this to was safety practices. That's not saying, of course, that we begrudge the safety improvement, only that it's usually impossible to have a free lunch.

RWM

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Wednesday, June 10, 2009 11:51 AM

HarveyK400
Braking horsepower would seem to have a wide range of values depending on the ratio of loaded and empty cars.  In explanation to other readers, braking force was limited to about 60% of empty car weight to avoid locking the wheels and sliding; but this may have changed more recently time. 

Braking HP per wheel would be brake shoe force X coeff of friction (shoe to wheel) X train speed. Brake shoe force is a function of brake pipe pressure and geometry of braking mechanism (cylinder diameter, levers, etc.) 

 

HarveyK400
Railroads seem to have gotten away from weighing cars, and the comment on the accuracy of the trainsheet complicates calculating train braking distance.  The railroads' care that the load limit is not exceeded; but seemingly have little interest other than for customer car supply to know how much the lading weighs or if it maxes car volume. 

Lots of stuff gets weighed, but not always for billing purposes.  Many coal loaders weight what they're loading on the fly.  RRs have weigh in motion scales to verify weights in some locations.  WILDs weigh cars as a secondary function of measuring impacts.  Automated hump systems have weigh in motion scales on th hump grade.  The problem is if you are going to use car weight as an input, you have to have it almost before you turn a wheel.  Most of the time, you don't know it until you are part way through the trip.  And, you almost never have anything useful for intermodal boxes.

HarveyK400
One thing you didn't mention was the number of cars and the time it takes to set up brakes for the whole train.  It seems newer brakes are faster acting, so that is another variable to deal with.

Not much of a variable.  Airbrake signal propogation is and has been pretty close to the speed of sound for quite a while now. (something like 80-90% of it, I think)  Time for the brake to actually set up on the car isn't much of a variable, either.

HarveyK400

With ambient temperatures near or below freezing, I understand how ice buildup is problematic and may need manual input/override if suspected.  The engineer can't really see what's happening behind him, although a passing train might.  Another possibility may be to make running brake applications periodically and prior to the distance required for dynamic braking to bring down speed.  Never having run a train or read an account of this procedure, I have no idea if this is done.

In the same vein, do air brake valves freeze up often from moisture in the line condensing and icing up with falling temperatures?

The problem of ice in the trainline almost always gets you before you leave the initial terminal.  You can't get the application and release and get the front to rear differential (or air flow) back to what it wants to be.  I have heard that there are instances when a brake valve will stick from frost, but this tends to cause UDEs rather than a "no brake" condition.

HarveyK400
Dynamic brake horsepower seems to be as readily available as locomotive horsepower and varies with speed, notwithstanding the small variations in actual unit performance on a given train.  Am I missing something?  Who cares what the future may bring?

I don't know if we'll ever get to a point where we're allow to count on the DB for anything other than convenience.  It's certainly not a safety device!

HarveyK400
Weather - temperatures and winds in the area of the train - might be obtained from the Weather Service.  It seems newer locomotives may measure ambient air temperature for more efficient operation.

Weather is sometimes the cause of the variable you're interested in, namely the wheel/rail adhesion during braking.  You could put a tribometer on the locomotive, I suppose, and measure it in real time, but you'd still not have what you really need.  You need to know what the rail conditions are up ahead where the braking will be taking place.  Could a weather forecast give you enough "9s" for your reliability calculation?

HarveyK400
It seems facetious for people to claim that a human can come in hot and the computer not for the very reason of assuring safety.  Sooner or later the hotshot engineer is going to overrun a block and maybe hurt someone.  I am not convinced that PTC would reduce capacity to any measurable degree, even with 50 trains a day. 

If the PTC system has to assume "all loads", it could really slow things down.  This could make a big difference getting trains into passing sidings, etc. 

There are lots of other variables that create variablility (duh!) in operations.  That hot shot engr. that you allude to, might get over the road quite a bit faster than Mr. Milquetoast.  He might try to keep it right on track speed or a mph or two over as much as possible, brake hard into speed restictions and put it back in 8 as soon as possible.  Mr. Milquetoast might be content to cruise along letting speed vary quite a bit below track speed before he reacts.  He might just coast into speed restictions, and might take his time getting back into 8 once he thought he was clear and then some.

There are attempts to reduce this variablility in train operation and minimize fuel consumption at the same time by providing the engineer with dynamically generated "suggested" train handling and measuring compliance.  NS's LEADER program is one of these.

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

  • Member since
    November 2007
  • 2,989 posts
Posted by Railway Man on Wednesday, June 10, 2009 12:27 PM

Don said all this much better than I could have.

RWM

  • Member since
    October 2006
  • 1,123 posts
Posted by HarveyK400 on Wednesday, June 10, 2009 4:00 PM

I misspoke in trying to summarize: it is the brake shoe pressure braking ratio that was the product of a (nominal, general use) brake ratio of 60% light weight assuming 70-ton cars [Hay].

"Not much of a variable.  Airbrake signal propogation is and has been pretty close to the speed of sound for quite a while now. (something like 80-90% of it, I think)  Time for the brake to actually set up on the car isn't much of a variable, either."

I understood it took about a second per car, 1-1/2 min for a 90-car train, for brake pipe pressure reduction (signal propagation?) which is the reason for the interest in electronic brake control.  Did I get this wrong; or are there separate issues?

"The problem of ice in the trainline almost always gets you before you leave the initial terminal.  You can't get the application and release and get the front to rear differential (or air flow) back to what it wants to be.  I have heard that there are instances when a brake valve will stick from frost, but this tends to cause UDEs rather than a "no brake" condition."

This is understandable sitting in the yard overnight; but what about changes in the weather in route or climbing to higher, colder elevations?

"I don't know if we'll ever get to a point where we're allow to count on the DB for anything other than convenience.  It's certainly not a safety device!"

I suppose dynamic braking may not meet an industry definition of a safety device since it poops out at low speed.  Even so, it would seem to aid in reducing train speed so partial disabling of train air brakes, for instance from ice build-up, would not be a total catastrophe. 

Conversely, are dynamic brakes operationally unreliable?  Isn't this similar to a unit being shut down in modeling performance?

"Could a weather forecast give you enough "9s" for your reliability calculation?" 

That's a good question.  Isn't weather condition more for adjusting for what might be considered worst-case operating conditions?  It's not so much a forecast as monitoring real-time weather reporting sites and interpolating data for the railroad ahead of the train.

It just came to mind that the railroads, at least the BNSF, have their own weather department to warn of severe weather and flooding dangers, perhaps more.

"If the PTC system has to assume "all loads", it could really slow things down.  This could make a big difference getting trains into passing sidings, etc."

This seems to be a very negative outlook.  Why on earth would PTC have to assume "all-loads" when tonnage and car data already is available and accumulated for the train?

"There are attempts to reduce this variablility in train operation and minimize fuel consumption at the same time by providing the engineer with dynamically generated "suggested" train handling and measuring compliance.  NS's LEADER program is one of these."

I've seen some stuff on this which gives hope that PTC can be a lot more refined than seems to be accepted at present.
  • Member since
    October 2006
  • 1,123 posts
Posted by HarveyK400 on Wednesday, June 10, 2009 6:04 PM
"PTC will "see" that signal aspect and make sure the train is at 30 before the head-end reaches the signal, which in many cases means the braking will begin to 30 long before the engineer can visually see the signal, and will follow a worst-case braking scenario to arrive at that target speed." 

RWM

Even if PTC used virtually the same worst-case algorithm as in establishing signal block lengths and allowable speeds, why would it be necessary then to brake before reaching the approach block?  This is equivalent to imposing an advance approach block which is not deemed necessary for current, assumed less reliable or compliant, non-PTC operation.

For train handling purposes, PTC might allow one train to drift and reduce speed to save fuel or to time a rolling meet well before the approach block.  Keeping a train rolling may save time overall starting and recovering speed after a delay.

A long time ago I read some stuff on rail profile and the cant of the tie plate.  This seemed inconclusive with pros and cons; so I hadn't worried about it.  What's changed with respect to heavy-haul and high-speed now that was refered to in recent posts?  How does it work in mixed traffic for the NEC at 125 and 150; and could it work in the Midwest at 110?

  • Member since
    July 2006
  • 9,610 posts
Posted by schlimm on Wednesday, June 10, 2009 7:42 PM

All of this has been very interesting.  BUT...why do we really need to invent the wheel?  To do so seems to not take advantage of what others already have had to struggle with.   Can we not examine the HSR and sub-HSR practices in France, Germany, etc. to see what they are using, their experience, what works, what doesn't, costs, etc.?  Or are we too parochial and thin-skinned to admit that other countries may know more than we do about some topics.

C&NW, CA&E, MILW, CGW and IC fan

  • Member since
    December 2001
  • 8,156 posts
Posted by henry6 on Wednesday, June 10, 2009 8:21 PM

It is not a matter of reinventing the wheel, or others knowing more, but rather adapting a different wheel.  As has been stated, North American railroads chose a heaver standard than did European railroads (for instance).  Therefore thier trains are lighter which leads to other variables and therefore it is not a matter of just bringing a European model over here, put it on the tracks and let it go, Adapting to our rail system is much more comlex than that.  That, at least in general, is how an Alstom VP explained to me why we can't just bring 'em over and run 'em.

RIDEWITHMEHENRY is the name for our almost monthly day of riding trains and transit in either the NYCity or Philadelphia areas including all commuter lines, Amtrak, subways, light rail and trolleys, bus and ferries when warranted. No fees, just let us know you want to join the ride and pay your fares. Ask to be on our email list or find us on FB as RIDEWITHMEHENRY (all caps) to get descriptions of each outing.

  • Member since
    July 2006
  • 9,610 posts
Posted by schlimm on Wednesday, June 10, 2009 11:32 PM
If you have dedicated, separate track for most passenger runs except near urban stations, and it sounds like most agree that is a given over 90 mph, it would seem that wheel profile/standards would not matter.  Most barriers to progress exist in the minds of the status quo folks.  Where have we gotten in 40 years?  Let's try something fresh or all we'll get is more half-baked NEC's.

C&NW, CA&E, MILW, CGW and IC fan

  • Member since
    January 2001
  • From: Atlanta
  • 11,971 posts
Posted by oltmannd on Thursday, June 11, 2009 7:36 AM

HarveyK400
This seems to be a very negative outlook.  Why on earth would PTC have to assume "all-loads" when tonnage and car data already is available and accumulated for the train?

The problem is the association of the waybill to the event reporting that tells you this car is on this train.  AEI scanners are pretty good at giving an accurate consist, but how do you know that that car in the consist that shows as empty is really empty?

The waybilling system knows about origins and destinations and what's in the car but it doesn't know where the car is.  The car reporting system knows where the car is (most of the time) but has no idea where the car is going or what's in it.  The association betweent the two is made in the back-office using some heuristic logic - and it doesn't always get it right. 

The error rate may be very small, but it can come in bunches.  You may get 20 cars out of one industry where erroneous car reporting or logic causes you to associate the previous empty waybill with the current load.  So, you think that they are empties when they are actually loads.  If they make up half the local train that pulled them from industry, the safe braking curve calculated from this information would not insure the train stopped short of movement authority. Very bad.

Part of the problem is drawing a bright line between the placment, release and pull.  The placment and pull information is manually collected by the train crew, faxed to the yard office, and manually entered into the car reporting system.  Despite edits in the data systems there are still lots of places for errors and omissions to occur.  Even when the data is correct, there are still timing issues that can cause problems.  You can get a release from the customer before you even get the placement event into your own system.

Elevating car reporting and waybilling systems to be part of a safety system leaves butterflys in my stomach.

I think the way to do this is to implement ECP braking and get data from each car directly.  The "smarts" on each car plus the commications train line could handle more than just braking info.  You could get car weight (good enough for a braking algorithm), bearing temperature info, car ride quality info, etc. by adding a few sensors.

-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 Thursday, June 11, 2009 1:01 PM

HarveyK400
I understood it took about a second per car, 1-1/2 min for a 90-car train, for brake pipe pressure reduction (signal propagation?) which is the reason for the interest in electronic brake control.  Did I get this wrong; or are there separate issues?

The brake valves react rather quickly to the pressure signal, but the brake cylinder gets up to full pressure rather slowly.  But, there is not much variablility car to car.

HarveyK400
Conversely, are dynamic brakes operationally unreliable?  Isn't this similar to a unit being shut down in modeling performance?

Every spring and fall there is a spate of UDEs due to frost in the brake valves for exactly this reason.

HarveyK400
Conversely, are dynamic brakes operationally unreliable?  Isn't this similar to a unit being shut down in modeling performance?

DB can quit and the train can still operate through.  Think of it as a "nice to have" train handling feature.  Whether it's working or not is really a moot point.  I suppose a rather sophisitcated train automation system could plan on using DB and if results were not sufficient, go to the air.

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

  • Member since
    October 2008
  • From: Near Burlington, WA
  • 380 posts
Posted by Maglev on Thursday, June 11, 2009 3:34 PM

Pardon me, but to some extent I feel that the technical discussion is a diversion from the intended thread topic.  The US needs to improve and expand passenger rail service. Airlines are going broke.  Planes are falling*.  Automobile manufacturers are going broke.  Highways are overcrowded.

Without a philosphical commitment, the economic and political barriers will always be too great.  Without a product to sell, market studies are meaningless.

---------------------

*(Slight digression--severe tropical thunderstorms, likely cause of AF 447 crash, will become more common as climate changes; this is a scientific consensus and my observation from living on Maui over 37 years.  Weather was one element of the Buffalo crash, economic stability of airlines another.  Christian Science Monitor has an on-line item today about possible terrorism in Air France crash; I doubt it, but for more on that topic see the "Railroads for Civil Defense" thread.)

 

 

 

"Make no little plans; they have no magic to stir men's blood." Daniel Burnham

  • Member since
    May 2002
  • 16 posts
Posted by natelord on Friday, January 29, 2010 11:52 PM

Why not see if Nystrom (Milwaukee Rd Hiawatha) bogie trucks could be fitted to Viewliners?  The patents have probably expired.  

  • Member since
    October 2006
  • From: Chicago, Ill.
  • 2,843 posts
Posted by al-in-chgo on Saturday, January 30, 2010 12:04 AM

natelord

Why not see if Nystrom (Milwaukee Rd Hiawatha) bogie trucks could be fitted to Viewliners?  The patents have probably expired.  

 

I was quite surprised the other day to see a Hiawatha zip past the Touhy Ave. (Niles) crossing.  Consist was a mixture of Amfleet and the square bodies I associate specifically with the Hiawatha fleet. 

Next thing you know, they'll be wanting reservations!  AFAIK Hiawathas are the only Amtrak trains that don't require them.

 

best

al-in-chgo

 

al-in-chgo

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