Oil Train

[BaltACD]

Euclid

The load/empty sensors would have to be reliable, but is this not possible?  How do these work when applied to cars with conventional air brakes?  Paul North mentioned unreliability relating to pneumatic components.  Are these components related to the pneumatic control of the conventional brakes, or are they part of the sensor itself? 

Load sensors for conventional braking require a battery power source, a wireless transmission, and apparently a battery charging means.  Considering these requirements, wouldn’t load sensors be simpler to execute with ECP brakes?  You would not need the wireless transmission or the battery power supply since these functions can be performed by the ECP control wire. 

You would also not need a finicky pneumatic control to adust the brake pressure according to the load/empty status. 

 

The P in ECP is Pneumatic - which is where the brakes get their brakeing power - the EC just transmits the braking signal faster.  You still have to deal with the finicky aspects of pneumatic controls to realize your actual braking power.

[Paul_D_North_Jr]

Euclid

The load/empty sensors would have to be reliable, but is this not possible?  How do these work when applied to cars with conventional air brakes?  Paul North mentioned unreliability relating to pneumatic components.  Are these components related to the pneumatic control of the conventional brakes, or are they part of the sensor itself? 

Load sensors for conventional braking require a battery power source, a wireless transmission, and apparently a battery charging means.  Considering these requirements, wouldn’t load sensors be simpler to execute with ECP brakes?  You would not need the wireless transmission or the battery power supply since these functions can be performed by the ECP control wire. 

You would also not need a finicky pneumatic control to adust the brake pressure according to the load/empty status. 

My understanding is that the load/ empty sensors and valves are strictly mechanical and pneumatic - no electric power required.  Again see Sec. 1.14.4 - Sensor Plates and Empty/ Load Devices of the TSB report. 

[dehusman]

Euclid

Load sensors for conventional braking require a battery power source, a wireless transmission, and apparently a battery charging means. 

Not at all.  They can be 100% mechanical and operate automatically, for that one car.  That's how they were done in the past.

Considering these requirements, wouldn’t load sensors be simpler to execute with ECP brakes? 

Load detection would be simpler with a power source.  Glue a strain gauge on the centersill and viola', load sensor.  No moving parts but it does require power.  Electronics to interpret the gauge signal would also require power.  The next part would be an electro-pnuematic device to change the setting in the brake valve, which would also require electricity.  So far none of this requires any communication with any other car to work.  All it requires is hardware on the car and an electricity source.  No batteries required.  If the power source the fails then the train should go in emergency and should not release until the power is restored, the brakes are cut out or the brakes are set to conventional operation.

You would also not need a finicky pneumatic control to adust the brake pressure according to the load/empty status. 

Sure you would since the brakes are still air brakes, and still have an pnuematic brake valve.  Its just a different method drives the controls of the pnuematic brake valve.  It changes from air operated controls to electric operated controls.

[Euclid]

Paul,

Thanks for working those numbers.  I will have to ponder that, but one thing occurs to me.  Are you determining the number of cars derailed correlating with the length of those cars in relation to the stopping distance?  I wonder if there may be other factors such as the resistance effect of the derailing cars. 

As I understand it, you are comparing stopping distance of ECP brakes with or without the maximization of brake force appropriate for loaded cars.  If that is correct, what then happens if you compare ECP brakes to conventional air brakes with that same maximized braking for the loads; and then factor in the instant application of ECP brakes versus the timed propagation of conventional air brakes? 

I was reading this reference:

http://onlinepubs.trb.org/onlinepubs/archive/studies/idea/finalreports/highspeedrail/HSR-51Final_Report.pdf

This approach to load sensing does indeed seem very complicated and costly.  It leaves me wondering why simply sensing the considerable difference in truck spring deflection would not be a lot simpler than the sophisticated strain gauge method shown in the link. 

But better yet, would be the load/empty switch to instantly convert the whole train between the two modes of braking. 

[Euclid]

Dave,

How does the sensor unit work in the 100% mechanical load sensing system?

When I referred to the “finicky” pneumatic control, I was referring to what Paul mentioned regarding dirt causing problems, and assumed that this was in the pneumatic control element that would determine how much pneumatic power to provide to the brakes in differentiating between loads and empties.  I also assumed that this control element is added as part of the load sensor system.  I understand that the braking power is developed pneumatically in both ECP and conventional air brakes.

[BaltACD]

While the derailment in this video is cause by a tornado, once the derailment happens, the resulting footage is indicatative of what happens to the inertia of a train, even with the brakes fully applied in emergency.

[BroadwayLion]

Euclid

Load sensors for conventional braking require a battery power source, a wireless transmission, and apparently a battery charging means.

Yadda Yadda Yadda!

If you use couplers of LION, power sourse is a gimmie. Laser was only a quick example. An air bladder and gauge would work just as well, as has been on the subways for half a century or more.

How would electric brakes work, if not through an electrical connection between the cars. Even if no special coupler is used, it is a simple matter to MU the cars with a jumper. Look, if you are building better cars, then you *can* build them from the track up.

There *are* issues I have not considered, but once considered can be solved.

ROAR

[BroadwayLion]

Euclid

This approach to load sensing does indeed see very complicated and costly. It leaves me wondering why simply sensing the considerable difference in truck spring deflection would not be a lot simpler than the sophisticated strain gauge method shown in the link.

How is this costly?

You have a bathroom scale do you not? Maybe an electronic one? They got electronic scales all over the place, truck terminals, etc.

It would not even take that. The CR or the car knockers walk down the train before departing a terminal, a glance at the springs will tell him if the car is loaded or not, and from there it would be a matter of shoving a lever this way or that for a load or an empty, and moving on to the next car.

IF you use automatic coulplers ala LION, this information would already be on the locomotive's computer via the conductors "paper work" and the brake setting can be adjusted from the conductor's computer screen.

ROAR

[dehusman]

BroadwayLion

How is this costly?

 

Hold that thought.

 

IF you use automatic coulplers ala LION, this information would already be on the locomotive's computer via the conductors "paper work" and the brake setting can be adjusted from the conductor's computer screen.

So your automatic couplers not oly require a new car they require a new engine to pull the car, one that is equipped with the new couplers, air system, control system, computer interface, etc. etc.

As you mentioned its all possible.  The question is is the price worth the benefit?  So far the railroads have tried ECP and it has not worked well enough or had enough benefit compared to the cost to warrant a large scale trials, let alone conversion.

[Euclid]

BroadwayLion

 

Euclid

This approach to load sensing does indeed see very complicated and costly. It leaves me wondering why simply sensing the considerable difference in truck spring deflection would not be a lot simpler than the sophisticated strain gauge method shown in the link.

How is this costly?

 

The complete context of what you quoted from me is this:

 

"I was reading this reference:

http://onlinepubs.trb.org/onlinepubs/archive/studies/idea/finalreports/highspeedrail/HSR-51Final_Report.pdf

This approach to load sensing does indeed see[m] very complicated and costly. It leaves me wondering why simply sensing the considerable difference in truck spring deflection would not be a lot simpler than the sophisticated strain gauge method shown in the link. "

 

You can see that when I said it seems costly, I was referring the appoach detailed in the link.

In any case, I think the advantage of load sensing brake adjustment on oil trains (with ECP brakes) could be achieved without load sensors because there is no question as to whether cars are loaded or empty, or anywhere in between.

[Wizlish]

Why is this discussion going all over the place with lasers and load cells in the suspension?

Previous posts have referred to one existing method of adjusting load/empty braking force for air-braked cars.  This uses a mechanical method similar to that used on some systems of automatic load-leveling for automobiles.  Here is an illustration of the mechanism:

This is from the Canadian accident report that was referenced earlier in this thread.  Referenced in that thread is the need for a 'sensor plate' to be attached to the truck frame -- this is not an active load cell or anything sophisticated that requires electricity or the communication of logic signals, it is just a solid flat surface that the 'sensor arm' of the apparatus bears on as the truck rotates relative to the car on curves. 

I found it interesting that so many of these plates were found damaged or missing when the accident investigation in 2007 was progressing.  This suggested to me that the preparation or welding operation used to attach them was somehow deficient -- I don't immediately see how any normal or even accidental force would tear such a plate, in such a location, loose from a sideframe, so I would have to suspect corrosion damage, or little more weld integrity than that for tack welds.  I can see how the assumption could be made that these plates only see minimal force applied to them in service, and that extensive welding on the sideframes themselves might cause HAZ damage.

There is little doubt that a setup like this could be adapted to oil-train service, and with little additional work it could be made to have multiple stages of actuation or even be made proportional.  In such a setup, the integrity as well as the proper setup of the sideframe reference plates becomes an issue for attention.

(Yes, I thought it was interesting that the improper 5/8" setup height of the plates was noted to counteract the effect of the plate breaking off with respect to actuation of the ride-height valve in the Lillooet accident!)

[BroadwayLion]

dehusman

So your automatic couplers not oly require a new car they require a new engine to pull the car, one that is equipped with the new couplers, air system, control system, computer interface, etc. etc.

Nope, not at all. Oil Trains require a sand car between the Oil Cans and the Locomotive. That is your transition car, it holds all of the necessary electronics, batteries, generators, terminals, and etc. for the operation of the train. It has a trainline coupler on one end, and a regualr coupular on the other. It could transpit wirelessly to a console in the locomotive (lap top or rugged tablet) used by the conductor.

Use any locomotives you wish, use locomtoives at both ends.

couplers open with a key rather than with a lift bar. Simple adapters on each car will connect iron to iron as necessary for shop, yard or setout movements.

ROAR

 

[zugmann]

BroadwayLion

Nope, not at all. Oil Trains require a sand car between the Oil Cans and the Locomotive. That is your transition car, it holds all of the necessary electronics, batteries, generators, terminals, and etc. for the operation of the train. It has a trainline coupler on one end, and a regualr coupular on the other. It could transpit wirelessly to a console in the locomotive (lap top or rugged tablet) used by the conductor.

 

Hazmat placement regulations restrict certain loaded hazmat cars (I think oil qualifies without looking at the chart) from being coupled next to cars with internal combustion engines (aka generators).

[Euclid]

Wizlish

Previous posts have referred to one existing method of adjusting load/empty braking force for air-braked cars. 

How common is the use of these load sensors on North American freight cars?  Why would a railroad company choose to install these devices on their rolling stock?

[cx500]

Paul North's calculations provide a fascinating comparison of the results of increasing the Net Braking Ratio to what is achievable in ideal conditions.  In less salubrious situations, as he also mentions, the reult may be a train of flatted wheels, still cheaper than a major derailment.  The flipside to this, and perhaps why the NBR is kept low, is that from time to time a train may experience an undesired emergency brake application due to a faulty brake valve somewhere on the train.  No derailment prevented, but still a lot of flatted wheels to deal with.

Locomotives can deal with varying factors of adhesion by their wheel slip sensing system.  In theory you could do something similar on each freight car (it was used on some passenger cars such as CPR's The Canadian) but the cost would be staggering to provide it on the North American freight car fleet.

I have no idea when the value for NBR was established so it may be a legacy of past experience.  Perhaps it can be raised to reflect the current more modern cars but in reality the brake/wheel/rail interfaces have changed very little in the past 100 years.

[Paul_D_North_Jr]

Euclid

[snipped - PDN] . . .

 

Are you determining the number of cars derailed correlating with the length of those cars in relation to the stopping distance?  I wonder if there may be other factors such as the resistance effect of the derailing cars. 

Yes.  The scenario behind my example was simple: a car ahead derails, comes to a complete stop instantly, and becomes an immovable object.  How many of the trailing cars will crash into it or each other before the rest of the train is stopped by braking force only ?  That's basically the distance traveled by a car subject only to that braking force.  As you imply, there are other forces that might reduce that distance, as a practical matter: the 1st derailed car might roll or slide ahead a little ways until it comes to a halt, providing a slight additional stopping distance; it and the subsequent derailed cars will experience further deceleration as they plow through the track structure and the ballast; and until they become uncoupled or jack-knife at 90 deg. to each other, the couplers and frames will transmit some additional reaction/ pushback force towards the other cars at the rear of the train. 

 

 

Euclid

As I understand it, you are comparing stopping distance of ECP brakes with or without the maximization of brake force appropriate for loaded cars.  If that is correct, what then happens if you compare ECP brakes to conventional air brakes with that same maximized braking for the loads; and then factor in the instant application of ECP brakes versus the timed propagation of conventional air brakes?  

Not quite.  I was comparing the effect on stopping distances on the basis of a former typical Net Braking Ratio, then 2 recent/ current maximum NBRs, and then 1 greater than is now currently specified, presuming it could be obtained with a more sophisticated empty /load device, faster-acting brakes, and a more powerful braking system. - regardless of the kind of controls used.  That likely would be ECP, but not necessarily - after all, someone could invent a way to do that with conventional brakes . . .  

I did not directly compare ECP vs. conventional because of the complexities of the conventional systems - all of which make them slower to react than the ECP.  I acknowledge those complexities as a qualitative issue, but quantifying them is beyond the scope of what I was trying to illustrate.   

- Paul North.

[Paul_D_North_Jr]

cx500

[snipped - PDN.] . . . I have no idea when the value for NBR was established so it may be a legacy of past experience.  Perhaps it can be raised to reflect the current more modern cars but in reality the brake/wheel/rail interfaces have changed very little in the past 100 years.

Per Sec. 1.14.2 of the TSB report, the NBRs were set at least once in 1999, then raised in 2004.  The quoted 2004 AAR Air Brake Association paper indicates that the Canadian Class 1's had been exceeding those values "for many years" to assure adequate performance. 

- Paul North. 

[Wizlish]

Euclid

Why would a railroad company choose to install these devices on their rolling stock?

The principal use for freight is on cars with a sizable difference between loaded and unloaded weight.  A good representative example was the car illustrated in the accident report, a center-beam flatcar used for lumber products.

With developments in reducing tare weight, the importance of adjusting NBR on loads vs. empties, in as automatic and 'car-specific' a way as possible, has taken on more significance.  The key has been to accomplish that in ways that best fit contemporary railroading practice.

It is also becoming more and more important to avoid even momentary wheelslide on low-tare-weight trains with high axle loading -- two examples being stack trains and coal-gon unit coal trains.  I think it is easy to comprehend why the use of 'automatic' proportioning would be of value on those consists.

Conversely, the advantage would be 'lower' for the specific case of oil trains built to increasingly armored standards, operated at lower speeds.  I'd still find the technology valuable, of course ... and would also think it would be political low-hanging fruit to mandate the use of brake proportioning on these "dangerous" consists in order to improve "safety".

[Euclid]

Wizlish

I understand those benefits and see their value.  I also definitely see your point about mandating load sensors on oil trains as being low hanging fruit in the current politics of the issue.  But my question is to ask why a railroad company would decide to add load sensors to a purchase of say 50 new freight cars.  If say two of those cars happen to be running in another company's train containing 100 cars without load sensors, it would be a slight benefit to that train, and no benefit to the railroad company that owns the two cars with the sensors.  In other words, why would a company spend money improving the stopping performance of another company's train?

This thought leads to another question:

Say you have a mixed consist train of loaded 100 cars, and there are 25 cars with load sensors and 75 cars without them.  Say the 25 cars with sensors are all together in the train.  Wouldn’t the extra brake force on the 25 cars have a bad effect on train handling? 

[BaltACD]

Euclid

Wizlish

I understand those benefits and see their value.  I also definitely see your point about mandating load sensors on oil trains as being low hanging fruit in the current politics of the issue.  But my question is to ask why a railroad company would decide to add load sensors to a purchase of say 50 new freight cars.  If say two of those cars happen to be running in another company's train containing 100 cars without load sensors, it would be a slight benefit to that train, and no benefit to the railroad company that owns the two cars with the sensors.  In other words, why would a company spend money improving the stopping performance of another company's train?

This thought leads to another question:

Say you have a mixed consist train of loaded 100 cars, and there are 25 cars with load sensors and 75 cars without them.  Say the 25 cars with sensors are all together in the train.  Wouldn’t the extra brake force on the 25 cars have a bad effect on train handling? 

 

Car owners pay for repairs to their cars, even when the repairs are made on another railroad.  There is and has been in place a uniform system for repairs to be billed back to the owner at at standard price for each type of repair made.  The benefit of load sensor equipment on cars would manifest itself as decreased billing for wheel change out from slid flat wheels.

In train handling, from the engineers perspective, he feels total train retardation - no matter how it is generated within the train.  So long as no wheels are sliding he doesn't care how the train retardation is being accomplished.  Handling freight train air brakes is not like braking ones automobile, as the only brake modulation that is available is to increase braking pressure or to fully release the brakes.

[Euclid]

BaltACD

The benefit of load sensor equipment on cars would manifest itself as decreased billing for wheel change out from slid flat wheels.

But why would load sensors help prevent wheel slides?  Do they reduce braking force to less than it would be on empty cars without load sensors?  I thought that they only increased braking force on the loads. 

[Wizlish]

OK, I think I see where this is going off the trolley a bit.

I don't believe these valves are used to INCREASE braking force above the 'usual' default calibration for a car, which is going to be for close to the nominal loaded weight.  The valve only derates braking force when the car is empty, so that the wheels will not slide.

The analogy with load-leveling is that the control is only applied in one direction; do not be confused by the fact that load leveling applies more force to create its effect, while railcar load-sensing reduces force to create its effect.  In both cases, the system usually doesn't compensate 'in the other direction' -- in load leveling, there usually isn't any active 'kneeling' action to force the rear of the vehicle down if, say, negative tongue load were applied to a hitch.

Some passenger train braking systems were set up to provide an increased level of braking at high speed -- the Decelakron, if I remember correctly, used a sliding inertial weight to give a control input for the physical rate of deceleration, and modulated the braking effort to achieve the higher brake cylinder pressure needed to produce that rate at higher speed.*  I wouldn't expect such a system to be of practical use on freight consists, even if it could be adjusted and maintained effectively in typical interchange service.

My understanding of how the antislide valve is calibrated is that it keeps braking effort 'well enough' below where wheels will slide for any expected combination of car weight and adhesion.  It's difficult to implement 'antilock brakes' on a foundation-braked car -- that's one of the reasons passenger cars went to truck-mounted cylinders with short actuating levers to the tread or disc-brake shoes.

 

*There was much more 'stuff' involved in a brake system with a Decelakron, including variable calibration of braking ratio, for example in four steps from 250% above 90 mph down to 100% below 20 mph on Union Pacific in the '30s.  I'm just illustrating how the control proportionality was done.

[Euclid]

My understanding is that without load sensors, freight car air brakes apply the highest braking force possible that does not exceed the wheel slide point of empty cars.  So when the cars are loaded, they could take more braking force without sliding, but none is available, so the loaded cars are under-braked.

In other words, I thought the point of load sensors was to distinguish between loads and empties in order to apply maximum braking force on empty cars, and then apply the higher maximum force on loaded cars.  The point would be better stopping power of the train to the extent that it has loads in it. 

Maybe this is not the correct interpretation.  My next question would have been to ask how the load sensing system in my understanding would add more brake force.

However, you are saying the opposite of my understanding when you say, “The valve only derates braking force when the car is empty, so that the wheels will not slide.”  You may be right, but if this were the case I would think that braking normally maximized for loads will always be excessive for empties.  If that is the case, how can equipment go without this load sensor option?  I would think wheels on empties would be sliding all the time during heavy brake applications if the brakes defaulted with maximum force for loaded cars no matter whether they are loaded or empty. 

Basically, it sounds like you and BaltACD are saying the load sensors are intended to prevent wheel slide on empties.  Whereas, my understanding is that load sensors are intended to provide more brake force on loads than would be the case on cars without load sensors.      

[tree68]

Euclid

Basically, it sounds like you and BaltACD are saying the load sensors are intended to prevent wheel slide on empties.  Whereas, my understanding is that load sensors are intended to provide more brake force on loads than would be the case on cars without load sensors.      

Al Krug

provides the answer.  Scroll down to "Some Other Embellishments."

 

[Wizlish]

tree68

 

 

Euclid

Basically, it sounds like you and BaltACD are saying the load sensors are intended to prevent wheel slide on empties.  Whereas, my understanding is that load sensors are intended to provide more brake force on loads than would be the case on cars without load sensors.      

 

 

Al Krug

provides the answer.  Scroll down to "Some Other Embellishments."

Well, what do you know, I had it exactly backward!

[Euclid]

Wait—I did not explain my understanding properly.  My understanding, like yours, would have the device derate the braking force for empties like you say.  It does say that in the report. 

But my understanding also is that adding this device would also entail the raising of the default braking force to maximum for loads instead of being maximized for empties as is the case on all cars without this device.  I suspect that default braking force would be raised by changing the brake mechanism.    

In that sense, the device itself would not directly increase brake force for loads.  But it would allow a higher default braking force for loads, and thus increase load braking in that indirect way by being able to reduce that higher braking force for loads to the point where it can apply to empties without sliding the wheels.  The point would be to provide greater braking force for loads.

So that is at least the proper explanation of my understanding of how the device works, although that may be incorrect in the case of this ELX-B 60 mechanical-pneumatic device that we have been discussing.  Maybe what I am talking about has not been invented yet.  I will look further into this.   

I will take a look at the Krug link, and a couple other things here. 

[Euclid]

Okay, my understanding of load sensors conforms to that described by Al Krug.  So the point is to add more brake force rather than to prevent wheel slide that would occur with the typical brake system without the load sensor. 

[Paul_D_North_Jr]

tree68

Al Krug provides the answer.  Scroll down to "Some Other Embellishments." 

Direct URL to that section, since it's not at Krug's usual website:

http://www.railway-technical.com/brake2.shtml#EmptyLoadSensors 

See why some of us like Al so much ?

- Paul North.  

[Euclid]

My understanding of load sensors has been this:   They are intended to provide greater braking power for loads. This goal is accomplished by two modifications to a freight car as follows:

1)    Modify the pneumatics of the brake system to increase the maximum brake force to as high as possible when the car is loaded.

 

2)    Add a load sensor that causes the pneumatics of the brake system to reduce the brake force when the car is empty.    

 

So adding a load sensor is not fundamentally intended to prevent wheel slide on empties, although this point can be confusing because in item #2, load sensor does do that in order to achieve the basic goal of item #1. 

So the purpose of the load sensor itself is to reduce the empty car brake force in order to prevent wheel slide.  But that is not the basic, underlying reason for adding a load sensor to a freight car.  The basic reason is to get more braking power on the loads.

So this gets back to my earlier question asking why you would want to add this system to your freight car.  It will have more braking power, but most likely, this will be realized when your car is operating in someone else’s train.  And even then, your improved car may be the only such car in that train.  So your improved car is working extra hard and wearing out brake shoes faster, but with little net benefit to the other guy’s train, and no benefit to you.

The only way this load sensing/enhanced braking makes sense to me is if all railroads agreed to install this on all of their cars.    

[zugmann]

Euclid

So your improved car is working extra hard and wearing out brake shoes faster, but with little net benefit to the other guy’s train, and no benefit to you.

 

It's a lot easier, quicker, and cheaper to replace worn brake shoes than to replace wheels.   Sliding wheels = flat spots.