battalion51
Engineer
Brake Pipe and Main Reservoir.
ACS-64 Testing Begins
- Thread starter rickycourtney
- Start date
Help Support Amtrak Unlimited Discussion Forum:
- Status
So: pressure gauges?
jis
Permanent Way Inspector
Staff member
Administator
Moderator
AU Supporting Member
Gathering Team Member
The two circular meters on the left are two pressure gauges related to the braking system.
The vertical bar graph is tractive effort.
The horizontal bar graph is rate of acceleration.
Somebody was holding down the throttle to make the thing move. The throttle and the brake controls are not visible in the video. Nor is the ACSES Display Unit.
you see the Alerter go on every so often. The response switch is also not in the video. but each time the alerter goes on the operator has to knock the response switch to keep the thing from stopping itself.
Last edited by a moderator:
JIS - as always - a man that knows his topic. Thank you.
The traction force I presume is in 1,000 ft lb units, ie, the max for the acs-64 is 72k ftlb, and it the bar graph is marked as 10, 20, 30... 70, 80 - that would fit?
On the traction force gauge - there is the bar graph - which I assume is the currently applied force, but there is a triangular pointer on its left side - possibly some software perception of what the current condition will permit? Or maybe an operator set desired value?
On the acceleration horizontal bar graph - any guess as to the units? In milli-G's? fps^2 or mps^2? ... but none seem to fit the digital readout associated with the bar graph.
I assume the "alerter" is effectively an old world "dead-man switch" ?
again, many thanks - greg
The traction force I presume is in 1,000 ft lb units, ie, the max for the acs-64 is 72k ftlb, and it the bar graph is marked as 10, 20, 30... 70, 80 - that would fit?
On the traction force gauge - there is the bar graph - which I assume is the currently applied force, but there is a triangular pointer on its left side - possibly some software perception of what the current condition will permit? Or maybe an operator set desired value?
On the acceleration horizontal bar graph - any guess as to the units? In milli-G's? fps^2 or mps^2? ... but none seem to fit the digital readout associated with the bar graph.
I assume the "alerter" is effectively an old world "dead-man switch" ?
again, many thanks - greg
Last edited by a moderator:
Fan Railer
OBS Chief
- Joined
- Jun 23, 2012
- Messages
- 887
I can answer two of your questions:
Firstly, the triangular pointer on the TE bar signifies how much TE is commanded by the current throttle setting (also factoring in current speed). Whether or not the blue bar (signifying actual TE) reaches the command triangle depends on adhesion conditions. You can clearly see in the video that for the low speed acceleration part, wheel slip does occur, and if you watch in HD, you can see the wheel slip indicator pop up on the screen.
Secondly, the accelerometer is measured in units of "miles per hour per minute." Once again, if you watch in HD and full screen, you can make it out. I will also post a popular high resolution image here of the cab display courtesy of the same video. Perhaps you will be able to make it out better:
Fan Railer - many thanks for the posting/reply - clearly by someone that know of which he speaks.
With your hi-res image in fact I was able to read what you pointed out - again, thanks.
A question about wheel slip - clearly unbridled wheel slip is rough both on the rails and the engine - the fact that the bar graph shows a value less than the commanded value - is that simply raw slip which is limiting what is applied and the drive wheels are slipping significantly; or is there software, much like an ABS unit in a car, but in reverse, putting down as much TE as the situation will allow, in attempt to save both rails and engine, ie, walking the edge of adhesion, but putting down/out no more torque than that? Related: if I read correctly, the -64 is capable of 72kftlb TE, but the highest commanded value I see is in the 64-65k range - is that a result of restraint on the engineer's part, or have the throttles been "detuned" and "100%" is derated to that 65k level?
again, many thanks for the much better image and thanks for the answers - greg
Fan Railer
OBS Chief
- Joined
- Jun 23, 2012
- Messages
- 887
In the video, I am guessing that the throttle is only at around 90%. Addressing the wheel slip issue, yes, the locomotive's software is limiting tractive effort as it detects wheel slip. Remember that this is one of the test locomotives at Pueblo, so I doubt any sort of limitation is applied to the amount of power that the locomotive will provide given the maximum throttle setting.
Yet again: thanks for the response.
W/re wheelspin - how is it detected? An unusual or unexpected increase in wheel speed, ie, they have an ODE model of wheel speed vs TE and anything outside of that is taken as wheelspin and the traction motors, or maybe just that pair (assuming two motors per axle), are throttled back until the ODE model is met? Or, do they use the slowest wheel as indicative of actual loco speed, and any driven wheels running faster are taken as slipping and are throttled back? Or, there is an independent loco speed sensor (since aren't all wheels on the -64 driven) that is used as the reference and any others are throttled back until they're running not faster or only acceptably faster than that reference? I'm assuming that all four sets of driven axles have their own drive logic/electronics and are driven independently, correct?
I'm assuming that the 90% was simply prudence? ... and at 100% wouldn't that have generated a 0-60 in the mid-thirteen second range - on par with most of the non-muscle cars of the 1960s or 70s? Pretty impressive.
Again: thank you - greg
W/re wheelspin - how is it detected? An unusual or unexpected increase in wheel speed, ie, they have an ODE model of wheel speed vs TE and anything outside of that is taken as wheelspin and the traction motors, or maybe just that pair (assuming two motors per axle), are throttled back until the ODE model is met? Or, do they use the slowest wheel as indicative of actual loco speed, and any driven wheels running faster are taken as slipping and are throttled back? Or, there is an independent loco speed sensor (since aren't all wheels on the -64 driven) that is used as the reference and any others are throttled back until they're running not faster or only acceptably faster than that reference? I'm assuming that all four sets of driven axles have their own drive logic/electronics and are driven independently, correct?
I'm assuming that the 90% was simply prudence? ... and at 100% wouldn't that have generated a 0-60 in the mid-thirteen second range - on par with most of the non-muscle cars of the 1960s or 70s? Pretty impressive.
Again: thank you - greg
Last edited by a moderator:
jis
Permanent Way Inspector
Staff member
Administator
Moderator
AU Supporting Member
Gathering Team Member
Greg, a radar unit is used to calculate actual linear speed of the engine. From that you can compute what the wheel rotation speed should be. If it is higher than that beyond a threshold then the Wheel Slip Management unit will take appropriate action to bring down the angular speed of the wheel to match that as computed from the linear velocity of the engine. These units have become extremely sophisticated and able to pretty much apply the absolute maximum torque/tractive effort just below wheel slip. Usually when they are active and running at the edge, you will hear an occasional screeching noise even on straight track.
The radar is in the same band as police radar in some cases so it is kind of fun to watch those over-speeding on parallel roads depending on their radar detectors.
BTW, even putting the throttle at 100% would not cause it to go any faster since the wheel slip management would intervene anyway.
And yes each axle is driven independently, though I believe each truck has a single set of inverters driving them, though I could be wrong.
The radar is in the same band as police radar in some cases so it is kind of fun to watch those over-speeding on parallel roads depending on their radar detectors.
BTW, even putting the throttle at 100% would not cause it to go any faster since the wheel slip management would intervene anyway.
And yes each axle is driven independently, though I believe each truck has a single set of inverters driving them, though I could be wrong.
Btwn you and Fan Railer, lots of leaning to be had - many thanks to both of you... as much as I like riding trains, have spent too many years as a lab scientist to not ask (need to ask??) the how and why questions. ;-) ;-(Greg, a radar unit is used to calculate actual linear speed of the engine. From that you can compute what the wheel rotation speed should be. If it is higher than that beyond a threshold then the Wheel Slip Management unit will take appropriate action to bring down the angular speed of the wheel to match that as computed from the linear velocity of the engine. These units have become extremely sophisticated and able to pretty much apply the absolute maximum torque/tractive effort just below wheel slip. Usually when they are active and running at the edge, you will hear an occasional screeching noise even on straight track.
The radar is in the same band as police radar in some cases so it is kind of fun to watch those over-speeding on parallel roads depending on their radar detectors.
BTW, even putting the throttle at 100% would not cause it to go any faster since the wheel slip management would intervene anyway.
And yes each axle is driven independently, though I believe each truck has a single set of inverters driving them, though I could be wrong.
Seems radar is a bit over kill - any idea why they went that route... or was it the best answer at the time and now that a system is in place, for inertial reasons, it'll stay? Is there some reason that I'm missing that some linear rail speed measure wasn't used instead?
W/re the 100% - at lower speed levels: yes, absolutely - one can only accelerate as fast and as long as there is sufficient friction.. but in the vid above some velocity it seems that the actual TE and the commanded were the same, ie, there was sufficient traction, and probably/maybe enough more to run it up to 100%... actually, if this was a test run, (to show what it could do) - one would've almost expected the engineer to push the throttle to 100% and simply let the software deliver what was available - hence showing "what it could do."
yet again - thanks for the learned response - greg
ps. related - given the level of software control/sophistication one sees in the -64: why isn't there a track obstruction detector/warning system? From the software perspective, such doesn't seem to be terribly difficult, and with a reasonable vid sensor (and the already present forward lighting) one should be able to pretty cheaply build one that is significantly better than engineer eyeballs? One would have to decide if it was to be given control over the throttles, or merely used as another set of superior eyeballs which are used to warn the engineer (though in this case the extra seconds that would be dissipated getting the human to think and then do something might reduce the usefulness of such a system)... maybe a hybrid where 4 or 5 sigma reliable detection takes control; and 2 or 3 sigma detection issues a warning.. or some blend of the like.
Last edited by a moderator:
jis
Permanent Way Inspector
Staff member
Administator
Moderator
AU Supporting Member
Gathering Team Member
I believe radar is used because it is a nice cheap self-contained unit. Heck My Prius uses a little radar to measure distance from and speed of the car ahead of me to modulate the speed of the cruise control to match that of the car ahead of me, to maintain the distance from it that I set through controls on the steering wheel.
With the Prius, using a micropower radar makes perfect sense, ie, looking for large highly microwave reflective objects with clean signatures... but with a train with so much clutter around, seems like a hit or miss proposition. Any idea when they went to radar for speed determination (era wise) - that might offer some hints as to why?
jis
Permanent Way Inspector
Staff member
Administator
Moderator
AU Supporting Member
Gathering Team Member
They don't look at clutter. They look at the track ties etc. It actually is quite reliable I am told. But I have no first hand knowledge or experience of this matter.
ericr
Train Attendant
jis
Permanent Way Inspector
Staff member
Administator
Moderator
AU Supporting Member
Gathering Team Member
That's a great picture. Thanks. You can also see the full ATC/ACSES display unit in that photo. In the other one you can only see half of it. Of course with both ATC and ACSES cut out, there is not much to see there.
Last edited by a moderator:
Many thanks for the post - truly easy to read all the controls and labels... when one can actually read stuff, one can find answers ;-)
cirdan
Engineer
- Joined
- Mar 30, 2011
- Messages
- 3,682
Back to the topic of wheelslip.
Wheelslip is not necessarily an evil in all situations. In fact a certain amount of wheelsip occurs all the time that tractive force is being applied and this is actually beneficial for adhesion. Without going into the physics and math behind this (which could fill volumes if you dig down deep enough) a certain amount of slip actually permits adhesion to be increased beyond the level that no slip occurs. On your car, the tires partially deform as they press down on the road, and without this you wouldn't be able to grip the road. Train wheels don't deform to the same degree, but this is compensated by the metal on metal interaction that moving surfaces create.
On modern locomotives, individual axles are even intentionally driven at different speeds and toruqe settings. For example under rainy conditions, adhesion can deteriorate, and to compensate this, the leading axle can be spun slightly faster than the rest to create a clean dry rail surface and permit normal adhesion for the following axles. Also, due to the train pulling from behind, the front axles tend to lift ever so slightly and the rear axles pushed down a little bit more. It's a bit like pulling a heavy trailer behind your car, where you can easily get problems if you have a front wheel drive and the car isn't very heavy for example. To make use of this phenomenon, the rear axles also do slightly more work, ie, they are controlled to output a slightly higher torque, as the extra pressure means they are less likely to slip, and so you can move closer to the limits of the equipment.
All this is controlled by the locmnotive's micropocessor which continuously monitors torque response and slip on the individual axles and infers from these what the problem is (such as is the rail too slippery or the load too heavy for the present algorithm), and reacts appropriately. So the engineer doesn't even have to type those details into the computer. It works it all out by itself.
Wheelslip is not necessarily an evil in all situations. In fact a certain amount of wheelsip occurs all the time that tractive force is being applied and this is actually beneficial for adhesion. Without going into the physics and math behind this (which could fill volumes if you dig down deep enough) a certain amount of slip actually permits adhesion to be increased beyond the level that no slip occurs. On your car, the tires partially deform as they press down on the road, and without this you wouldn't be able to grip the road. Train wheels don't deform to the same degree, but this is compensated by the metal on metal interaction that moving surfaces create.
On modern locomotives, individual axles are even intentionally driven at different speeds and toruqe settings. For example under rainy conditions, adhesion can deteriorate, and to compensate this, the leading axle can be spun slightly faster than the rest to create a clean dry rail surface and permit normal adhesion for the following axles. Also, due to the train pulling from behind, the front axles tend to lift ever so slightly and the rear axles pushed down a little bit more. It's a bit like pulling a heavy trailer behind your car, where you can easily get problems if you have a front wheel drive and the car isn't very heavy for example. To make use of this phenomenon, the rear axles also do slightly more work, ie, they are controlled to output a slightly higher torque, as the extra pressure means they are less likely to slip, and so you can move closer to the limits of the equipment.
All this is controlled by the locmnotive's micropocessor which continuously monitors torque response and slip on the individual axles and infers from these what the problem is (such as is the rail too slippery or the load too heavy for the present algorithm), and reacts appropriately. So the engineer doesn't even have to type those details into the computer. It works it all out by itself.
Last edited by a moderator:
- Status
