Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

loco/train acceleration vs notched throttle setting

955 views
14 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    July, 2009
  • From: somerset, nj
  • 2,038 posts
loco/train acceleration vs notched throttle setting
Posted by gregc on Tuesday, June 05, 2018 6:42 PM

i'm curious about how fast an individual locomotive or a train accelerates.   Acceleration is the change in speed over time, mph/sec.   

I'm aware that in RRing, acceleration may be thought of in terms of distance: what distance is required for a train to aquire a desired speed, mph/miles

i'm also aware the getting an entire train in motion requires some patience.  unlike flooring the gas in a mustang which causes the tires to spin, applying too much power and accelerating a locomotive may cause couplers to snap or the conductor to experience whiplash when the caboose coupler slack is finally taken up  and the rest of the train has already gained some speed.

 

specifically, i'm wondering how the notch setting might be used to increase speed depending on the notch assuming the entire train is moving relatively slowly on level grade.

for example, if it's current speed can be maintained in notch 2, how quickly can it gain speed if set to notch 4 versus notch 8?

Assuming there is a choice of notch setting (not withstanding any RR operating procedures) wouldn't a higher notch setting take less time to increase the speed of a train from 25 to 35 mph than a lower notch setting?   

Of course, the notch setting may be reduced once the desired speed is reached and in order to maintain that speed.

greg - Philadelphia & Reading / Reading

  • Member since
    September, 2003
  • 5,809 posts
Posted by Overmod on Tuesday, June 05, 2018 7:55 PM

There will be an effective train resistance, which for argument's sake you can calculate from the Davis formula or determine using a dynamometer car, for any speed of a given train at a given point.  Against this the locomotive exerts tractive effort, with the resultant giving an acceleration (or as appropriate deceleration or 'balancing speed'. 

The governor is set to produce fixed combinations of engine speed and load, within the constraint of the maximum power that can be produced out of a diesel engine.  Effective TE goes down as speed goes up, so in most cases you start approaching a balancing speed (and the rate of change in speed that constitutes acceleration begins to change more and more slowly as you approach that speed)long before you get up to whiz-bang speed.  That's when motoring.  If descending a grade, the resultant of gravity is an implacable source of acceleration that will produce speeds out of any proportion to self-powered engines, rather quickly...

Since as you know the notches (certainly the ones above 4) determine the locomotive output horsepower via the governor, the effective acceleration comes up quicker and reaches a higher peak value as the notch increases.  You can see this easily when a given train (often a commuter or passenger train) gets quickly to notch 8 and stays there as soon as the traction motors have come out of their slow-speed restrictions.

Once "the desired speed is reached" it will be maintained via appropriate manipulation of notch, dynamic, and air -- given the road knowledge that tells an experienced engineer when, and how far, to 'begin compensating' for changes in effective adjusted profile.  Again because of the rectangular hyperbola imposed by fixed peak power from the engine(s), the acceleration may be asynchronous from deceleration, and greater or less than the implacable resultant of gravity that is always pulling on the train along with the locomotive drawbar...

  • Member since
    July, 2009
  • From: somerset, nj
  • 2,038 posts
Posted by gregc on Tuesday, June 05, 2018 8:04 PM

Overmod
Again because of the rectangular hyperbola imposed by fixed peak power from the engine(s), the acceleration may be asynchronous from deceleration

can you explain what "rectangular hyperbola" means?

greg - Philadelphia & Reading / Reading

  • Member since
    September, 2003
  • 5,809 posts
Posted by Overmod on Tuesday, June 05, 2018 9:49 PM

gregc
can you explain what "rectangular hyperbola" means?

It's a geometric curve, one of the conic sections, that can be used to represent the part of the horsepower curve of a diesel-electric locomotive that is not restricted by traction-motor concerns (at the slow end) or back EMF in DC motors (at high rotational speed).  This is plotting tractive effort (usually in the USA, in lb) on the y axis, and speed (usually in the USA in mph, but it can be converted to fps where those units are more direct).  In this region, each point on the curve represents the TE developed at the corresponding road speed with the engine at maximum output at rated HP.

  • Member since
    July, 2009
  • From: somerset, nj
  • 2,038 posts
Posted by gregc on Wednesday, June 06, 2018 4:19 AM

i think the significant points of your explanation are

  • productive tractive effort must be maintained which limits the force that can be applied to the wheels.   no point is spinning wheels 
  • that acceleration decreases with speed because horsepower is the product of force and speed and therefore there is less force as speed increases.   (there's more aerodynamic drag, as well)

i assume that when changing speed, there's an appropriate progression of notches to maximize performance (i.e. get up to speed as quickly as possible).   Is this always the case, would an engineer accelerate to speed slower than possible?

how quickly to switching locomotives in a yard change speed?

 

greg - Philadelphia & Reading / Reading

  • Member since
    August, 2003
  • From: Collinwood, Ohio, USA
  • 5,969 posts
Posted by gmpullman on Wednesday, June 06, 2018 4:44 AM

Hi, Greg

Many of the locomotive operating manuals will explain some basics of train handling and may answer some of your questions about throttle positions, load meter, wheel slip, transition, power braking and such. Depending on locomotive model and the era you are interested in the sophistication of the controls have advanced, IE automatic wheel slip, dynamic brakes, load compensation, etc.

For instance here is the ubiquitous EMD F7 manual:

http://www.rr-fallenflags.org/manual/f7-om.pdf

Scroll down to section two for some operating basics.

Here's a large selection of manuals if there is a particular locomotive you're interested in.

http://www.rr-fallenflags.org/manual/manual.html

Hope that helps, Ed

  • Member since
    July, 2009
  • From: somerset, nj
  • 2,038 posts
Posted by gregc on Wednesday, June 06, 2018 6:47 AM

gmpullman
For instance here is the ubiquitous EMD F7 manual: http://www.rr-fallenflags.org/manual/f7-om.pdf

thanks.   that was interesting.   it seems that there circuits to protect the motors and maximize performance and i can see how that this complicates the answer to my questions.   But i assume an engineer may not always need maximum performance.

 

maybe my questions would be easier to answer if considering how a yard switcher is controlled without any cars attached.

in our cars, we don't "floor it" when the light turns green, but we might "floor it" when merging onto a a busy highway.    I know a car and locomotive are different, but the physics isn't.

greg - Philadelphia & Reading / Reading

  • Member since
    April, 2001
  • From: Roanoke, VA
  • 1,626 posts
Posted by BigJim on Wednesday, June 06, 2018 7:45 AM

With todays wheel slip systems, bringing up the EMD F7 is as dated as the loco itself!

On dry rail, once a train is started, you might be surprised at how quickly you can get into the eighth notch...and with no wheel slip!
I should also add that accellerating a train from a full stop is much different than trying to accellerate one that is moving at speed...especially if it is a GE loco!

Overmod
Again because of the rectangular hyperbola imposed by fixed peak power from the engine(s), the acceleration may be asynchronous from deceleration

Overmod,
Sometimes you need to remember that you are talking over most people's heads.

 

.

  • Member since
    September, 2003
  • 5,809 posts
Posted by Overmod on Wednesday, June 06, 2018 6:15 PM

gregc
It seems that there are circuits to protect the motors and maximize performance and i can see how that this complicates the answer to my questions.

It's really more simple than that: the armature current is high when in series connection near starting (as it is on many DC conventional electric locomotives) and there are both time and amperage restrictions on how the motors are loaded in that range.  (For the F unit, these are clearly spelled out in the operating manual with reference to the main ammeter).  The Baldwin BP20s in commuter service had sufficient reserve in the electrics -- if not as restricted in the manual! -- to alllow them to be accelerated with the ammeter 'pegged' for some time, station after station down the New Jersey coast lines where they ran.

Meanwhile, the main generator field current can also be controlled, and this directly governs (no pun intended) the rate at which full fuel rate can be applied to the diesel engine to make power.  This can be particularly dramatic in earlier GEs, where the engine acceleration under load was held to a low rate for pollution reasons and to reduce the visible and maintenance consequences of turbo lag on large CB engines run to high specific power.  (Most of the guys that ran 'em will tell you they hate slow-loading units...)

Big Jim is right in that modern AC units can be loaded fully at very slow speed, and that computer control over the engine and transmission would improve the quickness with which the locomotive could be accelerated if it were to be programmed that way.

But i assume an engineer may not always need maximum performance. Maybe my questions would be easier to answer if considering how a yard switcher is controlled without any cars attached.

In fact they are much better addressed if considering how a yard switcher is actually employed on most modern railroads for flat switching: here, the object is to kick long cuts of cars with high acceleration, followed by coasting (or energy harvesting) rather than by nice smooth acceleration and deceleration.  The original Green Goat, which was basically a very large battery kept trickle-charged by a small genset, could never really do the former without significant cumulative battery (and perhaps switchgear) damage -- I'm a bit surprised that Elon Musk hasn't adapted ludicrous mode to a modern battery switcher as, if the battery has been designed for proper active heat management (as there is ample space and weight-management provision to do), I suspect you could flat-switch all day on the smell of an oily rag.  A comparable comment could be made about genset engines that have the ability to predictively start and ramp up sets of their engines when needed; the wait to start... and then warm... and then accelerate slowly to load for pollution... being ridiculously long and all the effect essentially thrown away when the throttle is notched down for long enough.

In our cars, we don't "floor it" when the light turns green, but we might "floor it" when merging onto a a busy highway. I know a car and locomotive are different, but the physics isn't.

Where the physics differs so substantially is in the actions used to accelerate the engine to make power.  There's enough 'vanity cushion' and transmission kickdown, etc. in your car to make flooring it practical much of the time, but flooring most diesels is an open invitation to a component-damaging coal roll -- maximum acceleration being to accelerate the engine under minimal load to just over the anticipated rotational speed (and turbo speed) and then smoothly bring up the load through whatever gears you have and spooling up the turbo to boost mass flow reasonably quickly.  This takes more thought and discipline than in a throttled gas engine or, through the computer, a GDI engine.  And as observed in locomotives, many more careful steps in the right order.

Point about hyperbola well taken.

  • Member since
    July, 2009
  • From: somerset, nj
  • 2,038 posts
Posted by gregc on Thursday, June 07, 2018 11:47 AM

overmod

thanks much for the description of locomotive dynamics.    This kind of stuff is facinating

but i'm still without an answer.   It sounds like the notches on a locomotive are really just speed steps similar to our DCC controllers.   The engine electronics manages power to the motors to obtain the speed corresponding to the notch as efficiently (least amount of fuel and stress on the motors) as possible.

 

my motivation is a better understanding of the linear implementation of acceleration modeled by the momentum described by CV3. 

I remain skeptical that locomotive acceleration is constant (due to physics) up to the point where the speed corresponding to a notch is reached.  I find it hard to believe it is the same at some low speed where the force applied to the wheels is presumably higher than near the top speed in notch 8.

i think it would be more realistic if there was greater acceleration at low speed at DCC speed step 28 than in step 10 and that the speed step can govern that acceleration.

greg - Philadelphia & Reading / Reading

  • Member since
    February, 2002
  • From: Reading, PA
  • 24,643 posts
Posted by rrinker on Thursday, June 07, 2018 7:25 PM

 That's what a 2 point speed curve with CV2-6-5 does, adjusts the amount of change per speed step so you can configure a loco as a typical commuter engine which accelerates quickly but tapers off, or as a drag freight engine which accelerates slowly until the train is moving. Combined with momentum settings, the result is different acceleration rates at different speed steps.

Acceleration of the real thing isn't a straight line. Neither is your car. The forces opposing acceleration aren't all linear, so acceleration curves are not linear. Within the limits of adhesion, the locomotive can only deliver so much tractive effort to the rails. No matter what notch you're in. Even with modern anti-slip computer control - you will get maximum effort, but that may not be the total power the loco can deliver. Like traction control in cars - once you hit a certain horsepower, adding more won't make the car any faster unless you use more tiure for better traction. My car will chirp the tires and trigger traction control in 1st, 2nd, and 3rd. Maybe 4th but I'm not brave/crazy enough to try that on public streets. But that tells me, all else being equal, if I had 100 more HP, my car wouldn't accelerate any faster, I'm already at the adhesion limit for the tire size and compound I have.

 It's more complicated with a locomotive but similar in concept. Steel wheels on steel rails has a much lower coefficient of friction, which works FOR the train in terms of how much force is required to accelerate the cars, but also works AGAINST the train when it comes to how much power can be applied by the loco before the wheels slip. And since everything is so massive, inertia plays a much bigger role, both in the prime mover and the overall system. 

 IF the loco could apply 100% power and not slip, theoretically it would always accelerate fastest in notch 8. But if the limit of adhesion is reached in notch 5, going higher isn't going to do much. It does change the final speed (assuming no control changes are made). If the throttle isn't changed, the trainw ill continue to accelerate until the force applied by the loco is balanced by the drag of the train. If you had an infinitely long perfectly level track, that would be the case, but since even the best track has both up and down grades, this changes the forces acting on the train and so leaving the loco in one notch for the entire run doesn't work. Not considering that even on the perfectly level track, there's not much likelihood of any given notch balancing out a random train at the proscribed speed limit. 

                           --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July, 2009
  • From: somerset, nj
  • 2,038 posts
Posted by gregc on Friday, June 08, 2018 5:24 AM

hadn't thought of that -- that CV setting for seemingly unrelated features can be combined to realistically model the behavior of a locomotive

if i understand correctly, since DCC momentum describes the time between speed steps, making the motor voltage larger betwee lower speed steps will make the loco accelerate more at lower speeds than at larger speeds.

but doesn't this contradict the desire to have more speed resolution at lower speeds?   now speed step one is 1mph and speed step 2 is 5 mph AND there's only a 1 mph difference between speed steps 27 and 28.

 

i understand the limitations of wheel slip.   But isn't this less of an issue for a short commuter train?   Rather than just set it in the notch to achieve the desired speed, couldn't the engineer step thru the notches and accelerate more slowly?

i don't think this is possible with DCC momentum even with the speed curve adjusted; acceleration is a constant

again, i think the acceleration should depend on the difference between the current speed and the throttle speed setting.   

greg - Philadelphia & Reading / Reading

  • Member since
    October, 2001
  • From: OH
  • 16,382 posts
Posted by BRAKIE on Friday, June 08, 2018 7:55 AM

This I've seen with my very own peepers. I watch two NS GE  ES44AC lift a 9,000' train from a stand still to track speed in 30 car lengths. You couldn't tell the train was stopped a few minutes before.

Larry

SSRy

Conductor

“Shut one’s eyes tight or open one’s arms wide, either way, one’s a fool.” Flemeth-the witch of the Wilds.
  • Member since
    April, 2001
  • From: Roanoke, VA
  • 1,626 posts
Posted by BigJim on Friday, June 08, 2018 10:58 PM

BRAKIE

This I've seen with my very own peepers. I watch two NS GE  ES44AC lift a 9,000' train from a stand still to track speed in 30 car lengths. You couldn't tell the train was stopped a few minutes before.

 


What was the:

Train tonnage -
Length of cars -
Track speed -

???

.

  • Member since
    October, 2001
  • From: OH
  • 16,382 posts
Posted by BRAKIE on Saturday, June 09, 2018 8:39 AM

Jim, Here's what I know.

Train tonnage -No idea since I was not privy to that information..

Length of cars -Typical mixed  consist of gons,Centerbeams,coil,boxcars,bulkheads and tank cars

Track speed -35 through town..The train was a Eastbound and had stopped at Rt.98 siding to meet a Westbound.

The engineer put the whip to 'em and was up to track speed in 30 car lengths.

 

Larry

SSRy

Conductor

“Shut one’s eyes tight or open one’s arms wide, either way, one’s a fool.” Flemeth-the witch of the Wilds.

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!