Throttle behavior GP38 and SD40

[wato]

So I have been reading about throttle control of diesel-electric locomotives because I'm not convinced that the behavior simulated in TSW ATM is realistic.

In TSW each notch seems to produce power only up to a certain speed where current on the amp-meter falls of sharply and no more torque is produced unless a higher notch is selected.

Something seems to be off with the simulation at this point because from what I've read, I would expect power to remain constant at a level selected by the throttle position.
That would mean, since power = torque x RPM, that torque falls off inversely proportional to the locomotive's speed, albeit gradually up to the loco's top speed.

I'd like to hear other opinions on this, especially from people with real world experience.

[StratPlayer62]

I'm wondering about this also because it certainly is different from the previous versions of TS. I'm wondering if TSW is more accurate to real life because I've always noticed that trains going by at speed seem to be at a much higher RPM or notch than what I would need in TS. I've never driven a Train so I can't say for sure, but from what I have observed watching a LOT of trains, TSW seems to be more accurate. I hope someone who actually knows will respond.

[Delfino]

Originally posted by wato:

That would mean, since power = torque x RPM, that torque falls off inversely proportional to the locomotive's speed, albeit gradually up to the loco's top speed.

It's true but since there is still a generator and traction motors between I'm not sure. It all depends on construction details. Is it DC generator and motors? Remember that DC motors have quite different characteristic. Top power is achieved at half of stall torque.

[wato]

The SD40 seems to have DC generators and traction motors (http://www.kirara.co.uk/info/SD40%20-%20Operator's%20Manual/).
From there I also took my assumption about the constant power output:

Originally posted by Operator's Manual:

8. At locomotive start the throttle controls electrical devices that provide rapid power response at a level consistent with smoothly controlled starting.

9. During heavy-drag low-speed operation, as well as at moderate and high operating speeds, a load regulator operates to maintain power output at the specific level called for by throttle position. This prevents the engine from being overloaded or underloaded.

You're right though, that the formula I gave above is not accurate for a DC motor.
Max power of the traction motor is not really relevant though, I think, because power is limited by other factors like the maximum current and voltage the DC motors allow through, as well as the prime mover's power output.

I found this formula that should better describe the power output generated by the DC motors (http://lancet.mit.edu/motors/motors3.html#powercurve):
P(w) = -(t / w_max) w^2 + tw, w = angular velocity, t = torque
It follows that t = P / w (1 - w / w_max), thus at constant power we have the inversely proportional dependency on the angular velocity w.

And we can also see that torque is produced until shortly before w_max, the no load speed of the motor (when no resistant torque is applied to the output shaft). This speed has to be inherently higher than the speed at the loco's top speed because it must be able to reach top speed with load applied to the motor output shafts.

[canedriver]

The game is not accurate to real life in representing tractive effort. Traction motor amps do not suddenly drop to 0 at any speed, especially so low for the lower notches eg: notch 1 should still give amps at 50mph (for example), although a rather low amount.

[wato]

Thanks. I think that this needs fixing because it really takes the fun out of handling the trains, e.g., keeping precise speeds, stretching the train before applying power, etc.

[SteakOplomb]

Well, in the game I think the throttle notches are a mix between "target speed" and "generator RPM".
Ie, at notch 1, the generator will provide a set amount of power which would be drawn by the traction motors until the loco reaches pre-determined speed, explaining why the amp meter would drop when reaching said speed.

I have no idea how this works IRL. Before actually playing the game, I thought the throttle positions were either a "target torque" for traction motors (generator RPM being automatically adjusted to the needs) or "generator power output", directly feeding power to the traction motors.

[wato]

I was wondering the same. According to the operator's manual I quoted above, throttle position controls power output, leading to the torque curve I tried to describe above.

[Kyrah Abattoir]

Not an expert but in physic, the maximum tractive effort (your torque) is a function of speed:

(power/speed - rolling resistance) = TE

If your rolling resistance is lets say 13 N/ton and your train weights a thousand tons, that's 13kN that have to consistantly be overcome as the train is moving (assuming flat ground, so gravity is irrelevant), once the speed is high enough for (TractiveEeffort - RollingRresistance) == 0, you will effectively hit a state of equilibrium.

[SteakOplomb]

Kyrah, I may be wrong, but since you will always have to counter the rolling resistance (not to mention drag, but let's put this aside), you will still have to put power into your traction motors to keep your speed constant, and so, will still have a power draw on your generator. Having 0 acceleration does not mean you don't have to apply any power whatsoever.

Anyways, that's an intersting read. I always thought power curves for DC motors were similar to those of combustion engines. Although I knew DC motors have instant max torque, I never thought torque would decrease with RPMs, and max power actually be achieved around 50% max RPM. I'm having a pretty hard time wrapping my head around that, but the more I think about it, the more it makes sense.

[wato]

Originally posted by Kyrah Abattoir:

Not an expert but in physic, the maximum tractive effort (your torque) is a function of speed:

(power/speed - rolling resistance) = TE

I agree.
Although by torque I actually meant the torque on the output shaft of the motor, not tractive effort because that is irrelevant to the throttle behavior.
However, the problem in the simulation that I'm trying to point out indeed becomes apparent in the equilibrium points that you mention, which happen to be hit at all the wrong speeds because torque falls off so sharply at a certain speed per notch.

[SteakOplomb]

... and there by reading the answer above, I'm noticing the subtle difference between "tractive effort" and "torque output". Sorry Kyrah, I didn't fully get your point :p

[wato]

Now that I thought about it a bit, I probably also wouldn't include the rolling and other resistances in the tractive effort.
E.g., imagine a tractive effort curve with superimposed resistance curves (a plot of force over speed). You then get the equilibrium points of zero acceleration at the intersections between resistance and the tractive effort curve for the selected throttle position, i.e., clearly the resistance is not included in the tractive effort part of the curve, which rather is the conversion of traction motor torque into the force exerted by the locomotive on the rails - if I'm not mistaken.

[Latosusi]

power control feels more like cruise control so that first level of throttle gives 10 mph and so on

[Rickenbacker]

I noticed the same, and it feels really odd. The trains I normally drive are MUCH simpler to drive, with just a single lever for throttle and brakes, but they sure don't work like this! A certain throttle application feeds power to the traction motors at a certain rate, so to maintain a certain speed I have to back off the throttle, or even go to idle, but there's really no way I can just set the throttle at a certain position and get a certain speed...

[Bescot]

Definately a bug in the throttle physics. Really obvious when you're climbing the 1% grade sections, Notch 4 gives 0 amps and loses you speed while notch 5 gives 300-400 amps, and makes the train accelerate like a rocket. The current should drop off very slowly at each notch but this, as mentioned acts like a cruise control. Its the worst bug in the game and is spoiling it for me.