The short answer is there are 8 notches because the governor has 3 solenoids it uses to set engine speed from the throttle and that’s how many combinations you can make from 3. 2^3 = 8
It actually has a fourth solenoid that is used to signal the govenor to shut down the engine, and that solenoid gets exercised in the speed schedule so it won’t get stuck in place.
The Woodward governor - which is the standard on all locomotives, is a flyball governor that controls engine speed. Each solenoid acts against a lopsided triangular plate. The plate pushes against a spring that works against the flyball mechanism. The harder the spring pushes, the faster the flyballs have to spin to balance.
The plate’s geometry gives each solenoid a different length from the spring such that solendoid A is worth 1 increment of engine speed, B is worth 4 and C is worth 2. (D, the shutdown solenoid, is worth -2)
So, and engine speed schedule from the solenoid’s perspective looks like this:
Low idle AD (added circa 1980 - so there are actually 9 speed settings, and on units with low, low idle, 10)
Idle and Notch 1 no solenoids energized
Notch 2 A
Notch 3 C
Notch 4 AC
Notch 5 BCD (would be B if you didn’t have to “exercise” D)
Notch 6 ABCD (would be AB…"…)
Notch 7 BC
Notch 8 ABC
Each of these solenoids is controlled by throttle switches on a cam on the throttle and is assigned a pin on the 27 pin trainline.
Eight became the standard and fit the 27 pin MU scheme. There are all sorts of games that have been played with this speed schedul
In the very early days, there were some strange variants. Some Baldwins used pneumatic controls for the governor and some locomotives (switchers, mostly) had straight mechanical throttles where the control stand lever moved against the flyball governor spring directly giving “infinite” notches. This stuff is all before my time, so I don’t know the details.
This guy has a bunch of videos of load tests on various locomotives. The computer screens on newer units show the engine rpm and horsepower output in each throttle notch.
The SD70M-2 putting out over 4600 HP at 950 rpm impressed me, though that may have been total brake horsepower instead of traction output (or perhaps a bad sensor).
I was given a copy of the paper when I started my career as a railway mechanical engineer. I’ve read it and re-read it many times.
One thing that struck me much later was the comment that the 201A was limited by requirements from the US Navy. When I checked, the USN used the 201A in only one submarine, and later relied on Cleveland built engines that shared some 567 features. But the funding from the USN was probably vital in getting the engine built in a reasonable time…
The other things that stood out were the attention to detail in all areas, and the ability to recognise unexpected results (such as the good performance of cast iron pistons using only basic grey cast iron rather than the proposed special alloy.)
Moreover, he implicitly stresses watch for what the parts are telling you… and learn from it and from ‘all that that implies’ for your future work.
It occurs to me that Mr. Goding, or people he knows how to contact, would be able to write a comparable account of the H-block 265 engine, including a reasonably conclusive account of ultrasonic concentration in thin-wall cast blocks at high power and any remediation strategies GM tried. As I tried to point out during the British steam LSR challenge – unsuccessfully – documenting even blind alleys or mistaken designs can be highly valuable, to advancing the state of the art and its understanding as well as staying ahead of the edge of history.
As the HP increased, EMD eventually had to go to visous crankshaft dampers. The spring-pack ones didn’t have enough oomph. Engines speed schedule had to be worked around the “peaks”, for sure, but the governor design gives lots of leeway for setting the speed schedule.
That 567 paper is very interesting! The 567C engine was pretty bullet-proof and lots of the older A and B engines had some of the improvement incorporated when rebuilt. The GP9, with that 567C engine and battery field excitation might have been the most dirt simple, reliable locomotive ever made.
When a current GE is in Run 8, the engine isn’t always at maximum RPM? At low locomotive speed, prime mover RPMs might drop while the locomotive continues to produce maximum TE in Run 8?
That is a good question. In the days before FADEC EFI the governed maximum speed is what would be commanded in Run 8 (with the governor of course apportioning fuel to keep the engine at that rpm regardless of load). It would make sense for a computer to drop the rpm as noted for physical deloading, both for the significant aggregate fuel saving and reduction of wear and tear.
I suggest that the immediate path to a practical answer involves a BatLight call to Randy Stahl…
When operating in notch 8 for extended periods of time and additional traction motor cooling is needed, GE units will increase rpm to 1050, with no increase in HP.
And yes, if a GE loco is in the lead with an EMD trailing, under the right circumstances, when the throttle is increased the EMD will bump the GE telling it to quit picking its nose and start loading!!!
By the time the U25s with their half notches got to Conrail, all of that nonsense had been removed but the throttle still moved in half notch increments. There were still engineers who had all sorts of theories on how to use the half notches, even though they didn’t work anymore!
The 1-5-8 speed schedule. Trying to get the darned things to load a bit quicker. Notch 5 engine speed in notches 2-5, Notch 8 after that. A good chunk of PC U23Bs had this silliness. It was superceded with “skip 3, double 6”.
Only if the buttons were originally working but were then quietly disconnected as being an annoyance to vehicle operations…
Remember that these would result in weird loading mismatch with ordinary 8-notch units if more aggressive excitation is what the half-notches control and the engineer isn’t careful to count if he is engaging a half or whole notch in a one-notch throttle lever transition. If I remember correctly the engine speed changes were artificially dashpotted to avoid smoke shows due to turbo lag, and the purely loading transitions of the intermediate notches might not have had a comparable long ‘time constant’ – as a feature, of course.
The speed changes were fast. The loading was slow. If you wipe the throttle on a GE Dash 7 or 8, it’s 80 seconds to full load, and more than half of it occurs in the last 20 seconds.
If you have a train all stretched out and you are in a position where you can’t possibly generate enough TE to get a knuckle…wiping the throttle will just get you where you’re going faster
An EMD (pre-AC) - wiping the throttle would get you to full load inside of 25 seconds, nearly linearly.
GE CHEC had “three slope curve”.
God bless you if you had to start a train on the grade with a B36 leading and GP40s trailing…and you needed all three to get rolling.