AC vs. DC Question

[Denver Dude]

In this video, there is a lot of wheel slip on these DC locomotives- wet rails, curve, grade, etc.

Some people say that AC units wouldn't have made any difference. I wonder, though. I was under the impression that they have better adhesion and are a little heavier. Are the wheel slip controls the same on AC and DC?

I would think they'd make a difference. Not true?

https://www.youtube.com/watch?v=mLg_BrqZW3o" onclick="window.open(this.href);return false;

[Passenger]

I am under the impression that the traction motors are all DC in any case, regardless of how the power comes into the locomotive.

Is this correct?

[Denver Dude]

I am under the impression that the traction motors are all DC in any case, regardless of how the power comes into the locomotive.

Is this correct?

I don't think so.

[Allen Hazen]

Potted history.
With available technology in most of the 20th C, d.c. traction motors were better than a.c. So diesel electric locomotives were developed with d.c. generators and d.c. motors. (At least in the U.S., General Electric is the key player: power electricals and control systems evolved from GE's pre-WWI railcars, with improved versions on the first successful diesel-electric switchers and EMC's gas-electric railcars in the 1920s. Alco diesels used almost exclusively GE electricals -- a few "Hi Hood" switchers in the 1930s got W'house -- until Alco left the market at the end of the 1960s. EMC bought electricals from GE and W'house: after it was taken over by General Motors in the 1930s it developed its own electricals, based on GE designs.)

Big d.c. generators are problematic, so when solid-state rectification became feasible in the 1960s, locomotive builders switched to using -- on all 3,000hp and up units, and increasingly as time went on for lower powered units -- a.c. generators with rectifiers to provide d.c. current to the traction motors: this change was made by EMD with its 40-series models (introduced 1966) and by GE with the Alco C630 (introduced late 1965) and on some GE U28B and U28C units in 1966.

Further developements of solid state electricals and controls able to provide variable frequency a.c. allowed the use of synchronous a.c. traction motors. The locomotive has an a.c. generator, producing a.c. at a frequency determined by engine r.p.m., with the current being rectified so as to feed d.c. to the invertors which then produce a.c. at a frequency determined by the desired locomotive speed to feed to the traction motors. This technology was introduced in Europe (British experimental in the 1960s, German prototype in the 1970s) and in the U.S., by both EMD and GE, in production models in 1993 and 1994. Initially EMD and GE offered both models with d.c. motors and models with a.c. motors, but the advantages of a.c. have led to increasing use: I think all high horsepoer units in the past coulple of years have had a.c. motors.

[timz]

You meant to type "asynchronous a.c. traction motors", didn't you?

[Nasadowsk]

You meant to type "asynchronous a.c. traction motors", didn't you?

Both types were used. Asynchronous is all that anyone does now, but synchronous motors were used by Alstom and maybe a few others. This was done because it allowed earlier SCR technology to be used. With GTO and IGBTs, it's not needed anymore.

[Denver Dude]

Okay, but would AC locomotives have made a difference in the video?

[Allen Hazen]

Timz and Nasadowsk--
Apologies: my misuse of the terminology. What I meant to convey was that modern AC transmission diesels use traction motors in which the speed (rpm) of the motors (and so, since the motors are geared to the axles, of the locomotive) is determined by the frequency of the current (so the electrical system has to provide variable frequency AC). The contrast is with the sort of AC motors used on, for instance, the Pennsylvania Railroad's electric fleet (other than the experimental FF-1), where the motors always got current at the frequency (25hz) of the overhead wire.

Denver Dude--
I don't know. I think modern anti-wheelslip control on locomotives with DC motors (e.g. EMD 70 series, GE Dash-8 and Dash-9) may be almost as good stopping wheel-slip (at the price of a good deal of complexity): so I don't know whether an AC locomotive can exert more tractive effort than a comparable DC locomotive… for a few minutes. The big advantage of AC, I believe, is in continued operation under maximum tractive effort conditions: locomotives with AC motors can KEEP ON pulling at low speeds where the DC motors on an otherwise similar locomotive with DC motors would overheat.

[Denver Dude]

Thanks, Allen. If they (AC and DC) units both have the same traction control systems, and their weight is the same, adhesion would be the same, I guess.

[Allen Hazen]

The traction control systems can't be the same, since in detail they depend on the characteristics of the motors used, etc. But they may be almost equivalent in effect.
With DC motors, traction control (wheel-slip control) starts with detection of the beginning of slip. This can be done in any of a variety of electronic ways. (Will Davis, on his "Railroad Locomotives" blog, posted a history of the different systems GE used in the U-series period. I believe that in the "high traction era" that started with EMD's 50-series and GE's later Dash-7 series, GE went back to monitoring the speed of individual axles, not by means of axle generators mounted on the hubs, as on the early U-boats, but by means of "speedometer" circuits built integrally into the later versions -- 752AG and 752AH -- of the 752 traction motor.) Then sand is applied, and/or power temporarily reduced, to stop the slip. All occurring, in later systems, in a fraction of a second.
I'm not sure what happens with AC motors.

[Pneudyne]

The kind of AC motor used in the locomotives at interest, namely the polyphase induction type, does have an inherent advantage in that it has a natural shunt characteristic, that is a very steep torque vs. rotational speed curve. Series-wound DC motors of course have a series characteristic with a much shallower torque vs. speed curve. So with the AC motor, even a very small increase in rotational speed will result in a marked drop in torque, which means it has inherent slip–reducing properties. Whether this potential advantage translates into an actual advantage I do not know. But I suspect that at the adhesion limit, it could well make some difference (in favour of the AC motor).


Cheers,

[Jay Potter]

Having AC-traction locomotives on the train in the video would have made a difference because (1) adhesion is a function of wheel creep; and (2) wheel creep is regulated more exactly on AC-traction locomotives than on DC-traction locomotives because the AC regulation is performed by inverters. However similarly to Allen's comment, an additional -- oftentimes more significant -- advantage of AC traction is the ability of those locomotives to operate at low speeds for longer periods of time than DC locomotives can operate. If the grade in the video were longer than it apparently is, I suspect that the thermal protective circuits in the DC-traction locomotives would have eventually begun to derate the locomotives and caused the train to stall.

[Allen Hazen]

Thanks, Pneudyne and Jay Potter, for seeing I was out of my depth!
Pneudyne-- I had a feeling that the AC motors would have an inherent slip-correcting behavior such as you describe. To get the same effect with DC motors one would have to (i) detect the incipient slip VERY quickly and (ii) reduce power VERY quickly, and I don't know whether the latest control systems for DC-motored locomotives are good enough to mimic what an AC-motored unit can do.
Jay Potter-- EMD boasted that its control system for units with DC motors, introduced on the 50 series, allowed controlled creep to maximize tractive effort. My guess is that it is a lot EASIER to have controlled creep with AC because of the properties Pneudyne mentions.

(Definition. "Zero creep" is what you would get on a cog railway: the wheels turn EXACTLY once as the locomotive advances by the circumference of the wheel, so the point on the wheel rim in contact with the rail is stationary. "Creep" involves the wheel turning JUST A BIT faster than this. Maximum tractive effort comes with a certain, small, amount of creep, but this has to be controlled VERY precisely: if the wheel starts to turn too much faster than the zero-creep speed, adhesion is lost, and you have wheel slip.)

[Jay Potter]

Allen, as Pneudyne indicated an AC traction motor of a given size is, by design, stronger (i.e. capable of producing more rotational torque) than a DC traction motor of the same size. However the amount of tractive effort that either type of traction motor actually produces at any given time is a function of how exactly its creep is regulated at that time. As I indicated previously, an AC traction locomotive can regulate creep more exactly than a DC traction locomotive can because the AC regulation occurs at the inverter level, not at the traction alternator level. This enables an AC-traction locomotive to maintain a different creep rate for each truck (on most EMD units) or for each axle (on GE units). This is significant because optimal creep rate is a function of rail condition. If for a given rail condition (i.e. light moisture, rain, grease, etc.) the creep rate is either too high or too low, adhesion will be reduced. Furthermore, rail condition varies from wheel set to wheel set. For instance, the optimal creep rate for a number six axle is almost always going to be different than the optimal rate for the number one axle. This is because the sixth wheel set encounters rail that has already been burnished by the passage of five other wheel sets. Likewise, the trailing unit in a two-unit consist can be expected to produce more tractive effort than the lead unit can produce.

[Allen Hazen]

Jay Potter--
Thank you for your reply!
Trying to think through it…
(i) The greater torque of the AC motor design shouldn't effect the low-speed wheelslip issue: both DC and AC motors are capable of enough torque to slip the drivers at relevant speeds under realistic conditions.
(ii) Certainly the AC motors, with AC frequency produced by separate investors for each truck (early EMD AC locomotives) or each motor (GE locomotives, and I gather EMD has -- belatedly -- seen the light here had used separate alternators for each motor on some of their more recent production) allows more precise control of creep than conventional control systems for DCmotors.
(iii) But -- this is maybe a theoretical issue, not relevant toAmerican railroad practice -- separate control for each motor is possible with DC, and (with good enough micro-processor control and very precise monitoring of individual axle speeds) MIGHT allow a locomotive with DC motors to "mimic" the performance of an AC locomotive over a short period of time. This is done by having the motors "separately excited," as on the (Brush Electrical built) Class 60 locomotives of British Rail, introduced in 1989. From the Wikipedia article on "British Rail Class 60":
"Each motor has a separate microprocessor-controlled power supply (SEPEX in Brush's designation - from "Separately Excited"), a system that was first tried on the Class 58. One feature of this system is that if one set of wheels/axle/motor starts to wheelslip their speed can be reduced without affecting the other motors."
(iv) History of locomotive purchases by North American railroads in the past couple of decades, though, seems to show that using AC motors is preferable to trying to get similar effects with DC and fancy stuff!