General Electric Wheel Slip U25-U36

{I originally wrote and disseminated this document in mid-2005. At that time, I placed it on file in a number of locomotive forums. I have updated the information in this document and it appears here in this form only; consider this version to supersede others extant.}

Wheel slip control in the General Electric Universal Series.

When General Electric began demonstration of the U25B, it was already aware that the railroads would have concerns about adhesion, considering the high power per driving axle. In order to mitigate these concerns, and give the locomotive true drag capability, GE chose to employ a piece of equipment which had already independently been under test in the US for several years (both by Westinghouse Air Brake and by New York Air Brake) as standard equipment.

This device was known as the Slip Suppression Brake Valve, or SSBV. The principle of the SSBV was that wheel slip could be corrected not by reduction of power and application of sand, but by an extremely quick application and release of the independent brake on the locomotive wheels. This device had already been installed experimentally on EMD units on the Western Maryland and had shown an ability to allow increased tonnage ratings in drag service, due to its ability to arrest slip with the locomotive still producing full power. The idea was that any brief track perturbation or weight transfer between axles which caused the slip would be passed fairly quickly, and so simply identifying the slip condition and arresting it without reducing power would not allow speed to drop further, as it would if excitation were reduced, albeit temporarily.

GE also chose to employ a new speed sensing system on the U25B which could provide better monitoring of axle speeds, through individual direct measurement. Normally, wheel slip relays in locomotives were picked up by voltage imbalance between traction motor leads. The new Axle Alternator system used a journal-box-mounted alternator on each axle, whose output was then conditioned into a usable signal for detection of slip, and for transition control as well. This system, with theoretically more rapid detection of slip / incipient spin, coupled with the SSBV to stop it, was supposed to allow unusually high factors of adhesion. Tests with the U25B prototypes actually did show much better adhesion (over the normal 18-20%) in drag service.

With this system, there was a backup; if the slip continued for longer than three seconds, then sand would be applied, excitation reduced (through action of relay and energizing of the ORS solenoid on the governor, which moved the load regulator toward minimum field position,) and the warning light and buzzer sounded. Incidentally, this was not the original design; General Electric 751 and 752, when operating in 1960 as the experimental, prototype U25B locomotives had the slip suppression brake equipment but did NOT reduce main generator excitation upon detection of wheel slip {ref. GEJ-3807, Operating Instructions General Electric Model U25B 2500 HP Diesel-Electric Locomotive - for road no’s. 751-752, page 44.} These units only gave a light and buzzer indication of slip coupled with SSBV operation, without sanding or generator excitation reduction. This design didn’t make it to production units.

Early on, each railroad buying the U25B also bought it with the Slip Suppression equipment since it was standard. However, some roads began to complain of pinion slippage and worn wheels. In brief, the operation of the U25B in mountainous terrain did actually lead to still-unfavorable wheel slip conditions, even with Slip Suppression. In the cases where Automatic Power Matching was employed, this condition was made less severe, but as this system derated the locomotive when excessive traction motor field temperature was reached, the advantage of the high horsepower was lost. The problem continued, and worsened with the U28B, especially on the Pittsburgh & Lake Erie, where U28B units were often used in drag service. General Electric was already aware that refinement was necessary. (It also developed an optional Pinion Slip Alarm at the request of the Louisville & Nashville, although GE’s stated opinion on the subject was that properly applied traction motor pinions did not slip -- a statement condemning the maintenance performed by the railroads.)

In 1964, GE began using new control equipment in the U25B. A new Type FL7 Adhesion Loss Detection Panel replaced the original 17FM190 panel; a new Type FL10 Speed Sensing Panel replaced the original 17FM191 Automatic Transition Control Panel. It appears that the purpose of the new equipment was performance related, although details are not given; these changes are found only by comparison of a number of manuals. One alteration is the addition of circuitry in the FL7 panel that attempts to detect synchronous slip, not addressed with the earlier model, and which only operates upon a simultaneous high speed signal from all axles. This particular protective feature would be improved upon in later locomotives.

During late U25 production, GE had made available an optional All-Electric Wheel Slip System. This was basically the same as the original system without the SSBV, and included instant triggering of the slip relay. One GE representative noted, a few years later, that the actual air valve in the SSBV system was “like any other air component -- you have to maintain it properly or it isn’t going to work.” It seems clear that GE felt that the original idea was still good enough to retain, but also noted that some railroads had continued to buy the system on later units (U28, U30) and that some did not, and might have credited improper maintenance of the equipment with its operational shortcomings.

In the U30, GE kept the options of either Slip Suppression Wheel Slip Control, or else All-Electric Wheel Slip Control, but augmented both with new circuitry designed to detect and correct conditions not detected or corrected by the early U25/U28 systems. The new system could detect synchronous slip of all driving axles simultaneously at low locomotive speeds, and could detect simultaneous high-speed wheel spin of all axles as well. Two different circuits were used, one for each condition, while retaining the Axle Alternator detection system to monitor axle speeds.

With the new system, normally fitted without the SSBV, the ORS solenoid on the governor was not used in conjunction with the wheel slip detection to reduce excitation and thus power. Instead, the completely new excitation system (developed for use with alternator-rectifier transmission) employed inputs from the wheel slip circuit cards to immediately reduce excitation by action of reducing the output of the pulse-width modulator used to supply field to the exciter. This system was both more sensitive and much faster. (With the SSBV, only a slip condition longer than three seconds duration would trigger excitation reduction, sanding and operator warning; without, either in U25/U28 or the new U30 units, these latter actions were immediate.) In later U23 units with GT-581 (U23B) or GT-586 (U23C) generators, the new system was altered to, in fact, return to use of the ORS solenoid.

One of the features of the Universal Series locomotives from nearly the beginning was a system called Automatic Power Matching which reduced locomotive output under adverse operating conditions, which theoretically also had the advantage of reducing wheel slip. In the U25 and U28, this system was not triggered until an over temperature condition was detected in the traction motor shunt field windings (indirectly, by voltage drop.) This system pulsed the ORS solenoid on the governor to force the load regulator wiper arm to assume, on average, a lower field strength position; the pulses or cycles occurred at about 15 times per minute. The Power Matching Panel for the U25B was part number 17FM211C1, and our shop manual for the NYC’s U25B units instructs the setting of the resistor on this panel to correspond to a generator load of 700V / 2000A or about 1876 horsepower. The system remained operative until a timer circuit cleared, which ran seven minutes. Engine speed remained responsive to throttle setting allowing full ventilating air to the traction motors.

Statements by GE engineers at the time indicated that this Power Matching was as much to protect their warranty-covered traction motors on the U25B as it was to match the units to those with which GE units would operate. Regardless, this was the first system of its kind in the USA. This system was standard equipment on U25 and U28 units.

In the early U30 units for what appears to be several months of production, this system may have been retained. Mid-way through 1968, the locomotive manuals begin to describe the addition of a further static control panel in the electrical equipment cabinets: the 17FL24 Power Matching Panel. This panel was used to condition the reference signal for the aforementioned pulse-width modulator to effect a (constant, non-pulsed, non-timed) reduction in power at lower speeds when the locomotive was operating in series-parallel only. It appears by mention in the manuals, and in various statements by GE officials, that on the U30 the system was optional; on later U33B units it was standard, and optional on the U33C. U36B and U36C units followed the same pattern.

This card caused a limit of reference current with decreasing speed down to about 8 MPH in the U33B, below which the full normal range of current limit was available for starting. Above this speed, the reference current was limited essentially until the locomotive made transition to parallel. (Field shunting was not used on these locomotives, but transition was employed.) GE felt that no limit was necessary at speeds above that required for transition, and in fact, the temperature-triggered system on early units was also nullified at the time the locomotive made the transition from series-parallel to series-parallel shunted field (which for the U25B typically occurred at about 19 MPH with standard gears.) In the early system, if the time limit for power reduction (7 minutes) was not expired and the locomotive made forward transition to series-parallel with shunted field, and then back, the limit was still present; if the time had expired, and speed increased and then dropped, it would be necessary to have another over temperature condition to trigger the Power Matching again. In the later units, the system always operated (if fitted) in series-parallel. Both old and new systems could be adjusted, by rheostat, to finely tune the desired amount of power reduction and thus the actual minimum continuous speed.

One further feature included in the U33 (and added to the U30 after this) was the introduction of a rate of change limit on excitation. With this circuitry, there was a limit on how fast the excitation signal could increase, and was in effect at all times, and for any action (throttle manipulation, slip recovery, load regulator action.) One GE representative referred to this as “rounding off the notches,” meaning that no abrupt increase in excitation, and thus alternator output, was possible. This further improved (theoretically) the ability of the locomotives to hold the rail, especially the four-axle units.

As we can see, the systems fitted to the GE U-series evolved over time, and the changes did not necessarily exactly coincide with model changes either. At the time that the above improvements were introduced, one further addition was in the works as regards the wheel slip system, which first appears in manual GEJ-3868, an electrical equipment manual which is the first to cover models U23, U30 and U33. This addition is sometimes referred to as the Rate of Change system, and is also sometimes referred to (especially by GE engineers and representatives) as the Power Tie Circuit. These terms are proven synonymous by this manual.

Around the time that EMD introduced its new 645 line of locomotives, there had still been outside research going on not directly connected with the locomotive manufacturers. One such concept was the use of direct measuring coils, mounted on the traction motor leads, which could detect small but instantaneous changes in actual traction motor current. This system was developed and tested by outside companies and various railroads, but EMD essentially adopted it about one year after the new 645-powered locomotives were introduced, as its IDAC Wheel Slip Control System.

General Electric was still employing the same axle alternator units as primary detection as had been used since the U25. However, the challenges of maintaining adhesion with the U33B were proving daunting, and GE added a similar system in concept sometime around the middle of 1968 in its ROC or Power Tie system. This required the addition of yet another static panel, the FM262 ROC panel, and ROC slip sensing transformers. The power tie was connected between the midpoints of pairs of motors when in series-parallel; because of this location, with all axles rotating at the same speed, there should be either a tiny amount of current flowing through this cable, or else (usually) none. If one axle begins to slip, the amount of current drawn by the sets of motors will change relative to each other, and this would then cause current to flow in this power tie. The detection of slip by this indirect method was more sensitive than the axle-alternator detection. The normal wheel-slip system as previously described was still retained, but special circuitry was included to reduce the effect of the relay normally operated by the old wheel slip system when the ROC system relay was picked up. The new system also included the design consideration that rapid and frequent small slips would cause a proportional slight lowering of excitation (and an even more tapered power recovery) which would allow the unit to maintain the highest possible power in the given rail condition. In parallel transition, the ROC system was completely cut out.

This new system did not in any way require the use of axle alternators. Within several years, GE had developed it to the point that the old axle alternator system was deleted, along with its supporting circuitry, and all wheel slip detection performed by coils; this was the Current Measuring Reactor system, or CMR system, and while this system is normally associated with the later “Dash 7” units, one will note that a number of orders of late Universal Series locomotives completely lack axle alternators (except one for speed indicator/recorder drive.) This indicates application of this CMR system. Reportedly, this system could react in one thirtieth of one second.

This concludes our investigation of wheel slip control as applied to the General Electric Universal Series. GE entered the market with the advantage of high horsepower per axle, and in a sense bet its fortune on the concept of high haulage capacity combined with high power and nominally similar weight. The one Achilles heel of this concept was the ability of design engineers to cope rapidly enough with adhesion difficulties as the horsepower race accelerated between the builders, and we have seen that GE was more than mildly active concerning modifications to the U-series over time to address this problem. While many have faulted GE for locomotive quality issues in this early period, it is clear that GE was making every effort to correct the problems and still move ahead in power, allowing it to retain a competitive position -- which in the end helped to drive ALCO out of the business.

Source material for this document includes not only a number of GE locomotive maintenance manuals, but a series of lectures delivered to conventions of the Railway Fuel & Operating Officers' Association by GE engineers over a span of years covering models U25 through U33.