GE Export Diesels

My guess is that customer preference, coupled with engineering conservatism, was the reason why equalized three-axle trucks were preferred under some GE export locomotives for a few years after the unequalized type became available.

The civil engineers would have been accustomed to having equalizing gear on their steam locomotives, and some may have had (unsatisfactory) experiences of earlier British export steam locomotives that, per established domestic practice, were not fitted with equalizing gear. They might though have accepted unequalized two-axle trucks, as these had a degree of inherent equalization. Add to that that many export locomotives, designed with lowish axle loadings, had a relatively high power-to-adhesive weight ratio, and so were probably thought to be more in need of good equalizing gear to help minimize wheel-slip caused by axle unloading. To some extent the locomotive makers reinforced that notion; for example a feature of the GSC 1-C truck used under the South African U18C1 and U20C1 fleets was the extra-careful attention to equalization.

EMD used unequalized three-axle bogies on its export designs from the start, and their early reputation may not have helped the general spread of the type. One example;: New South Wales regarded its 42-class Clyde-GMs as having inferior riding and tracking characteristics at higher speeds as compared with its 44-class Goodwin-Alcos, which had Commonwealth double swing-bolster equalized trucks.

Still, it seems that eventually GE was able to convince its export customers that a well-designed unequalized three-axle truck, probably coupled with improved wheel-slip control, was an effective solution and also one that was simpler than an equalized truck and so lower in both first and maintenance costs.

What kind of trucks were fitted to the early Mozambique U20Cs I don’t know. But these were ordered at the same time as the Rhodesian fleet, and in both cases the operators were persuaded by the South African experience in which GE was seen have an excellent product, and one that was very well supported. So it seems likely that those Mozambique long-nose U20Cs had equalized trucks, like their Rhodesian counterparts.

Curiously, South Africa’s first EMD locomotives, GL26Cs in 1966-67, did not have the regular EMD trucks, but rather an outside-equalized design that did look somewhat like a longer wheelbase version of the GE long-nose U20C truck. What kind of bolster springing was fitted I don’t know. It would seem that to get its foot in the door in South Africa, EMD not only had to slim and trim its G16 design to the point where it was close to being a dead ringer for the U20C, but also fit similar trucks.

Cheers,

Another unusual GE export model was the U11B built for Costa Rica in the late 1970s.

This was essentially a lengthened U10B with a Caterpillar D399 16-cylinder engine in place of the standard D398 12-cylinder unit.
Note that unusually for a GE locomotive, the main generator was not a single-bearing machine close-coupled to the engine, but rather was a separate two-bearing unit mated to the engine via a flexible coupling. The Cat D399 does not appear to have been used very much in locomotive applications.

Cheers,

The Costa Rica U10B and U11B had KC108M2 master controllers:
Their Caterpillar engines had electropneumatic (EP) throttle operators, working with Caterpillar governors, as was normal on the small GE Universals. The four-solenoid EP unit might well have been a descendant of the earlier GE 17MK3 throttle operator that appears to date back to the late 1930s. Cat did offer the option of the Woodward PG governor, and this was used by some other locomotive builders.
These Costa Rica locomotives also had dynamic brake setup switches, whose correct position depended both upon the number of units in an MU consist and the position of each unit in that consist.
This was certainly unusual in a potential-wire controlled dynamic braking system, and I can only speculate as to why it was required. These smaller GE Universals had 1924 Lemp inherent characteristic load control, so did not have a load regulator rheostat. That means that dynamic braking could not be controlled by a micropositioner in conjunction with the load regulator rheostat. Micropositioner relays imposed but a very small loading on the potential wire, so within reason, the number of locomotives in an MU consist had but small effect on the braking curve. Absent this option, alternatively, the battery fields of all exciters in the consist could have been fed by from dynamic braking control rheostat/potentiometer in the leading unit, which would then be looking inot a non-negligible load. And given that the load on that rheostat/potentiometer would be proportional to the number of units in the consist, there may have been a switched resistor network that compensated for the effect that variations in unit count would have on the braking curve, that is the braking effort in relation to braking handle position. If so, this network had to be setup properly in each unit in a consist, and not just in the leading unit.

Cheers,

As mentioned upthread, during the 1950s, GE appeared to be less involved with licence-built export models than were Alco, Baldwin and GE. Already covered in this thread are the Alco 244-powered shovel-noses built by Baume et Marpent, Belgium, for Matadi-Leopoldville, Congo. Essentially, these were repeats of a batch built earlier by GE itself. Baume et Marpent was a GE licensee, as was SEM (Société Electro-Mécanique), who supplied some of the electrical equipment. These were supplied more-or-less at the same time as Cockerill supplied a batch of Baldwin licence-built AS616E models to the same railroad.

Three groups of GE locomotives built in Australia during the 1950s look as if they were licence-builds, but their situations may have been a bit more complicated. They were the New South Wales (NSWGR) 43 class, and what became the Queensland Railways (QR) 1170 and 1150 classes. The last mentioned was a repeat order, the first batch having been built by GE Erie.

In each of those cases the order was placed upon GE’s Australian subsidiary, Australian GE (AGE), who was thus the prime contractor. For the NSWGR 43 and QR 1150, AGE subcontracted building of the mechanical parts and assembly to Australian company Goninan, who had previously been subcontracted to build GE industrial switchers. Although Goninan later became a GE licensee, as best I can determine it was not such at the time these orders were placed. Rather it was simply a sub-contractor. For the QR 1170 class, the subcontractor was Walkers of Queensland, who had previously built steam locomotives for QR. Evidently QR had a preference for in-state construction. AGE was to build some of the electrical equipment for all of these in its own plant.

Before any of these locomotives could be built and delivered, in late 1955 AGE became Australian Electrical Industries (AEI), after Associated Electrical Industries (also an AEI), UK bought GE’s stake, as was noted in the Alco Export Locomotives thread; viewtopic.php?f=4&t=160339" onclick="window.open(this.href);return false;. Thus AEI (Australia) became the supplier of record for the three locomotive classes. Also as mentioned in the Alco thread, AEI (UK), particularly its BTH arm, had an established relationship with GE that allowed apparently easy exchange of technology. I have never seen anything to indicate that AEI was a GE licensee, although there might well have been a memorandum of understanding, or something similar.

AEI (Australia) issued the operating manual for the QR 1170 class. It did not include any mention of GE except in respect of the locomotive model designation, which included “4GE 756”.

These locomotives were in fact modified GE 70-tonners, with articulated A1A trucks and GE756 motors. So in part they derived from the C+C variant developed in the late 1940s, initially for Bolivia, I think, but perhaps better known for their use in Brasil and Chile.

More to follow.


Cheers,

Returning to the Goninan case, the 1961 August issue of the Australian journal “Railway Transportation” (RT) carried an article entitled “G-E Universal Loco Range Now Available for Local Manufacture”. Here is the first page:
The highlighted part indicates that the agreement between Goninan and GE was fairly recent, and that previously it had acted as a subcontractor to AEI (Australia).

It was perhaps unfortunate that this article, which referred to the U6B, U9B/C, U12B/C and U18C models from the initial Universal range (with mention of the U25B), appeared in print at about the same time that GE introduced its uprated or redesignated Universal range, U5B et seq., as shown in its 1961 August catalogue “Built-In Power”.

Emphasizing the anachronism, in the same issue of RT, Caterpillar Australia was advertising its then-new D398 engine, as used in the GE U8B.
As it turned out, Goninan did not obtain any GE line-service locomotive business in Australia until well into the 1970s, although it did build GE industrial locomotives in that period. By the time that Goninan was in the market as a primary supplier, the Australian state railroads had already committed to other builders; NSWGR to Goodwin-Alco, with token business to Clyde-GM; VR to Clyde-GM; SAR to Goodwin-Alco with English Electric (EE) as minority supplier; TGR to EE; QR to Clyde-GM and EE; and WAGR also to Clyde-GM and EE.

GE had gotten off to a good start with QR, supplying both mainline locomotives (alongside EE) and branchline locomotives (on a solus basis). But perhaps its supply arrangements hampered any capitalization on this. Whilst AEI (Australia) had taken over contracts originally held by AGE, it is not so clear that it (AEI) was free to bid GE locomotive designs for new business. Also, the arrangement whereby Walkers undertook in-state subcontract building of the 1170 class branchline locomotives was apparently did not proceed all that smoothly.

EE consolidated its claim on a good share of the QR business by setting up its Australian manufacturing operation in Queensland. Clyde-GM had gotten entry by offering its G12 model (with modifications to suit QR) on a trial basis, and shortly afterwards, building the six-motor GR12 (aka G12C) version to better suit QR’s needs. In the early 1960s it subcontracted building and assembly to Commonwealth Engineering (Comeng) Queensland to stay aligned with a major QR preference.

GE’s start with NSWGR, in the form of the 43 class, was a small one, and a locomotive based upon the Alco 12-244H engine was in some ways a dead-end. Neither was NSWGR completely happy with Clyde-GM 42 class (A7 model). The Goodwin-Alco DL500 (its 44 class) was the answer to its mainline needs, and from 1959 that model was available with AEI (at the time lower cost) as well as GE electrical equipment, NSWGR having had some difficulties with its GE761 motors.

Goninan is know to have bid on the NSWGR tender that resulted in its 45 class. This was in the later part of 1960. It offered not only the U18C, as might have been expected, but also, and rather strangely, the U13C. (This is evidence, inter alia, that the U12-to-U13C transition took place ahead of the general GE Universal range update in 1961.) Goodwin-Alco got the business with its DL541 model, the DL543 being still in the future. Amway, from the fact that it bid, we may infer that Goninan was a GE licensee empowered to bid for locomotive business by 1960.

I have not seen any evidence that any GE model was a contender in the earlier (1958) NSWGR tender that resulted in the start of its large 48 class fleet of the Goodwin-Alco DL531 model. On the other hand, there is no hard evidence that there was definitely not a GE model offered. (The U9C would have been a reasonable fit to the requirement.) Still, it does look as if in 1958, Goninan was not yet in a position to bid a GE model.

Apparently Goninan did bid, unsuccessfully, in a 1962 NSWGR tender for additional branchline locomotives, offering a locomotive described as having a Caterpillar engine with GE electrical equipment. (A C-C derivative of the then-new UM10B model (later redesignated U10B) could have been a fit, although the long-frame version of the small Universal was at that time still in the future.) Ironically Goninan was the successful bidder for what turned out to be the final tranche of NSWGR branchline locomotives, but here there was no GE connection. Rather Goninan teamed up with Hitachi to build (as the NSWGR 47 class) a model with a Caterpillar D399 engine, a combination that turned out to be rather unsuccessful.


Cheers,

Pneudyne--
Thank you for that nice summary of the Australian story! … I think I remember reading or being told that the U18C pitched to NSWGR would have been higher in price than the Alco-derivatives ultimately bought.
---
With the outline of the GE end-cab in the Caterpillar ad… The traction generator has a shape (cylinder, with a coned end) familiar from other GE direct current traction generators, but the placement of the auxiliary generator, sticking up (and so, I assume, run by some sort of angled gearing) is unusual: typically the auxiliary generators on GE-equipped domestic locomotives (both GE and Alco) are mounted on the end of the traction generator (on the flat end of the frustum of the cone). Given the drawing, one might think this placement was adopted for space reasons: wanting to fit everything onto a very short platform.

Hello Allen:

The GE small Universals all had an auxiliary generator and exciter on a common shaft, mounted on top and usually to one side of the main generator, and belt-driven from the end of the latter. This arrangement had previously been used on the 70-ton and 50/52-54-ton models. It is illustrated upthread in respect of the QR 1170 class (a 70-ton derivative) and the Costa Rica U11B.

It is shown fairly clearly in this picture of the powerplant for the Sorocabana (Brasil) metre-gauge C+C variant of the 70-tonner.
Another view is provided in this picture of the powerplant for a Robla (Spain) metre-gauge U10B. Not really visible, but these also had a 5 kW, 110-volt head-end lighting generator belt-driven from the free end of the engine.
The larger export Universals (including the U9B/C) had the familiar arrangement of the triplex gear drive at the end of the main generator to whose three outputs were attached respectively the auxiliary generator, the exciter, and a mechanical drive to a traction motor blower.

The belt-driven auxiliary generator-exciter combination was also used on some road locomotives, such as the QR 1150, WP&Y shovel-nose, and the RSR Thailand UM12C. Alco also used it on its DL531 and DL535 export models. Those combination machines had 5GMG-numbers, and show up as such in the GE traction motor list, which I think that you might have.

Usually the exciter was of the split-pole type, intended to be part of a 1924 Lemp inherent characteristic (“engine stalling”) load control scheme, although the WP&Y shovel-noses might have had a 1914 Lemp scheme, with load regulator rheostat. (Apparently the Woodward PG governor was the standard fitting on the Alco 6-251 engine from the start.)

Other pre-Universal GE road locomotive exports mostly had the triplex auxiliary drive, although in some cases with the third output driving a second auxiliary generator that powered motor-driven traction motor blowers and in some cases a motor-driven radiator fan. But the IRCA Guatemala C+B+C road switcher (and I imagine its Columbian clone) had its auxiliary generator and exciter mounted independently atop the main generator, each with its own belt-drive from the end of the main generator. This arrangement was I think copied from a concurrent Alco-GE domestic road-switcher design.


Cheers,

I should add that in the case of the QR 1150, the belt-driven auxiliary generator-exciter combination was mounted on the floor behind the main generator. In this case the space-saving aspect, as you say important in the smaller locomotives, would not have been an issue, as this locomotive was about the same length as the later export U18C (original 12-cylinder version), which it somewhat presaged. The U18C had sufficient internal room not only for the triplex auxiliary drive arrangement, but also for a train heating boiler.

Cheers,

This page from the 1956 GE brochure "Building Locomotives for the World's Railways" has in its lower picture a pair of export road-switchers, both the Columbian C+B-C design with Alco 12-244 engine and Amplidyne control, and the QR 1150 class C-C with Cooper-Bessemer FVL-12T engine and simple inherent characteristic control.
Cheers,

Pneudyne--
Thank you for reply!(*)
Do you know the model numbers of the generators used in various early GE "export" (incl. built by licensee) locomotives? I would expect anything built by GE using the Alco 12-244 engine to have electricals similar to domestic Alcos with that engine (GT564 generator up to early1950, GT581 later (with GT566 in the RSD-5 with its six 752 traction motors)). I ***think*** (though I don't think I could now find documentation…) that in the mid-1950s, for the 1954 4-unit test set shown in Erie colours, and I would assume also for the U12 line, GE decided to use the same generator -- GT581 -- with the 1200hp 8-cylinder C-B engine as used with the 1800hp 12-cylinder C-B. In which case, space permitting, the simplest course would have been to use the same mounting of the auxiliary and exertation generators as used on domestic Alcos with that generator.

(*) and for continuing to post scans of period articles and ads! They are much appreciated.

(For "exertation," read "excitation". Senior moment.)

Well, there seem to have been at least a couple of applications of GT581 with belt-driven auxiliary and axciter: you mention above that
"The belt-driven auxiliary generator-exciter combination was also used on ... the RSR Thailand UM12C"
which I am assuming would have used theGT581. And, back on the first page of this string, you report that the Alco Dl535 (six cylinder 251 engine, but rated at 1200hp) had the GT581 generator with belt driven companions.
And, once the engineering has been done, I suppose the choice of belt driven or "conventionally" mounted auxiliary and exciter for any given application would be open...

The version of the GT581 used in the Alco DL535 was the GT581E1, E2 This description is from the GE list:

E1 Similar to type GT581 except modified for belt-driven GMG167 exciter auxiliary generator. Gearing and gear box omitted. Magnet frame drilling for mounting on engine different than other GT581 generators. Armature and bearings are not interchangeable as a unit with any other form of GT581 except form E2. Superceded by form E2. Alco for India.

E2 Same as model GT581E1 except different covers and frame fabrication. Change from 10 frame arms to eight to improve access-ability to bottom brushholders. Supercedes form GT581E1.


The reference “Alco for India” probably refers to the fact that the DL535 was first built for India and may have been designed around an Indian requirement. This facet can be carried over to the Alco Export thread for more comment.

I’ll need to compile a list for the pre-Universal exports. As I recall, my data is incomplete, but I want to recheck it before I write anything.


Cheers,

Re the main generator fitments to the GE pre-Universal export road locomotives.

Argentina and Chile CB-engined metre-gauge A1A-A1A shovel-nose units: GT577 main generators with gear-driven auxiliaries and Amplidyne excitation.

Guatemala (two batches) and Colombia Alco-engined three foot-gauge C+B+C road switchers: GT564 main generators with belt-driven auxiliaries and Amplidyne excitation. The main generator type is inferred from the fact that the fact that twin belt drives were used and also from the fact that the engine-generator combination was said to be the same as that used for the contemporary Alco-GE domestic road-switcher.

Congo Matadi-Leopoldville Alco-engined Cape-gauge C-C shovel-nose units: GT564 main generators with gear-driven auxiliaries and Amplidyne excitation. The same for the repeat order built by Belgian licensee Baume et Marpent.

Uruguay Alco-engined standard-gauge C-C double-ended shovel-nose units: Unknown, but with gear-driven auxiliaries and Amplidyne excitation. Either GT564 or GT581. The SB-gauge motors (GE731) used in this model might have indicated the need for a higher capacity main generator than was acceptable in the CM-gauge units with GE756 motors.

Argentina Alco-engined broad-gauge C-C shovel nose units: Unknown, but with gear-driven auxiliaries and Amplidyne excitation. Either GT564 or GT581. Same comments as for the Uruguay case.

Argentina Alco-engined metre-gauge C-C shovel nose units: Unknown but with gear-driven auxiliaries and Amplidyne excitation. Either GT564 or GT581. GT564 seems more likely given that these were essentially copies of the Congo units. Plus the fact that the GT564 was used in the later metre-gauge Brasilian 110-ton road-switchers.

Australia QR CB-engined Cape gauge C-C road-switchers: GT567 main generators with belt-driven auxiliaries and inherent characteristic excitation. The same for the repeat batch supplied by AEI (Australia) (née AGE) and built by Goninan.

Brasil Alco-engined metre-gauge C-C 110-ton road-switchers: GT564 main generator with gear-driven auxiliaries and Amplidyne excitation. (These appear to have been the first locomotives fitted with GE761 motors.)

Indonesia Alco-engined Cape gauge C-2-C double-ended shovel-nose units: Unknown, but with but with gear-driven auxiliaries and Amplidyne excitation. Either GT564 or GT581. GT564 would be consistent with the other GE CM-gauge designs of the era.

Chile Alco-engined broad-gauge C-C shovel-nose units (two batches): GT581 main generator with Amplidyne excitation.

Australia NSWGR Alco-engined standard-gauge C-C streamlined cab units, supplied by AEI (née AGE) and built by Goninan: GT581 main generator with gear-driven auxiliaries and Amplidyne excitation. (These had GE761 motors, though.)

Philippines Manila RR CB-engined C-C streamlined cab units and modified road-switchers: GT584 main generator with belt-driven auxiliaries. Excitation system unknown, but either inherent characteristic or Lemp 1914 with load regulator rheostat. The main generator type is inferred from the GE motor and generator list, in which the GT584 was originally said to have been suitable for 1200 hp, and also in which the GT584E1 was said to be a special for the Manila RR. The units were later shown as “U12C” (not UM12C” in the GE World Users Lists, but as far as is known, were not so designated when built and delivered.


Cheers,

Ask and it shall be given, indeed! Thank you for that list of generator types!
--
GT581E1 and GT581E2 are weird: they must be later than the GT581 forms with gear-driven auxiliaries, and my sense is that everybody knew that gear-driven was better (less maintenance, fewer failures) than belt-driven. So WHY? Unless for use in a length-limited installation. (I ***think***, but will check, that the GT564 originally had belt-driven auxiliaries which were replaced by gear-driven for most production.)
--
In domestic applications, there is a clear demarcation between GT564 and GT581: some time in 1950. The GT581 replaced the GT564. Someone (I think on one of the Railroad.net forums years ago) said that the GT581 was a "productionized" GT564: no better, possibly even not as good, but cheaper to build.
--
And: "Australia QR CB-engined Cape gauge C-C road-switchers: GT567 main generators with belt-driven auxiliaries and inherent characteristic excitation." !?!?!?! In domestic applications, the GT567 was the generator developed for use with the Fairbanks-Morse engine: initially for the 2,000hp version in the Erie-built, but after W'house bailed out of the heavy traction market used on both 1600 and 2,400 hp models. The Queensland units were 1100 hp, weren't they? And the C-B engine would seem to be very different in its characteristics from the F-M engine (higher top r.p.m., for example). So why?
--
And again: thanks! I'll re-read and check what sources I have and get back if I find anything even potentially useful.

Checking my recollections about the GT564…
--Kirkland's Alco book seems clear that the GT564 was used on the Alco FA-1/FB-1 and RS-2, replaced with the GT581 on the successor FA-2/FB-2 and RS-3 models (RS-3 first built May 1950, FA-2 in October 1950). GT581A and GT581C seem to have been used concurrently.
--
--For the GT564… The Alco prototype/test freight units, the "Black Maria" ABA set of 1945 with the 241 engine, had (according to Kirkland) GT564A generators. Auxiliary and exciter were belt driven: the "spool" (? is that the right word?) for the belt is shown on the end of the traction generator in a drawing of the 241 engine in Steinbrenner's Alco book; from drawings of the Black Maria and FA-1 in Steinbrenner it would appear that the auxiliary and exciter were mounted on top of the main generator, with their belt "spools," like that on the main generator, on the end of the generator toward the cab end of the locomotive. FA-1 production started with the GT564B generator, also apparently with belt-driven auxiliary and exciter. Kirkland says of the second iteration of the FA-1/FB-1 (DL208A/ DL209A, first delivery in February 1947):

"Belt drives for the traction motor blowers and for the exciter and auxiliary generators were discontinued, in line with a policy adopted by all builders at this time of moving away from belt or mechanical drives for auxiliary equipment other than the mechanically driven air compressor. The traction generator was equipped with gear drive, to power the exciter and auxiliary generators. A second auxiliary generator was added to supply power to the traction motor blowers, which were now motor driven. The traction generator designation was changed at this time from model GT564B to GT564C."

So the combined wisdom of Alco and GE in 1947 seems to have been that gear drive was better than belt!
--
--In related developments…
(1) Another drawing in Steinbrenner's book, which purports to be of the initial production version of the PA-1 passenger locomotive, suggests that auxiliary and exciter generators were gear-driven, ousted in their now-familiar location on the end of the coned section of the ration generator housing, on even the first application of the larger GT566 generator.
(2) Kirkland's Alco book covers only domestic U.S. and Canadian models, and gives no details on export units. (Even the shovel-nose units GE provided for the White Pass and Yukon narrow-gauge railway, which were among the first locomotives to get Alco 251 engines, are mentioned without technical details.) All U.S. domestic Alco switchers with 6-251 engines (800 hp S5, 900 hp S6, and 1,000 hp T6) had GT584 generators. Their Canadian analogues (MLW S-13 and RS-23, both of 1,000 hp) had…GT740 generators: at least in nomenclature, Canadian General Electric was a law unto itself!
(3) A photo of a factory scene on p. 110 of Kirkland's Alco book supposedly shows "A model 251D, six-cylinder, 1,000 hp turbocharged engine coupled to its General Electric model GT-584 main generator, […] in 1962 being lowered into position in a Dl-535 export locomotive." The photo angle gives only avery poor view of the generator, but ***I think*** the cylindrical portion looks a bit shorter than that in a GT581. But something is wrong here: isn't the DL535 a 1,200 hp model?