• EMD Export Locomotives

  • Discussion of Electro-Motive locomotive products and technology, past and present. Official web site can be found here: http://www.emdiesels.com/.
Discussion of Electro-Motive locomotive products and technology, past and present. Official web site can be found here: http://www.emdiesels.com/.

Moderator: GOLDEN-ARM

  by Allen Hazen
Many thanks for these posts about EMD locomotives most North American rail enthusiasts know nothing about!
W.r.t. to the Egyptian AA16: how similar was this to the (visually similar) units built for a number of European railways (Scandinavia, Belgium come to mind)?
  by Pneudyne
As best I can determine, all of these AA16 units, in both A1A-A1A and C-C forms, were quite similar in respect of engines, electricals and key mechanical components.

Visually there were some differences.

For example, here is the Nohab version, as built for DSB Denmark:
Nohab-EMD AA16 for DSB Denmark.jpg
And here is the AFB version as built for SNCB Belgium:
AFB-EMD AA16 for SNCB Belgium.gif
  by Pneudyne
Here is the equipment layout diagram for the first batch of AA16 locomotives built by Nohab for DSB in 1954. These had 16-567B engines:
DSB Nohab-EMD AA16 Layout.gif
They had all-welded Flexicoil trucks, the choice of construction being based upon Nohab’s experience.
DSB Nohab-EMD AA16 Welded Truck.jpg

  by Pneudyne
The first batch of AA16 locomotives built by AFB for SNCB in 1955 also had 16-567B engines.
SNCB AFB-EMD AA16 Layout.gif

These though had Flexicoil trucks with cast frames, supplied by Haines St. Pierre. In this case the choice reflected the established expertise in Belgium. Henricot was perhaps better know as a supplier of cast truck frames, its customers including English Electric in the UK. These Belgian locomotives also had spoked wheel centres with tyred wheels, both features, but perhaps more so the former, being unusual on an EMD locomotive.
SNCB AFB-EMD AA16 Cast Truck Frame.jpg
  by NorthWest
It is interesting that the sides of the AA16 foreshadow the sides of the later Henschel-built J12s for ÖBB.

Halfway down the page here are a few interesting 'proposals' put out in Henschel literature. The top two both appear to have 16-567 equipment though one is in a body stretched over EMD's A1A Blomberg E Unit trucks. The second appears to be a slightly lengthened double-cab F Unit.
http://www.nohab-forum.de/t1869f9-nohab ... egypt.html
  by Pneudyne
Thanks for that link – that Henschel brochure is most interesting.

In respect of body styling, both Nohab and AFB deviated quite a bit from the EMD nose-end shape, and to my eye, not positively so. Henschel was somewhere in between with its AA units, although the TT12 nose ends did look more like a scaled-down version of the EMD original. In contrast, Clyde in Australia managed to retain close to the original EMD shape, even though it was working to what was more-or-less the same loading gauge as the European builders. One may ask what was Nohab thinking when it made its changes, particularly as the Clyde version was already extant. Actually, Nohab did a somewhat better on the nose ends of the SJ Ra class electric locomotive of 1955.

The shaped (two-plane) bodysides adopted by Henschel for the Egyptian AA16 and AA12 models were also found on its slightly later standard CM-gauge export diesel-hydraulic model, which was likely informed to quite an extent by the DB V160 class, which had been designed by Krupp although Henschel built some of the prototypes and I understand may have undertaken the work on optimizing the cardan shaft assembly.

Majorly different were the La Brugeoise-Nivelles (BN)-EMD models built for SNCB in the early 1960s. Here the body style was essentially prescribed by SNCB, and was very similar to that used by Cockerill for its builds of the same period, see the “Baldwin Export Locomotives” thread, viewtopic.php?f=5&t=160538" onclick="window.open(this.href);return false;. Here is the BN-EMD AA12, SNCB class 212. The BN-EMD AA16, SNCB class 205, had the same basic shape.
SNCB Class 212, BN-EMD AA12.jpg
By then, AFB was no longer building locomotives, and the EMD licence had passed to BN. As BN did not build electrical equipment, it subcontracted to ACEC, who was not only a Westinghouse licensee, but also a GE licensee by virtue of its acquisition of SEM, who inter alia had supplied the electrical equipment for the GE-design shovel-nose locomotives supplied to the Congo in 1954, mentioned in the ‘GE Export Diesels’ thread, viewtopic.php?f=8&t=159801" onclick="window.open(this.href);return false;.

Quite how ACEC managed those three licences for minimum interaction is unknown, although the existent extensive cross-licensing between GE and Westinghouse would have eased the situation between those two. These days there would be a whole corporate manual on the establishment and maintenance of firewalls. I can’t help wondering though whether the ACEC clones of the EMD traction motors were not a bit better. If you knew how to do a lot better – and clearly ACEC did - it would be hard to resist putting in a tweak here and a small refinement there.

  by Allen Hazen
Re: "In contrast, Clyde in Australia managed to retain close to the original EMD shape, even though it was working to what was more-or-less the same loading gauge as the European builders."
--- I believe that Clyde, in building "Bulldog" units for Australian customers, used stampings imported from EMD to get the complex shape right… despite the slightly narrower loading gauge(*). To make them fit, they trimmed along the centre line: the most visible evidence being the centre post between the two windscreens, which is noticeably narrower on Australian units than it is on domestic E and F types.
(*) From something said when the Melbourne commuter lines were testing a derivative of the Sydney "Tangara" E.M.U., I think the Victorian loading (or structure) gauge is even a couple of inches narrower than the NSW one, despite the Victorian track gauge being 5 and a half inches wider.
  by Pneudyne
Thanks, Allen. I was previously unaware of that facet of the Clyde designs, but now that I look again at the photographic evidence, the difference in the windshield centre pillar width is obvious!
NSWGR 42 Class.jpg
EMD F7.jpg
An aspect here is that EMD chose to address this part of the export market with what might be called a quasi-standard platform with significant regional variations. In all cases these variants were built by its overseas associates, including Clyde, Nohab, AFB, Henschel. Thus LaGrange was not involved with the production on non-standard models and variants, during the 1950s at least. That changed in a small way with the significant expansion of the home-produced export range in 1958, and then in a bigger way with the building of the very non-standard GA12 and GA12C models for India.

On the other hand, Alco started with a standard model, the DL500, that could be built at home or by licensees, and then allowed the licensees to make variations to suit regional and local requirements.

  by Pneudyne
NorthWest wrote:It is interesting that the sides of the AA16 foreshadow the sides of the later Henschel-built J12s for ÖBB.
Here is the Austrian (ÖBB) Henschel-EMD J12:
Austrian Railways (ÖBB) EMD J12.jpg
The trucks were apparently non-standard, and supplied by Austrian company Simmering-Graz-Pauker (SGP), presumably to provide some local content. The front end bears some resemblance to Austrian electric locomotives of the time. As best I can determine, this was the first instance of EMD’s use of the “J” prefix in the model number.

For comparison, here is the Danish Nohab-EMD AA12:
Danish Railways (DSB) Nohab-EMD AA12.jpg

And the Egyptian Henschel-EMD AA12:
Egyptian Railways Henschel-EMD AA12.jpg

One supposes that the Belgian (SNCB type 212) BN-EMD AA12 mentioned upthread did not have sufficiently flat ends to justify the use of the “J” prefix, so it was an AA12 despite its rather vestigial nose ends.

Its SNCB type 213 diesel-hydraulic counterpart, about which I’ll post more if/when I can find some worthwhile information, had the EMD model number DH12, one assumes meaning diesel-hydraulic with 12-cylinder engine, rather then the 12th diesel-hydraulic design.

  by Pneudyne
In respect of diesel-hydraulic counterparts, in the Baldwin Export Locomotives thread, Allen Hazen mentioned the diesel-hydraulic version of the EMD SW8. Apparently this was known as the DH2, being EMD’s second diesel-hydraulic design. Although strictly speaking this does not qualify for inclusion here, I think that it is pertinent when it comes to setting the scene for an AA12/DH12 comparison. Information on the DH2 appears to be hard to come by, but here is an outline “Diesel Railway Traction” (DRT).
DRT 195312 p.310 EMD Diesel-Hydraulic.gif

The DRT folk engaged in some speculation as to why EMD found the gearchange part of its transmission unsatisfactory. But it was probably not unreasonable to assume that Allison used a planetary-type gearbox with engagement by oil-immersed multiplate clutchpacks. That was what it used for its early on-highway automatic transmissions (which had a splitter gearset preceding a 3-speed main gearbox). Given that the clutchpacks for example do the work of pulling down the engine speed (if necessary) during upshifts, I imagine that doing that with the much higher rotational inertia of a medium-speed engine could be problematical. The torque converter details remain a mystery. In the Budd RDC transmission, Allison used a three-stage converter with a lockup clutch; in its on-highway automatic transmissions a polyphase converter-coupling with a lockup clutch, and in its V-drive city bus transmission a two-stage converter with lockup clutch, so evidently it adopted a “horses-for-course” approach in respect of this item.

  by Allen Hazen
Thank you for the discussion of, and the "Diesel Railway Traction" article about, the "SW-8-Hydraulic"!
The article has the first technical information I have seen on the locomotive… but the data is tantalizing rather than satisfying. A stock (diesel-electric) SW-8 weighed 230,000 pounds, so I suspect the number given for tractive effort (57,000 pounds) is just the conventional one-quarter-of-weight-on-drivers that was standardly quoted as the starting tractive effort of diesels at the time. (Diesel hydraulics have mechanical connections between the driving axles, so were often claimed to be better at low-speed tractive effort than diesel electrics, on which each axle could slip independently: I don't think we had evidence of that here.) Unless the transmission was VERY inefficient (or I've done my arithmetic wrong, always a possibility for an aging brain!), the locomotive should have been able to maintain that tractive effort at 2.6 mph at well under full engine horsepower: to get a data point relevant to comparing the efficiency of d.e. and d.h. transmissions, we'd need to know how much faster the unit could go and still exert that much tractive force!

I can't find (in the three books and one website I've checked) performance curves for a standard SW-8: if further search turns something up, I'll get back…
An SD-7 with optional, low-speed, 65:12 gearing(*) had a continuous speed of 4.6 mph(**). An SW-8 had about half the engine power (8-cylinder 567B instead of a 1500hp 16-cylinder 567B: the SW-8 was rated at 800hp rather than 750, but at the level of precision I'm currently aiming at I'm not going to quibble about the 50hp!), but 2/3 as many traction motors. My ***guess***, then, is that a (diesel-electric) SW-8 with this gearing would have had a minimum continuous speed of about 3.5 mph.

(*) This gear ratio was used for units intended for low-speed service, and nominally allowed a top speed of 50 mph. Most EMD freight and switcher units of the time had 62:15 gearing, for a nominal 65mph top speed.
(**) Number from Alvin Stauffer's "Pennsy Power II". The Pennsylvania Railroad had two SD-7 with this low-speed gear ratio, purchased specifically for service on the line to Madison (Indiana), which had the steppes grade on the North American network at the time.
  by Allen Hazen
F.w.i.w., George Elwood's marvelous "Fallen Flags" rail image resource has, at
http://www.rr-fallenflags.org/manual/sw8-om.pdf" onclick="window.open(this.href);return false;
an EMD operator's manual for the (diesel-electric) SW-8 (also covering the SW-9 and the transfer derivatives of these models). Data are included for 65/12 and 62/15 gear ratios: I assume this means that EMD thought both were options appreciable numbers of customers would like.
--Starting tractive efforts are shown for adhesion rations of 20%, 25%, and 30% (all simple factions of the locomotive weight).
--Nothing is said about continuous speed or short-time operation. I have a feeling that these may not have been an issue with these low-powered locomotives: that their traction motors were "protected" by adhesion limits, in that the drivers would slip if one TRIED to operate them at a speed and power that exceeded what the motors could take!
--For the SW-8 with 65/12 gearing, (automatic, but could be prevented if-- as might happen in yard switching-- one wanted to stay in series) forward transition from series to series-parallel was at 7.1 mph.
--By this time, top speed for units with 65/12 gearing was quoted as 55 mph (The Stauffer book gives 50mph for the PRR's EH-15: low-geared F-3 intended for helper service), but with a footnote noting that this is just the speed permitted by the traction motor structure: track conditions, etc, could mandate a lower speed limit.

All a bit off-topic, I'm afraid, but this is what the diesel-hydraulic design was "competing" against.
  by Pneudyne
Here the curves for the SW900. This was somewhat more powerful than the SW8, and I think that it also had higher capacity motors and larger traction motor blowers. As may be seen, the SW900 was viewed as being self-protecting, so that MCS and CTE numbers were not quoted.
EMD SW900 Curves.gif
Very roughly, the 57 000 lbf point on the SW900 tractive effort curve corresponds to a speed of 3 mile/h. Reducing the power by 11% (corresponding to the difference between the SW900 and the SW8) would nominally result in an 11% speed reduction for the same tractive effort, say from 3 to 2.7 mile/h. On that basis, 57 000 lbf at 2.5 mile/h looks to be not unreasonable for the DH2, at least within whatever precision one might reasonably assign to these very rough back-of-the-envelope calculations imply.

If we assume that the DH2 transmission was specially designed for it (albeit using established Allison precepts), then the designers were free to size the torque converter and its cooling system to meet the mission requirements. On the question of MCS and CTE, perhaps it was thought that there was no real advantage in having a CTE larger than that which typical adhesion conditions allowed, and so they settled upon the “25%” number as being a reasonable target. And on the curve, that corresponded to 2.5 mile/h, quite reasonable for the MCS.

Returning to the SW900, 57 000 lbf at 3 mile/h corresponds to 456 hp, well down on the 900 hp available at the main generator input. Whilst transmission efficiency is lower at low speeds where traction motor currents are higher, and so consequently are ohmic losses, in part , perhaps in major part, that difference is attributable to the fact that at the bottom of the speed range, the transmission cannot fully load the engine, so the utilization factor is low. Looked at another way, at zero speed with full tractive effort the power is zero. So the power curve ramps up from zero to maximum in the lower speed range, usually with a turnover point below about 10 mile/h, and probably at the low end of the range for switching locomotives.

In the DH2 case, 57 000 lbf at 2.5 mile/h corresponds to 380 hp, again well down on the 800 hp available. Here a reasonable proportion of that difference would be attributable to power lost to heating the torque converter fluid. Unavoidably, at very low speeds, hydraulic transmissions must operate on the low efficiency part of the torque converter (or 1st torque converter as appropriate) curve. Here are a couple of examples that illustrate that point.
Voith L36r Curves.gif
SCG Torque Converter Transmission Curves.gif
The SCG transmission comprised a single-stage, two-phase torque converter-coupling and a four-speed Wilson planetary gearbox. Note that torque converter efficiency in 1st gear had dropped to 50% at around 2 mile/h. And the tractive effort curve leans to the left of the parabola below 5 mile/h.

The Voith L36r transmission was of the classic Voith triple converter form. 1st converter efficiency has dropped to 50% at around 10% of maximum speed. By the way, Voith is being generous in allowing a 100% utilization factor down to zero speed for the electric transmission comparison case. Rather it would decline somewhat. I’d say that the utilization factors in this case are “gross”, in that they reflect the total engine loading, including power lost to heat, as well a useful power.

  by Pneudyne
Not at all to sideline the discussion on the EMD DH2 diesel-hydraulic, which wherever it leads I think will set the scene for more discussion on the export DH12 in the near future, but continuing the EMD export theme, this advertisement shows that there were some exports of domestic models, in this case the GP9:
DRT 195801 p.09.jpg
This GP9 variant supplied to Venezuela does appear to have been equipped with dynamic braking. That makes me wonder – did it have field-loop control as per standard domestic practice, or did it have potential wire control as per standard export model practice. (EMD dynamic brake control details were discussed in this thread: viewtopic.php?f=6&t=152156" onclick="window.open(this.href);return false;.)

I imagine that in a situation where exported domestic models might have to mix with export models, potential wire control would have been preferable. In the just mentioned DB thread it was determined that in 1957 the EMD -9 series was changed over from high-current rheostat DB control to potentiometer and load regulator control, with what was effectively a potential wire between the potentiometer and the load regulator positioner relay. So adapting a GP-9 so fitted to full potential-wire control would appear to have been relatively straightforward. The Venezuelan GP-9 fleet looks to have just post-dated the change.

  by Allen Hazen
Two posts up, in the post with the SW900 curves, you say
"Whilst transmission efficiency is lower at low speeds where traction motor currents are higher…"
How big an effect is that? The ballpark figure I have seen for American diesel-electric locomotives in the 1970s and earlier is 18%, so that, e.g., a 1500hp locomotive (*) should be expected to have a horsepower at the rail of around 1230. I noticed, in the Stauffer book, that the PRR's low-geared SD-9 had a continuous horsepower of 87,700 pounds, which, at the 4.6mph continuous speed quoted, comes to about 1076 hp. Is this plausible? … Since the SW-8 was roughly contemporary with the SD-7, one would expect it to represent a similar level of technology. The speeds and tractive efforts we are looking at, however, translate to significantly road h.p. even than the 71% or so of the SD-7: so, as you argue, "utilization factor" rather than engine power and transmission efficiency seems to be the limiting factor.

EMD seems to have thought that their transmissions approached being "self protecting" on many of their early freight locomotives. An operator's manual for the F-7 freight locomotive, at the beginning of the section on short-time operation, says that the traction motors are almost self-protecting on this model for the 65/12 and even the 62/15 gear ratios: this was a 1500 h.p. locomotive with four traction motors of the same model used on the SW-8!

(*)Numbers for American ratings, where the engine power quoted is input to traction generator.