counted for by repairs; 27 per cent by crews; and 23.9 per cent by fuel. All other items were comparatively minor, ranging from depreciation (11.1 per cent) to water (0.3 per cent). To backstop the implication of these figures, GE hired a railroad mechanical officer to ask 33 major roads what they thought of their diesels and he got an earful of what C.M.E.'s did not like (anything that broke down or needed excessive maintenance). So once more the old dictum was verified: "If you can't prove it's essential, take it off; if it isn't on there, it can't fail." That, in turn, led to some interesting conclusions:

The more complex a design, the more specialized must be the training of maintenance personnel; yet a high turnover of maintainers is a fact of life for many roads. Therefore, if a design is simple, it's understandable - even to a new man.

¶Simplicity can sometimes be achieved only at the expense of greater fuel consumption. But "fuel expense is clean-cut; that is all there is to it," reasoned GE, whereas "while a locomotive is being maintained it is neither making mileage nor producing revenue and there is the recurrent expense of training men to maintain it. Capital charges continue while the locomotive is out of service. Furthermore, a device that requires maintenance may cause road delays and other intangible expense never allocated to the part in question." Lesson: keep it simple.

To illustrate: air supply. Locomotives gulp in huge volumes of air - for engine combustion, for generator and traction motor cooling, for cab comfort. It should be dry, clean air. Which, until the U25B, led the designer into a vicious circle ending in inevitable compromise. Ordinary filters require removal and cleaning every 2500 miles or so; each blower must have its own motor or be driven by shafts, belts, or pulleys; and engines themselves foul the air supply with their oil fumes, exhaust, and heat. All this adds up to dirt or maintenance, which spells expenses either way. Confronted by this age-old dilemma, GE's engineers first joked, "We'll have to design the air system first and then build the locomotive around it," and then took themselves seriously. To suck in air and propel it under pressure to ventilation outlets, an axial-flow blower, engine-driven off the same gear box as the twin radiator fans which flank it, was mounted at the rear of the locomotive directly above a primary air cleaner. The cleaner is a rectangular box infested with 2-inch-diameter plastic tubes-1470 of 'em - working on the principle of centrifugal force. Incoming air swirls into the center of

the tube; 99 per cent of all dirt particles larger than 8 microns* in diameter are forced out one end; cleaned air goes out the other. Though such cleaners are expensive to build, they are self cleaning and require no maintenance. The cleaned air for cooling uses is then directed through a duct between the main sills of the locomotive's platform or frame. GE refers to it as "inviolate," for not a single pipe or wire runs in or out of it at any point; hence there is no opportunity for oil seepage or other contamination. Thus ventilation air for main and auxiliary generators, traction motors, cab, and the control cabinet (mounted under the running board on the left-hand side and pressurized) for resistors, contactors, and relays completely bypasses the hot, oily atmosphere of the engineroom - and supplementary blowers and filters, cab-heat and defroster motors are obviated. This air supply system, so simple as to be ingenious, has been cited as the finest of several fresh design departures built into the U25B; it has won quick endorsement by the railroads, not to mention emulation in principle by a competitive maker.

The same start-from-scratch thinking shows up in the locomotive's engine cooling system. Here the problem was primarily one of cold weather. To begin with, a 16-cylinder engine requires about 220 gallons of cooling water; heat transfer takes place in radiators over which air is blown by adjacent fans. A hot day and a wideopen throttle create what designers refer to as "maximum heat-transfer requirements," but when the engine is idling on a February night it may require only 2 per cent of its total cooling capacity. Summer, then, requires large shutter areas and numerous radiator fans; but on a zero day these shutters must be mechanically closed up tight and fan motors must slow or stop. An automobile, on the other hand, sidesteps such complexity by recirculating all or part of the water in the warm engine, and antifreeze allows cold water to remain in the radiator. This would work on a locomotive if a railroad could afford the Prestone. What GE decided to do in essence was to imitate the automobile (i.e., to control the water flow through the radiator rather than the air flow). The problem of freeze-up was resolved by use of a radiator with selfdraining tubes. A combination of temperature-sensitive and pressureoperated valves automatically determines, on the basis of outside weather and engine heat, whether the cooling water should be recirculated through the radiator or not - or only partially. If these valves dictate that the radiator A micron is a unit of length about 0.000039 inch.

HEART of the U25B is the largest and most powerful diesel engine in active railroad service today, here or abroad. The FDL-16 is a 4-cycle, 16-cylinder (each of 9-inch bore and 10%1⁄2-inch stroke), 45-degree V, 1000 maximum r.p.m. power plant of 2500 h.p. input to the generator. It is the end result of search and research, for GE checked 46 locomotive-size American and European engines (including those "quick-running" Germans) before making up its mind, then revised virtually everything but the basic concept of the engine it did select before introducing it to the U25B.

Briefly, a V-type design was selected for its excellent maintenance accessibility and space utilization characteristics; and a 9-inch bore creates the biggest V engine that can be accommodated within the 6-footwide hood of a road-switcher. Sixteen, in turn, was the maximum number of cylinders which could be coupled to one crankshaft without risking torsional stress problems.

So far as foreign "quick-running" engines went, their higher r.p.m. did not compensate for their smaller bore and stroke, hence the required locomotive horsepower couldn't be produced off a single crankshaft. Electric transmissions favor single-engine units. Moreover, GE felt that the ratings of the overseas engines were optimistic, being based upon European rather than American field conditions; and sturdiness was questioned.

The fundamental concept which sired the FDL-16 was drawn up by Cooper-Bessemer in the early 1940's, and at the end of World War II development had reached the point where a 6-cylinder inline version could be placed in a GE export locomotive. The Erie Works people deemed the engine "fairly satisfactory" in operation but too expensive and heavy to be competitive with other makes and there the matter stood until 1953, when Alco-GE quietly dissolved their hyphenated partnership. At that point it became imperative for GE to obtain rights to and to develop an engine if the builder expected to enter a line of standardized units in the expanding world market. Of all the 2- and 4-cycle engines studied, the CB appeared to have the "greatest potential" for high-horsepower application; in particular, GE liked its "iron pistons, large bearing areas, good

a

Having settled on a 130-ton B-B layout for its U25B, General Electric decided to concentrate all its road research on that one size and model rather than to spread its developmental efforts less effectively across wider catalog. The builder is convinced that the specialization perhaps necessary to eradicate steam power is dead and that the market is now focused upon high-horsepower locomotives which can effect significant unit reductions in moving heavy trains faster. Which leads to the 2500 h.p. U25B's operating in multiples up to a total multiple-unit output of 10,000 h.p. or more.

IN service, the U25B demonstrators have now been at large long enough to acquire some meaningful statistics. On grades, for example, a four-unit 10,000 h.p. team has registered these performances:

On 1 per cent compensated, 3952 tons moved at 31 mph.

On 1 per cent again, 5762 tons (109

cars) moved at 22 mph.

Also on 1 per cent, 9200 tons (144 loads, 12 empties) moved at 14.5 mph.

On 1.42 per cent, 5258 tons (116 cars) moved at 18 mph.

Again on 1.42 per cent, 5614 tons (128 cars) moved at 16 mph.

[On 2.2 per cent, 3950 tons (82 cars) moved at 15 mph.

On dry rail in the West the U25B's have recorded adhesions of as high as 24 to 25 per cent at speeds from 11.5 to 12 mph. They started a 3950-ton train on 2.2 per cent without any wheel slip indication and achieved 10 mph within 4 minutes under full throttle.

On demonstration, a single U25B unit started a 2500-ton train out of a yard on a 1 per cent grade, maintaining 30 per cent adhesion for 5 minutes with some mechanical sanding.

On a 110-car, 11,900-ton ore train, three U25B's totaling 7500 h.p. held 19 mph on an average grade of 0.45 per cent, whereas the train would have required four of the road's 1750 h.p. units and speed on the grade would have been 11 mph.

As one would expect of any original piece of engineering, the U25B has experienced its full measure of teething problems, for no lab test can quite duplicate prolonged 50 mph running across a dusty desert at, say, a 120

degree roadbed temperature, or the throttle hand of an impatient engineer moving big tonnage out of a yard on frosty 1 per cent. Yet the performance consensus is that the locomotive puts out its advertised 2500 h.p. and that not only are its engine and electric transmission unequaled in capacity now but they have a substantial margin for future growth.

There is no use in pretending, to be sure, that General Electric has anything but an upgrade fight to stake out an appreciable share of the market for the U25B.

First, GE is new in the domestic road diesel field. Although 5000 power assemblies (cylinder, rod, valves, and so forth) are at work in Universalmodel export units overseas from South America to South Africa, the U25B's FDL-16 engine remains fresh to U. S. roads and hence carries with it implication of new parts inventory, maintenance procedures, personnel orientation. Similarly, U. S. roads possess no historic knowledge of the FDL-16's long-term repair cost or fuel consumption. Until the U25B, GE was a supplier of traction motors, generators, and electrical control gear, primarily to Alco, and was a builder only in terms of such specialties as straight electrics, turbines, export motive power, and smaller industrial and shortline diesels. Aware of this image, the Erie design team has kept the design of the U25B as unsophisticated as possible using, for example, stock railroad storeroom hardware for all expendable parts - but nevertheless first-class salesmanship is mandatory, supported by in-depth field service.

Second, in the midst of a market dominated by trade-in thinking the U25B is without an ancestor. Yet to remain competitive GE must offer an allowance on other makes which, aside from trucks equipped with GE motors and certain GE electrical components, offer little but their weight in scrap prices.

All this totals up to the penalty paid by the pioneer. If comparatively few U25B's have been sold to date (see customer specs on this page) when gauged by the success accorded Electro-Motive's GP30, then at least General Electric has signed up a cosmopolitan corps and attained the foot-inthe-door milepost. If and when these original buyers place substantial U25B repeat orders we will have an opportunity to put General Electric's unique design effort in perspective. Bear in mind that upon occasion the men at Erie have waxed eloquent about an "advancing technology [which] may someday produce diesel-electrics of 4000 or even greater horsepower per four-axle unit." If so, the U25B will be found in their family tree. I