Chris,
I'm not sure what happened, but there was a derating of the D938B in later GE locomotives. The Costa Rica units got derated to 950 hp, and I recall seeing later U10B units advertised at 950 hp which just might be copying the Costa Rica specs or was this the general spec? Horsepower ratings can be misleading, and often generally over optimistic. GE seems to have advertised their industrial switchers as the short-time gross/brake horsepower rating of the engines, and I believe common international practice has been to use the continuous bhp rating. Typical US practice (except the smaller industrial) has been to use the traction hp input as a rating, though the "gen-sets" of recent years have been misleading because they use the bhp rating, so you need to subtract the auxiliaries which between radiator fans, low voltage supply, air compressor and traction motor blowers can add up to a 200+ hp fairly easy.
The D397 in the GE 50, some 52 and the 54 tons was advertised by Caterpillar as a 500 hp engine, but GE rated it as 450 bhp/400 traction, but only 390 in the 54 ton (traction motors blowers and a larger radiator fan ate the extra 10 hp). An odd one was the 65/80 ton that used the Cummins NHBIS-6 engine. It was advertised as a 550 hp locomotive (2 x 275), but the specifications for GE only show them as "470 hp for locomotive use" and 420 for traction. An 80 ton spec for BLH/Whitcomb with the same engines rates them as 550 bhp and 470 for traction. I'm not sure if the lower GE rating was due to traction motor or main generator limitations, or the diesels themselves (GE used the GE747 self ventilating motor with double reduction, BLH/Whitcomb was using axle hung blown Westinghouse motors). If you really want to be confused, look at the horsepower ratings for marine engines - for short time use (short ferry runs or harbor tugs) they have a fairly generous rating, and then there are one or two intermediate ratings down the lowest "continuous" rating.
As far as track - I've run way too much on light rail with poor ties and no anchors. Typical North American practice using cut spikes doesen't provide any rail restraint besides gage. For welded rail, they have to heat the rail to the proper temperature, vibrate it, then anchor it in place at the correct tension factor taking into consideration the seasonal temperature extremes. All the track with concrete ties uses Pandrol or similar clips which provide both gage and tension restraint. The only screw spikes I know of in US practice has been limited use to attach tie plates to wood ties, then use a clip to actually retrain the rail. There was actually an oil train derailment here in Oregon blamed on this system. Union Pacific has installed screw spike tie plates on wood ties on some curves to allow transposing and replacing rail, but for some reason the heads of the screw spikes broke and the track inspections failed to find this before the gage failed.
There was a shortline I was working on this winter (originally built as 36") where they had been running trains of 20-40 loads the same direction on old light rail, very poor ties, sporadic sections of anchored track, primarily relying on extended range dynamic brakes which are capable of 35,000-40,000 pounds retarding force from about 4-20 mph per unit. You could see where the rail was running (moving longitudinally in the direction of loads and braking) around 6-18" in some locations over the previous 8 years or so. The previous operator derailed often, and only ran like one train a week at that with the resulting damage. We limited trains to 25 loads and no dynamics on the poor track. So far the trains have stayed on the track better and the only derailemnt so far might have had a mechanical issue contributing.
Think about it - if a K-28 is rated at 28,000 pounds tractive effort (more or less), then at nominal 25% adhesion (factor of 4 if you prefer) then it should weigh about 56 short tons on its drivers. Likewise a K-37 would be about 74 tons on its drivers. So any of the end-cab GEs, unless they were very heavily ballasted, should be in the approximate weight range of 50-70 tons, the same weight as the steam locos in present use. The only reasons to go beyond a B-B (Bo-Bo) design would be to lower the weight per axle (not a requirement here) or to spread the amps across more traction motors. There might be a slight advantage to creating a bigger "footprint" with more wheels in some low adhesion conditions, but you really don't need a loco crawling at 5 mph in exchange for paying for and maintaining the extra motors. Now if you are getting into bigger locomotives (weight or horsepower), then use, the extra axles/traction motors serve a real purpose.
Assuming a U10B at 950 hp at a nominal 60 tons - One with GE761 motors (meter+) at 36,200 pounds continuous tractive effort would be at a minimum continuous speed of about 8 mph and would require 30% adhesion - possible, but likely to slip as poor adhesion can be as low as 20%. Same engine with GE764 motors (36"+) at 26,400 pounds cte at 11 mph mcs would require 22% adhesion so would be more likely to hold its footing at its full rating. Don't forget that there is a schedule to keep, so I doubt most tourist railroads want to run their trains much slower than 10 mph.
Dan
