I don't think that there's any point in trying to scale down weight when there are so many other things that we don't really scale down properly, like our curve radii, or our grades, or in many cases, even the speed at which we run trains.
As for how the prototype selects power for a train: each locomotive type is assigned a tonnage rating for each division or subdivision on which it is allowed to operate. This can sometimes vary depending on the season (for instance, wet leaves on the rails can play havoc with tonnage ratings). Steep grades (up or down), curvature (sharp or otherwise) and weather conditions can all have a negative effect on tonnage ratings. A locomotive rated for 5500 tons on straight and level track will not have that same tonnage rating on a mountainous division. I'm not sure how the prototype arrives at their tonnage ratings, but as modellers, it's fairly simple.
On my layout, the Erie Northshore sub is either level or downhill going west, and level and uphill going east. A modified Athearn Mikado, travelling west, can pull probably a couple dozen cars (funny, I've never done the tests for west-bound trains), although such a train would be too long for the passing sidings). East-bound, the same loco will handle 12 randomly selected freight cars (some roll freely, others less so, and sometimes, there'll be a real brick in the consist - car weights are approximately around the NMRA recommendations). I know this because I've tested these locos upgrade, starting with a train that the loco can move on level track, but not up the grade. I simply remove cars until the loco can consistently make the grade. If one loco in the same class cannot pull as many cars as the others, then the whole class is assigned the
lower rating. If the train that needs to be moved exceeds, even by one car, the tonnage rating for the loco available, then a second loco is added to the consist. These same locos, when working on a different subdivision, may or may not have the same tonnage rating: that will be determined by on-site tests. I have two Moguls, one from IHC and the other a re-motored brass import. The smaller brass loco will outpull the IHC, but both are rated for the lower tonnage rating of the IHC.
I also do another tonnage rating test, this one because I run trains of hoppers with "live" loads. Each of the cars weigh 8 ounces, quite a bit more than the NMRA recommendations, and enough to severely tax just about any loco. The power plant requires three trains of twelve cars each, per week. To move a 12 car train of the loaded hoppers, plus a caboose, requires two Athearn Mikados, or two Bachmann Consolidations, or one of each: all have the same tonnage rating for coal trains. (One of these loaded hoppers in a train of mixed freight doesn't affect the tonnage rating, as there's a bit of a built-in "cushion". Two or more can be enough to push a train over the tonnage rating of a single locomotive).
My locomotives were modified to achieve these tonnage ratings because that was the heaviest train that will be regularly operated. In actual practice, most trains are usually overpowered, as my passing sidings limit train length.
As a general rule-of-thumb, the maximum pulling power of any locomotive is approximately 25% of the locomotive's weight, so the heavier the locomotive, the more it will pull. I have three modified Athearn U-boats (A bit out-of-place on a 1930s-era layout
) that each have two can motors and weigh just over 33 ounces: their measured drawbar pull is 8.3 ounces, about 25% of their weight. Traction tires can increase this percentage, but I don't use them. Most locomotives can stand to have some weight added to them: those Athearn Mikes, when I was trying to determine how much weight I could safely add, were still able to slip their drivers with a 20 oz. saddle of sheet lead draped over the boiler. Unfortunately, I couldn't find room inside the boiler shell to hide that much weight.:cry: If you're going to add weight to a locomotive, it is important that the loco, when pulling a train that exceeds its pulling power, is still able to slip its wheels when the weight of the train overwhelms it. An easy way to ensure that this will happen is to place the desired amount of weight atop the loco (like my experience with the lead saddle, you may not be able to fit all of it into the shell when you actually set about modifying your loco), then couple one car behind the loco. While you hold the single car in place on the track, open the throttle: the loco will try to move the car, but because your hand prevents this, the loco's wheel will start to slip (spin). If this occurs, you can safely add that weight to the loco. If the loco's wheels do not slip when you open the throttle, you have used too much weight, and you should remove some, in increments, until the wheels do slip. Wheel slippage is like a safety valve: it lets you know that the train is too heavy for the locomotive. When the train stops moving, but the wheels don't slip, all of that power that you're applying through the throttle is still trying to make the motor run: when it can't, it produces heat, which will, sooner or later, destroy the motor. If you're running a train, and the wheels slip somewhat on a grade or curve, you might want to add another locomotive, or remove a couple of cars from the train, as excessive slipping will, over time, remove the conductive plating from the wheel treads.
As usual, I've blathered on well past my two cents worth, but I hope this adds something useful to the discussion.
Wayne
