Gee, golly and gosh-a-roo, Nate. Just look what you‘ve done!
Many variables abound around the railroad, which is one of the things that make them so unique, and much has changed since I left it all behind nine years ago. But, as near as I can tell, physics, including gravity and the other forces of nature have remained the same everywhere on the planet (except Omaha) from that time until this. Here is what was at play at that time and how I was taught by some people who definitely knew their business.
There are many types of draft gear and they vary in tensile strength from type to type. High capacity draft gear is usually found on the cars of unit trains, grain and coal in particular. As long as I worked “standard” E-Type (which should be the benchmark, yes?) draft gear was understood to be rated at 240,000 lbs. This figure includes a 10,000 lbs margin for error above the 230,000 lbs. referenced above. If you have mixed freight, you better figure less than 240,000 lbs. if you want any kind of assurance to keep the train in one piece. (Handy Dandy Note #17: “F” type knuckles are rated nearly the same but will fit into a loco draw bar in a pinch).
But, even though that figure can vary between standard and higher capacity draft gear, the point is moot and the information is useless unless someone knows what to do with it. So the question becomes, “How do we know how much force will be exerted on the draft gear by the tonnage for the train we have and over the territory it will move?” Someone a lot smarter than you and me with a very sharp pencil came up with the “Rolling Train Resistance Formula” and the “Starting Train Resistance Formula.” The two are almost the same and quite simple to understand.
The rolling train resistance formula is: 20lbs for each ton in the train, multiplied by grade, plus 5 equals the draw bar force, or [(20 x T) x %G] + 5 = F. Starting force is the same except substitute “30” in the place of 20 lbs.
Ok. I have 6,000 tons I need to get up a 2.5% grade. 20 x 6000 = 120,000 x 2.5 = 300,000 + 5 = 300,005 lbs. Oops. We’re over the limit by 60,005 lbs (300,005 - 240,000 = 60,005). Now what?
You must reduce draft gear force. You do that by reducing tonnage (rarely ever happens) or add helper power, either manned or DPU, pushing some of that excess trailing tonnage, which reduces the draft forces on the head end. When speaking of a multiple unit helper with three units and up, and entrained (two units can usually go behind the last car but certain restrictions may prohibit this placement), they must be placed such that, of the tonnage they handle, 1/3 must be ahead of the helper with 2/3s trailing. If you do otherwise the helper can shove the cars ahead of it right off the track.
This helper placement creates the “node” which Andy properly mentioned. The node is that point in the train where there is neutral slack. That is, between the car which is last pulled by the road engine and the first car of the helper cut, where the slack is“floating,” if you will. The node will vary according to changes in speed or grade as well as slack adjustments and the finesse Andy refers to is definitely the order of the day, every day.
Now this is not a definitive answer since no one knows everything (don‘t ever bet your life on info from unofficial sources, such as Wikipedia and Yahoo!Answers), but it is the principle under which every engineer I’ve ever known or had the pleasure and good fortune to learn from has operated for at least the last 70 years or so, whom also happen to have been experts in grade territory railroading and the power requirements necessitated by the immutable laws of the physical universe to overcome those grades and, quite frankly, the ones that learned only two generations removed from first hand from those who wrote the book on grade operations and the demands such operation imparts to the draft gear. And the same people taught Andy as well, which is why most times our answers to questions are usually in agreement. I can say with pride that SP engineers were the best trained and other railroads availed themselves of the Simulator center that was then in Cerritos, Ca., when available for training their engineers as a part of the promotion process.
This isn't taking into account track / train dynamics ("L/V" ratio), either. Then it starts to get a bit more complicated, but the formula to determine horsepower needs is applied in conjunction with the rolling resistance formula when trying to determine power requirements for trains.
Using our same example, how much horsepower is needed to get that same tonnage up the same grade at, say, 16 mph? Here the formula is horsepower per ton (arrived at by dividing tonnage by horsepower), multiplied by 12, divided by the grade will equal speed, or HPT x 12 / %G = S. So, we need at least 21,000 horsepower. Why? 21,000 divided by 6000 is 3.5 HPT. 3.5 x 12 = 42 / 2.5 = 16.8 mph.
Now we know our power requirements and how it must be distributed so as to not produce more force than the draft gear can handle. We also know that capacity varies between types of couplers, which goes back to the heart of your question. I'll stand on 240,000 lbs. as standard. If I'm not mistaken, those 32.000 ton trains are dedicated iron ore hauling on track that is flat as a pancake and straight as an arrow for 1,000 miles across the barren Australian outback. Not my idea of "standard."
Some one was pulling your leg about five mile long trains and the Big Boys. But you can get back at 'em. Ask that person which was the most powerful locomotive built in America and he'll happily (and incorrectly) tell you it was the Alco Big Boy. Not true. They were a little over four feet longer in the engine with tender wheelbase, and 30 tons heavier, overall, but it is the Baldwin built 2-8-8-4s of the Duluth, Missabe and Iron Range, that hold the title for most powerful, and but for the two differences I cite above, they outclassed the Big Boys in all other categories, cylinder size, boiler size, grate area, etc.. Get him to bet on it if you possibly can, the more the wager the better.
And I'll certainly be the first to admit I could be wrong about any or all of the above, as my three remaining brain cells only hook up every so often these days.
Buuuuuut...
Wanna make a bet?