My favorite NTrak train is a string of MT Twin Bay 33' hoppers, all C&O. I own 127 now (always looking for more, dang out of print stuff). If the NTrak layout is _flat_ and doesn't have crazy curves too close a pair of GP-7's (Atlas) will pull ~110-120 without slipping. A four pack will pull all 127, no matter what the layout it seems. The short cars also tolerate very bad trackwork, for the most part. It's a huge crowd pleaser. I'll generally let it go 2-3 hours at each show, sometimes more, and have never had maintenance issues with m Atlas diesels.
I want to pull it with my LifeLike Berks. Mine are better pullers than most, one is good for 40-45 cars on a flat, non-curvey layout. Double headed they will often hit 100 cars.
Having 20 kids trying to count your train and then asking if the count is right is awesome. I want more steam that can pull long trains.
Near the end of the Ohio N Scale Weekend about three years ago just prior to tearing down of the layout we took all of the rolling stock on our layout and formed on long train. It totalled 143 cars. We tried various combinations of Kato and Atlas engines but they all would slip and lose traction. I had an ABBA set of split frame LifeLike FA/B1's that were draw barred together and we backed them up to the train. We coupled them and those engines just walked off with the whole thing without slipping or loss of traction. The layout was in the form of a modified "j" and had seven 3 foot corners and one 4 foot corner. Most of the time the train occupied six of the corners and sometimes occupied seven corners. We made six orbits then had to pack up. I would have liked to remove the two B units but that was not possible in the short amount of time we had. I do think it is possible to have conventionally configured steam locomotives with good pulling power without resorting to traction tires. The GHQ conversions for the Kato 2-8-2 is a good example. Sure the conversion with the traction tire will still out pull the one without the traction tire but it still has a very respectful tractive effort. Another possibility is a stoker drive design where the motor and gearing are in the tender and a drive shaft extends from the tender below the floor of the locomotive cab [as a stoker would] to the drivers. This would enable filling the area in the locomotive that normally housed the motor with more weight.
