A couple of questions about NYC's Mohawks (Particularly the L-4)

[Aurora SL 1]

In the book Loco Profile #20: The American 4-8-4, author Brian Reed makes this observation as he begins his discussion of the Niagaras:

'The Niagaras came late, but with a high pedigree. Here was a 4-8-4 that grew out of a 4-6-4, though through an intermediate 4-8-2 stage that was scarcely up to the superb standards of predecessor and successor, [my emphasis] though the last 4-8-2s, the L-4, did attain 120 000 miles a year.' 

And

'All that made the Hudsons so outstanding was put into the 4-8-4; all that had dropped the standard of the Mohawks was left out. [My emphasis].Would that Kiefer had gone almost straight from the J-3s to the S-1b...' 

I have to admit, that I don't know much about the Mohawks but Reed's observations strike me as pretty blunt. So, my questions: 

* Was there a 'problem' or shortcoming with the Mohawks? 

* Is Reed more suggesting that with 550 Mohawks already on the books, the road didn't need another 50, in the form of the L-4s? 

 

[Overmod]

Aurora SL 1

* Was there a 'problem' or shortcoming with the Mohawks?

No, not really, although the 'real' Mohawks only really commence with the two higher-speed rebuilds (2998-2999) and then the L3s in 1940.  The vast number of older 4-8-2s were really little better than high-augment drag freight engines, and one reason for the failure of the first NYC 4-8-4, the H S1a, was designing it as a more thermally-efficient drag follow-on rather than explicitly as a 'dual-service' engine with good acceleration followed by reliable speed as fast as normal trains needed to go.

As with the PRR M1as, the other great 'road not taken' with 4-8-2s that could do most of the business, 72" was perfectly adequate; the correct experiment was even tried with lightweight rods on two of the L4bs (3144 and 3148) which would in many respects have made them less extreme versions of the Niagara.  Richard Leonard has pointed out that NYC didn't really need more than a 2-wheel trailing truck net of all improvements applicable to the last L-class engines; the pin-guided lead truck made them as stable as any passenger engine.

We are fortunate that Tom Gerbracht is bringing out volumes 2 and 3 of "Know Thy Mohawks"  which will bring out the story in likely full detail.

Part of the "problem" is that the Niagaras were so dramatically famous with their 79"-driver upgrades, in what was still a volume-guaranteed long-distance sustained-high-speed express service, that there was little interest in 'pushing' a new set of Mohawks or Niagara-like rebuilding of more of them until it was 'too late' for Kiefer to get approval; he had to be content with the A-2-A Berks which were a splendid design as far as they went, but optimized around 2-8-2 specs, with built-to-a-price omission of roller bearings and 63" drivers those would never be effective in fast freight ... which is a pity for what was a pretty good design for December 1946 but a disaster only a couple months later (by which time P&LE had conclusively recognized that even first-generation diesels were vastly preferable to any reciprocating steam).

[Aurora SL 1]

Thanks for the detailed response, Overmod, fascinating to read. I hadn't realised that the first couple of series of Mohawks were closer to drag-freight locomotives than dual-purpose machines. 

On your comments about driver size, I can't recall where I read it (it may have been in the book by Reed quoted in my original post) but I have read that the 79" drivers on the Niagaras were superfluous as the 75" drivers on the prototype were more than adequate, even at the highest speeds. 

[Overmod]

Aurora SL 1

I have read that the 79" drivers on the Niagaras were superfluous as the 75" drivers on the prototype were more than adequate, even at the highest speeds. 

Precisely the opposite was Kiefer's and NYC's conclusion: the prototype was rapidly uprated to 79" and no subsequent Niagara was built with the lower drivers, nor was any retrofitted with smaller drivers even though the capital cost to do so would be relatively limited (much of the expensive rod and pin equipment being the same, for instance).  At least some of this must be associated with the engines being used in high-speed service so extensively in the early part of their lives, but I would also note there is no tendency for ATSF to reverse the 80"-driver conversions of 3751 class engines when better counterbalancing solutions permitting smaller drivers -- some of which, like angling, were developed to sophistication on that road.

Remember that keeping machinery speed limited was more important than 'enhancing 'tractive effort, slip recovery, etc. merely with mechanical advantage and slightly greater stroke frequency at a given speed.  People sometimes tend to confuse high drivers with increased speed; they need to look at locomotives like Golsdorf's 2-6-4 to understand why that doesn't need to be the only, or even a significant, reason.

[Aurora SL 1]

Overmod

Remember that keeping machinery speed limited was more important than 'enhancing 'tractive effort, slip recovery, etc. merely with mechanical advantage and slightly greater stroke frequency at a given speed.

Yes, I see, I hadn't considered that at all. I had always associated larger drivers soley with higher over-the-road speeds, as opposed to limiting machinery speed. 

[daveklepper]

But then why the driver size on my overall favorite locomotive, the N&W J?

[Overmod]

daveklepper

But then why the driver size on my overall favorite locomotive, the N&W J?

We've been over this fairly often.  Periodically it gets trotted out in some context, for example eith Withuhn conjugated duplexing in the early '70s and then again when his book came out recently.

The J is one example of optimizing locomotive balancing vs. providing adhesion and power.  It does this in thoroughly sensible ways, and the result can be acknowledged effective in real-world testing; on the other hand there is not free lunch' in other respects, most particularly in machinery speed and inertia concerns of piston valves, which were likewise evidenced in real-world testing.

In primitive days the idea of 'diameter speed' caught on in the English-speaking world: by a happy coincidence the highest speed a locomotive could conveniently reach was similar to its driving-wheel diameter.  Better methods including dynamic wheel balancing cane into use in England, but the generally-acknowledged beginning of scientific augment reduction only became 'public' with Eksergian, in 1928.  Some of the follow-on results, such as the general dodge of small piston diameter combined with what seems suicidally long stroke (and angular rod displacement) practices notably at Lima, took less than a decade to become recognized practice, and some countries (Australia notably among them) started experimenting with very low or even zero overbalancing using 'other methods' of addressing the resulting yaw couples.

Glaze's balancing on the relatively ponderous J class is interesting precisely in how he addresses the overbalance issue.  Some of the 'conventional wisdom' held that the J design has low overbalance with the yaw handled with much stronger and stiffer lateral compliance from lead and trailing trucks ... this is intellectually interesting and I myself subscribed to that faith for a while.  It turns out that the 'current' balancing on 611 is far from zero net overbalance (there is something like 2200# in the coupled wheels) but there is almost none in the main driver (and what is there is proportioned to the calculated peak vertical component of piston thrust, about 80# as I recall).  This implies that the main driver will have little tendency to bounce or oscillate, but that effects such as surge or developed hunting will continue to be ameliorated simply via use of overbalance as covered by Johnson et al.

Effective nosing is of course partially addressed by high effective polar moment of inertia and long rigid wheelbase; fancy lateral motion for better guiding or less leading flange wear can detrimentally affect the 'effective' rigid wheelbase and engine length... in which case you go to larger tenders, 'detuned' coupling and radial buffer, and perhaps Langer balancer(s) to deal with any inertial surge considerations and then invoke stiff lateral, etc., to deal with leftover yaw...