In the nineteenth and early twentieth century steam locomotives that were designed for high-speed running were almost exclusively built with a four-wheel leading truck for high-stability. This trend started to change in the 1920’s however when increasingly more locomotives were designed for high-speed running with just a two-wheel leading truck.
Can someone explain the advances the occurred to allow this to take place? What exactly were the technology advances to allow a two-wheel leading truck to be remarkably stable at high-speeds?
Keep in mind that this is a strictly relative thing. While it is certainly possible to design an effective two-wheel lead truck for ‘true’ high speed, it’s still geometrically unstable (perhaps a better term would be ‘metastable’) meaning that anything that perturbs the truck laterally has an increasing tendency to keep it turning further and further. This is particularly concerning if lateral resonance of any kind develops (think about what happens if you’re walking with an appliance and the cord starts to bang back and forth between your calves) and, while there were some interesting experiments in damping some of these tendencies in primary ‘lateral’ suspension (most notably with Fabreeka pads and later ‘chevron’ composite springs) you are still left essentially with (1) three-axis progressive control, and (2) effective three-axis damping of developed motion, in order to make high speed practical.
A good place to observe the ‘right’ approach is the lead truck of preserved N&W A 1218 in Roanoke. Here you can see, for example, how the equalization weight is provided right over the axle, via a single center-mounted bar, so that lateral roll of the axle doesn’t selectively load and unload wheels and promote dynamic couples in the pivot. It still depends largely on high applied weight preload and friction for its ‘damping’ however, which works reasonably well in vertical accommodation up to a point and provides reasonably strong progressive correction for deflections far from ‘center’, but does little for small-period oscillation off ‘center’, particularly in the regime before obligate flange-bearing commences (which it does fairly rapidly in most steam locom
As to “what changed” to allow higher speeds with single-axle lead trucks, Semmens and Goldfinch “How Steam Locomotives Really Work” claims that a change from “swing links” to a spring to provide the centering force helped avoid the derailments blamed on the lead truck. They cite the example of the British V2 class, a 2-6-2 that we would call a Prairie type but in Great Britain, they just called it the “V2 Class.”
The explanation offered by Semmens and Goldfinch is that you need some kind of centering force on a lead truck, especially a single-axle lead truck. If you provide the restoring force from gravity acting on lateral swing links, pushing those links sideways has the effect of lifting the lead driver off the rails, leading to the derailments. When the V2 was changed over to a centering spring, they more-or-less stayed on the rails.
As to the stability or lack thereof of a particular wheel arrangement and suspension system, you would have to run a simulation in one of those fancy software packages to make sure, and even the software packages are making simplifying assumptions regarding complicated effects of friction and wheel-rail “creep.” But in later-generation steam, you could figure that based on experience, later designers had design rules to follow that gave them something that worked, but one of the things they were doing was using in some cases rather stiff centering springs.
Centering springs are such a wrong answer (except perhaps in Britain, where locomotive suspensions characteristically and often painfully lack full equalization) that it almost beggars the imagination.
There is a very famous story about the Reading experimenting with this ‘approach’ on one of their 2-10-0s, right around the time of their dramatic fiasco with building a 4-4-4 express locomotive (with a couple of those “stable” pin-guided trucks helpfully fore and aft of the short driving wheelbase – HAS to be genius, right? … oops). They were ‘smart’ enough to know that the spring had to be relatively long (similar in practice to what’s in one of those Wagner bypass valves on ATSF locomotives) and they neatly had opposing springs, nice strong ones to counteract the nosing from high piston thrust, to keep the truck on center all the time and make sure it was self-restoring through chassis guiding on curves.
What they forgot was oscillation. The damn thing slammed so badly that someone had to come out and autogenously weld the spring arrangement on both sides even to get the test over the road. Perhaps tellingly we never see this approach being tried on the Reading … or elsewhere in American practice … again.
There are actually three separate things involved here (remembering that the truck is not dynamically stable even in small-period perturbance): there’s a need to keep the truck tread-guiding for very small excursions, then a need to provide relatively stiff correction outside of that small range up to a few inches, then the need to provide strong and progressive compensation as the angle between truck and frame increases in spirals or curves. By comparison, look at the forces involved in a good Delta-style truck arrangement, where rockers or even sector gears are provided to use the weight of the back of the locomotive as the ‘restoring force’ as the truck takes up an angle and the rear moves out relative t
ShroomZed
Can someone explain the advances the occurred to allow this to take place? What exactly were the technology advances to allow a two-wheel leading truck to be remarkably stable at high-speeds?
Keep in mind that this is a strictly relative thing. While it is certainly possible to design an effective two-wheel lead truck for ‘true’ high speed, it’s still geometrically unstable (perhaps a better term would be ‘metastable’) meaning that anything that perturbs the truck laterally has an increasing tendency to keep it turning further and further. This is particularly concerning if lateral resonance of any kind develops (think about what happens if you’re walking with an appliance and the cord starts to bang back and forth between your calves) and, while there were some interesting experiments in damping some of these tendencies in primary ‘lateral’ suspension (most notably with Fabreeka pads and later ‘chevron’ composite springs) you are still left essentially with (1) three-axis progressive control, and (2) effective three-axis damping of developed motion, in order to make high speed practical.
A good place to observe the ‘right’ approach is the lead truck of preserved N&W A 1218 in Roanoke. Here you can see, for example, how the equalization weight is provided right over the axle, via a single center-mounted bar, so that lateral roll of the axle doesn’t selectively load and unload wheels and promote dynamic couples in the pivot. It still depends largely on high applied weight preload and friction for its ‘damping’ however, which works reasonably well in vertical accommodation up to a point and provides reasonably strong progressive correction for defl
I don’t know of a good source of detail pictures, but I wouldn’t be surprised to find they exist, and that someone N&W-centric like Big Jim or feltonhill can provide a link to them.
The initial point of a ‘Bissel’ truck, properly defined, is to provide a laterally-pivoting frame that has its pivot point aft of its wheels, and upon which some portion of the locomotive’s weight can be imposed. The geometry of pivot point and frame length is arranged so that the truck’s flanged wheels can remain precisely parallel to the track as the engine negotiates curves (usually this is taken as relative to the midline of the engine’s rigid wheelbase, not the first driver pair). This of course is part of the operating ‘principle’ of good radial-steering diesel-electric trucks.
Added to this is the sense of providing self-centering, as early Bissels which lack this quality are dynamically unstable – the further they hinge outward, the more they tend to diverge further, like backing a two-wheel trailer. This can be done via springs, but for a variety of mechanical reasons (some of which I gave) this is not the best way to provide centering; arranging some method of swing links (either hangers or inverted) upon which the weight of the engine can be provided is the ‘usual’ method of performing “steering” return to center, but you will recognize that any lost motion in the link mechanism will produce some “wandering” around center which can be driven into oscillation. (Prof. Milenkovic is one of the world authorities in kinematic linkages, so I defer to him both in characterizing this better and in producing better methods of ‘hinging’ a two-wheel lead truck across
Is the OP’s hypothesis even correct. Even until the end of steam, the really high speed passenger locomotives were still built with 4 wheel lead trucks. Witness the Hudson and Northern. In articulateds, the Big Boy and Challenger also had them. There were exceptions, like the NKP Berks and N&W A’s.
Depends on the definition of ‘high speed’ you want to use. By normal railroad standards even 45-50mph might be considered ‘high’ and I suspect that most freight speeds in the pre-truck-competition era, outside M&E or the organized fast-freight lines and services, was as slow as bean-counting cheapness could allow it.
Naive approaches to centering a dynamically-unstable truck were not very happy ones. The ‘revolution’ in high-speed Prairies and ‘bicycles’ was not very long in duration, despite some amazing propaganda for the likes of the LS&MS Ks; the answer of course went in two famous directions here (the Pacific being a Prairie with a four-wheel lead truck, and the Atlantic being an engine as boiler-capable as any six-coupled Ten-Wheeler but with higher-speed machinery at the time).
The effective safe limit of a weight-centered truck is high, but not very much above the 70mph or so a good AMC Berk could reach. I suspect the finest flower of this would have come with the use of the lightweight-rod-equipped N&W As in competition with the C&O steam-turbine Chessie train to Cincinnati – the thing I think they were built to provide. Perhaps there is some record at NWHS of testing of these locomotives, which could – nominally – have been as fast mechanically as the comparably-equipped J class, with less rotating and reciprocating mass per main, and which would have allowed easy analysis right up to the limits of any contemporary approach to 2-wheel leading-truck design. In particular the application of Fabreeka transverse shear ‘springing’ between axlebox and riding springs takes on added interest in such a context… as would the arrangements made to isolate weight centering of the truck as a whole from short-period motion of the wheelset itself.