Compound Mallet Question

[Anonymous]

Mark -

You state that UP's financial results were difficult to break out of all the intricate corporate structure and that's true. But the statistics such as GTM/TH and operating ratio wouldn't have anything to do with the oil and coal operations, and would provide some comparisons.

Paul Milenkovic - Baldwin beat Porta to the skinny-boilered 2-10-0 concept by about eighty years. They were selling light-axle-loading Decks to short lines beginning in the late 1920s. Strasburg 90 is a living, breathing example and several that were owned by the Gainesville Midland in Georgia still exist, although none operate. The old Russian Decapods were even earlier examples of the concept, and the IRM at Union, Ill. has one of those that has operated in the recent past if it's not running right now.

Old Timer

[Overmod]

Might also be mentioned that the British 5AT project, to develop an advanced modern locomotive, will be using a narrow firebox...

But typical American practice used wide, and sometimes long, grates to give the desired combustion area... and most late European practice, especially where coal quality was lower, tended to conform to this. Wide fireboxes are, I think, somewhat easier to fire heavily, as there can be more 'heel' at the back and better loading, shaking, etc. of the grates at sustained hp output. Wide fireboxes generally require a trailing truck for stability -- engines like PRR G5s that put a wide grate above the drivers tend to have decidedly poor riding qualities!

Will Woodard decided, I think correctly, that there would be benefits in articulating a long trailing truck under a BFF (that's a big firebox to those with gentle minds ;-}), but somewhat less correctly he essentially tried to run the locomotive drawbar through the truck frame -- this produced some rather interesting weight-transfer and suspension issues! What got used instead was progressive weight loading (sometimes through actual oval gear-toothed rollers or roller segments!) on the rear of a pivoted trailing-truck frame; to my knowledge, this provided sufficient lateral guiding on the trailing-truck frame for whatever high speeds the engines themselves could reach in service. (I have my doubts that Fabreeka springing would 'live' effectively in trailing trucks due to the heat due to the proximity of the ashpan and grate; a number of roads chose not to use roller bearings on trailing-truck axles for I think this same reason.)

It's possible (as on the 5AT proposal) to use a modified Franklin radial buffer to snug the tender up against the locomotive, which essentially uses the first tender truck as if it were a trailing truck to control and guide the motion of the frame and drivers of the locomotive. This should dramatically relieve problems like those encountered on the G5s.

Paul, the question about the leading truck requires a somewhat more detailed answer, and a bit of 'what-if' imagineering.

Traditionally, the pin-guided four-wheel truck has been used on high-speed power, and the 'Bissel'-style two-wheel truck on freight power. The reasons have to do with geometry. The Bissel truck is inherently dynamically unstable in yaw ... the further it deflects, the greater the deflecting force on it, as opposed to the inherent restoring force on the same geometry when the axle is behind the pivot, as in a Delta trailing truck. That means that without lateral compensation the 2-wheel truck will develop high forces and perhaps sudden oscillations or instabilities without warning at higher speed... and most traditional 2-wheel leading trucks only used some variant of weight or friction to control their lateral movement. In the pin-guided truck, thrusts from the leading and trailing wheels 'balance out' to keep the truck frame aligned with the track, and the lateral motion of the pin relative to the locomotive frame (which does the actual guiding of the locomotive proper in curves) is not explicitly tied to yaw moment. However, this geometry is NOT what you want under the end of the locomotive behind the drivers, as there's no inherent means of damping oscillations or hunting of the pin-guided truck when constrained to follow the frames... the Germans tried this approach with some of their 'bidirectional' express locomotives and had to limit top speed to about the equivalent of 81mph as a result of rather substantial oscillation and other problems that even their engineering couldn't cure...

The Reading had a couple of sad failures that point up both these tendencies. A progressive engineer thought a good cure for Bissel instability would be to mount a pair of heavy lateral springs to keep the truck aligned with the frame by default -- good thinking, as far as it went: the springs would provide the desired kind of restoring force, but would be comparatively neutral for small-arc truck movements. Unfortunately, nobody seems to have recognized that with no damping other than friction, the springs would build up oscillations -- the 2-10-0s with these devices became notorious for hard riding, and apparently one guy got so fed up that he welded the spring gear up so it didn't "cushion" -- wow, instant improvement, I wonder why...

Meanwhile, at about this time, the Reading built a 4-4-4 with symmetrical pin-guided trucks under both front and rear... a nice improvement in capacity for big Wootten fireboxes over what an Atlantic's single axle could support at reasonable track-saving axle loading. It did not take them long to remove it, and the history I've been able to find is mercifully silent on their experience convincing them to do so.

The correct approach was at least tried by N&W, which was to use composite springs, similar to those used on pedestal tenders, on the leading-truck axle. These "Fabreeka" springs are sandwiches of rubber blocks and metal sheets, similar to the 'chevron' springs seen in some of the high-speed trucks of a quarter-century ago, and have very good lateral restoring force coupled with very high inherent damping. On pedestal tenders, they are used to locate the (otherwise laterally floating) tender axles, to give fully flexible lateral motion on curves and during equalizing, while providing a fixed location for the actual vertical suspension to bear upon. In the two-wheel leading truck, they allow a degree of highly-damped lateral motion, and consequent good guiding, for small lateral movement, and also damp any oscillations that may develop on curve entry and negotiation. Further angle of the truck in curves can be accommodated by a progressive weight-transfer system, or by variable-wound or other variable-rate springs accompanied by dashpots or other appropriate forms of damping. There is no theoretical reason why such a truck wouldn't be capable of equal or better stability compared to a 'typical' pin-guided lead truck (which normally doesn't have anything other than accidental friction and, to an extent, viscous damping about its axis of rotation)

Hope this is helpful and relevant to what you were asking!

[Anonymous]

O, I guess in short you are concluding that a two wheeled leading truck can be just as good as a four wheeled leading truck given proper damping.

One advantage that I can see that a four wheeled truck might have over a two wheeled truck, is that any movement or forces transmitted to the frame from the truck is an average of the movement and forces acting at the extremities of the truck.

It seems that trailing trucks have a much longer distance from the axel center to the point of connection to the frame. It seems that would increase lateral movement for a given curve, possibly a bad thing? Or is this distance actually shorter? When compared to a leading truck's distance. It seems in either a leading or a trailing truck, a long reach to the trucks center of movement would increase lateral movement and also increase forces transmitted to the frame for a given curve?

[M636C]

Paul, and Overmod,

Most two axle leading trucks used a swing link arrangement to increase the vertical load on the truck proportional to its displacement. This had an inherent problem that the load on each wheel varied with displacement, and this could lead to derailment.

On the British London and North Eastern Railway, a large number of 2-6-2 locomotives of class V2 suffered from problems due to their swing link trucks. When he took over from Nigel Gresley, Edward Thompson rebuilt many of these locomotives with spring centred trucks, using the same centring action as most four wheel bogies. Thompson did substitute a 4-6-2 design, class A2, for new construction however.

In Europe, two designs of truck were developed that coupled a single leading axle with the leading coupled axle, allowing the usual form of leading bogie location and centring to be used.

The older design, the Italian "Zara" truck, used plate frames in what was a conventional two axle truck, except, of course that the wheels in forward and rear wheelsets were greatly different in size. Italian steam locomotives used plate frames, and those on the Zara truck were sufficiently narrower than the main frames to fit inside them, and the main frames had no direct attachment to the leading coupled axle. The most common Italian express locomotive, class 685, was a 2-6-2 with a front Zara truck.

The later design, the German "Krauss-Helmhotz" was a different design more suitable for bar frames, used widely in Germany (as in the USA). The truck had the usual arrangement for the leading axle and the pivot , but at the leading coupled axle the truck was connected to a pivot below the leading coupled axle, which was carried in conventional bearings in the main frames. These allowed additional lateral movement to allow the axle to act as the trailing axle of the bogie.

These Krauss-Helmhotz trucks were used on most German standard locomotives, including the very large number of 2-10-0 War Locomotives of classes 52 and 42, which were not intended to run at high speed. They were also used on 2-6-2 class 23 and 2-8-2 class 41 which were used for passenger traffic.

In this respect, European design was well in advance of British and American designs.

Peter

[Overmod]

Just as a note, Krauss-Helmholtz (note sp) was a reasonable approach to guiding... in Europe. I think you meant that bar frames, not the bogies, were "used widely ... as in the USA" -- I don't know of any successful use of a Krauss-Helmholtz in American practice (lateral-motion devices being used instead). There isn't doubt that they worked nicely in a wide variety of European locomotives, some operating at considerable speed (some Golsdorf locomotives come to mind)

Note what is required for the siderods going to the driver axle in one of these bogies. Even spherical-pivot roller bearings won't be happy with such loadings. I have seen a couple of frankly ingenious arrangements of rods and levers that can, in theory, maintain the rods in quarter for a hinged axle, but again none of these made it to heavy US practice.

We might also mention the Cartazzi axlebox, which implemented lateral motion via horizontal curves in the boxes and pedestals that matched the 'swing' of the axle on curves while maintaining reasonably tight longitudinal tolerance.

jruppert, you're correct in presuming that the swing at the rear end of some trailing trucks could be extreme. In practice, this didn't matter in most cases, since the bearer connections between truck and locomotive frame at the rear didn't have to be all the way at loading-gauge clearance, and devices (such as toothed rollers) could easily accommodate substantial amounts of swing and still maintain weight transfer. The longer the distance from pivot to leading axle, the greater the 'self-centering' tendency and resistance to instabilities with the engine running nose-forward... and, in some cases, the added length was needed to position the axles correctly under a longer firebox in order to get the axle loadings correct.

[M636C]

Overmod,

Sorry about that! I've been out chasing trains and I'm too tired. I'm sure I hit the "L" in Helmholtz both times, but didn't notice that it didn't appear. And I did mean the compound sentence about use in the USA to refer to the frames and not the trucks.

The Cartazzi worked well enough in England, but was a complete disaster in India, combined with insufficient initial spring compression in the bogies. E.S.Cox, who accompanied W.A.Stanier to India (basically as a French translator) has described the problem very well.

I think the problem with the 4-4-4 "Reading Type" was that the trailing axles in the Atlantics were virtually rigid (compared to the bogie), and provided a much longer effective wheelbase (much the same problem as the Indian "Pacific" locomotives).
The firebox of the 4-4-4 was shortened by 12" on conversion to Atlantic, which may also have helped by reducing the rear overhang and the mass distribution.

Peter

[Overmod]

Do you have references, preferably to an online-located resource, for E.S. Cox's observations on Cartazzi problems?

My understanding of the Reading 4-4-4 was that it was an "improvement" on the existing Atlantics, which dated back a long way on the Reading -- they had one in the early 1890s with Vauclain compound cylinders and 81" or taller drivers. I had also understood that most of the early Atlantics used an extension of the locomotive's rigid frame to position the trailing axle... but that the axle had substantial lateral motion. Some of the earlier 'built-up' trailing trucks, for example the Cole type that Alco used, were IIRC much more intended as devices for controlled lateral motion than for radial swing from a pivot, were they not? At least in older and presumably slower locomotives, as soon as you have lateral motion, the more significant of your 'rigid-wheelbase' worries do not apply, and some other advantages (such as easier equalization) apply. It's my opinion that only a substantial cast-steel frame makes a Delta-type trailing truck practical...

I think what you're tacitly getting at in the Reading discussion is that the design had a high ... probably much TOO high ... trailing polar moment of inertia, relative to the short driver wheelbase. I can re-create some of the 'logic' behind using a pin-guided truck back there: the lateral spring guiding would inherently provide control of the swing excursion of both front and rear of the locomotive dynamically... and this was known to occur 'correctly' on a wide variety of successful locomotives in practice, on the front, so why shouldn't it do so on the rear...

Now, I don't know whether the Baltic locomotive (of 1912), which essentially used this approach on a six-coupled locomotive, was an inspiration to the P&R for their locomotive. Nor do I know whether the Baltic itself was successful in achieving good guiding -- although it seems memorable enough that Milwaukee thought to apply the name to its earliest true-4-6-4 designs (to this day, a Hudson is a 4-6-4 with a hinged trailing truck, and a Baltic is a 4-6-4 with pin-guided trailing truck). Seems to me you'd get better results with the longer driver wheelbase for a given spring rate and strength in the lateral springs for the bogie pivots.

I also think that the problems with pin-guided trailing trucks can be solved by using weight-actuated outside bearers, or for that matter other forms of bolstering. The principal reason to avoid them is that they interfere with the ashpan, and with any structure under the firebox area, and (imnsho) don't provide any benefits whatever compared to alternative types with proper detail design, in almost all prospective service.

Must say I wi***hat I could have been chasing trains too!

[Anonymous]

For a good writeup on the Reading 4-4-4 see Fred Westing's APEX OF THE ATLANTICS - his history of the Pennsy E6 (one of the greatest locomotive books ever written, BTW).

It seems that neither the leading or trailing trucks of the 4-4-4 were equalized with the drivers. This allowed loss of adhesive weight when the drivers were in low spots, and overloading of the driver axles when they were in high spots. It also contributed to instability at high speeds. The 4-4-4s were all rebuilt into conventional 4-4-2s with the trailer equalized with the drivers.

None were ever equal to the E6, which, interestingly, had the lead truck equalized with the #1 driver and the trailer equalized with the #2; the two drivers were not equalized together.

Old Timer

[daveklepper]

Agree with the assessment of Fred Westing's book. And note the extremely thorough testing of the E-6 before serial production. Compare that with PRR's Metroliners, Amtrak's Acela, the Boeing LRV, and the most recent Breda light rail cars in Boston.

But back to locomotive truck design. Is one reason the various European designs mentioned never were tried in the USA the difference in quality and design of the track structure?