Wheel Bases

[Oregon_Steamer]

Okay, we all should know that short wheel bases have pros from fitting arround tighter radii, but what are the pros for a longer wheelbase? I know a certain type of 0-6-0 that had a 16 foot wheel base used for switching at a dock. (You can guess how well that turned out.)

[Paul Milenkovic]

According to the Alfred Bruce book, a short wheelbase can be a problem with "nosing" or "hunting" induced not by wheel-track dynamics but by unbalanced reciprocating forces from a 2-outside-cylinder locomotive.  Bruce states that the 2-6-2 Prairie type was badly affected by this, especially with the arrangement of the rods acting on the middle drivers.

A longer wheel base gives more of a lever arm for the wheels to resist being moved side-to-side this way.

[Flintlock76]

The Prairie type also had problems at speed due to the two wheel pilot truck, again causing "nosing" and "hunting," which head-end crews found a bit disconcerting, to say the least.

Eventually 2-6-2's settled into branch-line service, in some places commuter service, where like most six-coupled locomotives they performed quite adequately.

[54light15]

I can only assume that Pacifics were the answer to that problem? I have heard of the term "dynamic augment" which I think refers to hunting or is it something else? 

[Flintlock76]

Right, six-coupled performance at speed definately improved with a four-wheel pilot truck, but locomotive builders should have known that from the 4-6-0 days.  Why they thought otherwise with the Prairie type is beyone me.

"Dynamic augment" refers to the pounding the drivers will give the rails if they and the rod assemblys aren't balanced properly.  If that's the case the only cure is a speed restriction placed on the locomotive's operation.

"Hunting" is a side-to-side shimmy of the whole locomotive at speed.

[Overmod]

Flintlock76

Why they thought otherwise with the Prairie type is beyond me.

There is nothing dramatically wrong with a properly designed Bissel leading truck provided it has proper progressive lateral combined with strong damping.  The LS&MS had some of the fastest locomotives in the world just before the turn of the 20th Century, which were essentially early Pacifics from the drivers back but with shorter and more direct drive thanks to the absence of a carrying axle between cylinders and forward driver pair.

Of course a leading Bissel is dynamically unstable, something much less true of a pin-guided Adams truck, and if the lateral 'stabilization' is provided by, say, metal springs with a characteristic rate, inadequate damping can rapidly run away.  The Reading provided perhaps the most egregious example, on a 2-10-2, so bad that only about 50 miles into the test the spring arrangement had to be gagged by welding.  (This was in the same general period that the Reading discovered precisely when, and how, pin-guided trucks were extremely poor replacements for Bissels... at the other end...)

Note that by the time of the Erie and AMC Berks, and then the N&W A, and certainly by the Alleghenies, proper definition of guiding, centering, and equalization made the two-wheel truck suitable for good speed.

"Dynamic augment" refers to the pounding the drivers will give the rails if they and the rod assemblies aren't balanced properly.

To be honest it's the pounding that you get when you DO 'balance properly' for one aspect of the 'rod assemblies' and thereby complicate another.  For example a relatively late school of practice, notably in Australia, came to recognize zero overbalance (in other words, balancing the rotating mass perfectly) had advantages for high speed on light rail.  On the other hand, toward the end of the drag-freight era, heavy rods and high piston thrust made high overbalance a necessity (in part because of severe hunting, which is a couple of nosing (yaw) and rolling) and thereby made vertical augment both more severe and sharper with increasing speed... not that much speed... right up to the point the imbalance overcame equalized weight on the axle, at which point hammer blow escalated into dramatic levels (cf. certain difficult-to-find films of certain C&NW locomotives in the late '30s).

An interesting exercise is to see how Riddles resolved this design quandary in the balancing design of the 9F class 2-10-0s, which for a very small sophistication in overbalance could reach ridiculously high speeds on low drivers.

If that's the case the only cure is a speed restriction placed on the locomotive's operation.

A better cure is to use a relatively long wheelbase and large polar moment of inertia combined with progressive lateral and relatively stiff lateral compliance exerted out toward the ends of the chassis to maximize both lever arm and travel.  A good Delta-type truck, of course, does this at the rear, but a good Bissel can do it a bit further forward than, say, a typical Adams pivoted under the cylinder block once proper progressive resistance means are provided...

[Flintlock76]

Overmod

If that's the case the only cure is a speed restriction placed on the locomotive's operation.

What came to mind when I said that was the Central Vermont's 2-10-4's.  Once the locomotive's built, it's too late for a fix, (mostly) and you just have to live with it. Hence the CV's speed restriction on their "Texas" types, 35mph, and no faster.

[Overmod]

Flintlock76

Once the locomotive's built, it's too late for a fix, (mostly) and you just have to live with it.

That's just plain silly.  A perfectly valid counterexample, of a much worse-augmented 2-10-4, is the T&P 600s; there was a reasonably good article in Trains (which Kalmbach reprinted in one of their 'free' bundles) about how the critical speed of those dogs (which were lengthened versions of the A-1 design) was increased well over 45mph.  (In case you are wondering, part of that involved the use of Baldwin Disc mains, which 610 wears to this day)

Other approaches are possible, although not fully cost-effective in modern practice.  CB&Q had an interesting approach to improving augment by adding bobweights to the inner axles of one class of locomotive, just before WW1; this could be made compatible (within limits) with cannon-box construction using roller bearings.  

Whether we like it or not, the Niagara was just a so-so medium-size 4-8-4 design until the advent of low-mass running gear combined with a valid need to make high sustained horsepower at speed.  I wish more could have been made out of the sequential rebuildings of the C&NW H class in 1941 and 1948, particularly the latter, as I think they were even more portentous than the modifications in ATSF high-speed 4-8-4 practice after a decidedly slow start.

It would have been relatively cheap and easy to have implemented a better lead truck on the CV locomotives, which were at the small end (I believe Erie 2-8-4s, among others, were bigger) had there been a perceived desire for greater peak speed -- and I'd argue that with the NH acceleration of 'named' freight service after about 1930, just such a desire might have been merited... but we all know what happened instead.

[Paul Milenkovic]

If I may offer a Bissel quibble.

According to Alfred Bruce, the true Bissel 2-wheel leading "engine" truck had this arrangement of inclined sliding planes to provide both centering force (from gravity) and friction damping.  He states that it required a lot of maintenance.  What Bruce favors as a later arrangment involves those "rocker cams" providing tailored, variable centering force.

There is a well-developed theory of those cams as there is of the ordinary rocking chair, which has to lift you up as you rock forward, otherwise the fine thing will dump you out of the chair and onto your face.

I wonder, however, how those cams are kept from just falling out every now and then, which I guess is a question live-steam modelers ask with their locomotives.

[Overmod]

The true Bissel refers to the geometry of the truck, more specifically to the placement of the pin relative to the rigid wheelbase and the arm from that pin out to the axle centerline.  The idea is to provide true radial steering, which the controlled proportional lateral (however applied) then transfers to the chassis.  The arrangement of 'heart-shaped rockers' that works for a four-wheel truck... laterally... obviously has to be modified to serve effectively for a 'radiating axle' (consider the various composite trailing trucks, and the Cartazzi axle, which purport to allow this swing but do not have proper positive resistance -- then compare to the longer-travel and much better steering advantage of a second-design Delta with sector eccentric gears or cams at the far rear corners.

Part of the fun is that the lead truck actually does several things, and for correct approximation to 'control theory' this means relatively low stiffness a fraction off straight, progressive but more than linear compensation outward from that, then capped at linear for full curve following without oscillation or excessive 'flange' wear.