Bill Withuhn book

[Lithonia Operator]

Thanks, Overmod!

A crank in the axle, and a cyinder hidden underneath the smokebox! Why didn't I think of that? 

I believe there is an illustration in the book of a Glehn-design loco with the LP clylinders aft, behind the drivers, with the main rod extending forward to one wheel set, and the HP cylinders in the normal spot forward, connected to a different set. And I think that engine was not articulated, maybe only had 4-6 drivers total. I don't have the book right here.

One really cool engine shown is a 2-8-8-8-2 with the aft set of drivers under the tender!

[Overmod]

On compounds:

Do NOT get into the lazy railfan habit of saying compounds 'use the steam twice'.  What they do is expand the steam in multiple stages, and thereby get the most expansive mechanical power out of the heat contained in the steam.

Instead of the usual 'optimization' of a simple engine, where admission is as close to boiler pressure as you can cost-effectively make it and exhaust is at as low a back pressure as you can get while making adequate draft, the compound picks a high admission and intermediate-exhaust pressure; the power from the HP cylinders is calculated and then produced as occurring over this stated pressure difference.

That means that the exhaust steam, gathered and 'flow-smoothed' somewhat via a receiver, is available as 'supply' (just as it would be from a lower-pressure boiler directly) at that admission pressure in the LP cylinders.  These normally exhaust to atmosphere with the usual considerations about back pressure and condensation, and the usual ghastly saturated-steam inefficiencies that characterized pre-superheater operation.  Some people tried to reheat the intermediate steam, to introduce some 'resuperheat' into the LP steam to match the effect of regular superheat on the HP steam.  This generally didn't work too well.  (N&W probably came closest in this country, with an arrangement that used live steam to reheat the HP exhaust very slightly (I think on the order of about 5 degrees of effective superheat) to cut some of the losses in the LP cylinders as well as providing some pressure boost.)  See the Chapelon 160 A1 boiler for perhaps the best place to put an IP reheat arrangement, and proportion it.

You could, with money enough and time, do multiple expansion on a locomotive.  The last of the D&H compound drag locomotives was this: triple-expansion from something around 450psi throttle pressure, four cylinders (at the corners of the driver wheelbase!) and remarkably efficient operation... at about 5mph.  Nifty for Red Edge firing while operating a long, long string of anthracite hoppers, but perhaps not the best for making actual ton-miles with expensively engineered and complicated equipment.  (Compare the four-cylinder replacement class, which were Challengers ...)

[Overmod]

On three cylinders:

The two you're familiar with are still there, on the outside, and now we add a third, right in the middle between the frames.  It works on a crank let into the main driver axle (there are pictures in the Tornado repair thread, among many others available). 

Instead of the cranks being 'quartered' (set at 90 degrees or a right angle relative to each other) as on a standard double-acting two-cylinder engine, the three-cylinder engine sets the outside cranks and the inside crank at 120 degrees to each other (or effectively so, as the third cylinder is often raised slightly above the 'line' of the other two, for clearance) which promises to give more even torque (google "Swiss Drive" for some of the mechanics)

England had numerous four-cylinder locomotives ... inside the strictures of the British loading gage! -- and here you have two cylinders inside the frame and two outside.  A compound version of this is the de Glehn-du Bousquet arrangement, where the LP cylinders usually rode forward (driving on the leading coupled axle) and the HP cylinders were further back, out to the sides, driving on the second driver pair. 

In case this doesn't make your head hurt enough, there were tandem compounds (ATSF was notorious for them) which put the HP and LP on a common piston rod to a typical outside-rod arrangement -- even when designed for reasonable maintenance (a few of the ATSF engines actually had provision for a crane to lift cylinders up and swing them out at maintenance time!)  You can imagine the augment effects from this sort of thing.

American compounds could do some weird things.  Look at examples of the first kind of Vauclain compound, where the HP was coaxial with the LP on each side, one over the other, with the piston rods driving on a common crosshead.  (Usually the HP was on top, but this could be inverted for more ground clearance).  The fastest locomotive in the world in 1894 was built this way.  Meanwhile, the second kind of Vauclain compound, and the Cole compound and some other developments of the immediate pre-superheater period had all four cylinders in a line across the front, which could result in some very interesting arrangements to get the inside ones cranked onto a main driver axle while clearing the one in front.  (In some cases the rod was made in two bolt-together pieces; in some cases the leading axle was actually bent to clear the angle of the rods as the axles turned)

Finally there was Bulleid's Leader, which had three cylinders side-by-side inside the gage.  Naturally they were small, and had sleeve valves requiring a minimum of valve-gear width.  All three drove cranks in the 'middle' driver axle, and the axles before and behind were connected by Morse roller chains (in Bulleid's favorite oil-bath arrangement)

[Lithonia Operator]

I recently bought this extremely impressive, and very informative book by the late William Withuhn:

American Steam Locomotives: Design and Development, 1880–1960

I'd like to comment on the book, and I also have a question the book has spawned.

I am sure that this book will become an instant classic, and be revered in the future as a must-read, milepost scholarly work on steam. It's an epic accomplishment, no doubt, and a great addition to my library and probably yours, too.

But although I am enjoying it thoroughly, I must confess that quite a bit of it is Greek to me, as they say. There are many references to engineering and physics concepts that I just don't grasp. I am not a technical guy, and this is a technical book. I knew that going in, but it is much more technical in parts than I was expecting. (There is an entire chapter solely on the subject of counter-balancing driver wheels!) I would say that if you have an engineering or physics background, OR if you are a person who is technically-minded, OR if you are a person who is simply something of an expert on steam locomotives already, you won't have any problems.

However, if you are more like me, who is drawn more to the human and operations side, and the romance and "poetry" of railroading, you too might be left scratching your head at times.

I expected more of an introduction to how steam locomotives work, for the lay reader, and that isn't there to any real extent. So I have now ordered this book:

How a Steam Locomotive Works

(used, as it's out of print), to get me more up to speed. When it arrives, hopefully tomorrow, I will take a break from Withuhn, and devour it. It's only 96 pages. Interestingly, Withuhn wrote the intro for the Parker book. Then I'll re-join Bill.

With the above caveats, I heartily recommend the Withuhn book. Even when I don't get all the details and concepts of something, I still come away understanding the general gist of many things I had been totally unaware of beforehand. It's a very well written and edited book, IMO. And often quite fascinating.

Now the question. In passing he refers to three-cylinder steam locos. I can't get my head around that. Two cylinders on one side, one on the other???

There is a chapter on compounding, which I came to learn, is about using cylinders - two, or even sometimes three - in series, to maake use of otherwise-wasted steam. At first when I read "three cylinders," I thought he meant per side. And in fact, some such engines were built, apparently, although most compounding was done with two cylinders, usually two on each side; or with articulateds, four. In any event, even numbers. Symmetry.

But reading carefully, I don't think he was referring to an engine with triple cylinders on each side. I think he was referring to an engine with three cylinders, period.

If someone could explain this, I'd really like to understand this. I simply cannot picture such an engine. A link to a clear photo or illustration of a three-cylinder steam locomotive would be most appreciated!

Thanks!