• A couple of technical questions...

  • Discussion of steam locomotives from all manufacturers and railroads
Discussion of steam locomotives from all manufacturers and railroads

Moderators: Typewriters, slide rules

  by Allen Hazen
 
Thinking abut some late steam locomotive designs, a couple of questions occurred to me:
(i) About the sides of the firebox. There is water between the inner firebox (the metal compartment containing the fire) and the outer shell of the boiler. With wide fireboxes (various late big steam locomotives had grates nine feet wide), there is a limit to HOW MUCH water space there can be along the sides of the firebox (I think in some of these same late designs, the lower firebox was at least tied for being the widest part of the locomotive). I remember reading (so: since I don't remember WHERE I read it, any references would be welcome!) that some of the Norfolk & Western's late steam locomotives (A? J? Y-6? all of the above?) had unusually thick water spaces here: this was praised as a design feature. I assumee the idea is that you getter water circulation around the firebox if the space the water circulates in isn't too narrow, and that this improves heat transfer from fire to water, and so boiler efficiency. Does anybody here know more abut this issue?
(ii) About boiler pressure. There are losses in the steam pipes. Does anybody know (a) a general estimate of how the boiler pressure compares to the pressure of steam at the inlet valves of the cylinders and/or (b) how different designs compare in this respect? (The importance of the issue was recognized: late steam designers payed careful attention to "internal streamlining," to prevent "wire-drawing" of the steam in its passage from boiler through engine.)

Any information, bibliographical or otherwise, would be appreciated. Even if a bit tangential to the questions as I have asked them.
  by Pneudyne
 
Hi Allen:

Re your item (ii), I think that this excerpt from Bruce (*) provides at least a broad-brush answer.

Bruce pp.390,391.png

On item (i), I don’t recall seeing anything substantive on the size of the water space around the firebox. We could use mudring width as a proxy for water space width, although the latter usually steadily increases above the mudring. The width of the mudring width seems to be a rarely quoted parameter in specification tables. It is sometimes shown in boiler diagrams, but not regularly so. A cursory look through the data on hand found the following numbers:

5 inches Pennsy 4-4-4-4
6 inches DM&IR 2-8-8-4, UP 4-8-8-4, NYC 4-8-4
7 inches N&W 4-8-4

Based upon a single example in a very small sample, one might say that the N&W opted for a slightly wider mudring than was common practice.

Steam locomotive boiler parameters seem to be variable over quite wide bands, no doubt with much interdependence, within which satisfactory performance can be obtained. Really, they have to be, as the boiler has to fit the chassis on to which it is mounted, whose dimensions in part are determined by the chosen wheel arrangements and driving wheel diameters. One might say of boilers that in unconstrained applications, “short and thick” is better than “long and thin”, but locomotive boilers have often, of necessity, veered towards the “long and thin” end of the spectrum.

In respect of mudring width, I suspect that there is a point where the gains from further increases become incremental, and do not offset the debits, such as the need for longer staybolts. I am not surprised that the N&W used what appeared to be a wider-than-customary mudring width. In some ways it [the N&W] operated towards the boundaries of parameter bands, and (heresy alert!) whilst that was fine in situ , I don’t think that its practices would always have been a prototype for wider application. For example, I doubt that the very stiff lateral characteristics of its 4-8-4, necessary to gain stable riding at very high speeds given its relatively small (for a fast passenger locomotive) drivers, would have been acceptable on roads where this type of locomotive had to be “jack of all trades”. Nor might the very high piston speeds have been liked by roads whose passenger locomotives habitually operated for extended periods at around 90 mile/h. The N&W also pushed the limits on factor of adhesion, and for example I expect that its 2-6-6-4 might have been rather slippery in for example typical Wyoming winter conditions.


(*) A.W. Bruce; The Steam Locomotive in America: Its Development in the Twentieth Century; Norton, 1952.


Cheers,
You do not have the required permissions to view the files attached to this post.
  by Allen Hazen
 
Thank you! (Again!)
Thinking about it: will respond in due course.
The 7 inch mud ring from the Norfolk & Western is what I remember (from wherever I read it). It's remarkable, since these locomotives (I'm pretty sure) also had nine-foot-wide grates. (The J boiler was similar in size to the ATSF 3776 boiler, with a 108 sq ft, 9 foot by 12 foot, grate.) Add in the thickness of the firebox wall and of the outer boiler shell and the lower firebox is getting about as wide as the U.S. loading gauge permits!
What got me thinking about it was wondering about the Pennsylvania Railroad S-2, the experimental 6-8-6 steam turbine locomotive. It had a 120 foot grate area, but with the six-wheel trailing truck the grate was long rather than wide: 15 feet long by 8 feet wide (so: same grate width as, say, a New York Central Niagara). And it occurred to me that, IF there is an advantage to having a wide water space beside the firebox, the S-2's configuration (or the 4-8-6 configuration Lima proposed for their "Double Belpaire boiler" design) would have allowed full advantage to be taken of it.
  by AllenHazen
 
Another couple of data points...
I recently found (the only pure model railroad store left in the province of Alberta -- "Trains & Such" in Calgary -- has a corner with old books...) some early volumes of the "Trainshed Cyclopedia" reprints of material from ancient trade publications.
"TSC" #9 (Newton K. Gregg Publisher, 1973), "War and Standard Locomotives and Cars (1919)," contains a 25-page description (apparently from the "1919 Locomotive Dictionary and Cyclopedia") of the U.S.R.A. standard designs of steam locomotive. This section has a lot more detail than most Locomotive Cyclopedia descriptions I have seen-- detailed diagrams of clearance envelopes, and of wheel-bases and per-axle weights, and tonnage rating graphs.
(Some of the descriptions of individual types seem to have been re-used in the 1922 edition of the Locomotive Cyclopedia, reproduced in "TSC" #1 (N.K.G. Pub, 1972).)
---
Anyway... Unlike every other steam locomotive description I recall seeing in these Locomotive Cyclopedia (and "Railroad Mechanical Engineer") reprints, the tabular data accompanying some of the drawings includes information about the water space around the fire-boxes. Both the Light and the Heavy 2-8-2 boilers are listed as having 5" water space on sides and back, 6" on the front. The Light 2-10-2's description says "Front, 6 in.; Sides and Back, 6 in; 5 in." (I suspect that one may have been mis-edited...). The Heavy 4-8-2 says 6" Front and 5" Sides and Back.
... So it looks as if the U.S.R.A.'s locomotive design team (It would be fascinating to know how that worked: how many engineers from which companies were involved, whether they all got together in a hotel to hash things out and how long it took...!) aimed, at least for the larger types, for 5" Back and Sides, and 6" at the throat.
... Not that different from the 5" Pneudyne reports for the PRR T-1 and the 6" for the NYC S-1 and the big DM&IR and UP articulateds. (Possible straw in the wind: the NYC Niagara had the reputation of being a very refined, technologically cutting edge, design: so maybe Kiefer and the Alco team thought more water space around the boiler was a good thing.) And the N&W J continues to look like an out-lier. (Did the locomotive designers outrank the civil engineers in the Norfolk & Western corporate hierarchy? Keeping axle loadings down doesn't seem to have been a priority in Roanoke!)
  by Pneudyne
 
The Railway Age, 1945, September 22 article on the NYC Niagara provides a bit more detail on mudring width. It was 5½ inches at the back, 6½ inches at the front, and 6 inches at the sides. By way of comparison, the corresponding dimensions for the UP FEF-1 were 5, 6½, and 6 inches.

That article suggested that the Niagara was a derivative of the NYC L4 4-8-2. But Alco’s input probably drew from its work in developing the UP FEF-2, and perhaps the D&H K-62. The latter has also been cited as an inspiration for the Niagara. Interesting though is that NYC settled for the earlier Alco lateral motion arrangement as on the FEF-1, and not the full Blunt deal as on the later UP and on the D&H locos. In thermal and performance terms, the Niagara was probably another refinement increment beyond those earlier Alco 4-8-4s, although still with a driving axle load under 70 000 lb.

I suspect that the mechanical vs. civil battle played out on quite a few roads – it certainly happened here in NZ – with mechanical winning more often. Track distress usually takes a while to show up, and can be blamed on other causes, such as poor track gang performance. In the interim, the mechanicals have returned glowing reports of the performance benefits and cost reductions attributable to their latest machine, such that the civils now have an uphill battle persuading management if they think it is the cause of their problems.

Still, there would seem to be more finesse involved in designing a large, fast 4-8-4 within an axle loading of no grater than 70 000 lb, rather going the “brute force” route and letting it lie where it falls at some higher level.

Re the USRA design process, Huddleston gives some, but by no means a full insight in his book “Uncle Sam’s Locomotives”.


Cheers,
  by Engineer Spike
 
It seems to me that the length of time to heat a large volume of water might have been detrimental. Other factors come to mind like water circulation through the boiler, Maybe the size and intensity of the modern (now about 80+ years old fireboxes offset this.