Le Massena's "Big Engines" article (1968 Trains)

Trains Magazine Steam & Preservation
1

In the other thread a few guys mentioned Le Massena’s famous article in June 1968 Trains that tried to rank the largest US steam locomotives, both by power and by merit. As he commented on the engines’ merits he naturally included comments on the railroads operating those engines-- railroad A did a fine job, railroad B clearly was run by pointy-haired bosses.

He calculated the “potential power” of each engine in arbitrary units: square feet of grate area times pounds per square inch of boiler pressure. A UP 4-8+8-4 had 150 square feet of grate and 300 psi, so its potential power was 45000, putting it in second place behind the NP 2-8+8-4, which (initially?) had 182 sq ft and 250 psi.

Third place was one of those D&H experimental compound 2-8-0 from circa 1930, which had 82 sq ft and 500 psi. This raises a question, namely: Huh?

Fortunately he explained.

“Why select boiler pressure and grate area as measures of potential power? The reason is not difficult to comprehend. Going back to the high-school physics class…”

Mark that. It’s simple, he says-- no higher education required. In a couple sentences he’s going to make a clanger of a mistake, which the reader will be baffled by: could he really have meant that? I must be misunderstanding him-- it must be more complicated. But it isn’t.

“…recall that power is defined as how much energy is released or developed in a specified time. For steam, this energy is expressed as volume multiplied by pressure. Hence, steam “power” is boiler pressure multiplied by the quantity produced per hour or minute.”

I’m not sure that’s quite right, but close enough for us: power is pressure times volume of steam produced per unit time.

"Now, since the quantity of steam is related closely to how much fuel is burned per square foot of grate area, the fuel aspect can be eliminated by saying that for equal fuel consumptions [per square foot of grate], the larger grate area

2

The article kind of made sense when I first read it at the end of 8th grade, but less so after taking a engineering thermodynamics class in my senior year in college.

Just for grins, I dug out my thermo textbook and looked up the steam tables - assuming isentropic performance from an engine, one could get maybe 30% more power out of steam at 500psi & 700F versus 240psi & 700F. This implies that a more accurate formula would have been grate area times pressure to (say) 0.4 power.

  • Erik
3

The NP 2-8-8-4’s grate area was initially 182 SF, but it was later reduced to 161.4 SF because of drafting/combustion problems.

4

The question that I wanted to ask was referencing the Quality of Coal Burned in the various steam locomotives. By no means am I a technical person, so I may be overly simplistic, but it seems to me that the Quantity of coal, plus draft, and combustion creat the BTU’s needed; the quality of the available coal to be combusted would be a consideration in the final performance delivered. and the one thing missing from these conversations as well as the ones on the previous link is exactly the bearing of coal quality on the equations.

reference from this link: http://geology.about.com/od/mineral_resources/a/aa_nutshellcoal.htm

"Grades of Coal### Coal comes in three main types, or grades. First the swampy peat is squeezed and heated to form a brown, soft coal called lignite. In the process, the material releases hydrocarbons, which migrate away and eventually become petroleum. With more heat and pressure lignite releases more hydrocarbons and becomes the higher-grade bituminous coal. Bituminous coal is black, hard and usually dull to glossy in appearance. Still greater heat and pressure yields anthracite, the highest grade of coal. In the process, the coal releases methane or natural gas. Anthracite, a shiny, hard black stone, is nearly pure carbon and burns with great heat and little smoke…"

The UP’s Big Boys, as I understand it were designed to burn a lower grade of Wyoming coal. The Allegeheny was designed to burn various grades of Appalachian coal ( either Bituminous or Anthracite?). Illinois Central burned available coal from Illinois mines and Kentucky mines ( my guess it was bituminous, but I am unsure about its quality). The Katy, as I had said burned a local coal from company mines in Kansas ( it was poor quality but readily available.) Santa Fe utilized oil, and may also

5

The quality of the coal as fuel definitely had an effect. The NP Yellowstones had their enormous fireboxes because they were designed to get the most of the low-grade coal that they burned. Lloyd Arkinstall, in one of his articles about firing for PRR in New Jersey, commented on the difference between the Westmoreland County coal supplied to passenger power and the West Virginia clinker coal he was stuck with on his local freight.

6

So Le Massena calculated the “potential power” (pressure time grate area) for a bunch of engines, along with their “power” per ton of engine weight, which was his criterion of merit. Since in his view merit was roughly proportional to boiler pressure, his favorite engines couldn’t have much less than 300 psi. The SFe 3460-class 4-6-4, 3765-class 4-8-4 and 5001-class 2-10-4 all had 300 or more, and they all were at least tied for largest grate area of their type (except for the NP dirtburner 4-8-4s, which used lower pressure). So he concluded “AT&SF was the uncontested leader in three-, four-, and five-driving-axle locomotives of conventional design.” Uncontested! Nobody can accuse the guy of being afraid to stick his neck out.

His locomotive ratings didn’t seem to make much impression, but his railroad ratings have stuck pretty well-- they’re now conventional wisdom, unfortunately. Like lots of fans he figured the railroads would do better if they’d just run their trains faster:

“Meanwhile, however, a few railroads were relying more on speed than on tonnage to give themselves better operating statistics while supplying better service to their shippers. Among these heretics were Nickel Plate, St. Louis Southwestern, Santa Fe and Union Pacific.”

“Like the Santa Fe and unlike many other roads which misapplied their high-power engines, UP held down tonnages, allowing its locomotives to run at high speeds and thus deriving from them the most work in the least time at the least cost. The financial records of these carriers are convincing evidence of their unorthodox operational philosophy.”

They got better stats and lower costs by being heretical and unorthodox… we need details on that. They will follow.

7

I don’t have access to Railway Engineering (Hay), but I do remember rated boiler horsepower factors in grate area, direct heating surface, firebox and flues, indirect, superheater steam tubes in the flues, and boiler pressure. I think a factor for bituminous coal was assumed. It seemed the efficiency was measured as the proportion between pressure in the boiler and in the piston having a lot to do with steam flow. FYI, I calculated that a fairly large, later series C&NW E-Class Pacific put out about 2,500hp.

The grate area alone would yield different results than for boiler power; and La Massena had an argument for discounting the other factors in his rating system.

8

Would a measure of current fixed utility power plant’s boiler pressure give a good balance of higher pressure vs more complex piping??

9

Say we’re deciding what tonnage to give to our new 2-8-4s on a railroad where the ruling grade isn’t too tough-- 0.5 %, let’s say. We try a 3500-ton train and the engine makes the grade at 12 mph; we try a 2500-ton train and it does 25 mph over the summit. If we assign 2500 tons instead of 3500 our ton-miles per train-hour on the grade will be 49% greater.

Far as we can tell from his article, Le Massena takes it for granted that more ton-miles per train-hour means lower total cost to run the railroad. Better service for the shippers and lower costs-- what’s not to like?

Seems simple enough, doesn’t it? But not simple enough for lots of railroads, who unaccountably persisted in assigning 3500 tons to their 2-8-4s. Le Massena’s conclusion: railroads were run by blockheads.

I can’t blame you for assuming I must be setting up a straw man-- surely he couldn’t be that serenely obtuse? And maybe he can’t be-- but the article gives no hint of any analysis beyond the seeming parody given here. More quotes to follow.

10

“Lima Locomotive Works, however, was not in agreement with this method of railroading; its 1925 experimental 2-8-4 with large boiler, big grate area, small cylinders, and bigger drivers was designed to deliver its maximum power at higher speeds. Excellent over-the-road performance was in direct opposition to the prevalent maximum-tonnage, minimum-speed gospel of current popularity. Despite the 2-8-4’s immediate success, any number of tonnage worshipers refused to accept the new doctrine of speed first, tonnage second.”

The new doctrine, fortunately conceived by Lima. But I’m guessing Lima wasn’t trying to force any new doctrines on the railroads-- they were in the business of selling locomotives, and the way to do that is to build locomotives that do the job the railroad wants done. They intended the 2-8-4 to pull the same drag tonnage as a 2-8-2 while burning less fuel at drag speed and making better speed once past the ruling grade. Some railroads would have “fast freights”, and the 2-8-4 was intended to be well suited to them too-- but Lima didn’t ask the railroads to reduce train tonnage, on the “fast” freights or the drags.

The 2-8-4 wasn’t actually “designed to deliver its maximum power at higher speeds”. It was designed to deliver higher power, which would inevitably be at a higher speed than on the older engine. Maybe the old engine was good for 2500 dbhp at 25 mph and the 2-8-4 could do 3500 at 35 mph-- that’s great, but if Lima could have found a way to get the 2-8-4 to do 3500 dbhp at 25 mph they would have jumped at the chance.

11

Dear Tim,

is this correct, as I may understand you, he made mistakes on purpose? Maybe it was just a typo?

12

I do not have the article, maybe he meant running it hours for hours just on a grade, the higher HP-engine is no bonus, there.

lars

13

Timz, the Lima design is typified by the NYC A-1 class (and used also by IC and ATSF) doesn’t seem to have been as successful as the design created by the “Advisory Mechanical Committee” (Van Sweringen roads). The Lima design used larger cylinders and 63" Driving Wheels, and used Limited Cutoff to keep TE within reasonable adhesion limits.

14

It wasn’t a typo-- the full quote is

“The year 1937 represents the high-water mark of steam locomotive development and construction. During the previous two decades power-producing capabilities had doubled; of vastly greater significance was the fact that power per ton of locomotive weight had also virtually doubled. One specific example is revealed by comparing the latest Santa Fe 4-6-4 (300 pounds pressure, 99 square feet of grate, 412,000 pounds total weight) with Illinois Central’s 4-6-4 conversion from Lima’s A1 model (265 pounds pressure, 100 square feet of grate, 692,000 pounds total weight).”

Could he really have thought he was comparing engine weights?

Railfan books/articles like the later engines better, and limited cutoff sure went out of fashion after 1930 (except on SFe). Aside from that, what do we know about their success?

15

I meant to correct the reference to Santa Fe Berkshires and change it to Missouri Pacific. The Santa Fe’s Berkshires were just enlarged Mikados. The Lima designed Berkshires on NYC’s B & A amounted to 55 locomotives in one order. They were never duplicated, and were replaced by diesels fairly soon after WW2. Both Missouri Pacific and Illinois Central rebuilt their fleets of Berkshires to totally different types of locomotives, which clearly indicates that they were unsatisfactory as built.

16

I never could figure out where the “total” weight figure for IC’s 4-6-4 came from in the article. The best I have is that it weighted 405,600 lbs, excluding tender. The predecessor 2-8-4 weighed 393,500 lbs, also excluding tender. Typo? Maybe, but I doubt it was an intentional “gotcha” figure. If anyone has ever checked the complete table in the Trains article, there are several inaccuracies, perhaps due to rounding or other causes. I never found them until PC’s came into play in the 1980s and I could put the entire table into one spreadsheet very quickly.

AFAIK, Santa Fe had 10 early-design 2-8-4’s purchased from B&M - 4101-4104 and 4193-4198. These weren’t extended mikados.

17

The IC 4-6-4 weight would be very close to that of the NYC A-1 2-8-4. It’s a matter of weight redistribution, moving the driving wheels forward, in part because of their larger diameter. My guess is that the IC engine had 216,000 lbs on the drivers, 2/3rd of the difference on the trailing truck and the remainder on the lead truck.

18

As a follow-up, I see you gave a weight for the A-1 of about 393,000 lbs. That would put approximately 59,000 lbs on the lead truck (29,500 lb axle load) and 118,000 lbs on the trailing truck (59,000 lb axle load) with a 2:1 ratio. A 3:2 ratio would result in 35,400 lb and 53,100 lb axle loads on the respective lead and trailing trucks.

Didn’t the IC engines get cast steel beds and cylinders when rebuilt?

19

More quotes from the article–

“The [CB&Q] 2-10-4 was operated at speeds well below that for peak power (about 20 mph)…The Burlington 2-10-4’s big boiler, 107-square-foot grate, and 250-pound pressure, thus misapplied, were more ornamental than useful in the production of ton-miles at minimum cost.”

CB&Q got the 2-10-4s to haul 8000-ton coal trains out of southern Illinois; ruling grade against the loads may have been 0.3%, which would indeed slow them to 20 mph or so-- when they were on that grade. But most of the time they wouldn’t be on that grade, so most of the time they’d be making better speed, maybe even good enough to satisfy Le Massena. No doubt he’d prefer that they reduce the assigned tonnage, but he still makes no attempt to explain how that would lower costs.

“When the [SP] cab-forwards were used in passenger service over difficult profiles at high speeds, their performance was magnificent; yet their primary use was in low-speed freight service, three or four to a train, operating well below the ultimate capacity of boiler and machinery.”

In the 1940s SP rated its newest 4-8+8-2s at 1450 tons eastward over Donner Pass, with which they could apparently make 15 mph or better on the steepest part of the climb. Not good enough for Le Massena-- one wonders what tonnage rating he would have approved. Note that if they cut trailing tonnage by, say, 40%, they’d only be cutting total tonnage by 30%, so the increase in ton-miles per hour won’t be that great.

20

I was underwhelmed when I first read about this method of power ranking, and the more I thought about it, the more exceptions I could find, even when I held the caloric value of coal constant. (This thread’s discussion of BTU values has been very useful to explore a different area of disagreement.)

A central problem was the dismissal of the boiler. Same grate, same pressure and I was pretty sure you’d have different results with different-sized boilers.

Then I wondered about the effect of superheating the steam that’s produced? So I used the “equivalent heating surface” calculation of 1.5 x the superheater’s sq footage divided into the total heating surface. That brings the boiler into play and, as I’ll show in a moment, takes a long step toward making this a useful way to compare locomotives.

But then I thought of oil-fired locomotives - they don’t have “grates” in the sense of spreading the fuel out over a single surface. They burn their fuel in a volume bounded by the firebox. So I substituted firebox heating surface for grate area, which I think allows for an apples-to-apples comparison of oil burners with coal burners.

And, I thought, if grate area is your only measure, then clearly Wootten anthracite-burning fireboxes take the palm. Their grates are relatively huge - at least twice the area and often more compared to an equivalent narrow-firebox design. What I discovered is that these fireboxes usually had very similar direct heating surface areas, so soft-draft, thin-fire apparatuses such as these can be compared.

After tinkering some more, I came up with a formula that takes into account the superheater’s contribution as outlined above,

the higher rates of evaporation per square foot of heating surface for fireboxes compared to tubes (about 6 to 1),

driver diameter (fewer RPM means fewer “lungfuls” per minute or per mile).

It was a lot of fun and the resulting number works pretty well for all kinds of steam l