K4 - Top Passenger Steam Locomotive?

Trains Magazine Steam & Preservation
21

Roundhouse queen was my expression, and I don’t have any first-hand knowledge, but Trains Magazine articles on the T1 and on Pennsy passenger power from the late 50’s suggested low availability and difficult maintenance owing to the inaccessible exotic valve gear. Some of the articles were in the “first person” voice of people who worked for the Pennsy, although any single employee may only give an anecdotal rather than a big-picture statistical view of what was going on.

Deferred maintenance perhaps stands to reason – the decision had been made to go with Diesels, and perhaps the effort was not made to keep the T1 highly available.

As to the high level of experimentation, there were perhaps only two experiments – the poppet valves and the divided drive. Other than that, the T1 seemed like a high-output rigid-frame four-driver-axled steam locomotive comparable to many others of that time frame.

Or make that one experiment – the poppet valves. The notion of using paired simple-expansion engines on high-speed passenger locomotives had already been done on the Challenger, only 6 axles articulated instead of 4 axles rigid-frame.

22

I never really understood the theory behind the Duplexes anyway . Why the T-1 instead of a normal 4-8-4? It would seem to me that the duplication of machinary inherent in four cylinders for a locomotive of approximately the same wieght and horsepower is unnecessary. And most railroads and builders would seem to agree. Is there an optimum size for cylinders and valves and having the cylinder capacity and using four instead of two actually increases efficiency? Aren’t four driver axles better coupled together than split into two halves? Undoubtadly there are some good technical papers on this, but a good summary might help a lot.

And did anyone else experience the slight oscillation in forward motion when riding at some particular high speed behind a T-1 that I noticed?

Anyway, the K-4’s outlived the T’s by six years. Andf were reliable locomotives to the end.

23

I think the reasoning for the divided drive was partly to reduce hammerblow on the track. Two sets of drivers with four wheels each could have lighter running gear than a locomotive with all four axles coupled together. Also cylinder efficiency could be improved by getting the same power by four smaller cylinders with proportionately larger valves. I remember reading that Baldwins chief engineer at the time the original duplex locomotives were built believed that the 8-coupled, 2 cylinder locomotives of the time were at or near practical limits in terms of steam flow as well. Using four cylinders was a way to get around that. Theoretically such a locomotive would be more powerful and efficient than a conventional two cylinder design.

Keep in mind that steam had been gone from the mainlines in the US for 26 years before I was even born so none of my knowledge is first hand. So I can’t answer your question about the oscillating motion you noticed. From what I read they were slippery engines, but only if yo

24

Hello feltonhill,

please allow me to add that interesting link about the T1, a very extensive article, for the other readers, if it is OK for you.

I do not know much about the K4, but must say with its Belpaire firebox and bullish front arrangement it was a very distinguished, but maybe not in a technical way outstanding machine.

Were the myths about it just hyped, or was it just more or less nothing more than a ubiquitious Pacific?

I read in "Last of the Steam " from Joe Collias that a K4 with 20.000pounds less tractive effort could start a train easier than a T1. I do not believe that, except the engineneer would have been a mule than certified/qualified one. Probably some people at the PRR did not liked them, 'cause the T1 could almost twice the job of a K4 and would give less people work at the railroad.

Especially in the last years, Pacifics played only a role in commuter service. For example, the last Pacifics aquired for the UP was 1920, and by '53 only round about 30 of those rabbit looking engines were left.

That the K4 survied the T1 was obvious: cheap and simple and not complex and expensive.

Happy New Year to everybody!

Lars

25

Lars - I’m not sure what link you have in mind, but go ahead!

I believe the reference to a K4 being able to start a train easier than a T1 may be based on the observation that passenger trains were rarely heavy enough to tax a locomotive’s starting tractive effort, unless the station stops were on grades. A K4 was probably much easier to handle and required less skill to get under way than a T1, sort of turn on the sanders, get up to 5-10 mph then open 'er up and get out of town!

Strangely enough, when the T1’s were tested on C&O and N&W under very close scrutiny, there was no mention of any adhesion problems whatsoever. On C&O, there are many surviving memos and reports that specifically state the locomotives did not slip when starting or attempting to start under difficult conditions. On N&W there are several runs on Christiansburg grade where the throttle loss was in the 3-4 psi range, certainly indicating wide open operation with relatively high drawbar pull readings under 10 mph. PRR engineers interviewed for several articles in The Keystone (PRRT&HS magazine) also indicated if you learned how to operate them, they were very capable performers.

For those who may be interested, the off-line tests have been written up in magazines published by C&OHS and N&WHS. PRRT&HS also has an article in process.

26

Thank you feltonhill,

I had those article of the C&OHS in my mind, puplished here:

http://findarticles.com/p/articles/mi_qa3943/is_200505/ai_n13642634

You will find many more by browsing this site. Yery interesting to read. Some from Mr. E. Huddleston are available there, also.

Best Regards

Lars

27

That statement alone should tell you that there was something “rotten in Denmark”.

28

Just a moment!

140 feet would be the length of the engine and tender, or maybe the total wheelbase. The S-1 had 84" drivers and each pair wouldn’t be more than eight feet apart. Even allowing four feet for the cylinders the pairs wouldn’t be more than twelve feet apart, so about twenty eight feet rigid at the outside. Two streamlined passenger cars would have a (non rigid) wheelbase of about 140 feet and the S-1 wasn’t that big. It was about the same size as a Big Boy.

As far as adhesion was concerned, the loco with four driving axles and six carrying axles suggested that it wasn’t using its weight efficiently.

It was literally too big for clearances East of Crestline so it stayed at the West end of the system.

M636C

29

I’m not so sure the S wasn’t anything more than a glorified test bed. I believe it was gussied up and put on display at the 1939 NY World’s Fair. The real production locomotives were the T’s and Q’s.

30

As I said, there was something “rotten in Denmark”.
I would tend more to believe Oltmannd’s way of thinking that it was a “glorified test bed”.

31

PRR’s Altoona test plant reported a peak of 3550 horsepower for the K4. Experimental improvements to K4 5399 included poppet valves, front end throttle and an improved superheater, which yielded a peak of 4200 HP. Both numbers would of course be reduced at the rear of the tender. Neither seems especially remarkable for a passenger locomotive.

32

3550 HP at the Drawbar IIRC the dyno at Altoona was compensated to show Drawbar HP. So getting that from at the time of the testing more than likely a NON SUPERHEATED HAND FIRED Pacific Type Locomotive. Remember that the Best 4-8-4 got maybe 5000-5500 HP at the Drawbar and remember this at the time the K-4 wer built they used regular bearings no rollers on anything that right there cost them 10% or so. When they were rebuilt with Superheaters and stokers and Roller Bearins which some were they might have reached 4000 HP. I know I read someplace that alot of the improvements from the K5 prototypes were later retrofitted onto the K4’s

33

A standard K4s under test produced 3,520 IHP (at the cylinders) at about 60 mph, rated evaporation 70,000 lbs steam/hr. Locomotive drawbar HP (on the test plant) was less than that and drawbar HP at the rear of the tender was even less, about 2,720 at 40 mph. The first poppet valve K4, 5399, produced about 4,180 IHP at 80 mph, also rated at 70,000 lbs of steam/hr. Other 4-6-2’s may have done better, but the K4 has been analyzed microscopically by comparison to any other of that wheel arrangement. Does that make it the best or highest HP? No, just the highest measured under certain test conditions.

The K4’s were always superheated if they had the designation K4s, the “s” signifiying superheater-equipped. No. 1737 may have been built saturated, but I doubt it. All other were built that way, not rebuilt. They were built handfired, not doubt about that, and stokers helped immensely!

Unless someone is privy to more test data than I have here, no 4-8-4 ever tested at 5,500 dbhp. Most of the argument revolves around two top contenders, 5,050 dbhp (NYC S1b) and 5,250 dbhp (N&W J), and even these figures aren’t comparable. The N&W J and ATSF 2900’s may have been capable of 5,500 dbhp or more, but no such reading was ever recorded, AFAIK. Neither was pushed to that point by their respective owners. I’ve run this idea into the ground on other threads here.

For a 1914 design, the K4 did very well for itself over something like 40+ years of service. On the NY&LB, it took two diesels to equal the performance of one K4 in 1956-57.

A lot of the improvements incorporated into the K5 (higher boiler pressure, feedwater heater, larger boiler barrel) were not incorporated into the K4’s. The K4 was a much better balanced design because of the K5’s drafting and steaming problems. They never lived up to their potential.

34

A PRR memo dated April 3, 1951 almost answers your question about cost per passenger mile. I say almost because the memo gives cost per loco mile, regardless of the number of passengers. It says the T1 cost $1.5816 per mile. The K4 cost $1.1554 per mile. So, clearly you are exactly right - one T1 was cheaper than doubleheaded K4s. By comparison the cost of 6000 diesel horsepower was $1.1450. And most trains didn’t need 6000 horsepower.

My source for this info is Eric Hirsimaki’s “Black Gold - Black Diamonds,” volume 2.

35

I’m afraid this raises some questions for me. In a thread, a heated exchange, about two years or more ago, it was stated that diesels were found to have a replacement life cycle at just under 14 years. No one knew this when they came on board. That wasn’t the case with steam, not by a long shot. Also, I wonder if financing costs were included in the operating costs for the diesels since GMAC made their initial acquisition more palatable to CEO’s and CFO’s at the time.

36

The cited memo accounts for costs of fuel,water, lubricants, other supplies, enginehouse expenses and repairs. There is no mention of initial acquisition cost.

Another memo, dated October 21, 1947, compares T1, Q2, GG1 and 6000 HP diesel locos. There is a line titled Relative First Cost per Locomotive. T1 is the baseline at 100, the GG1 is 114 and the diesel is 220. On the freight side of the table, Q2 is 100, GG1 is 140 and diesel is 267. Apparently the Q2 was cheaper than the T1.

At the bottom of the table is a line :“Fixed Charges.” This may be where acquisition cost comes in, but I’m not sure of that. Anyhow, that line shows fixed charges of .2984 per mile for the T1 and .3120 for the diesel. Total costs per mile: T1 = $1.6657; diesel $1.2960. It may have been much better had the T1 been a better runner. Its annual mileage was only 79200 vs. 228760 for the diesel. Incidentally, the GG1 managed a respectable 141240 miles per year costing $ .9736 per mile. This, too, is from the Hirsimaki book.

I remember that thread. There is a certain former poster whose name has not been seen since then. It was quite entertaining.

37

Thanks for your informative reply. Interesting figures, to say the least.

You can doubtlessly see my quandary; if a single steamer could perform at the same level as the two, three, or four diesels it would have taken in that era to match the steamer’s performance, instead of an even match with the diesel having to be given a major, third line, overhaul at 14 years, or replaced outright, there really would be, on a mile-for-mile basis, the need to replace two, three, or four such platforms. Would their costs have been fully amortized to that point, or would GMAC have rescued them yet again with newer engines? What was the cost of a life-cycle replacement of up to eight diesels, or fewer of their more potent replacements, while the lone steamer was merely in need of yet another washout and piston rod packing? I ask this not knowing the answers, and don’t mean to give the impression that I am leading you and onlookers to a Grand Slam…but the questions pertain, I think.? A whole bunch of diesels, with the costs of two initial purchases, were needed to replace a single steamer, with no changes of crewing numbers between the them, which still had more than half of its useful life ahead of it.

As the King of Siam was fond of saying, “Is puzzlement!” [%-)]

-Crandell

38

Again it’s Hirsimaki to the rescue. And the answer is that in practice a single steamer did not equal several diesels. Indeed, it was just the opposite. In 1947 PRR put three ABBA sets of F3s to work on trains between Chicago and Enola, e/b Train CG-8 via Columbus and w/b LCL-1 via Crestline. The diesels ran straight through without change, needing 2 steam helpers w/b from Altoona to Gallitzin. To completely dieselize this run they needed one more diesel set, so it was often still a steam operation. With steam they used five locos, relay style, from division to division, plus six helpers, in ones and twos at various points along the way. Around Horseshoe, the steamers needed three helpers. In short, a single set of diesels did the work of nine steam locomotives.

“The three F3 sets made 33 eastbound and 32 westbound trips by September 23. Each took about 48 hours. This is a reduction of 586 steam engine handlings during this period and provided a good example of the hidden economies of dieselization,” to quote Eric Hirsimaki.

By June 1948, 18 diesel locomotives (generally 3 or 4-unit sets) were doing the work of 90 M1s, J1s or Q2s because they could run straight through without change.

A memo from PRR President James Symes of May 9, 1949 titled “Five Year Diesel-Electric Acquisition Plan, 1950 -1954,” speaks to your question about costs of acquisition. The bottom line shows a planned investment in 749 units totalling just over 1 million HP. The projected cost was $124.3 million. Operating savings, before depreciation, were $33.2 million per year. That is a return on investment of 26.7%. Again, that is before depreciation. Not bad.

I don’t know whether this answers your questions, but it is a good indication of the logic of the time as seen by at least one railroad.

Finally let me recommend the “Black Gold - Black Diamonds” volumes

39

Thanks for that. I guess I should put BGBD on my wish list for next Christmas. It’s time I got a decent RR book anyway. [:)]

-Crandell