Who Built The Highest Quality 4-8-4's?

[selector]

Duuuuccckkkk!!

[feltonhill]

No need to duck (QUAAACK!!!). This sort of thing happens. Lets clear away some of the smoke and noise and answer the questions raised.

I may or may not know anything. What I do know is this: steam locomotives are one of the most dramatic and dynamic machines ever created. That’s why I try to figure out how they did what they did.

The Allegheny had a very large boiler with the correct proportions for producing large quantities of steam. In the case of locomotives, and possibly in other venues, size matters. You can offset the effect of lower boiler pressure by developing more lbs of steam per hour. In this way you can equalize the potential energy in the steam available to the engine sets. The large quantity of steam was fed into four relatively small cylinders, which are ideal for extracting the potential energy from the steam provided by the boiler. This is why the Allegheny was capable of punching out 6,500 DBHP all day at relatively economical firing and evaporation rates. Nothing else matched that.

Ross can say anything he wants; it’s his loco, he’s in charge. That doesn’t mean he’s exactly right when it comes to horsepower estimates. If 5,000 is his estimate, so be it. I don’t know how he arrived at that figure. Maybe he was referring to indicated HP, I can’t say and neither does the 614 website. It just says approximately 5,000 hp. I do know, that based on actual tests 5,000 DBHP is relatively rarified atmosphere for a 4-8-4, and that 614 does not have the specifications, accessories, or proportions to match the locomotives that developed DBHP readings in this range during tests. If it's IHP, that's another discussion.

Please realize that it doesn’t take huge amounts of horsepower to run 79 mph with 25 commuter cars. It certainly does take a darn fine loco to do it, though, and 614 is every bit of that. I’ve seen the videos of the NJT trips and the C&O trips many times. Very spectacular. But let’s not be unduly impressed by large numbers. Back in 1938, AAR estimated that it took about 3,000 hp to get 1,000 tons to 100 mph on level track. I use similar estimating techniques to those employed back then: Johnson/Keisel equations for locomotive evaporation, resistance and performance, Davis equations for train resistance. The former two are generally accepted estimating techniques used by the railroad industry in the 1940's at the end of steam development. They also work very well as far as developing estimates of real-world locomotive performance when compared to actual tests. Davis equations in various forms have been accepted by AAR and American Railway Engineering Association (AREA) as reasonable methods for obtaining train resistance. Most of this is found in textbooks and test reports, printed on paper. So, it’s not my say-so, but the wisdom of the experts that went before. I have a great deal of respect for their knowledge. I believe we in this generation should too.

I have never run steam; only a privileged few have. However, operating steam locomotives and estimating their performance are two different skill sets. They can coincide, but don’t necessarily have to. FWIW, I’ve passed information back and forth with Ross Rowland over the past several years. He’s been very helpful in furnishing actual information on coal and water usage for 614 as well as its performance. He certainly didn’t have to do this, but the fact he volunteered his time to respond says a great deal positive about him, IMO. Rich Melvin also generously replied to my questions with real information of 765's performance and coal/water usage. I’ve obtained information from the engineers that ran N&W 611 during excursion years. Actual tests from ATSF, PRR, NYC, N&W, WM have also survived, and I’ve used them as sources of real-world data. Theory without a reality check is useless. That’s why I use actual information to measure the reasonableness of my estimates. Thermodynamics may be complicated, but it’s not a black art, and it does not play favorites.

Unfortunately 614 and its sisters appeared very late in the game. AFAIK, C&O did not run dynamometer tests on them or its final 4-6-4's, and we’re the poorer for it. As a result we’re stuck with no actual data. It may be interesting to note that 614's boiler proportions are somewhat different than the first two series, and more of less follow the tube/flue setup found in NYC’s Niagara. However, it has less direct heating surface (firebox, combustion chamber) than the Niagaras, and along with the lower pressure, accounts in part for the lower DBHP estimate. Recall that the Niagaras achieved about 5,050 DBHP, one of the higher figures recorded for a 4-8-4. That’s why I said that 5,000 DBHP is exceptional.

So, without a dyno car test, we’re free to believe whomever and whatever we want.

If my comparatively modest DBHP figure for 614 is interpreted as a slam, it’s out of my control. There’s nothing contradictory between that figure and its observed performances. It’s a very good 4-8-4.

As we get further and further from the steam era, the grand old men who really knew what performance was all about (and how to estimate it) are disappearing. A most recent and great loss was Vernon Smith some months ago. Further, number inflation is taking over. Indicated HP becomes drawbar HP, equivalent evaporation becomes evaporation. Adjectives are dropped in the cause of conciseness or entertainment, and facts become more and more obscured.

I just try to make decent estimates using the best historical information I can find.

Yielding the soapbox to the next speaker.......

[nanaimo73]

jlampke
Thanks for starting this thread. Have you read this one ?
www.trains.com/community/forum/topic.asp?TOPIC_ID=31139

[CSSHEGEWISCH]

Comparing various steam locomotives is always to have a high proportion of subjectivity in it because of the nature of the beast. The misuse of the Alleghenies is fairly well known and one could reasonably argue that they could have outperformed Big Boys if they were operated on UP in fast freight service instead of in drag freight service on C&O. It has also been posited that the NYC Niagaras were too powerful for the service in which they were used, otherwise how could a 4000 HP A-A set of E7A's successfully replace a 5050 HP 4-8-4? The steam historians among us could probably cite other similar situations.

An otherwise high quality steam locomotive of any wheel arrangement would be an underperformer if it was not used in the service for which it was designed.

[selector]

As I said some time back, if we were to agree on a standardized set of criteria, the solution would long since have been reached.

Pesonally, calorific content consumed per ton-weight at 60 mph on level track would be a good start. Now, if we could just get those figures.

[feltonhill]

How about lbs of coal and water per drawbar-horsepower-hr at peak DBHP speed? This is only a slight variation of selector's recommended standard. The reason I changed the speed is that 60 mph would favor higher drivered locos. A loco can be forced to all sorts of readings, but the coal and water required increase very rapidly. So any measurement standard that links output performance with consumables should produce interesting results. This sort of thing can be synthesized fairly readily, but I'm not sure such methods would be well received here.

[feltonhill]

This information is for GP40-2. I found some specifics on the C&O J3s. According to Gene Huddleston's book, C&O Super Power Locomotives, the 5,000 figure for a C&O 4-8-4 was been taken from Railway Mechanical Enginneer, February 1936, which states, in part "....is expected to provide a locomotive capable of developing not less than 5,000 cylinder horsepower." Cylinder HP is the same as indicated HP. This figure was related to the first series of J3 4-8-4's which were built at that time. They had a higher total heating surface than the final J3a's, with a different direct to indirect heating surface configuration. Firebox and boiler design proportions changed during the last 10 years of steam development. However, 614 as originally built would probably be able to match 5,000 IHP. Likely this is where Rowland's estimate of 5,000 (non-specified) HP that appears on the 614 website came from. Keep in mind that 614 no longer has its feedwater heater (actually an exhaust steam injector) so that costs it about 8 to 10% in steam generating efficiency. In its present state, it would be more hard-presssed to equal either 4,000 DBHP or 5,000 IHP.

[Anonymous]

Just want to say, I do not know a lot about the 4-8-4's but they are impressive to watch run no matter who built them. I am glad to see such an interest in the locomotive. Truely one of the finest built.

[GP40-2]

Feltonhill,

First, sorry if I sounded so harsh

Second, I believe when contrasting the earlier J3's to the later J3a's you are comparing apples to oranges somewhat.

There are several references from Lima and the C&O that indicated the changes to the J3a's boilers (which make them look somewhat smaller on paper) actually optimized the combustion efficiency, and greatly increased the boiler ability to absorb the BTU's from the firebox (whose efficiency was increased from the eariler versions also).

There are also references that indicated the steam circut and exhaust nozzle geometry was optimized over and above what was accomplished by Alco on the NYC locomotives.

I'll have to dig out the exact specifications, but I seem to recall the J3a actually had a larger grate area and heating surface (but a slightly lower operating boiler pressure) than the NYC engines any way.

I also think you place too much emphisis on items such as grate area, heating surface, and operating boiler PSI in determining your estimate of locomotive HP.

In a nutshell, grate area figures do not take into account the efficiency in which the coal is burned, heating surface area does not take into account how efficiently a boiler absorbs the heat to produce steam, and pressure does not take into account how efficient the steam circut delivers the steam to the cylinders, or how efficient the exaust system is. There are plenty of mechanical engineers out there that will tell you the same thing.

Your statement that modest boiler pressure worked against high DBHP is also not entirely accurate. Pressure and volume of steam produced are inversely proportional in a boiler. Yes, you increased the potential energy of the steam, but you proportionally reduced the volume of steam that boiler can produce. Might be a thermodynamic increase, by HP output is more than likely a wash at best.

I also recall Mr. Rowlands statement of "approximately" 5000 DBHP did not occur until after the 614 proformed the ACE tests pulling frieght on CSX. If memory serves me right, the 614 was hooked to a dynometer car that recorded drawbar pull at various speeds and firing rates for the test. I really don't believe he got the figure from the eariler / less powerful J3's, as Mr. Rowland would tell everybody within earshot that the J3a's were much more advanced and powerful that the older versions.

While I admire your interest in comparing steam locomotives, I think you are trying to accompli***he impossible. As you know, steam locomotives were all custom built for their respective railroads. I really think you are stretching things too far with your realitively thin historical data.

On paper, the NYC engines don't look as impressive as many other 4-8-4's, but we know it was the top of the heap HP wise (well, at least from what historical tests were actually done correctly and documented correctly--you always need to ask what parties conducted the tests, and what personal advantages those parties had to gain from, um, fudging said data) Maybe the N&W J and NYC engines did preform as stated, or, maybe the testers had some other motive--like protecting their jobs from the efficient and evil job destroying diesel-electric? At this point in time we can only guess and take the reports at face value. On the other hand, there were plenty of steamers that looked incredible on paper, yet were total stinkers in service.

Finally, what do we really know about the J3a's. 1) The IHP you stated was from the J3's which were a generation removed from steam technology of 1948 when the 614 was built. 2) There is Lima and C&O documentation that the J3a's were much more optimized than the orginal versions. 3.) Without extensive engineering data of the total systemic efficiency of the NYC, N&W, and C&O engines, a direct comparison using just PSI, grate area, heating surface is meaningless. and 4.) (not to rub this in) you have no first hand knowledge of actually pulling tonnage with the 614, while Mr. Rowland has thousands of hours behind the throttle pulling thousands of tons at speed.

Last, you statement of the HP needed to pull 25 commuter cars on level tangent track is also somewhat meaningless. I don't know a whole lot about the NJ trips, but I know the track certainly was not level. On the CSX trips, the track was anything but level, and the 614 certainly was not pulling light weight commuter equipment!! Look at the mainline route maps on the B&O the 614 traveled. They are much more severe in gradient along with curvature than anything else in the East. I don't know of any excursion locomotive, including the N&W J that had to contend with the high speeds, along with the high gradient and curvature that the 614 had to content with while operating on the old B&O routes. Obviously, you never rode on the 614 while it stormed up Sand Patch or the West End Grades at speed. Too bad, your loss.

I admire your courage in tackling this (maybe impossible to answer) question, but at this point : Advantage Mr Rowland.

P.S. Ross, if you are out there, it would be great if you would chime in on this issue.

[Tim Burton]

Are their any existing Niagras?

[GP40-2]

QUOTE: Originally posted by Tim Burton

Are their any existing Niagras?

Nope. All scraped.

Only one N&W J - The shell of the 611 in a Virginia Museum (I say shell, because I don't think it is in running order anymore)

Only one C&O J3a - The 614, which is in semi-running order and in storage in PA. I believe it needs around $500,000 in mandated FRA work at this time.

[Tim Burton]

Ugh, as bad as PRR destroying all the T-1s. I think a pissing match on certain grave stones would be due....

[feltonhill]

Take heart, Tim (well, a little anyway). 611 was returned to VMT intact, if in need of FRA recertification. It's still moved back and forth periodically. 1218 is the shell. It was under rebuild at the time of NS' discontinuance of the steam program. All the pieces are still available, but there are no grates in the firebox, no flues in the boiler, and no pistons in the front (??) engine. Sad, but at least they both still exist.

[feltonhill]

GP40-2,

Hmm, lots of items to address here, and I gotta not sound like the Bickersons. To quote a well-known person discussed previously and below, “Here we go.”

No doubt about the changes to the J3a’s boilers being in line with the final and likely best thinking of the late steam era. They were an improvement of the first two groups of J3's.

Do you recall the source for the steam circuit and exhaust nozzle optimization comments? Someone is going to have to go a long way to beat NYC’s Selkirk front end design. I’d like to read what it said.

For others that may not have access to specifications, the grate area, direct and indirect heating surface, superheating surface, and flue length and configuration of the C&O J3a and the NYC S1b are almost the same. The biggest difference is pressure: 255 psi for C&O and 275 psi for NYC.

My estimates consider more than the items you suggest. I use most of the information at the locomotive diagram level: boiler proportions and internal configuration, machinery factors (cylinders, drivers, bearing type), and weight distribution. None of this makes the method correct, but it reflects an effort to make it inclusive. I can only estimate a reasonably expected average, with flexibility to include specific performance data when it is found. If you want to see what the entire setup looks like, it was published as a set of unexplained workpapers in NYC Historical Society’s (NYCHS) magazine, The Headlight, third quarter 2002 issue. The factors you mention (e.g., burn rate, heat absorption, steam circuit design) are not generally available unless actual tests have been run. I incorporate this information when I can. But in 614's case, we’re faced with the strong probability that it was not definitively tested and we’ll never have a reference point. I wish we did. It would be better than haggling over details.

The effect of raising boiler pressure while leaving the rest of the locomotive the same was illustrated by tests on another 4-8-4, N&W #604. Constant speed tests were run in mid-1945 to determine the operating economies of raising the boiler pressure of the Class Js from 275 psi to 300 psi. The entire test report has survived intact (N&W Historical Society and NRHS-Roanoke archives). A summary of the test results were printed in the December 1946 issue of Railway Mechanical Engineer. A graph on pg 657 shows two DBHP curves and DB pull curves for 275 psi and 300 psi. Both curves for 300 psi are higher than for 275 psi at all speeds. Based on this graph, raising boiler pressure (while changing nothing else) increases output and lowers coal and water consumption.

During the ACE tests, 614 was equipped with various instruments, but no dynamometer car was used. The car trailing 614 was a tool car.

The NYC Niagara tests are pretty much the gold standard for modern locomotive road tests in the U.S. No other 4-8-4's performance was so thoroughly documented. We’re fortunate that the test report survived (it’s in the NYCHS archives), because much of Central’s engineering heritage was trashed. The entire test was run in accordance with standards of the American Society of Mechanical Engineers (ASME), which were also used by N&W, PRR and other roads. The document, which includes detailed data and calculation for both 6023 and 5500 is a bit over 180 pages long. I’ve entered the test data into a spreadsheet and developed equations for most of the calculations (the typed test report is essentially a spreadsheet converted to values). None were out of line. Therefore, allegations that von Nagle’s theorem was applied or this information was otherwise fudged are unwarranted. I would recommend Paul Kiefer’s book, Practical Evaluation of Railroad Motive Power. If you can find a copy, read it. He compares steam, diesel and electric power as it was in 1947.

As far as N&W is concerned, their Class J 4-8-4 tests, although not as extensive, were run before the diesel was considered, so they would have had no motivation to play games. There was no need in either case. Performances of the Niagara and J were good enough as is.

For purposes of comparison, it should be noted that NYC was interested in the ultimate capacity of its locomotives, and as a result, the road’s maximum performance figures are just that. They pushed boiler and machinery to the limits, and used high evaporation and firing rates to achieve the figures. This makes some of the results difficult to use as a basis of comparison with other roads that may not have tested their locomotives the same way. Knowing the specifics of tests is important.

You can rub in what you want, it doesn’t matter. I’ve found that practical operating experience and engineering expertise are complementary. The best results regarding locomotive performance evaluation are achieved when there is a cooperative effort between the two bodies of knowledge. That’s why I contact those who have hands-on experience. However, I don’t subscribe to the idea that one can supplant the other.

Ross used to participate on another forum, but got fed up with trolls, imposters, and other disagreeable types. Who knows, he may visit this one if he has the time.

Observation, and please don’t take it the wrong way - You seem to defend 614 as if it can somehow be diminished through the existence of contrasting comments by others. Can’t happen. It survived, ran for years, performed admirably under all sorts of conditions, including many for which it was not designed. I’ve spent hours on watching 614 videos (17-Mile Grade, Sand Patch, ACE tests, New River trains, NJT), sometimes just to enjoy the spectacle, other times to analyze the performance as carefully as possible. C&O 614 is a winner, regardless how much we push and shove and drive everyone to distraction with minutia.

[selector]

QUOTE: Originally posted by feltonhill

GP40-2,

You can rub in what you want, it doesn’t matter. I’ve found that practical operating experience and engineering expertise are complementary. The best results regarding locomotive performance evaluation are achieved when there is a cooperative effort between the two bodies of knowledge. That’s why I contact those who have hands-on experience. However, I don’t subscribe to the idea that one can supplant the other.
.

Very true. I was raised in a mining family where my Dad made a darned good living developing exertise in low-grade copper ore metallurgy. He used the engineering firms for the 'rough' work, and they did good work. But getting what was ultimately installed in the way of grinders, mills, floatation circuits, separators, etc, was more an art than engineering. The guys in the cabs and in the ops offices had to work as a team to get the most out of the locos in the real world.

[nanaimo73]

How much would DBHP be effected if one of these 4-8-4s were switched to oil ?

[daveklepper]

I liked them all, and rode behind Niagras on the Wolverine, Empire State Express, and New England States (west of Albany) many times, but a N&W J was one of the very finest locomotives built anywhere and gets my vote.

[feltonhill]

Switching from coal to oil - There's very little information on this subject relative to 4-8-4's. ATSF 3751 was changed from coal to oil, UP 844 was also changed over. No known RR-run comparison exists for either. Change was apparently successful because both seem to be running fine. RDG T1 #21?? was recently switched over in Canada, no known comparison. The change sometimes goes well, sometimes it doesn't. UP routinely changed some of its 4-6-6-4's to oil and back, with positive results. However, the same change on a Big Boy was not called a success by the railroad. Mixed bag, no good evidence, AFAIK.

[Gunns]

The 2900s from AT&SF were oil fired as built, also the 5001s (2-10-4) wrer oil from the begining. Oils is much easer than Coal, cleaner and no stoker to fail.

Our oil bunker,


Fitting dunnage to the oilbunker,


all my restoration photos are here http://www.railimages.com/gallery/kevinevans

Gunns

[jlampke]

Whew!! Great response, fascinating and lots to learn....
Selector: No question about it. After all these years, where and how the locomotive has been stored and cared for would certainly have to have become one of the primary considerations when choosing a restoration project.

Gunns: The site you referenced, yes I have seen it before. Very interesting, and not so involved that I had to read it 2 or 3 times to absorb the contents. Just fun. I'd like to see it include more of the 4-8-4's surviving in North America.

nanaimo73: I checked out the topic you recommended. Darned interesting. Thank you.
One aspect of this that I didn't expect but has definitely made it even more interesting is the loyalty some of you gents seem to have for a particular locomotive, or regard for a particular Engineer. Ross, for example. I've never met the man, myself. I have met Doyle McCormack and had a few conversations with him down at the roundhouse on weekends. It's too bad there aren't more people like him around. He is one of the greats..... My largely subjective admiration for 4449 (and 4460) and Doyle is largely due to the fact that I am from that area, have met some of the gang, like the looks of SP's GS locomotives, have a huge amount of respect for Doyle, and I am happy to hear various people who actually know a lot about steam locomotives say that Lima built a good product. Steam forever!!!!!! [tup] [tup]

[Anonymous]

Feltonhill,

Do you have any estimated performance data on the proposed NYC C1a Duplex? I've read that it was to have Baker valve gear rather than poppet. Were the 5500 6023 tests done in part to determine the design of the C1?

[feltonhill]

The only information I have is probably what you saw in the Nov 1959 issue of Trains mag. It looks like the C1a would have had a stock Niagara boiler with a T1 type running gear. No valve gear was specified, but the cylinder drawing looks like it had conventional piston valves. The details underlying the illustration were dated 3/28/45, according to the caption. I've seen another very similar dagram for this loco, but can't find it. I don't believe it was ever seriously considered by NYC. It existed onlyas a proposal or concept.

Its performance would probably have been little different than the Niagara except at higher speeds. Also, NYC was extremely disciplined about speed and operation, and most if not all of its mainline locomotives were equipped with a device called Locomotive Valve Pilot, which among its other features, I believe had a speed tape included. So the opportunity to get out and run unobserved was very limited compared to other roads with did not keep this kind of record. Actually, the Niagara didn't need much help for normal service, so the C1a would probably not produced any great improvement. In fact, It probably would have been less useful in regular service. Duplexes were always more sensitive to rail conditions (although the PRR T1's tribulations were considerably exaggerated in most past accounts). The extremely short stroke (26" in the C1a's case) may have compromised cylinder efficiency at lower speeds (say under 60-65 mph) because of the large amount of clearance volume compared to a locomotive with a 32" stroke. This is not my opinion, but that of N&W's Robert Pilcher, who had considerable expertise and success with N&W's locomotives over the last years of steam. On the other hand, NYC had no significant grades outside Albany Hill, so the ability of a duplex arrangement to use steam more economically while cruising at 80-85 mph for prolonged periods may have been beneficial. Locomotives like the Niagara and N&W J set a very high standard of performance and operating economy and would be very hard to top in their respective services.

Somewhere along the line (wish I could remember the source), there was a statement like this: it was harder to build a bad 4-8-4 than a good one. Apparently the proportions were just about right all the way around. You could make them bigger (e.g., ATSF, N&W, WM, UP) or smaller (e.g., NC&StL, NdeM, GTW), but they would end up pretty good regardless.

[CSSHEGEWISCH]

At the risk of being drawn and quartered by a mob of steam enthusiasts, I will venture this little tidbit: There were less 4-8-4's built for service in North America than SD45's, it was a relatively uncommon wheel arrangement. That being said, I would agree with feltonhill, any 4-8-4 was a first-rate steam locomotive, I find it surprising that more roads didn't own them.

[Anonymous]

QUOTE: Originally posted by feltonhill

The only information I have is probably what you saw in the Nov 1959 issue of Trains mag. It looks like the C1a would have had a stock Niagara boiler with a T1 type running gear. No valve gear was specified, but the cylinder drawing looks like it had conventional piston valves. The details underlying the illustration were dated 3/28/45, according to the caption. I've seen another very similar dagram for this loco, but can't find it. I don't believe it was ever seriously considered by NYC. It existed onlyas a proposal or concept.

A sketch, preliminary diagram, and a short description of the C1a was included in Al Staufer's "NYC Later Power" p.269.

The book also mentions that the 1946 tests included a comparison of 6023 with 79" drivers and with the 75" drivers saved from 6000, but doesn't say much about the results.

[nanaimo73]

feltonhill,
I would be interested in your thoughts on Belpaire boilers. Do you think the extra costs were worth the gains ? Would the ATSF 4-8-4s been even better with them ?

[feltonhill]

UP829,

Thanks for the source. I'll see if I can round up a copy of the book next month. NWHS has a copy in its archives and I show up there there three days a month, supposedly to get some work done, but all that information around.... Sorta leads to a lot of diversions.

[feltonhill]

Belpaire boilers - Well, PRR and GN thought they were worthwhile here in the US but that's about it. Elsewhere in the world they had quite a few devotees.

Various advantages have been alleged for the Belpaire firebox (boilers are about the same). The most obvious is the configuration of the staybolts and the more direct stress pattern through them. In a Belpaire, the crown sheet and top of the firebox (the visible part that gives it a flat-topped look) are roughly parallel to each other, and the stress through the staybolt heads is more or less a straight line. In a conventional boiler, the crown sheet is slightly domed but the boiler is cylindrical. The staybolts have to compensate for one radius at the top and another at the bottom, and the forces at each end are skewed, particularly at the boiler shell. So the Belpaire is theoretically less expensive to maintain. I've never seen an actual cost comparison. There are also supposed to be big gains in direct heating surface, but I can only see an increase in steam volume above the crown sheet. This is a positive point, but I may be missing something else.

ATSF 4-8-4's were oil burners and as such were configured a little differently than coal burners. ATSF and SP GS 4-8-4's like 4449 tended to have less direct heating surface (firebox and combustion chamber) and more indirect heating surface (tubes and flues) than similar sized coal burners. I've read, but can't recall the source, that this has something to do with the burner characteristics, flame front/shape, and radiant heat transfer.

Would a Belpaire improve ATSF 4-8-4's? Don't know. As a practical matter, I doubt it. Kip Farrington's book, Santa Fe's Big Three, contains a reprint of a detailed performance test of 3766 transcribed in large part from ATSF Test Report 87312. One of major deficiencies that could have been corrected was noted by E. E. Chapman, Engineer of Tests. 3766 had a considerable pressure loss between the boiler and cylinders. It reached 170 psi at high speed and capacity output. Whether this was corrected on the 3776 and 2900 classes I don't know. Their specs are identical. However, ATSF locomotive diagrams indicate 3766 had two syphons and the newer classes only had one syphon. Compared to other large 4-8-4's 3766 never reached high IHP or DBHP outputs at speed because of this pressure drop (5450 IHP at 65 mph and 4550 DBHP at 50 mph). Really though, these figures are high enough to ensure very good performance in daily use, so it probably didn't matter very much unless ATSF was going to enter the "my 4-8-4 is bigger and badder than your 4-8-4" contest. Something tells me the road didn't care about that sort of thing. They moved the trains, that's what counted.

[nanaimo73]

Thank you.[:)]

[GP40-2]

Feltonhill, I really don't want to beat this to death either (this is more than likely my last post on this subject), but as you said "here goes".

In one of your early posts, you stated matter-of-fact that the J3a did not have the "specifications" and "proportions" of the NYC engine. Then you back- pedeled somewhat and stated the C&O/NYC locomotive's specs where "almost the same". What you neglected to say was that the 614's grate area was almost the same as the NYC (within 1% of each other), but the 614 has 14% more evaporative surface area and 10% more superheater area.

Yes, I have read and studied the N&W J tests when they raised the pressure. Yes, you are right that the HP was raised on the margin. You make it sound like raising the pressure from 255 to 275 (C&O vs. NYC) or 275 to 300 (N&W J) is going to pay huge dividends in HP. Again, what you neglected to say was that the N&W tests showed only a 3%-4% increase in maximum HP from the pressure increase. Like I said, pressure and volume are inversely proportional. Choose your poison, but the HP difference is going to be marginal.

Fluid Mechanics plays a much larger role in how much HP a steam locomotive makes than raising the thermodynamics by relatively small pressure increases.

I also disagree with you that the N&W tests were not done in response to the diesel-electric's increasing market penetration. Yes, they decided to dieselize later in the game, but they were very well aware of the diesels efficiency early in the game. The J tests were to increase fuel efficiency, not to increase power. Steam locomotive builders were lucky the bean counters at GM wouldn't let EMD put turbos on the 567 earlier, the increase in power and efficiency of a turbocharged diesel vs. a non turbo is huge. So large, that the steamers would of been junked by the early 1940's, and no J, S1, and no J3a would of been built for us to even have this discussion.

Last, I really am having a togh time understanding your estimate of the 614 DBHP. First, you try to use the IHP of a generation older J3 to establish your estimate of the 614, then admit the J3a was the result of more modern thinking. To quote you "No doubt about the changes to the J3a’s boilers being in line with the final and likely best thinking of the late steam era." Not only that, but the J3a's running gear was also in line with " the final and likely best thinking of the late steam era" so I really have a hard time believing a free rolling engine like the 614 is going to loose 1000 HP from the cylinders to the coupler.

My problem with your comments is not with contrasting the 614 with other 4-8-4's. As you said the 614's proformance speaks for herself. My problem is with you making matter-of-fact statements, then admitting you don't have all the engineering specifications , or use older data from a different locomotive, or when you do have some facts, you leave out the magnitude of the difference.

I think your estimate of the 614 DBHP is absurdly low, and while I admit the 614 may not have (barely-like the N&W and NYC engines) broke the "magical" 5000 DBHP barrier, it is certainly (in top running order) closer to 5000 DBHP than your estimate.

[feltonhill]

For those who may be getting tired of this exchange, please take a look at the recommended reading list at the end, if nothing else. There are several general and specific sources that will help all who are interested to understanding what’s going on inside steam locos. As you can tell, they’re a complicated animal to figure out.

Now, in response to a diminishing number of requests, let’s give this another try.

The following figures are from the respective railroad’s locomotive diagrams:

C&O J3a NYC S1b Difference
Grate area , sq.ft. 100.0 101.0 +0.1%
Direct htg surf, sq.ft. 482 499 +3.5%
Indir htg surf, sq.ft. 4,339 4,320 -0.4%
Total hgt surf, sq.ft. 4,821 4,819 -0.1%
Superheater, sq.ft. 2,058 2,073 +0.7%

These figures are almost the same.

The table on page 658 in December 1946 RME shows a 2.7% increase in DBHP (comparing the average of two runs for 275 psi and 300 psi) and is a record of the constant speed tests. The two curves on page 657, on the other hand, indicate the effect of increased boiler pressure on maximum output characteristics of the locomotive at all operating speeds. The Class J curves were calculated by N&W using a method developed by Baldwin, at 275 psi and 300 psi, a 9% increase (Test Report Chart 1). The maximum DBHP at 40 mph was about 4,660 at 275 psi and 5,100 at 300 psi, a 9.4% increase. So the table and graphs reflect two different results of raising the boiler pressure.

For corroboration of this increase, Paul Kiefer’s book A Practical Evaluation of Railroad Motive Power, (mentioned earlier) shows the same thing at page 13. DB Pull and DBHP curves are shown for the Niagara at 275 psi and 290 psi, a 5.5% increase. The maximum DBHP at 60 mph was about 5,050 at 275 psi and 5,300 at 290 psi, a 5% increase. The curves show the same configuration as the N&W-developed curves, i.e., raising the boiler pressure while changing none of the other physical characteristics of the locomotive will raise both DB pull and DBHP.

Total resistance losses are a peculiar thing. Most formulas indicate that total resistance rises significantly with speed. However, with modern 4-8-4's, tests seem to indicate that total resistance responds more to operating conditions than speed. The Niagara test data indicates that the average HP loss at 63.8 mph was 1,524 HP (average of data line 1207 in the report summary). The range of all tests, including the maximum power run was 1,210 HP at 57.5 mph to 1,680 HP at 65.1 mph. The average fits within this range. The running gear of 614 and the Niagara look very similar, so I suspect that they would give the same results under similar load conditions.

Another way of calculating total resistance is to use IHP and DBHP curves, also found in the Niagara test report. These curves show a range of 1,200 to 1,248 HP loss at 60 mph. So although 1,200 HP loss from cylinders to rear of tender may seem like a lot, it’s supported by actual test results.

My 4,000 DBHP estimate for the J3a reflects everyday operation, dispatchable horsepower if you will, not test conditions. These would be: evaporation rate of 85 lbs of water per SF of direct heating surface, or about 80,000 to 82,000 lbs of water from the tender per hour, and 250 degrees superheat. These are all very moderate, easily achievable figures for any 4-8-4 of comparable size. According to the Feb 1936 RME, the original J3's, which had a greater total heating surface that the J3a, were estimated at an evaporation of 79,640 lbs per hour.

My estimate of IHP and DBHP at this evaporation level are as follows:

614 as built - 5,300 IHP at 80 mph, 3,900 DBHP at 40-50 mph
614 presently (without ESI) - 4,900 IHP at 80 mph, 3,600 DBHP at 40-50 mph

Could 614 do better than this at higher evaporation rates? Most likely, but what’s an appropriate figure to use? Because there is no test data for 614 as built, I’m reluctant to predict what its maximum output would be. This is because I have no information regarding boiler steaming at high demand, pressure loss in the steam circuit between boiler and steam chests at high flow rates, the ability of the valves to handle large quantities of steam at speed, and the drafting characteristics that balance back pressure in the exhaust with the amount of vacuum produced in the smokebox. If this information is out there somewhere, now’s the time to bring it forward. That way we can get some real-world information on this last of the commercially built 4-8-4's

Recommended reading

Likely still available for sale:
Central Headlight, August 1975, Road Testing of the Niagaras by Richard Dawson (was available from NYCHS)
NYC Locomotive Diagrams - Niagara specifications (NYCHS)
C&O Locomotive Diagrams - J3 and J3a specifications (C&OHS)
C&O Greenbrier Type 4-8-4 Locomotives - Nuckles and Dixon (C&OHS)
The Red Devil and other Tales for the Age of Steam - David Wardale. Part of Chapter 6 covers the 614T tests during ACE program, pgs 384-402.(still in print)

Out of print, have apparently become “collectibles” and can be expensive:
The Steam Locomotive, Ralph Johnson (full range of locomotive design)
A Practical Evaluation of Railroad Motive Power - Paul Kiefer (includes information on all NYC modern steam power)
Railroad Engineering, First Edition - W. W. Hay (generally accepted standard text on the subject)

Probably available only in libraries and railroad historical society archives:
Railway Mechanical Engineer, February 1936, first series of C&O J3's
Railway Mechanical Engineer, December 1946, N&W J test
Locomotive Data, 12 Edition, pg 25, method used by N&W to compute Class J curves
The full Niagara and J tests survive in HS archives, but are not for sale AFAIK

Support the railroad historical society of your choice. It’s the only way this information will be preserved and kept available for research. Money and volunteers are always needed.