Belpaire fireboxes-any safer ? (and PRR T1) (any PRR steam)

[nanaimo73]

Were Belpaire fireboxes just as dangerous with low water as normal fireboxes ?

Another question-Why didn't N&W have any Belpaires ? Or did they ?
I have always read that they were more efficient,and the PRR had financial control of N&W.

[jchnhtfd]

so far as I know, yes... differences in the degree of danger (from pretty d__n bad to catastrophic) seem to have more to do with the steel used (some of the late alloys cracked more under rapid temperature change; some of the earlier ones yielded at lower temps) and such details as staying arrangements...

[Hugh Jampton]

They had a bottom narrow enough to fit between the frames of a locomotive. They were wider higher up the boiler barrel, this gave them a larger steam space and greater water surface area in the hottest part of the boiler resulting ingreater efficiency.
They were no different when it came to low water, probably worse because of the greater heat exchanged in the firebox.

Read he blurb at http://www.prrths.com/PRR_Belpaire.html for a bit more info.

[eastside]

QUOTE: Originally posted by nanaimo73

Another question-Why didn't N&W have any Belpaires? Or did they ?
I have always read that they were more efficient,and the PRR had financial control of N&W.

Financial control doesn't mean operating control. Even then, the PRR's share holdings only reached about one-third of the N&W, not enough for majority control. From what I've read the PRR looked at it as an investment (a good call, don't you think) and kept their mitts out of the N&W's affairs.

[espeefoamer]

N&W did follow PRR practice of painting their passenger equipment tuscan red.

[Overmod]

One big point about Belpaire fireboxes was that the staybolting could be done with many more common-length bolts, and the incident stress angle between the bolts and sheets could be kept very close to 90 degrees. That would mean, in theory, that there might be a little bit more support for a weakened sheet before it started to let go, and a developing breach in the crownsheet might be a bit less likely to rip open to the same extent you'd see in a comparably-sized wagon-top (although this would be very little solace to the crew in the adjacent cab, who'd be cooked and probably lifted just as 'lethally' in either case!)

To get a truly 'safe' box, you need to go to something like a welded Jacobs-Shupert, which doesn't have any nominal geometric planes along which destructive force will magnify catastrophically (as is the case in any boiler structure I know that involves staybolted sheets as pressure members, including 'classic' thermic syphons). Note that I say 'welded' because riveted connections were the death of the original Jacobs-Shupert design... and company...

With respect to N&W, I want to hear Old Timer's take on this boiler question; if anyone has quick access to Louis Newton, they might ask if he has an opinion or knows something. Who out there has experience with, or knowledge of, Belpaires operating at pressures above 300psi, and whether or not they encountered more than their share of maintenance problems in that range...

[selector]

And, on a related note, what is the old-timers' (all of them) take on water troughs and their efficacy? SDR mentions sludge due to water, and that could be from mineralization, strearates due to interactions with organics, and sediment from temporarily suspended particles. Wouldn't the troughs have been reservoirs for generally crappy water?

[Anonymous]

N&W DID have Belpaire fireboxes on some pre-1900 locomotives.

ALL fireboxes would fail if you let the crown sheets be uncovered long enough. Even the Jacobs-Shupert firebox was tested to destruction after running dry.

Belpaire fireboxes were considered more expensive to build - the exterior shoulders at the front corners were more difficult to form - but more staybolts of equal length could be used. PRR backers claim more efficiency from Belpaires, but other railroads including N&W were not willing to pay the extra money for what must have been only a marginal advantage.

PRR's stock ownership of N&W did constitute control, but PRR recognized that it would be better off not interfering with N&W's affairs. It was right. During the period of stock ownership, N&W paid PRR almost $407 MILLION (data from N&W annual reports, figured on dividends N&W paid out during the years of PRR ownership of 39% of N&W's common stock - this includes every year during the depression, some of which PRR made no money on its own operations). PRR let N&W do its own thing, being content to ca***he dividend checks. This might have been the smartest move PRR ever made. N&W fans have a basis for saying that PRR's status as the "Standard Railroad of the World" was paid for, at least partly, with N&W dividends.

N&W did paint passenger cars red like Pennsy, but it didn't come out of the same can.

N&W did use position-light signals, like Pennsy, but signal rules were different.

- Old Timer

[daveklepper]

Didn't GN or NP at one time have some Belpaire firebox locomotives?

[feltonhill]

GN had a lot of Belpaire fireboxes, including 4-8-4's and some large 2-8-8-2's.

[Paul Milenkovic]

What ever happened to the fusible plug? I suppose any mechanical system can be made to fail and fail catastrophically is subject to certain patterns of abuse, and I would imagine that there is no such thing as an explosion-proof boiler. But what is supposed to happen is that if you lose water over the crown sheet, this plug was supposed to melt, letting steam flood into the firebox and put out the fire.

I am not saying that the fusible plug would prevent all boiler explosions, but isn't that a fail-safe mechanism that is supposed to at least disable a steam locomotive and prevent it from exploding at least most of the time?

[Overmod]

Ahhh...fusible plugs, the 'magic bullet' of the 19th Century...

Take a look at the bore of a typical fusible, even a 'drop' plug that opens a relatively clean full bore when it lets go. How much mass flow do you expect the thing to pass? Certainly not enough to relieve pressure quickly enough to keep a larger part of a softened crown from going anyway... or, for that matter, to make much of a dent in the fire in the time that action would be particularly meaningful... the water dissociates at these temperatures, then re-flashes to steam a bit further along; if anything transferring the heat a bit more effectively than dry combustion gas. Remember that the fusibles are way up above the arch, with draft tending to move the evolved steam forward rather than down, so there has to be proportionally more mass flow (even at high nominal boiler pressure) to get enough steam down through the firebed to start 'cooling' the fire down... and even then, you're engaging in a big defective GPCS with that steam long before it actually chills down the combustion to stop it -- carbon monoxide out the wazoo, anyone? (You oil-firing boys can speculate on whether or not puffs of steam flow back down from the crown area might cause explosive pockets if the oil doesn't get shut off pronto!)

About the only thing that fusibles will do 'effectively' on very large locomotive boilers is cause nucleate boiling and surging in the physical water level -- which will probably cause repeated quench and heating of portions of a previously-dry crownsheet (the thing would have to be largely dry for the plug to drop in the first place). That's about the LAST thing you want to happen to you in all too many situations.

I've read a number of discussions about the number of fusibles that would be needed for meaningful release of steam -- it's a fair number, and when you put them in, you necessarily weaken the sheet area (both because you've bored out to install the mounting thimble or whatever, and because you're giving up staybolting in the same area). Oh yes, when they let go, they blow high-pressure live steam into the firebox... from whence it is just as likely to cook the crew or lineside people as it is to start extinguishing the fire. Hard to proceed to drop the grates when there's no skin or muscle left on your arms to do it...

Now, on little bitty locomotives, it does make a bit more sense to have fusibles. But not on big, high-powered ones... which is why you don't see them implemented any greater than the letter of the law requires. To me, they make about as much sense as the (equally government-mandated) block valves at the exit of boiling-water nuclear reactors -- the thing that makes a PEA possible when it would otherwise not be -- looks like common sense but really is more the opposite.

I have wondered, though, whether there would be a use for a fusible valve actuator, which would for instance operate a nest of pop valves atop the turret if the plug drops, or dump grate sections. That might let you get enough valve opening to give meaningful mass flow to reduce pressure... without also pressure-cooking crewfolk and fans.

[dknelson]

The N&W ran some PRR Pacifics -- K2s or K3s, and I am pretty sure this was in the early 20th century. They also famously swapped (temporarily) a J class Northern for a T Class 4-4-4-4 and from I have read the N&W guys couldn't keep it from slipping any better than their Pennsy counterparts did.
Dave Nelson

[feltonhill]

Slight mid-thread correction. There was no swap involved between an N&W J and a PRR T1. N&W J 610 was tested on PRR from 12/5/44 through 1/2/45. PRR T1 5511 was tested on N&W 6/9/48 through 6/26/48, over three years later. BTW, there's no mention in N&W's test report that the T1 had any adhesion problems. There were two dynamometer car readings of 62,000 lbs and 63,000 drawbar pull at 10 mph and under. Not too shabby for a loco with a tractive effort of 64,650 lbs. Drawbar pull at the rear of the tender is usually about 2,500-3,000 lbs less than TE at very low speeds.

Most of what has been written about the T1 in the more distant past was not well researched. It wasn't nearly as flaky as some accounts would have us believe.

[Anonymous]

Feltonhill -

The PRR T1 on N&W did well enough east of Roanoke, after it got over Blue Ridge Mountain. But it did slip still westbound up Alleghany Mountain's 1.3% grade with a J-sized load.

The speed of the J on the Pennsy surprised several engineers, who sneered at the comparatively low drivers. The unofficial top speed of J 610 on the Fort Wayne Division was 115 MPH, at which point the drivers were turning around 525 RPM.

I had several discussions about the T1 with the late Vernon Smith, a former neighbor, who worked for Franklin Poppet Valve Co. (Franklin Railway Supply); Vernon made the very true statement that for their brief lives they generally served in the passenger pools and took what came their way very nicely.

I've read the Pennsy Society's recent writeups (the Keystone?) on the T1 performance which generally affirm what Smith said.

But I believe that the T1 had trouble in the hands of engineers who weren't as skilled, or who, perhaps, resented that they and their locomotive were replacing two K4 Pacifics and two crews. This has been the case in several areas where doubleheaded small power was replaced by big power; the crews resented it and didn't want to operate the big power to its fullest capacity.

But it's obvious from the writeups that when skillfully handled, the T1 would do what it was supposed to do.

It's also entirely possible that the Duplex configuration caused weight distribution problems and the reduction of back pressure (the reason for poppet valves) applied in the same package was a bit much. If these concepts had come along ten years earlier and were more fully developed, it's possible that the T1 could have been a much more fantastic machine than it really was.

Old Timer

[feltonhill]

Old Timer said it well. The T1 was an example of incomplete development. However, under difficult rail conditions at lower speeds, the duplex's two sets of two-axle engines under the control of one throttle was asking for trouble. This can be illustrated arithmetically, but since this is supposed to be about Belpaire fireboxes, better let that ride. I don't believe any duplex (let alone a high drivered, short stroke setup like the T1) could slug it out in the same way as a 4-8-4 at low speeds unless conditions were almost perfect. And out on the road, they never are.

[Overmod]

I think the principal "problem" with the T1 was that the valves were too good and the spring rigging too 'poor'. The divided-drive made it much more likely to get a slip on one 'engine' or the other than would have been the case with all eight drivers connected or conjugated. The Franklin Type A gear was specially designed not to throttle the steam passages at high rotational speed, so a slip would rapidly, rapidly develop into a spin up to very high rpm, aided by the generally good balance and crossbalance, the lightweight roller rods, and the 'perfect' timing up to high speeds that the Type A gear could allow. Mind you, this was as much or more a problem at 80mph speeds as at starting...

Now, the T1 didn't have separate throttles for the two engines, so you had to correct for a slip by tinkering with the reverse, provided you could identify which of the engines was the culprit. I've speculated it wouldn't be difficult to incorporate either a differential throttle cutout (a la Q2, perhaps, in the main steamlines) or some form of wheelslip light located over a 'trim' reverse control for each of the two separate 'engines'. This might have happened if PRR had resolved to keep the engines -- and certainly seems more 'cost-effective' than what the PRR actually did with one T1 in converting it to conventional Walschaerts and piston valves -- T1a, #5547, contrast-enhanced picture at

http://prrsteam.pennsyrr.com/images/prr5547.jpg

I wasn't there at the time and I'm not going to speculate one way or the other about what PRR's priorities in the Forties were. I do wish I'd had the chance to ask Mr. Smith why PRR didn't do the rebuild with Franklin type B instead of conventional (not that they wouldn't have had a world of hurt with wear on those variable cams!)

Perhaps the 'best' solution would have been a hypoid conjugating shaft (like the approach proposed for the Qs by Riley Deem?) which would have cleared the intermediate cylinder block. This would have at least kept the individual engines from getting any significant rotational inertia built up in a slip.

How you get the suspension problems (both vertical and lateral) fixed is more difficult. I do not think the T1 used the relatively-sophisticated N&W approach to eliminating overbalance (stiff lateral compliance on lead and trailing trucks), perhaps in part because the duplexes had that somewhat excessive rigid wheelbase already (and weren't exactly poster children for lateral-motion devices on driver axles!)

As mentioned in several places, notably Trains Magazine a couple of months ago, using smaller drivers in connection with lightweight rods, good valves, and good balance would allow steam locomotives to run quite as fast as normal train speeds would justify. The T1 was a very good 120mph locomotive on a railroad with a whole center section that wasn't ever going to be optimized for that kind of speed, and a passenger fleet that I don't think would ever be consistently designed to run safely (or, for that matter, very comfortably) up in that range. I'd have liked to see what would have developed from using something like the 72" driver diameter of the Ms (and a slightly fatter boiler with better front-end aerodynamics and a properly-sized grate and firebox!) to produce a divided-drive engine optimized to run in that area above 85mph where two-cylinder 4-8-4s are starting to shake and pound excessively, but where lots and lots of PRR trains were actually run. (I'd also be prepared to bet serious money that a locomotive with this wheel diameter and Franklin type B would run as fast as PRR could legally go, as well as work long mail trains and even freight 'at least as well' as the M1s could...

Reading between the lines, there was still plenty of interest in 100+mph passenger trains all the way through the mid-Fifties, so there'd be a potential niche for the design. Although I suspect that E8s would still wind up doing the job better... ;-}

But it wouldn't have replaced the Js, which were perfectly suited for... well, end-to-end trips over something not much different from a theoretical end-to-end trip on PRR from Harrisburg to Chicago, come to think of it. N&W didn't have the luxury of two different classes of passenger power, one for the mountains and one for the relative 'flats' where the A's ran best.

Of course, if you want an even 'better' solution, consider an A with Franklin type B, Fabreeka springs in the leading truck, and some other mods to give better control of lateral compliance and of relative motion of the leading engine. Of course, THAT would lead to the question "where would you find a 100mph train long enough for the locomotive to pull..."

[nanaimo73]

Feltonhill-The thread title has been changed.

The January 1979 Trains has an article about the two orphan E7s showing up at Harrisburg on the same day 5501 and 5502 arrived. After 6 months of going to Detroit and back they needed their wheels turned.The General Foreman could not believe they had 69,000 miles on them. The most on any T1 was 2800.

[feltonhill]

Overmod -

T1 5500 was rebuilt with Type B gear. According to some of the enginemen's accounts in The Keystone, PRRT&HS magazine, it was the best of the lot with very smooth starting characteristics.

PRR also modified the spring rigging on the production run to reduce the amount of lift induced by the lead truck on the front engine in curves, as well as equalizing both engine sets together.

Keep in mind there were two EMD E7's on the property in late 1945 and we all know where that would go. PRR had already made the decision to dieselize its passenger fleet by the time the T1 production order was completed in 1946. There were a variety of reasons, none of which had anything to do with the T1. At that point, it was a matter of getting them to work as best they could (and there were a lot of modifications made according to what I have here), and buying all the diesels they could find to reduce costs as fast as possible (not to mention compete with the already dieselized trains on NYC). PRR had its hands full in the immediate postwar period. The company lost money for the first time in 1946, which seemed to get its attention. PRR was also used to maintaining dirt-simple locos which could be ignored or dispatched with minimal attention to detail. Couldn't do that with a T1. You either got it right, or it would bite back. A Ferrari isn't the same as the old stove-bolt Chevy 6!

[feltonhill]

OK, with expanded topic, there's some more room for discussion

The mileage example is another one of the items that need to be looked at more critically. I read Mr. Gardner's article in Trains and I have to raise a question regarding T1 mileage during the time from Sept 1945 through April 1946, the 6-month period he refers to. The new E7's ran 69,000 miles during that time, about 11,500 miles per month. He was told that the highest mileage T1 during that period ran only 2,800 miles. According to mileage reports I have here from the CMP's office, 5504 was the highest mileage T1 at the time, posting 40,642 miles since its in-service date of 12/5/45. This is about 8,294 miles/month. For the month of April 1946, it posted 10,791 miles, only slightly less than the E7's at that time. Also during April, 5512 posted 11,442 miles and 5508 posed 10,942 miles, also about the same as the E7's. Maximum mileage figures for the T1 fleet stayed in this range until March 1947. I could never understand where the 2,800 mile figure came from. It certainly isn't supported by the available evidence.

Offsetting this, the fleet average for the 30 T1's in service in April 1946 was about 7,244 miles/month, substantially less than the two E7's. For comparative purposes, the K4 fleet was averaging about 6,200-7,000 miles per month, based on the previous 3-month's data.

More telling than that, diesel mileage increased as time went by, so that by Oct 1947 the fleet average was about 19,620 miles per month. By this time, T1 fleet's mileage had slowly declined to about 6,738 miles per month. The K4's decline was more significant; they were down to about 4,808 miles per month. So the T1's had a bit more staying power than is sometimes thought. However, the real trend was clear enough with diesels setting the pace going away. It was all over but the shouting...

[daveklepper]

I bet the K-4's reduced mileage had nothing to do with availability, just that the K-4 had lost the high speed long distance assignments, Of course, somewhat the same happened to the T-1. Interesting to know what the mileage figures on the K-4's in New York and Long Branch service from South Amboy to Sea Girt were, despite it being mostly a rush hour operation.

[Overmod]

feltonhill, you will know this:

What's the correlation between T1 miles run and available coal quality?

Seems to me the boilers were designed with the idea of burning high-quality passenger coal, but all too often received lower-quality -- weren't the troubles that eventually led up to the UMW strike at the end of the Forties cropping up already by '47, too?

That bitty little grate guaranteed that you'd get 'burning of Rome' smoke effects if you tried to realize any kind of serious power with less than ideal coal ... and the streamlining guaranteed that the soot would wind up down the boiler flanks and in the crew's eyes...

What was done at PRR to enhance combustion effectiveness, enlarge the grate, or improve the lifting aerodynamics on the T1s? (I think I know the answer already, from the post above regarding Es)

[nanaimo73]

Here is a question for you guys.In the June 1968 Trains is an article by Robert A Le Massena called "The Big Engines". He states American Steam locomotive design peaked in 1938. He seems really impressed with the ATSF 2-10-4 and 4-8-4 engines built in Philadelphia by Baldwin. The PRR must have had a good look at these two designs as they were being built.Did the PRR ever think of buying copies of these designs as their traffic exploded in 1942 or 1943 ? And which would have been more help to the PRR ?

[Overmod]

Well, as you may know, the PRR DID make use of the 2-10-4s a bit later, in the traffic surge of the '50s. For their own use, they adapted a C&O 2-10-4 design into their class J, which some authorities rank as the best of the PRR freight locomotives.

By the late '30s, the PRR was firmly in the 'modern' design camp, which among other things had concluded that lightweight divided drive was the only way to go for large high-speed express locomotives. They probably wouldn't have looked at a 4-8-4 during that period; the T was more optimized in almost every respect. I've always seen the Qs as being the logical (I use that term in a very specific sense!) application of the divided-drive idea to big and fast ten-coupled power, given a passenger pilot truck; in a sense, the same evolution over a 2-10-4 that a T class would represent over a conventional 4-8-4.

Some of that 1938 business is (imho) a bit overrated -- it's the era that brought us nickel-steel boilers with all kinds of trouble, the divided-drive boondoggle, the overbalance-formula fiasco that essentially doomed the otherwise-beautiful ACL R-class 4-8-4s, etc. I think a very convincing partial refutation of this was provided in the recent article about late-'40s developments in steam power ... depends on how you define what it was that 'peaked' in '38. Personally, I think quite a few of the most interesting developments came after that time.

[feltonhill]

Overmod -

In response to your questions on coal quality, etc, I have some much shorter answers than usual.

Correlation of miles to coal quality - no data that I've found at this point. The N&W tests in 1948 indicated that the T1 used more coal and water per unit of output than the J at low speeds, about the same at 70 mph and less above 75 mph (very approximately).

The T1's relatively small grate was a compromise to reduce weight. It was thought that the divided drive and poppet valves would save enough steam to offset the demand on the boiler and also reduce the firing rate. PRR ended up with a boiler that was about the size of a very large 4-6-4, e.g. C&NW's E4s or Milw's F7s.

PRR apparently knew that the grate size was a compromise, but no plans were made to enlarge it, AFAIK.

Will have to get back on combustion efficiency. It may be somewhere in the Altoona test plant data. I'm out of town all day today starting in about an hour so I'll look when I get back.

No further efforts were made to improve aerodynamics to lift smoke. Some of the crews found that running slightly longer cutoffs helped.

For some reason the ATSF 4-8-4's did not realize their power potential. The 3776 class suffered from a significant pressure loss through the steam circuit (superheater?) at high flow rates (see Farrington's book, Santa Fe's Big 3). I believe there were some improvements on the last 2900 class, but I have no definitive data on them (See Lloyd Stagner's article in Feb 1987 Trains).

PRR's J1 was probably a better alternative for its use than a modified ATSF 2-10-4. PRR was basically a 50-mph freight road, and the J1 was a very good fit. Plus it could work the Pittsburgh Div grades at lower speeds without getting grumpy. ATSF 2-10-4's were not very surefooted at low speeds on grades. This had somethg to do with the valve timing and limited cutoff not producing the reduction in tractive effort when starting and at low speeds. The 2-10-4's actually produced significantly more TE than rated using the boiler pressure factors recommended for the valve setting (See Lloyd Stagner's article in Aug 1975 Trains).

The 50-mph freight limit was one of the reasons that the Q2 didn't really have a purpose in PRR's freight serivce. They performed best above 50 mph, and their increased operating cost were not offset by benefits in operating economy.

Best I can do for now. Will try to find the thread later tonight or tomorrow AM.

[wccobb]

Another slight mid-thread correction: Returning to http://www.prrths.com/PRR_Belpaire.html the final paragraph reads, in part: "...Collin raised the firebox above the frames..." This during the 1800's. Thus, all "modern" PRR power (even the lowly A5 0-4-0) had the firebox grates above the frame.
To return to the original question of safety: Were there any discernable difference of safety between the two firebox designs, it is a safe bet (very bad pun intended !!!) that NOBODY would have used the un-safe design.

[feltonhill]

I left combustion effectiveness question hanging two days ago. Didn't find anything earthshaking to add.

PRR tried 14 different front ends during the Altoona tests in 1944 in order to get the drafting as close to right as possible. It picked one but was apparently not fully satisfied with compromises that were required. However, one complaint I've not heard directed at any of the T1's was the boiler's ability to produce steam at very high superheat. Keep in mind although 92 SF sounds like a small grate area for a 4-8-4, the T1's boiler was about the same size as a large 4-6-4, or a UP FEF3. Both the NYC Niagara and N&W J had greater heating surfaces and superheaters. From this, I have to conclude that the balance between the grate area and heating surfaces (both direct and indirect) were acceptable, and that the superheater was satisfactory.

However, during the N&W tests the penality of extracting large quantities of steam at low speeds (30 mph and less) by forcing the boiler were reflected in the high coal usage per unit of output, and lower boiler efficiency figures for the T1.

Hope this helps.