Jones1945I am not sure about the accurate term of it, but wasn’t burning of coal dust has been proofed unsuccessful on the German BR 05 003; the one of a kind Cab Forward Class 05, early in late-30s? I wish I have the knowledge and a chance to understand why Yellott and his research team thought it would work in their new coal burning turbine project not more than 10 years later and even patented it.
What you're looking at is coal-dust firing of steam locomotives -- what we call pulverized-coal firing. It was not proven 'unsuccessful' (there were quite good technical results, for example, in Australia in the 1950s); the problems with it were more economics ... and long-term safety ... related.
BCR and Yellott were involved with something else altogether - using pulverized fuel in a GAS turbine, in place of things like burner cans, kerosene/Jet A and whatnot. This isn't unthinkable with a solid fuel that has been de-ashed, say with solvent refining (google SRC), but the idea was to use mine-run coal (!!!) by using a combination of techniques to separate ash from combustion gas and make any residual content so fine that it would neither erode nor 'glass' the leading edges of the turbine blading.
It didn't work, and didn't work, and didn't work, but was always cheerfully promoted as 'just around the corner' year after year after year.
Didn't work when UP tried it in their Mad Max locomotive (so big they had to number it twice during development), more's the pity.
Wasn't there a Hirsimaki article on the coal turbine in one of the early issues of Classic Trains?
MiningmanPRR reports highest revenue for passenger and freight ever in 1946, this on top of a 17% increase in freight rates and in the very next sentence reports its first loss ever. Like what??? Some vague mumble explanation that is nothing but Orwellian doublespeak. Ten years later everything is junked, duplexes are staples and razor blades, passenger trains have regressed shockingly, Ike is warning the Nation of the "military-industrial complex" because he saw and knew things that we don't know, even the incredible Pacific Electric in LA is going gone. A tragic incredibly unbelievable loss. A future destroyed and replaced by a few men behind a curtain. Too bad we could not expose this.
Ten years later everything is junked, duplexes are staples and razor blades, passenger trains have regressed shockingly, Ike is warning the Nation of the "military-industrial complex" because he saw and knew things that we don't know, even the incredible Pacific Electric in LA is going gone. A tragic incredibly unbelievable loss.
A future destroyed and replaced by a few men behind a curtain. Too bad we could not expose this.
Thank you very much for the reply, Miningman. The historic annual income loss of PRR in1946 did sound very suspicious, if the last year(1945) PRR had only 17% less revenue compared to their highest revenue record (1946) in probably a few decades, let alone PRR’s passenger service income was doing extremely well during the high of wartime traffic from 1941 to 1946 (they even needed to sold lounge car and dining car seats to handle the wartime traffic), It was hard to believe PRR reports its first historic loss this early.
According to the book "Penny Streamliner", gross annual revenue of the Trail Blazer was $2,260,000 in 1940, which was enough to build seven T1 or three S1, it is not hard to imagine tons of money PRR had earned from their Blue-Ribbon Fleet as well as their freight service during the war. Since I don’t have their financial reports, I can only assume that maybe the PRR Board made some reckless investment or banks and debtors suddenly call loans after the war. PRR had many stockholders who were noted person, I believe it wasn’t easy for PRR to do anything shady.
However, it was a fact that PRR was very good at wasting money after the war. I always think that it was unnecessary and too early to retire their 52 T1s in 1949-1952. Construction cost of 6110 was $310,676, assume the cost didn't change in 1945, 52 of them would cost $16 million! Note that I haven’t calculate the construction cost of 27 Q1 and Q2 yet. Horse power of one T1 (6000hp) and one Q2(8000hp) equal to three units and four units of EMD E8 respectively. In order to replace all T1 and Q2, PRR required another 264 units of E8, let alone they need even more diesel to replace those worn out K4s and I1s, M1s etc.!!
PRR already spent almost 25 million for the construction of T1 and Q2, in 1944-45, after not more than 5 years, they unnecessarily spend another tens of millions for Dieselization, let alone they bought some problematic early diesel from BLW and ALCO, this was no different than openly digging their own grave under broad daylight. If the PRR Board willing to spend an hour to read the passenger survey about the dieselization of Southern Pacific's Daylight, they should know what should had done, given that RR could never beat the speed of plane, even a DC-4.
Jones 3D Modeling Club https://www.youtube.com/Jones3DModelingClub
Jones1945... their engineers managed to build the most successful 4-8-4 ever in America railroad history and their overall performance and operation cost was good enough to slap EMD E8 in the face
Except that it wasn't. It was barely better than E7s, based on a tremendously skewed service density and essentially mid-Forties factor and labor prices. None of those things still applied by the time the E8 became so many railroads' chief passenger power. Watch what happens when you run the numbers as in Kiefer's report, but with details from 1949 or 1950...
Overmod What you're looking at is coal-dust firing of steam locomotives -- what we call pulverized-coal firing. It was not proven 'unsuccessful' (there were quite good technical results, for example, in Australia in the 1950s); the problems with it were more economics ... and long-term safety ... related. BCR and Yellott were involved with something else altogether - using pulverized fuel in a GAS turbine, in place of things like burner cans, kerosene/Jet A and whatnot. This isn't unthinkable with a solid fuel that has been de-ashed, say with solvent refining (google SRC), but the idea was to use mine-run coal (!!!) by using a combination of techniques to separate ash from combustion gas and make any residual content so fine that it would neither erode nor 'glass' the leading edges of the turbine blading. It didn't work, and didn't work, and didn't work, but was always cheerfully promoted as 'just around the corner' year after year after year. Didn't work when UP tried it in their Mad Max locomotive (so big they had to number it twice during development), more's the pity.
Overmod Except that it wasn't. It was barely better than E7s, based on a tremendously skewed service density and essentially mid-Forties factor and labor prices. None of those things still applied by the time the E8 became so many railroads' chief passenger power. Watch what happens when you run the numbers as in Kiefer's report, but with details from 1949 or 1950...
Jones1945I don’t know if the author is saying the pulverized coal delivery system consisted of a pulverised-coal feed pipes was unsuccessful or the method itself was not successful
If you have access to Gottwaldt's books, I believe he covers some of this better than Doug Self did. One issue that is seldom discussed is that the type of PC firing that was used 'mimicks' oil firing, with all the differences in detail design and firing 'strategy' which that implies. However, the firebox design on 05 003 appears to have been little modified for the test, and the burner complexity required for high, variable, unpredictable turndown (as generally seen in even the highest-speed German passenger service) does not appear to be either present or "designed for" by the engineers involved.
It is interesting as a 'thought experiment' of sorts to work through what would have been needed to make levitated PC firing a good solution for the class 05 locomotive (I would argue it needed conventional stoker firing and grates for baseline power, with PC injected over the grates for variable higher power output as with some dual-fuel schemes, and much better design of the radiant section than Wagner seemed to ever figure out how to do. I would also argue that the first best use of such a system would have been not on the high-speed six-coupled power but on the class 06 Mammuts, which "failed" in part because their steam generation was inadequate for the capabilities of their running gear...)
I invite you to work through the 'primary secondary' references in the trade press to the BCR developments, including the growing rumble of dissatisfaction at the progress of development as you get into the Fifties. I have just e-mailed Battelle (still in Columbus, OH) to see if they have or will provide records of the work done on the project. There are some very clear contemporary discussions of the technology that describe the operative problems with the designs, particularly the time-of-flight concerns at necessary combustion density within the constraints of the 1940s loading gage and weight restrictions. Note that there is a fundamental difficulty with Hilsch tube operation in that the 'fine' ceramic ash constituents will be kept levitated in the hot-end flow almost regardless of what can be practically done to filter them or segregate them centrifugally, but they will still glass or erode typical hard-metal alloy turbine-blade leading edges.
Jones1945IF I were the PRR HQ, I would apply Franklin type B poppet valves on S1 after its retirement in mid-1946 instead of throw[ing] it on the scrapheap
One potential problem here was that the amount of work to accomplish this would have been as heroic as the T1a conversion "in reverse", with the added requirement of adjusting the port areas above the 'line of section' to flow-match what would come from and go to the relevant poppet valves, with minimum dead-space effect. By that time the effective drawbacks of the eight-valve version of the Franklin System were becoming pretty well known to PRR, and although they used B-2 on the T1 to good effect, the situation is quite different from voluntarily adopting, then fabricating a new eight-valve setup four times for a locomotive that is already too heavy, too long, probably too slippery in the hands of most crew, etc. In addition (assuming a world in which it would make sense to rebuild steam expensively for service) the opportunity cost of any money spent to adapt the S1, dollar for dollar, would involve money remediating T1s ... where in my opinion it would be better spent.
It is tempting to state 'because PRR was actively considering replacing Franklin System poppets with piston valves even on the T1, why consider poppets?' But it probably should be remembered that when PRR made the T1a modifications they fully understood this would greatly reduce the true high-speed capability of the locomotives, and to me all the subsequent test results (and some of the analysis done by Joe Burgard) bears this out. The true question is slightly different, and could be framed as 'did an 80mph railroad like PRR actually need locomotives that made distinctively high power above 115mph' ... and that idea itself, very attractive in the first postwar years, fell off very quickly in the years of ICC order 29543 and the decline of market for even the best high-speed long-distance trains that was observed. So T1s with 'reliable old piston valves' (and the back pressure issues that retarded slip propagation) were a perfectly good "4-8-4 replacement with lower augment" in the speed ranges the railroad needed to use ... and yes, in the absence of good diesels it would have been highly interesting to have seen if the 1948 'remediations' provided a better reliability for engine effectiveness than the T1a conversion offered.
https://patents.google.com/patent/US2197731A
Think of Nonemaker's system as a differential governor, rather than a mechanical conjugation system. It senses the speeds of both engines in a duplex (with Spicer-drive-like shafts to a differential output) and generates a control 'error' signal when speeds differ by more than a set amount (this takes care of differential wheel wear in a way that mechanical conjugation needs slip to accommodate), You may recall that the Q2 slip system used small wheels on driver treads for its speed sensors; the '731 patent avoids this ... but as drawn, could not be implemented on a T1 not converted to B-2 (as the governor and shafts would interfere with the rear engine type A cambox).
There is no reason why the two Spicer drives cannot be linked directly to DC proportional generators, or rotary encoders of a variety of designs, and the differential action then performed electrically (e.g. via the Q2 approach or as in contemporary gun directors) which can then include both absolute speed differential and rate-of-change in "computing" and controlling the response.
There are better ways of implementing the differential control 'trim' activation. You will note that this shares the problem of the Q2 system in having to support a rotary bushing that must retain both tribology and seal in the presence of a wide temperature range including much higher than 'expected' levels of superheat. It would be much simpler to implement this with Franklin Precision air throttle actuators, more directly, or better still with Wagner fluidic valves throttling part of the steam flow in a branched section of manifold.
Nonemaker's idea was followed up by one by Steins, which was issued in 1948:
https://patents.google.com/patent/US2440124A/en
Note that this is not the slip control applied to the Q2s, which used a completely different method of detection.
Here the idea is that back pressure increases with speed, and (certainly in theory) the two engines in a duplex could be designed so their rate of back-pressure increase with speed were close across the range of speed. (Remember here that PRR is using common cutoff and throttle for the two engines of a duplex, and this system is removing the necessity to provide separate control of either.) The system measures back pressure for both engines and generates a control-error signal that moves a valve to progressively impede steam flow in the 'faster' engine. The patent language explicitly recognizes the need to damp the response both in speed and 'recovery' and mentions where this comes from.
This system retains the idea of rotary vanes acting to reduce the free area of some part of the steam supply piping, but this time keeping the bearings for the vanes entirely inside the pipe and actuating via a rod completely sealed in the steam space from an attached cylinder. It might be interesting to see what issues would have come up in maintaining these, as there are two bearings, two pin joints, and a rod gland exposed to the effect of superheat. I don't think any 'stiction' or heat expansion effects have any influence on effective function of the system, as both the sensing and the differential actuation are independent of the position or speed of operation of the actual control actuation. It's an interesting 'practical' approach taken in comparison to the other methods.
Inside rods a la PLM (or the Withuhn conjugated duplex) are a poor idea mechanically for any large (or fast) steam locomotive, as it is difficult to implement them with roller bearings. Note that on the ACE 3000 the inside cranks conjugate the two main driver pairs, which implies that these would not have cannon boxes (or the usual kind of lateral-motion device). The only way I could figure out to get around this is to construct a stiff cage outside the sweep of the inside cranks between the two roller-bearing axleboxes; this is facilitated by using tunnel cranks (and large-diameter bearings) for the mains, but the unsprung mass of the arrangement becomes substantial although it does remove one of the critical limitations on achieving short stroke on the outside cylinders.
Jones1945 ... only Q2 installed another anti-slip system and that system wasn’t maintained by the PRR constantly and it sometimes didn't work properly (IIRC.)
It may bear repeating that the system installed on the Q2s involved cutting-edge technology at the time, some of which could almost be considered 'classified' as it was used in gun directors. Some of the problems involved the sensors for the driver rims. For example, the drivers would neatly offset-print curve grease and traction sand up to the little wheels; this was not as serious on a Valve Pilot (which displays an advisory signal continuously) as on the Q2 where it must track both speed and acceleration for feedback control with short but 'debounced' latency. The real problem appears to be the maintenance of the 'butterfly' valves themselves, about which I've already noted some of the "issues", and the nature of the control actuation itself, which apparently went between 'open' and 'closed' position quickly (to ensure no sticking of the valve, perhaps) rather than modulating continuously or in small steps -- bang-bang control is NOT a good idea on a high-horsepower duplex working near its effective adhesion limit.
S1 and T1 didn’t even have any device to ensure two set of separated drivers, four cylinders operating at the same speed.
What truly amazes me is not so much that it wasn't installed as that no particular organized attempt was made to retrofit it when the slipperiness problem reports became so pronounced.
I think it wouldn’t need a complicated device to at least let the crew know which set of driver is slipping, 2 speedometers with a red light or alarm above each of them, detect the speed of both front and rear engine separately, an electric device connect and monitor the speed of both speedometers, if the device detect a huge speed different between two speedometers, the red light or alarm above the speedometer showing a higher speed will be turned on or activated.
It could be done a bit simpler than that. The existing Jones-Motrola speedometer is driven off the rear engine, which conveniently is not the one that usually slips; the mechanical drive can easily be fitted with a centrifugal switch to detect abnormally high acceleration of that engine (which would light the respective display light for a slip on that engine).
Meanwhile, the mechanical drive can (relatively) easily be given a wiper potentiometer arrangement, for example close to the pointer, that produces a reference voltage associated with speed. The front engine drives its own speed reference (probably an instrument generator directly producing a voltage proportional to speed in that era; easier to use a rotary encoder and circuitry later) which is then easily bridge-compared to the rear-engine voltage so that the forward-engine light illuminates ... you decide whether you want it to come on when a given number of volts' difference is detected, or illuminate brighter and brighter relative to the voltage (i.e. speed) difference between the engines.
The arrangement on 5550 will likely involve a couple of rotary encoders (4096 or 8192 steps) each with a small processor that generates speed and delta-speed (rate of change, or acceleration/deceleration) to run the lights, and (to the extent installed) control the antislip systems.
Note that the easiest way to adjust slip on a T1, which has only the one throttle and little provision for effectively shifting valve gear on only one engine, is to split the independent brake foundation rigging so each 'half' works on only one engine, and provide a small proportional valve that (1) allows the shoes to be brought right up to the wheeltread, to trim the latency out of actuation, and (2) proportionally applies and graduatedly releases pressure to control both slip and tendency to break away. A 'proportionally-dimming' light control signal could just as easily be a magnetic-valve actuation signal.
It requires the engineer to control the throttle manually when wheel slip occur. But I understand that PRR wanted a system which can detect and adjust the speed of both engines on a duplex automatically, like the anti-wheel slip system they used on Q2.
It's a little of both ... wound up with the idea of spending as little as po$$ible while eliminating any tendency for failure of the system either to disable the locomotive's operation or 'fail unsafe' by not indicating a potentially damaging slip condition. Note that the phenomenon of high-speed slipping is by definition a transient situation that cannot be 'corrected' by moving the throttle (even a Franklin Precision air throttle set up to have very small control latency in both opening and closing) as speed must be maintained, and all an incremental closing of the throttle will produce is a slower road speed down to the point that transient loss of adhesion doesn't propagate effective slipping -- a speed that may be comparatively low in many cases where train resistance is near locomotive capacity.
Needless to say, any device that eliminates the issues with responding to transient slips without explicit engineer distraction is reasonable to consider (even if it costs a considerable amount to implement) provided it does not break or fail in the wrong ways when needed. That is particularly so when the inertial forces build up as quickly as they do, and overspeed consequences can be as severe as they often are, in direct-connected reciprocating locomotives with high effective axle load. Having a system that responds effectively in controlling and then removing overspeed in about 1/4 revolution, far quicker than any human reaction time even to a light in the cab, becomes much more important than PRR acknowledged.
If not, it was no different than telling the whole world that the idea of duplex had an unsolvable defect.
And therein, of course, lies a somewhat paranoiac point: since there are so many relatively simple answers to make the problem stop without expensive redesign (e.g. installing two front-end throttles on the T1s) ... why was not one of them tried? Not for the first time do I wonder if things were being set up to fail. PRR couldn't make any use of their wonder wartime power once they went back to 50mph commodity freight in an anticipated postwar recession. Even before the consequences of the Great Boiler Steel Debacle reared their heads.
Jones1945, in part:https://patents.google.com/patent/US2128490A/ ... I believe many people have seen this patent drawing long time ago.
Angus Sinclair used to comment on certain 'crank' inventions regarding steam flow by noting that there were usually arrows in their patent drawings showing which way the steam flowed. The problem being that the steam couldn't read, and therefore didn't follow the neat little arrows ... so the idea didn't work as stated.
Keep that in mind when examining this patent, and the apparent lack of enthusiasm for its teachings on PRR subsequently. The T1 ranks as one of the least aerodynamically-satisfactory designs of all time from the standpoint of keeping smoke away from the engineer's line of sight or out of the cab. And to my knowledge the difficulties were never meaningfully addressed...
Now take a look at the Quellmalz patent conveniently referenced for you in the Google Patents link Jones1945 provided. (For those that read German, the original, on Espacenet, is here. The link in Google Patents has the two-edged 'feature' of being translated, presumably via the Google Translate engine so be prepared to do some on-the-fly reinterpretation of words and grammatical structure as presented.
Overmod It may bear repeating that the system installed on the Q2s involved cutting-edge technology at the time, some of which could almost be considered 'classified' as it was used in gun directors.
It is still hard to believe for me that Pennsy only used Q2s, the most powerful non-articulated locomotive for merely 4-6 years which was only equal to about 13% of GG1’s average service life despite PRR put so much effort to develop duplex for freight hauling with cutting-edge technology. We can see PRR had high hope of the duplex project from the conceptual drawing of Q1, which was a 4-6-4-6, with a streamlined design by Raymond Loewy. I wonder if there were any other technologies in late-40s which could be applied to Q2s to decrease their operating cost, assume that PRR given up the V1 project.
Overmod What truly amazes me is not so much that it wasn't installed as that no particular organized attempt was made to retrofit it when the slipperiness problem reports became so pronounced. Exactly, the PRR Board probably knew the slipperiness problem of their duplexes since the day when S1 had her first road test in 1939. T1s was what PRR really wanted but after two years of testing (#6110 & #6111)from 1942-1944, there was enough time to determinate this idea was practical or not. If PRR thought it was impractical, they still had “plan B” like building a PRR 4-8-4 base on the design of N&W Class J 4-8-4 or building “Super K4s” like #5399 (both of them were tested by PRR and powerful enough), if not, they had enough time to train their engineers the trick and skill to handle the T1 properly, but they didn’t. https://sites.google.com/site/pacific462steamlocomotive Overmod ...It could be done a bit simpler than that. The existing Jones-Motrola speedometer is driven off the rear engine, which conveniently is not the one that usually slips… ...The arrangement on 5550 will likely involve a couple of rotary encoders (4096 or 8192 steps) each with a small processor that generates speed and delta-speed (rate of change, or acceleration/deceleration) to run the lights, and (to the extent installed) control the antislip systems… ...And therein, of course, lies a somewhat paranoiac point: since there are so many relatively simple answers to make the problem stop without expensive redesign (e.g. installing two front-end throttles on the T1s) ... why was not one of them tried? Not for the first time do I wonder if things were being set up to fail. PRR couldn't make any use of their wonder wartime power once they went back to 50mph commodity freight in an anticipated postwar recession. Even before the consequences of the Great Boiler Steel Debacle reared their heads. Very true, Overmod. so many things about PRR’s duplexes were incomprehensible, so many things should have done and could have been done. I really wish I can see the 5550 running again in person and take a ride on it...... I don’t think PRR couldn’t afford another 78 Jones-Motrola speedometer and two front-end throttles on their duplexes after their passenger ridership hit record high year after year until 1946! When Baldwin told PRR to not apply the Franklin poppet valve gear to all T1s in one single order, the PRR stubbornly rejected, when people asked the Board why not using Type B poppet valve gear instead of older version, they couldn’t provide any persuasive reason. When a PRR mechanical engineer asked why using 77’ driver on Q1, a freight engine, the Project Manager told him the engine was “well proportioned” but it was proofed wrong later. There were mistake after mistake, wrong decision after wrong decision, tons of money wrongly invested, a lot of check was unnecessarily signed to different parties. Who or which parties from different company, from PRR itself to Baldwin, Franklin, EMD etc, could benefited from this kind of messy management? I would say a lot. I remember you mentioned that you have some documents about “the dark side of PRR” which can be used as evidence, this further reinforced my belief that something went wrong in PRR, at least on the T1 duplex project. Compare to the development of Electric Locomotive like GG1 and K4, K5, M1 in the 1920s to 1930s, PRR never act like a lost cause like this before. Yesterday I found a pic on eBay which showing a page from a book about PRR passenger equipment. It was surprising to know that the first postwar order of 90 lightweight coaches built by PRR Shop and ACF in Nov 1944 had quality problem which made them needed to retire in mid-1960s, meanwhile, many rebuilt P70 managed to continue serving in PC or even Amtrak era. Degenerate! LIFE magazine 1940
Overmod What truly amazes me is not so much that it wasn't installed as that no particular organized attempt was made to retrofit it when the slipperiness problem reports became so pronounced.
Exactly, the PRR Board probably knew the slipperiness problem of their duplexes since the day when S1 had her first road test in 1939. T1s was what PRR really wanted but after two years of testing (#6110 & #6111)from 1942-1944, there was enough time to determinate this idea was practical or not. If PRR thought it was impractical, they still had “plan B” like building a PRR 4-8-4 base on the design of N&W Class J 4-8-4 or building “Super K4s” like #5399 (both of them were tested by PRR and powerful enough), if not, they had enough time to train their engineers the trick and skill to handle the T1 properly, but they didn’t.
https://sites.google.com/site/pacific462steamlocomotive
Overmod ...It could be done a bit simpler than that. The existing Jones-Motrola speedometer is driven off the rear engine, which conveniently is not the one that usually slips… ...The arrangement on 5550 will likely involve a couple of rotary encoders (4096 or 8192 steps) each with a small processor that generates speed and delta-speed (rate of change, or acceleration/deceleration) to run the lights, and (to the extent installed) control the antislip systems… ...And therein, of course, lies a somewhat paranoiac point: since there are so many relatively simple answers to make the problem stop without expensive redesign (e.g. installing two front-end throttles on the T1s) ... why was not one of them tried? Not for the first time do I wonder if things were being set up to fail. PRR couldn't make any use of their wonder wartime power once they went back to 50mph commodity freight in an anticipated postwar recession. Even before the consequences of the Great Boiler Steel Debacle reared their heads.
...It could be done a bit simpler than that. The existing Jones-Motrola speedometer is driven off the rear engine, which conveniently is not the one that usually slips…
...The arrangement on 5550 will likely involve a couple of rotary encoders (4096 or 8192 steps) each with a small processor that generates speed and delta-speed (rate of change, or acceleration/deceleration) to run the lights, and (to the extent installed) control the antislip systems…
...And therein, of course, lies a somewhat paranoiac point: since there are so many relatively simple answers to make the problem stop without expensive redesign (e.g. installing two front-end throttles on the T1s) ... why was not one of them tried? Not for the first time do I wonder if things were being set up to fail. PRR couldn't make any use of their wonder wartime power once they went back to 50mph commodity freight in an anticipated postwar recession. Even before the consequences of the Great Boiler Steel Debacle reared their heads.
Very true, Overmod. so many things about PRR’s duplexes were incomprehensible, so many things should have done and could have been done. I really wish I can see the 5550 running again in person and take a ride on it......
I don’t think PRR couldn’t afford another 78 Jones-Motrola speedometer and two front-end throttles on their duplexes after their passenger ridership hit record high year after year until 1946! When Baldwin told PRR to not apply the Franklin poppet valve gear to all T1s in one single order, the PRR stubbornly rejected, when people asked the Board why not using Type B poppet valve gear instead of older version, they couldn’t provide any persuasive reason. When a PRR mechanical engineer asked why using 77’ driver on Q1, a freight engine, the Project Manager told him the engine was “well proportioned” but it was proofed wrong later.
There were mistake after mistake, wrong decision after wrong decision, tons of money wrongly invested, a lot of check was unnecessarily signed to different parties. Who or which parties from different company, from PRR itself to Baldwin, Franklin, EMD etc, could benefited from this kind of messy management? I would say a lot. I remember you mentioned that you have some documents about “the dark side of PRR” which can be used as evidence, this further reinforced my belief that something went wrong in PRR, at least on the T1 duplex project. Compare to the development of Electric Locomotive like GG1 and K4, K5, M1 in the 1920s to 1930s, PRR never act like a lost cause like this before.
Yesterday I found a pic on eBay which showing a page from a book about PRR passenger equipment. It was surprising to know that the first postwar order of 90 lightweight coaches built by PRR Shop and ACF in Nov 1944 had quality problem which made them needed to retire in mid-1960s, meanwhile, many rebuilt P70 managed to continue serving in PC or even Amtrak era. Degenerate!
LIFE magazine 1940
There is a Keystone article (90 mph and beyond) in the Fall 2014 volume. The 140 mph speed issue was discussed in the context of the T-1 and the Franklin poppet valve system. Apparently Franklin warrantied their equipment for continuous 120 mph running. Nevetheless the valve systems were failing and Franklin/Pennsy wondered why. So Franklin sent mech engineers to ride T-1s in regular service in the cab with train crews and note speeds. They could not use speedometers since the needle was usually already buried. So they used stopwatches and quickly found crews were runnng the T-1s at 140 - on the jointed rail of the Ft wayne Div. with no ATS or cab signals. The implication was these speeds were a regular occurrence. Maybe the crews were playing with the poor Franklin guys, or were always making up time - who knows.
Overmod Keep that in mind when examining this patent, and the apparent lack of enthusiasm for its teachings on PRR subsequently. The T1 ranks as one of the least aerodynamically-satisfactory designs of all time from the standpoint of keeping smoke away from the engineer's line of sight or out of the cab. And to my knowledge the difficulties were never meaningfully addressed...
Thank you for the link, Overmod. I did note that T1s wasn't subject to any wind tunnel test just like their sister S1 #6100 aka Cleopatra VII Philopator. I shared my thought about it in this post. I think It's time for me to create an index on the first post for our readers convenient.
I didn't know if this topic was raised in the past or not. Railroads took aerodynamic engineering very seriously when designing their streamlined steam locomotive in 1930s, example like DRG Class 05, DRG 61 (002 is one of my favorites) of German and the Autorail Bugatti of France. NYC’s first streamlined Hudson was also subjected to wind tunnel test multiple times and filed some detailed Patents for it. But after RRs noted that streamlining was more for advertising than improving speed and efficiency, it seems that RR didn’t arrange the wind tunnel test for heavy steam locomotive design since 1940s.
From Wiki
For a high-speed steam locomotive like T1 who always operated at 120mph or above, I think it worth the time and money to arrange wind tunnel test for them since they were fast enough for the advantage factors of streamlining to take effect, improve speed and efficiency above 80mph, but PRR didn’t think so!
From my understanding aerodynamic engineering is not only about streamlining, things like smoke deflector or how to design the front end of a steam locomotive to ensure the crew having a good visibility also involved knowledge of aerodynamic, PRR S2 and Q2 were two good examples. As Overmod stated, “The T1 ranks as one of the least aerodynamically-satisfactory designs of all time from the standpoint of keeping smoke away from the engineer's line of sight or out of the cab…” If PRR and Raymond Loewy never suggest or request a wind tunnel test for T1, they were negligence of duty, irresponsible and unprofessional in my book.
LAWRENCE SMITHFranklin sent mech engineers to ride T-1s in regular service in the cab with train crews and note speeds. They could not use speedometers since the needle was usually already buried.
See previous discussions about T1 speedometers, and the assumption in a famously quoted anecdote about reading 120mph on one.
All the T1s used Jones-Motrola speedometers, which were only calibrated to 100mph. The S1 went to 110mph. (Hudsons went to 120mph, but that was a 'standard' Valve Pilot component). The strong conclusion was that Mr. Crosby forgot, in the heat of storytelling, that the 120mph speedometer standard in most automobiles was ... not standard on the T1s. I have yet to find any indication (no pun intended) that a 'faster' speedometer was ever fitted to any T1.
Now look at the timing difference, in seconds, between, say, 120mph and 140mph, and consider the likely ride on jointed rail at anywhere near those speeds. Tell me anyone will be extracting consistent data showing higher continuous speeds in those ranges.
So they used stopwatches and quickly found crews were runnng the T-1s at 140 - on the jointed rail of the Ft Wayne Div. with no ATS or cab signals. The implication was these speeds were a regular occurrence. Maybe the crews were playing with the poor Franklin guys, or were always making up time - who knows.
Do you have a better original source than the passage in One Man's Locomotives? Because the implication there was, as I suspect is closer to the truth, that what was causing the 'overspeed' was transient high-speed slipping (i.e. something that would show up on typical rotary-encoding speed sensing devices) and not actual road speed (which is what the stopwatch method would provide if accurately done). From the standpoint of the Franklin engineers the transient-slip equivalent of "140 peak mph" causing mechanical failures of several predictable kinds was enough of an answer, as that's what they would see on their indicator traces. (Add to that the rapid deceleration stress/shock of repeated slip recovery under heavy load at high speed -- cf. Harley's description of high-speed slipping as 'felt' back in the train')
It's a lovely fantasy to assume that if a T1 could be run up to 132mph or so, with reports of increasingly bad riding in the last couple of mph, that it's just a pull of the throttle away from 140 (or 141.2 or whatever) mph. But the various physics involved don't work that way.
"It's a lovely fantasy to assume that if a T1 could be run up to 132mph or so, with reports of increasingly bad riding in the last couple of mph, that it's just a pull of the throttle away from 140 (or 141.2 or whatever) mph. But the various physics involved don't work that way."
An English website (whose name escapes me right now) has a report from a crewman on 05 002 on the day they got to 200km/h.
They'd been running tests all week, and they'd just stopped to check some recording gear and the locomotive appeared to be running more smoothly than usual, for no discernable reason, so they just kept increasing the speed until they got to 200, but then thought that they didn't want to push their luck and came back down to more normal speeds (140 km/h) which were officially sanctioned.
As someone who had run various tests (none related to speed) that agrees with my experience. Sometimes, it all works. Other times you don't even walk back down the track to collect the expensive equipment that has fallen off for unknown reasons.
Peter
Overmod ...Because the implication there was, as I suspect is closer to the truth, that what was causing the 'overspeed' was transient high-speed slipping (i.e. something that would show up on typical rotary-encoding speed sensing devices) and not actual road speed (which is what the stopwatch method would provide if accurately done). From the standpoint of the Franklin engineers the transient-slip equivalent of "140 peak mph" causing mechanical failures of several predictable kinds was enough of an answer, as that's what they would see on their indicator traces. (Add to that the rapid deceleration stress/shock of repeated slip recovery under heavy load at high speed -- cf. Harley's description of high-speed slipping as 'felt' back in the train') It's a lovely fantasy to assume that if a T1 could be run up to 132mph or so, with reports of increasingly bad riding in the last couple of mph, that it's just a pull of the throttle away from 140 (or 141.2 or whatever) mph. But the various physics involved don't work that way.
...Because the implication there was, as I suspect is closer to the truth, that what was causing the 'overspeed' was transient high-speed slipping (i.e. something that would show up on typical rotary-encoding speed sensing devices) and not actual road speed (which is what the stopwatch method would provide if accurately done). From the standpoint of the Franklin engineers the transient-slip equivalent of "140 peak mph" causing mechanical failures of several predictable kinds was enough of an answer, as that's what they would see on their indicator traces. (Add to that the rapid deceleration stress/shock of repeated slip recovery under heavy load at high speed -- cf. Harley's description of high-speed slipping as 'felt' back in the train')
I agree that this was what happened about the "140mph" thing; high-speed slipping causing mechanical failures instead of T1 operated at 140mph road speed to make up time. As many forum members pointed out in this post before that at such high speed, several hundred tons of passenger stock behind them would have shaken like a house during a magnitude 7 earthquake, passengers would complaint about that if they noted the ride quality got worse on a speeding train or worried about their own safety, these were something that any RR wouldn’t want to see. Moreover, the average schedule speed of many if not all prime named trains in 40s seldom excessed 60mph in the States, I don’t think S1 or T1 needed to operate at their design top speed, 100mph or above on a daily basis to staying on schedule. They could go faster than that, but they didn’t need to. Safety comes first!
M636C ......An English website (whose name escapes me right now) has a report from a crewman on 05 002 on the day they got to 200km/h. They'd been running tests all week, and they'd just stopped to check some recording gear and the locomotive appeared to be running more smoothly than usual, for no discernable reason, so they just kept increasing the speed until they got to 200, but then thought that they didn't want to push their luck and came back down to more normal speeds (140 km/h) which were officially sanctioned......
......An English website (whose name escapes me right now) has a report from a crewman on 05 002 on the day they got to 200km/h.
They'd been running tests all week, and they'd just stopped to check some recording gear and the locomotive appeared to be running more smoothly than usual, for no discernable reason, so they just kept increasing the speed until they got to 200, but then thought that they didn't want to push their luck and came back down to more normal speeds (140 km/h) which were officially sanctioned......
I believe they were experienced crews, their professional insight and rich experiences told them that 200km/h was very close to the limit after a couple of runs which probably can apply to most of the reciprocating steam locomotive all over the world.
http://www.gerdboehmer-berlinereisenbahnarchiv.de/Bildergalerien/19850000-db/1985oooo-851509.html
From ebay:
Raymond Loewy with his S1 model.
Kuhler was referring to American streamlining attempts, and specifically had Kantola's and some of the other 'bathtub' shrouds in mind (I had this discussion with him back in the early '70s, and he did mention the 'wind tunnel' issues with 3768 producing too much enclosure then). He was referring to those attempts that shrouded the works in many thousands of pounds of unstyled tin, or implemented streamstyling 'tropes' that made maintenance or operation worse (including smoke-"lifting" devices that didn't do the right job) -- you might remember the quote about him looking under the running boards of one of those amusing Lackawanna Disney-style jobs "to see the mechanism that made the wings flap".
We can argue about his aesthetic taste compared to others in the period; I don't care for some of the designs (including the B&O engine pictured). On the other hand I don't like the original parabolic-nosed Loewy T1s nearly as well as the late postwar engines, so we may have to agree to disagree on that point.
It has been my experience that most United States railfans don't care much for the Reichsbahn full streamlining; it's worse aesthetically than the usual inverted-bathtub designs and that's saying a mouthful. Arguments about how effective it actually was at economically-achievable train speeds (this justifying use of a pure aerodynamic design over a 'streamstyled' one) are somewhat dubious even in the case of the 05s, arguably the fastest European locomotives built. (BTW, want to see an illustration of a working definition of 'cognitive dissonance'?look at the unshrouding job done on members of that class after the war, with the 'bottom' completely open while the 'top' remains contoured...)
I believe many would agree that it is almost impossible to reach any consensus when different people trying to share their own aesthetic perspective, just as some if not many railfan dislike any streamlined or shrouded steam engine no matter where they came from or what the purpose of the streamlining was, some railfan love the same engine class but like or dislike their different appearances in different time period.
If Kuhler was referring Kantola's works on the first NYC streamlined Hudson, I agree with him partially that it was not the best-looking engine at the time, but I wonder if he ever imagined what it would be like if all decorative paint jobs and elements of his streamlined MILW Class A and F7s were removed? It would just looks like another "Commodore Vanderbilt". A renowned scale model manufacturer tried to make NYC's "Commodore Vanderbilt" a product since 2 years ago but still haven’t got enough pre-order for it, their Loewy’s K4s model shared a similar fate, Kuhler’s engines is not even on their product list, meanwhile scale models of EMD’s E8, UP’s Big Boy, SP’s GS-5 are still selling like hotcakes for decades.
If the main goal of the shrouding, especially on those prewar engines, was trying to save energy or improve efficiency and performance base on aerodynamics principles, I would choose to not judging it simply base on an aesthetic perspective, even though these goals were considered impractical by many RRs after the war; those shrouding were not a piece of art in the first place. Unlike the streamlining on DL&W 4-6-2. If DL&W really thought those Disney-style wings alike decorative plates could make the train run faster, they were world-class fool, but I think even Kuhler himself knew DL&W didn’t, so what was the point to name names or tried to mock other’s works in his own article with such an euphemistically way?
Class 05 was never mass-produced and the “semi de-skirted” thing DRG did to them were really funny, but it wasn’t a permanent measure. I think many readers already knew the popularity of these German streamlined locomotive in the scale model market (in EU) and the reason behind it. Speaking of 'cognitive dissonance', it was hard to beat the PRR in terms of scale; from their treatment on their duplexes, execution of their dieselization plan to their insight of the industry, were so embarrassing that even many PRR fans or “Pro-PRR” authors gave them harsh but fair criticism.
Edited for horrific typo.
More:
Sep. 13, 1944
PRR Board authorizes the expenditure of $375,000 for a 4,000 HP EMD E7 A-A set; track changes at “Q” Tower at Sunnyside Yard; approves a contract between the railroads and the General Electric Company for a coal-burning steam turbine electric locomotive of 5,435 HP. (MB)
Dec. 5, 1944
PRR begins one month of tests with borrowed N&W Class J 4-8-4 No. 610 in freight and passenger service on Fort Wayne Division; tests made at request of VP-Western Region James M. Symes, who is not impressed by performance of T1's and Q2's; makes 2 freight and 12 passenger runs at speeds up to 94 MPH; less power than a T1 at speeds over 42.5 MPH but better acceleration. (Hirsimaki)
Dec. 20,1944
PRR Board authorizes the construction of 50 Class T1 locomotives for $14,125,000.
Bonus:
July 29, 1948
Raymond Loewy’s office makes final settlement for its work on Penn Center and the models and plans are turned over to the PRR’s Chief Engineer; the PRR is unable to obtain financial backing or commitments. (CDY)
Interesting that Chris Baer has PRR Board of Directors approving production of the GE steam-turbine electric, but not the internal V1 direct-drive project. All the 'conventional' documentation I remember seeing (admittedly often with that red-flag caveat, use of the same uncommon term 'greenlighting for production' for what took place) indicates the V1 received approval in 1944, and I have to wonder on other grounds if the WPB would approve a STE locomotive design with, for example, the same copper requirements as a comparable diesel-electric over a mechanical system with lower war-priority diversion cost...
It would be interesting to see the details of the GE design, as it would have relatively little in common with the Union Pacific/GN high-pressure condensing locomotives of the late Thirties. When I was researching boxes at the Hagley I saw no engineering references at all to a coal-burning steam-turbine design developed enough for construction competitive to the V1, which of course is no guarantee the documentation involved had not been lost in acquisition or simply filed in different locations. On the other hand, substantial lengthening of the trackpans would be exactly something I'd expect to be required for the vastly increased water rate of something like an 8000hp V1 in appropriate (wartime appropriate to its horsepower and speed capability) service.
Just as with the Q2, much of the rationale for a high-speed very-high-horsepower single locomotive in freight service 'went away' when war priority traffic did. Thereafter, a 70"-converted J class did a steam-locomotive's work just about as well, and more cheaply, for the short while steam still made economic sense.
I was not keeping careful notes on this, but the V1's official cancellation did not come until later in the '40s, and largely over the water-rate operating economics (far less favorable vs. economics of F units even net of acquisition costs and so forth). What I have not seen is explanation why actual construction of parts of a V1 was not undertaken between 1944 and the late '40s, even though PRR was promoting the idea of by-then 9000hp turbine locomotives (not practical with a Q2 boiler, please note) that late. I consider it unlikely that this was due to a GE design for which little if any detail design appears to be documented in the same period (whereas we have a detailed Bowes-drive discussion for the V1 chassis in 1947) but that would certainly be near the top of a list of potential reasons why even a smaller-scale test of the V1 approach was not made.
I assume you have acquired and read the PRR report of the N&W J testing. It is interesting reading, and I believe at least a synopsis has been provided here in the forums. Contrary to some of the typical railfan wisdom, the report seems quite fair and does not disparage the J design, although some of the finer points of the balancing philosophy seem not to have been comprehended well. One point is that PRR still expected reasonably larger drivers for high-speed service superior to M1s (obsolescent even at 72" unworn) and adapting a J even for 77" drivers would put it as far out of many PRR clearances as the duplexes were. (See our previous discussions on why a double-Belpaire required 76" or lower drivers as designed.)
What might have been fun would be to adapt Glaze's balancing, and Timken rods, to one of the PRR J's (the 2-10-4s) and see the effects of the faster running. I suspect there were very good reasons this was not seriously undertaken, starting with Chapelon's implicit conclusion regarding roller-rod practicality that very serious cycling lateral flexure in the Timken rods was a necessary everyday operating occurrence. Much more discretion and care against short-term overstress would have been required for lightweight rods on a ten-coupled engine of that size and power than typical PRR crews were likely to maintain consistently without exception. If that gets solved, we can move to issues of flange force and two-wheel lateral guidance under conditions of limited overbalance...
... but considering what the English found they could do with speed of ten-coupled power with a two-wheel lead truck a decade later, it might have been interesting.
Overmod…All the 'conventional' documentation I remember seeing (admittedly often with that red-flag caveat, use of the same uncommon term 'greenlighting for production' for what took place) indicates the V1 received approval in 1944, and I have to wonder on other grounds if the WPB would approve a STE locomotive design with, for example, the same copper requirements as a comparable diesel-electric over a mechanical system with lower war-priority diversion cost...
Overmod It would be interesting to see the details of the GE design, as it would have relatively little in common with the Union Pacific/GN high-pressure condensing locomotives of the late Thirties.
Overmod Just as with the Q2, much of the rationale for a high-speed very-high-horsepower single locomotive in freight service 'went away' when war priority traffic did. Thereafter, a 70"-converted J class did a steam-locomotive's work just about as well, and more cheaply, for the short while steam still made economic sense………….why even a smaller-scale test of the V1 approach was not made.
OvermodI assume you have acquired and read the PRR report of the N&W J testing. It is interesting reading, and I believe at least a synopsis has been provided here in the forums. Contrary to some of the typical railfan wisdom, the report seems quite fair and does not disparage the J design, although some of the finer points of the balancing philosophy seem not to have been comprehended well.
Overmod …What might have been fun would be to adapt Glaze's balancing, and Timken rods, to one of the PRR J's (the 2-10-4s) and see the effects of the faster running. I suspect there were very good reasons this was not seriously undertaken, starting with Chapelon's implicit conclusion regarding roller-rod practicality that very serious cycling lateral flexure in the Timken rods was a necessary everyday operating occurrence... ... but considering what the English found they could do with speed of ten-coupled power with a two-wheel lead truck a decade later, it might have been interesting.
Jones1945Two experimental Class K5 4-6-2 passenger locomotives built with poppet valves and Caprotti valve gear; designed by W. F. Kiesel.
Just as a note, only one of these was built with the Caprotti poppet-valve gear (and was perhaps inevitably nicknamed 'Mussolini'); there were several other test installations of this kind of valve gear in this general time period, with none being markedly successful. (This was around the time that Baldwin acquired the rights to Caprotti gear and tried it on a wide range of product -- equally unsuccessfully.) If I recall correctly it was in connection with K5 testing that Caprotti noted 'your locomotives pull houses, not cars' -- and the detail design of the Italian version of the gear was in some respects not up to North American loads.
I have found it interesting that no attempt was made, even for test purposes, to fit any version of the Franklin System to one of the K5s, or for that matter to one of the M1as where the better volumetric efficiency might have meant something important. How much of that was a bad taste left from the Caprotti 'experience', I can't directly say.
Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!
Get the Classic Trains twice-monthly newsletter