You folk have been very helpful to me in the past, and I’m posting this in the belief that you will have some further useful commentary. Its aim is to throw some light on the development trajectory of latter day US steam.
I was recently sent copies of the PRR test reports on the Franklin valve K4, the T1, and Q2- something of a holy grail for me finding these. There are also snippets of information on the tests of the K5, which I see as the missing link in the PRR chain.
The particular interest for me was whether the supposed benefits of the Franklin poppet valves were real. I have written a report on this, and what is below is a rather lengthy summary of what I discovered. The bottom line is that much of the hype about the poppet valves is unjustified. They did change performance characteristics, but, at that time, the leakage problems that had dogged such designs since they were introduced in the 19thcentury were not solved, and they provided no advance in engine efficiency. There were in addition well known mechanical issues on the T1s.
In addition to the test reports themselves, my additional sources are primarily a random selection of ‘Keystone’ articles, and several from Railway Mechanical Engineer, plus articles by distinguished US authors in the post steam era.. But some of you may have access to other primary sources.
Please do not interpret this as in any way a criticism of the T1 new build project. This is an amazing effort which deserves everyone’s full support.
I would be happy to forward the full report to anyone interested, though fear its primary use would be as a cure for insomnia.
Summary
· This paper analyses plant and road test data in Altoona Bulletins on the PRR K4 and Duplex T1 classes with Franklin A poppet valves, to see what benefits the valves gave. Comparators are the original K4, a 1930s modified K4 and a Walschaerts Duplex, the Q2.
· The context of this work was the belief in the late 1930s that US passenger trains of the future would need to run at 100mph, and load to 1000 US tons.
· A 1930s piston valve K4 was tested on a 1000 ton test train for the AAR. It developed about 2800IHP at 75-80mph in ca 40% cut off, a steam rate in the high 40000s lbs/hr. It could not reach 100mph on the relatively flat PRR main. It was noted that the steam rate achieved was some way short of the 65000lbs/hr achieved on the test plant, and that this could not be achieved without use of excessively long, inefficient cut offs.
· The resistance of the coaching stock was established on these tests, which showed that over 3000EDHP would be needed to move a 1000 ton train at 100mph on the level. With 2000HP for the locomotive, plus acceleration needs, this gave target cylinder power around 6000IHP.
· It was believed that poppet valves were to be preferred for high speed running and would use steam more efficiently in part because more is admitted at a given cut off, allowing more economical cut offs to be used. This also helps to deliver the enormous power required.
· Lentz poppet valves, initially with oscillating and then rotary cams had been tested in the UK, with Gresley taking the lead. Oscillating cams were unsuccessful, rotary cams on the D49 showed more promise, but showed no efficiency savings, and were quickly removed from P2 2001. Leakage was the suspected problem. The D49s would likely have been rebuilt had other priorities including WW2 not intervened.
· In the US, the Franklin valve company visited Lentz in London, and produced their own oscillating cam version, the Franklin A valve. The Franklin Valve company fitted K4 5399 with these valves and improved steam circuit, and it was road tested in a similar fashion to the standard K4. It produced about 3400IHP, and achieved speeds of 85-90mph. Later plant tests imply the steam rate was 55000+lbs/hr, much higher than the original K4, with cut offs again approaching 40%. That is, the higher power was mostly due to the higher steam flow the poppet valves allowed, in turn allowing boiler to be steamed harder.
· The increased performance of 5399 meant it was sufficiently powerful to eliminate wasteful double heading of K4s on premier trains. So although the boiler would be operating inefficiently at the high steam rates required, it was likely overall much more economical.
· Before being tested on the Altoona plant, 5399 was then fitted with a new sine wave superheater, which eliminated much of the serious pressure drop to the steam chest on earlier versions, and a much wider exhaust.
· It proved possible to steam this machine at 77000lbs/hr (1100lbs steam/ sqft grate/hr). It showed marked improvements in engine efficiency over the Standard K4 at a steam rate of 70000lbs/hr, with particularly marked advantages at high speed. This was taken as a resounding endorsement of the suitability of Franklin valves for a future high speed design, and was a critical factor in the decision to fit the duplex T1 with Franklin A valves.
· There was clearly a great sales push by the Franklin Valve company to recoup their development effort, but it reads like the PRR’s desire to gain a breakthrough technological lead made them more than willing partners.
· However, as a number of contemporary observers had suggested, analysis of the test plant results with modern techniques says:
o The increase in engine efficiency on the test plant over the Standard K4 was entirely due to higher steam chest pressure and wider exhaust, not the Franklin valves.
o The higher steam flow at a given cut off allowed by the Franklin valves ought to have given a further significant improvement in efficiency by reducing the very long cut offs needed to work at high steam rates. This does not appear, and analysis of the data says that, as with other poppet valve set ups of that era, steam leakage was occurring which negated any benefit that the lower cut off gave.
o The enhanced efficiency benefit of 5399 at very high speed was again not due to the poppet valves, but rather that 5399 was operating at much lower back pressure, thus lowering the over compression that occurred at high speeds on the original K4.
o In any event, 70000lbs/hr is quite impracticable in service from a 69 sqft grate.
o By inference, the superior performance of 5399 in road testing was almost entirely due to the fact that the greater steam flow at a given cut off given by poppet valves allowed 5399 to be steamed at higher rates than the standard K4. A similar result could have been achieved by raising the boiler pressure to 250 psi, and indeed Altoona testing of a 250psi K5 Pacific showed it had superior efficiency to all K4s tested, but changes to the boiler design on the K5 made it a temperamental beast.
o It is not possible for 5399 to have averaged 100mph for 25 miles on end as claimed, nor does the time course data support this claim.
· Based on the contemporary interpretation of the 5399 results, two prototypes of a futuristic high power high speed Passenger locomotive, the T1, were built, incorporating both Franklin A valves and also a four cylinder Duplex drive. Both features would enhance the flow of steam at a given cut off, hence increase power at that setting. Driven in the same cut off as a K4, a T1 would produce nearly twice the power at speed.
· Prototype 6110 was tested on the Altoona test plant. It proved able to meet the 6000+IHP at 100mph brief, although the boiler performance was deteriorating quickly at the high steam rates needed, with low efficiency and high levels of black smoke. Maximum service IHP would likely be have been no more than 5000-5500IHP.
· The high steam rates required needed a narrow blastpipe, acknowledged to be sub optimal.
· Even at 5000-5500IHP, and with a target load now reduced to 880 US tons, a T1 could have comfortably averaged 100mph over long stretches of the PRR Crestline to Chicago main, had the necessary investments in infrastructure been made. (They never were).
· At a more realistic 4000-4500IHP, major schedule improvements with plenty of 100mph running would still have been possible.
· Its higher pressure, superheat and large grate meant the T1 was far more economical in coal consumption than double headed K4s, hence a major step forward. Note however Altoona felt the need to test 6110 with reduced superheater area, reducing efficiency, because superheat was ‘somewhat higher than is desirable from a maintenance perspective’.
· The massively high powers achieved by 6110 were far higher than anything needed on PRR schedules of that era. The Altoona report acknowledges this in as much as it was recognised that at the very low cut offs needed in service at full boiler pressure (due to valve design and four cylinders) ‘the action of the valve was not entirely satisfactory’. For this reason, tests were done at 150-170 psi steam chest pressure to establish what was needed in practice, this negating the efficiency benefit of 300 psi boiler pressure. At these lower in service steam rates, too high superheat and back pressure would not have been problems.
· As with 5399, analysis of the results says that the engine was not as efficient as it ought to have been. Using test results on the Walschaerts Q2 and K4 as a comparator, it can be shown that the Franklin valves on the T1 were also leaking to a degree which significantly reduced overall engine efficiency, negating the benefit of shorter cut off, so contrary to even recent commentary, did not provide an efficiency benefit, even at full boiler pressure. This coupled with the fact that four valves and cylinders will inevitably leak more than two means that the T1 was likely no more efficient than a PV 4-8-4 working in longer cut off, and under typical service conditions, quite possibly worse.
· The combination of two radically new technologies, Duplex drive and Franklin valves led to a number of operational issues with the 50 T1s purchased on the back of the 6110 results. Development work was done to eliminate these problems, including a newly designed rotary cam Franklin B valve, but the T1s were made redundant before they were even built by the purchase of diesels, so it was all irrelevant. The leakage problem was not recognised, and whether it was a tractable one is not known, nor whether Franklin B solved this problem.
· Overall, one can say that the misinterpretation of results on 5399 meant the speed and economy logic for poppet valves on the T1 was quite spurious. Further, leakage of the poppet valves, which paralleled earlier UK experience with both oscillating and rotary cam valves, had not been solved at that time.
· More fundamentally, one may speculate that the creation of 5399 was a consequence of the failure of the K5, which could have matched 5399. This failure has been linked to deficiencies in its boiler design, consequent on the firmly held belief that, since greater evaporative capacity was needed the evaporative surface area of the K4 boiler had to be increased. Modern analysis says this is not so; heat transfer does not limit evaporation and power.
· The British persisted with poppet valves, and a class of about 30 4-6-0s was built with valves to a modified Caprotti design in the 1950s. They worked in service for a decade without adverse comment, the right kind of compliment! A Pacific with the same set up showed good efficiency when tested. It was suggested that the modified Caprotti valves might seat better (i.e. be more steam tight) than the Lentz arrangement because their weight is not carried on the valve spindle.
I'm way out of my depth, but I do have a question that we may get around to discussing, perhaps needing an entirely separate thread...but bear with me for one moment, please:
You mentioned that the cut-off for the K4 would be near 40%. This sounds high to me for a steamer at full throat near 70 mph. Is that figure realistic? Why do many steamers have cut-off capability all the way down to 15%? What effect would a mid-range cut-off impose on a steamer attempting to get up to track speed with 700-1200 tons behind it, assuming good traction, steaming, and about 4000 hp working for it?
DreyfusshudsonIt is not possible for 5399 to have averaged 100mph for 25 miles on end as claimed, nor does the time course data support this claim.
selectorWhy do many steamers have cut-off capability all the way down to 15%?
(Edit: it was a 1951 Oregon Div Spec Instr, and it said not to work them at less than 33%, to avoid "hot main pins". Page 17 of
http://wx4.org/to/foam/maps/1-Ogle/1951-02-01UP_Oregon_SI_10_OCR-Ogle.pdf
Haven't seen that on any other UP division; lots of other UP Spec Instr on that site.)
Thanks, timz. I have laboured, perhaps in error, under the impression that the figure of 15% was available in valve gear design because it was probably going to be useful at least some of the time, and that low it would have been near the 4 cycles per second (?) of the piston where the most horsepower is being produced on most steamers. But now that I think of it, there was reference to the 'company notch or position" for cut-off that the engineers were encouraged to use when up to speed, so what you say about the UP's policy makes sense.
The figure of around 15% is largely empirical and reflects the amount of admission needed to get the engine past the following dead center under load; it's got little to do with 'practical' economical running. UP and other roads don't like short cutoff for another important reason (similar to why they may not like using reverse rather than partial throttle when starting or running heavily loaded at slow speed) - the torque produced by a typical quartered two-cylinder simple becomes very peaky as cutoff is wound toward mid, and when the throttle is worked open quickly and things warm up so that steam pressure (net of superheat pressure gain) at the piston face is an appreciable high percentage of theoretical, the engine will become prone to slip with very high short-period acceleration at the periods of highest applied force through the main pins.
British Caprotti famously boasted that it offered precise events down to somewhere in the 3-5% cutoff range. This of course isn't a measure of economy, it's a measure of how precise the admission and cutoff can be timed when the gear is set up (and the example on Duke of Gloucester 71000 is, in fact, that precise). You would never use cutoff that short - the engine would stall. Note that at high speed, the steam mass flow requirement turns around at some point and then increases to where the exhaust characteristics balance practical admission flow -- on Mallard this was, famously, about 40% at 'record' speed, and you can almost take that neighborhood of cutoff setting as the one you'll observe producing practical top speed during testing.
There are a couple of meanings for 'company notch' depending on who's spewing the tale: the one I remember hearing from old Railroad Magazine stories was the 'down in the corner' position on a lever reverser, where the engine would be working steam like crazy but producing maximum torque to haul the longest train up the steepest grade at what would eventually be slow drag speed but never quite a stall. The 'college boy' version of what the company notch was is the reverser setting that provides the most efficient running (best water rate, least smoke, or whatever) and presumably saves the company money -- this is for example what a properly-set-up Valve Pilot device would show and then help maintain when running.
Thanks for that, really appreciate it. Now, back to our OP's intentions for this thread...and I thank him for his forbearance.
1949 PRR DESCRIPTION AND OPERATION OF FRANKLIN TYPE A ON T1
The Pennsylvania Railroad's 1949 Machinery Examinations For Locomotive Fireman had several pages and a number of large fold-out plates describing the "Franklin System of Steam Distribution - Type A - Class T1 Locomotives".
An excerpt from the operating information may be of interest to forum readers.
" Instructions for Operation "
"There is very little difference in the handling of a locomotive equipped with the Franklin System of Steam Distribution as compared with a conventional piston valve locomotive.
This system allows the use of very short cut-offs (approximately 10 per cent). For most economical operation at speed, the throttle shoiuld be kept wide open, and the locomotive should be operated with as short a cut-off as possible. The shortest running cut-off is mid-gear (marked on reverse gear plate Mid-gear Forward and Mid-gear Backward).
Due to short cut-offs, the exhaust of a poppet valve locomotive is likely to be soft. Therefore, the depth of the fire must be regulated accordingly and should be as thin as possible.
When starting the train, it is nesessary to have the reverse gear in the maximum cut-off position (full gear). As soon as the train is under way, the cut-off should be shortened gradually.
Upon shutting off the throttle, move handle of three-way drifting control cock into drifting position.
During long drifting periods, move the reverse gear into the position indicated on the reverse gear dial by "Drifting". When running in one direction, never pass drifting position (located centrally between mid-gear forward and mid-gear backward on reverse gear indicator) toward the other direction.
When the engine is left by the crew, the reverse gear and the drifting control valve should be placed in drifting position."
The operting instructions continued with information on how to operate the T1 with only one unit in case of damage. Also included was information on lubrication such as, "The cam boxes are flooded with valve oil pumped through the distributor pipes in the cover plate." etc., etc.
Selector- The 40% cut off I quote was on a test train when they were trying hard to deliver maximum power. I would think this was uncommon, but there is very little detailed timing of the running of US passenger trains that would allow an assessment of how much power was normally being developed, hence what the cut offs were. The K4s were pretty out of date and underpowered by the 1930s (sorry PRR fans), so much so that they sometimes had to be double headed. So I would guess single headed they would need to be worked pretty hard. At 30-35% cut off they would develop 2200-2600 cylinder HP at 80mph. Power in this range will allow you to run at 75-80mph with an 800 (US) ton train on level track, though would take some time to get there.
As for the 15% cut off question, that's what the valve gear allows, and if you want to reduce power, shortening cut off is one thing you can do. Some UK designs were able to be worked in short cut offs like this, but many suffered from rough riding worked this way, and so if reduced power was required, it was more common to stick with ca. 25% cut off and reduce the steam chest pressure (Throttle). Working this way leads to a small loss of efficiency.
Many North American designs I suspect needed to be worked in longer cut offs, since two cylinders were pretty universal, and at maximum permissible tractive efforts (determined by adhesion considerations) you need longer cut offs to get the requisite steam flow and power with two cylinders.
From the very limited data there is, I suspect 4000HP (perhaps 55000lbs/hr evaporation with a feedwater heater) is about as much as was normally developed. The NYC report on Niagara testing said that 4000HP was 'above average working'- about 30% cut off for a Niagara at 80mph. 4000HP is sufficient for a big 4-8-4 to cruise at 90-95mph on level/undulating track with 1000US tons. With a grate of 100 sqft, the evaporation rate required to achieve this would be not too taxing.
There are some details of a couple of runs with ATSF 2900s working uphill from Dodge City to La Junta with 900-1000 ton trains. The gradient, though very gentle, needs about 1000HP to overcome. 2921 began at about 4000hp, rising to 4500hp for the last 100 miles. This gained half an hour on the schedule. 2909 made a similar time, but gained an hour on its schedule. Working was less uniform, but after Syracuse had a couple of spells at 5000HP for 5 to 10 minutes. With the massive cylinder capacity and long lap of the 2900s, this will have required no more than about 30% cut off. Given the time gains, I suspect that this was exceptional working, showing what could be done, not the norm, but there is just not the data to be sure.
Some very high powers were also developed by 3776 climbing single handed to Cajon on test, though I think trains on this section were often assisted.
Timz
By time course I mean the passing times at all points along the route.
nhrand.
Many thanks for the additional info.
I have to say that i find the statement that 'There is very little difference in the handling... compared to a normal piston valve locomotive' very surprising in view of the fact that they go on to note that the system allows use of very short cut offs. As noted in an earlier response, a K4 would have to be driven hard in 30+% cut off to deliver even 2200-2500HP at speed. At 11% (nominal) cut off a T1 would deliver 3600HP at the same speeds- chalk and cheese if ever there was such. What were they trying to say?
The steam flow to the cylinders at a given cut off hence power is significantly increased not only by the Franklin gear, but also the fact that at equal TE, there is much greater flow into four cylinders. And, the cylinder capacity of the T1 is of course much greater,which increases steam flow further. All of which accounts for the massive difference in the power vs cut off characterstics of the T1 and K4
So, it seems to me, someone who had spent a lifetime driving K4s would need to drastically modify their driving technique for the T1.
It is also interesting that they say maximum cut off should be used on starting and then 'gradually' reduced. This would lead to very high wheelrim powers when accelerating, more so than on a PV 4-8-4 because of the enhanced steam flow from the Franklin gear and four cylinders at a given cut off, and if driven this way, it is perhaps not surprising they tended to slip.
It seems to me that the PRR operating information meant that the crews had a lot to figure out for themselves about these very different beasts- this based on their own test plant data.
Sorry BTW to all for the delays in replying- the automatic e:mail notification of inputs hasn't been working
You are correct on the lack of training. After reading seveal articals in the Keystone Mag. on T1, this consensus came out. Unlike how management views training of new equipment (think flight simulators for a Boeing 777), back then you learned on the job and because you can run a K4 you go run a T1. Some Engineers learned as they went and where very good handeling a T1, while other's just run them hard and the rest be darned.
Dear forum members,
There is a post in Classic Trains forum "PRR Duplexes and Experimental Engines Discussion ( S1, S2, T1 etc.)" http://cs.trains.com/ctr/f/3/t/271182.aspx discussing different experimental steam engines, including Pennsy's duplexes. Please feel free to check it out if you like. Your input and participation are always welcomed! Thank you for your attention.
Jones 3D Modeling Club https://www.youtube.com/Jones3DModelingClub
DreyfusshudsonIt is also interesting that they say maximum cut off should be used on starting and then 'gradually' reduced. This would lead to very high wheelrim powers when accelerating, more so than on a PV 4-8-4 because of the enhanced steam flow from the Franklin gear and four cylinders at a given cut off, and if driven this way, it is perhaps not surprising they tended to slip.
This actually means something a bit different, which is well-established wisdom in some parts of the world like Australia. That it's wisdom developed after the recognition of low- and high-speed slipping (and adjustment of things like FA and equalization to overcome the problem) on T1 locomotives should tell you there's more to it.
Typical "thermodynamic expert" revealed wisdom is to get the throttle fully open as fast as possible and then 'drive on the reverse' -- this produces theoretical minimum throttling loss/wiredrawing in the admission steam and micrometric cutoff control of developed power. The 'advocated' strategy is saying just the opposite: keep the reverse 'down in the corner' in the desired direction for a while, and drive on the throttle (which is the thing I think you're missing in the criticism here).
The problem with early advance of the cutoff is the problem you note: much peakier torque (with conventional valve cutoff methods) resulting in more propensity to slip under any given adhesion conditions at a given developed horsepower (or average demanded torque, to use a different physical metric). Keeping the cutoff 'long' does two things: it keeps the average torque more normalized over the stroke, and it provides a vast sink of steam (at a given throttle opening) so that any slip once it starts to propagate becomes quickly self-limiting (in the same sort of way that the LP of a Mallet is self-limiting).
Problem #1 with Franklin-modified Lentz gear is that the gear can have, and usually does have, inherent limited cutoff, and problem #2 is that the gear can, and usually does, have very positive and precise action up to comparatively high rotational speed (one of the salient tested advantages observed for the T1 over the T1a). This means that even with the cutoff as long as it can be 'wound' there may still be sufficient steam to keep a slipping engine out of re-establishing static friction ... until the aggregate throttle is closed enough that the non-slipping engine loses power or stalls.
Now, the T1 is fitted with Franklin Precision air throttle, so very fine modulation of the throttle at any point is technically easy, regardless of how the multiple cams are set or if the effective throttle opening is smoothly progressive with physical throttle opening (which we won't know until 5550 is actively tested under steam; my guess is that some fun empirical calibration will be needed to get this effect). So driving the engine right at adhesion limit becomes fully practical once you have reasonable slip indicators, a thing we have covered in other posts here... that is, it becomes fully practical with appropriate haptic feedback IN the physical throttle.
That is not something likely to be easy in the grapevine setup as applied to the T1, which I think needs appreciable 'augmented force' (as applied to the F107 controls) in both servo directions to be particularly good. That right there is the antithesis of a typical K4 throttle, which despite postdating the practical implementation of Wagner-throttle servo action required quite an amount of banging open and closed at small opening. (I never had the chance to run one, but I suspect there are some here who have, and can confirm this.)
One approach that I can NOT advocate is something Neil Burnell mentioned as a 'coping mechanism' some engineers tried -- letting the throttle pressure fall off approaching and at a stop, so the locomotive would start with, say, 185psi gauge during the critical runup to 35mph or so (in the process mirroring some of Tuplin's observations on sliding-pressure firing Niagaras). As you get to the critical speed where the T1 has rapid acceleration, the increased draft etc. will have developed fire and hence steam pressure up to where desirable... and off you go.
I trust you see (1) that this perfectly mirrors the idea of simply keeping the throttle part closed until up to reasonable speed, and (2) what delightful effect this will have on things like staybolts in parts of your boiler, with the pressure and necessarily concomitant thermal cycling at each start. As people accuse Livio Dante Porta of finding out, sometimes thermodynamic expediency leads to substantial maintenance despair, wiping out any 'where's my big savings' on the coal and water bill perhaps many times over...
DreyfusshudsonBy time course I mean the passing times at all points along the route.
Jones 1945 -thanks for the tip off- a great set of postings that it will take me some time to digest. I think there are a couple of points i would like to add, but need time to engage brain fully on these rather different topics.
The data I have are times- to the nearest minute only-at Vandale, Warsaw, Plymouth (these two the stretch in question), with an overall elapsed time of 91 minutes for the 118 miles from Fort Wayne to Liverpool- an exceptional effort after a slow start to Vandale.
The first point is that you cannot make any meaningful claim to average speed when times are only to the nearest minute- there could easily be an error of a minute between two points, which has a big impact on the average speed.
The second point is that if you work out the power needed to achieve what was claimed, it works out at about 4500 cylinder horsepower for a quarter of an hour. This is way in excess of what they were able to achieve on the test plant with a much improved version of 5399- and then only for a short period. Steam rate would be in excess of 80000lbs/hr. So, we are in dreamland.
The log, such as it is, analyses more sensibly if you assume maximum power was around 3500HP- as on the earlier tests with the 1000ton test train when they were trying to push the loco to its limits. The average speed then works out at about 91mph on this stretch- remember it is ever so slightly up hill, so this is a great effort. It could be a bit better than this, but nowhere near 100mph.
I suspect a bit of judicious rounding up of the Warsaw and Plymouth times was made to support the claim. Franklin were clearly in 'sales' mode and this was great publicity. As I read the papers, the PRR folk were equally keen to believe and support the claim. Times had been very tough and they needed something to cheer.
Overmod:
I agree that, for the reasons you give, full gear and part regulator is the way to start.
My only point was ( apologies if I wasn't clear) that if you drove a T1 in this fashion, in the way you would drive a K4, the wheel rim horsepower on an individual wheel would be much greater than on a K4. The load on the drivers is not that different, so other things being equal (they are not) you might expect more slippage. Also, because of the different characteristics of the valve gear, and the presence of four cylinders, you would also get more HP at each wheelrim than on a Niagara driven in identical fashion.
As to whether Duplexes, by their nature, slip more, I am quite unable to comment
I remember a Trains "Second Section" piece about a cab ride on the 20th Century Limited, where the locomotive got into a high speed slip and the correspondent was "guest engineer". The correspondents first impulse was to back off on the throttle, but the engineer said "no" because it would cause slack action. The engineer cured the slip by spinning valve gear wheel to lengthen cut-off. One explanation was this would soften the impulses and thus relieve the slip.
Dreyfusshudson Jones 1945 -thanks for the tip off- a great set of postings that it will take me some time to digest. I think there are a couple of points i would like to add, but need time to engage brain fully on these rather different topics.
You are welcome, Dreyfusshudson. Take your time, It's not urgent.
I am a user of Trainz simulator, I understand that it is not a professional mechanical or automobile engineering program but it is the most realistic simulator available for trains simulation. I believe many of our forum members using this software like me. One of my projects is to simulate PRR S1 6-4-4-6 #6100 by using all actual figures of the real engines specification I gathered from books, forums, and internet.
Starting the engine (*1460 tons passenger stock behind ) with a maximum cut-off at full throttle caused the engine wheel slipping endlessly and stuck in the same spot. The only way to get the train moving forward is to adjust the cut-off to 26%-32%, full throttle, the S1 (in the computer) will start moving gradually and reach 60mph by increasing the cut-off from 26% to maximum cut-off. The engine can reach 105mph or above after 12-15 mins on level track by gradually decreasing the cut-off to 41% or less, depending on the gradient of the track.
In the realistic mode which allows derailment, my S1 derailed at every single sharp curve and it is not easy to restart the train once she stopped in random places; sometimes I have to spend 5 mins to get her moving forward again by adjusting the throttle and % of cut-off.
erikem
Passenger comfort before what I believe is the tried and tested way of dealing with this situation? interesting!
Jones 1945,
Thanks for this- I haven't paid much attention to the S1 in the past, but have had a quick attempt to simulate its performance.
My simulations are on Excel, using equations for resistances of coaches and loco, and the power developed by a loco in a given cut off and speed. The coach resistances are taken from standard sources (US trains are a nightmare because of the variety of coach types and the power needs for air con), loco resistance from equations of my own devising based on UK data (the US data I have seen on this subject is not entirely reliable, but the ball park is clear), and the power equations are computed from first principles- I get good agreement with a wide variety of locos on both UK and US test plants, and I think the extrapolations to types that weren't tested will be pretty good. For the S1 I have had to make a few guesses because I don’t have all the data, but these will be second order effects that won’t change the big picture much I think.
We do know what the resistance of PRR P70 stock on PRR track is from the AAR tests (a bit less than the Davis formula), and using this I get the following powers needed for an S1 with loads of a) 1460 tons and b) 730 tons to sustain various speeds on level track -indicative rather than precise figures, not least becasue all the data on resistance runs out at 80-90mph, and so the high speed estimates are based on a big extrapolation. We do know that resistances tend to increase with the square of speed, so that 140mph should be roughly twice 100mph is not unreasonable.
S1 plus 730 US tons
speed mph
IHP
EDHP
Loco Res
80
2656
1326
1331
90
3478
1729
1751
100
4465
2214
2253
110
5633
2790
2846
120
7000
3467
3537
130
8583
4253
4334
140
10404
5157
5246
S1 plus 1460 US tons
4016
2652
1365
5250
3458
1795
6734
4428
2310
8493
5581
2917
10554
6935
3626
12943
8507
4443
15684
10315
5379
The sources I have give the maximum cylinder power (IHP) as 7200 at 100mph, which in my hands needs about 32% cut of and 100000lbs/hr steam- should be well within the capacity of the 132 sqft grate. The above says that an S1 could cruise at 100mph on level track with 1460 tons, but it would take quite a time to get there. The most favourable stretch of track on the PRR from Crestline to Chicago is perhaps from Bucyrus down to Upper Sandusky, 15 miles at 1/470 worth about 2000HP at 90mph with a 1460 ton train, 1300HP with the lighter one. You will be able to work out from the above Table what the prospects for the claimed speeds of 130+mph are!
Using my equations, I can get 1460 tons up to 90mph on the level in 9 miles working flat out, but the acceleration to 100mph is painfully slow and takes another 10 miles.
So there would seem to be some disparity between your calculations and mine. Happy to try to work this through off line if that would be of interest- it’s tedious stuff! Depends if you know the provenance of the equations in the TRAINZ simulator?- I am completely unfamiliar with it.
In my computer, locos neither slip nor derail by the way- my tracks are all arrow straight, but do go up and downhill according to gradient profiles.
Dreyfusshudson Thanks for this- I haven't paid much attention to the S1 in the past, but have had a quick attempt to simulate its performance. My simulations are on Excel, using equations for resistances of coaches and loco, and the power developed by a loco in a given cut off and speed. The coach resistances are taken from standard sources (US trains are a nightmare because of the variety of coach types and the power needs for air con), loco resistance from equations of my own devising based on UK data (the US data I have seen on this subject is not entirely reliable, but the ball park is clear), and the power equations are computed from first principles- I get good agreement with a wide variety of locos on both UK and US test plants, and I think the extrapolations to types that weren't tested will be pretty good. For the S1 I have had to make a few guesses because I don’t have all the data, but these will be second order effects that won’t change the big picture much I think. We do know what the resistance of PRR P70 stock on PRR track is from the AAR tests (a bit less than the Davis formula), and using this I get the following powers needed for an S1 with loads of a) 1460 tons and b) 730 tons to sustain various speeds on level track -indicative rather than precise figures. S1 plus 730 US tons speed mph IHP EDHP Loco Res 80 2656 1326 1331 90 3478 1729 1751 100 4465 2214 2253 110 5633 2790 2846 120 7000 3467 3537 130 8583 4253 4334 140 10404 5157 5246 S1 plus 1460 US tons 80 4016 2652 1365 90 5250 3458 1795 100 6734 4428 2310 110 8493 5581 2917 120 10554 6935 3626 130 12943 8507 4443 140 15684 10315 5379 The sources I have give the maximum cylinder power (IHP) as 7200 at 100mph, which in my hands needs about 32% cut of and 100000lbs/hr steam- should be well within the capacity of the 132 sqft grate. The above says that an S1 could cruise at 100mph on level track with 1460 tons, but it would take quite a time to get there. The most favourable stretch of track on the PRR from Crestline to Chicago is perhaps from Bucyrus down to Upper Sandusky, 15 miles at 1/470 worth about 2000HP at 90mph with a 1460 ton train, 1300HP with the lighter one. You will be able to work out from the above Table what the prospects for the claimed speeds of 130+mph are! Using my equations, I can get 1460 tons up to 90mph on the level in 9 miles working flat out, but the acceleration to 100mph is painfully slow and takes another 10 miles. So there would seem to be some disparity between your calculations and mine. Happy to try to work this through off line if that would be of interest- it’s tedious stuff! Depends if you know the provenance of the equations in the TRAINZ simulator?- I am completely unfamiliar with it. In my computer, locos neither slip nor derail by the way- my tracks are all arrow straight, but do go up and downhill according to gradient profiles.
We do know what the resistance of PRR P70 stock on PRR track is from the AAR tests (a bit less than the Davis formula), and using this I get the following powers needed for an S1 with loads of a) 1460 tons and b) 730 tons to sustain various speeds on level track -indicative rather than precise figures.
Thank you very much, Dreyfusshudson! I have been waiting for a professional calculation to simulate the performance of S1 for so long and I finally got one! This means a lot to me! Thanks a lot! TRAINZ simulator is not a professional software so I could only get a brief and inaccurate result from it.
Under normal circumstances, the PRR S1 #6100; the preferred engine for the prewar premier overnight all-coach train the Trail Blazer ( Standard consist was 14 cars), was needed to haul a PB70ER passenger baggage car (154500 lbs), Ten P70kr or P70gsr betterment coaches ( 146100 lbs ), a twin unit dining car consisted of a D70dR Kitchen Dormitory (191400 lbs) and a D70cR Full dining car (169700 lbs), a POC70R - Parlor-Observation Car (149500 lbs), thus the whole consist of the Trail Blazer was weighed 2126100 lbs (1063.05 US tons). The reason I used the 1460 tons figure was that I tried to test the limit of PRR S1.
Base on your professional calculation result, can I just simply adjust the figure of IHP (as well as EDHP and Loco Res) by multiplying them by 0.728 (1063/1460) from the figures you provided like the following table?
S1 plus 1063 US tons
EDHP(unedited)
Loco Res(unedited)
2924
3822
4902
6183
7683
9423
11418
I would be really grateful if you could help me to calculate the time and miles of PRR S1 needed to reach 100mph (or above) with 1063 tons behind her! Thank you very much! It seems that PRR S1 was capable to reach at least 110mph hauling 1063 tons, this is compelling for me!
Wiki
His table estimates the power needed to pull the train at a given speed. He doesn't claim to have a good idea of the engine's power -- we're all guessing at that. You can guess as well as anyone, and you can calculate the result of your guesses. You'll assume the engine has so many pounds of drawbar pull at 1 mph, and the train has so much resistance, so you can calculate the rate of acceleration from 0 to 2 mph, and the time and distance for that. Do the same for 2 to 4 mph, and 4 to 6 mph...
An example
http://cs.trains.com/trn/f/740/t/242844.aspx
Thanks a lot for the link, timz. I think I could figure it out myself. : )
Move from other post:
Thanks Jones 1945.
As tim z points out below, it’s not quite as you think- the resistance of the locomotive would not change, only the coach resistance would fall in proportion to the weight, and the figures themselves are not directly useful.
I have two answers to your question about how long it would take 6100 to accelerate 1065 tons to 100mph on level track.
The first is, if we assume that flat out working is 7200HP (I don’t know how this reported figure was arrived at), and that when starting you can put down about 70% of the rated TE without slipping (Pure guess, about the maximum normally applied ion the UK), then you reach 100mph in 11 miles, about 9¾ minutes. After that you can maintain 100mph with about 5400HP. During the acceleration, steam rate is in excess of 100000lbs/hr, and you are burning about 17000lb/hr top quality coal. This kind of time is what you might claim for PR purposes.
However, my sense of the way locomotives were actually expected to perform is somewhat different. Starting is about TE, sustained performance is about how hard you are prepared to steam the boiler. As far as I can tell, it was unusual for US boilers to be steamed much above about 600lbs steam/sqft/ grate/hr, sometimes less. (This figure refers to top quality Coal of about 13500BtU/lb; it would be less for lower calorific values). Above this specific evaporation rate, unburned coal losses rise sharply, and it may be that this was a prudent level to avoid high boiler maintenance costs, e.g. due to cinder cutting, by not asking too much of them- the draughting usually allowed up to at least 900lbs/sqft/hr- a whopping 120000lbs/hr for the S1.
Looked at this way, the Table below shows how a number of different locomotives might be expected to perform with a 1065 ton train on level track, again working at 70% maximum TE up to 20mph, maximum steam rate about 600lbs/sqft/hr beyond that. All numbers are just indicative of relative potential, not precise.
Speed after 20 miles and elapsed time
final speed, mph
elapsed time, mins
sustained HP
PRR S1
96.8
16.50
5500
NYC J3a
84.6
18.75
3750
Niagara
90.4
17.25
4500
T1 Franklin
86.9
18.00
4100
T1Walschaerts
86.2
K4
71.2
24.50
2400
Not even the S1 can reach 100mph, though all are still accelerating slowly, and the S1 would eventually get there. The Niagara comes out better than the T1 versions, because the PRR banked the supposed engine efficiency advantages of the Franklin valve and reduced their grate size from 100 to 92 sqft, thus, according to my criteria, reducing the maximum sensible steam rate.
Of course, on any real line there will be gradient fluctuations, which might enhance acceleration. On the Water Level Route and PRR Crestline to Chicago the gradients are gentle, but the eight coupled designs would all get up to 100mph after a while, and cruise in the low 90s. And, if a train is overloaded or running late, there is likely plenty of upward power reserve to knock a hole in these figures.
To explore all these factors, I have spreadsheets with a simple flat test track, also the PRR main from Crestline to Chicago and back. I can send you these if you wish, together with the equations which will allow you to drive any of the above locomotives to see how all this works for yourself- there is an enormous range of possibilities.
An important input in all this is what the speed limit on the PRR main was on this stretch. 1940s and 1950s schedules from Crestline to Englewood only require average speeds in the mid to low 60s, I suppose after the 1946 Naperville disaster, speed limit was 80mph- is this so?- What was it before then? The answer to the second question determines what the maximum power you need from you locomotives is. If it was 90 to 100mph, then faster schedules would seem possible with T1s and diesels.
To summarise, the answer you are looking for depends on exactly what question you are asking. Are you are asking ‘What would happen if an S1 was pushed to its limit? The Niagara, T1, Mallard and French Compounds were all thrashed to find out the answer to this question for them. You really would have had to do the experiment to find out the true limit of the S1. But, I would argue that the results obtained with the four types mentioned when tested this way tell you very little indeed about what their in service potential is. Steam fans quote these efforts as indications of their relative prowess. Chapelon used his results to try to convince his bosses that steam was a viable answer. But in all cases, these feats do not represent what they did in service.
If you are asking what kind of performance an S1 might reasonably deliver in service, then I think my second estimate is more realistic. The S1 is the only type that could realistically whizz 1065 tons along the level at 100mph, thanks to its massive fire grate. This would however only be possible if the line was fettled up to handle that, and other trains are making similar speeds. Until that happened, the S1 would have been constrained by other traffic.
I am no professional, by the way, just an amateur with no axes to grind, at least I hope so!
Some more thoughts in a follow on a note to timz.
Timz;
Well I think we can estimate the power the engine will deliver reasonably accurately. The Table below shows part of my first pass attempt at modelling the S1 engine output.
Cut-off
Speed
St.cons
BackP
Effy
%
mph
lb/hr
lb
25
60
65218
4689
5.47
13.8
30
76585
5447
8.12
13.5
35
88571
6129
11.74
13
40
101686
6740
16.12
12.4
45
113822
7343
20.71
12
50
128126
7874
25.81
11.4
15
70
46258
3283
2.47
13.9
70290
5126
6.48
82604
5958
9.68
13.6
95618
6688
14.1
13.1
110247
7369
19.08
124914
8025
24.44
11.9
48887
3530
2.77
74331
5496
7.37
14
87744
6357
11.21
101796
7138
16.18
117330
7883
21.71
51024
3732
3.03
14.2
77941
5785
8.24
90607
6741
11.98
107158
7492
18.04
124637
8268
24.08
12.3
52816
3892
3.27
14.3
80859
6009
8.99
95911
6887
13.95
13.4
111581
7762
19.61
12.9
130779
8560
26.11
12.1
83185
6179
9.65
98641
7064
14.91
115239
7961
20.92
12.7
84934
6300
10.17
100794
7193
15.66
13.3
118182
8092
22.05
12.6
Two things need to be said;
· This is a first attempt. I do not have lap, lead, exhaust lap, clearance and volume nor superheat vs steam rate for the S1- I have estimated them; these will have some modest impact on the efficiencies I calculate. I also assumed steam chest pressure was 290 psi across the board which is decidedly optimistic. Crucially however, I set the blastpipe nozzle area as 64 sq ins. This is very generous and allows low back pressure at high rates. If it is significantly less than this, back pressure will rocket at high outputs and the very high powers shown at 100000lbs/hr+ will be curtailed significantly. If anyone can supply further details on these dimensions, most especially the blastpipe design, that would be helpful to improve matters.
· It can be seen nonetheless that the engine is capable of delivering enormous power. You are therefore right to imply that these figures don’t tell you a great deal- the limiting power is all about the boiler output that can be achieved, which can only be established by trying!.
DreyfusshudsonI think we can estimate the power the engine will deliver reasonably accurately.
(Maybe when you say "engine" you don't mean "locomotive"?)
Rwy Mech Engr 8/39 says 6-4-4-6 valve travel was 7-1/2, lap was 1-7/8, lead was 5/16. Don't think the article gave the exhaust lap/clearance. A Zeuner calculation says that's 70.6% maximum cutoff.
timzYou think 8092 IHP at 120 mph is "reasonably accurate"? Why?
I believe he is using Bill Hall's Perform software (and has discussed how to use it with Prof. Hall). If you have a better package for analyzing reciprocating steam locomotives, or advice in using it or setting initial conditions, please provide it.
I think he is indeed using 'engine' to mean 'locomotive', as 8092ihp to just four wheels is patently nonsensical. Yes, technically the S1 has two 'engines' and the results are a slightly weighted sum of the two as indicated -- I suspect more adjustment of duplex performance from an 'equivalent' conjugated equivalent might be warranted but much of this would likely involve empirical factors...
Can you scan and post the corresponding Zeuner diagram, as it would be useful in a number of ways here? There are a number of comparable ones in the steam_tech Files section (calculated and drawn by Mick Duncan) but I can't link them here as it's a closed group. Maximum cutoff in his table is well shy of the built-in limited cutoff, I think.
Dreyfusshudson I have two answers to your question about how long it would take 6100 to accelerate 1065 tons to 100mph on level track. The first is, if we assume that flat out working is 7200HP (I don’t know how this reported figure was arrived at), and that when starting you can put down about 70% of the rated TE without slipping (Pure guess, about the maximum normally applied ion the UK), then you reach 100mph in 11 miles, about 9¾ minutes. After that you can maintain 100mph with about 5400HP. During the acceleration, steam rate is in excess of 100000lbs/hr, and you are burning about 17000lb/hr top quality coal. This kind of time is what you might claim for PR purposes. However, my sense of the way locomotives were actually expected to perform is somewhat different. Starting is about TE, sustained performance is about how hard you are prepared to steam the boiler. As far as I can tell, it was unusual for US boilers to be steamed much above about 600lbs steam/sqft/ grate/hr, sometimes less. (This figure refers to top quality Coal of about 13500BtU/lb; it would be less for lower calorific values). Above this specific evaporation rate, unburned coal losses rise sharply, and it may be that this was a prudent level to avoid high boiler maintenance costs, e.g. due to cinder cutting, by not asking too much of them- the draughting usually allowed up to at least 900lbs/sqft/hr- a whopping 120000lbs/hr for the S1. Looked at this way, the Table below shows how a number of different locomotives might be expected to perform with a 1065 ton train on level track, again working at 70% maximum TE up to 20mph, maximum steam rate about 600lbs/sqft/hr beyond that. All numbers are just indicative of relative potential, not precise.
Our community is FREE to join. To participate you must either login or register for an account.