Franklin B rotary cam poppet valves - Trains Magazine
The translation is by Carpenter. It does not redraw the French graphs and exhibits, much as the Wasatch translation of 'Rear Boiler Knowledge' from Glasers Annalen does not. Among other things it contains Chapelon's useful Sherlock Holmes deduction on how Timken thin-section rods can possibly work with roller bearings.
All the calculations were to be in 'part 2' of LLAV, which was never reprinted after the original edition (in French) and as I recall much of it was never printed at all. Discussion of this on steam_tech was extensive but now 'lost to science' with the collapse of Yahoo Groups; it remains to be seen if the archives or collateral are restored on groups.io. I recommend that you review anything remaining on the Web by Thierry Stora or Claude Bersano before it, too, gets taken down...
I was sent a page scan of the 1938 edition to do a full translation of the 'missing' part ... and should get back on it!
GPCS is in part founded on a mistaken understanding of heat transfer in a locomotive boiler; it resembles some of the misunderstanding of the thermodynamics of Besler tubes. Effective heat release is only part of the concern; it must then be taken up effectively and then communicated to water in ways that facilitate generation of suitable 'power steam'.
Products of combustion transparent across a wide portion of the emission spectrum are not as effective at radiant heat transfer as the luminous flame from vaporized (but unoxidized) carbon. Uptake from transparent gas is likewise peaky in the tubes; hence heating of a blackbody followed by re-radiation nearly normal to a (similarly black) tube wall has enhanced heat transfer benefits in addition to increased convective transfer in an annular restriction. In a boiler where there is no practical effort to recover the heat from the (transparent) H2O in the gas stream -- in GPCS this deriving not only from combustion but the 'process steam' introduced in the cellular windbox -- still more Rankine efficiency is thrown away.
Wardale acknowledged at one point that GPCS was a chemical reaction, not a firing method. Just as chemical reactions benefit from intricate process control, and often go haywire without it, so we can reasonably expect large-scale GPCS in tight packaging to be; if we now expect it to follow what is already a complex load-following steam demand it may become clear that no human fireman can consistently make all the various often interrelated adjustments in realtime... or perhaps better stated, shouldn't be expected to.
This isn't to say that some of the ideas of GPCS are mistaken. The idea can be adapted to produce a luminous primary plume with the addition of preheated primary air in addition to process steam, and to preclude clinkering and other ash-glassing problems (it would be nice to have cyclone firing on a locomotive but there are considerable issues!). Part of the reason for a cellular windbox is to permit different parts of the bed to be selectively 'blown' at different times and at different rates.
The angled stacks are not 'new' to Wardale and Girdlestone; N&W famously used them in a considerable range of power, although offhand I can only remember Porta using one purely to increase the effective accelerating entrainment length in the front end purely for loading-gage reasons (the others usually being to clear parts of the feedwater heater, as with the Australian Hudsons) Note that American use of stack extensions, particularly on ATSF, accomplishes the 'same thing' where loading gage permits.
Fascinating in a way that I get all the way here without mentioning the Franklin System at all, let alone the flavors of RC actuation. Note that one of the issues with OC Franklin, as with Lentz/Lenz before it, involved a certain amount of scam. The premise was that radial valve gear could be used to drive the System's poppet valves just like piston valves, with shorter cutoff producing different mass flow to the cylinders but with timing precision. The problem is that with poppet valves, opening them incompletely or slowly doesn't produce good effects; a poppet valve needs to lift cleanly and return sharply without bounce or float, and no SHM valve gear provides this unless the actuating cams are cut with steep profile (in which the necessary modified-trapezoid cam profile is difficult to achieve) and this in turn introduces nightmare-box complexity into the drive mechanism to work.
Meanwhile the RC system can only approximate the continuous cutoff control of Walschaerts or Baker if it uses a continuous-contour cam (which uses spherical followers) and in more sophisticated variants variable actuation or 'variable geometry' on the cam-follower linkage. To my knowledge no system of fixed cams was particularly successful in practice, and the best of the 'workarounds' was an almost-appalling-to-purists dodge (actually patented!) which used 'wiredrawing' intentionally to give "cutoff" effect with increasing speed ... this was fine for War Department 2-8-0s that could be driven by inexperienced soldiers, but not so good for railroads concerned with best efficiency...
