3 Cylinder Steam- Why Europe and Not the US?

[Dr D]

Wow!

"4 four cylinder locos - and four T1 four cylinder locos fortified by preservation!"

And she has such big strong muscles!

Dr. D

[erikem]

Juniatha

I still prefer the idea they are four - except for in this case they should have four cylinders each .

Mmmm, maybe we can bring back the D&H 1403???? And the machine shop to keep her running...

First issue of Trains I bought had the article on the D&H high pressure compounds.

- Erik

[CSSHEGEWISCH]

As long as we're talking four cylinders on a rigid (non-articulated) locomotive, we should not forget any of these:  PRR S-1, Q-1, Q-2's, and B&O N-1.

[Overmod]

CSSHEGEWISCH

As long as we're talking four cylinders on a rigid (non-articulated) locomotive, we should not forget any of these:  PRR S-1, Q-1, Q-2's, and B&O N-1.

Or the PLM 10-coupled locomotive that predates them all.  (Not a "duplex", but clearly divided drive for the 'right sorts of reasons' on a rigid frame...)

[Paul Milenkovic]

(Channeling Homer Simpson)  Mmmmmm!  6 cylinders.  And 4 of them inside the frame!  Compounding and steam cylinder jackets too!

[Overmod]

Paul Milenkovic

(Channeling Homer Simpson)  Mmmmmm!  6 cylinders.  And 4 of them inside the frame!  Compounding and steam cylinder jackets too!

Isn't that one the twelve-coupled 160 A1?

[Juniatha]

Guys ,

The one was P.O.Midi 160.A.1 - a 2-12-0 , had a six cylinder compound unit with steam envelopes which - as I have understood vague enough descriptions in literature - on top of it could be set up in various ways for steam to enter before / or after working in cylinder and with superheated or saturated steam , probably a variability allowed for testing , not for a practical service application .  The boiler included a re-heater for receiver steam , too .   The engine did *not* have triple expansion in two cylinders each for high / medium / low pressure stages but the idea rather was to split up low pressure cylinder volume because the engine’s main purpose was to improve the steam loco’s perpetual weakness :  low speed performance and efficiency , which was attacked both by steam jacketing and by increasing tractive effort at short combined cut-off via increase of HP and namely LP cylinder volume and their volumetric relation . 

The 160.A for an engine of 20 t axle load was enormously powerful at already very slow speeds on rising grades without consuming undue amounts of fuel yet would run very smoothly at relatively elevated speeds ( though not tested for maximum speed attainable ) .  Wear in drive proved exceptionally low and in fact never needed a classified repair until – premature – shelving and scrapping .

The other was series PLM 151.A - a rather clumsy 2-10-2 ( no , not a 2-4-6-2 because it had those odd inside coupling rods , proposed in retrospective now and then as a cure-all for Pennsy’s T1 - a tongue-in-cheek effort at eating the cake and keeping it .  I’m not going to enter topic insane , sorry , inside coupling here , that may remain another story ).   The type was ‘successful’ because proud PLM by default never had locomotives built turning out a failure .   That’s why they didn’t dump their early ‘attempts’ in 4-8-2 type passenger locos but duplicated them faithfully ( PLM series 241.A , not to be confused with later SNCF 241.A of EST origin ) and only in the following tried to discretely improve them by revamping the drive layout mainly ( series 241.B ) and continued unswerving with further production including then further minor changes (241.D , skipping their one-off prototype 241.C ) and then rebuilding the D type into E type ( still no ways as dynamic as a Jaguar E type ) .  

Performance of the 151.A was in the range of DR 44 three cylinder Decapod if the latter was not fully extended , yet at higher fuel consumption with the compound .   Still , it worked and thus was successful if you looked at it from a suiting angle .  It soon got way out-performed by the 1936 Alsac-Lorraine three cylinder 2-10-2 with a formidably steaming combustion chamber boiler which preceded the DR 45 class which again proved quite economic in fuel consumption yet was ailing off production floor since born with a ‘weak heart’ by Wagner’s dictation , namely a boiler with way too small firebox and radiation heating surface . Some of them were rebuilt with combustion chamber boilers around 1951 – however like all other 20 bar engines rebuilt they were knocked ‘back in line’ with German ever-standard 16 bar boiler pressure , which robbed them 20 % of their up-hill ‘punch’ .

Regards

Juniatha

[Overmod]

Juniatha

The one was P.O.Midi 160.A.1 - a 2-12-0 , had a six cylinder compound unit with steam envelopes which - as I have understood vague enough descriptions in literature - on top of it could be set up in various ways for steam to enter before / or after working in cylinder and with superheated or saturated steam , probably a variability allowed for testing , not for a practical service application . The boiler included a re-heater for receiver steam , too . The engine did *not* have triple expansion in two cylinders each for high / medium / low pressure stages but the idea rather was to split up low pressure cylinder volume because the engine’s main purpose was to improve the steam loco’s perpetual weakness : low speed performance and efficiency , which was attacked both by steam jacketing and by increasing tractive effort at short combined cut-off via increase of HP and namely LP cylinder volume and their volumetric relation .

Some recent work by Claude Bersano and Thierry Stora has established that the full flow of steam to the high-pressure cylinders first passed around them, through the 'jackets'.  This eliminated much of the need for high superheat to keep HP wall and nucleate condensation losses especially toward the end of the stroke.  What this in turn allowed was a very small degree of primary superheat (the HP superheater was less than 72 sq.m!).  There are some discussions in which Chapelon indicates the engine could be worked 'saturated' with almost no thermodynamic losses due to the lack of 'conventional' elevated superheat in the HP steam.

The LP 'reheater' actually used a substantial number of Schmidt-type superheater elements, of shorter length than usual elements, in the lower portion of the boiler.  This resuperheater had much larger area than the primary superheater, which may have been partially an 'artifact' of the large cross-sectional area needed for LP steam flow.  Interestingly, the 'ideal' ratio between LP and HP swept volume was adjusted by making two of the four LP cylinders different in size (larger; the other two were the same dimensions as the HP cylinders).

Here is a longitudinal elevation/section of the locomotive (courtesy of Thierry Stora's site at chapelon.net).  I believe there is a better copy available and will provide a link if I can locate it.

[Firelock76]

SIX cylinders?  Ye gods, the mind reels...

[Overmod]

Firelock76

SIX cylinders?  Ye gods, the mind reels...

It's actually common sense -- if you need big LP cylinder capacity, divide it between multiple cylinders instead of having big dustbins.  A clever thing on 160 A1 is that it has one 'pair' of LP cylinders the same dimensions as the HP (for parts and machining commonalty, among other things) and that allows you to tinker with either or both dimensions of the 'other' ones to get your expansion ratio just where you want it.  As a fringe benefit you can phase the pairs of LP differently if you want smoother torque peaks...

[Juniatha]

 

Hi -

 

Well , since the thread is *not* about the 160.A.1 but about the question of three cylinder .. ( oh come on , see  by yourself ) I wanted to keep it concise , no more than just sorting out which loco was what .   I should have known though Prof O would jump at it , filling in what I had left out for shortness and since it was not topical here .  Ok .

 

>> the full flow of steam to the high-pressure cylinders first passed around them, through the 'jackets'. <<

Actually , as I wrote , that was just one option of arrangements put up and it was for saturated HP steam in fact – the same would not make sense with superheated HP steam because what would be saved of condensation inside of cylinder would be lost by condensation in the envelope around the cylinder , or actually more than that because the envelope’s walls towards the outside of necessity had to have a larger surface than the cylinder itself , resulting in a net loss by the effort .   As far as I know the one set-up that brought the rather unexpected result of practically no loss in thermal efficiency when totally omitting HP superheating was when the HP cylinder envelope was arranged to be filled by receiver steam downstream of the secondary superheater .   This way , not only was there no condensation in HP cylinder in spite of saturated steam but there was an *increase* in steam temperature during work – at least a relative if not an absolute one – by enthalpy boost provided by the envelope's superheated steam.  

Still , heating the envelope by the very mass of steam that was used for work in cylinders must have cost some heat drop and I’m not fully convinced test data in this rather special arrangement of steam flow always provided a complete and fully true image of what happened – mind this was at a time long before electronic equipment for precision measurements had become available and it must have taken quite some ingenious rigging up to get valid data with the rather plain means of that time .   

Superheating steam also increased volume .   Without it in the HP stage and having it on receiver steam would ask for a more stretched out relation of LP to HP cylinder volumes than with the regular pracetic of superheating boiler steam and leaving it alone in the receiver stage .   Giesl in his book on loco types of high degree coupled axles remarked HP volume looked somewhat small to him – this might just have been well befitting working in this unusual setting-up of having but secondary superheating and that again could have helped improve specific steam consumption in that special mode of working .

 

>> This resuperheater had much larger area than the primary superheater, which may have been partially an 'artifact' of the large cross-sectional area needed for LP steam flow. <<

An artifact – uhm , *g* , okay ..  Well , sure the lower pressure receiver steam asked for a larger cross section of elements for adequate steamflow while to avoid excessive steam temperature which would only result in a lot of superheating left in exhaust steam , elements were not to be extended nearly as far down to the firebox tube plate as usually , although dimensioning also had to take care of the special arrangement of tubular section of boiler including preheating in the 160.A.1 .

 

 

>> Interestingly, the 'ideal' ratio between LP and HP swept volume was adjusted by making two of the four LP cylinders different in size (larger; the other two were the same dimensions as the HP cylinders). <<

No , actually that was not the case .   First of all it’s not just the displacement volume that matters in relation between HP and LP stage rather than it’s total volume including clearance volume plus of course characteristics of valve gear , namely as concerns degrees of compression used in HP and LP .  

While in a typical four cylinder compound the LP unit worked much like a two cylinder simple expansion engine on rather mild steam chest pressure , compression in HP cylinders was always crucial for good working balance in a compound and usually involved an increased % clearance volume .   

Secondly , the differing cylinder volumes in LP inside / outside cylinders were not meant to >> adjust << volumetric relation – the same relation could have been realized with all LP cylinders of identical size if only that size was itself chosen to suit intended relation .   Having LP in two different sizes simply was a consequence of crank displacements :  90 degrees with HP unit making torque profile similar to that of a two cylinder SE engine ;  120 degrees with the outside LP cylinders and 180 degrees with the inside LP cylinders which in this case provided a total of piston thrust same as one of the outside cylinders making LP unit torque effectively similar to that of a three cylinder SE engine , again if working on but milder steam chest pressure .  All in all , the 160.A thus worked rather like a five cylinder engine , a configuration proposed by A. Wolff for a high speed 4-8-4 in an academic investigation on possible high speed rail traffic .  

Actually , since in the 160.A inner LP cylinders had been made of same diameter and stroke as HP cylinders and their combined piston thrust was supposed to be the same as that of one outside LP cylinder , volume relation should have been 3 : 1 straight .   Yet it was noted officially as 2.82 : 1 – seemingly an odd value it resulted from usual plain calculation of cylinder volume as r² p x s without deducting piston rod volume and then as piston swept volume only , without accounting for larger clearance in HP cylinders of a compound engine .   Yet , accounting for these factors still left a small volume surplus with inside LP cylinders over one outside cylinder and that was due to the former being made identical in d x s to HP cylinders although with identical diameters alone a small adaptation could have been made in stroke since each the crank axles were individual pieces differing in crank displacement anyways .  Yet again it looks prudent to have left a small surplus with the inside ‘split’ LP cylinder to help account for larger total steam leakage losses wth two smaller than with one larger outside cylinder .

 

In view of the topic of this thread actually being the SE three cylinder engine not very special compound engines and lest it will become rather special reading , I’ll stop here , hoping Prof O will also respect that .

Regards

            Juniatha

 

 edit : sorry , somehow HP / LP got swivelidizzied at two points when writing , corrected in blue

[Overmod]

Juniatha

In view of the topic of this thread actually being the SE three cylinder engine not very special compound engines and lest it will become rather special reading , I’ll stop here , hoping Prof O will also respect that .

I concur.  At least I got you to post it!

I think we should have a new thread on this topic, or revive an old one that is apropos, and you should reproduce these comments there.  I know of no source on the Web, in English, that describes this so well, and I think it's valuable to have it, but I'm not going to cut and paste someone else's post into another thread myself to do so.