I do not know how the control was arranged on 60000; perhaps Mike (wanswheel) can find something definitive. My naive idea would be to have a 'trim' control on the center cylinder that would allow it to be adjusted separately from the LP cylinders, but then have all the gears worked running by the same power reverse. Fortunately the engine still exists and someone doing research could at least in theory gain full access to the cab and running gear, and watch the result of adjusting the controls that are there, with HR video running to observe the valve spindle motion outside and inside, to see how the trick is done.
ACY The center cylinder is cranked to a driving axle, with the connection located between the frame members. It is angled in order to provide clearance. There were simple, as well as compound versions of this arrangement, but the compounds generally proved impractical in North American use. All survivors in this country are simple engines, as far as I know. Surviving examples include Alton & Southern 0-8-0 no. 12 in Kirkwood, MO; Baldwin 4-10-2 no. 60,000 in Philadelphia; Southern Pacific 4-10-2 5021 in Pomona, CA; and Union Pacific 4-12-2 9000, also in Pomona.
The center cylinder is cranked to a driving axle, with the connection located between the frame members. It is angled in order to provide clearance. There were simple, as well as compound versions of this arrangement, but the compounds generally proved impractical in North American use. All survivors in this country are simple engines, as far as I know.
Surviving examples include Alton & Southern 0-8-0 no. 12 in Kirkwood, MO; Baldwin 4-10-2 no. 60,000 in Philadelphia; Southern Pacific 4-10-2 5021 in Pomona, CA; and Union Pacific 4-12-2 9000, also in Pomona.
21. Walschaerts valve gear is used, with an independent motion for each cylinder, but all controlled by one type B Ragonnet power reverse gear. The valve for the left-hand cylinder is operated from the left-hand main pin and crosshead in the usual way. The right-hand valve receives its lead from the right-hand cross-head, but the link for this cylinder is operated through a transverse shaft, by means of a connection to the left-hand crosshead. The return crank on the right-hand main pin is set to operate the valve for the inside cylinder, and this valve is given lead through a connection with the inside crosshead. The arrangement of the links, arms and rockers through which these connections are made, is shown in the accompanying drawings. The valve motion bearer is a single steel casting supporting practically the entire valve gear.
Reading this in conjunction with the side view of the valve gear Fig 13a at the link Timz provided above:
So my interpretation is:
I had experience with Baldwin built 4-6-2s in Thailand in the '70s when they were in service. Originally they were used on the premier passenger trains. By the time I saw them they were working local service due to the introduction of newer engines. They had equal size cylinders with the steam chest to the left of the middle cylinder. There were two sets of Walschearts motion on the left side. The chug was heard in a synchopated beat. They must have been successful because all other locomotives purchased about the same time were long retired.
Never have tried to figure out BLW 60000's valve gear-- try
http://cwrr.com/Lounge/Reference/baldwin/baldwin.html
There's a side view drawing, which you'll have a hard time deciphering. Maybe the text will explain.
The Franklin Institute acquired Baldwin 60,000 in 1932
M636C Does that mean that the [BLW 60000's] HP cylinder was 180 degrees ahead of the right hand LP cylinder, and the left cylinder 90 degrees behind that?
The outside cylinders were 90 deg to each other-- the center cyl was 135 deg, plus or minus its inclination.
S. ConnorThis is probably more than you wanted to know
Not at all. Good post.
(Now come the nit pickers.)
ChuckAllen, TX
M636CDoes that mean that the [BLW 60000's] HP cylinder was 180 degrees ahead of the right hand LP cylinder, and the left cylinder 90 degrees behind that?
Several French compound 3- and 4-cylinder designs also drove the lead driving axle, including a class of PLM 4-8-2's built in the 1920's and early 1930's.
M636CNote that there are separate lifting links for the right and centre valve gears, suggesting that the cut off might be varied separately for the high and low pressure cylinders, although how this might be done isn't clear to me...
I would expect that a Smith compound setup like the one on 60000 would need separate HP cutoff, if for no other reason than the fixed expansion ratio will not be ideal. This would be best addressed by trimming cutoff and throttle opening on the HP vs. cutoff on the LP to best use the available 'exhaust'. I would suspect that a four-cylinder de Glehn-du Bousquet would have its HP and LP cutoff manipulated in analogous fashion.
Who's going to Philadelphia next?
M636C The D&RG Mountains had the same valve gear as 60000
The DRGW 4-8-2s had two ecc cranks and rods on the right side-- BLW 60000 had one set
http://digitalcollections.smu.edu/cdm/ref/collection/rwy/id/634
http://www.cwrr.com/Lounge/Reference/baldwin/fig13.gif
See what happens when you rely on memory. I haven't been to Philadelphia in more than twenty years.....
I'd never seen that drawing, of course... I've got copies of the photo of 60000 but never checked them closely. Enlarging that at the link to 100% is quite helpful.
60000 has three valve gears, but two of them are driven from the same eccentic crank.
Does that mean that the HP cylinder was 180 degrees ahead of the right hand LP cylinder, and the left cylinder 90 degrees behind that?
60000 has no conjugating gear, since the centre cylinder's valve has no connection to the left valve gear, only the right valve gear...
The link and crank visible on the right side of the locomotive look very similar to the drive for the Henschel conjugating gear, so the drawing is very important in showing what is actually there... The rocking shaft is a bit too far forward for the Henschel conjugating gear to fit there anyway....
Note that there are separate lifting links for the right and centre valve gears, suggesting that the cut off might be varied separately for the high and low pressure cylinders, although how this might be done isn't clear to me...
M636C
I'm surprised nothing has been said about the seven 3-cylinder 4-6-2s rostered by the Mexican Railway (FCM) which lasted in service until the early 1960s. They were wonderful to see and to listen to in regular passenger service. Before the FCM was absorbed by the NdeM these engines were well maintained and proudly cared for by their engine crews. They were assigned to daily passenger trains between Mexico City and Esperanza, with tight schedules requiring speeds of up to 100kph. I was privileged to two cab rides in 1960 on these beauties. The first four, numbers 130 through 133 were built as simple three cylinder engines with Gresley gear by Alco-Schenectady in 1928. Three more, 134 thriough 136 were built in 1938 by Montreal Locomotive Works to the same specs. Unfortunately, NdeM, while still heavy in steam in the early 1960s, they shoved the FCM Pacifics off to local freight service to run out their service life before scrapping them.
S. Kistler
pajrr Shop men from Baldwin actually changed out the bearings on the 60,000 at the museum years ago. As you stated, the limited movement made the original bearings egg-shaped. I don't know if they have been changed again, but I do know that they have been changed out at least once due to the wear and tear of "museum service".
Shop men from Baldwin actually changed out the bearings on the 60,000 at the museum years ago. As you stated, the limited movement made the original bearings egg-shaped. I don't know if they have been changed again, but I do know that they have been changed out at least once due to the wear and tear of "museum service".
That must have been quite a few years ago, no later than 1970 or so (If I remember correctly Baldwin closed its doors in 1972).
When did the 60000 go on display and start rolling? Her bearings are likely going or have gone bad again.
-S. Connor
M636CThe D&RG Mountains had the same valve gear as 60000
Who would change them after Baldwin left the steam locomotive business?
THREE CYLINDER locomotives in America were occasionally built before 1910 which heralded modern construction of 4 locomotives for the Philadelphia and Reading Railroad. These engines were 4-4-2 "Atlantic Type" locomotives and included one 4-6-0 "Ten Wheeler Type." These Reading Railroad engines were superheated and used Walschaerts outside valve gear while the inside cylinder used inside Joy valve gear.
THREE CYLINDER steam locomotive development in the United States was not such a unique event in railroad history as one might assume. THREE CYLINDER POWER design as we know it really deals with the product development of ALCO Locomotive Works and its competiton with LIMA and BALDWIN Locomotive Works in the years from 1922 to 1930.
In the 1920's when LIMA Locomotive Works was under the influence of William Woodward with his much heralded SUPER POWER concept. ALCO developed and was going strong with its own heralded THREE CYLINDER POWER concept as the way of the future.
Denver and Rio Grande Western in the market for Rocky Mountain pulling passenger power settled upon the 4-8-2 "Mountain Type" three cylinder design. BALDWIN Locomotive Works produced this successful three cylinder D&RGW class 1600 series of 4-8-2's.
This 4-10-2 demonstrator locomotive as it was envisioned by Baldwin was an extremely high technology engine that was never duplicated. The Baldwin 60000 demonstrator engine remains today brand new yet never sold at the Franklin Institute in Philadelphia, PA where Baldwin got rid of it through donation in what appears as a tax write-off.
I think the wrong idea is created regarding three cylinder locomotives if consideration is limited to those within the United States, since they followed a trend started overseas, and the great majority were based on the patented conjugated valve gear developed by Holcroft. As I indicated earlier, all the locomotives built in the UK and the USA using the Holcroft gear were outnumbered significantly by German built locomotives using a Henschel patent conjugating gear. In addition, there were a large number of three cylinder locomotives built in Germany and in the UK that used three separate valve gears, a number with the Baldwin arrangement (or a variation thereof) but many with the third gear inside (like the early Reading locomotives).
The German three cylinder locomotives were built during or just after the First World War, and were thus contemporaries of the USRA types, and like the USRA locos were a standard design supplied to most of the main line railways in Germany before the nationalisation of the early 1920s.
Apart from the Union Pacific, the three cylinder locomotives tended to be discarded and rebuilt as two cylinder locomotives, and the UP superseded the three cylinder locomotives with their well known simple articulated locomotives.
The Baldwin locomotives, while not discarded, lasted longer in general but were not repeated.
A few Baldwin 3 cylinder pacifics made their way to Thailand, where they probably met British built pacifics of the Malayan Railway, which included a large class fitted with rotary cam poppet valves which needed no conjugating gear (Although Gresley manage to build a class of 4-4-0 locomotives with oscillating cam valves driven by the Holcroft gear - fortunately behind the cylinders where is sat for the piston valve versions).
The D&RG Mountains had the same valve gear as 60000, which while unsold is hardly new. It seems to have its original paint under layers of lacquer, and the feet of generations of schoolchildren have worn away the footplate decking. Whether or not Baldwin benefited from the donation to the Franklin Institute we must be grateful that it is there.
"Gallop apace, you fiery-footed steeds"
The Bard of the High Iron does it again!
Thanks again for a most enjoyable post, Doctor D!
At ALCO this began in 1922 with the development of the MOHAWK 4-8-2 rebuilds of New York Central NYC 2568 and NYC 2569. Going from a 2 cylinder to a three cylinder design in the rebuilding increased the hauled tonage and tractive effort capacity on these locomotives by 25%.
Following this in 1924 Lehigh Valley Railroad contracted an order for 6 engines, the first "New Built" of the ALCO THREE CYLINDER POWER with their 5000 series 4-8-2 "Mountain Type" designs.
ALCO Locomotive Works continued the new concept in developing the THREE CYLINDER 4-6-2 "Pacific Type" and 2-8-2 "Mikado Types." Increase of tractive effort without total increase of axle loading was featured as well as the use of "Gresley" valve gear for the inside cylinder motion. Purchasers were Missouri Pacific RR, Louisville and Nashville RR, Delaware, Lackawanna and Western RR, Wabash RR, and the New Haven RR. Mexico with light rail loadings and Brazil's Sorocabana RR narrow gauge were also customers for THREE CYLINDER POWER.
Delaware Lackawana and Western used its first 4-8-2 three cylinder locomotives for heavy passenger work. This was followed by a second series of 25 three cylinder engines numbered DL&W 2201-2225 which were used for heavy freight, and which order was followed by an even larger class of three cylinder designed engine series of the DL&W 2226-2235 class.
Switch engines also came in for "three cylinder" design. New Haven Railroad tried the 0-8-0. Also of note was a large 0-8-0 transfer engine designed for Belt Railway of Chicago.
------------------------
THREE CYLINDER POWER then took off in the form of "really, really big power" design projects. Because there was a point beyond which regular two cylinder locomotives could not be built. This was the "really, really big power" possible with the 4-10-2 "Overland Type" wheel arrangement developed in 1925 for the Union Pacific but made more famous in 1926 with a demonstrator model constructed by Baldwin Locomotive Works.
The 4-10-2 "Overland Type" continued to be built only as THREE CYLINDER power, albeit without water tube firebox, high pressure boiler and compound cylinders of the Baldwin 60000 demonstrator.
Strangely the THREE CYLINDER POWER 4-10-2 "Overland Type" also came to be called the 4-10-2 "Southern Pacific Type" owing to the 1925 building of 49 engines of the 5000 class for Southern Pacific Railroad. Union Pacific RR who first ordered 10 of the 4-10-2 three cylinder locomotives were glad to keep the classification of them as originally given as the 4-10-2 "Overland Type."
Further, not to be so outdone by Southern Pacific RR and its 49 copies of "Southern Pacific Type" so renamed, the Union Pacific RR then undertook the development of its own named THREE CYLINDER POWER design using the "one more step format" of the 4-12-2 "Union Pacific Type" 9000 series engines of "really, really, really big power" locomotive. Eighty eight of these massive three cylinder locomotives were constructed which had an operational speed of 60 miles per hour and a very long service life. Some passenger usage of this astounding 4-12-2 three cylinder locomotive was reported.
It would appear that ALCO and its THREE CYLINDER POWER concept needs to take its place historically along side of LIMA and its SUPER POWER concept!
Doc
M636C Gresley called Holcroft to his office in London to describe the conjugated gear.
I thought this was much later than Holcroft's invention of conjugated gear for four-cylinder engines, and concerned something particularly germane to the present thread: how to make the 2:1 lever three-cylinder gear work correctly with an inclined center cylinder.
Holcroft's gear worked entirely behind the cylinder block so all the valves and valve rods expanded in the expected direction. However if I recall correctly when there was not sufficient space behind the cylinder block, it would all be installed in front of the cylinder block, leading to some interesting rod arrangements to drive it.
There has never been any conjugated gear for four cylinder locomotives: all that is required is a 1:1 rocking lever to drive the adjacent valve. This was widely used by the GWR, for whom Holcroft worked during the early years of the 20th Century. Holcroft was at least partly responsible for the detail design of Churchward's standard locomotives, which drew heavily on contemporary designs from the ALCO Brooks works. The GWR four cylinder locomotives followed Du Bosquet De Glehn compounds but remained simple expansion (and used inside valve gear rather than outside). While with the GWR, Holcroft designed his original conjugated gear which used two 2:1 rocking arms and a central oscillating beam with the drive to the centre valve in the centre of the oscillating beam. This was intended for a 4-4-0 much like Churchward's County which was known for rough riding and heavy hammer blow, and Holcroft expected that a 3 cylinder would run more smoothly and be easier on the track. The problem with that gear was that all the cylinders had to be parallel and in line fore and aft and the cente valve had to be above the cylinder.
Gresley produced his first O2 2-8-0 in May 1918 using a mechanism that might have infringed the existing Henschel patent (possibly not a big problem in early 1918) but which didn't work very well. Holcroft had continued to work on alternative conjugating gears and came up with the "Gresley" arrangement about this time. Gresley had been advised of Holcroft's earlier patent which had expired in 1913 and asked him for advice sometime in mid 1918. The new design was prepared in time for the K3 class 2-8-0 in 1920. I think Holcroft was the one who realised that a horizontal valve could feed an angled cylinder but that that valve could not be central, and that realisation led to the design with the single large 2:1 lever. Two designs on the LNER used Holcroft's later design behind the cylinders, the D49 4-4-0 and a rebuild of an NER 4-6-0, the B16/3 and these gave pretty much trouble free service, which I attribute to the elimination of the valve rod expansion problem and the resulting uncertainty in valve events.
The two Maunsell locomotives used the Gresley arrangement with outside links replacing the valve rod drive, presumably to Holcroft's own design, other wise identical to the "Gresley" design. These two were soon rebuilt to use three sets of Walschaerts gear, and a dozen or so similar locomotives were built that way. Holcroft presumably lost interest in the gear, as did nearly everybody else.
The initial response to the A4 middle big end was to insert a glass capsule of a liquid which released a pungent odour on overheating so that the crew were aware of the problem and could ease off before serious damage occurred. I believe different designs were used into post BR days when more accurate optical measuring techniques were introduced and it was found that by mor accurately aligning the whole chassis in conjunction with a heavier inside big end effectively removed the overheating.
M636CGresley called Holcroft to his office in London to describe the conjugated gear.
I have always thought that part of the differential-expansion issue on Gresley-gear-equipped engines is that actual superheat under some running conditions, particularly WOT at high road speed, was much higher than expected, a likely effect being greater expansion of the valve relative to the port areas in the cylinder block.
Peripherally, but on topic: Can someone post (or re-post) drawings and details of the original 'defective' big-end arrangement on the A4s, and how the design was practically improved after Mallard's 'experiences'? I have read about this (I think in the 'three designers' book) but there were no illustrations of the specific design details.
Wizlish One of the Australian classes got around the issue of shaft whip and bearing play by using a system of gears that turned cross-shafts in bearings. Unfortunately the shafts 'wound up' a bit, as torsion bars in Chrysler suspensions do, and this allegedly produced much the same timing problem as the version with the levers I think there is an actual difference between Holcroft's and Gresley's gear, although they use the same basic idea. Holcroft's arrangement was for a four-cylinder locomotive; Gresley's for three using the 2:1 lever we all know and loathe. For some reason I remember learning that Holcroft himself accepted this difference, but I may just have heard it during a lecture on valve gear minutiae.
One of the Australian classes got around the issue of shaft whip and bearing play by using a system of gears that turned cross-shafts in bearings. Unfortunately the shafts 'wound up' a bit, as torsion bars in Chrysler suspensions do, and this allegedly produced much the same timing problem as the version with the levers
I think there is an actual difference between Holcroft's and Gresley's gear, although they use the same basic idea. Holcroft's arrangement was for a four-cylinder locomotive; Gresley's for three using the 2:1 lever we all know and loathe. For some reason I remember learning that Holcroft himself accepted this difference, but I may just have heard it during a lecture on valve gear minutiae.
The Australian locomotive with rack and pinion valve gear was the D58. This class started off as an order for 25 further D57 class for wartime traffic in 1942. The cast steel beds arrived from GSI but it was decided that the new locomotive should meet tighter clearances requiring smaller cylinders. The whole design went downhill from there. The rack and pinion gear was intended to allow longer valve travel, since Gresley gear had caused problems with overtravel in the centre valve, and the new gear was intended to eliminate this. When the first locomotive was steamed for the first time expansion locked the gear solid, and the gear was removed from the frame and cantilevered off the cylinder casting. Only 13 of the 25 D58 class were ever assembled starting in 1949 and their place was taken by Beyer Garratts.
Gresley called Holcroft to his office in London to describe the conjugated gear. The meeting is described in Holcroft's autobiography. The layout used was favoured by Gresley, although Holcroft outlined the theoretical disadvantages at that initial meeting but Gresley felt that ease of access for maintenance outweighed the problems. Gresley wanted Holcroft to work for him on the GNR but R.E.L. Maunsell, CME of the SE&CR refused to release him.
Partly to appease Holcroft, Maunsell built two three cylinder locomotives with Holcroft's conjugating gear, a 2-6-0 class N1 and a 2-6-4T class K1, the latter later rebuilt as a 2-6-0 class U1. These had an arrangement similar to that used by Gresley but the conjugating gear was driven directly from the valve gears with outside links rather from the valve spindles as Gresley did.
I believe this was a significant point. The outside valves were located by the valve gears behind the cylinder and valve castings but the centre valve was located by the conjugating gear forward of the castings. As they heated the outside valves expanded forward while the inside valve expanded rearward. Since the valves had to be set when cold, only estimates could be made of the correct position at operating temperature.
When this is combined with wear and flexibility, it becomes clear why Gresley gear is inherently imprecise.
I examined the indicator diagrams of road trials on loco 5711 made during 1929 when the loco was new but run in. The indicator on the centre cylinder was inoperative for most tests but one card was made at 45 mph and the diagrams were very strange suggesting that that that speed the valve was mistimed and the power from that cylinder was about 30% greater than the outside cylinders.
Edward Thonpson lined up the centre cylinders of the A4 class to 17" from 18" to reduce this effect.
tomikawaTTThe main effect of the raised and angled center cylinder was a syncopated exhaust (Chuff ---- chuff --- chuff -- Chuff ---- chuff...) instead of the evenly-timed exhaust blasts of a two cylinder loco.
That syncopation is only the result of age, or incorrect valve setting. As built by anyone competent, the crank angles and valve actuation were arranged to give equal inlet and exhaust, and while the exhaust was triple it was regular.
(On Smith compounds and the like, as I mentioned, there was audible syncopation due to one out of every three exhaust events being internal, but that's not what you're talking about.) EDIT - I 'mentioned' this wrong. The exhausts on Baldwin 60000 were timed 90 degrees apart, so the audible 'beat' of the exhaust was as normal for a 2-cylinder DA engine...
Some of the audio recordings of three-cylinder engines, particularly a couple of the Nines near the end of their time, are almost terrifyingly mistimed. Here is a famous example.
The main effect of the raised and angled center cylinder was a syncopated exhaust (Chuff ---- chuff --- chuff -- Chuff ---- chuff...) instead of the evenly-timed exhaust blasts of a two cylinder loco.
Another interesting effect, seen on Alco-built JNR C52 class Pacifics, was 'out of position' counterweights on the main drivers. They had to balance the inside crank and rod end as well as the usual outside rod, so the counterweight was not opposite the visible crankpin.
Union Pacific had some Gresley conjugating mechanisms with plain bearings, and some later ones with needle bearings. When wear problems showed up (plain bearings much more so) some of the plain bearing mechanisms were fitted with needle bearings. Most had the conjugating mechanism replaced with a third Walschaerts valve gear.
Chuck
The Reading, early in the 20th century, built several 3 cylinder locomotives. They used Joy valve gear for the central piston valve, Walschearts on the outside cylinders. I believe there were 2 Atlantics and 2 ten wheelers, all Camelbacks, constructed for high-speed heavy passenger service. All were later converted to 2 cylinder power and enjoyed long careers.
M636CInterestingly, in his first use of the gear on GNR 2-6-0 1000, Gresley used needle roller bearings on the main bearing of the 2:1 lever, although plain bearings were used later.
Something to remember here is that the 'typical' kinds of rolling-element bearings are not good for oscillating motion that does not give at least one full revolution of the rolling elements start to finish. They wear funny. For very small motion, the result can be very like the damage older roller bearings could suffer via fretting when exposed to vibration or shock when not rolling and not oiled.
Now, a needle bearing isn't really a proper roller bearing, as there is no cage and often no full rolling contact on all the pressure surfaces. Think of a Multirol as a hydrodynamic bearing which has hardened steel for all its bearing surfaces, but which cannot gall or seize under load because there is always an open reservoir of lubricant between the hardened line contacts of the rolls and the inner and outer races. This should have been a good technology for the pivots in a conjugated valve gear like Holcroft or Gresley; I can't speak to the quality of the lube or the steel or the fabrication quality of the bearings that were used ... but it's possible that cost and perhaps NIH played some part. My understanding was that it turned out not to be play in the bearing joints that was the principal high-speed issue, but physical deflection of the lever under inertial loading (I have always wondered if there was a resonant component at a critical speed). Of course, once the bearings (particularly plain bearings) started to open up a little, the forces rapidly beat the bearings to where the clearances were excessive -- a little slop anywhere would rapidly force a little slop everywhere, and the slops add up in that gear.
Something my father suggested for using Timken bearings on conjugating levers was to put a pawl and a friction of some sort on the cage of each bearing, so it kept preferentially turning slightly in the same direction (with the rollers bearing properly but scuffing a bit against the edges od the cage slots). I do not know if this would work any better than grease lube in a Multirol for the limited time before the inside valve needed maintenance or attention.
S. Connor PS- Not all 3 cylinders used Gresley valve gear. Gresley gear wore quickly and was not very suitable for high speeds. Many 3 cylinders (If not simplified to 2 cylinder operation) were rebuilt with Walschaerts valve gear on the center cylinder, or had the Gresley rebuilt with roller bearings. Some of the UP 9000's had this done to them, but none survived save the 9000, which was not altered. Now I believe the only 3 cylinder with "special" (Not Gresley) valve gear for the 3rd cylinder is Baldwin 60000. Not sure what valve gear is used on that, though.
PS- Not all 3 cylinders used Gresley valve gear. Gresley gear wore quickly and was not very suitable for high speeds. Many 3 cylinders (If not simplified to 2 cylinder operation) were rebuilt with Walschaerts valve gear on the center cylinder, or had the Gresley rebuilt with roller bearings. Some of the UP 9000's had this done to them, but none survived save the 9000, which was not altered.
Now I believe the only 3 cylinder with "special" (Not Gresley) valve gear for the 3rd cylinder is Baldwin 60000. Not sure what valve gear is used on that, though.
S. Connorwhere did you find the tilt angle for the A&S #12?
S. Connor[quoting someone else] Another feature of the Gresley design was the general need to angle the third cylinder so that the inner main rod would clear the first driver axle.
Our community is FREE to join. To participate you must either login or register for an account.