Steam loco cylinders - more is better

[kevinlms]

9 minutes ago, melmerby said:

There's at least one class of locos where the first axle is cranked, although the drive is to the second axle, this was done to give clearance for the inside connecting rod(s).

Thereby creating a fine line between an experimental loco and one with modifications carried out, to (usually or hopefully!) improve the efficiency a type of loco.

[Artless Bodger]

There was a single German experimental express loco with 8 cylinders, but uncoupled wheels, so a 1'Do1', no 19 1001. It was built as a steam analogue to the 1'Do1' electric express locomotives of classes E18 and E19. Each driving wheel set was powered by a 2 cylinder V steam motor. It was not particularly succesful afaik but given it was completed in 1942 was probably not able to be tested in ideal conditions.

 

 

[rogerzilla]

3 gives smoother power delivery.  Its main problem is that you need an extra set of valve gear or a Gresley derived arrangement.  4 cylinder locos normally use 2 sets of valve gear with rocking levers, since the inside cylinders have the opposite valve events to the outside ones.

 

4 doesn't give smoother power delivery (in virtually all cases) because it gives the same power pulses as a 2-cylinder loco, although it reduces surging and hammer blow because the reciprocating masses mostly cancel out.

 

4 is better for large, fast and heavy locos because you can fit more cylinder volume in the loading gauge and it's easier on the track.

 

The BR Standards eschewed 4 cylinders because they used a high running plate to fit 2 large ones, and E S Cox reckoned the reciprocating mass issue was overstated, preferring not to balance half of it with wheel weights - as was normal practice but causes hammer blow - and instead tweaking the drawbar springing to reduce the surging effect.

[Jeremy Cumberland]

1 hour ago, rogerzilla said:

E S Cox reckoned the reciprocating mass issue was overstated, preferring not to balance half of it with wheel weights - as was normal practice but causes hammer blow - and instead tweaking the drawbar springing to reduce the surging effect.

One thing I have wondered is how much of the surging is due to reciprocating masses and how much is due to cyclical forces between wheel and rail.

 

What were his conclusions regarding drawbar springing, by the way? What needed to be done to reduce surging?

[ciderglider]

6 hours ago, rogerzilla said:

3 gives smoother power delivery.  Its main problem is that you need an extra set of valve gear or a Gresley derived arrangement.  4 cylinder locos normally use 2 sets of valve gear with rocking levers, since the inside cylinders have the opposite valve events to the outside ones.

 

4 doesn't give smoother power delivery (in virtually all cases) because it gives the same power pulses as a 2-cylinder loco, although it reduces surging and hammer blow because the reciprocating masses mostly cancel out.

 

4 is better for large, fast and heavy locos because you can fit more cylinder volume in the loading gauge and it's easier on the track.

 

The BR Standards eschewed 4 cylinders because they used a high running plate to fit 2 large ones, and E S Cox reckoned the reciprocating mass issue was overstated, preferring not to balance half of it with wheel weights - as was normal practice but causes hammer blow - and instead tweaking the drawbar springing to reduce the surging effect.

The USA has more generous loading gauges than we have in the UK or the rest of Europe. Union Pacific's FEF class 4-8-4 was a large express loco, with just two cylinders. So it seems to me that the main reason for going beyond two cylinders was staying within the loading gauge, rather than achieving smoothness. 

[LMS2968]

1 hour ago, ciderglider said:

The USA has more generous loading gauges than we have in the UK or the rest of Europe. Union Pacific's FEF class 4-8-4 was a large express loco, with just two cylinders. So it seems to me that the main reason for going beyond two cylinders was staying within the loading gauge, rather than achieving smoothness. 

C.J. Bowen Cook designed the Claughton class with four cylinders, all driving the leading coupled axle, specifically to eliminate the need for balance weights in the wheels as these would cause hammerblow. Since there wouldn't be a dynamic supplement to the static axle load, he intended to take that static axle load above what would normally be acceptable to the Chief Civil Engineer. Unfortunately, this was beyond the CCE's understanding and he rejected the design, which then had to be redrawn with lower weights, most particularly in the boiler. 

 

Perhaps C.J.B.C. was ahead of his time?

[Jeremy Cumberland]

6 hours ago, ciderglider said:

So it seems to me that the main reason for going beyond two cylinders was staying within the loading gauge, rather than achieving smoothness. 

I think so, but I don't know enough about early use of 3 or 4 cylinders.

 

By the early 20th century, it was clear that you could have 20" outside cylinders, giving 628 square inches of piston area. I can't think of any 3- or 4-cylinder locomotives with a smaller piston area than this (3x 16¼" or less or 4x 14⅛" or less), so it looks like 3- and 4- cylinders were only resorted to when 628 square inches of piston area was not enough.

 

The LNWR George the Fifth class later had 20½" outside cylinders (660 square inches). I don't know when they were enlarged from 20" Oops. Back to school with me. They were 20½" but were inside.

 

In 1934, the Gresley P2 class had 21" outside cylinders, for 693 square inches (bigger than several previous 3- and 4-cylinder locomotive classes), but I can't think of any other class with outside cylinders this big apart from the Midland Compounds. Midland Compounds had the connecting rods inside the coupling rods, allowing the outside cylinders to be several inches further inboard, so they don't really count. Both the P2 and the Compounds were three-cylinder locomotives, of course. I don't know of any 2-cylinder locomotive with 21" cylinders.

Edited July 18, 2023 by Jeremy Cumberland

[The Lurker]

The first batch of N15 had 22 inch cylinders the second batch 21 inch and subsequent batches 20 1/2. I seem to recall that most were rebuilt to the smallest diameter- and Hornby had to change the intended production of Red Knight to another engine because it was the one class member that never received the smaller diameter cylinders. 

this info is a combination of the Sremg page and Wikipedia.
 

I would guess the reason for the smaller cylinders was the loading gauge on the Central and Eastern divisions.

 

EDIT - the other Urie 4-6-0 classes also had 21 inch cylinders as built although the Maunsell versions also had smaller cylinders 

Edited July 18, 2023 by The Lurker

[LMS2968]

2 hours ago, Jeremy Cumberland said:

The LNWR George the Fifth class later had 20½" outside cylinders (660 square inches). I don't know when they were enlarged from 20".

Really? When did that happen? It's news to me. Some Prince of Wales class got outside Walsheart's valve gear, but the cylinders stayed inside.

2 hours ago, Jeremy Cumberland said:

II don't know of any 2-cylinder locomotive with 21" cylinders.

Try 245 Horwich Crabs!

[Jeremy Cumberland]

8 minutes ago, LMS2968 said:

Really? When did that happen? It's news to me. Some Prince of Wales class got outside Walsheart's valve gear, but the cylinders stayed inside.

 

I am a fool, aren't I? They were inside.

[LMS2968]

No, we all make mistakes. It's called being human.

[jim.snowdon]

On 18/07/2023 at 08:47, Jeremy Cumberland said:

One thing I have wondered is how much of the surging is due to reciprocating masses and how much is due to cyclical forces between wheel and rail.

 

What were his conclusions regarding drawbar springing, by the way? What needed to be done to reduce surging?

It was solved by adjusting the pre-load on the engine-tender drawbar springs. The effort was not to reduce the cause, but to reduce the degree of resonant coupling in the springing of both the drawbar and the tender to train coupling.

On 18/07/2023 at 07:29, rogerzilla said:

The BR Standards eschewed 4 cylinders because they used a high running plate to fit 2 large ones, and E S Cox reckoned the reciprocating mass issue was overstated, preferring not to balance half of it with wheel weights - as was normal practice but causes hammer blow - and instead tweaking the drawbar springing to reduce the surging effect.

The high running plate on the BR Standards was the result of a conscious decision to attach it and all the other fittings to the boiler, where they wouldn't shake themselves loose.

The decision to adopt only outside cylinders was largely for reasons of accessibility, following on from the work of Ivatt on the LMS, and bolstered by the lack of any noticeable performance difference between the 2- and 3-cylinder 2-6-4s designed during Stanier's tenure on the LMS.

On 18/07/2023 at 07:29, rogerzilla said:

4 cylinder locos normally use 2 sets of valve gear with rocking levers, since the inside cylinders have the opposite valve events to the outside ones.

In Britain and Ireland, the only 4-cylinder locomotives that used only two sets of valve gear were the Churchward 4-4-2/4-6-0/4-6-2 family, the Stanier Coronations and the Midland 0-10-0 Lickey Banker. Everything else used one set of valve gear for each cylinder, not that apart of the LMS Princess Royals there weren't many other 4-cylinder locomotives anyway.

[Compound2632]

15 hours ago, jim.snowdon said:

not that apart of the LMS Princess Royals there weren't many other 4-cylinder locomotives anyway.

 

How was the valve gear arranged on the Dreadnoughts (a class of 70) and Claughtons (a class of 130)? (A total roughly equal to the number of Castles and Kings.)

 

One should also perhaps mention the various classes of Webb and Robinson 4-cylinder compounds. Of the Webb engines, there were 80 4-4-0s of the Jubilee and Alfred the Great classes, 30 4-6-0s of the 1400 or Bill Bailey class, and 170 Class B 0-8-0s - that's a total of 280, which does exceed the total number of GWR 4-cylinder engines. So the LNWR had over 400 4-cylinder locomotives but not all at the same time.

 

EDIT: and not to forget the Lord Nelsons either, for all that there were only 16 of them.

Edited July 20, 2023 by Compound2632

[34theletterbetweenB&D]

On 18/07/2023 at 14:05, ciderglider said:

The USA has more generous loading gauges than we have in the UK or the rest of Europe. Union Pacific's FEF class 4-8-4 was a large express loco, with just two cylinders. So it seems to me that the main reason for going beyond two cylinders was staying within the loading gauge, rather than achieving smoothness. 

Now you are in the territory that Chapelon complained of, when comparing what was possible in North American compared to Western European steam traction.

 

That final generation of 4-8-4s with two cylinders cast integrally with the locomotive frame weighed in as loco and tender between 400 and 450 tons, so there's more mass damping of the piston thrusts. And they had to be started carefully, and could be, because the regulator ('Throttle') design in North American practise had necessarily become very refined to avoid violent slipping, especially at starting.

 

You can still ride behind the likes of the FEF and J class and observe that the starts are typically accomplished very slowly and smoothly, only needs a few hundred horsepower. The 5,000 hp and up power was for high speed on an all up train weight that could be 1500  tons. Once running the flywheel effect of the driving wheels adds to the mass damping.

[LMS2968]

9 hours ago, Compound2632 said:

How was the valve gear arranged on the Dreadnoughts (a class of 70) and Claughtons (a class of 130)? (A total roughly equal to the number of Castles and Kings.)

Both - the rebuilt dreadnoughts at least - used rocking levers from outside to inside; the Dreadnoughts as built had the valve gear between the frames driving the outside valves. The Claughtons' rocking lever was ahead of the cylinders; the rebuilt Dreadnoughts had it behind. The way this was arranged was sufficiently robust to be used for Stanier's Big Lizzies.

[rogerzilla]

Why was the LN 135 degree valve arrangement not repeated?  In theory it was a sound idea but it never seems to have given any advantage.  Was it actually flawed (apart from the obvious issue that you can't use only two sets of valve gear), or was any advantage just masked by the other deficiencies of the design, mainly that it was hard to fire*?

 

*which is ironic given that two very successful locos use at least some of the LN firebox design!

[Michael Hodgson]

Rather off topic, but when I worked for Perkins Engines I heard of an order (from Ireland) for one of our 4-cylinder diesel engines which was to be built with only 2 of the cylinders used for power.  The other two cylinders were to be used for pumping water.  Engineering told the customer it won't work.  Apparently they went ahead with the order anyway, so the factory were instructed to file off the company name or logo from any castings which showed such markings!  I never did hear whether or not the customer was happy.

[Artless Bodger]

On 20/07/2023 at 08:28, Compound2632 said:

 

EDIT: and not to forget the Lord Nelsons either, for all that there were only 16 of them.

The LSWR had a few 4 cylinder locomotives too. 

[Compound2632]

25 minutes ago, Artless Bodger said:

The LSWR had a few 4 cylinder locomotives too. 

 

Alas not Dugald Drummond's finest hour.

[rodent279]

4 hours ago, rogerzilla said:

Why was the LN 135 degree valve arrangement not repeated?  In theory it was a sound idea but it never seems to have given any advantage.  Was it actually flawed (apart from the obvious issue that you can't use only two sets of valve gear), or was any advantage just masked by the other deficiencies of the design, mainly that it was hard to fire*?

 

*which is ironic given that two very successful locos use at least some of the LN firebox design!

More expensive, because it requires a more complex crank axle?

[Artless Bodger]

20 hours ago, Compound2632 said:

 

Alas not Dugald Drummond's finest hour.

Proved that more is not always better.

[Artless Bodger]

On the question of sets of valve gear, would rotary cam poppet valves have solved the problem of needing separate sets according to crank angles and room to fit rocking levers etc? With just one drive shaft (or chain) to the cam shaft, presumably the cams could be set to permit any crank angle per cylinder? I may be wrong, wasn't one black 5 fitted with a single rotary valve gear drive take off on the centre of one coupled axle?

[LMS2968]

Apart from the final two, all the Caprotti Black fives had a single central drive shaft to drive the valves on both sides; 71000 had a shaft each side but one also drove the inside valves. Whether or not you could make a single shaft strong enough to drive three or four sets of valves is a debatable point. 

 

It took time to get this gear right and there were problems even with those same Black Fives in the late 1940s. How it might have behaved earlier is a moot point, although it seems to have worked well enough on the LNWR Claughtons in the 1930s.

[MarkC]

On 03/08/2023 at 19:42, Michael Hodgson said:

Rather off topic, but when I worked for Perkins Engines I heard of an order (from Ireland) for one of our 4-cylinder diesel engines which was to be built with only 2 of the cylinders used for power.  The other two cylinders were to be used for pumping water.  Engineering told the customer it won't work.  Apparently they went ahead with the order anyway, so the factory were instructed to file off the company name or logo from any castings which showed such markings!  I never did hear whether or not the customer was happy.

Well, Lister produced a diesel powered air compressor by basically replacing one cylinder head of their air cooled two cylinder diesel engines with one fitted with compressor valves, & removing/blanking off the running gear - push rods etc. It worked just fine.

 

Mark

[johnofwessex]

13 minutes ago, MarkC said:

Well, Lister produced a diesel powered air compressor by basically replacing one cylinder head of their air cooled two cylinder diesel engines with one fitted with compressor valves, & removing/blanking off the running gear - push rods etc. It worked just fine.

 

Mark

 

I believe - and Mark you should know, it was fairly common on some diesels to be able to use one cylinder as a compressor for starting air