2 Cylinders V 3 Cylinders

I think some of you have mis-quoted Holcroft re the Schools. The problem with starting from London Bridge was due to the piston valve design and leakage from the then solid piston valve head in advance of the packing ring.This was corrected and no further problems occured. It is detailed in some depth in Vol 2 'Locomotive Adventure' by Holcroft.

Holcroft was always of the opinion that a 3 cylinder loco provided a more even turning moment/torque when starting, compared to a 2 cylinder loco. Churchward agreed, but took the view that his standardisation programme was more important. Maunsell allowed Holcroft to exploit his ideas on the SECR then SR.

The SR Z class was a very clever design. Although in print Holcroft never claimed any credit, the 3 cylinder design and novel valve gear on the inside cylinder is pure Holcroft. Again see 'Locomotive Adventure' Vol 2 for an assessment.

My personal view is that Holcroft should have succeeded Maunsel as CME. Why he did not I do not know. Perhaps a topic for another thread.

Cheers,
Julian

 

As a matter of interest do you know why it is customary to refer to 135degrees, rather than 45degrees, in relation to the LNs? Is this the relationship between the right hand outside motion and the left hand outside motion? (I read somewhere the right hand leads on the LNs).

 

Yes, it's the relative angle between left and right hand crank pins.

Tom

 

Thanks for that. So it looks like the peak torque on starting is less than in the 90deg arrangement, but broadly comparable to a hypothetical three cylinder equivalent; but also the variation between max and min torque over the first revolution is even less (which is presumably a good thing).

Did anyone else try a four cylinder loco with equivalent crank settings?

Tom

 

A CME was primarily an administrator though. I'm not sure Holcroft would have been particularly happy or well suited in that role. Trouble is do we know enough about second echelon people who didn't get CME jobs but perhaps should have done?

Incidentally, for starting trains, forgive me if I'm wrong, because this is way out of my areas of knowledge, but it seems to me that its possible that where torque is variable wouldn't the minimum figure on the graph be the most significant for starting the train, and the maximum most sigificant for incipient slip?

 

I'm not sure that's correct. If you look between the frames on Lord Nelson, I seem to remember there's the surprising sight of a conventional 90 degree crank axle. I think the arrangement is that the outside cylinders are at 135 degrees to their neighbouring inside cylinders, therefore putting the outside crankpins at 90 (or 270) degrees to each other. The advantage of doing it this way is that the coupling rods are at the most advantageous angles (when one is doing nothing the other is at the max torque position, thus minimising the load) and switching from this arrangement to a more conventional one requires less modification. I think the reciprocating balance may be more acceptable too, but I haven't done the workings yet.

 

Mea Culpa... Reading the descriptions in Nock and Bradley it is somewhat ambiguous, but I found a GA drawing and it appears as you say, as well as I can tell.

Tom

 

Another aspect of two versus multiple cylinders, that might be trivia for some, could be worth mentioning.

The cyclic heat loss to the cylinder walls (internal cylinder condensation) causes higher losses the smaller the cylinders are. Thus 3 cylinders are less efficient than 2 (keeping the total volume constant).
According to Dr. Adolph Giesl’s book Lokomotiv-Athleten the relative cylinder efficiency varies approximately as the 9:th root of the cylinder volume. Thus a three cylinder loco has a 4.7% higher steam consumption than a corresponding two cylinder at the same output. A four cylinder is 8% less efficient.
Of course, this effect might partly be offset by more efficient valves and smaller wheels with higher rpm for a multiple cylinder engine.

This volume effect is most pronounced for small narrow gauge engines; if you scale down a mainline engine to 18” gauge the cylinder efficiency will drop 25-30%. I wonder if this has been confirmed for the RHDR locomotives?
And the steam consumption is about doubled for a 7 ¼" gauge loco.

 

I didn't know it was a 9th root relation, although I was aware of the general principle. Presumably cylinder bore/stroke proportions also have an effect. I'd cite the Bulleid light pacifics as the classic example in Britain of the high power possible from multiple small cylinders and large piston valves - I think as big as any other class that ran in the UK. I'd maybe not cite them as paragons of efficiency though.

In a locomotive context it's possible that multiple small cylinders could be better thermally insulated than two large cylinders where there's barely any clearance between the cylinder casting wall and the structure gauge, although I don't think this was exploited very often. Compounds also have an advantage because the reduced pressure ratio across each cylinder means less cyclical variation in steam temperature, and of necessity their cylinders tend to be rather large, except the HP cylinders on four cylinder examples.

 

The 9th root is my own fit to a published curve, but its looks as it was used to create the curve. It gives good predictions for e.g. 20 hp narrow gauge locomotives. But I would like to know if it works also for live steam miniatures.

The external insulation of the cylinders have only a very minor effect as the losses only contribute with a few percent of the total losses.

What I am talking about are the internal losses, where the hot steam is rapidly cooled against the cylinder walls and the piston. When the expanded and cool steam leaves the cylinder it is it steals back the heat from the walls. The effect is fully cyclic and affects the walls only millimetre deep. But the result is a very drastic heat loss, and from the outside it looks as the cylinder is much bigger than it is in reality, more steam goes into and comes out of the cylinder than corresponds to its volume. In the order of 20-30% of the steam is kept from doing any useful work in the cylinder this way under ideal conditions. Much more at low speeds, for small cylinders and for saturated steam.

Both superheating and compounding address this problem, the superheat makes the steam dry and isolating and the compound keeps the temperature differences down. But the only really efficient way to avoid the problem would be to have a perfectly insulating material inside the cylinder. A composite piston and cylinder heads together with a ceramic cylinder liner would increase the efficiency quite a lot.