Cylinder Size/Draughting (ex Braunton thread)

This is something I don't understand very well. I've seen various references to locomotive designs being over cylindered, and unsatisfactory until the cylinders were sleeved down smaller or otherwise modified. But I haven't seen any such references in the case of the Churchward standard 2 cylinder locos, even though the 18" standard front end was used with boilers varying from the Std2 with 20.4 sq ft grate right up to the Std 1 with 27 sqft grate. Its well documented that, for instance, the 4300s with the std 4 boiler and a 20.6sqft grate could run short of steam if asked to run fast with heavy loads (the reason for the conversions to Granges) but it doesn't seem to have been regarded as an issue requiring reduced cylinder size.

Maybe semi relevant is another interesting quote from Durrant on GWR and Bulleids elsewhere in his book. Durrant was a major steam enthusiast: he spent his apprenticeship holidays touring round Britain and later the continent seeing everything he could... Anyway

 

On its own, the size of the cylinder is largely irrelevant. It is the amount of steam that is used and that is essentially a product of cylinder size x strokes/time period. So, small cylinders coupled to small wheels can be the equivalent of large cylinders coupled to large wheels. The boiler will produce so much steam and, if the demand from the cylinders exceeds this, down will go the pressure and water level.
It's not quite as simple as that, but it forms the basics, everything else being equal.

 

Steve's explanation covers it all really, but I'm going to risk expanding on it. If the GWR had sleeved down the cylinders on the 4300s, then the results would have been (1) less brute tractive effort at low speeds and, (2) they would have to run longer cut-off at speed to maintain performance, thus still running out of steam. I suppose it might have been slightly happier at (say) full regulator and 15% with smaller cylinders instead of 10% on large cylinders when running fast at the boilers limit of continuous output, but maximum loads and acceleration would have been compromised. All just a question of what the loco is designed for, and if the boiler simply won't deliver the required power then no amount of fiddling with cylinder sizes will help.

Maunsell reduced the cylinder bores on the most of the Urie N15 class, but my guess is that with the relatively short travel valves they didn't respond well to being notched up to match the boiler output. Smaller cylinders would allow them to be run at a cut-off the gear was better suited to. Mind you, he also revised the draughting fairly heavily. Bulleid's best results modifying Arthurs seem to have been with the Urie engines; 755 got valves with double porting on the exhaust side in addition to a Lemaitre, and was very well regarded. I'd love to know more about those modifications, but essentially they'd have compensated for the poor exhaust events of short travel valves by freeing up the exhaust for the duration of the valve opening.

Since Bulleid keeps coming up on this subject, I wonder if perhaps he learned his approach more from Ivatt than Gresley or Chapelon. Ivatt found his original Atlantics were having to be eased to avoid over-extending the boilers at speed, so put a much bigger boiler on basically the same running gear (with less tractive effort than the final Stirling singles), thus giving the crews no excuse for shortness of steam. Bulleid would no doubt have observed this in his GN days.

 

Maybe Tom was referring to 4 cylinder GWR types, the obvious comparison is a Saint with a Star, the latter having the extra demand of another pair of cylinders, but still fed by the number one boiler of the former.

Then again, short sharp burst of high power was desirable in certain cases like the short but steep South Devon Banks.

 

Yep, I understand the basics of steam demand, as Steve says, but I don't understand why, for instance the GWR considered it was useful and important to reduce the cylinder diameter on the Dean singles and some of the pannier tank classes, but not on the standard outside 2 cylinder locos. Surely that means that the size of the cylinder was extremely relevant for some reason, otherwise why did they do it? Was it something to do with the valve gear characteristics perhaps?

 

The reduced cylinder diameter on the Maunsell N15s was to compensate for a rise in boiler pressure from 180psi to 200psi. The valve gear was also changed to give longer travel valves and the cylinder design changed to streamline the steam ports - so there is more difference between the Maunsell and Urie 4-6-0s than is at first apparent just looking at raw dimensions. In Holcroft's words, the locos were "Ashfordised" as far as possible while still allowing the design to be produced very quickly as a "stop-gap" pending the longer lead time needed to get the Lord Nelsons into production.

Tom

 

Tom, I was referring to the modifications under Maunsell to get the performance of the LSWR-built Urie N15s up to acceptable standards: cylinders mostly reduced from 22" to 21" when renewed, small diameter stovepipes replaced with King Arthur draughting (derived from the H15 setup, I think), and enlarged steam ports with increased exhaust clearance (or negative exhaust lap, if you prefer), while using the original valve gear. All of these are just ways of improving the exhaust with suboptimal valve gear, since the N15s as built would not steam well when notched up at speed. There's nothing wrong with the basic Urie 4-6-0 boiler, as 777 and various S15s demonstrate to this day.

As you say, with the engines built under Maunsell the changes went further: higher pressure, 20.5" cylinders, improved valve events. The Urie volume of Bradley's Locomotives of the LSWR is good on the subject.

Jim, I don't really know anything about the GWR classes you mentioned, but there are a few potential reasons for reducing the cylinder sizes. (1) Too much tractive effort for available adhesion (Dean singles, perhaps). (2) Cylinders too big for valves and gear, so poor breathing, as discussed for the N15s above. This would also apply if the cut-off to avoid draining the boiler in the normal intended operating conditions was simply too short to be effective, particularly with low superheat or saturated engines If it can't stand more than full regulator and 5% (to take it to extremes) at the intended operating speed, the cylinders may be a bit large for the boiler. (3) Excessive piston thrust damaging frame structure. I seem to remember the rebuilt Patriots, with the excellent rebuilt Royal Scot boiler, had cylinders 1" smaller than the Scots to avoid this problem.

The standard GWR 18ishx30 cylinders and their valve gear were pretty good for their time, so getting steam in and out wouldn't have been a particular problem. The GWR seems to have used wheel diameter to tailor the intended speed and tractive effort, so the only remaining question is whether the boiler can supply enough steam to make the desired power. Evidently the 4300 chassis could use more steam at speed than a No4 boiler could make, but if the requirement was for more power then putting similar running gear under a No1 boiler was the correct answer, not reducing the cylinders.

First find the maximum sustained power output a loco needs to produce. That dictates your boiler, since it needs to make enough steam. Size the cylinders and drivers to have a sensible range of rpm and cutoff in your desired operating envelope. Then make sure you have can get enough weight on the driving wheels to be able to generate the required drawbar pull without too much slipping. This is probably where all the loco designer's problems really begin...

 

To get the best out of a locomotive you really do need to install the very best exhaust system in order to extract the maximum power from the boiler. As time has gone on the need for a wide variation in driving wheel diameter has receded. Better valve and valve gear design coupled with a more enlightened approach to cylinder design has allowed higher rotational speeds to be more readily achieved. Balancing techniques for both rotating and reciprocating components again allowed for improvements in this and in other areas. Some fairly unrefined techniques were used to try and improve engines in the past. They reflected the knowledge base of the time.
Designing a locomotive is not easy. Not if you want to make a really good one. Nearly all engines can "beat the boiler" the question is how much power can you obtain before this starts to happen.

 

Ah, this is starting to make some sense to me... Most of the incidences of "too large cylinders" I've come across are older locomotives, pre the Churchward revolution of long travel valve gear and generous sized piston valves. So it seems reasonable that is was possible to have cylinders that couldn't reasonably be filled with steam in the time available, even though it was available in the boiler.
With better valve gear available the cylinders could always be filled if steam was available, so the only penalty of having larger cylinders was in weight and internal resistance, so it made sense to have big cylinders to accelerate the train away - assuming of course the regulator design allowed the driver to control the power and avoid excessive slipping - and standard sized cylinders and valve gear to reduce the spares requirement and increase interchangeability. Wouldn't stop you running short of steam if the task was beyond the locomotive, but did mean that power was there for starts and hills when you needed it provided the boiler could be allowed to catch up again.

 

It's hardly surprising that some of Dean's engines were considered over cylindered. The singles as built had a cylinder 'capacity' nearly as much as Churchward's original 2 cylinder 4-6-0s while the Armstrong class 4-4-0s with the same boiler were comparable with the Stars as built.

It is notable that as experience was gained, and superheaters introduced, the cylinder diameter of Churchward's larger types was gradually increased and that the Castles famously successfully incorporated a large increase in cylinder diameter without a comparable increase in the boiler size.