The LNER Encyclopedia

The Liverpool & Manchester No.9 locomotive 'Planet' had a number of innovations that locomotives built a year earlier did not have: A multi-tubular boiler; a smokebox; & horizontal cylinders at the front of the locomotive. Brilliant improvements to locomotive design, which I assumed had been taken up by locomotive builders & operators henceforth. But not Timothy Hackworth, William Lister or the Stockton & Darlington Railway, who continued to build machines of antiquated design. S&DR No. 9 'Middlesbrough' and S&DR No. 25 'Derwent' were such locomotives. S&DR locomotive 'Locomotion No. 1' though, managed to raise enough steam for its boiler to explode, despite that boiler being of conventional design, and not multi-tubular. Perhaps those type of boilers weren't as inefficient as they've been made out to be.

When was the last non-multitubular locomotive boiler built, does anyone know?

Middlesbrough No.9: https://www.flickr.com/photos/60292162@ ... ed-public/

The working Lancashire boiler at Coldharbour Mill in Devon was manufactured by Galloways in 1910 – long after multi-tubular boilers had become a commonplace feature of railway locomotive design. This suggests to me that something about the multi-tubular boiler particularly suited it to locomotive practice but was thought unnecessary as a feature of factory steam generation.
My guess is that the twin flue Lancashire design was sufficiently economical to maintain its suitability for meeting a steady, protracted demand for steam whilst still being simple and cheap to build and maintain, compared with a multi-tubular boiler.
I also guess that that multi-tubular boiler was, by contrast, much better suited to handling a rapidly fluctuating steam demand such as that found in the railway environment. The multi-tubular design strikes me as being much better able to absorb the stresses of relatively rapid heating and cooling of the furnace than a twin-flue boiler. Getting a Lancashire boiler online involves warming through for a day before you get any pressure on, simply to avoid stressing of the boiler that a railway locomotive would take in its stride.
Horses for courses.

West_Stanley wrote:The Liverpool & Manchester No.9 locomotive 'Planet' had a number of innovations that locomotives built a year earlier did not have: A multi-tubular boiler; a smokebox; & horizontal cylinders at the front of the locomotive. Brilliant improvements to locomotive design, which I assumed had been taken up by locomotive builders & operators henceforth. But not Timothy Hackworth, William Lister or the Stockton & Darlington Railway, who continued to build machines of antiquated design. S&DR No. 9 'Middlesbrough' and S&DR No. 25 'Derwent' were such locomotives. S&DR locomotive 'Locomotion No. 1' though, managed to raise enough steam for its boiler to explode, despite that boiler being of conventional design, and not multi-tubular. Perhaps those type of boilers weren't as inefficient as they've been made out to be.

When was the last non-multitubular locomotive boiler built, does anyone know?

Middlesbrough No.9: https://www.flickr.com/photos/60292162@ ... ed-public/

Surely a multi-tubular boiler would be constructed of concentric rings?

The Lancashire boiler is, strictly, also a multi-tube boiler in that there are two circular furnace tubes running between the end plates; the type with a single furnace tube running end-to-end is the Cornish boiler.

It may perhaps be that mill boilers do not work at a particularly high pressure. To display my embarrassing ignorance in some areas of steam power; I visited the MOSI, Manchester when a lot of their engines were working. Very quietly. I assumed they were being driven by electric motors, but couldn't see how. It was gently explained to me that they were condensing engines and therefore the 'exhaust' was water, not steam. Oops.
All that is required of the boiler for these engines is to provide steam to fill the cylinder on the admission stroke, and this will be drawn in by the momentum of the flywheel; obviously steam has to be always available: the volume produced rather than the pressure is possibly the important factor, though of course there may be a relationship between the two. I don't know whether condensing engines were replace in mills and factories by ones that used expansion, though.

The bank of multiple flue tubes brought three major advantages at introduction: a large increase in boiler heating surface through very much thinner metal walls than the equivalent flue cross section in a single thicker walled tube, a resistive effect on the forced draught, and the tubes efficiently provide most of the staying between the firebox and smokebox tube plates. (Fortuitously, it would go on to provide a support structure for superheater elements.)

The resistive effect of the tube bank is most significant in providing a steam raiser which responds proportionally to variation in steam demand and thus the forced draught created by the exhaust through the blast pipe, with control over the risk of lifting the fire off the grate at high combustion rates and in the event of a slip with the regulator well opened; while allowing sufficent draught to keep the fire going when the in service loco is standing and only a small smokebox vacuum is provided by the blower. (Turbulence increases in the flue tubes as the gas velocity increases which means the resistance is variable, increasing as steam demand does, it's a very neat self limiting effect.)

As with so much on the Stephenson locomotive format, it was part of a synergistic whole. The proof is that for as long as steam locos operated as dominant railway traction, it remained the format from which all significantly successful development of the type commenced. The only 'deviation' which looked to have the potential to outright replace an element of the Stephenson locomotive format was the turbine: it came too late to maturity in a form compact and reliable enough for the locomotive application to displace the reciprocating engine in my opinion.

Hatfield Shed wrote:As with so much on the Stephenson locomotive format, it was part of a synergistic whole. The proof is that for as long as steam locos operated as dominant railway traction, it remained the format from which all significantly successful development of the type commenced. The only 'deviation' which looked to have the potential to outright replace an element of the Stephenson locomotive format was the turbine: it came too late to maturity in a form compact and reliable enough for the locomotive application to displace the reciprocating engine in my opinion.

The fundamental point to remember is that each different boiler type evolved to satisfy a particular set of conditions. Stationary factory boilers and marine boilers generally had to provide an substantial amount of steam at even pressure for days at a time, whereas the locomotive boiler was more akin to an internal combustion engine, being required to respond to acceleration, sustained work and idling on demand. However fire-tube boilers were limited in practical working pressure which is why experiments were undertaken with water-tube boilers which, incidentally, were also the source of power in Sentinel locomotives and lorries.

There are perhaps two other 'inventions' to note, one is of course the superheater - a means of overcoming limitations in boiler pressure and drying the steam, and the other the unique 'Paget' engine, which had a quite different arrangement of firebox and boiler, let alone the means of transmitting power to the axles.

You can add to the Paget the Bulleid Leader steam bogies, and numerous similar attempts outside the UK. But none ever 'made the break' and significantly replaced construction on the Stephenson layout.

The water tube boiler is an interesting failure. Small units such as used by Sentinels are adequate; but scaled up to a couple of thousand horsepower and more, not so much. Our home grown example on the LNER's 'Hush-hush' was a very careful piece of design work by Yarrow, with all their experience; removed from the locomotive it put in another near thirty years as a static steam supply at Darlington works, outlasting the loco frame in service. And it is the same story elsewhere, they were extensively tried in the USA. For all the care taken in design, this boiler format just doesn't work so well when rattled about all the working day on a loco frame. Static, or in the reasonably stable ship application, very successful indeed. Not synergistic with the rail locomotive application.

By contrast the multitubular locomotive boiler was regularly observed to perform better 'on the road' as opposed to static test whether as boiler alone, or in a complete locomotive on the rollers of a test plant.

Yes it was a quantum leap in locomotive design. As a locomotive boiler it was much better at raising steam quickly. Hackworth on his passenger locomotives did use multi tubular boilers as they suited the requirements better. In fact many of them were fitted with a multi tubular return flue to aid with steam raising. Cf Derwent. The Stockton and Darlington had a lot of slow moving goods traffic and frequent stops so the return flue boiler could have time to raise steam. With the tiny grate area and long flue they were found to be very economical when used on this type of work. They also cost significantly less to build and could be maintained with limited facilities so they kept building them up until around 1845. If it ain't broke...

JonR wrote:.... If it ain't broke...

....make it even better, like Andre Chapelon did.

An important difference between the locomotive boiler and industrial steam raising plant is their maintenance requirements. A locomotive can be stopped for tube cleaning along with other routine tasks at relatively frequent intervals, while its industrial counterpart has to keep going. The single flue type was much less prone to blockage and reduced performance than the multi-tube type, and its lower efficiency was probably an acceptable penalty in most industrial applications when fuel was cheap and before water tube types were fully developed.