British Railways Class 8P Pacific (4-6-2) "Duke of Gloucester"

British Railways produced a large number of BR Standard classes of steam locomotives, but it never produced a class of large express Pacifics (4-6-2s) - the designs of the Big Four post-Grouping companies continued in service until dieselisation in the mid 1960s. A class of medium, 2-cylinder Pacifics was produced, the Britannia Class 7P, as well as a class of light Pacifics, the Clan Class 6P.

There was one, solitary exception to this; a single experimental locomotive was produced, of Class 8P, designed, like the other standard classes, by Robert A. Riddles. The justification was the destruction of the experimental LMS Pacific "Princess Anne", the former Turbomotive, in the Harrow and Wealdstone disaster in 1952; Riddles gained permission to build an experimental three-cylinder heavy Pacific to replace it. The intention was for this locomotive to be the prototype for a new generation of large express locomotives to replace the pre-Nationalisation locomotives until the planned electrification of the main lines.

Much that was new was tried in this locomotive, the most obvious innovation being in the locomotive's valve gear; British Caprotti valve gear was used. Unlike normal steam locomotive valve gears, that operate through pivoting and sliding links, the Caprotti gear works along similar principles to an automobile engine's valves; a rotating shaft (in the locomotive case, driven by gears on the driving axle and through a long, slanting shaft to the valve gear atop the cylinders) with shaped cams on it that depress poppet valves at the appropriate points in the cycle. Unlike most automobile valve gear, the Caprotti gear allows for completely variable valve timing of both inlet and exhaust valves. The common Walschaerts valve gear and others allow this to a degree to achieve variable cutoff, but the Caprotti gear allows inlet cutoff to be reduced to six percent or less without affecting the exhaust timing; on the Walschearts gear, the exhaust and inlet timings are linked. This allowed much greater power and efficiency.

Early Problems

The locomotive emerged from Crewe Works in 1954 and was numbered 71000 and named Duke of Gloucester. Soon after construction it was taken to the Swindon roller test plant - a dynamometer for locomotives - and was found to be capable of great economy and efficiency.

Out on the road, however, things worked out not as happily. The boiler, which theoretically should have been as good as the best practice on other large Pacifics, proved deficient and steamed poorly, especially at high outputs, while the locomotive proved difficult to fire well. It seemed to 'choke' when hard work was asked of it, and sometimes even had to be replaced by other locomotives. These mysterious problems were never fully investigated, since by the time they were discovered, the decision had been made to abandon steam locomotive development. The previous plan to continue using steam locomotives until electrification could be completed were abandoned in favor of dieselisation. The locomotive was deemed a failure, and after only eight years of occasional use, was withdrawn from service in 1962.

As the last express passenger locomotive built in Britain, serious consideration was given to preserving it in the National Collection, but in the end, its unsuccessful running sank this plan. The outside cylinders and their sophisticated valve gear were removed, and one was placed, sectioned, in the Science Museum in South Kensington (London). The rest of the locomotive, stripped of its valuable non-ferrous metal parts, was sent for scrap.

Fortunately, the scrapyard it was sent to was Dai Woodham's Barry Scrapyard, in Barry, South Wales; Woodham's bought locomotives in haste but scrapped them at a pace much more leisurely, especially since he found scrapping obsolete freight wagons to be a much more profitable and less labor-intensive process. The locomotives sat at the far end of Dai Woodham's yard for years while the easy profits from wagon scrapping were worked at first. So great, in fact, was the flow of obsolete wagons to scrap that Dai's scrapyard workers never got around to scrapping more than 200 locomotives, and by the time railway enthusiasts had realised what treasures lay there for the purchase.

Among the more than 200 remaining there was Duke of Gloucester, by that time a forlorn sight, rusted, missing major components including both outside cylinders, and seemingly an impossible task to restore to even display condition, let alone running. A small group of volunteers decided to rise to the challenge, however, and despite derision from some circles, and indeed accusations that resources would be better spent on more 'hopeful' locomotives, they decided to buy the Duke and attempt to restore it. The locomotive left Barry Scrapyard on the 24th April 1974 (by truck) for a restoration that would take twelve years.


In the 1970s, British manufacturing industry was in a much healthier state than more recently, and the restoration was assisted by much commercial sponsorship. It was decided that since there was clearly something wrong with the Duke in its previous existence, that attempts should be made to discover what caused the boiler's mysterious inabilities to keep up steam. Problems were indeed found.

A new ashpan had to be constructed, as the old had rusted through, and it was found that a mistake had been made when the original was constructed; the damper door air spaces, which allowed air up through the firebed, were much smaller than the original construction drawings specified. That in itself would have restricted the combustion, since not enough air would have reached the fire. It was decided, though, while this was significant, it couldn't have been the only problem.

That problem was found when comparing the Duke's proportions to those of two similarly sized designs; Bullied's Merchant Navy class, and the Peppercorn A1. Most dimensions were comparable except for one critical area; the locomotive's exhaust. Both the blastpipe (US: nozzle) area, and the chimney (US: stack) area were significantly smaller than the other locomotives. The blastpipe area was about 20% smaller, while the chimney area was only two-thirds that of the Peppercorn A1 and half that of the Merchant Navy.

Could such a restriction in area have 'choked' the locomotive at high outputs? It was certainly quite likely. If the blastpipe and chimney were undersized for the amount of steam being ejected and the required flow of hot gases through the boiler tubes and flues at times when the locomotive was working hard, then firstly the back pressure, the pressure against which the exhaust steam had to work in order to be ejected, would have been much higher than optimum. Since the effort steam can exert in the cylinders is the different between the inlet pressure and the exhaust pressure, this obviously severely restricted the power of the locomotive -- like if a car's tailpipe was partially blocked. Furthermore, this restriction also limited the draught through the fire and tubes.

Tom Daniels, the chief engineer for Associated Locomotive Systems who designed the British Caprotti valve gear, informed the volunteers that he had recommended a Kylchap exhaust, like that which was fitted to the Peppercorn A1, but his suggestion had been overruled. It was decided to - this time - follow his recommendation, and using the Peppercorn A1 design as a template, a complete Kylchap exhaust system was produced and fitted, with a substantially enlarged double chimney arrangement.


In 1986 restoration of the Duke was complete and it was steamed again for the first time. The modifications made proved totally successful, the locomotive finally proving that the vision of its designers was a true one, just having been let down by detail design of the front end and poor manufacture. It proved easy to fire, even with less experienced crew members shovelling; stupendously powerful and speedy, and overall possibly the most impressive steam express locomotive ever to have run on British metals.

Further modifications in restored service included electric lighting, which was pretty much never fitted to British locomotives before dieselisation, using a turbogenerator providing 24 volt AC to modern headlamps and sidelamps, fitted inside the casings of oil lamps to present an authentic appearance. These allow the locomotive to be certified to run at 90mph. A support coach contains an additional generator in case of problems, as well as an intercom with the footplate crew. The crew also have a standard Railtrack radio-telephone so they can converse with signalling staff.

Overhaul, and further improvements

Following restoration, the Duke of Gloucester carried out admirable work for a number of years, but in the mid 1990s the locomotive became due for a major overhaul. A substantial amount of money was obtained from the Heritage Lottery Fund to enable both the overhaul and an ambitious plan of further improvements to be carried out. Among these include increasing the tender water space, fabricating a new, steam powered coal pusher to replicate the one originally equipped, the building from scratch of a new Davies and Metcalfe exhaust steam injector of adequate size (the original having been melted down, and borrowed ones of insufficient size used after restoration), various tunings of the valve gear, and finally air brake equipment to allow the Duke to haul air braked stock.

This restoration is almost complete, and the Duke is expected to re-enter service sometime in the Spring of 2003. The overhauled locomotive has already steamed successfully once more. No doubt more good things are to come.

Sources: and as well as books in my collection.