The Caprotti Poppet Valve Gear
The Caprotti poppet valve gear for steam engines was invented by an Italian, Arturo Caprotti, and patented in 1919. Caprotti was an engineer with experience of automobile engineering, and he sought to apply and adapt the principles of the valves commonly used on the internal combustion engine to the steam engine.
The flaws Caprotti found in the typical steam locomotive valve gear included:
- Because the steam engine is double acting, the same valve ports were used for both inlet and exhaust, even though steam expands in the cylinder and thus has a much greater exhaust volume.
- Since the same piston valve is used for both inlet and exhaust, as the cutoff is reduced and the inlet valve is closed earlier and earlier in the piston stroke, for expansive working and using less steam more efficiently, the exhaust valve timing is necessarily also altered, even though this is not desirable at all. To some extent this can be mitigated, but there will always be a point beyond which the exhaust valve starts being closed too early in the cycle, and the remaining exhaust steam will be under compression for the rest of the stroke. This means that the more the cutoff is reduced, the greater the engine will be 'choked' by working against compression in the exhausting side of the double acting cylinder. This gives a practical limit to how far the cutoff can be reduced in a conventional steam engine using slide or piston valves, even though the demands of efficiency would prefer otherwise.
- The only run-time adjustment permitted by a conventional valve gear is to adjust the length of travel of the valve piston, which is used to affect the amount of time the valves are open. No other adjustments are possible.
Caprotti's patent valve gear worked quite differently to common locomotive valve gears like the Walschaerts valve gear which used eccentric cranks, rods and levers to produce a backwards-and-forwards valve motion.
Instead, gearing mounted at the center of a driving wheel turned a diagonal, rotating shaft (fitted with universal joints to allow for the up-and-down motion of sprung wheels] that connected to the valve gearbox mounted atop the cylinder. Here the rotary motion of the shaft was translated using gearing and cams to actuate four valves - a seperate inlet and exhaust valve for each side of the cylinder.
In Caprotti's original design, the exhaust valve timing was fixed, always opening and closing at the same points in the cycle. The inlet valve timing was adjustable to allow for the valve to close sooner or later in the stroke, allowing different degrees of expansive working as demanded. Of course, as demanded by the steam locomotive, the valve gear was fully reversible to allow the locomotive to move backwards - many, but by no means all, marine and stationary engines only run in forward. The Caprotti gear also allowed the locomotive to be as efficient in reverse as in forward motion, while a conventional valve gear was tuned for efficiency only in forward motion.
Service Use in Europe
In 1921 the FS (Italian Railways) began to install Caprotti gear on some of its locomotives, and this was followed by experimental applications all over the world, though it would be fair to say that nowhere else in the world was the Caprotti gear quite as popular as in its inventor's native land. In addition, the gear was applied to numerous marine engines.
American Experiences
In the United States it was tried by several railroads, but with no conspicuous success. The first to try it was the B&O in 1927, on both a 2-8-0 freight locomotive and a 4-6-2 Pacific (No. 5320 "President Cleveland"). It was removed fairly quickly, as were similar attempts by the Reading, the Southern Pacific and Union Pacific.
The Pennsylvania Railroad tried Caprotti gear on one of its two experimental K5 Pacifics built in 1930. Arturo Caprotti himself was invited to see the locomotive in action and was impressed by the gigantic size of American railroad equipment compared to what he was used to in Europe. He is reputed to have said, "Your locomotives haul houses, not cars!" The K5 in question was eventually converted to Walschaerts valve gear, but the Pennsy definitely had a curiousity about poppet valves after that, and they tried numerous other designs, mostly on modified Pacifics of the K4s class; in particular they were impressed by the American-designed gears produced by the Franklin company, eventually purchasing a whole class of 52 locomotives, the T1, fitted with such gear.
Caprotti valves were also tried on a Santa Fe 4-8-4, again with little success; it's possible that in many of these American implementations of the gear, the valve sizes were not scaled up in proportion to the much increased size and power of the engine.
The bigger problem, and the greatest reason why the Caprotti gear was not adopted in the United States, is that its potential greater efficiency was bought at the cost of much greater complexity. The Walschaerts valve gear, for all its problems, was simple. All its components were large, and the vast majority of them were out in the open on the side of the locomotive, in plain view; easy to maintain, and if anything were to go wrong, it would be obvious. The Caprotti gear was made of a multitude of small parts that had to fit to tight tolerances, hidden away in a box atop the cylinder. This often ended up being a maintenance nightmare, and one in which problems were rarely discovered before total failure on the road.
British Use
The Caprotti valve gear became the most commonly used poppet valve gear used in Britain, though it never acheived widespread adoption. Among the first users was the London and North Western Railway, who fitted Caprotti gear to ten 4-6-0s of the Claughton class.
After the Second World War, an ex-GWR engineer named Tom Daniels took the Caprotti gear and made modifications and improvements to it, naming his improved version the British Caprotti valve gear.
Mr Daniels' main innovation was making the exhaust valves also adjustable, rather than the fixed exhaust
valve events of the original Caprotti design. British Caprotti valve gear was fitted to a number of British locomotives, including several
LMS Black Fives, thirty
British Railways Standard 5MT 4-6-0s, and the last express passenger locomotive built in Britain, the only
Standard 8P,
Duke of Gloucester.
In addition, it was fitted to a large number of export locomotives, for India in particular, from the North British Locomotive Company among others. These exported locomotives numbered in the hundreds.
Conclusions?
The Caprotti valve gear suffered, as did many improvements to the Stephenson locomotive, from the fundamental problem that the steam locomotive's primary advantage over other forms of traction was its simplicity and ruggedness. The introduction of precision machinery to the steam locomotive, especially in places where it could wear and break, always cause problems.
Maybe the biggest secret behind the triumph of the diesel locomotive over the steam locomotive was that, being a wholly new form of traction, it could succeed in changing the way the industry looked after its locomotives, change the whole railroad infrastructure. No improved steam locomotive could do that. On the railroads which DID introduce modern maintenance arrangements for their steam locomotives, much better performance and reliability figures were witnessed.
Today
In Britain, two preserved locomotives exist with Caprotti gear; Duke of Gloucester, and one of the Caprotti-fitted Standard 5MT locomotives. A number still exist in Italy, and some Indian locomotives fitted with Caprotti gear are also in preservation.
Thanks to the 71000 Duke of Gloucester site at http://www.dukeofgloucester.co.uk/caprotti.html and a number of other sites for small details.