Remontoire escapements GRAVITY (Denison double three-legged type), 1860
This type of escapement is in a class sometimes called 'remontoire escapements'. It should be noted that this definition is different from that of a remontoire. Those operate as constant force devices independent of the type of escapement they are powering. They contain their own auxiliary power supply in the form of weight or spring, and if the main weight is removed, the escapement will continue to be powered until the remontoire is ready to cycle. Typically 15 to 60 seconds. On the other hand, remontoire escapements operate as part and parcel of the escapement itself. So if the main weight is removed the escapement will cease to be powered on the next swing of the pendulum.
Remontoire escapements are distinguished from the prior escapements since the impulse is not given to the pendulum directly by energy from the main weight through the clockwork, but by some other small weight lifted up, or small spring bent, always through the same distance, by the clock train at every beat of the pendulum. Some also work on every other beat, leaving the pendulum free one half the time. This allows the pendulum to be largely detached from the rest of the clock. Many illustrious makers had tried to perfect the gravity escapement - Berthoud, Mudge, Cumming and Hardy. Bloxam had come close in 1853. All of these prior attempts suffered from various problems, chief amongst them the fact that the pallets had tended to bounce off the escapement locking surface; known as 'tripping'. Edward Denison (later Lord Grimthorp) perfected the gravity escapement in 1860 by eliminating the tripping problem. He did this through the connection of a fly, a.k.a. an air brake, directly to the escape arbor via a friction clutch. It allows the fan to advance slightly after the escapement engages the pallet. The inertia provided by the weight of the fly keeps the escapement seated against the pallet during locking; in essence acting as an 'energy sink'. This escapement provides a nearly detached pendulum from the rest of the clockwork and is particularly important in tower clocks where wind and weather can cause disruptions to the movement through the exterior hands. Another special feature of this escapement is that there is no sliding friction so it does not need oil on the escapement. Again, due to the environment in which tower clocks are found oil contamination is a problem; severe temperature changes can cause oil to thicken and thin beyond their normal intended characteristics. Because of these features afforded by the gravity escapement, the use of a train remontoire is rarely seen in conjunction with it. While there was some improvement in accuracy over the deadbeat escapement, its real virtues lie in its stability, the design’s ability to keep the pendulum largely detached from the rest of the train, and the lack of need for oil.
The negative characteristics are the fact that it is a 'power hog'. The escape wheel arbor rotates 1/6th of a revolution per pendulum swing compared with 1/30th for a standard recoil or deadbeat escapement - a five fold loss. This is even more severe in the case of a 4 legged escapement, (1/4th revolution) - a 7.5 fold loss. There is usually the need for an extra wheel in the train or a very large wheel ratio - say a very large great wheel. Both of these solutions involve extra manufacturing expense. The escapement itself is also more expensive due to the number of parts used as well as the more complicated fabrication processes needed in manufacture as compared to a standard deadbeat escape wheel.
Denison's clutched fly acts as a non-linear system automatically adjusting to the varying energy demands of and upon the movement thus shielding the pendulum from those influences. The table below shows the conjectured energy flow in 'Big Ben', in millijoules per (two second) beat.
Hands under heavy ice load Good Weather, no wind Large Wheels 100 100 Hands 200 45 Fly 10 160 Pallet stops 4 9 Unlocking 1 1 Pendulum 5 5 Total (from main weight) 320mJ 320mJ
In essence the fly acts as an 'energy sink'. Accuracy requires an escapement system that provides a constant impulse to the pendulum despite varying energy demands from the hands (a big factor in the case of tower clocks with ice buildup or wind) and to a lesser degree variations in the wheel train due to oil viscosity induced by temperature and environmental contaminants. When demands are high the fly slows down, however, as long as it moves though its' allotted 60 degrees (in the case of the Denison double three-legged type) within the time it takes the pendulum to make one swing the escapement is unaffected. When demands are low the fly moves quickly, partially dissipating the excess energy as an air brake (i.e. heating the air). But because the fly is attached to the escape arbor through a friction clutch any additional excess energy that would be dissipated by slamming into the pallet stops is instead lost through the sliding of the clutch (again, negligible heat). This last issue is what distinguishes the Denison gravity escapement from all earlier attempts to solve the tripping problem. The fly device is a non-linear system making it well suited to varying demands. The table neatly shows how a large change in demand from the external forces acting upon the movement is made into an even stream though the escapement as it reaches the pendulum. The Denison is indeed a remontoire and escapement in one device - one of a very few practical 'escapement remontoire‘ for clocks. (There are a number for watches).
Table from Antiquarian Horology, vol. 11, no. 6, Winter 1979. "The Fly in the Grimthorp Gravity Escapement", by Henry Wallman, pg. 629-631.
|
