The SeeSaw Gravity Escapement part 2
by Roger Bunce

THE DESCRIPTION that follows is of my proof-of-concept model. I emphasize that the methods and materials used are not necessarily those recommended for use in a 'real clock'. I chose the simplest, cheapest, quickest ways to explore the validity of the invention, without regard for optimization or longevity. Nevertheless, to give the escapement the best chance of working, ball bearings were used in 'important places'. I have used simplified drawings in this description for clarity.


Parts

Figures 3a – 8c show a SeeSaw Gravity Escapement (SGE) in which impulse occurs twice during each complete cycle of the pendulum. Figures 9 – 14 are photographs of the proof-of-concept model.

The components of the mechanism are mounted on and between front and back plates. The pendulum assembly is shown pivoting on knife-edges (in the actual concept model pointed screws were used for simplicity and adjustment). The escapement was tested with 1 second (0.994m long) and 0.68 second (0.462m long) pendulums (time of swing in one direction).  Extending each side of the pendulum are LH and RH arms. Sharing the same pivotal axis as the pendulum are LH and RH weight supports (Figures 4a – 5b, and Figure 12). The weight supports are provided with V-notches in which periodically rest the LH and RH weights (Figure 13). Each weight is provided with an inverted V-notch, and these locate on the pendulum arms. The flanks of the inverted Vees are slightly convex so the weights are self-aligning to allow for any slight misalignment in the aforementioned Vees.

As the pendulum oscillates, the weights are alternately lifted from the weight supports and then replaced, as described above (Energy interchange). The weight supports are fitted with counterweights so that the distal ends are biased gently downwards. Fitted to the underside of the weight supports are 'flat' cams. The LH and RH props act on the cams to raise or lower the weight supports (Figures 6a – 7b). Adjustable stops restrict the movement of the weight supports and the props.

The upper ends of the props terminate in rollers, which incorporate one-way ball bearing clutches (MF-CB-10, modelfixings.co.uk). The clutches are arranged so that as the upper ends of the props move towards the centre of the mechanism, the rollers are free to roll against the cams. Conversely, as the ends of the props move away from the centre, the rollers are prevented from rolling against the cams. Hence, the frictional resistance between rollers and cams is greatly increased in the outward direction.

Fitted to the props are locking arms. The ends of the locking arms terminate in ball bearings (SMR83ZZ EZ0, SMB Bearings Ltd), which act against the teeth of the escape wheel. Extending downwards from the props are reset levers. On the opposite side to the locking arms are adjustable counterweights. These lightly bias the LH and RH props anticlockwise and clockwise respectively.

Below the prop assemblies, is the escape wheel assembly (Figures 8a – c, and Figure 14). This comprises a rear escape wheel with three teeth, and a front escape wheel, also with three teeth. Interposed between the two escape wheels are three ball bearing resetting rollers (S604ZZ EZ0, SMB Bearings Ltd), which act on the reset levers. Mounted on the shaft of the escape wheel assembly is a gear, which is driven by a drive means (not shown). In the actual concept model the gear was replaced by a pulley, which is driven via a line and weights. In an actual clock, the shaft would be fitted with a seconds indicating hand. Further hands, indicating minutes and hours, would be incorporated into the drive means (not shown). The shafts of the prop assemblies and escape wheel assembly are supported on ball bearings (S604ZZ EZ0, SMB Bearings Ltd).


Materials

As I explained at the beginning of this section, the methods and materials used in my concept model are not necessarily those recommended for use in a 'real clock'.

The fame of the escapement is made of wood. This is not as silly as it sounds – John Harrison, probably the world's most famous clockmaker ever, began his working life as a joiner. His early clock mechanisms were made of wood, and several remain working after nearly 300 years! 

The plates are made from 9.5mm thick, 7 ply, birch plywood. The traditional material is of course brass, but for a concept model brass was considered far too expensive. In any case, I am not a great fan of brass in 21st century clocks, and prefer anodized aluminium or stainless steel. Alternatively, carbon fibre plates may be considered for thermal stability and visual interest.The plates are spaced apart with pillars made of beech, and the blocks holding the adjustable stops are also beech. The whole frame is held together with TurboUltra stainless steel wood screws (ScrewFix Ltd). The frame was given a coat of Osmo TopOil (Sykes Timber Ltd), to seal the grain and stop it getting too grubby. Of course, in a 'real clock', these aspects would be 'sympathetic' to the material of the plates.

The pendulum pivots each comprise two TurboUltra wood screws, the pointed end of the upper screw pivoting in the cross head of its partner below. I am not suggesting this as an alternative to proper knife-edge pivots – it certainly is not. However, it is quick, cheap and adjustable, and ideal for the concept model. For 'a proper' knife-edge suspension, I would turn to Riefler astronomical clocks of the early 20th century for inspiration.

The adjustable stops were also made from wood screws. Ideally, in a 'real clock', adjustable stops should be avoided since they can result in dimensional instability. Furthermore, to minimize wear, the stops may be jeweled. 

The pendulum arms were made from aluminium tube (which I happened to have). A better material would be carbon fibre tube; its very low coefficient of thermal expansion being more in keeping with a precision, temperature compensated, pendulum.

The weights were made of brass, but stainless steel would be better, to avoid oxidation.

The weight supports were made from 1mm thick aluminium and skeletonized for lightness. Carbon fibre might be a better alternative since it is light, strong, and thermally stable. I confess to not having bushed the weight supports. In a 'real clock' these would have brass or bronze bushes, or be jeweled.

The props were made from beech for lightness, and the locking arms from grey PVC, again, for lightness. I decided against wood for the locking arms in case the delicate regions by the rollers split off. The reset levers were made from aluminium alloy. The prop parts were held together on polyacetal hubs. Again, carbon fibre components may be incorporated for thermal stability, strength and lightness.

The shafts were made from silver steel, but stainless steel is preferred.

The escape wheels were made from aluminium alloy, but titanium is preferred.

Go to part one

Go to part three

9.
 
10.
 

11.

12.

13.

14.

Figure 3a. Front view of the SGE (front plate not shown for clarity).

Figure 3b. Side view of Figure 3a in the direction of Arrow A.

Figures 4a and b. RH weight support, where 4a is a front view and 4b a side view. Figures 5a and b. LH weight support, where 5a is a front view and 5b as side view.

 

Figure 6a and b. LH prop assembly, where 6a is a front view and 6b a plan view. Figures 7a and b. RH prop assembly, where 7a is a front view and 7b a plan view. The props are shown in their relative positions.

Figures 8a to c. Components of the escape wheel, where 8a shows the assembly with the front escape wheel removed for clarity, 8b shows the front escape wheel, and 8c a side view of the escape wheel assembly.

Figure 9. Front view of concept model.

 
Figure 10. Rear view of concept model.
 

Figure 11. Side view of concept model.

Figure 12. Top view of concept model.

Figure 13. Close up of weight, and weight support.

Figure 14. Close up of escape wheel and prop assemblies.

WATCH IT RUN!

http://www.youtube.com/watch?v=A1yH9vJN__whttp://www.youtube.com/watch?v=A1yH9vJN__wshapeimage_3_link_0