Grasshopper Lego Clock Escapement W20

Grasshopper Escapement (John Harrison, ca 1722)

The contact between pallet and tooth is not without friction: the tooth tips are rounded, rather than pointed.

No recoil shock absorbers (composers) have been installed. The shocks are absorbed by torsion of the pendulum axis (not a good idea).

The pallets are bulky and heavy. An alternative would be the toothed toggle joint connector #4273, with the teeth filed off, but I did not succeed in getting it properly modified. Improved pallets have been installed in W21.


Geometry: The distance r between the pendulum axis and the scapewheel axis can be calculated from r**2 = (R1-R2)**2 + (R+)**2 with R2 the radius of the scapewheel, and R1 the length of the upper-legs (that meet at an angle ALFA at the pendulum arbor. The length of the lower legs, R+ and R-, can be calculated from R+/- = R2/s +/- R1*s, with s = tan(ALFA/2) Here, R1=6.0 and R2=4.5 (in Lego Units; 1 LU = 7.937 mm); ALFA = 60 deg.
This gives r = 11.4, which can be approximated by SQRT(7**2 + 9**2), giving the scapewheel and pendulum arbor positions.

Note: The swing, BETA, of the yellow frame is inversely proportional to the number N of teeth: BETA = (180/N)*(R2/R1). Here N=14, so BETA = 9.6 deg., which is too large by a factor of ~3 to give a smooth operation. The swing of the pendulum (ca 25 deg.), and the torsion in the pendulum axis, could be reduced considerably by a scapewheel with 30 or 60 teeth.

Category:

Science & Technology

Tags:

• Harrison
• grasshopper
• lego
• clock
• escapement
• pendulum

[keyboardjeff]

It is fantastic! How do you come up with these designs? Is it accurate?

[BenVanDeWaal]

@keyboardjeff I don't think it is very accurate. The torsion in the pendulum axis, and the large amplitude are complicating factors. To make it really accurate, I think I have to redesign the whole thing with special attention for these things.For accuracy, you have to go to the gravity escapements.

[KEvronista]

confound you, ben! it's not enough for you to simply build a gh for yourself. no, you always have to go and re-invent the entire mousetrap. what next? you gonna re-invent the mouse, too?!

am i seeing this correctly? is the crank plate detached from the pendulum? what led to pursue that aspect?

"toothed toggle joint connector #4273"

i'm not quite grasping the concept there, but would connector #44 be the piece you need?

you never fail to astound, ben. bravo!

KEvron

[BenVanDeWaal]

@KEvronista Yes, #44 would have been a much better choice; I was unaware of it's existance. However, there are other problems with this type of pallet: they easily slip off sideways from the scapewheel. When you put in an axis an uneven surface results, which has to be made flat. Finally, the axis can touch the scapewheel, which is impossible now, because I used 2 axes, on bothe sides of the wheel.

The yellow plate is fixed on the pendulum axis, but the pendulum can rotate freely about this axis

[BenVanDeWaal]

@BenVanDeWaal (contnd) if not for the brown strip which is attached to the upper part of the pendulum and (fixed angularly) to the axis. This is necessary to give pendulum and escapement their mutual orientation. So, effectively, pendulum and crank plate are fixed to the same axis, and they should swing unisono. But the angular motion of the plate is not harmonic while the pendulum motion (nearly) is, so the axis is twisted. I have not a good solution for this problem.

Thanks for your comments.

[vinesthemonkey]

I look forward to your innovative videos. :)