The Beall Clock Forum

General => Other escapements => Topic started by: dcunningham2 on December 20, 2006, 10:07:25 PM

Title: Single leg gravity escapement -- case of the missing energy
Post by: dcunningham2 on December 20, 2006, 10:07:25 PM
I've built a paper clock from a kit (The Peace Tower from Wrebbit) that uses a single-leg gravity escapement.  It was damaged during a move and now I'm trying to repair it and get it going again.  It used to function consistently, but now I find that I cannot keep it running.  

I don’t know if all single-leg gravity escapements work the same and I’m unsure of the proper nomenclature, but I’ll attempt to describe its operation in my layman’s terms.  Real horologists are asked to refrain from smirking, thank you.

The escapement consists of a six-toothed escapement wheel and two legs that dangle on each side of the pendulum shaft.  One of the legs is weighted and the other is not.  When the pendulum swings against the non-weighted leg, its outward displacement releases the escapement wheel.  It rotates one sixth of a revolution until contacting a stop on the weighted leg.  The effect of the rotation is also to raise and “cock” the weighted leg so that it is poised to drop against the pendulum arm.   As the pendulum contacts and displaces the weighted leg, the wheel is released again and rotates another sixth of a revolution until it contacts a stop on the non-weighted leg.  The weighted leg is also released and free to bear against the pendulum arm as it swings back to its original position, providing an impulse to the pendulum.  The cycle then repeats.

Or so it used to.  Now, the escapement appears to trigger and release as before; however, I find that the pendulum arc gradually reduces in magnitude until it fails to contact the "inactive" or non-weighted leg.  At that point, the weighted leg is not lifted or cocked and so no more energy is provided to the pendulum.  I cannot find obvious energy sinks, such as the pendulum rubbing against a surface, or a change in the geometry between the wheel tooth and the “trigger stops” that might increase the releasing friction.

I'm thinking there may be one of four causes for the insufficient energy:  

1.  The pendulum arm has lost the necessary stiffness to transfer enough of the impulse from the gravity arm to the pendulum, or
2.  The non-weighted escapement leg no longer swings freely enough, such that energy is absorbed by its friction throughout its stroke, or
3.  The weighted escapement leg no longer swings freely enough, such that less impulse is delivered to the pendulum, or
4.  Some weight has been accidentally lost from gravity arm so that less impulse is delivered to the pendulum.

I’m interested to hear if anyone has any theories as to where the energy has gone.

Also, I’m curious if anyone has also built this clock, and what results they obtained.

Dave Cunningham
Title: gravity escape problem
Post by: rabbit on December 22, 2006, 06:52:02 PM
i've built a wooden clock with almost the same escape mechanism. even though a single impulse in the pendulum period doesn't technically have a "beat", this clock is sensitive to having equal travel in each direction. (instead of "tick-tock", this escape goes "click-click pause tock") have you checked/verified that it's hanging level?
any or all of your supposed energy losses could be correct - excellent analysis.
if it is #4 ("lost weight"), try adding weight to the gravity arm until just before the point where the train doesn't have the power to lift it anymore. this will maximize the impulse for the power that's available.
if it's #1-3, just plug away - ah, the wonderful world of clocks; in search and kill of friction.
i would love to try my hand at a paper clock - yes, they too were once trees.
Title: Thanks for the help
Post by: dcunningham2 on December 23, 2006, 10:59:43 AM
It's interesting to hear that this escapement is sensitive to being level -- I had tried hanging the clock at various angles with but got no satisfaction.  

It strikes me that if I lean the clock "towards" the weighted leg such that it delivers impulse over a greater arc of the pendulum, some of the additional impulse will be consumed in recoil before the pendulum reverses direction.  Also, the pendulum will need to swing farther in order to contact the unweighted leg.

On the other hand, if I lean towards the unweighted leg, the pendulum arc won't need to be as great to contact the unweighted leg, but there will be less impulse available to get it there because the weighted leg will act over a shorter arc.

AHA !!  I think you have helped me uncover another source of possible energy loss.  One of the problems with paper structures is that they are less than dimensionally stable, to say the least.  Most of my work in restoring this clock has been in trying to unwarp it.  It now occurs to me that the geometry of the escapement may have changed -- the distance between the legs may have increased.  Specifically, with the pendulum arm removed, I think the distance between the contact points on the arm pallets when hanging freely (I may using the wrong terminology) should be equal to the width of the pendulum arm.  I suspect that the arms may be deformed such that this gap is increased and the distance the pendulum must swing is now greater than the what the available energy will sustain.

I'll check it out.

I know that I can get it running again if I increase the weight in the arm, but I'm holding that out as the last resort.

Thanks again for your thoughts!