Simulation of Denison double 3-arm gravity escapement

[BillM]

Art

Attached is a .vcp file that models a Denison double 3-arm gravity escapemen t.  I know you had published a gravitic escapemen t mech file but I knew that a true gravity escapemen t does not use a spring between the arms.

My first attempt at a gravity escapemen t was based on an animated gif of such a movement that can be seen at:

http://theconversation.com/the-remarkable-accuracy-of-the-trinity-college-clock-and-what-makes-it-tick-57195

My first attempt sort of worked  just had an irregular beat.  I was definitel y over confident at that point.  In trying to refine the first attempt resulted in a total disaster. ..quite a humbling experienc e.

After numerous hours of trial, error and lessons learned, I came to a better understan ding of the vector processor .

I now better understan d group collision control and other settings available to chains and groups of chains.

I hope the attached file works properly.  To run the model, start it and then pull the pendulum to the left sufficien tly to release the vane holding the left arm.  I'm not really sure how long the pendulum will swing.  Initial trials only ran for about 10 seconds.

I do plan to revise the design in the future.

Bill Michael

[ArtF]

Bill:

  Works pretty good, I got near 40 seconds here.

Art

[BillM]

Art

As I look at the simulatio n it is still not quite right.  The spin of the vanes should be much faster.  Any increase in the rpm results in either failure or erratic behavior.  Too high of a driving rpm and collision s are missed or the arms do not get released properly.

I think some of my problem lies with fine tuning the placement of the vanes, stops and pendulum.

I will soon have access again to a real gravity escapemen t that will help me with the redesign.  Keep posted... I do not give up easily.

Bill

[ArtF]

Bill:

  It may be granulati on, where the motion is not fine in time slice enough to properly react
to the hits on either side. If I redo the kinetics Ill have to look deeper into ray tracing to
capture those intersect ion events of fine proportio n.

 Good job with what you have though...

Art

[Jerrywi]

Good Morning Bill, Art,
I was going through the forum this morning and came across your post and remembere d that many,many years ago I build a clock made completel y out of paper. It uses, I believe, is the same escapemen t as you are describin g. It doesn't use a spring but rather a small paper box that I filled with sand (as a weight) . when the pendulum pushed the left arm out and released one leg of the (Star wheel) I believe that's the term the instructi ons used to described the escapemen t wheel) the weighted side would return the leg.   

My intention s where to dig the paper clock out of the catacombs and duplicate it using wood and recreatin g it using Gearotic Motions. (that hasn't happened yet, LOL)  I thought this would be of interest to you.

Thanks for letting me share
Jerry

 

[ArtF]

Very nice.. from paper yet...

art

[BillM]

Jerry

Thanks for sharing the info about the paper clock.  The escapemen t that the paper clock uses is similar to the model that I created using Augie's vector processor .  I don't know if you had a chance to download and run the model...i t does not use any springs.  The model is being designed based upon a GIF simulatio n that I found on the internet.

I think you should dig into the catacombs and recover your paper clock.  You have an ideal situation where you have the actual gears of a running clock that can be modeled using Gearotic Motions.  I think such an effort to learn Gearotics will be worth the effort.  A while ago I used Gearotics to model/document the entire main train of a tower clock that I have access to.

The particula r gravity escapemen t is not modeled in Gearotics but you do have the original that you can measure and make wood equivalen ts.

Modeling the escapemen t using the vector processor is a challenge .  Getting the timing correct between the vanes that lock the arm relative to the pins that lift the arms is a problem that I'm currently working on.

Thanks again for sharing your photos.

Bill

[Jerrywi]

Bill,
Thanks for the words of encourage ment. I admit the project is long over due.
I reviewed your model in augie and it looks great. I'm no where close to having the skills or abilities of designing a project by scratch.
My intention was to take the plans from the paper clock and scan them into a dxf format  then import them into CamBam software to produce  the framing
 then use gearotic to produce  G-code for the gears since I have the gear info at hand. 
Let me know how your model is coming. I always appreciat e someone that can take a thought or idea and turn it into a working model
Thanks again
Jerry

[BillM]

Jerry

Keep at it and don't be afraid to experimen t with Gearotics, Vexx, and Augie (especiall y the vector processor's simulatio n capabilit y).

I did refine the gravity escapemen t simulatio n to the point that it was running with a regular beat and ran for a 1/2 hour.  I then decided to stress the vector processor more by combining my motion works simulatio n with the gravity escapemen t model.  The combined model is posted as a new topic.

Bill

[Jerrywi]

Thanks Bill, I ran your model again and it's amazing
Great work
Jerry

[John T]

Hi Guys,
I have done what you guys are talking about.  I had the original paper clock so I was able to reverse engineer the unit.  The gravity escapemen t is not exactly the same as the one  you posted here - but it does work as a gravity escapemen t without any springs.  I you are intereste d in my experienc e you can read about it here

http://jtwoodenclocks.com/peace-tower.html

John

[BillM]

John

I checked out your website.  Quite an impressiv e collectio n of clocks and figits.  It looks like you have a lot of fun dreaming up and creating those machines.

I like the way you reverse engineere d the Paper Peace Clock model. As far as the gears, did you use Gearotics to reverse engineer the various gears/gear ratios?

I am quite impressed with the capabilit ies  of Gearotics .  A while ago I used Gearotics to build a model of an 1860's E Howard Tower Clock. I had the chance to actually measure each of the gears in the clock and then used Gearotics to model the actual clock.

Bill

[John T]

I do have other programs for gear design, but Gearotics is the only one used for the tower clock  I still keep and use a Gearotic 1.0 to make and design “imaginary” gears. Otherwise it’s current gearotics all the way.