Angular momentum drive system timing?

JohnUlett

In my recent post i mentioned adding energy to the drive system by having an off balance disc rotating in the y-z plane. As the weight goes through the bottom of its rotation it would be at the same time the robot is in a stride and thus add energy to the springs in the legs, as the robot come off the ground, the weight would be going through the top of its rotation giving the robot energy in the positive z direction. how would you design a control system that changed the rotational speed of the weight to match the period of the stride and flight sequence?

iamdenteddisk

I can see this absolutly working,great concept..such a bot could theoreticly hop through the grass like an impalla on the african plain.
plauseable but I belive a centrifical clutch like that on a gocart and a digitaly controled brake calliper would need be used to pull the sequence into alignment such a mechinisem would shred meshed sprockett's at a direction or speed change and the disk itself would have to great a centrifical force to be reduced and sped back up in a useable time frame useing any motor without a brake ,at the same time the amount of applied brake and the durration of brake will also change angular momentum like a motocross rider aplies rear brake to bring the nose down on landing or gases it in the air to bring nose up.
love to see this in action..goodluck.

watching horses ,dog's any fourleged creature in stride youll notice that the neck&head extend upward as the ground is pulled to the aft end then the neck&head ducks down as they reach out for new ground this is while they are runing"galloping" and front and back legs are paralell with each other, but when a horse starts to trot the neck goes sharply back and the aft end rolls under as the back streightens and action becomes slightly out of sequence like two men carrying a sofa.. spend an hour playing fetch with a dog you'll see my point ,just maybe this walk method can be duplicated without sophistocated mechanizems by simply following form&proportion..
either way I like where your going with it..

Post Edited (iamdenteddisk) : 6/13/2008 10:22:23 AM GMT

JohnUlett

What about a linear weighted system. (vertical tube with an iron core that goes up and down) I could use coils of wire around the slug to make it go up and down and the timing system would be much simpler then rotational. Could a minimal electric current provide the necessary magnetic field to lift the iron core vertically to provide a worthwhile change in momentum?

iamdenteddisk

I see no reason it couldn't be experimented with, best advice here is shielding to prevent the fields of one coil from affecting a different coil used for another part and revers current prevention as coils induce a secondary voltage. the idea in it's self is a very intelligent concept. mimicking nature in robotics. seen in Leonardo Devinci's work and the ancient Egyptian relics.

mostly the question is, do we have current technology capable of doing it this way. if not we need someone to try(experiment), so technology can catchup.
P.S., sorry to take so long to come back to this forum and reply..

kwinn

In nature the energy is transferred from the body's downward velocity into a spring, and then transferred back as upward velocity. The galloping/hopping gait/frequency does not vary much with speed, the length of the stride increases or decreases more.

If you wanted to increase the amount of energy stored/released with an oscillating mass an off balance disk driven by a motor whose rpm is controlled by a pulse width modulated driver would be mechanically simpler.