Society of Robots - Robot Forum

Mechanics and Construction => Mechanics and Construction => Topic started by: penfold99 on April 28, 2012, 02:26:08 PM

Title: Hopping leg actuation
Post by: penfold99 on April 28, 2012, 02:26:08 PM
Hi

I'm having a bit of trouble with ideas. I am creating a robotic hopping leg.. or two. The issue of having two legs is not the point at the moment.
The problem being, how to create the linear actuation to push the robot off the ground.
A variety of methods have passed through my mind and been investigated, these being: pneumatic, hydraulic, electric motors and some sort of coiled spring based trigger system.
The problem with all of them except the trigger idea, are too slow, too bulky or too expensive. It's a personal project so budget is limited, and technical machinery is also.
The trigger system, I can't really figure out in my head, but I have a vague idea.
So, I was wondering if anyone has any ideas of linear actuation methods, which would provide a brief but reasonably powerful to the leg (imagine a leg of two cylinders, one slides into the other, the top leg pushes upon the bottom), the actuation distance need not be great, one cm or two... but it would need to be fast. The weight of the robot would be a few kg at most.
The hopping amplitude would be built up over time to a harmonic peak.

Thanks
Title: Re: Hopping leg actuation
Post by: Soeren on April 28, 2012, 08:03:25 PM
Hi,

[...] the actuation distance need not be great, one cm or two... but it would need to be fast. The weight of the robot would be a few kg at most.
The hopping amplitude would be built up over time to a harmonic peak.
You already have most of the solution then - harmonics is the way to go, to keep the average power requirements bearable.

Not having seen what you plan, my immediate thought is to use a large (home made) solenoid, with a (magnetically shielded) coil in the main body, with the force being transfered through the leg via a (carbon fiber?) tube to a reasonably hardness synthetic "rubber" "shoe" ( a brake pad fom a side-by-side roller skate sounds about right).
The outer shell of the leg doesn't need to be super light, but the actuating rod should be as light as conveniently possible and don't make the leg too short, that will make it less "stable" - think of it like balancing a hammer on one finger; with the head up it's easy, the other way it's "regulation time" becomes too short.

You should experiment with the stroke length vs. supply power - longer strokes can be relatively slower for the same result (jump-vise).
Trigger timing will take either optic- or Hall switches to determine the position of the "kicker" (both for the trigger signal and to determine when a full stroke has been reached). Done that way, it will automatically adjust to the harmonic frequency of your jumper, even if you add or subtract some weight or a tether (and you'd need an accelerometer to determine which way to angle the leg of course and the leg needs a 360° moveable hinged connection to the body.

Add plenty of capacity to the solenoid supply, to make it kick fast and powerful. The coil will depend on your voltage and power switch element.


For short run times, a small CO2 bottle could be used for pneumatic actuation, but pneumatics valves and such are costly.
Title: Re: Hopping leg actuation
Post by: penfold99 on April 29, 2012, 07:25:44 AM
I didn't really consider solenoids.. silly me.
Done a bit of research and as long as I can create one powerful enough, and shield it enough so it doesn't interfere anywhere in the circuit, it does seem like the perfect idea.
Thanks.. I'll return if I have any more revelations :P