Kicking Mechanism

Hey everyone,
We are creating a kicking mechanism that has to kick an 8cm diameter sphere that weighs 100grams and the ideal speed would be between 1 and 2 m/s, the faster the better. We have approx 17cm long and 9cm wide to work with on the shape of the kicker and smaller is better. As light as possible is good too.
We are currently using an overcharged solenoid (12V push type @ 36V) but it isnt quite giving us the kicking strength we desire.
Any ideas?
TIA
Jeff

Fr335tyl3r <fr335tyl3r> wrote:
  Quoted Text Here  
- Wind a spring with a gear motor, and then let it fly. - Pump up a container with air and dump it through a valve into a cylinder. - Get a wheel spinning real fast, and press the ball onto it to get it flying.

Well, how about some elementary physics ?
To get a 100g ball moving at 2m/s takes about 0.2 Joules (wd = ke = 1/2 . m . v^2)
Small solenoids at nominal volts seem to offer something like 5N force average over motion of perhaps 5mm, which implies a 'kick' of about 0.025 Joules. (wd = force x distance)
So we are down by a factor of 8, assuming everything works perfectly, which it won't.
How to get round this ? Firstly store energy and suddenly discharge it - so use a fairly big electrolytic capacitor charged to a good voltage. Secondly, consider rewinding the solenoid to reduce its resistance and inductance, which will also affect the rate of discharge. Note also that the solenoid produces much more 'push' at one end of its travel (depending on type), and so a good strategy is to discharge the storage caps at this part of the travel.
Life is never this simple of course, so you will also be considering what happens when you push the ball hard - if it just deflects, much of the energy will go to waste.
Dave
"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  

Do some experiments. Find out just how much you are getting now (kick straight up and measure distance travelled, then do the obvious calculation to get launch speed). Vary the drive voltage and see what happens to the kick. Draw some graphs and extrapolate.
I doubt that just increasing the drive voltage by a large factor is a good answer.
Dave
"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  

"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  
I used to use solenoids to fire shotgun shells in a tank fire simulator. I charge a huge cap to 90V. The effeciency of the solenoids drops with use. I think I was breaking down the insulation, or magnetizing the core. What works a few times may not work a whole match.
Somehow you have to store and deliver sufficient energy for whatever your mechanism, whether by charing up a cap, or winding up a spring.

Like Randy said, use caps. Maybe even some diodes. You are probably not actually driving the solenoid at 36V for very long. For a really nice kicker, you could use a brushless torque actuator from lucas-ledex (BTA) THese are heavy, but awesomely strong. Pinball machines may also have some nice parts, since they rely heavily on solenoids.
Many solenoids suffer lousy torque vs stroke curves, since the plunger leaves the magnetic field. The BTA's keep in the field the whole time. Time to complete the 45 degree stroke is in the ms range.
Personally, I like the spinning wheel idea. This is how baseball pitching machines work. I also had a HowWheels car set when I was a kid that used counter-rotating rollers. Imagine my surprise when we connected that 3V motor to a model RR train transformer, and gave it 12 Volts. HotWheels rocket launcher. You can store up an awful lot of energy in 2 motors spinning at 8000 rpm, and use a servo to "squeeze" the ball in between them.
http://www.ledex.com / look for the BTA's
Mike

Go to the salvation army. Buy a spring wound alarm clock. Take it apart. Draw inspiration from it.
Any of these ideas presented can be automated. A lever with a torsion spring can be drawn back with a cam, and brought past the cam which raised it , to release it.
Mike "Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  
on
good
but

I'm happy to work with pneumatics but it has to be electronically activated as well as keeping in mind that we cannot recharge the gas/air containers until halftime (10 minute halves). It is likely to want to kick 20-30 times in a single match, so it should be able to kick this much.
Would this system just use an air pump and a pnuematic cylinder or something like this?
TIa
Jeff
  Quoted Text Here  
the

The ideal would be to kick the ball at 2m/s all the time, although a bit slower (1.5m/s) occasionally would probably be acceptable
Kick doesnt need to vary...it is for use in a 2v2 robotic soccer game so basically line up to the goal and belt the ball as hard as u can at the gap as quik as u can
The kicker might kick anywhere between 10-20 times in a half (10 mins) and we have time at halftime (5 mins) to rejuvenate the kicking system whatever it might be
Kicking 2 times in rapid succession so the 'draw back' time doesnt have to be exceptionally short but under 30 secs would be good
When the kickers power supply is exhausted (assuming we are basing on non electrical power...if electrical whole robot would shutdown) we would operate as per normal without the kicker which isnt realli a bad thing but could be better...probably just change to a system wheere it runs to the goal and simply lets go
Spring style systems seem to be the wya to go for the power but the mechanics behind them will take a little bit of brain power on my part cause I am not sure how it all reallyworks...I think i'll take that suggestion and go and get a clock and check how all the bits go
Keep suggestions coming though, there must be something out there that is even better!
TIA
Jeff
  Quoted Text Here  
m/s,
with on
  Quoted Text Here  
good
but

For a spring type mechanism, think 'crossbow' rather than 'clock' !
(Although I suppose you could consider something like the striking mech from a chiming clock.)
Dave
"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  

Yes, we would be thinking more along the lines of crossbow.
Mayb I can get some working diagrams of crossbow's off the net.
TIa
Jeff
  Quoted Text Here  
gap
and
whatever
to
non
but
cause
and

Think.
Coiled spring with foot on ouside of circumference winding. Solenoid latch release mechanism.
Bidriectional motor on inside winding spindle/axle with torsion/time/motion cutoff to self limit/set torsion.
Power motor until foot mechanism latches and then reverse motor until desired tension is acheived. Fire latch when kick needed. repeat wind before next kick required.
"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  
bit
so
the
have
the
suggestion

Fr335tyl3r wrote:
  Quoted Text Here  
does it have to be a 'proper' kick? could you have a verticaly mounted spinning fly wheel thing? 2 weighted feet to hit the ball with a motor to constantly spin it. i found this via the robot wars site...
http://www25.brinkster.com/teamvertex/index.html
craig

If you can live with the kicker being off center, think about human leg (ultimate kicking apparatus Robocup is trying to match) analogy. Use motor with long arm and massive foot at the end. Accelerate foot during half or more rotation and transfer accumulated kinetic energy to the ball. 2 m/s should not be a problem and mechanics of it is pretty simple.
Paul.
"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  

"Fr335tyl3r" <fr335tyl3r> wrote in message
  Quoted Text Here  
sphere
the
too.
it
I was digging through my junk collection (it's all good junk) and found some leftovers of a central locking kit for a car. 12Volts. These are made for after market fitting inside the car doors, attaching to the existing locking mechanism by clamping onto the small shaft between the 'Button' and the door latch. I'll call them Linear actuators for want of a better description as they are not solenoids but a PM motor with gear deduction onto a rack. (Think rack and pinion arrangement) There is about 20mm of travel and plenty of _grunt_ (Technical term for balls) It draws a heap of current and as I found out, it's best not to get your fingers in the way.... The gear reduction is not much. ~6:1
As for how much current, how much push/thrust and kicking 100grams @ 2M/s, well I'd have to do some testing but if you know the things I'm talking about it might be an option worth looking at.
Perhaps someone else is more familiar with these devices and could comment on their suitability for this particular application.
HTHs Greg the Grog

Fr335tyl3r wrote:
  Quoted Text Here  
Search on Google for "robocup kick spring", for example. There are published designs of solutions to this problem.
Mitch Berkson