Robotic Braking- Quick Stopping Inquiry

[kam427]

Currently, I am in the process of designing a robot which will have "surveillance" purposes- generic observations through a wireless webcam.  The technical aspects of the transmitting device have been worked out- however, the chassis is a different story.  Specifically the locomotion of the bot- it will be making quite a few turns, and will be navigating high-traffic areas (hallways, classrooms).  In addition, the final design will weigh about 30 pounds.  The dimensions of the base are approximately 24" by 18".  The wheels themselves are a bit oversized with a diameter of 1 foot, each.  Four wheels will be used, and the two on each side (right, left) will be connected (so that they spin together).  Two motors will be used- one per side, to enable differential steering.  However, the current problem is on braking.


The Problem

Basically, I do not know how to proceed with the braking system.  Every other schematic has been thought out of- but there seems to be a major problem.  I would like to use frictional braking, however, I do not know how to implement it.  Should a locking system be used that LOCKS the wheel in place (though the vibrations may damage the equipment on board), or should enough force be applied by a high-torque motor of some sort that pushes rubber upon rubber?  Or should it go far enough to push a pad of rubber against the ground, providing great friction and vibration?  Or was something overlooked, and that a MUCH more efficient method exists?  Moreover, I would like to emphasize that simply stopping the flow of current to the motor will not suffice- some sort of direct braking mechanism is definitely needed. 

[pomprocker]

check out the stairbot braking

http://www.stairbot.de/en_beschreib.htm

 Both omniwheels use simple disc brakes. The brake is made from a flexible plastic part covered by a lining (black/red) normally used for table-tennis bats (= ping-pong paddles for american readers).

Moving the servo mounted cam pushes both brake discs ( via the rods) against the omniwheels. The brakes will be released by retaining springs. The polyurethane-rollers of these wheels (  60mm TRAPO-rollers) provide sufficient adhesion even on smooth ground.

[GHF]

If you connect the terminals of a motor together, forces caused by Lenz's Law will resist motion and brake very effectively. Many motor controllers have this feature, and systems built with them require no extra mechanism in order to brake. Since the braking force generated is proportional to the motor's RPM and the current it is caused by can be controlled, these impromptu "eddy current brakes" can be adjusted to be harder or softer.

~edit~
But, sometimes you do want mechanical brakes, mostly if you traveling relatively fast or with a lot of inertia, or if you want to stop somewhere and stay planted so other things (people?) can't budge it. Even then, you'd probably want to combine it with motor braking.

[kam427]

Hmmm...

The first design using omniwheels would work- however, is there a way to implement that design onto larger, non-omniwheel type designs?  And how exactly would the servo mounted cam be moved in the first place?


As for the connection of motor terminals- wouldn't that also strain the gearing system?  Since the motor itself is geared down, and the motor now is locked in place- then the gears would have to deal with extreme amounts of tension/torsion.  Wouldn't this cause extensive wear? 

[Iron Man]

No expert on the subject.

but from experience we used motor breaking on a 130lb robot that got about up to 8-10fps and it was geared down.

we also did similiar in an arm that lifted a 10lb ball from the end of a 6 foot arm. which was also geared down using a planetary system

We didn't see failure in either situation. However the robot only had about 2 weeks of heavy competition, lots of robot to robot contact, but idk how they would have stood up over long duration day to day use.

The motor breaks are definetly a strong choice, even for a geared system.   imho.

just make sure your chain is tight if you're using chain. 

[kam427]

Theoretically then, if motor breaking was used- which particular gears should be used?  (to gear the system down)

[ArcMan]

Like Iron Man said, motor braking is very effective.  I plan to use it on my 100 lb. lawnmower robot.

Note that a worm gearbox is the most effective at holding a stationary position.  It is virtually impossible for the wheels to spin the motor when connected through a worm gearbox.  These gearboxes are very common - windshield wiper motors, power window motors, wheelchair motors.

[kam427]

Thank you so very much for your assistance... It's greatly appreciated... xD

[Admin]

There are motor drivers that electrically break motors, basically by shorting them.

Another possible option would be to put a PID control loop (using encoders) that will suddenly reverse motor current to break the motor.

As for mechanical breaking, why not imitate automobile breaking systems?

And how reactive/accurate does your breaking need to be? Whats the weight of your robot?

[kam427]

Imitating automobile brakes was attempted at first, but creating the hydraulic disc brake system posed a challenge at first.  The braking needs to be VERY accurate- stopping the robot within 1-2 feet, since it will be navigating amongst humans in classrooms, etc.  And as stated before, it weighs about 30 pounds.  xD