A slightly different B9 Robot
Pneumatic Claws of Death!
Home
Bluetooth sound
Testing & Mobility
Cost (never done)
Treads DONE!
Power DONE!
Control DONE!
Structure DONE!
Knees DONE!
Legs DONE!
Torso Rotation DONE !
slip rings
Torso
Not neon
Not Neon Sound interface
Collar
Slightly different radar
Pneumatic bubble lifter
Claws & Arms
Pneumatic Claws of Death!
Center of Gravity
Robot pneumatics
RoboPhone

2_03_ghost_planet_06_jpg.jpg

Can Crusher?

Nice looking claws, good. Nice looking claws that actually work? Very very nice...

 

Having working Robot claws opens up many practical applications even for a Robot that just stands in the corner. When I was working out the details I had to decide how strong the claws and their mechanism would be. Too strong would be a problem: Someone could get hurt or the claws could break themselves. Too weak and an object held in the claw could be dropped. In the end I decided that a design requirement would be the ability to crush an (empty) aluminum can. It takes about 2 lbs to start and 8 lbs to finish crushing a typical aluminum beverage can from the sides. Not enough force to hurt anyone but enough to do some useful tasks, crushing a can being just one of them!

claw_paint.jpg

The claws were "painted" with red Plasti-dip spray, only 3 coats. More coats seemed to build up unevenly and lost some of their gloss.

 

If you made the claws per the claw page: Slightly different Arms & Claws, with the linkage holes in the correct location the following design will work well. If not I can’t say they will work at all, they could be dangerous or self destruct on first use.

 

This is a fairly simple linkage but a few design particulars:

 

  • The air cylinder is set up so that the rod extension is stopped by the cylinder itself instead of the claws and linkage. This saves the claw and linkage from damage.
  • Length of the links is based on the geometry of the claws and the position of the cylinder. If the links are made shorter they may form a toggle linkage and could go over center and create very high (damaging) forces.
  • Cheap nylon bearings are used to make the mechanism reasonably durable.

 Wrists were made from PVC pipe caps and assorted PVC material. Claws themselves are painted with 3 coats of red plasti-dip spray. Wrists are painted with Rustoleum  #7784 Charcoal gray.  Aluminum brackets support the claw parts and the air cylinder. At the moment the wrist plates can be repositioned to orient the claw in any direction. For now the air cylinder is 1 1/16" dia with 1" stroke.  It is from McMaster Carr: #6498K171

claw_link2.jpg
claw_link3.jpg

In the above picture one of the side plates is removed. The white material under the moving parts in the assembly is 1/16" thick self adhesive Teflon. Note that the thickness of the Teflon is the same as the thickness of the flange on the nylon bearings that support the claw parts. When the assembly is bolted together there is maybe .020" side play between all parts.

 

For a fairly well detailed drawing of the mechanism here is an AutoSketch .SKD file of the claw assembly:

Claw AutoSketch file

The air cylinder uses the same compressed air supply system that the bubble lifter does. Once the air system was in place it was very economical to extend the system to include operating the claws.

 

The on board pneumatic system is here: The pneumatic system

 

These particular air cylinders are single acting, spring return. This means that when air is applied the rod extends and when air pressure is removed and the air is allowed to escape the rod retracts due to the spring force alone. This also means that the claws normal position will be “Open”. With a different air cylinder this could be changed, a “spring extend” cylinder would normally have the claws closed but would also be limited to the spring force to hold the claws closed. Other possibilities would require air cylinders that are double acting but this would also require different solenoid valves

 

A great thing about using pneumatics is that within reason the claw will grip based on force, not position. No feedback or force measurement is required. Say the claw is gripping an irregular object and said object shifts a little. The claw will just maintain its grip on the object and hopefully not drop it. At the moment there is a single pressure regulator that maintains the air pressure in the claw system. It would also be simple to have 2 regulators, one for low pressure and one for high pressure. Each regulator would also have to have its own solenoid valve. This could be set up so that normally only low pressure air was used, just enough pressure to make the claw move. When necessary the high pressure could be used for times when “Crush, Kill, Destroy” forces are needed...

Click below for a video of some claw crushing tests:

Enter supporting content here