-- StephenHibbs - 08 Dec 2009

Skin fabrication

* Outer skin layer made from Wacker Elastosil 3162 (black conductive rubber) as the tough outer layer. Three layers of scotch tape (roughly 6-7 thousandths of an inch thick) was generally a good thickness. The Wacker layer is baked on at 130C for 15 minutes.

• Wacker Processing Info
• Conductive Rubber info

Zoflex PL45 comprises the conductive inner layer.

• Zoflex Datasheet
• Resistivity of Zoflex, measured in layers of scotch tape, across roughly an inch of material. One layer of scotch tape is roughly 2 thousandths of an inch thick.
• All samples also had 3 layers of Wacker added.

	#Layers		4 layers		5 layers		6 layers		7 layers		8 layers		10 layers
	resistivity(Ohms)		2.3		1.3		1.5		1.5		.6		.5

• A 7.5cm * 30cm sheet made with 5 layers of Zoflex and 3 of Wacker has a little under 9 Ohms resistance from end to end. For such a sheet, 15.2grams of Zoflex was made, and 20grams of Wacker. The zoflex was baked for ~5 minutes, and the wacker for ~17 at 135C. This is the standard sheet size we've been making, and it's generally twice as wide as our gripper sensors.

Waterproof sensor:

The first version of the waterproof sensor was made with SmoothOn Ecoflex 0010, a softer version of the Dragon Skin used for StickyBot fabrications, and we have some in the lab. It's easy to mix and pour. It is a very soft and has some tendency to stay depressed when it has been pushed on, but mounting it and attaching the outer rubber helped the problem. Encasing the ends of the gel in rubber as well as the top also helped a lot.

The gel was molded to a semi-circle, and was wrapped with a layer of the skin (3 layers of wacker, 5 of zoflex) and the whole unit was glued together and to a sensor with a silicone glue. The sensor was mounted on a slightly larger piece of acrylic material to avoid bad effects at the edge of the sensor. The sensor was also embedded on the acrylic in epoxy to make the gel layer sit flush with the sensor. When there was no shielding on the bottom or sides, there were issues, but wrapping the whole sensor in grounded silicone fabric shielded it very well. Later, skin was glued to the exposed sides and bottom with zoflex, making sure to maintian the electrical connection with the rest of the skin. The sensor package will then be coated in rubber to prevent water getting into the package.

Other gels that have been played with, but not molded into a sensor, are the DowCorning 3-4170, 3-4680 and 3-4150. The 3-4170 requires heat curing and was not ever fully cured, and seems impractical. The 3-4150 and 3-4680 look good, but don't seem very distinct. They pour like water, cure softer than the EcoFlex, and seem to regain their shape faster and more consistently. They are very sticky though, even when cured, so I think it's best to mold them directly to the skin and sensor.

• SmoothOn's ordering site with more info
• Data sheet for Dow Corning Gels
• The 3-4680 is a 30 minute cure, the 3-4150 is a 90 minute cure at room temperature