Your tentacle

For posts about your own soft robot designs. Look here for Super-Releaser official releases, robot downloads, and details on specific designs.
Post Reply
JonCard
Posts:8
Joined:Sat Mar 08, 2014 4:41 pm
Your tentacle

Post by JonCard » Sat Mar 08, 2014 4:50 pm

I have been trying to reproduce the tentacle on your site, and a 4-fingered gripper I found on Instructables. Chronicling those adventures here: http://instagram.com/p/lRKLfwAuUf/. Just finished my first attempt at the tentacle, and am debating moving forward, or attempting a better casting. Not sure how to get a good seal through the silicone with this as-is.

Thanks for posting your stuff; it's been great.

User avatar
Gianteye
Site Admin
Posts:19
Joined:Fri Feb 14, 2014 8:30 pm
Location:NYC
Contact:

Re: Your tentacle

Post by Gianteye » Sat Mar 08, 2014 5:29 pm

Howdy,
Seeing the experiments you're producing is pretty exciting. It really validates what I'm doing when other people are able to pick up one of my designs and run with it. I think I might be able to help you with some of the flaws you're seeing in the castings.

You should try casting the cores in a neutral colored jewelry wax. I'm using this aqua wax from Freeman which melts at a relatively low temp, doesn't tint the final silicone skin, and is relatively cheap. Parafin is a pretty pure wax which makes it pretty weak and a bad option for long structural stuff like the tentacle cores. Jewelry waxes like the one I'm using have a plastic component to them, which makes them terrible for turning into candles but super strong and rigid.

When I cast the tentacle cores, I wipe the hot wax off the surface of the mold with a flexible sheet of plastic or a bit of cardboard so I don't have to trim the flashing off afterwards. It just saves extra work and hassle later.

I assemble the cores with the help of an alignment jig and a hot knife for melting extra wax into the join between the base and the long core elements. The hot knife is just a sculpting tool jammed into a soldering iron with some steel wire to close the gap. You might get some sticky wax to make the process of joining the base to the cores easier.

I still haven't found a way to get all the wax out cleanly. There always seems to be a little skin stuck to my robots no matter what I do. I've been considering getting a washing machine and dryer just for tumbling robots and squeezing out the wax, but it might be a matter of finding the right cleaners. The best process has been heating the silicone casting in the oven until the wax is liquid, squeezing out the wax with water resistant heat proof gloves and then repeating the process a few times with soapy water.

You're doing awesome work and I'm eager to see what the next round of experiments look like.
Work stuff - Super-Releaser
Personal Stuff - HAR.MS

JonCard
Posts:8
Joined:Sat Mar 08, 2014 4:41 pm

Re: Your tentacle

Post by JonCard » Wed Mar 12, 2014 1:16 am

Cool. Once I got started with the paraffin, a couple of people here at the Artisan's Asylum in Somerville, MA, suggested jewelry wax, but I already bought all of this paraffin.... :)

I skipped the alignment jig for no good reason except that I didn't know what to do with the AI file. I figured out it was Adobe Illustrator, but didn't understand until this picture above what it was used for. That makes a lot of sense. Some of the guys here are putting a laser cutter together; I'll have to see if I can get this piece going.

What you said above is basically what I did. The wax was molten, I basically just "milked" the tentacle. Once it solidified, squeezing it broke off the big chunks, too.

I was able to inflate Mark 1 today. One chamber had a hole from one core being too close to the edge at the tip, and one core had broken off at the base a few times and I repaired it too close to the edge, so it tore near the base, but one inflated. I'd thought I couldn't get a seal, but Aaron W. here told me to just put the tube in with a zip tie, instead of playing around with fittings and such for now. That worked.

The ultimate goal is to combine the tentacle with a four-fingered gripper, and create a kind of "Venus flytrap" that can grab things out of the air. I'm hoping to combine that with a class I'm taking in kinematics to try to sense the location of something, and solve the control problem to try to grab it in a sphere around the stalk. Another thing I was thinking of was etching designs into the "flower petal" so that it inflates funny, so that when it inflates an "eye" lifts out of the "leaf". I'm not sure if this will break the flexing power of the "leaf".

JonCard
Posts:8
Joined:Sat Mar 08, 2014 4:41 pm

Re: Your tentacle

Post by JonCard » Wed Mar 12, 2014 2:30 am

I've been looking at the Aqua Freeman Flakes. One of the things that has me interested in them is the low shrinkage. One of the issues, in addition to the brittleness, has been that I need to fill the molds twice because the wax pulls into the mold so far as it cools.

User avatar
Gianteye
Site Admin
Posts:19
Joined:Fri Feb 14, 2014 8:30 pm
Location:NYC
Contact:

Re: Your tentacle

Post by Gianteye » Wed Mar 12, 2014 4:11 pm

Excellent! Glad to hear the tentacle worked.

I also had trouble with getting fastenings and air fittings on to the silicone bots in the beginning. My solution for that has been to laser cut a wooden base for the bot with holes slightly smaller than the hoses that I'm putting into it. When I push the hoses in they're pretty well secured, being squeezed by those holes. Then I coat the surfaces that are going to touch (the bottom of the silicone robot, the top of the wooden laser cut base, and the ends of the tubes sticking out of the base) with Sil-Poxy and press them together, trying to make certain I'm not trapping any air bubbles. That's been really effective for the conical tentacle, the trefoil tentacle, and the early quadrupeds.

I've also been experimenting with 3d printed luer fittings. Currently I've been printing bases for the quadrupeds and sticking them on with Sil-Poxy. It's been working out pretty well.

As for the wax, shrinkage is always a problem. Wax expands more than anything I know when it melts. Most of my wax candles and jewelry experiments have needed lots of trial and error to figure out how to compensate for it. One solution has been heating the mold, so everything cools pretty evenly and the wax shrinks all at once. This makes for a simpler casting process but means the part ends up smaller than the mold. The other option is to cast into a room temp mold that has a sprue so that when the outer surface of the wax hitting the mold crystalizes and the interior shrinks, there's more wax in the sprue to flow down into it. Unfortunately with the trefoil molds this isn't possible, but I haven't had too much trouble with topping them off as they cool. I made myself a little wax pot for heating and pouring wax easily.

I think adding a gripper on to the tentacle is a fabulous idea. You might even be able to get a hole in the center of it so you can pipe air lines through to the gripper on top. Have any sketches of what you're planning?
Attachments
BhnC9r8IUAAePc7.jpg
Printed luer fitting closeup
BhnC9r8IUAAePc7.jpg (43.61KiB)Viewed 16332 times
BhrFINoIgAAU37E.jpg
Quadruped with printed luer fitting
BhrFINoIgAAU37E.jpg (84.38KiB)Viewed 16332 times
Work stuff - Super-Releaser
Personal Stuff - HAR.MS

JonCard
Posts:8
Joined:Sat Mar 08, 2014 4:41 pm

Re: Your tentacle

Post by JonCard » Sun Mar 16, 2014 2:29 am

I put the sketches of the flower design in a separate topic, under Designs.

JonCard
Posts:8
Joined:Sat Mar 08, 2014 4:41 pm

Re: Your tentacle

Post by JonCard » Tue Mar 18, 2014 2:46 am

I have a question about your article on the Trefoil Tentacle here: https://www.adafruit.com/products/997#Description

I had assumed that the use of PWM control of a maximum pressure, rather than a variable pressure valve, would require the off state to be vented to the atmosphere, probably through a needle valve or similar restricted flow device. Wouldn't that require the off state in the control valve to be open? Would it require a 3-port, 2-way valve that ventilates in the off state, which the linked valves aren't, or am I making it too complicated?

User avatar
Gianteye
Site Admin
Posts:19
Joined:Fri Feb 14, 2014 8:30 pm
Location:NYC
Contact:

Re: Your tentacle

Post by Gianteye » Tue Mar 18, 2014 4:03 am

The valves I'm using just bleed when there's pressure on them and I have a needle valve inline with the air supply to allow the pressure back out. I'm not certain if the Adafruit valves bleed when there's back pressure. I'm using a board that came with a bunch of these on it, but they're pretty scarce on ebay, etc, so didn't want to specifically recommend them.

I'm sure there are more precise ways of doing this, but my goal has been to have as few parts as possible in getting these bots working.
Work stuff - Super-Releaser
Personal Stuff - HAR.MS

Post Reply