Care and Feeding of your Spiny Feet



Care and Feeding of your Spiny Feet

By Alan, Aaron, Sangbae

Shipping contents:

6 of Sangbae’s feet

4 assembled feet with Alan’s toes

2 assembled feet with Aaron’s toes

2 sets of spare Alan’s toes

1 set of spare Aaron’s toes

6 nuts and bolts for attaching Alan’s/Aaron’s feet to the robot

Assorted running feet for you to play with.

[pic]

Notes in general about the feet:

If you have any questions about anything, do not hesitate to call us at lab or email us. We will be happy to explain anything to you and would rather you called first rather than possibly break something.

You should not attempt to use the feet on surfaces on which they do not work well. This is the primary cause of toes breaking, which should be avoided as much as possible because it is a major pain making more feet. Treat the toes like your firstborn children. Give them names if you want. Be very sad when you break one. The feet are the result of much, much painstaking manual labor.

YOU SHOULD TEST THE FEET ON EVERY SURFACE YOU WANT TO CLIMB by hand before using the feet on the robot on that surface. To tell if the feet will work well on a surface or not, do the following:

1) Hold the foot IN YOUR HAND by the ankle and press it lightly against the wall, until there is some deflection of the toes, but you’re not applying tons of force.

2) Drag the foot downwards parallel to the wall about a centimeter. At the bottom of the centimeter, you should pull away from the wall slightly. Hopefully a bunch of toes will have caught on surface asperities. IF ONLY A FEW TOES CATCH, IT WON’T WORK! For Alan’s you need to have at least 10 toes catch on average for it to not break under a 1 kilogram load. For Aaron’s toes, you need to have at least 4 catch. For Sangbae’s, you need at least 7 to catch (or 25%).

3) You should familiarize yourself with how far the toes can stretch before breaking. I was going to send you some test toes to break, but I forgot. So, do the following: For Alan’s toes, take one of the spare toes (with the black stripe on top--see below) and stretch it out slowly (with your hands) until the flexure breaks. There are enough spare toes that you can do this with one of them and it won’t cause problems. As you stretch the toe until it breaks, carefully observe how far you are stretching it. If the robot is climbing on the wall and not enough toes catch on the wall, the foot will move downwards until the caught toes stretch too far and break. So, if you carefully observe how far the toes can stretch until they break, and pre-test the toes on the surfaces before trying to climb them with the robot, you can prevent toe breakage and have the toes last a long time. For Sangbae’s and Aaron’s toes, there aren’t as many spares, so you should just pull on them and observe how far they are stretching, but not so much that they break.

4) If you use gaits with different numbers of feet on the wall at once, you may be able to get away with climbing smoother surfaces on which fewer toes attach. Here are the breakage limits for our feet: Alan’s toes—break between 100-120g, and there are 38 toes/foot except for if you make feet of the spare toes, which would have ~34 toes/foot. Sangbae’s toes—most of them break between 100-150g, and some of them (the original batch) break at 200+g. Aaron’s toes break between 350-400g, and there are ~12 toes/foot. Note that the furthest-extended toe will always be the one to break first. Good rough surfaces on which the feet will work well on are: rough rocks embedded in concrete (see picture below), stucco with lots of rough bumps, very rough concrete like cinder blocks, the bricks we had at SwRI, and probably tree trunks and wood planks (though we haven’t tried them on these surfaces yet). Surfaces on which you might think the feet would work well but they actually won’t are: surfaces where there are discrete holes or bumps every so often, with relatively smooth material in between, such as the concrete board we had at SwRI, some walls with smooth rocks embedded in them, smooth concretes.

[pic]

One of Alan’s feet hanging on a rough-rocks-in-concrete wall (a good surface). Note that although a few toes are stretched pretty far out, there are about 20 toes that have stretched out just a little (hard to see in the picture) and are carrying most of the load. If these 20 toes hadn’t caught, the few that stretched far would probably break due to the additional stretching that would occur.

To disengage the toes, be sure to move the foot forwards and up at the same time, ideally in an arc starting maybe 10-15 degrees from the wall and finishing at maybe 30-45 degrees from the wall. If you just pull directly out from the wall, the toes are much more likely to jam. Note that the toes will be extended different lengths so you need to move the foot semi-parallel to the wall for a while to disengage them all.

Notes about Sangbae’s and Alan’s feet:

For both Sangbae’s and Alan’s toes, the 200 μm spines were used. Our supplier changed the spines and made their pointed tips a lot more slender than before (smaller cone angle), so we had to trim them so that the ends of the spines wouldn’t break (because, as we all know, the strength the spine can support before bending goes as the diameter squared). Sangbae sharpened his with a file after cutting them, Alan did not. So, Alan’s spines are now strong, but they are not very sharp. Consequently I think they will function mostly by mechanical interlocking with rough surfaces. As they get used, they may actually become sharper, since the spine tip is kind of a strange shape right now. If you want to make them sharp so they can climb smoother surfaces (for a while, until they get dull), you can file down the tips of each of the spines by hand. I expect this would take multiple hours and not be much fun. Otherwise, we will just climb rougher surfaces until we get better spines.

Also, you should take careful note of the angle of the very tip of the spines (for both Sangbae’s and Alan’s toes). The spines may get bent through use, if a very large load is on one of the spines. In that case you will need to re-bend the spine back to the way they are now—with the very tip of the spine at about a 45-60 degree angle from the wall. You can bend the spines using needle-nose pliers.

Notes about Alan’s and Aaron’s feet:

The ankles for Alan’s and Aaron’s feet should be mounted on the robot in the following way: put the side with the electrical tape, which is the side with the lower Aluminum wall, towards the inside of the robot. The metals rods will then stick out on the outside of the robot and won’t run into anything. I cut down the inside Aluminum wall of the foot to reduce the feet interfering with the lower legs.

You should screw the feet onto the {outside, front} hole on the lower leg for the front two feet, and on the {outside, back} hole on the lower leg for the middle and back legs. This will give the feet enough room so they won’t hit each other for I think any existing gait.

If you want to make all 6 feet have Alan’s toes (useful for climbing some surfaces), you can take Aaron’s toes out of the feet and put the two sets of spares into the two ankles. This should be pretty easy. You should slide Aaron’s toes and dividers onto one of the rods that are currently holding Alan’s spare toes, so you don’t have to put them all back together again (big pain).

You should currently NOT mix Alan’s and Aaron’s toes on the same foot. This is because Aaron’s toes extend down too far in the current version and would prevent Alan’s toes from touching the walls with light normal pressure on the foot. We will make more Aaron’s toes that have a different design and can be mixed with Alan’s toes.

Aaron’s toes sometimes get tangled, even with the dividing walls. You’ll see how this happens. They will need to be untangled reasonably frequently with the current version. Alan’s toes will occasionally but infrequently get tangled.

Alan’s feet have been colored different colors with a marker, as you will see. There are two color codes. The first is the stripe on the TOP of the toes, near the small pin that attaches the toes to the feet.

|Color of stripe |Meaning |

|No line |These toes are the best, in terms of their stretchiness and damping and breaking strength. |

|Orange |These were from the other batch, that had bubbles and didn’t cure as well, but these ones came out |

| |pretty well and should be almost as good as the no line ones. |

|Black |These were from the bad batch, and have bubbles in the main flexure, such that they can probably only |

| |support 50-100 grams before breaking (instead of 100-120). Some are pretty good, some are pretty bad. |

| |You should break one of these when you test-break a toe. Use one with few bubbles to get an accurate |

| |picture of the other toes’ breaking strength. |

The second color is near the bend in the outer part of the toe, and is on the side of the toe. The possible colors are: Clear, Orange, Purple, Black. These colors were to indicate different toe designs. The designs have relatively minor differences in the back flexure length and width, and shape of the middle top flexure. If you notice one of the feet consistently breaking more toes than the other feet, please tell me (Alan) the color of those toes. The clear toes should be able to support a slightly higher load than the other toes—I suggest putting these on a front foot. Otherwise, don’t worry about these colors.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download