This week we made things by casting into molds. My original plan was to make an experimental rappel device modelled off of a figure eight device, but in the shape of a goat's face. I was going to use a lost foam casting technique to cast my part out of aluminum. I CADed my part in Rhino and then milled a positive out of foam on the ShopBot. The lost foam technique uses tightly packed sand bonded with either oil or water as a mold. When I tried this, the sprue hole ended up being too long and so the aluminum froze in the sprue and the part didn't come out, but it did melt part of the foam.
After I did this first trial with aluminum, it was suggested to me that I should make a silicon mold and cast into that, because it would give me more experience making molds than the lost foam casting technique. So I then decided to switch to babbitt and an oomoo mold. I switched tracks pretty late in the week, so I decided to stick with the positive that I had already milled so that I could have enough time to make the mold and cast into it. I made a box out of hardboard and put my foam positive in it and then filled around that with oomoo. After the mold hardened, I cast the final positive shape with babbitt.
While the piece was designed to be a prototype rappel device, it was never going to actually be a reliable device for this purpose. Rappel devices are frequently made from aluminum, however casting doesn't create a piece with the full strength that can be attained from aluminum that's been forged, for example. At it's original planned size and material, the device probably would have held a persons weight, but would not have nearly the margin of error that is required for a structural piece such as an 8 type device. In the end, the piece is much thinner than planned and made of babbitt making it much weaker and definitely not fit for use as a rappel device. In addition, the melting point of this alloy is in the 200-300 degree F range. With the localized heating that a rappel device is subjected to at the points of contact with the rope, the piece could potentially start to melt onto the rope during a long rappel. From some rough calculations based on how the device is loaded when rigged, one gets that it can support about 1000 N which is a little less than twice my weight. However, casting gives very unstandard results and so this is just a rough guess. I might pull test it at a later time and see how much force it can withstand. I did do two test runs in very controlled conditions, and it withstood both of those tests. In the first test, I started about three feet off of the floor, so if it started to fail, I could just stand up. If it failed suddenly, I wouldn't fall that far. In the second test, I rappelled a further distance, however I was on belay on a second line. If it had failed then, the belay line would have caught me. Afterwards, the rope did not appear to have any metal worn off onto. The device started to get a small rope groove, which is normal for belay devices, but usually takes longer to form from a dry rope. Videos of these test runs can be found in my public folder.