Rubik's Puzzle: Pyraminx Barrel

19-8-2008

The Rubik's cube is an icon of the 80's. Nowadays, it is much less common, but still an icon. In the past few years, a sport, called SpeedCubing, has expanded vastly. Competitions are held all around the world, and so far nearly 3500 people have solved the Rubik's Cube in one of these competitions The world record is currently set at 7.09 seconds (subject to change) by the Dutch Erik Akkersdijk. My best time being a little over 25 seconds, does not come close to his standards.

There are many variations on the well known 3x3x3 cube, the most obvious being the 2x2x2, 4x4x4 and 5x5x5. Recently Mr. Verdes has produced 6x6x6 and 7x7x7 cubes, which can be purchased at www.v-cubes.com. Besides these, there are also many other puzzles, such as the megaminx (the dodecahedral equivalent of the 3x3x3), pyraminx (the tetrahedral equivalent), Rubik's Clock and much more.

In 1982 Mr. Meffert (the manufacturer of the previously mentioned pyraminx and megaminx) published a booklet with proposals for new puzzles. Most of these where never made. One of these is the Pyraminx Barrel:

The picture shown above is an extract from the booklet, and is a cardboard mockup. There is a community of puzzle-builders, who like to spend their time creating new puzzles. You can visit them at www.twistypuzzles.com/forum. A member of this forum, the.drizzle has created a design for the puzzle, and published the STL files, for anyone to use, you can find it here: http://www.twistypuzzles.com/forum/viewtopic.php?f=9&t=9618

While working prototypes of the design have been made, I wanted to try and build one as a way of spending my summer vacation. ProtoSpace in Utrecht has a wonderful 3D printer, the ZPrinter 450. While being able to print in full-colour, this was not needed. I am going to print one of each part, then create mould using silicone RTV rubber and cast the pieces with polyurethane resin.

I made an appointment at ProtoSpace, and when I arrived there I was warmly welcomed and helped by Joris, the lab-manager. He helped me to set up the files for printing, and also to clean up and treat the parts with superglue. He also showed me some of the things the printer can create, such as chainmails or puppets with your own photograph on them or even full-colour architectural models. This is a video of the printing process:

The ZPrinter works by depositing a thin layer of plaster, which is then bonded together using ink that you would find in any household printer. The excess plaster is simply vacuumed away and recycled. The models are then taken for a swim in epoxy or superglue, to make them more rigid. The picture below shows the parts after their superglue-treatment:

The parts seemed quite smooth, though a closer look revealed a textured surface. Before moulding I will have to sand them using fine (and possibly wet?) sandpaper both on the outside surfaces for cosmetic reasons and on the inside to make the parts slide better. The central part with the screw-hole has some superglue in the hole which I will have to remove before moulding.

Tomorrow I will sand the pieces and make the first half (and possibly the second) of the mould. I hope to have a working prototype by Friday, as I'll be participating in the 2008 Dutch Masters, and I would love to show off what I've made. I will probably pay ProtoSpace another visit on Thursday, to use the plotter to create some stickers for the puzzle.

20-8-2008

Today I visited a nice little hardware store nearby. I took the pieces there and they explained me about different techniques for sanding. I ended up buying some 400 grit wet sandpaper. I also purchased some M3 50mm screws, nuts and washers. These will hold the finished puzzle together.

Wet sanding worked quite well on the pieces and you can really feel the difference. The sanded pieces are now much smoother. Sanding the inside will improve the turning, while sanding the outside will make the puzzle look better. It is important to sand the pieces before moulding, as any imperfections will show up in all casts and require work on all of them.

The pieces turned out to be a little bigger than they should have been, I think this is due to the impregnation process. They will require further sanding, as some pieces that should slide together wouldn't. For the harder to reach area's, I glued some sandpaper onto a spatula. And although I'm not quite done yet, I can see it is working. However having to remove so much material leads to a painful wrist! I will also have to spread some extra superglue over the surface, as it only penetrates 2mm deep into the material. The holes for the screws where also too small so I've had to use a drill to remove the excess material.

The 4th piece I started working on had a major flaw: There was a big bump on it in a vital place. I probably failed to remove it before treating with superglue. As it was in a hard to reach area, trying to sand it away damaged the piece and rendered it useless. I used this piece to test the strength, and it broke on impact with the floor. I'll go to ProtoSpace tomorrow, and redo the piece.

21-8-2008

Today I went to PS again and printed another copy of the faulty part. I also experimented with the laser cutter, I will upload some pictures and maybe another video after the weekend. I have done some more sanding, but these parts really require a lot of work to get right. I guess I didn't remove enough of the excess plaster before the superglue dipping.