In my extended family, there’s been a longstanding tradition that we have an ornament exchange which takes place at Thanksgiving each year. Although the ornaments that come up in the exchange each year range from the mundane to the sublime, there are a lot of crafty people in the family, which means that store-bought ornaments tend to get a bit outclassed. Over the years, I’ve gone mostly with the store-bought stuff anyway, with a couple of ill-advised attempts at handmade ornaments thrown in for good measure as well. I actually kind of enjoy making ornaments myself, although to be honest, when I try to make stuff by hand, usually the results don’t quite add up to what I originally had in mind. To give you some idea of this, I tend to consider a craft project to be successful if nothing got set on fire or maimed during the process.
It is because of this tendency that last year, I decided to take a different approach to the problem. As outlined in the post that I wrote about last year’s ornaments, I found a place over on Capitol Hill in Seattle called Metrix Create:Space that has all sorts of cool toys like 3D printers and laser cutters. Although I’m pretty sure the 3D printers are a bit out of my league for the time being (at least until I manage to get some idea what the heck I’m doing when it comes to 3D modeling) I found that through a combination of open source tools and following instructions found in several FAQs posted on the Internet, creating a design that would work with the laser cutter proved to be a lot less daunting task than it might sound. The ornaments which resulted from this project turned out very nice, and were very well received. The only real problem I had with these ornaments was the fact that they ended up being far more expensive than I had originally planned on. In fact, I think I spent about twice as much on the project as I had originally budgeted, and given the costs of laser cutting, I had planned on quite a bit.
As a result of this, although the tools for this year’s ornaments were pretty much the same as last years, the approach taken to the project was quite different, as was the end result. After the jump, you’ll find a description of the process used to create my ornaments for this year’s family Thanksgiving ornament exchange.
With last year’s ornaments being so well received, this year’s ornaments were going to have a tough act to follow, but at the same time I also needed to put a priority on keeping the costs down as well. The first step in this would be to switch from acrylic to wood, which not only costs less than the plastic does in material costs, but also cuts more quickly on the laser and doesn’t come with the “stink” surcharge that the shop has for acrylic laser cutting (can’t say I blame them on that one.) At the same time, instead of cutting out 2 or 3 big complex parts, I decided a better approach would be to use a number of less intricate parts that could be cut more quickly, and use less time on the laser. In particular, I found this design for an Old School Space Rocket which had been posted by a member of Thingiverse, a site which people use to share pieces they have designed to use with laser cutters, 3D printers, CNC routers and other such tools. If you’ve ever had one of those wooden dinosaur skeleton puzzles as a kid (or still have one, they still seem to be pretty common) the principle is much the same. You have a number of different pieces to the puzzle (in this case, the rocket) which press fit together in a certain order to make the desired structure, without the need for any glue or fasteners. In a way, it’s a (relatively simple) puzzle as much as it is a decorative piece, and aside from the fact that the thing is 15 inches tall (or 3o inches tall if you’d prefer the larger version) it looked like a good candidate to turn into an ornament.
Of course, even with the precision afforded by a laser cutter, scaling something that complex and with that many pieces down to Christmas ornament size was going to be impractical, so the first order of business would be to significantly simplify the design. Digging up some ancient knowledge acquired from a basic drafting class that I took in high school (for the most part this was still a pencil-and-paper affair, AutoCAD was a different class I never got a chance to take), I sketched up the stuff you see above. Obviously this wasn’t intended to be done with any sort of precision, mostly I was doing this as an exercise to determine the proportions, and even then, I didn’t have a ruler available when I was doing this, so I’m sure it’s no more than an educated guess really.
With a vague idea of what I wanted this to look like, the next step was to fire up Inkscape to start putting things together. Fortunately, Inkscape is well-suited to creating the shapes that are required for the center rings. I made heavy use of the snap-to-grid feature (on my initial prototype, I had my gridlines spaced 1/16″ apart, the smaller gridlines don’t show at this scale) to make sure the sizes and positions were correct. The notches that the pieces would be inserted into were added by placing them in the proper spot at the top of the circle and using Inkscape’s difference tool to “subtract” them from the shape, then rotating 120 degrees (or 60 degrees for the 6-notch pieces) and repeating the process. I also used the grid to set the positions of the notches on the side pieces in order to ensure that this would come together properly, then used the curve tool to build the shape around them, differencing the notches out of the finished shape. At this time, I was working in inches, which ultimately ended up causing some problems as the plywood pieces I was working with were measured in millimeters (3mm, where 1/8″ is closer to 3.125mm.) I also created rings in quite a few different sizes so I could play around with the proportions of the final piece a bit.
After creating some of my initial designs, I decided I might save myself a few bucks on prototyping by cutting out pieces in Foamcore. A few minutes of messing around with an X-Acto knife and mostly making a mess of things put that notion to rest. At least I didn’t spend a fortune on a cutting board…
After figuring out all the specifics, the next step was to put together a “kit” that could be run on the laser cutter to make a prototype. While it would be easy to just throw the pieces together, in this case I would also need to come up with a layout that would allow the maximum number of pieces to be made off of a 12×18″ sheet of birch plywood, so an effort was made to get all the pieces down into the smallest possible footprint.
With layout in hand, the next step was to make a trip over to Metrix to cut out a copy of the layout above and try it out. After cutting, it became quickly apparent that there were a number of problems with this version. Aside from the fact that the top of the rocket doesn’t work too well in this version (something I had intended to fix later,) the proportions just don’t work out well for a Christmas ornament on this version. But an even bigger problem was the fact that although I could get the pieces together without too much trouble, they just simply didn’t want to stay together, a result of the notches being too wide. This was, as noted above, caused by using inches on a task that was really much more suited to metric units. There were also some problems with the cutter not properly cutting out a couple of the pieces which ultimately required a couple of the fins to need to be recut, but that had nothing to do with the design. On top of all that, this particular version required just a bit over four minutes of laser time to cut out, which was going to end up being more than I wanted. I will probably go back and rework this version somewhere along the line, but for the time being it was back to the drawing board. This particular photo makes this look much shorter than it actually is, in reality the rings are spaced about an inch apart from each other. In this particular design, I figured I should be able to get roughly enough pieces for eight rockets off of a 12×18″ sheet of plywood.
Starting again from basically square one, I decided that the next (and ultimately final) version of this would be a fair bit smaller. Once again, I started out by making the rings in a number of different sizes, and using those to determine proportion again, and also to determine how close to each other I could get the notches without pieces interfering with each other. Another big change I made on this version was to use an asymmetrical layout of ring sizes to improve the proportions. I also altered the shape of the fins somewhat, to make it a bit easier for the finished product to stand upright. Perhaps most importantly, I used smaller notches in this version (2.9 millimeters, as opposed to the 3.125 mm in the first version) so the pieces would fit together more tightly. In the end, I think they ended up being a little too tight on this one, but at least this version is a lot less likely to fall apart when assembled. I could probably mess around with that a bit more too, but in the end, I don’t think it would make much difference.
Here are the parts for the second version. As you can see, they take up a lot less space than the first version.
With the second prototype ready to cut, I took another trip out to Metrix. After a slight mix-up that resulted in an accidental second copy of the first version being made, the parts for the second version were cut out and assembled for the first time. the result can be seen above. Aside from being much better proportioned, a much better size for an ornament and actually staying together, this version also significantly reduced the cutting time needed from the version 1, down from over four minutes to just under three. given the high price of laser cutting time, this is pretty significant.
For comparison purposes, here is version 2 shown next to the significantly larger version 1.
Just for curiosity’s sake, I did also have one of these cut out in acrylic, but quickly abandoned this idea after finding out that the notches were incorrectly sized when cut out in plastic (way too loose) and had one of the side pieces break during my attempt to assemble it. Between this and the inadvertent second copy of version 1, I had a decent pile of extra pieces sitting around when everything was said and done.
Finally happy (mostly) with the design, the next step was to prepare for production. From this point, no changes were made besides the addition of a place for the ornament hook to be attached, as you can see above. In some initial messing around with the design before adding the ornament hook to it, I found that I could easily get the parts for 12 complete kits off a sheet of plywood using the parts layout for a single kit above. When I split up the individual parts to their own separate sections (as seen here) I was easily able to get the parts for 14 kits (plus some spares) on a sheet. With the ornament hangers added, it became a bit trickier, but ultimately I was able to get 14 kits off a single sheet. I think I’d be hard-pressed to get much more than that though.
With the final design sorted out, the next step was production. This photo was taken toward the end of the cutting process, which ended up being a little over 42 minutes total (which was charged at the production rate which is cheaper than the prototype rate, but includes an initial setup fee.) Fortunately, the initial problems with the laser not cutting pieces properly was resolved with this batch, and every single piece was cut out on the first try with no issue.
Here’s what some of the pieces looked like after cutting while they were being collected from the machine. The circles at the center of each of the rings just popped right out with no hassle.
And here’s what the sheet of plywood looked like after all the pieces were cut out of it. It’s actually kind of a cool piece in its own right so I kept it, although I’m not sure exactly what I’m supposed to do with it. You can see that especially on the pieces with the ornament hangers some of the tolerances were closer than I had expected, but everything worked out pretty well in the end.
Although I was reasonably certain that most people would be able to figure out how to put these together without too much trouble, I figured it would probably be useful to have some sort of instructions included just in case. As a finishing touch, I adapted the initial drawing of the second version into a quick illustration of all the pieces to show what goes where. I would have liked to do a 3D model as well, but my brief attempts at using Google Sketchup for the purpose proved largely fruitless as my attempts to import the SVG file with a Sketchup plugin did little more than throw cryptic errors.
Even though I didn’t spend anywhere near as much on this year’s ornaments as I did on last year’s version, I didn’t want to get too fancy with ribbons, bows and bags, so I just opted for simple Ziploc bags (which, if you think about it, is probably more in line with the do-it-yourself feel of the whole project anyway.) This was the form in which the ornaments were handed out to the rest of the family.
All in all, even though I can think of a few things that I might do to improve on this project the next time I make something like this, I think this turned out pretty well. I haven’t heard anything back from any of the recipients of the ornaments on any difficulties with assembly or any problems with stuff falling apart, so hopefully that’s a good sign. Once I get a few things sorted out I intend to put the SVG files for this up on Thingiverse so anyone out there who has access to a laser cutter can make these for themselves (I’ll be sure to add this here once it’s available.) I’ll probably also go back and revise the larger first version into something that probably still won’t work too well as an ornament, but should still be OK to use for something to stick on the desk, or possibly throw a few pencils in. In the meantime, now I’ve got to figure out what I’m going to make next year. Fortunately I’ve got some time to figure this out…