I was asked by my wife to make a credible steampunk fan. We decided to make it out of brass sheet with a gear train pattern etched onto the front side. Now, I must admit, these gears are purely decorative, which is a bit against my core values. But at least they mesh properly, have a reasonable tooth profile, and they connect in a series that actually make sense. So my obscure artificer's ethics were sufficiently assuaged. :)

There were several challenges in making this project. First, we had to design the decorative pattern, and then transfer that to 28 individual fan blades. They all overlap slightly, so the pattern had to be cut up into overlapping strips. Next, the blades had to be etched and cut. All 28 of them -- and 28 times anything is a lot of work! They needed to have some kind of patina to make the background pop. They each needed 4 very small holes drilled, through which to lace the monofilament which keeps them bound together. And finally, they needed to be sewed together into a single fan. All of this took about a month to get done, most of that dominated by an interminable etching process.

Step 1: Design

The first step in designing the pattern was to simply draw it out in the shape of the unfolded fan, using a cheap bamboo fan as a reference. Because this was destined to be etched into brass, it was drawn in negative. Black areas will not be etched, resulting in a raised, shiny brass area on the final product. White areas will be etched, and thus be the darker color of the patina you chose. Also note the solid arc in the middle of the fan. This was added to help mask the lines of monofilament which bind the fan together.

We measured the shape of a single blade and drew that in Photoshop, being sure to include the pivot hole at the bottom. Making 28 copies, each was rotated by 6 degrees more than the previous one. These were used to create masks which cut up slices of the full design image. The final result was 28 separate blade patterns. In the full design I've included here, I turned up the transparency a bit so you can see how much the design overlaps on successive blades.

The blades were each numbered at their base, where it wouldn't show when stacked together with the other blades. This was very helpful when assembling them in order later. But be sure not to number the final blade, as that one will be visible! (We did not think of that.)

The example set of blade patterns included here is backwards, as anything you're going to etch will get reversed during the transfer process.

Step 2: Etching

I won't cover the basics of etching brass here, as many guides for that have already been written. The key is experimentation, to find the transfer process that works best for you.

This wasn't my first project involving brass etching, but it was by far the largest I had attempted. I decided to try an electrolytic process, as I had heard good things and not having to buy ferric chloride was very appealing. Unfortunately, it was a disaster. The etch quality wasn't great, and I continually blew fuses in my power supply. I'm sure people are telling the truth when they say it's a good method, but I couldn't get it to work.

After struggling with electrolytic for about 10 of the blades, I switched back to ferric chloride. This ended up being a good decision, improved by two discoveries. It turns out ferric chloride is considerably cheaper at real electronics supply houses (as compared to Radio Shack or Fry's). I also came up with a nice way to etch many blades in parallel. I stuck each blade in a separate zip-top bag and poured in some ferric chloride. I placed the bags inside a larger plastic box, to serve as secondary containment in case of a puncture. (Ferric chloride isn't toxic, but it stains like crazy!) This also provided a convenient way to agitate the solution in order to speed etching. I held the box on my lap and jogged my leg while watching TV. It also simplifies disposal. You stuff in paper towels until all the liquid is absorbed, reseal the bag, and throw it away. Kind of genius, if I do say so myself.

Step 3: Cutting and Drilling

So, yeah, the problem with doing 28 of anything is that it takes forever. Cutting the blades after they were etched was another instance of this problem. Originally I planned on doing it with a jeweller’s saw, but that was going to take about half an hour each. Each! Instead, I borrowed a friend's metal shear in the middle of the worst snowstorm Seattle had seen in decades and roughed them out using that. The rounded edges were then finished with the jeweller's saw. You could do it with tin snips, but make sure to get the kind without serrations.

Each blade also needed 4 small holes drilled into it, to allow them to be stitched together later. These were done with a #60 drill bit, to match the thin monofilament I intended to use. When drilling holes that small, you'll really need a drillpress. Even so, I'd recommend buying a couple drillbits as a backup. Breaking ones that small is pretty much inevitable. If you do go to an electronics supply house to buy ferric chloride, you might be able to find packs of drill bits in this size range, as they are sold for drilling out printed circuit boards. I also recommend making a simple jig with the hole pattern drilled into it. This will save a lot of time laying out the positions of the holes onto each blade.

Finally, the pivot hole at the bottom of each blade had to be drilled. I did this by drilling the entire stack at once to save time. That was a mistake. The hole drifted noticeably towards the bottom. If I were doing it again, I would drill them individually.

I made a pivot out of a section of brass rod, which I threaded on either end, and two acorn nuts. At this point the fan could be temporarily assembled. Nice, but the pattern wasn't very visible.

Step 4: Patinization

The blades needed a patina, to distinguish foreground from background. I decided to go with a simple verdigris recipe, with which I'd had luck in the past. There are dozens of options, though, so do some searches and find one that you like.

First, the blades had to be cleaned. Dirt and oils can act as a resist, preventing the patina from forming in areas. After all this work, you don't want a big thumbprint on the final product! Normally I would wash the part in the sink, but I decided that it was easier to just send all 28 of them through the dishwasher.

For the patinization, I placed the blades on a rack in a closed container, with a small bowl of ammonia in the bottom. This will create a nice green patina, but on larger surface areas like this it will be very slow and inconsistent. If you brush some salt water (1 tablespoon salt in 1 cup of water) onto the parts first, this will make the process much faster and more consistent.

Once all the blades had a patina, I lightly sanded them with a fine grit sandpaper. This cleaned the patina off the raised, unetched sections, leaving them a shiny brass. The lower, etched sections of the design weren't touched, and kept their darker patina.

Step 5: Assembly

The final step was to thread monofilament through all the holes. This is what holds the blades into the proper orientation when the fan is open, or else each blade will flop around unconnected to the others. I ran two lines of monofilament to add redundancy in case one breaks. That's not unlikely, after all, with the sharp edges of the drilled holes and the considerable mass of the metal blades. You'll have to use very thin monofilament, too, or else it adds too much thickness between blades and the fan won't open smoothly. I used 4 pound fishing line on mine.

This ended up being a slow, tricky process. All together the blades have some real mass, which makes working with them awkward. I found it was best if I used spring clamps to hold the bits I wasn't working on, or else they'd flop around and make it very hard to get even tension on the lines.

Starting at one end, loosely stitch one blade to the next one. I've (crudely) illustrated the stitching pattern in the attached diagram. Adjust the angle back and forth until the design lines up properly between the two. Pull the monofilament loops snug, so they can't slip. This prevents the blades from shifting relative to each other, which would distort the design. Adjust the clamps as needed and move on to the next blade.

Step 6: Final Product

It ended up being a bit more work than I had expected, but the final result was worth the effort. The motion is smooth and satisfying, and it has a serious heft to it. (The weight does mean you're unlikely to want to carry it around for long periods of time. We had a leather sheath made for this one.) I made this one several years ago, and it has held up well, including being passed around as an example at many convention panels. I definitely don't recommend flicking it open dramatically, though. Doing that is a lot more likely to sever the monofilament holding it together.

The best thing about this project is that the etched design is the real star. The metalwork is really just providing a blank canvas for you to work with. Come up with your own design and have fun with it!