If you want to climb ice, you're going to need a pair of ice axes. The state-of-the-art is this carbon pair from Black Diamond, which will run you a cool \$340 per axe. Or, you could just make your own, like this guy did, saving himself over 500 grams in the process.

Nils Cuinat-Guerraz is a composite materials researcher at Canada's McGill University. He also doesn't have \$680 to drop on a pair of carbon-fiber axes, so when he got into ice climbing, he set out to build his own.

Left to right: Shallow/medium/steep pick angles

The first step for DIYing ice axes is to decide on a shape. The geometry of an ice axe is hugely important, completely changing the way the tool feels and works. A walking axe (above, left), where the pick is only going to be used for self-arrest, has a very shallow pick angle (the angle the pick sits at relative to the handle). A mixed-climbing axe, like the BD Fusion (above, right), has a steep pick angle, with a steep ice tool sitting somewhere in the middle.

When an outdoors company designs an ice axe, working out the geometry — pick angle, handle, tool length, shaft curve and all — is one of the hardest parts. According to Bill Belcourt, the head of R&D at Black Diamond, the team starts out with a set of goals for a project, like "make an axe that's under a pound but strong enough to climb ice," and then tries out 15-20 concept designs in the lab and in the field, to see what works.

DIY axe next to its commerical cousin, the Petzl Quark

Without being able to prototype different shapes, Nils stuck to a tried-and-tested design, in this case taking the shape from Petzl's successful Quark tool, and copying it wholesale in carbon fiber. This also let him use standard Petzl picks and finger-rests, instead of trying to mill his own.

Once he'd settled on a shape, it was on to a computer to test out the strength. Finite element modeling (FEM) is basically computerized stress testing, allowing engineers to apply different loads and forces to a model, see what breaks as a result, and then (hopefully) fixing the problem.

Foam mold overlaid with a carbon-fiber sock

Computer testing done, and with a basic idea of construction in hand, Nils moved onto the first prototype. To make his carbon-fiber axe, he was using a process called Vacuum-Assist Resin Transfer Molding (VARTM). Basically, it involves making a rigid mold, in this case out of foam, with the end shape you want to achieve. Then, carbon fiber is wrapped around the outside. In this case, Nils used three layers of a carbon-fiber 'sock', each with a different weave patterns to achieve better overall strength across multiple directions.

Then, the carbon-wrapped mold is covered with resin, and placed in a vacuum bag, which uses pressure to impregnate the carbon-fiber with the resin, resulting in the super-light, super-rigid carbon-fiber structure.

The process is relatively cheap and easy — you don't need much more machinery than a vacuum bag and the carbon fiber itself (the expensive part). But it's a slow process, and one that can lead to imperfections in the final product. It's also difficult to make intricate shapes or designs.

So, when making carbon-fiber tools for customers, Black Diamond uses Resin-Transfer Molding, a similar idea with different execution. Carbon-fiber is packed into a steel mold, and then an air bladder inserted into the middle and inflated, which pushes the carbon-fiber (and resin) into the steel mold with a high level of precision. It's a much more accurate process — BD's ice tools are laid-up with 54 different pieces of carbon-fiber, as opposed to Nils' three — but involves much more expensive and complicated machinery.

Another big difference between home DIYers and real companies is the level of testing. Belcourt told me Black Diamond goes through hundreds of axes in lab stress tests, plus any ice axe sold in shops is certified by the UIAA, which does its own series of tests.

All credit to Nils, he tried to replicate the main strength-test on an ice-axe shaft: fixing the ends, applying a weight in the middle, and trying to break it. Commercial axes need to withstand 2kN to pass the basic rating, and 4kN for a technical tool. Nils pull-tested his first prototype at about 1.4kN (although he thinks his final models would probably go to around 3).

Once Nils had two carbon-fiber shafts, he just had to drill holes in the head for a pick, wrap the handle with grip tape, and he had something you can actually climb on. I've taken them up a few of the (rapidly-melting) waterfalls in Quebec, and they're surprisingly OK to climb on.