I got into rock climbing around 2019 and have been getting deeper into the sport ever since. One of the disciplines I got into is trad climbing where one climbs while placing protective gear in cracks and features to catch them in case of a fall. One of the coolest protection devices is the cam which has several moving parts. Being an engineer and climber, I want to combine my hobbies and try to make my own gear.
One note before we get too far - even though I am building and testing safety gear I don't actually plan on actually using this device in my climbing. All of the fall testing I did had backups with real safety gear. With proper testing and quality control (Out of the scope of this project but it doesn't seem too hard to make something compliant with EN 12276) somebody could make a cam that can be considered safe enough to use while climbing.
I started off the design by fighting solidworks to make a parametric logarithmic spiral using the equation driven curve feature and milling a single lobe. This was the hardest part of the design and a good test for the CNC mill so I wanted to de-risk it first. I knew that the geometry was going to make it difficult to hold in the mill so I made a fixture to hold a raw piece of stock for the first operation and a half complete cam lobe for the second operation.
Once I knew that the lobes can be machined, I spent some time on the rest of the design. Most cams are rated for 8-12kN of force with extremely small or large cams falling out of that range so my target was for the homemade cam to break somewhere around that force. My napkin math said that a 1/4" 4140 shaft shears at 13kN and I found some interesting heavy duty spiral snap rings that claim to be good for 500lbf axial load.
For the loop that attaches the cam shaft to a carabiner most manufacturers use a steel cable with swaged or soldered connections. Although I have access to the tools, I would want to create a process and test a few swaged connections before committing to it for the design. Luckily I found an alternative - dyneema cord. This cord can meet the tensile strength of steel but rather than swaging connections you can splice them. There are well-defined instructions from the manufacturer on how to splice dyneema to ensure the connection is as strong as the cord itself that saved me time developing a crimping process.
Finally I had to design the stem and trigger system for the cam. The stem is usually flexible on cams to allow them to bend around rock features without breaking. Older cams used to have rigid stems and in some placements it was possible for them to create a lever arm and snap when a climber falls on them. To allow the stem to flex I used 3D printed TPU. I was also able to 3D print the trigger bar but that was PETG to keep it rigid.
To hold the cam in place, there are springs that push the lobes outward to hold it in place until the cam is weighted. I could not find an exact fit for my lobe design so I decided to make my own springs. The process was pretty simple - I got some music wire and turned a spring jig on my lathe. After a few attempts I was able to control the diameter of the spring. I tried a few different wire thicknesses to get my desired spring constant without plastic deformation of the spring and found a combination that feels similar to my other cams.
Assembing the cam was hard due to the spiral snap rings. In the future I would look for alternatives (possibly threading the ends of the shaft?). Once I completed the cam, I headed out to the crag to test it out. Here is a short video on the process of making the cam and me taking some lead falls (falling from above the cam).
After doing some of my own testing and not breaking the cam, I sent it over to Ryan at HowNot2 and he tested the cam out on his break test machine.
Thanks Ryan! I'm glad I got to see it break.
I'm super happy that the cam broke in the right range and slightly surprised at how well the dyneema held up. It looks like a stress concentration in the axle at the snap ring groove caused the cam to break in that test. I think a fun future project could be building 2-3 cams in parallel and try to make them all pass the EN 12276 testing.