When she co-founded Dame Products in 2014, Janet had worked as an engineer at a variety of hardware companies, including MakerBot, Quirky, and Nanz. Since then Dame has shipped 80,000 units. They've done so partly through their focus on sex-positive, intimacy-forward products, and partly on bringing the same attention to quality and customer experience that's present elsewhere in the hardware world.
- A PechaKucha that Janet did on what it's like being a female engineer.
- A case study presentation Janet did on how they prototype using 3D printers.
- The Wikipedia page for Brushed DC motor design configurations.
- From a recent issue of The Prepared's newsletter, Deb Chachra on how advances in silicone manufacturing spurred huge changes in consumer products.
- Terri Gross' 1987 interview with Tom Wolfe, and Spencer's takeaway from it.
As we were saying goodbye to Janet, she hinted at the fascinating story behind the engineering of Eva's wings, which tuck behind the labia are used to hold Eva against the clitoris. Janet was extremely generous to describe the whole process here:
One of the biggest engineering challenges for Eva was designing something that would fit in the majority of womens' labia. There's a lot of geometry challenges that went into that. One fairly straightforward part of the challenge, though was dialing in the spring force of the wings. In order to fit a variety of women, and even the same woman in different positions, Eva has flexible wings. Those tuck into the folds of the labia, and stay in place by always being a little bit bent (acting as a spring). I knew that getting the force right was going to be a big part of success or failure. I needed to prototype as close to production as possible, so we did all our testing with machined samples. Given vulval variety, I was expecting a higher force to be better for some women and a lower force for others, so you can imagine my surprise when the data started indicating there might be a fairly universal optimal force. There was a round of testing in polypropylene, and the testers said it didn't hold well enough. I then found out that the silicone wouldn't adhere to polypropylene, so we switched to nylon. The force still wasn't high enough, so we tried ABS. The ABS seemed to be about right, but fatigued too easily - those wings snapped off after a few uses. So we broke out the old Solidworks FEA tools to try to figure out what nylon geometry would give the same spring force as the ABS (and also tried to confirm that the nylon would be able to handle the fatigue). When I tried out that prototype, it seemed about right, but I couldn't be sure if it could be better with a higher force. So I stepped it up a bit again, but then the phrase "popped out" started coming up more and more on feedback surveys - I'd gone too high.
So, going into manufacturing, we were in a pretty good position. I had a sample for the spring force we wanted, and samples from the past rounds that could act as limits on either side. I asked the manufacturer to make a gauge for measuring spring force, and it was ready when I got there. It was basically a clamp and two force meters - it clamped a sample down that gave you readings from both sides. I took measurements off the optimal sample and the two edge cases to set my range. Should've been easy to measure the production injection molded parts and adjust the tooling to match the same force.
However, what I was not ready for was moisture content. When you injection mold nylon, all water that may have been absorbed into the nylon evaporates. Then, over time with exposure to air, the nylon will pick up moisture again. That moisture level actually affects the material properties. Machined parts are made from stock that's been exposed to air, so they're already stable. The first samples we tested had been sitting around for a few days, and they measured a little low. Then we made a 0.1 mm tool change, and suddenly the force readings were waaaaay too high. We chased our tails for a while before we figured out that the force on the day it was molded would be more than double what it would be three weeks later when it evened out (which, obviously, couldn't be figured out in less than three weeks). Now they take measurements for every batch, but the target is 4 N higher than the eventual goal.
Thanks so much to Janet for the detailed explanation, and for joining us!