ME 203: Design and Manufacturing
Winter 2024, Stanford University
In Winter 2024, I took the class ME 203: Design and Manufacturing at Stanford. It's a famous class among mechanical engineers, because it gets students really familiar and comfortable with the Product Realization Lab (PRL), the main (amazingly stacked) makerspace on campus. The class is taught by Andy Switky and essentially covers the process of designing, prototyping, and fabricating a physical object. There are various labs to teach students how to conduct various fabrication processes, such as milling, lathing, welding, metal forming, sand casting, and many others. The final project of the class is to build a meaningful product using the design and machining process learned from the class.
For my project, I made a camping stove. I really love both exploring nature and cooking delicious hot food. Usually these two loves are not compatible... unless you have a camping stove. I've always wanted a camp stove and I finally decided to just make one myself! Also this way, I could design and model it after my favorite cooking assistant with an attitude, Calcifer from Howl's Moving Castle.
There are two main components to this project: 1) the thin-walled steel stove and 2) the cast aluminum pan. Starting with the double-walled stove, I decided to make this with mild steel, to be more durable than aluminum and easier to machine, form, and weld than stainless steel. Overall, the 28 gauge steel I used (thinnest available from Alan Steel) provided a strong and relatively lightweight structure for my stove.
All the pieces were cut with the metal laser cutter (from the Lab64 makerspace at Stanford) and then formed and welded together.
With the cut strips of steel, I then had to form them into a cylindrical shape, which I did using the slip roller in the PRL. The stove is slightly conical - the top has a larger radius than the bottom to better mimic the shape of fire - so I attempted to achieve different radii during slip roll, which was a bit challenging. However, since the steel was so thin it was relatively easy to correct any imperfections with hand bending.
After forming, I then moved on to acetylene welding, which was incredibly challenging with steel of such a small thickness. The torch frequently just burned right through the metal and the circular shape made workholding a delicate craft. Eventually with a lot of filler rod, I welded all pieces together and cleaned up the seams so they would fit together well.
For the final touches, I used high-heat resistant spray paint to coat the metal and paint on Calcifer's silly, mischievous face :).
The next step was making the cast aluminum pan through sand casting. This was my first time ever sand casting, and I was really excited to try it. To start, I designed the pan on Solidworks and 3d printed a plastic pattern (positive mold) for the sand to form around. When creating the pattern, I also added a gate and runner, which is just extra cavities that allow the molten metal to flow smoothly into the pan cavity when the aluminum is poured.
My first iteration of the pan pattern did not cast as I expected, as shown on the left, and I attributed this to 1) an insufficient draft angle and 2) not coating the pattern with shellac (a finish that helps it pull from the sand). After creating a new pattern and correcting these two errors, I tried again and was able to successfully mold and cast my pan! The second cast is shown on the right.
After I finished casting, I then finished the part: cut off the gate and runner, cut and tapped a hole for the wooden handle, and polished the pan to be free of blemishes and look cleaner.
Making the wooden handle was the next fun experience because I got to use the wooden lathe. Unlike the metal lathe, the wood lathe requires a lot more freehand work: the tools are hand tools and you can position them wherever as the piece spins (instead of being fixed in place by the machine). This allowed me to slowly craft the perfect handle shape I wanted just by feeling it out instead of precisely changing parameters. It was a totally new approach to precision machining and one I really enjoyed.