Rockets
Solid fuel and very small amounts of liquid fuel (at bottom of page)
The full system in work
For a very long time my friend, brother, and I have worked on solid fuel rockets. We have mostly focused on a sucrose kno3 solid fuel in a verity of ratios and with supplemental ingredients such as red iron oxide. We never had great or predictable success, most ended in CATO or rapid unscheduled disassembly seen in the photo above. This mostly stemmed from a lack of an accurate scale to get consistent and correct mixtures. This obviously made finding and correcting issues a challenge. On top of the mixture issues we would often find the rocket failed at the casing which was usually made from paper and tape or 3d printed casings. Despite the issues this project gave me great insight on failure in engineering and how to deal with them, I also learned a lot about implementing an iterative design process. While I would say we never were fully successful the knowledge about solid rockets, nozzle design, chemistry, and the aforementioned skills were far more valuable than getting one to fly further than a few tens of feet.
Below are picture of rockets, rocket building, launches and failures.
3d printing made making nozzles much easier
My math professor and I worked on designing a kerosene/gox engine after I became interested in the subject and asked for his help. We made some progress on the engine, we decided on fuel and oxidizer, engine size, thrust and other simple stuff but ran into some difficulty with the actual engine geometry due to its inter-dependency. For example the chamber pressure dictates the injector design but the injector's performance is influenced by the throat area which in turn is dependent on the overall mass flow rate but the mass flow rate is also tied to the combustion efficiency which is affected by the chamber pressure. We found that we had to essentially select initial values somewhat arbitrarily, then refine them through successive iterations. This was quite difficult with very limited time to work on it and without much background knowledge or a defined end goal other than having designed a liquid engine. Despite the annoying nature of the codependency of the design I found it very fun and an very interesting problem because there was no formula or any real concrete way of going about it in the beginning.