Computer Generated Design
To test his theory and algorithms, I was asked to design a simple 2-phase bi-stable snap through element. This was something I could easy rapid prototype and test repeatedly.I came up with a decent design made of nylon that functioned for a limited amount of time dependent on the amount of plastic deformation experience during the snap-through action. Subsequently, he would use his program to generate a design a design to serve the same purpose taking into account similar boundary conditions and material properties.
After small design tweaks of mine, the computer generated design did exhibit bi-stability similarly to my original design, but was much more reliable.
My cousin is a professor at Urbana. Knowing my interest in design history and ability to rapid prototype, he recruited me to assist him with an ongoing research topic of his. He proposes that computer algorithms could be used to generate optimal designs based of inputted boundary conditions and desired outcomes and functions. This practice is known as topology optimization.