Machine Learning-based Design of Quick Return Slider Crank Mechanism

This project was about 5 months of work in a team of 5 students. The goal of this project was to prepare us for our actual capstone, hence why many students refer to this class as "mini capstone". We had the choice of designing any type of quick return slider crank mechanism that had to meet a few requirements such as, needing to push a 0.5 kg mass, have a 250mm nominal range of motion, have an input speed of 30 rpm, material needs to be aluminum, a quick return ratio of 1.5 times to 2.5 times faster then the forward notion, optimized motor power and as small as possible. 

In the download file below you'll see the course outline that explains the project in depth and of course all the hard work done by myself and my teammates. Thankfully for this class I was paired with a team where everyone pulled their weight and the culture of the team was, if one of us finished our parts in advance, we would offer our time to help a fellow teammate who could benefit from the extra man power.  This culture ultimately led to us being one of two teams to receive an A+ in this class. My main role in this project was the development of the excel tool that had tens of thousands of lines. To view this, in the download folder go to "Excel Tool with macros and DOE" and for the final data we used go to the "ValidOnly" sheet or the first sheet in the Excel. This tool was used to create data that would be feed into a machine learning model, created by us, that would help us determine the optimized design, the tool has a coupe dozen formulas that were built to determine the different stresses on the different links of this mechanism and we then worked backwards to find the best length and cross sectional area based on the materials mechanical properties.