Thread by @StevenHCollins

Here are materials for a Junior-level course on mechanical design that we've developed over the past decade. It teaches a design process that combines technical and creative skills, using carefully scoped projects, without fancy equipment:

biomechatronics.stanford.edu/mechanical-sys…

Thread:

The focus of the course is learning how to incorporate both creative and technical skills into design, in a process that combines intuition, ideation, simple mathematical models, computational analysis, prototyping, technical writing, and iteration.

biomechatronics.stanford.edu/sites/g/files/…

We particularly emphasize back-of-the-envelope analysis using simplified models such as free-body diagrams, a hallmark of expert mechanical designers. The most important design decisions are made at this stage, which is underappreciated by beginners.

Course projects are the critical element. They provide the opportunity to practice the intended design skills and the motivation to think hard and struggle, which is needed for learning. To avoid encroaching on project time, we don't have any homework.

biomechatronics.stanford.edu/sites/g/files/…

The first project is a warm up, giving students a chance to get familiar with the course structure, review technical materials from prereqs, and make sure their making hardware is ready (3D printers in this case). The secret goal is to prep for Project 2.

biomechatronics.stanford.edu/sites/g/files/…

We like shake-down projects better than milestones. Any deliverables not rigorously assessed can feel like busywork. They also give the misimpression that design is a simple linear process, like a toy HW problem, rather than the messy, iterative process of solving a mystery.

The second project is the design of a single component for mass efficiency. It is deceptively challenging, with an infinite design space. Simple models of statics and stress are extremely useful. It's variants are my favorite class projects of all time.

biomechatronics.stanford.edu/sites/g/files/…

The third project is another warm-up, this time introducing new technical content: how electric motors and gears work. The objective is to characterize the motor and gearbox, select an operating point, and minimize time to lift a small mass.

biomechatronics.stanford.edu/sites/g/files/…

The fourth project brings it all together in the design of a simple machine that uses a motor, gears and custom components to accomplish a surprising task, like climbing up a door. (This past quarter it had an Inception theme, hence the trippy diagram.)

biomechatronics.stanford.edu/sites/g/files/…

One pitfall we work to avoid is scoping projects too broadly, which forces students into unprincipled hacking. We design projects such that students' existing technical skills have real value in finding a good solution, and that there is enough time to design each aspect well.

Another pitfall we work to avoid is scoping projects too narrowly, such that they become glorified homework problems. We make sure that the design space is sufficiently large and open-ended that students must engage in real-world levels of ideation and iteration.

We try to draw students in from many angles. Each project has: an attention-getting theme; an objective, performance-based grade component, e.g., based on component mass; and several awards. The last is a team project, bringing in social motivations.

These projects don't require fancy facilities.

The pandemic forced us to reimagine how designs could be tested more simply at home. While this introduces some variation in setups, it also makes the projects feel more real. We'll keep this going forward.

The projects *do* require precisely fabricated custom components. In the past we've used laser cutters, but this year we used cheap ($200) 3D printers, sent to each student, and they were great. Here is a battle-tested guide to assembling and tuning them:

drive.google.com/file/d/1PzxoTd…

We've tried truly cheap fabrication techniques, like hot gluing popsicle sticks, but this precludes use of many technical skills we want our students to develop and pushes them to use only intuition, prototyping and iteration. A laser cutter or 3D printer is worth the investment.

For Projects 3 & 4, students also need a gear-motor kit (we use Tamiya 72001, $20), some cheap bearings and shafts, and access to an adjustable power supply ($75). For all projects, they need access to a good scale and some common household items.

p. 4: biomechatronics.stanford.edu/sites/g/files/…

Students also need access to a computer with CAD software, including finite-element analysis. SolidWorks is ideal if your institution has a license, but the most easily accessed is @autodesk Fusion 360, which has a free student version: autodesk.com/education/abou…

We provide starter CAD for some common components, like a capstan that connects to the output shaft of the gearbox and an editable spur gear (with proper involute teeth), so that students can sidestep some of the fussier tasks and focus on the intellectual core of the project.

At the beginning of each project, we assign a few topic readings that teach skills that will be useful in that project: biomechatronics.stanford.edu/mechanical-sys…

We then reinforce them using in-class exercises. This year I made videos of all the lectures and exercises: youtube.com/playlist?list=…

This past year we've also had regular meetings of small 'coaching sessions', each led by a graduate-student instructor. Everyone has loved this; the students form a tight bond, the TAs get real teaching experience, and I get to teach how to teach, which has been awesome.

The course has been well received, garnering teaching awards over the years, but the best signs that we're doing well are: (1) the excellent designs students produce, and (2) those emails, years later, saying how valuable the course was... or even asking for the readings :P.

Many excellent designers and teachers have contributed over the years, including Juanjuan Zhang, Kirby Ann Witte, @MyungheeKim19, Rachel Jackson, Evan Dvorak, Daniel Chan, @g_m_bryan, Isaiah Drummond, Fareeha Safir, Hojung Choi, Amar Hajj-Ahmad, @GuanRongTan, and @p_franks_.

My hope is that posting these materials will inspire you to try something like this in your own course. Feel free also to use any of these materials outright, and let me know if you'd like other scaffolding, like submission forms.

Happy (course) designing out there!