This project is the vehicle by which I learned many areas of mechanical engineering through a hands-on approach.

I always wanted to build something large and complicated, but the opportunity never arose. When the original Robot Wars from the Bay Area and Battlebots grew popular, I saw the perfect opportunity.

I never entered a competition, nor did I want to build something destructive or destined for destruction. Having gone to several of the competitions armed with a camera and many rolls of film, I gained insight as to what made all the machines tick - some to success, others not so lucky.

Many competitors had websites through which I absorbed much hands on advice about these machines which were overbuilt radio controlled cars and tanks.

The two features I never saw in any of the robots were timing belt drivetrains and composite frames. Naturally, I was eager to find out why, knowing that timing belts are incredibly efficient, and composites have very high strength/stiffness to weight ratios.

Little did I know what I was getting into. While the above characteristics are still true, timing belt drivetrains are expensive, and several stages are needed for any significant speed reduction. Composites are also very expensive, but on top of that, quality control issues are severe when one is trying to create complex structural shapes in a home garage.

The timing belt drivetrain eventually came to fruition, but after two years and significant investment in equipment and time for a process called vacuum-infused resin, the composite tubular space frame showed little hope of success.

Then one winter break at UC Berkeley, I decided to finish the robot once and for all and abandon composites. Structural aluminum was the solution.

Two long nights with Pro/ENGINEER, and I had the component layout and frame design completed. I ordered all the aluminum according to the CAD specifications, and the material ended up costing less than one eighth of the amount I had spent on composite experimentation.

The student machine shop in the Department of Mechanical Engineering opened up the week following New Year's.

I cut the shapes to their rough lengths and proceeded to transfer the CAD files via MasterCAM to the CNC mill.

Two business weeks later I had 500 holes drilled and tapped, all the frame components milled and ready for assembly, and all the hardware components to mount to the frame.

There was one and only one mistake with this design - this was the first time I had used timing belts, and I was intent on avoiding belt tensioners. Therefore, I calculated the distances required between two pulleys depending on belt length and pitch and pulley diameter. To make sure I wouldn't need a tensioner I rounded all the distances up to the nearest tenth of an inch. One inch wide fiberglass-reinforced timing belts do not even like to stretch one hundredth of an inch. The correction that I had to make was drill the hole pattern offset from the original, so there was a slot instead of a hole to adjust the tension.

With everything ready to be assembled, school started. Unfortunately, I did not have the 14 hours a day free that I had during winter break. With school in progress, I finished the robot on 12 March 2005. After several years of thinking and tinkering and experimenting, this project took three months to complete once I decided to complete it and move on to other projects.

This project is not the type of project I would consider making for a company, but the type I would consider making for a pure learning experience. I became thoroughly acquainted with industrial suppliers, and developed skills such as project management, taking a sketched idea to CAD to CAM to assembly to finished product, project presentation such as this website, and most importantly, I learned that I love doing this.

This project was completed independently and was self-funded - this was not for school.

Copyright ©2007, Tom Brown
   Robot School Trebuchet Submarine Speakers About MechRAND