8.6 | Conclusions & Future Work
Results:
All in all, our Trireme was able to successfully row two sets of oars at the same time without any human assistance, proving the concept of automated rowing with synchronized oars being possible. We were able to adjust the speed at which the Trireme rowed by increasing or decreasing the motor’s speed in our code.
Lessons Learned:
Our prototype attempted to implement the dual slider crank mechanism our rowing relied upon. However, this result had way too much friction and was ultimately unusable because of it. We also realized we wanted to scale our design up significantly from our first tiny prototype. We learned a lot about the engineering process: Taking ideas from previous iterations, keeping what works, and reworking what doesn’t. For instance, we had a problem with the shafts moving around inside the bearing and the bearings popping out of place from inside the arms. Both of these were also problems we had with the very first slider-crank project, so we took inspiration from those learnings and used wide washers and shaft collars to better secure the entire assembly together.
Potential Improvements:
The shaft collar on the inside shaft of the driving link ran into the supporting arms. This was just a small oversight, and luckily it did hold up pretty well without it, but the bearing did eventually pop out during our demo (actually the only issue we had at Demo Day, if you can believe it). We’d shape that arm to remove the intersecting area to fix that part.
Additionally, we ran out of time to order all the washers we needed, so in the future we’d likely remove a lot of the 3D-printed spacers in place for the more reliable washers.
By replacing the 3D-printed spacers with washers, we would also reduce the shaft lengths of a few connections too.
Finally, we’d affix the oars to the frame with a more reliable solution. Our current solution was to add screws along the frame to keep the oars in place, but this solution was pretty last minute, and I think given another week or so, we could create a better mount for it to settle into.
The oar orientation was slightly off as well, so we’d like to reprint the oars to be facing the correct direction, add more, and overall try to slim up the design a bit.
Tips for future groups:
If there’s a simple or already-known solution: use it!!! We spent our first prototype trying to reinvent the wheel with a complicated pin slot solution, but all we needed was a rail and linear bearing system (exactly what we used in Project 1).
Take advantage of RMD’s resources. Besides the LEDs, shaft collars, and specialized washers, all of our electronics and hardware were already in the RMD bins!
Acknowledgements:
We want to say a special thank you to Dr. Symmank and the TA’s Min-Geun Park and David Gutierrez Moreno for their feedback during our proposal and check-ins. Thank you to the AHG printers for clutching up in the last 12 hours. Thank you to Cameron’s roommate Landon for giving him replacement motors the night before the car project was due. We couldn’t have made this project and succeeded in this class without the support of these and so many other people.