For the Design and Manufacturing (2.007) County-Fair Robotics Competition, I built an autonomous and remote controlled robot, the "Scorpion King". Events on the playing field included hitting a high-striker, spinning a ferris wheel, and pushing a button. With a need for speed and power, I decided to hit the high striker with a spinning flywheel hammer. The most lucrative point-making strategy after that was to spin the ferris wheel for a multiplier.
Video of the final round (I did much better in the seeding round, with the top score of the night! I could not find the recording though.)
"The Scorpion King", a fitting name with its spinning flywheel hammer! This robot was eliminated in the quarter-finals after winning three rounds. The design had the potential to win, I just needed a little more time to do some key tweaks to improve consistency. CONSISTENCY! so important.
The camera cut out right on impact, so annoying, but you can hear the bell ringing afterwards.
Unlike the Brave Little Toaster, the majority of my time spent was designing, machining, and building, while I only left myself a week to code the onboard Arduino. I did not expect to have problems coding leading up to the last week, so I spent most of my time refining the mechanical components.
The first module I built was the hammer. First, I did simple energy calculations to find how much mass I would need at the operating rpm of the DC motor, given the gear ratios I had available.
Equating the potential energy gained by the high striker slug to the energy stored in the flywheel hammer. Simple, but good enough given the requirements and selection of parts/gearing.
I could have included the damping from gearing, delrin-aluminum contact, the lever of the high striker, and the mass that traveled up a shaft towards the bell. Since I only had limited gearing, motors, and time, simplifying things was advantageous and sufficient for the requirements. Knowing the necessary mass, I made sure the motor could accelerate the hammer at a reasonable pace.
The only flaw with this hammer was that it had two striking ends and spun too fast, reducing the chance of properly hitting the high striker. The robot did not approach the high-striker at a high enough speed, causing the hammer to skim the striking platform with one end before hitting it properly with the next, once things were out of whack. With more time, I would have shortened one of the ends and added a counterweight (proportional to R^2) to increase the chances of proper contact (CONSISTENCY is key). One crucial fix I did have time to make was replace the belt drive. The belts provided by the staff were very inconsistent and did not provide the "clutch" effect during spin-up that I needed (to keep the motor from heating up). So, I made my own belt out of a rubber cord and pulleys out of two wheels with turned down grooves (you can see these upgrades in the first picture). Worked great!
The hammer was driven by a brushed DC motor, mounted on a gear box with a pulley that connected a belt to the shaft of the hammer. The delrin-aluminum "bushing" contact had very low friction, and near-perfect alignment allowed it to spin freely.
The second stage in production was to make the chasis. The main structure, where the Arduino, battery, hammer, and ferris wheel spinner would mount was made out of a piece of aluminum sheet metal. To make all of the rivet attachment holes, I used a waterjet. The low temper aluminum (2024-O, annealed "dead soft") bent very easily, unlike the aluminum square tubing used for the drivetrain (6061-T6, I believe).
The last component I built, in a rush, was the ferris wheel spinner. With two high-torque servos in series, it had a lot of force to spin the ferris wheel against the opposition. The gearing was supported on both sides to mitigate misalignment when the wheels were under load. In retrospect, the spinner should have been mounted lower on the robot to allow it to wedge tighter under the ferris wheel, increasing the normal force. Because of size contraints, this was impossible to do without major changes that I did not have time for.
Ferris wheel spinner. Two high-torque servos in series provided a ton of torque (can't find the specs right now!) .
Video of the final round (I did much better in the seeding round, with the top score of the night! I could not find the recording though.)
"The Scorpion King", a fitting name with its spinning flywheel hammer! This robot was eliminated in the quarter-finals after winning three rounds. The design had the potential to win, I just needed a little more time to do some key tweaks to improve consistency. CONSISTENCY! so important.
The camera cut out right on impact, so annoying, but you can hear the bell ringing afterwards.
Unlike the Brave Little Toaster, the majority of my time spent was designing, machining, and building, while I only left myself a week to code the onboard Arduino. I did not expect to have problems coding leading up to the last week, so I spent most of my time refining the mechanical components.
The first module I built was the hammer. First, I did simple energy calculations to find how much mass I would need at the operating rpm of the DC motor, given the gear ratios I had available.
Equating the potential energy gained by the high striker slug to the energy stored in the flywheel hammer. Simple, but good enough given the requirements and selection of parts/gearing.
I could have included the damping from gearing, delrin-aluminum contact, the lever of the high striker, and the mass that traveled up a shaft towards the bell. Since I only had limited gearing, motors, and time, simplifying things was advantageous and sufficient for the requirements. Knowing the necessary mass, I made sure the motor could accelerate the hammer at a reasonable pace.
The only flaw with this hammer was that it had two striking ends and spun too fast, reducing the chance of properly hitting the high striker. The robot did not approach the high-striker at a high enough speed, causing the hammer to skim the striking platform with one end before hitting it properly with the next, once things were out of whack. With more time, I would have shortened one of the ends and added a counterweight (proportional to R^2) to increase the chances of proper contact (CONSISTENCY is key). One crucial fix I did have time to make was replace the belt drive. The belts provided by the staff were very inconsistent and did not provide the "clutch" effect during spin-up that I needed (to keep the motor from heating up). So, I made my own belt out of a rubber cord and pulleys out of two wheels with turned down grooves (you can see these upgrades in the first picture). Worked great!
The hammer was driven by a brushed DC motor, mounted on a gear box with a pulley that connected a belt to the shaft of the hammer. The delrin-aluminum "bushing" contact had very low friction, and near-perfect alignment allowed it to spin freely.
The second stage in production was to make the chasis. The main structure, where the Arduino, battery, hammer, and ferris wheel spinner would mount was made out of a piece of aluminum sheet metal. To make all of the rivet attachment holes, I used a waterjet. The low temper aluminum (2024-O, annealed "dead soft") bent very easily, unlike the aluminum square tubing used for the drivetrain (6061-T6, I believe).
The last component I built, in a rush, was the ferris wheel spinner. With two high-torque servos in series, it had a lot of force to spin the ferris wheel against the opposition. The gearing was supported on both sides to mitigate misalignment when the wheels were under load. In retrospect, the spinner should have been mounted lower on the robot to allow it to wedge tighter under the ferris wheel, increasing the normal force. Because of size contraints, this was impossible to do without major changes that I did not have time for.
Ferris wheel spinner. Two high-torque servos in series provided a ton of torque (can't find the specs right now!) .
Comments
Post a Comment