Monday, May 19, 2014

Toss Up Poster

Here is the poster we made for 2014 Worlds.  It describes our final Toss Up robot, and shows how the robot evolved from "Bot 1" in November to the final robot that we took to the Worlds in April.






Notice the logo applique created by Ora Lin!


This part of the poster describes the features of our final robot








The transmission that connects six motor drivetrain to the hanging winch.



This part of the poster describes our design process, starting from "Bot 1" that we built in November



Bot 1

We only had two weeks to develop our first competition robot.  With that limitation, we decided to focus on stashing BuckyBalls, since they are not descorable.  We also wanted to be able to go under the barrier and over the bump. 
We used a front lever arm and conveyer belt to lift BuckyBalls  and deposit them into the goals.  A motorized roller with rubber flaps pulled BuckyBalls from the ground into the ball carrier.  The U-shaped chassis left space for the lever arm and made it easy to line up with the goal.  The wheels were attached to legs to gain clearance for crossing the bump.  It was four wheel drive, using direct drive on the back wheels and geared drive on the front. 
Bot 1 performed reliably, but there were points that could be improved.  It sometimes took several seconds for the conveyer flaps to transfer a BuckyBall into the goal.  The robot tipped over easily because the rear wheels were too far forward.








Bot 2

We wanted the robot to deposit BuckyBalls in the goals more easily, so we redesigned to a scissor lift with a gravity-fed intake tray that quickly deposited balls into the goals.  When the lift was raised, a rack gear pushed a slide out the rear of the robot, where it acted as an anti-tipback.  We added Mechanum wheels for improved maneuverability and traction.  The Mecanum wheels were driven by chains that meshed with sprockets attached to the motors.  These sprockets acted like raised wheels to help the robot drive over the bump without needing the high clearance of Bot 1.  This allowed more vertical space inside the chassis while allowing the entire robot to fit underneath the twelve inch barrier.
The new ball carrier was effective, and the Mecanum drive worked well.  Weaknesses were that the scissor lift took several seconds to reach goal height, and the robot crossed the bump very slowly.  The scissor lift swayed from side to side because there was no cross bracing between the two sides of the lift.  It was clear that the lift, which was powered by two motors, was not strong enough to lift the robot off the ground for a hang.


Bot 3 (no photo)

Hoping to make the lift strong enough to hang, we used four 1:1 motors instead of two.  To reduce the tendency to tip over, we substituted steel beams for aluminum in the lift to lower the center of gravity.  We changed to a larger sprocket on the intake in order to spit out balls faster.  To cross the bump more easily, we elevated the chassis two inches.
Despite these well-intentioned “improvements,” Bot 3 was a disaster.  The scissor lift was too weak to handle the power of the four motors, and the aluminum beams bent under the force.  The ball carriage, which was the same one we used in Bot 2, was too wide and got stuck on the scissor lift when the lift went up. The drive wasn’t very powerful, and the resistance of the foam game mats stalled the robot. 








Bot 4

After the slow scissor lifts of bots 2&3, we wanted to bring BuckyBalls up to goal height more quickly.  We decided to change to a six bar linkage lift driven by four motors. A six bar lift uses parallelograms to keep the end of the lift in the same orientation as the chassis of the robot. The six bar lift did prove to be faster.
We also wanted the robot to be able to spit BuckyBalls across the field, so we could pass BuckyBalls from hanging zone to allies in the goal zone.  We thought that we could gear our top roller so that it would spin fast enough to launch bucky balls across the field.  The robot did spit balls out, but not as fast we wanted it to.  It turns out that the motors are internally geared down so that when we geared them up there was too much resistance to achieve a very high speed. 
Another problem was that the Mecanum drive had too much resistance and not enough torque to adequately drive the robot, so the robot stalled very often.  It was very slow:  It took about 5 seconds to cross the field. 









Bot 5

We were satisfied with our lift design, but wanted to find a better way to throw BuckyBalls, add the ability to manipulate Large Balls, improve the drive system, and enable the robot to high hang with a ball.  We designed a new ball carriage made of aluminum and lexan, with a one-piston BuckyBall launcher mounted at the top of the intake.  To intake Large Balls, we added vertical side rollers with 6-tooth sprockets on the bottom and 36-tooth gears wrapped in rubber bands on the top. The robot could score BuckyBalls in the goal, but the rollers meant for large balls were not very effective. The BuckyBall launcher was also disappointing.  It could only fire the top ball of three, and only about 3 feet in front of the robot.
To improve speed and torque and reduce stalling, we switched to a six motor drivetrain with omni wheels, geared 1.5:1, that was more compact and powerful than the Mecanum drive.  After the changes the robot moved around the field more quickly, and had extra torque to push other robots and cross the bump. 
To enable the robot to high hang, we used a pneumatically activated transmission to connect the 6-motor drivetrain to the lift.  The transmission could not handle the load of the 6 drive motors, though.


We also added a spring-loaded hanging hook to the back of the ball carraige.  The hook could knock balls off the barrier, but prevented the robot from crossing under the barrier backwards.  








No comments:

Post a Comment