Thursday, March 27, 2008

Robot Suction and Crane: Design and Development

Westminster WiredCats

March 9, 2008

Suction Device

Parts before assembly

Assembled Suction Dome

Key Features and Innovations

· Suction created with off-the-shelf handheld vacuum cleaner (Dirt Devil)

· Dome created from large low-weight kitchen utensil (mixing bowl)

· Sealing ring at rim of dome created from soft, easily deformed tubing (pipe insulation)

· Flexible glue to secure sealing ring (silicone RTV)

Thought Experiments

· Initial design meetings focused on picking up the ball using tongs or grabbers on the sides or underneath/above the ball. The main problem with these methods was that the tongs would have to extend out from the robot, making them vulnerable, heavy and awkward. Thought experiments were done to simplify and shrink a “grabber” of some sort.

· Suction was an early contender, after calculating that it be would only a small vacuum per square inch to hold and pick up the ball, if this vacuum could be applied to many square inches. Although several students and mentors insisted that this method could not succeed, we performed an actual experiment, as opposed to a thought experiment, and found that suction could work as a means of picking up the ball.

Tests and Prototypes

· The first demonstration of the use of vacuum to pick up the ball was done with a
simple, plastic funnel. The microwave antenna attached to the funnel was too shallow to grip the ball, but the funnel was enough to show us that it could work, even with a cup of very small radius.

· The next test was with an old household vacuum cleaner, using the stock hose connected to a microwave radome. The dome (about 18” diam.) turned out to have too large a parabolic profile, such that the ball contacted only the center of the dome, at the suction hose fitting. However, even though the inner diameter of this fitting was only 3”, the vacuum produced enough force to lift the ball from the floor. This experiment proved that the concept was viable.

· The next test used another microwave antenna made of aluminum, with a smaller diameter (14”) and smaller parabolic profile. The vacuum cleaner was replaced with a smaller, battery-powered machine made by Dirt Devil. This proved to have more than enough force to pick up, hold, and vigorously swing the ball around.

· One important feature of this second prototype was the sealing ring that contacted the ball. This ring was made of a soft rubbery urethane, originally intended to be pipe insulation.

Issues and Problems

· Although the Dirt Devil motor was the same size, voltage and speed as the acceptable Fisher-Price motor, switching the two motors created some machining and mechanical challenges.

· Despite the fact that the Dirt Devil motor’s specs indicate that it should run at 12 volts, a 15.6 volt battery actually powered it in their rechargeable device. This higher voltage resulted in a faster speed (est. 30-50% faster), which created more suction. We guessed that the 12-volt speed would be inadequate.

· The next experiment used a 19” diameter dome, which increased the suction area by 47%. A geared transmission was built to increase the output shaft speed by 50%. When tested, this combination provided excellent holding power: it was nearly impossible to get the ball off this larger dome.

· However, after vigorous testing, the Fisher-Price motor transmission rattled itself apart. In addition, it was too heavy to be placed at the end of the crane (see below). After testing the ungeared motor, the results showed that despite the weaker suction, it was still adequate to pick up and hold the ball if the larger dome was used. Thus, the motor-only system replaced the motor-transmission system.

Final Solution

· Stock Dirt Devil case and impeller, with motor changed out to Fisher-Price motor

· Mixing bowl from kitchen supply store

· Lightweight wood attachment ring

· Soft compliant sealing ring made from pipe insulation

Crane

Thought Experiments

· Once the suction dome concept had been confirmed to work, a strong simple mechanism for lifting the dome and ball over the overpass was needed. After sketching out the dimensions, we realized that a simple crane could fit into the robot frame and still extend high enough to be able to hurtle the ball over the overpass. A 4-foot arm with a pivot at the top of the 5-foot robot frame would hold the center of the ball at 9 feet, such that the bottom of the ball would clear the 6 ½ foot overpass by a few inches.

· The first plan to pick up the ball was to have the dome facing forward, and drive the robot up to the ball. However, the ball would roll away as the robot approached and pushed into it. After more thinking and sketching, we realized that the crane could rotate 270° such that the dome faced DOWN. This would trap the ball between the dome and the floor and prevent the ball from rolling away while the robot maneuvered to pick it up.

· We realized early that the weight of the ball, crane arm, and suction dome would create significant torque, considering the four-foot moment arm. Calculations revealed that the van-door motor supplied in the robot kit would have sufficient torque, if it were geared down. We estimated that an 8:1 gear ratio would provide plenty of torque.

Crane in stowed position, trapping the ball, & in the up position

Tests and Prototypes

· The crane arm’s first prototype had a steel framework. This framework could support the weight of the dome and ball, but was too heavy. Aluminum turned out to be more rigid for the same weight, so that less overall material and weight was used.

Issues and Problems

· The crane gears and shafts were too heavy, which required redesign and reducing the weight of the robot frame. To reduce weight in the larger gears and frame, holes were drilled throughout them.

· From the drivers’ perspective, control of the crane was difficult, due to difficulty judging the angle between the frame and crane. To solve this problem, a control system using an angle-sensing potentiometer was attached to the crane and a computer program that would drive the crane to a pre-specified angle was implemented. The drivers actuated this program by means of several large On-Off buttons.

Final Solution

· Simple, one-axis crane with aluminum frame

· Van-door motor drive, geared down 8:1

· Computer control of crane position with manual override

Building the Green Bot

Wired Cats at the Peachtree Regional





























Reaction to the Wired Cats FIRST Regional Tournament

I was surprised about how intense all of the teams seemed to be. Not only did we get to see some cool designs, but also the energy in the atmosphere was exhilarating. It really showed a lot about what FIRST is supposed to be. Even during the awards ceremony everyone stood up and clapped for the winning teams. Overall a really cool experience.

David (9th grade)

Wednesday, March 26, 2008

Reaction to Westminster's Robotics Rookie All Star Win

I really enjoyed watching the robotics competition. My mom and I were sitting behind "1533-Triple Strange Team" with metallic orange capes, who were rooting for the blue alliance. (Which I later learned in match 49, that Westminster was a part of.)

The thing that I found the most interesting was robotic design. I noticed that teams went about making the "body" of their robot different shapes. I thought about all the research that must go into creating a sturdy "body" to support your "grabbing" apparatus. After watching our team and one other team with a triangular prism "body", I noticed that it was the most efficient.

I loved our robot's design the best. Not only was it lime green (which is awesome by the way), but it was the FASTEST, most sturdy looking, tallest robot that I saw. The suction used to push down and put up the blue ball was a great idea.

I was really impressed with our team. It takes a lot of talent from VERY smart people to create a robot that successful.
-Elizabeth (9th grade)