Mechatronics Final Robot - Placed 3rd out of 40 in Final Competition
More detailed information on our Team Page, Located HERE
Our overall design involved a two-tiered laser-cut shape, held together by simple laser-cut “jigsaw-puzzle” supports. Using OnShape, compatible with any operating system, allowed our design to be accessed by multiple teammates simultaneously, as well as to utilize imported functions (including the laser-joint function). Our design, with relevant functions and dimensions, can be found here.
We selected four 3.5” VEX Omni Wheels with square axels to characterize our powertrain, and the corresponding VEX 100-rpm 2-wire motors to drive our wheels (subsequently circumventing the need for shaft-couplings and saving dimensional space), oriented as depicted. We mounted the motors simply using several zip ties, pillow blocks on either side of each wheel, and shaft collars to simply support and constrain wheel-translation. This drivetrain configuration allowed Raisin Bran to operate quickly, utilizing two motors and two wheels when moving in either the x or y direction (forward/reverse and left/right), with minimal rolling resistance and additional motor calibration. This eliminated the need to turn the robot to shoot at a target.
The second tier of our robot was dedicated toward circuitry housing and our ball storage and launching mechanism. Our launch mechanism utilized a flywheel, gravity, and a servo-”gate” that, when opened, allowed our bot to rapid-fire up to five nerf-balls. The positioning of our flywheel, motor mounts, and supports were modeled accurately in OnShape, with final alignment important to reduce the added impedance to our flywheel motor.
Overall Circuitry Schematic
Schematic of our circuitry including all four ultrasonic sensors, three motor drivers running five motors, a voltage divider, and our circuit breaker.
We needed a higher voltage to drive the flywheel at our desired speed with sufficient torque to launch the ball. To do this, we put our batteries in series, creating a 14V power supply. The circuit breaker was used to power on/off the system by interrupting the current flow.
Motor Control - Wheels
Two H-Bridge L293 motor drivers were used to control the robot’s movement. This allowed us to use different voltages to run our Teensy LC microcontroller and motors, as well as change the direction of the wheels. One L293 controlled forward and left motion, and one controlled backward and right motion. For all enable pins, we used pull down resistors to disable the motors from running during initialization and startup.
Because the operating voltage for our ultrasonic sensors and servo is 5V, we used a voltage regulator, which divided the 14V input and to output 5V.
Flywheel Mechanism - Motor Control
A third H-Bridge L293 motor driver was used to control the flywheel motor. Similar to the motors for the wheels, the motor driver enabled us draw currents large enough to power the motor. Unlike the wheels, we did not alter the digital output for the direction pin.
Flywheel Mechanism - Servo
The rotation of the servo was manipulated via the signal input pin, and was connected to a 5V input and ground.
Robot Start (short circuit)
The robot’s program began to run when the designated "run" pin was pulled to an open circuit.
We used 4 ultrasonic sensors (one on each side) to sense distance from the wall. All sensors were connected to a 5V input and ground. Two 10KOhm series resistors were used to reduce the signal input from the Echo pin to the Teensy, as we found that the Teensy pins were not 5V tolerant.
We hoped to make choices throughout the process that would allow the project to be tackled as feasibly and smoothly as possible whilst preserving the integrity of the assignment. This included trying not to “hard-code” when possible to ensure a robot that operated consistently. We chose to focus on a “wall-hugging” strategy, utilizing ultrasonic sensors and their pre-existing arduino libraries to navigate the world whilst avoiding the potential complications of filtration, with the initial goal of successful and accurate navigation and the destruction of at least one tower.
These choices allowed us to focus on the implementation of a functioning robot before deciding which tower to target. Because of navigational ease and the geometry of our robot, we subsequently chose Casterly Rock and were able to consistently raze our target in 38 seconds or fewer. We were then able to take advantage of the reliability of our robot and state machine to successfully target an additional tower, King’s Landing, minutes before the quarter-finals. The formerly-untested code worked, and Raisin Bran placed third overall in the showcase. Key lessons learned include:
(1) Keep your wiring clean and color-coded for easy troubleshooting, from the very beginning!
(2) Working in parallel (with effective communication and timing) saves energy and effort
(3) Make design choices early that save time troubleshooting later, and build in stages!
Final CAD for Our Delivery Drone
The following slides detail a quarterlong project to ideate, refine, design, and test a specific drone delivery mission, from user need to airfoil and drop mechanism design.
VIGA - Butterfly Whistle
Final Project for ME 203 - Design & Manufacturing
Description: A combined-function safety whistle capturing the beauty of the common mime butterfly, VIGA is a useable artistic piece conveying a message about pacifism and personal protection.
Components: I machined the whistle, its press-fit components, and the brazed wing. Not depicted are initial prototypes of whistles to study cut angle, length, and pitch. Final whistle reaches up to 2 dB.
An image of the final product - a fully machined and cast safety whistle with a swing attachment, brass, silver, and cast-bronze.
Working in Onshape
I decided to try working with OnShape's cloud-based CAD platform for this project - I found that it was missing a number of key features but enjoyed the convenience of working on the cloud.
an Onshape Drawing of the assembly
I carved modulan using belt sanders and rotary tools to create my pattern, gate, and runner, and the dimensions of the casting collars for the pattern board.
Early Cast Part
I fine-tuned the pattern for a smoother cast-finish (primarily addressing sand-wash using draft angle, a finer sanding grit, more layers of shellac).
Silver, with a lower melting point than brass and bronze, was the material I used to braze the parts; the cast bronze piece required more time to heat than the thinner-walled whistle body.
I tested the feasibility of specific processes, the primary focus being the method of connecting a cast-piece to a machined piece.
Press-Fitting a Turned Endcap
Having machined the whistle, I also machined two pieces to subsequently press-fit - a thipple for the mouthpiece, and an endcap. Each piece needed to be within a 2 thou tolerance to be properly press-fit.
RAMBO - Bluetooth Robot Car
Rambo Robot Car - Calibrations
Final Project for ENGR 40B
A bluetooth controlled robot car, self built, functionality written in arduino using an adafruit microcontroller and bluetooth controller.
Phone Tilt coded using phone's quarternation values converted into roll, pitch, and yaw and mapped to the servos.
MYNOCH MECHANIC - Jumping Robot
Final Project for Mechanical Systems Design
Self-Actuating Jumping Robot
OH(L)MS - Crawler
Save Our Salamanders - Oh(l)ms Crawler
Midterm Project for Mechanical Systems Design
How do heat and thermal shock affect the properties of a Raku glaze? A reflective experiment.
I built this kiln myself using the following:
Wire Fence, 20 ft 17 gauge High Temperature Wire, 100 ft 14 gauge Wire, 20 sq. ft. Ceramic Fiber Blanket, 9 Cinder Blocks, 4.5 gal. Propane Tank, Weed Burner, Ceramic Kiln Shelf, 3 Ceramic Pillars, Dust Masks, Sunglasses/Glasses, Welding Gloves, Gardening Gloves Utility Knife, Tin Snips, Pliers