Project Page
Thermal Resistance Circuits -
Heatsink Design
(WIP) Showcasing heatsink efficiency can be calculated through resistance.
Example:
Given
Device Heat Load: 5 W
Device Max Temp: 50 °C
Ambient Temp: 25 °C
Өj-c: 2 °C/W
MG 8617A Thermal Paste:
3 W/m-K (k) ; thermal conductivity
0.33 m-K/W (ρ) ; thermal resistivity
2 mm = 0.002 m (t) ; thickness
1200 mm2 = 0.0012 m2 (A); contact area
Өc-hs = ρ(t) / A = 0.55 °C/W
Өsa = (Tj – Ta / Qdide ) – (Өj-c + Өc-hs )
[
Өsa = 2.45 °C/W ; Max heatsink thermal resistance
ref:
Heat Sink Design Facts & Guidelines for Thermal Analysis – Wakefield Thermal
Load Cell
A load cell is a force transducer, this means for a given force it can read and generate a proportional charge allowing us to measure said force. Designed simple model for load cell using strain gauges in Wheatstone bridge configuration. Explored half and full Wheatstone configurations.
Rapid Prototype Robotics Projects
The Robotic Arm:
The task was to build a robot arm that could write our initials autonomously.
My team and I managed to build an arm that could trace initials (using a laser pointer) from given coordinates. Though tracing the letters in a straight line/ grid pattern proved difficult.
DH parameter convention was used to sort out the inverse kinematics of each joint. I believe our error comes from our DH parameters and that given another week of prototyping/testing we could've fixed this issue.
I worked on design, assembly, and assisted with coding on this project.
Robot Arm.mp4Balancing Robot:
The task was to build a robot that actively balances using an IMU. The robot must be able to freely rotate/fall +/- 90 degrees around one axis once control is turned off.
My team and I decided to build an inverted pendulum on a cart. This featured a linear belt system. An encoder and lidar sensor were used to track position and derive velocity while an IMU located on the pendulum was used to track angular acceleration. Our control scheme included a Kalman filter which would create accurate predictions for our states (position, velocity, angle, and angular acceleration) as well as a Linear Quadratic Regulator to control said states.
At the end of two weeks we built a robot that attempted to balance but found that our motor couldn't move our cart fast enough. (Another solution to a stronger motor would've been changing the weight/size of our pendulum)
I worked on design, fabrication and assembly on this project.
Balance.mp4HRI Robot
The task was to build a robot that could accurately receive and pass a tennis ball to at least two humans space 90 degrees apart. The robot must determine which human, from given cues, to interact with.
We decided to use a gear and spring system to launch the tennis ball. Our gear system pulls back a rack attached to a spring, once disengaged the spring shoots the rack and pushed the ball out of the launcher. In terms of HRI our robot uses a camera, microphone, and lidar sensor. The robot works by scanning radially until a face is centered in its frame. Then once given a vocal que it begins the process of launching the tennis ball by springing back. Once the tennis ball is launched and the user returns the ball into the top holder, a lidar sensor senses that the robot has been reloaded and starts the process again.
We ran into a lot of trouble with getting enough torque to pull our springs. Additionally we had to test different disengaging methods to get it to work. We wanted to be creative and use springs, but they proved to be unreliable/deformed after a few tests. Additionally I was tasked with designing the base and body, but it proved difficult since we were iterating and testing the launcher (which rests on top of the base/ inside the body). This meant I had to go back and redesign parts so that they fit, and at the end had to improvise for the body.
I worked on the design, fabrication, and assembly of the base, body, and the gear system tilting our launcher.
HRI.mp4Brachiation Robot (Final Project)
The task was to build a robot that can traverse a brachiation/ monkey-bar course. The robot should be able to move from rung to run autonomously.
I had the idea of trying to solve the problem statically by splitting the robots movement into steps. The design features two arms that extend forward and back using a rack and pinion system. Attached to each arm are two hooks that close together, as well as an ultrasonic sensor. The body of the robot is where we stored the electronics as well as an IMU, and a third hook and ultrasonic sensor pair.
The robot functioned by having all three hooks attached at once. It initiated by detaching one hook arm and extending that arm to the next rung. Once the ultrasonic sensor on the arm detects an object (the rung), the hooks on that arm would snap closed. Next the middle hook supporting the body would open, and using the racks on either arm, the body will traverse until it reaches the next rung. Once the middle ultrasonic sensor senses the rung above, the middle hook closes on the rung. Finally the hooks on the back arm will open, and the back arm will be driven forward in the same process. With this design there are always at least 2 points of contact between the robot and the brachiation course. To prevent as much swing as possible, and to keep the arms aligned, I designed a rail that the arms slide in and out of. I incorporated ball rollers at the bottom of the rail that the arms would slide over for supported but frictionless movement as well as two guiding gears (the same size as the pinion gear) that would help align the arms.
The main issue came from friction in the arms caused by misalignment of the guiding gears. I had designed, fabricated, and assembled the entire robot except for the body and hooks. When it came to putting the different parts together though, the holes in the body didn't align with the rail or the arm. The laser-cutter we were using for fabrication had stopped working as well and so we weren't able to cut another box. The solution was to remove the guiding gears, but we found that the arms wouldn't stay aligned.
If I were to redo this project I would have communicated with my team better. The error came from a small misalignment in parts made by different members. The robot was able to move and grab onto the next bar with assistance. There are six gears that need to be aligned (two driven gears and four guiding gears on either side). With all of the gears aligned I believe the arm would also have been steady enough for the robot to work autonomously.
I worked on the design, fabrication, and assembly of this project.
Briach.mp4Braille Clock
One week robotics project in which we were tasked with developing an analog clock using a raspberry pi and a servo motor. My team and I decided to take a nonstandard approach and tried to design a clock that could tell time in braille.
Each motor corresponds to a digit, the first two being for hours and the second two being for minutes. Thus the workings of the clock are simple, I developed a script in python that grabs the current time in whatever given time-zone and separates it into four different variables. Each motor is assigned a variable and moves to the state matching that variable. These states match the numbers in braille which are located on a 3D printed piece attached to each motor. These pieces feature braille cells/ numbers for 0-9. Once the current time is read and the time variables are updated the motor moves to the corresponding state matching the braille cell for the current time.
The video shows a sped-up version of my code in which I change the motors indicating each minute to indicate seconds.
Clock Video.mov2020 James Dyson Award Challenge
During the height of the COVID-19 pandemic, when my original 2020 internship and summer plans were cancelled, a friend and I decided to join the Dyson Design Award Challenge as a team. Run by the James Dyson Foundation, the international design award challenge occurs every year. Though we began the project just a few months before the deadline and were working remotely with no physical tools or materials, we were able to successfully complete a prototype design.
The focus of our design was assistive technology for individuals with Cerebral Palsy. After interviewing a focus group of several people who live with the disorder, my partner and I decided to design a comfortable in-ear headphone that doesn't fall out and is less irritating to the skin than standard earpieces. Durability, a snug fit, and comfort were the main problems that our focus group noted when discussing the practical use of headphones by individuals with Cerebral Palsy. To address these issues, our design employs elements of athletic earbuds, such as durability and a snug fit in the ear, while making use of a more comfortable material that will not cause irritation.
Though our design did not win the Dyson Design competition, the process of designing a product (not to mention doing so remotely while in the midst of a global pandemic) to directly address a problem experienced by disabled individuals was a valuable and rewarding experience. My partner and I believe that our in-ear headphone design, if fabricated, could provide a real-world solution to an everyday difficulty faced by individuals with Cerebral Palsy. See our design below.
By: Cindy Siu, Jared Jaramillo, and Mischael Anilus
Acrylic Structures
Project completed for ME 42: Machine Design course at Tufts University. For this class project, I designed two structures made from acrylic that should deflect 0.2" when 5 lbf is applied vertically. This would result in a spring constant of k = 25 lbf/in. We assumed linear yielding under low stress levels, as most polymers (such as acrylic) have non elastic performance.
Shown below is one of my structures, designed in SolidWorks. I used Finite Element Analysis (FEA), through SolidWorks Simulation, to create a design that would closely deflect under our specific conditions. The actual laser cut acrylic piece is shown in the Instron machine, in addition to a chart showing its deformation.
Voltage Regulator
Class project for ME 93: Electronics for Mechanical Engineers course at Tufts University. Designed PCB for a Voltage Regulator.
Specifications:
- Accepts power from a 2.1 x 5.5 mm plug from a 12 V wall adapter
- Emits 12 V, 5 V, and 3.3 V simultaneously
H-Bridge
Project for ME 93: Electronics for Mechanical Engineers course at Tufts University. Designed PCB for a H-bridge.
Specifications:
- Accepts power from a 2.1 x 5.5 mm plug from a 12 V wall adapter
- Screw terminals for power and control lines
- LED that lights up when motor power is available
- Makes a DC motor spin in both directions
Bio-mimic Toy Design
Project for EN 01: Applications in Engineering Foundations of Design course at Tufts University. Toy functioned as a 3D maze in which the user would guide a ball from top to bottom. Each cylinder piece could be connected together and rotated, while more cylinders could be connected to make the maze longer/ more difficult. The design was based off of anthill structures.
High School Capstone
High school Capstone project at the Academy of Information Technology & Engineering, in which we were told to invent a product that solves a problem. My partner and I decided to focus on the filtration of contaminated water, as well as clean water storage. Our final design was a water barrel with an inner filtration mechanism. This would allow for the filtration of acidic rain or other contaminated water that is poured into the barrel. Our product would allow for filtration of contaminated water, in addition to a means of storing clean water. See our design presentation below.