Thursday, September 19, 2013

Current FSAE Work

My work on the MIT Motorsports, or Formula SAE Team has changed scope quite a bit since the last time I updated. I have worked quite extensively on the Accumulator Enclosure for the electric race car. The enclosure must securely hold the batteries in case of a collision, and be able to maintain its rated strength in the case of an electrical arc or electrical fire. This is a crucial component on the car, and its required that we verify our calculations and that the enclosure meets the prescribed rules. If the team is not able to prove this in a way that is satisfactory to the judges, the car will fail E-scrutineering and be disqualified from competing in the dynamic events. Our enclosure is going to be a kevlar composite layup with a nomex honeycomb core that contains several threaded aluminum inserts to bolt to mounting tabs on the frame. Natalie and I performed several destructive tests on composite beams to determine the material properties of our layups for use in design calculations. A short report of our work can be found here, but here are some picture of our test setup
The test piece is mounted so one can load the AN4 bolt in shear, and assess the strength of the aluminum insert embedded in the layup.

The above 2 pictures are of the test to assess the force required to push the aluminum insert out of the composite layup.
 The test specimens are 2 ply kevlar fabric coated in epoxy resin on either side of a nomex honeycomb core. We tested both 1/4" and 1/2" thick core layers. I also made Aluminum inserts on the manual lathe in the N52 shop in varying diameters to embed in the layup to simulate mounting points. These inserts were etched prior to insertion in the layup. All of these layups were vacuum bagged for at least 8 hours while the epoxy cured and used roughly a 50:50 mix of epoxy resin/hardener to kevlar fabric by weight. Natalie and I made a short video of our layup process of the tests to determine an effective elastic modulus of our layup.


The next project that I am hoping to finish is a small 12 Degree of Freedom (DoF) Hexapod. The frame consists of 12 HXT500 5gram Servos, lasercut acrylic, and the hardware to hold everything together. I plan on entering this in an instructables contest before October, so keep an eye out for it. I am doing a custom circuit board with a high current power supply, a "Master" Microcontroller, and a "slave" mcu that acts as a serial servo controller. I am eventually going to release all the design files under a Creative Commons license when I iron out more of the bugs, but for now here is a sneak peek.

 The CAD files were all done in Solidworks, with many iterations of test parts cut to find the optimal size. I currently have the schematic and about 10% of the board layout done, and the entire mechanical frame constructed.

Hexapod with tiny Fedora. The green PCB's are Pololu Micro Maestro Serial Controllers I was using to test.

Underside of the Hexapod

Close up of the leg mechanism.

I was going for a Jake and Elwood Blues color scheme, and am going to use the Fedora to cover up the finished PCB. I know that the leg joint design could be improved greatly by replacing the cantilevered limbs with better supported ones, but I am currently focusing on ease of assembly and economy. I think that this could be  packaged as a kit for the amateur roboticist, as a lower cost alternative to current hexapod chassis. The only tools required to assemble the final version will be a soldering station, needlenose pliers and a hex key.

Tuesday, April 16, 2013




I have started on a new project: a two-wheel, self-balancing electric skateboard. This project is largely inspired by our recent success at an electric vehicle with Melonkart, coupled with a desire for my own personal vehicle that I can ride around. I am basing this current iteration of Segboard on an Instructable written by XenonJohn. I designed the frame and mechanical systems myself, and am planning on using his code to control the board until I have a better feel for it and can try writing my own. This project is also in dire need of a better name that Segboard (suggestions?).

I am using two 8 in pneumatic tires from that have are belt driven. I also am using two MY1016 DC brushed motors from them as well for propelling the skateboard. The controls are implemented through an arduino pro, taking readings from a 5-axis analog IMU from Sparkfun Electronics and outputting signals to a Sabertooth Motor Controller.

Thursday, September 20, 2012

The post where I update EVERYTHING!

This summer and the end of last year has caused me to be a little lax about posting on the various things I have been working on and what I was building (finals and summer jobs...). I will post full build logs in the next week or so, but this can be a little preview.

So this summer I went to Alaska:
Deploying the Glider from a fishing trawler.
  The Marine Robotics Team traveled to Ketchikan, Alaska for the initial sea trials of our autonomous underwater glider. The DoD was running a STEM outreach program for teachers and middle to high school students where the goal was to construct an AUG out of a Nalgene bottle. We were invited to attend this "SeaGlide" workshop and showcase our glider while running sea trials.

I also helped the (almost?) creation of an electric go-kart

Group Photo of the last day of the workshop
I was a mentor at this summer's Engineering Design Workshop. Local high schoolers are put into groups and have 4 weeks to complete the design and fabrication of an engineering project.

Initial BrainStorming

Learning about the Machine Shop
My group completed the rolling chassis of an electric Go-Kart. They did not manage to successfully have it propel itself, but the parts will be put to good use as the next generation of EV's to terrorize campus come into being this fall.

I built a SegBoard

Testing the frame...

 I built a self balancing electric skateboard as a procrastination technique from studying for finals. I based the electronics off an Instructable by XenonJohn, but did the mechanical system myself. It utilized a 5 DOF IMU unit from Sparkfun, homebuilt A124 battery pack, electric scooter motors and an arduino as a controller.

 It took a little bit of practice to get used to riding it.

 Shifting Gears

I also worked on my first project for the MIT Motorsports team at the end of last semester. I had approximately three days to build a shift indicator from scratch before the final competition. Dario and Erik did the programming, while I did the hardware in a hectic 30 or so hours.The goal was for an embedded microcontroller to read a potentiometer mounted to the shifter and display the current gear on an LCD screen.

It was an interesting experience in rapid prototyping to say the least. We ended up using a mini breadboard for the final version, but then someone at competition plugged it in backwards, and the judges wouldn't allow a breadboard on the car for racing. Next time...

It was a pretty jam packed summer. I also worked at the Lahontan Water Quality Control Board doing some data analysis, but I still fit in some fun
Lake Tahoe-Home, Sweet, Home
Fishing at Blue Lakes with the parents
Diving at Catalina Island
That is a good recap of the summer. More detailed reports will come soon!

But wait, there's more!

a sneak peek of the next round of projects:

Robot Claw
Light-Seeking BristleBot
RC Bomber Plane

Renovations for the Floor Lounge-New brackets and a working Soda Machine
I also will be documenting the stuff for my floor, the Burton Third Bombers at Bomber Blog

That's it for now!

Monday, May 14, 2012

2.007 Grand Prix

The end of a long semester with 2.007 is finally here! We will have our final race this Sunday at the Albany St. parking garage and we will also be having a drag race in the nearby area. This also marks the end of phase 1 in the development of MelonKart. The project has come a long way from a concept in early February to a fully-realized, functional vehicle ready to race. Here is a general wrapup and assessment of the project in my own opinion, and a list of improvements that I would like to make or things that I wish we would have done differently.

Acceleration and Speed: Exceeds Expectations
MelonKart was much lighter than planned, so the acceleration is significantly more punchy than predicted. I am still amazed with the how fast it feels to drive the go-kart, since it is only using one model airplane motor.

Handling: Acceptable BAD
I would like to accurately model the steering geometry in the future with MATLAB or SolidWorks and refine it to make it more linear in nature. I also would like to add a steering ratio similar to a real car into MelonKart. There was also some manufacturing error in the uprights that severely increased the camber of the front wheels, and I believe that the handling would have a substantial improvement if this was corrected. I flipped MelonKart twice in the competition; steering needs improvement.

Robustness: To be determined, but so far so good
We have accidentally blow several of the fuses in the batteries by accelerating too hard and having the software limit in the motor controller set too high and have suffered no ill effects in the electrical components of the vehicle. We have also hauled the Go-Kart around campus to be filmed for 2.007 and for the DeFlorez Competition (We won a hundred bucks!!), as well as demonstrated it at the MAES extravaganza and MelonKart has held together (even with Radu and Actanner taking it for test rides) so the mechanical components seem to have a certain amount of durability to them. It also did not suffer damage from a few crashes.

The final race video

MelonKart at the DeFlorez Design Competition

Future Improvements

Add a second MelonMotor (if they ever come back in stock on HobbyKing)
Switch to a belt drive that has an auto-tension, or redesign the tensioning system for the chain
Add more structural rigidity to the motor mount and support the motor on both sides
Consider a roll bar?
Weatherproof the system
Increase the ground clearance to not drag on a speed bump


I think I took a lot away from this class. First, I now possess half of a functional MelonKart and an extremely messy design notebook covered in chips and cutting fluid. Beyond these physical takeaways, I really feel more confident as a mechanical engineer. I got to collaborate on a project, go through an entire design process, create a working CAD model, source parts for our design, create a manufacturing plan, build it, troubleshoot, and finally walk away with a working thing. I think that out of everything that an engineer does, the design process is what really stands out to me as a quintessential engineering trait. This project made me feel more like an actual engineer than any pset or test ever has. I really enjoyed the team aspect as well, and feel that the sum of our two ideas on the Go-Kart made it a much better project than either of us would. This project gave me more confidence in working with a higher voltage and a much more powerful system than I have ever worked with before. This class also renewed an interest in building things and applying the knowledge taught in my classes, hopefully resulting in many more successful projects!

Friday, April 20, 2012

Adventures in Woodworking and Swapfest

...or playing with scrap wood. This weekend in was the marathon weekend in Boston which meant it was 4 days of freedom from MIT. We ended up taking MelonKart to  SwapFest, an electronics based flea market sponsored by MITERS
There was a ton of cool stuff there. I want to go back in May with a clear plan of a project in mind (...maybe a Tesla coil?) and hunt with a clear purpose so I don't buy everything! Melonkart got lots of attention, especially when Jackie blew a 40 amp fuse trying to go off road in it. We probably should set a max current draw limit in the Kelley to prevent further abuse to the A123 batteries. I also undertook 2 miniature projects: an air conditioner mount for the GRT on my floor and a prototype of  a floating bottle rack.

My mass > Air Conditioner mass

 The Mount rest on a little ledge beneath the window and the actual AC unit goes where I am.

MAGIC!...or moments = 0

 I want to work out all the kinks and maybe mass-produce them as gifts when they are really stable. The bottle is filled with water, of course!