Robert McMullen’s Fuse Machine
Robert McMullen, a sophomore at Olin, has a passion for machine design and pyrotechnics. He decided to combine these in a passionate pursuit to build a fuse making machine.
The goal of the fuse machine was to make visco fuse (a.k.a. safety fuse) as fast as it is made commercially while ensuring a fuse of the highest quality. Visco fuse was invented in 1831 by William Bickford as a safer fuse for use in blasting. It’s improvement over earlier fuse is its constant burn rate and reliability. Once it is ignited, visco fuse burns at a constant rate and is difficult to extinguish, allowing the time of the blast to be known, and not causing delayed explosions. In modern times, it is used almost exclusively in fireworks, but the machines that make it are still quite primitive. This machine attempts to modernize the design and make it more compact, robust, and better looking, while ensuring it is entirely non-sparking for safety. Robert designed the machine at Olin using SolidWorks using a few pictures for inspiration. In the spring he machined it in the Olin machine shop, and then tested it.
Front view of the fuse machine.
The machine works by feeding black powder through the blue funnel into the top brass die (magnified view), where 12 threads wrap around it as it is falling through the small gap between the two pieces of brass. The fuse then travels through the top plate down to the bottom die, where 6 threads wrap it in the opposite direction to prevent it from unraveling. This travels through the bottom plate into a nitrocellulose lacquer bath (hiding behind the take-up reel), and out onto the take-up reel. The black and white threads (tracers) at the top travel through the black powder to help pull it down into the fuse, preventing any empty pockets. Not shown is a pulley clamped about 10 feet away that the fuse can go to on its way from the lacquer to the reel in order for it to dry. The lacquer layer or the outside helps hold everything together and gives the fuse its waterproof qualities. The finished fuse is 1/8” in diameter.
The top die where the falling black powder (in this case sugar) is wrapped up in string. The inset photo was taken while the machine was running.
After a few trial runs, the machine was producing fuse similar to commercial visco fuse, except faster burning (caused by the burning speed of the black powder used). To test the quality of the fuse, Robert conducted an unrestrained burn test and underwater burn test. Quality visco fuse should be self-propelling and burn underwater, which shows that the black powder core is powerful and consistent. The machine itself is much more compact than any commercial machines with that can be seen online. Its biggest weakness is the take-up reel, which is fairly small and limits the machine’s capacity. It is made of mainly aluminum and brass, both of which are non-sparking, and the motor is properly shielded. As far as atheistics are concerned, you be the judge.
Weather Balloon
watch?v=rvfdeCIpSC8Projects are an integral part of the freshman curricum. The weather ballon is an example of a freshman project. Juliana Nazare, Noam Rubin, Cypress Frankenfeld, and Dan Kearney designed and built a weather balloon system to measure the high-altitude atmosphere for our final Real World Measurements project.
The apparatus consisits of a 20-foot- diameter meterological ballon, the parachute, and the payload. The payload is a box made from GreatStuff expanding foam and blue foam. Inside the box was a circuit board and DAQ, an RC logger, a Cannon PowerShot camera, a SPOT GPS, a phone with Boost Mobile GPS, external batteries and multiple foot warmer to ensure that the batteries in each device functions in extremely low outside temperature. The circuit was logged pressure and temperature data. The apparatus was tested extensively before launch.
The balloon was launched in Pittsfield Massachusett.According to predictions, the balloon was predicted to land near Lowell Massachusetts. However due to the weather conditions that day, the ballon was landed in Lunenburg, Massacusetts. The balloon was found in a small forest loudged in a 100ft tree. The ballon was able to be retreived with minimal damage to the payload.
Hovercraft
The hovercraft is an example of the many independent projects Olin students participate it. Silas Hughes and Joe Gibson are working towards producing a full size, ride-able hovercraft. However, due to space constraints, they decided to construct this half-scale prototype.
The hovercraft can be broken down into three major components: the propulsion source and duct, the skirt, and the base. Although some hovercrafts utilize two propulsion sources. This hovercraft uses a Eflite Power 160 Brushess motor, typically used for large model airplanes. The motor and the propellor are secured in the propeller duct. This helps direct the air. More importantly, the motor protect the propeller blades from obstruction. Much of the air flow from the propellor is sent backwards for thrust. The rest of air flow is channelled down into the skirt. The particular skirt used for this design is called a bag skirt. Air is pumped into is and it inflates like an intertube or a doughnut. Small holes line the inside parameter of the "doughnut" and release the air into the center of the ring or the "doughnut hole". That air escapes from underneath the bag and produces the thin film of air that the hovercraft floats on. It is this film of air that makes the hovercraft hover.
As typical of many Olin projects, the hovercraft has gone through several stages of iteration. The students initially wanted to build the hovercraft from insulating foam. They decided it would be too dangerous to use the motor with a flimsy material. Instead, they built the hovercraft out of wood. This change in material produced a much heavier hovercraft than initial predicted. The hovercraft was too heavy to fully function. From this experience, the students learned a lot and have plans to build a full scale hovercraft as time permits.
“Snake Norris”- A CompArch Project by The French Invaders
This was a Computer Architecture project by the French Invaders, Jeremie Tardivon and Gabriel Villenave, two French exchange students. Using what they learned about Verilog in CompArch, they decided to pay tribute to an old fashioned video game, Snake, by making it themselves.
The Setup
Using an FPGA (Field-Programmable Gate Array), they built a game that was more than just a simulation. An FPGA is an integrated circuit that has components called “logic blocks” that can serve many functions from simple logic gates like AND to memory elements. The way they programmed the FPGA to get a video signal is based on the integrated clock. By “dividing” the clock period, they obtained the horizontal and vertical synchronization signals at the optimal frequencies, based on their FPGA board clock, set up at 100 MHz. They also used two counters (one horizontal, one vertical), with limit values corresponding to the desired resolution (basic 640x480 pixels). To draw the snake and moving objects, they specified the wanted coordinates on the screen, and refresh them at the desired speed (the maximum refresh rate being the HV sync frequencies). At first, they tried to model collisions with the code from Pong, but the results varied so they wrote a code that checked the position of the head in relation to the bounds of the screen within the border. The user can control the directions of the snake with two buttons (left and right).
The Snake
Overall, they learned that snake and pong are vastly different, and given the challenges of using new things like FPGA and Verilog, it took them over 120 hours to put it together. If they had to do it again, they would have built their game from scratch, not try to build it from a base. If you’re interested in learning more about classic video games, one of their resources they found for the code to Pong is fpga4fun.com.
2.14.11 The Making of Snake Norris from Olin Projects on Vimeo.
The LockCracker: Cracking the Combinations You Forgot
Just about everyone has an old combination lock lying in a drawer or hidden in a cabinet somewhere. Until recently, such locks were worthless to owners who had lost or forgotten the combination. A team of Principles of Engineering students, Jessica Bethune, Jessica Noglows, Aiswarya Kolisetty, and Robert Sobecki, decided to create a solution for this irritating problem. They call it The LockCracker. The LockCracker is a mechatronic device that holds a combination lock and physically dials in hundreds of combinations to get it open. When it senses that it has opened the lock, it stops spinning the dial and displays the correct combination on the screen of the computer to which it is connected.
So far, the team has consistently found long-lost combinations in less than 40 minutes, and estimates that the maximum possible time required for their device to find combinations will be one and a half hours. While designing the LockCracker, they learned that many people have not forgotten the entire combination for their lock; they often know one or two of the numbers. Therefore, they included a feature in our program that allows the users to select from several different lock-solving options based on how much of the combination they know. Knowing even one number of the combination cuts down the time required to solve the lock by between 50 to 95 percent.
As of now they have only tested the LockCracker with MasterLock brand locks with a forty-position dial, though their flexible system can easily be modified to fit with just about any turn-dial lock. A description of their mechanical, electrical, and software systems can be found in more detail on their website: http://students.olin.edu/2013/jnoglows/The_LockCracker/The_Project.html . If you have a MasterLock and can’t remember the combination, feel free to contact them and they’ll help you out!
Here's a video demonstrating how the LockCracker “cracks” combination locks based on user input. Towards the end of the video there is continuous footage (shown faster than at actual speed) of the LockCracker finding the entire combination to a now-useful lock.
2.11.2011 The LockCracker from Olin Projects on Vimeo.
Autofrost: Making Custom Cupcakes, One At a Time
The AutoFrost automatic cake decorator is a frosting printer which moves automatically in two axes while dispensing frosting, with a manual third axis for height. A user designs a cake up to 11x13” in our paint-like Python GUI, then presses a button to send commands to two Arduinos. These control two stepper motors and a servo, which effect the position and dispensation rate of the frosting nozzle. The project was the result of a semester’s effort by Tim Raymond, Kelsey Breseman, Tara Krishnan, Ilana Walder-Biesanz, and Karan Kanodia for the Principles of Engineering class. Though born out of a misunderstanding of the “Cupcake CNC” 3D printer, the idea piqued interest; the finished product was innovative and successful, and was featured on many online tech blogs, including Wired.com, PopSci.com, and Hackaday.com.
The mechanical system for this project uses two stepper motors, which turn drive shafts in both the X and Y axes. The X axis drive shaft moves the cake back and forth along the X axis, while the Y axis shaft moves the nozzle head assembly along the Y axis. The motion along each axis is facilitated by drawer slides. A servo motor is used to drive a rack and pinion system, which dispenses the frosting at a constant rate onto the cake. There is also a manual wheel which allows the height of the nozzle to be adjusted for different cakes.
We chose to write the software side of things primarily in Python. We made a Microsoft Paint styled drawing GUI, where the user could draw on a canvas proportional to the size of the cake in a variety of colors and styles. Once this was completed, the drawing was converted into a list of points, each containing information about frosting color and distance from the next point. These points were then passed along to the arduino using Pyserial. The arduino then interpreted these distances, and in turn drove the stepper motors such that the drawing would be properly scaled to fit on the cake. A second arduino was used to control the frosting dispensing via the servo motor.
1.25.11 Autofrost: Making Custom Cupcakes, One At a Time from Olin Projects on Vimeo.
If you would like to learn more about Autofrost, here's their website
Making a Milkshake Machine
This was a Principles of Engineering (POE) project by Margaret-Ann Seger, Tanner Reid, Heidi Nafis, and Hannah Sarver. Their goal for this project was to streamline the process of making a delicious milkshake. They initially aimed to accomplish this by focusing on four main goals: the final machine would rotate to dispense toppings, and it would be clean and efficient for the user. They hoped to dispense different toppings/mix-ins in an efficient manner. They decided to do this via a rotational mechanism (like an automated Lazy Susan) with multiple topping containers mounted to the top of the mechanism. To dispense a set amount of a given topping from a simple user input, they decided they would use a trap-door mechanism that opened and dispensed the contents of the container for a certain amount of time depending on the amount of topping specified by the user.
"Team Milkshake" ended up scaling back their original design to something we could reasonably accomplish within the time scale of the project. They took out plans for a lid-closing mechanism, a whipped cream dispenser, and interfacing the blender to blend directly from the graphical user interface, all of which were included in our original project design. What they ended up with was an automated topping dispenser that is able to dispense user-controlled amounts of up to 4 different toppings into any separate container that is placed correctly under the machine. They used a large servo to control the rotational motion of the topping platform, along with small servos to open the appropriate dispenser door when each particular topping is above the blender. They used a Python GUI that sends commands to a PIC microcontroller programmed in C so that the user directly tells the servos when to turn (although the distance turned is a predetermined amount). Thus each user of the machine is able to have a customizable milkshake experience and put the optimal amount of each chosen topping into his or her final product. In addition, their GUI can play the Milkshake song when the user presses the 'Bring All the Boys to the Yard' button, which was a key part of the system's functionality.
Their milkshake maker was altered significantly from its original goal to accommodate for the multiple problems and challenges encountered throughout the construction and testing phases of our PoE project. Our first major issue was with the large servo that rotates the topping platform. Since servos aren’t symmetrical, it was hard to center in the central column, and part of it collided with the pieces it was attached to when it attempted to rotate, and so the rotation was jerky and limited. To fix this problem, they built a support structure to constrain the servo within the stationary column. Additionally, they also found during testing that if the servos used for opening the dispensing doors were mounted directly on the platform, the weight of the toppings on the doors would bend them downward and pop the doors off of the servos. To combat this, they inverted the servos and suspended them with brackets so that the weight of the toppings would instead push the doors down against the servo, theoretically securing them on as opposed to popping them off. After installing tracks for the doors to slide into for additional support, the process ran much more smoothly. They also made simple structural fixes to account for the glass jars falling over and wires tangling, and to help place our circuit and blender optimally.
1.24.11 Making a Milkshake Machine from Olin Projects on Vimeo.
Understanding Microprocessors Through K’Nex
At the heart of every microprocessor are millions of microscopic logical components that manipulate electrical pulses. One of the most fundamental is the AND logic gate. It’s a very simple device. It asks if both of its two inputs are “1” or “true”. If they are, the gate’s single output registers a “1” or “true”. In all other cases, the gate’s output shows a “0” or “false”. When you combine this simple operation plus several others you can create extremely complex systems from simple adders all the way up to the modern programs that run on computers today.
To explode the microscopic into the macroscopic, Steve Zhang (class of 2012) and Evan Morikawa (class of 2011) created a single logical AND gate using the toy building blocks, K’Nex. If you look at the final video here, you can see the two inputs of the AND gate coming in from the top, and the single output leaving from the bottom. The center piece even looks like the schematic drawing of a traditional AND gate. Each wire holds its state as either a “1” or a “0”. The horizontal colored flags sticking out of each wire (“1” = “green” and “0” = “red”) represent the state of that wire. A single K’Nex ball represents a change in the state of a wire. This is why when you drop a ball down a wire that was once a “0”, it changes state to a “1”.
In theory, millions of these stacked together could create all of the complexities of a modern microchip. However, millions of AND gates at this size would be about as large as the Boston metropolitan area and would process millions of times slower than a real electronic chip. Nonetheless, the physical facsimile of K’Nex holds to the behavior of its electronic counterparts. When we were building the device, we had to construct large trussed support structures in order to assemble the main unit in three dimensions. We also had to deal with the “physical” nature of K’Nex, such as the bounciness of the rods and the support structures required to hold such a large device up. Nonetheless, the resulting product you can see in the video attached to this post shows a functional logical AND gate and is the beginning of a computer age made out of K’Nex.
1.20.11 Comp Arch "K'Nex Differential AND Gate" from Olin Projects on Vimeo.
This project was done under the inspiration of Olin College’s computer architecture class. This class covers the intricate boundaries between the hardware and software components of a computer at their most basic level. In addition to a theoretical understanding of Boolean logic, processor design, and several other standard topics, it has a heavy focus on hands-on exploration of topics such as the Boolean logic one investigated here.
Creating A Talking Hand
Principles of Engineering (POE), is a class most take their sophomore year. At the end of this class they create their own final project, and this is an example of the work accomplished by Tom Pandolfo, Ijeoma C. Ekeh, and Michael Sullivan. In their own words:
Our mission at the start of this project was to make a mechatronic hand that would be able to mimic human sign language. By using a Python-coded GUI that allows the input of a letter of the alphabet, our circuit would drive the output to multiple servos that wind up moving the fingers to spell what is desired. In addition to that, we had hoped to allow the hand to “learn” signs by equipping it with a function that tracked the servo positions for a duration in which the fingers were manually moved. Once the movement was finished, the entire motion could be saved as new sign that would then be able to be used in addition to the 26 letters of the alphabet. In essence, we were trying to conceive an awesome mechanical hand that had sentient AI and could easily take over the world.
Unfortunately, as we quickly found out, making a mechatronic hand is much harder than we thought. We found out that despite being able to easily conceptualize what was needed, such as cords to pull the fingers and a lovely arm “body” to make everything neat and organized, actually hammering out what specific materials we needed for everything is an intensely grueling process. We waited too long before we actually made prototypes to test the kinds of materials, and as a result, we were pigeonholed into using the first thing that came to mind, having little to no time left over to change it out for anything more efficient. Fortunately for us, the materials we did end up choosing work, but we know that if we had planned our time more efficiently, we would have been able to use other materials, and perhaps been able to more strategically planned the shape of our hand so it would less bulky.
In the end, however, we actually were able to make a mechatronic hand that carried out several of the functions we had intended it to do. It does sign several of the letters of the alphabet, but it only does those letters that do not require any multi-planar finger movement. For example, to form I, J, and Z, you need the hand to move out of position. We did not have the time or idea of how to easily incorporate that, so we could not sign those letters. Fortunately, we were able to figure out how to get the scissoring motion of R to work, which required the fingers to bend towards each other, instead of forward, towards the palm. Our final product is a hand that can sign around 21 of the 26 letters of the alphabet quite efficiently, and it can do all this by receiving input from a Python-coded GUI.
12.19.10 POE "The Talking Hand" from Olin Projects on Vimeo.
Entrepreneurship at Olin: The Midnight Bakery
The Midnight Bakery is an ongoing student business founded in the spring of 2010 by Samantha Becht and Hannah Sarver. Building on their passion for baked goods and a penchant for late nights, they ran a summer business out of the Olin Dining Hall, which shipped ordered goods to customers nationwide, performed local delivery service of fresh-baked goods until five in the morning (catering to the college-student lifestyle), and participated in community events including the Natick Farmers’ Market and author talks at the Wellesley Public Library.
Due to the success of the summer business, the Midnight Bakery decided to continue operations into the school year, hiring new employees Kelsey Breseman, Becca Schutzengel, Blair Emanuel, and Amy Whitcombe. They started off the semester selling at the farmers market during the fall season. Since the Natick market closes for winter, they are currently exploring alternate markets, including the return of online mail-ordering, sales to clubs and events, and in-college sales after dining hall hours. Building on the model of FBE business "The Flour Garden," the Bakery has also expanded to retail ingredients to Olin students in the dormitory kitchens.
Some fun facts about the Midnight Bakery:
In the average week, they use over 25 pounds of flour, 6 dozen eggs and 12 pounds of butter making close to 200 cookies and 15 loaves of bread.
Each week the members of the Midnight Bakery log over 45 women hours in planning, shopping (we have become friends with the employees at Restaurant Depot), baking, and answering the deluge of emails.
Even though sometimes our work takes us from early in the morning until late into the night, they still retain our passion for baking and motivation to continue our business during our time at Olin.
For more information, or to order delicious baked goods from the Midnight Bakery, please see their website.