About Me

IMG_20190817_181218.jpg

Hello! My name is Jason. Welcome to my portfolio.

 

I am a Chicago based designer and mechanical engineer. I graduated from Northeastern University in 2017, BSME.

As an engineer, I aim to leverage my innate curiosity to learn more about the world, and then apply technical knowledge and creativity to solving real world problems. 

It is my hope that this portfolio will give you a sense of both my capabilities and the projects I've worked on over the years.

Thank you for taking the time to check it out. You can reach me at jasonmarktoby@gmail.com if you are interested in working together.

Projects

 

To view project details, click any of the images, or simply scroll down the page

LED Protection Kit 

Company: Bevi

 
bevi machines.jpg
Bevi's Standup and Countertop machines.
2018_bevi-078_1.8mb_2.png
The Countertop Bevi uses LEDs to illuminate the dispense area to improve the user experience.

Background

The Bevi machine is a 'smart water cooler' that dispenses flavored still and sparkling beverages. The mission of Bevi is to provide an alternative for bottled drinks that cause significant plastic waste, and is primarily used in office environments as a perk for employees. 

In particular, this project pertains to the bright LED panels located near the nozzle. These color changing LED panels are an important part of the Bevi brand, and help to add a fun touch to the experience, as well as illuminate the dispense area in low lighting situations. 

A few months after the release of the Countertop Bevi, we received reports of malfunctioning LED panels on some of the machines. I was tasked with diagnosing the root cause and coming up with a solution. 

11_edited.jpg
In this benchtop test, I applied sparkling water to LED strips mounted below steel plates, observing failure mechanisms.
Graog.PNG
This graph shows the temperature of the LED panel rising as time passed and more fluid was applied. The symbols represent changes in the type of fluid as well as observable failure modes. 

Diagnosis & Testing

 

It was suspected that these failure modes were related to a nozzle misalignment issue. This misalignment caused liquid to spill onto where the LED panels were mounted. To investigate this, a colleague and I set up a few tests to see what happens when fluid was applied directly to the LED panels. 

As expected, these tests did cause failure of the LED panels. The symptoms were very similar to what customers were reporting from their machines, which included blinking lights, color changes, and ultimately complete failure of the LED panels. 

It was found through this testing that the LED panels didn't fail right away. It took anywhere from several hours to multiple days for them to start showing symptoms of failure. The rate at which they failed depended on the level of exposure and the type of drink applied.  

20180806_120454.jpg
Using pliers and a heat-gun, I experimented with various types of heat-shrink to find the right fit. 
20180806_121640_edited_edited.jpg
The finished assembly, with the LEDs and wiring completely enclosed in heat-shrink tubing.

Solution

 

At this point, the design of newer Countertop Bevi models had already been adjusted to eliminate this issue. However, older models were still vulnerable. We were tasked with developing a retrofit kit that could be installed in these machines to protect the LED panels. 

After testing out a few different ideas, the most effective solution proved to be a layer of clear heat-shrink tubing, applied over both LED panels and a small section of wiring. While providing excellent protection against fluid damage, the added thickness prevented the LED panels from fitting into the cutouts in the machine. 

To compensate, I designed a small metal bracket that housed the LED panels, which was secured to the machine via double-sided VHB tape. This formed the basis for the retrofit kit, which was made in a small production run by a local contract manufacturer. 

As of June 2019, this kit is still in use by Bevi, and has helped to reduce the incident rates of Countertop LED related failure modes. Additionally, data from this testing was used to improve the LED protection in future designs. 

CIMG2015_edited.jpg
Sample retrofit kits from the contract manufacturer, including installation instructions. 
20180831_170706_edited.jpg
The finished assemblies use double-sided tape to secure themselves to the Countertop Bevi.

Quad Cap Fixture

Company: Ambri

 
Battery Cells (WBUR)_edited.jpg
Ambri's liquid metal batteries aim to improve grid power management and energy storage for solar and wind.
MVP Sprint 250.png
A CMM, or Coordinate Measuring Machine, uses a probe and camera to measure parts and assemblies.

Background

Ambri is an energy startup looking to change the way we store energy. They are pioneering a liquid metal battery technology that is ideal for clean energy storage as well as smoothing out demand changes in the energy grid. 

A critical part of assembling each test cell was assembling and the 'cap', which contained the seal preventing the anode and cathode from bridging. As the Quality Engineering Co-op, one of my responsibilities was to inspect each cap before it goes into the battery cell.

 

However, the inspection fixture only allowed a single cap to be inspected at a time, which made the process fairly labor intensive. I was tasked with coming up with a new design that could automatically inspect up to four caps at once. 

Old Fixture.png
The existing fixture for measuring the 'cap' assembly, which sealed the battery at the final stage of production.
012329 Iso Exploded 4.PNG
The new fixture design, composed of an acrylic base plate that used dowel pins to key into the main fixture.

Designing the new fixture

 

The existing design located the cylindrical seal into a cavity in the middle, and then used three ball-ended set screws to seat the top plate. I adapted this design into a 2-by-2 array, and modified the corner geometry to further restrict the motion of the cap while it was mounted. To save cost on materials, the material was changed from aluminum to Delrin. 

To improve alignment, an acrylic base plate was added to the design. The CMM's stage had a steel corner bracket which allow fixtures to snap in place using magnets. The existing design did not take advantage of this, and the manual alignment often caused errors during inspection, adding additional time to the process. 

To make assembling the two components easier, I added dowel pins to the acrylic base plate, which mated with slip-fit holes underneath the Delrin block to align the two parts. These were largely inspired by my time at Bio-Rad (see Glue Robot project). After reviewing the design with my manager and creating drawings for fabrication, it was ready to be built. 

20161206_104445_edited.jpg
After receiving parts, I inspected critical dimensions to ensure the fixture would function as designed. 
20161209_093116_edited.jpg
Ball-ended set screws were used to seat the flat plane of the cap assembly. They had to be installed just right...

Assembly & Result

 

After creating fabrication drawings, I sent them out to a small plastics shop in South Boston that Ambri had worked with before. After inspecting the parts for dimensional accuracy, the dowel pins, magnets, and set screws were all installed. The set screws were particularly tricky, as each had to be carefully aligned relative to eachother so that they would not contribute to flatness deviations. 

Next, the inspection routine had to be adjusted to accommodate the new parts. This was fairly straightforward, as I had taken the travel path into account during the design process. Because the CMM used both a camera and probe, it was important that the fixture fall into a specific area where all measurements were possible and no collisions would occur. 

 

Finally, after conducting a GR&R check, the fixture was ready to be used. The introduction of this fixture made it not only possible to inspect four assemblies at a time, but increased the hands-off time from two minutes to eight minutes. The additional location features improved the accuracy and reduced errors. I'm proud to say this assembly is still in use, helping Ambri to streamline their inspection process so that they can revolutionize the energy industry. 

Assembled Fixture.PNG
The final assembly, ready for part inspection. 

Solder Test Station

Company: Bio-Rad

 
Solder Fill Station (Bio-Rad).png
Screenshot of the CAD assembly in SolidWorks.
IMG_6687.JPG
Photo of the assembled test fixture.

Background

Gnubio was a biomedical startup utilizing microfluidic technology to streamline the DNA diagnostic process. They were acquired by Bio-Rad in early 2014, but unfortunately was closed down in 2017.

As a Mechanical Engineering Co-op, I worked on the automation team, with the goal of automating the manufacturing process of a consumable cartridge and microfluidic chip used to analyze the DNA assays. 

Part of the manufacturing process for the consumable cartridge involved injecting molten solder to create electrical contacts. This was being done manually, and was fairly time intensive. 

The goal of this project was to test out a process for automating solder injection that was being developed in a much larger machine. I was tasked with designing the test fixture and validating this process to see if it worked. 

IMG_6768.JPG
Solder ingots were loaded into a cylindrical chamber and heated until fully liquid. 
IMG_6659.JPG
The syringe block pulls the molten solder through a series of tubes and interlocks.

Design Overview

The proposed process involved melting cylindrical solder ingots and then pulling the molten solder into a syringe. The cartridge is then positioned over the syringe, and solder is pushed up into the cartridge. 

The solder flow pathway was complex, using stainless steel tubing, silicone gaskets, and numerous threaded connections.

To create the fixture, I worked with my manager to adapt the larger design into a smaller stand. I spent a few weeks laying out all of the components in CAD, the layout slowly evolving in complexity. Many parts were recycled from prototypes of the larger machine, most notably the clamp holding the heating assembly in place.

This was my first experience ordering CNC machined parts that I had drafted. Design drawings were reviewed and approved, and I gained some machining experience creating the 80/20 base for the station. 

 

IMG_6493.JPG
CNC machined and anodized aluminum parts, fabricated by a vendor in China.
IMG_6690.JPG
This assembly contained temperature controllers for regulating the output of the fixture's heating elements. 

Testing Results

Once the test fixture was constructed, I carried out a series of tests was carried out to evaluate the solder filling process. In order to control the various heating elements, I used Omega temperature controllers that another design engineer had assembled into a box. The goal was to see if we could pull solder into the syringe, and then push it back up to form the electrical leads in the cartridge. 

 

Initial testing showed a huge problem. The solder ingots successfully melted, but no matter what I tried, I could not pull it into the syringe. After extensive troubleshooting, it was found that it took an additional 60 PSI of nitrogren to pull solder into the syringe. An analysis revealed that the channel diameter was too narrow (~0.8mm) to pull the viscous solder through. 

While the tests were ultimately not successful, I learned a lot from this project. I presented these results to the team and heard from later interns that the design was changed to widen the channels significantly. Unfortunately, I do not believe the design ever made it to production, as Bio-Rad shut down Gnubio in September of 2017. 

Glue Robot

 

Company: Bio-Rad

Capture 7 - Robot w-o Cart.PNG
A SolidWorks screenshot showing the 3-axis robot mounted on the base plate of the automation cart.
Capture 2 - ISO w- robot.PNG
My work focused on the support structure, which used machine design elements to support and align it.
Capture 4 - Rear Iso w-o robot.PNG
Capture 3 - Exploded View.PNG

Background

Gnubio was a biomedical startup utilizing microfluidic technology to streamline the DNA diagnostic process. They were acquired by Bio-Rad in early 2014, but unfortunately was closed down in 2017.

As a Mechanical Engineering Co-op, I worked on the automation team, with the goal of automating the manufacturing process of a consumable cartridge and microfluidic chip used to analyze the DNA assays. 

As part of this process, the two sides of the microfluidic chip needed to be glued together. I was tasked with designing the support structure for a custom dispensing system for the glue, which was going to replace an off-the-shelf dispensing robot. 

The 3-axis robot was aligned to the base plate through a series of aluminum blocks and dowel pins. 
This exploded view shows all the parts and screws composing the main support structure. 

Design Overview

My initial task was to design support structure for the robot to sit on. I started by familiarizing myself with it's mounting requirements, which involved a series of 'toe clamps' that constrained it's horizontal movement. I spent a lot of time plotting out the travel required for the movements of the glue path, and figuring out how large the automation cart needed to be to accommodate it. 

This project introduced me to a modeling technique called master modeling, which was frequently applied by Gnubio. Instead of creating each part independently, there was a master sketch that controlled the dimensions and location of each part. This was used to control the table dimensions. Initially the table was designed in 80/20, but after a particular design review it was decided to use solid aluminum blocks instead.

I also designed parts that held the syringe, which had their own design considerations to keep in mind. The syringe itself had to be swapped out, so I designed a small plate that used a thumb screw to bend a C-clamp and secure it. I utilized a Finite Element Analysis in SolidWorks to ensure that the mechanical advantage from the thumb screw would be enough to bend the part closed. 

Syringe Mounting Assembly (low res).PNG
Capture 5 - Syringe Clamp FEA.PNG
This assembly held the syringe that dispensed the adhesive, and was attached to the Z axis of the robot.
I ran FEA simulations to adjust the part geometry, checking to make sure it would secure the syringe tip.

Results

Unfortunately, I did not ever see these parts make it into production. The original plan was to contract out the assembling of the robot, but progress stalled and my internship ended before I could see anything happen.

 

However, I learned a lot about machine design from this project. It introduced me to GD&T, dowel pins, ISO tolerances, basic FEA, and working in PDM. I completed production drawings for all of the machined components before starting work on the Solder Test Station. 

The last I heard from my former manager, the parts were being ordered in April of 2016. Gnubio was later shut down on September of 2017, and all projects were either shuttered or absorbed by Bio-Rad for other purposes. 

Sorry Flip

 

Company: Hasbro

Sorry Flip no background.jpg
Sorry Flip Logo.PNG
Sorry Flip Logo.PNG

Background

My first college internship was on the Kre-O team at Hasbro, working as a Project Engineering Co-op. Hasbro is one of the largest toy companies in the world, responsible for famous toy brands such as Transformers and My Little Pony.

Alongside our normal responsibilities, all of the co-op students were given a chance to design a new toy or board game, and pitch their idea to the executives and designers. This was called the 'Grand Idea Fair', and we were encouraged to spend around 10% of our time each week working on this. 

Over the six months of my co-op, I worked with four other interns to brainstorm ideas and concepts. We were specifically challenged to create a new board game platform that could be applied across multiple games.

Sorry Flip GIF.gif

Design 

 

After deliberating between many ideas, we settled on a new platform where traditional game boards were expanded with mechanical elements. Our most promising idea was a magnetic game board that flipped from one side to the other.

To achieve this, the board was raised about 6 inches off the table. Each side of the board had a small clip that was attached to it. These pieces were critical to the design, as they contained a small ball and detent which locked the rotation every 180 degrees. It took several iterations of prototypes and designs to accomplish the desired balance, which we assessed as a team. 

In addition to this, we created a new version of Sorry! where players had to get their four game pieces to their base on the other side of the board. Only the side of the board facing up was active, so players had to strategically use the FLIP ability to restrict the motion of others. 

Leg Support Snapshot.PNG
Axle Clip Snapshot.PNG
Clip and Board soft edges.PNG

Result 

To create the prototypes, I was given access to an Up Mini 2 desktop 3D printer, which allowed me to work through iterations quickly rather than waiting for them to be manufactured. I learned both how to operate it as well as how to design parts for it, teaching me skills that I value to this day. 

This project culminated in a presentation to various executives at Hasbro. In addition to creating a functional prototype, our team also performed a comprehensive cost analysis to project what the design would cost to manufacture if it was released to the market. 

As my first foray in a corporate environment, my time at Hasbro taught me many things. It taught me the value of creativity, no matter how large the company- new ideas are always important. I am grateful to have been given the chance to collaborate with both my peers and mentors within the organization to make Sorry Flip. 

Solar Tracking Mechanism 

 

Senior Capstone Project, Northeastern University

CAD - Parabolic Mirror Assembly.png
Senior Capstone_edited.png

Background

Each year at Northeastern, mechanical engineers are teamed up for the senior Capstone course, with the goal to take on a complex engineering problem by applying theory learned in class. Students are allowed to submit their own project or work on an existing project proposed by a professor or the department. 

My team's project was a solar desalination machine, designed to provide a low-cost freshwater solution for coastal countries with easy access to seawater. Previous Capstone groups had created a design using a parabolic mirror and solar tracking mechanism, and our task was to further improve and optimize it.

 

The existing design proved difficult to operate in practice. The solar tracking mechanism was not reliable enough to operate on it's own, which meant we had to rotate the mirror manually. Similarly, users had to pour seawater into the inlet every hour or so, instead of a continuous input that would maximize output. After discussion with the team, I opted to take the lead on repairing the solar tracking mechanism, while my group focused on other improvements to the machine. 

Graham Escapement.gif
image14.gif
Research - Choosing a suspension_edited.

A Lesson in Clockwork

 

The solar tracking mechanism was essentially an adapted version of the grandfather clock. A pendulum swung back and forth, each swing allowing a falling weight to transfer some energy into the rotation of the parabolic mirror, creating a slow single axis rotation to track the sun. However, the escapement movement was complex, and for several months I was stumped. No matter what I tried, the clock-like contraption would only run for a few minutes before grinding to a halt. With only two months to go until graduation, I decided to go to our school's library to see if I could find any insight from the past. 

A breakthrough didn't come until I was several hours into Phillip Woodward's My Own Right Time. The author had included a diagram of the anchor escapement that the previous Capstone group had adapted. As in their design, the swinging pendulum and anchor were coupled by a suspension spring and crutch. However, the diagram in the book looked slightly different than theirs. The crutch was much longer, and the pendulum spring was mounted higher up. Onto something, I continued to search for answers. 

I finally found the solution in Ward Goodrich's The Modern Clock. Published in 1905, it was a  comprehensive text on clock design that also provided insight into the history of timekeeping.  With the assistance of both this text and a helpful internet guide interpreting it, I revised the anchor to correct the lift angle, extended the crutch to provide more mechanical advantage, and aligned the axes of the suspension spring and anchor. I added my own improvements as well, adding D-profiles to the shaft to reduce slip, and redesigning the wooden frame to make it more robust. 

image25.gif
indoor testing.gif

Results

After many rounds of 3D printing and hours in the shop, I finished the modifications to the design. To the team's immense relief, the changes worked. The escapement motion, previously erratic, became smooth and stable. With nearly a week to go until our final testing, the machine could rotate autonomously over the course of the day. In parallel, my teammates had been hard at work. They developed a way to test the machine indoors without sunlight, using heating elements typically used for thawing frozen pipes. They also integrated a float valve to introduce seawater into the pipe from a larger reservoir. 

Unfortunately, our final outdoor test did not yield the result we wanted. We only produced 1 liter of freshwater, which was impressive for April in New England, but was not even close to the 18 liters we wanted, even when adjusted for the differences in solar radiation. While disappointing, we had made solid progress that improved the reliability of the machine and made it possible to run fully autonomously. As others did before us, we left our work behind for future Capstone groups to build upon.

Although our group did not achieve the output goals, I learned a lot from the experience. The decision to use a Graham escapement to rotate the mirror was a neat idea, but required a great deal of care and precision in order to work properly. When my initial attempts failed, I found a lot of value by taking a step back and researching the construction of mechanical clocks. I think this is a good lesson that I hope to carry with me as I progress with my career. 

Flat Pack Guitar Stand

 

3D Fundamentals: Structure and Drawing, Northeastern University

20170206_010156_edited.jpg
Guitar Stand.png

Background

In the senior year of my undergraduate studies, I signed up for a class called 3D Fundamentals. Typically taught to freshmen art students, it was a project-based class that was taught in a beautiful design studio in the art department. Looking for a breath of fresh air from the rigors of the engineering curriculum, I was excited to try something new. 

 

Our first project was to design a piece of furniture out of a single piece of 4' x 2' plywood, using as little waste as possible. After sketching out a bunch of different ideas, I grew most excited about an idea for a guitar stand. I wanted it to not only hold the guitar itself, but also have room for accessories and cables that were cluttering up my room. 

As the project moved along, we gained access to the laser cutter, and our professor introduced us to flat-pack furniture. This type of furniture is designed to be assembled with no tools and fold up flat. I decided to integrate this into my design in the second iteration, which forced me to forgo wood glue and think carefully about how each part would fit together. 

Handmade Iteration 1.PNG.png
20170129_175228_edited.jpg
Lasercut Iteration 1.PNG.png
20170206_005028.jpg
Lasercut Iteration 2.PNG.png
20170210_203030.jpg

Design Process

For the first iteration, I was mainly concerned creating a functional design that would also minimize waste. I created a cutout for the neck of the guitar to sit, and that shape was used again for the endstop where the guitar body would rest. This design was shaped on a bandsaw, and glued together using wood glue.

The second iteration took a much different form. We were allowed to use laser cutting, and the restriction that we had to leverage 90% of the plywood was removed. I spent an afternoon sketching out different ideas, and the design started to change from a very blocky design to a more organic form. I adapted to the 'flat-pack' challenge, instead of using wood glue, everything assembled via tabs and slots. I laid everything out in SketchUp and then used an extension to create .dxf drawings for the laser cutter. 

The final iteration was focused on functional improvements. The second iteration's tabs and slots were changed to hooks, based on a suggestion from my professor. This added additional structural strength that prevented the design from falling apart and made it much more sturdy. The new design was adjustable to accommodate to different guitar sizes, and had several hooks for cables and accessories. 

20170206_010156_edited.jpg
Guitar Stand.png
Final back view.jpg

Result

The final iteration was well received by my professor and the class. It was able to support my guitar's weight, and the geometry changes complimented the curvature of the guitar very nicely. I had created an additional part that secured the two sides of the neck, but found this to be largely unnecessary. 

Something we discussed in class that was really important to this project's success was to "let the process guide you". This meant being accepting some uncertainty, and focusing less on perfection and more on learning from subsequent iterations. This was a wonderful lesson as an engineer. It is often that we forget to step back and let the creative process take over. That being said, other times we must focus on the constraints of the project, or data from test results.

Overall, I felt that this class was really important for my self development as an engineer and designer. Being around art students was huge for me. I found myself drawing more, and being more engaged in my engineering classes, having a creative outlet. If I could go back and do it over again, I would have taken this class much earlier in my undergraduate education. I went on to complete two more projects in this class, see also 'Deadbeat Escapement Sculpture'. 

Deadbeat Escapement Sculpture

 

3D Fundamentals: Structure and Drawing, Northeastern University,

Escapement Mechanism Assembly (1).gif
V3 Iso View_edited.jpg

Background

For our final project in 3D Fundamentals, we were tasked with creating something personal, that represented what we had learned throughout the class thus far. We had already created a piece of furniture (Guitar Stand) and a motion sculpture that showcased a particular emotion (not in portfolio). For this project, I wanted to do something different.

At the time, I was very focused on my Senior Capstone Project, where I was working on fixing a mechanical clock that was made by a previous Capstone group. I figured it might be neat to construct a new one for this project. In particular, I had spent a lot of time working to understand the 'deadbeat escapement', or Graham escapement. 

I enlisted my friend Kevin to join me, and we worked out a design that brought the Graham Escapement to the center of the sculpture. It did not have any gears, it just focused on the interactions between the anchor and escapement wheel. Excited to see if we could replicate the motion on our own, we began work. 

V3 CAD with labels.PNG
Deadbeat Animation.gif

Design & Construction

The main job of any escapement mechanism is to control the release of energy. In grandfather clocks, it balances the release of potential energy from a falling weight with the periodicity of a swinging pendulum. This is accomplished via the interaction between two parts; the anchor and escapement wheel. 

In order to release the potential energy of the hanging weight, it must rotate the escapement wheel. The anchor prevents it from freely spinning, only allowing it to advance at the apex of the pendulum's rotation. This is what governs the duration of each 'tick' of a mechanical clock. 

We decided to place the escapement wheel and anchor on the outside of the device. Consequently, there is no 'clock' in our design, just a slowly falling weight. This omission was intended to make it much easier to build. However, even getting the escapement motion to work correctly proved to be a serious challenge.

V1 Prototype.PNG
V1 back .jpg
V2 Iso View.jpg
V2 back view.jpg
V3 Iso View_edited.jpg
V3 Side View.jpg

Iterations & Results

Through three round of prototypes, we explored the idea and how it might be constructed with what we had available. The first prototype was made mostly with power tools and rough measurements. The second prototype was made almost entirely by laser cutting. The third and final prototype was made using a combination of machined parts, 3D printed parts, and procured parts- including some lead buckshot to create the weights. 

Each iteration successfully replicated the escapement motion we were looking for. The challenge was keeping it going. The housing was extremely light, so we had to balance the weight of the pendulum with an additional weight on the escapement side. However, no matter what we tried, the escapement motion would eventually come to a stop.

We realized that we had left out a critical part- the pendulum's suspension. This assembly comes between the pendulum and escapement assembly, and is what keeps the pendulum's swing in check. It is my belief that this is what was preventing the design from working properly. We received an 'A' on the project for our efforts, but I hope to return to this project someday to test my hypothesis. 

5-A-diagram-of-a-pendulum-clock.png
caring-for-your-clock1.jpg

Freelance Mechanical Design 

Lighter Holder Keychain

 
Mouse Trap Assembly.PNG
14518442_10211110801592741_930570193_n_e

Overview

The goal of this project was to create a device that secured a standard BIC lighter on a keychain. My client wanted a design he could take to manufacturers to produce at higher volumes. My job was to take it from a napkin sketch to a final prototype. 

Over the span of a year or so, we iterated through several ideas. I used my desktop 3D printer (a Printrbot Makers Edition) to make some simple prototypes in PLA plastic. Eventually we sent a revised design out to Shapeways for 3D printing in their steel/bronze material

After three iterations in metal, I created a drawing package and compiled CAD files for a manufacturer to look at. We did not go through a design-for-manufacturing (DFM) stage, but I felt this project was a great learning lesson in producing designs on my own. 

14397349_10210984981847326_633265398_n.p
Mouse Trap Iso 1.PNG

Design & Prototyping Process

The first thing I did was to model the lighter itself. I wanted the design to be able to be slid on and off the lighter easily, and it was important to get the fit just right. I noticed that the cross sectional area decreased from the bottom to top, which meant that the device needed to be able to flex in order to secure it. 

I ran a basic FEA analysis on SolidWorks. Using a load of 10 lbf, I measured the deflection of the part relative to it's stress, and made adjustments to the geometry based on the results. As I worked through iterations, the biggest change came when transitioning from PLA to Shapeways' steel. 

 

Using data from their material sheet, I ran the analysis again. In order to maintain the same flexure, I needed to open up the previously closed keychain hole. However, I was able to thin out the material a lot. At that point, we sent the design out for the first 3D printed metal prototype, hoping for the best. 

14518442_10211110801592741_930570193_n_e
video-1475345015 GIF.gif

Results

 

After receiving the parts, I found myself surprised by how well they worked. We went through a few more iterations to reduce the material. Even a small reduction in material decreased the cost significantly, so it was worth it to figure out where material could be conserved. 

My client was very happy with how the design came out. I put together a mechanical drawing and a package of CAD files in various formats, so he could take it to a manufacturer if he desired, or simply sell it via Shapeways' marketplace. 

I feel very proud of this project. The FEA analysis was something I had done in my internships, and with this project I was able to incorporate that into my own process. The journey from idea to finished part was a satisfying one, but took a lot of work. Nevertheless, I would happily do it again, and am grateful to have had the opportunity to work on this project.