For the midterm project, Steve and I worked to create a mechanical structure. We were inspired by the Hoberman mechanism and decided to create a 2D version of it.

Process

Phase 1: Base Structure

The very first step was brainstorming. After exploring several ideas, we decided to build a model that mimics a “burst” with parts moving in and out repeatedly as a person turns a crank. To gather inspiration, we went through 507 Mechanical Movements and searched online, where we came across the Hoberman sphere. It is a spherical structure that dynamically expands and contracts, often seen in toy designs. We realized that a 2D version of this expandable structure was exactly what we wanted to create. To bring it to life, we planned to use laser-cut wood for the main frame and incorporate gears in both the horizontal and vertical directions to achieve the expanding and contracting motion.

 

 

 

 

 

 

 

 

After finalizing the mechanism, we moved on to CAD modeling to test how everything would work together. We built an Onshape assembly to visualize the motion, and it generally performed as we had planned. The structure mainly consists of a cross-shaped base structure that holds dowels for the gears, 9 evenly distributed gears with the central one serving as the power source, and 4 sets of arcs attached to the corner gears. Once the design was complete, we prepared vector files and laser-cut them out of cardboard for our low-fidelity prototype.

We then tried to assemble our cardboard parts together, but it turned out to be too fragile to properly demonstrate even the basic motion of the system. A part of that was also because the gear teeth were too small. After discussing the material issue, we laser-cut the new gears and arcs using plywood and stacked the base with cardboard copies to accommodate the dowels. This version performed much better as the arcs rotated as expected, and the gears meshed smoothly with correct spacing. However, as shown in the video, we had to use four threads to pull up the connection points between each pair of arcs to prevent them from sagging and interfering with the gears’ movement. Additionally, there was nothing to keep the gears from slipping off the base when the model was tilted, causing the structure to fall apart during handling.

 

 

 

 

 

 

 

 

Video of the Medium Fidelity Prototype

To address the issues found in the medium-fidelity prototype, we looked at the model in detail. For the arc problem, we realized it was mainly caused by the tolerance between the hole diameter on the arc and the actual diameter of the dowel. When we first designed the holes, we assumed each arc would simply rotate around its axis, so I intentionally added extra clearance to make the rotation smoother. The dowel we used had a diameter of 0.3125″, while the holes on the arcs were set to 0.33″ to provide that additional space. However, this larger tolerance unintentionally created another degree of freedom, causing the arcs to wobble vertically due to the gap between the dowel and the hole. To fix this, we used washers as stoppers to limit the wobbling while still having smooth rotational motion around the dowels. We applied the same method to the gears to keep them in place. The successful use of washers as stoppers became a key part of our iteration and troubleshooting process, effectively addressing the instability of the arcs and the gears.

 

 

 

 

 

 

 

 

 

Phase 2: Crank and Outer Arc Layer

After finalizing the main structure, we began building the crank and the second layer of arcs. Initially, we considered placing the crank on the side of the structure, but due to time and material constraints, this was too difficult to achieve. As a result, we decided to mount both the center gear and the crank onto the same rod, allowing the crank’s rotation to directly drive the center gear and power the entire structure.

In our first version, we used the 0.3125″ rod and put the crank under the base structure. This design required the base to be hand-held in order to turn the crank from below, which meant the model couldn’t stand on its own. We initially planned to build a rack to support it, but after consulting with Dr. Wettergreen, we revised our design to place the crank on top instead. This adjustment allowed the base to remain still, making the model much more stable.

Another improvement we made to the crank was redesigning the connection between the metal crank bar and the rod. Our initial plan was to simply press-fit the rod into the hole and reinforce it with super glue. However, this method relied heavily on precise press-fitting and the glue’s strength to prevent the bar from rotating around the rod. Dr. Wettergreen suggested a more reliable method, which is creating two semicircular holes with a central slit on the metal bar and cutting a matching slit along the rod’s center. This interlocking design mechanically prevents the bar from rotating relative to the rod. We fabricated the metal bar using a water jet cutter from a 0.25″ aluminum sheet, used the 0.5″ diameter rod, and successfully assembled the crank.

Next, we attached the outer arcs to the existing inner arcs. Initially, the outer arcs were too short to fit properly, but after adjusting their angles and lengths, we aligned them properly. Using the same dowels and washers, we connected the arcs and tuned the tolerances to ensure a smooth motion.

Phase 3: Post-processing and Decoration

At this stage, the main structure was fully assembled and functional, so we moved on to post-processing and aesthetic details. We initially considered staining the entire wooden structure but were reminded that staining could significantly change the friction between moving parts. Since this mechanism depends heavily on smooth gear interactions and arc rotations, we decided to limit the surface treatment. Instead, we applied a clear coat only to the gears’ surfaces and added decorative star patterns on wood circles using vinyl masks.

We also made use of some of the small gears from earlier iterations by gluing them onto the arcs as flower-like decorations. Finally, we attached our engraved nameplate to the side of the base to complete the model. The mechanical movement we used was no.24, spur-gears.

Video of the Hoberman Model

Cleaned Workspace

Cost Analysis

• Raw Materials

Birch Plywood 32″x24″ – 4 : $20

Aluminum Sheet 5″x5″ : $8

Wooden Rod – 1 : $1

• Tools (based on usage time, approximate)

Laser Cutter : $36 (4 hour for $9/hour)

• Labor

35 Hours – $10 per hour – 2 people : $700

Total Cost: $765

Reflection

Through this project, we completed a functional 2D Hoberman mechanism model and learned several lessons about the process of creating a mechanical model. Rather than following the original design all the way, the process is a continuous process of iteration through many trials and errors. From adjusting laser-cutting parameters to figuring out arc connection methods, most solutions were discovered along the way. We learned how to identify new problems more quickly and respond effectively to fix them. We also realized that physically prototyping often yields better results than long theoretical discussions. It’s also important to test run first. For instance, we once set the wrong hole diameter on the arcs and ended up wasting many copies, a mistake that could have been avoided with earlier testing.

Beyond technical lessons, we came to appreciate the importance of teamwork in an engineering project. Our collaboration went smoothly because we communicated effectively and made sure we understood one another’s ideas. The good communication not only kept us organized but also inspired creative solutions. Overall, this was a meaningful and enjoyable experience, and we are grateful to everyone who helped along the way.