Spinning Cube

This was a university micro-project for a "Fast Prototyping" class. Our task was to create a project that incorporated at least two different fabrication processes, included at least one sensor and one actuator, and it also had to be fully dismountable.

With these constraints in mind, we decided to create a cube that could rotate and move using the inertia of an internal flywheel.

1. Arduino Uno
2. Battery 12V or DC power supply
3. Brushless motor
4. PLA for some elements 3D prints
5. Controller
6. Inertia flywheel
7. Accelerometer
8. Bolts and nuts
9. Cables needed

First, we need to check that the connections, cables, and components available are fully functional. To do this, we use a multimeter, which allows us to test the status and integrity of the components.

Additionally, we must use some of the stoch Arduino code to establish an initial connection with the motor, verify that it works correctly, and ensure that it has enough power to rotate the flywheel.

Here is the code. It is designed to control a servo motor that rotates the cube if it is not resting on its -Z face. The setup includes an LIS2DH12 accelerometer to detect the cube’s orientation and a brushless motor to spin the flywheel, which then generates the necessary force to rotate the cube when needed. 👍

To minimize unnecessary weight, the cube was designed to be as small as possible while still allowing enough space for unobstructed rotation. The structure of the cube was created using 3D printing.

For the flywheel, we used laser cutting, incorporating multiple open slots to position heavy bolts and nuts based on our testing results.

Here are the complete files:

For assembling the larger components, we used only bolts and nuts to securely hold everything in place. As for the Arduino, we arranged it as shown in the attached photo.

Final Product and Challenges Encountered

In the previous steps, we presented the final product. However, we encountered many challenges that we find important to share in case you want to try this project yourself.

Original Idea

Our original goal was to create a cube that could rotate in any direction without external connections. This required a battery, a motor capable of spinning fast enough to act as a flywheel, and an additional motor to change the flywheel's initial direction.

Finding a Suitable Motor

At the university, we have access to a makerspace with various resources and materials. However, finding a motor with enough force to meet our needs was tricky at first. We had to test multiple motors and fine-tune their settings to find the right one. We chose a brushless motor (drone one) since it was the most powerfull and fast amoung all.

Dimensioning the Flywheel

After selecting the appropriate motor, we focused on choosing the right flywheel.

1. First Attempt - Metal Flywheel:
2. Our initial choice was a metal washer, but it lacked sufficient inertia, making it ineffective.
3. Second Attempt - Ceramic Flywheel:
4. We then tried a ceramic flywheel, but it was not heavy enough and didn't work as expected.
5. Final Solution - Custom Flywheel:
6. We decided to design our own flywheel with an eccentric weight distribution to have better control over its performance. We first modeled it in 3D, then built prototypes using wood and acrylic. Ultimately, we found that the best combination was wood + acrylic, as it provided the desired balance of weight and stability.

To fine-tune the weight distribution, we experimented with heavy bolts and nuts, testing different configurations until we achieved optimal performance.

Rotating the Inside Mechanism

For the second motor (the one not attached to the flywheel), we encountered several problems:

1. Component Sizing: Balancing between a small cube (to reduce weight) and a large enough design to allow smooth internal rotation.
2. Stability Issues: The cube became too unstable when trying to rotate in certain positions. This was due to:
3. The flywheel being off-center
4. The motor being too weak to stabilize the movement

Because of these limitations, we had to reduce the cube’s functionality from a two-axis rotation to a single-axis rotation. This led us to modify our original idea, so the cube now moves in only one direction.

Testing the Accelerometer

The accelerometer works by measuring instantaneous acceleration. However, we faced an issue where, after the cube rotated and collided with the floor, the impact caused errors in acceleration readings.

Solution:

To correct this, we added a small delay that allowed the system to ignore noise caused by the impact and record more accurate data.

Testing the Cube (and Its Unexpected "Explosion")

When we thought everything was under control, we conducted our first real test to adjust the flywheel's weight distribution using different bolt and nut configurations. However, we encountered a huge problem:

1. When the flywheel was activated at high speed with only one bolt, it generated so much inertia that the cube flipped violently.
2. The force was so intense that none of the materials could withstand it.
3. The PLA cube and wooden flywheel broke apart, causing the cube to practically explode.

Rebuilding the Cube and Flywheel

After this unexpected failure, we rebuilt everything from scratch.

1. This time, we permanently switched to the acrylic + wood combination for the flywheel.
2. We then tested it again with improved configurations.

Battery Failure

When we restarted testing, the system suddenly stopped working. After checking every component and connection, we discovered that the battery had died—and unfortunately, we didn’t have any spares in the lab.

Solution:

To keep the project running, we decided to switch from battery power to a traditional DC power supply to ensure a stable power source.

Final Conclusions

Through this process, we encountered several unexpected technical and mechanical challenges, including motor selection, weight distribution, stability issues, and power supply failures. However, each problem led us to refine our design and improve our understanding of the system.

While we had to simplify our initial concept, the final cube successfully rotates using a single-axis flywheel mechanism. Future improvements could focus on stabilization methods, better impact absorption, and a more efficient power system.