Embark on a visual transformation journey as we integrate captivating isometric infographics into the fabric of your rebranding initiative.
Desktop Wind Tunnel: Engineering Log & Design Iteration
Introduction & Project Goals
Our goal for this project was to design and build a functional desktop wind tunnel. We wanted a reliable way to test fluid dynamics across custom 3D-printed shapes while viewing clear, real-time airflow patterns. To make the wind tunnel truly useful for analysis, we also needed to integrate sensors connected to an Arduino system to log temperature, pressure, and airflow data automatically.
While the concept seemed straightforward at first, actually generating a thick, readable stream of smoke without melting our setup or soaking our materials proved to be a huge engineering hurdle. It took five distinct phases of trial, error, and rebuilds to get a design that actually worked.
[INSERT PICTURE HERE: Overall setup photo showing the full wind tunnel, fan, dry ice chamber, and Arduino sensor setup]
Phase 1: The Cardboard & Steam Proof of Concept
We started out with basic materials to see if our initial concept was viable. We built a rough frame using a sheet of Plexiglas for the viewing section and attached a cardboard intake cone using hot glue and masking tape. To generate wind, we placed a standard household fan right in front of the intake cone.
To visualize the air moving through the tunnel, we aimed a standard facial steamer into the fan intake to channel warm vapor into the test chamber. This setup failed almost immediately. The facial steamer produced a very weak, uneven vapor stream that was almost impossible to see on camera. Even worse, the heavy moisture from the steam rapidly soaked into the cardboard intake cone. Within minutes, the cardboard became completely soggy, leading to a structural collapse of our entire front section.
[INSERT PICTURE HERE: Early Phase 1 prototype made of cardboard and tape, or a photo of the collapsed cardboard intake]
Phase 2: Upgrading to Bambu Lab 3D Printing
After the cardboard disaster, we realized that our wind tunnel needed precise tolerances and water-resistant materials. We completely shifted our approach to additive manufacturing, redesigning the entire chassis into modular parts that could be printed on a Bambu Lab 3D printer. This allowed us to quickly manufacture custom sensor housings, stream aligners, and sturdy enclosure mounts.
While the printer was running continuous print cycles, we worked on the electronics in parallel. We wired a BME280 sensor matrix to an Arduino ecosystem and wrote our first data-logging scripts so we could record environmental changes inside the tunnel during tests.
[INSERT PICTURE HERE: CAD models or close-up photo of the 3D-printed modular components hot off the printer]
Phase 3: The Smoke & Thermal Meltdown Phase
The Incense Stick Approach
Our first attempt was building a small dedicated smoke box using standard incense sticks. We quickly ran into a density problem. A single stick barely made a visible trace, and we realized we would need around 20 sticks burning simultaneously just to get enough smoke volume to track on camera.
The High-Efficiency Incense Pods
To fix the density issue, we upgraded from sticks to high-efficiency incense pods inside a new burn chamber. The smoke output was much better, but the pods produced an extreme amount of localized heat. The thermal energy burned straight through the top of the smoke box before the vapor could even travel through the guide tubes into the main tunnel. We tried wrapping the chamber in aluminum foil to reflect the heat, but the internal temperatures still warped the plastic and ruined the chamber beyond repair.
The Aquarium Air Pump Experiment
For our third attempt, we tried using a miniature aquarium air pump to actively push smoke from an external burn box through heavy-duty tubing and into the chamber. While this fixed the airflow speed, we made the mistake of leaving the prototyping setup next to the heat source overnight. The intense heat buildup caused thermal degradation across the whole assembly, melting our enclosure for a second time.
[INSERT PICTURE HERE: Photos showing the heat damage, melted smoke box, or the incense setup during testing]
Phase 4: Switching to Cold Vapor (Dry Ice Breakthrough)
Realizing that heat-based smoke would require expensive fiberglass or resin insulation that our budget didn’t allow for, we completely flipped our strategy and moved to the opposite end of the thermal spectrum: Dry Ice (CO₂).
Using dry ice brought a few unique challenges. First, dry ice sublimates quickly and cannot be kept long-term in regular freezers, meaning we had to carefully coordinate our testing windows right after buying the ice. Second, cold CO₂ vapor naturally sinks, which goes against traditional horizontal wind tunnel channels.
To overcome the physics of sinking vapor, we submerged small chunks of dry ice directly into hot water inside a sealed chamber. This reaction created a thick, dense, and perfectly consistent white fog without generating any heat. To direct the flow, we designed a custom-angled 3D-printed attachment with stabilization tubes. This nozzle forced the naturally sinking cold vapor upward and straight into our main air stream. Because the fog was cold, it was 100% safe for our plastic prints and never degraded the structure.
[INSERT PICTURE HERE: The dry ice fog generator running, showing the dense white vapor stream moving through the wind tunnel]
Phase 5: Code Optimization & Electronics Calibration
Once our physical wind tunnel was reliable and producing visible air streams, we spent our final phase refining the software side.
Our initial Arduino scripts were causing lag and occasionally missing data logs. We went back into the code to remove unnecessary delays, streamline loop execution, and adjust our sensor sampling intervals. We then ran hardware diagnostics to make sure our real-time graphs lined up accurately with what we were capturing on video.
[INSERT PICTURE HERE: Screen capture of the Arduino IDE code, serial monitor, or plotted sensor data graphs]
Next Steps & Future Modifications
Blacking Out the Interior: We plan to paint or line the internal test section walls in matte black. This will maximize visual contrast when photographing the bright white CO₂ vapor stream.
Expanding Water Capacity: We want to build a larger fluid chamber for the dry ice so we can increase the water-to-volume ratio, allowing us to run much longer continuous testing sessions without needing to refill the hot water.
[INSERT PICTURE HERE: Final hero shot of the team or the complete operational setup during a test run]
Specifications & Comparison
Chassis Material
Phase 1 (Proof of Concept): Cardboard & Masking Tape
Final Iteration (Current Setup): Modular 3D-Printed Parts (Bambu Lab)
Flow Medium
Phase 1 (Proof of Concept): Hot Water Steam (Weak & Erratic)
Final Iteration (Current Setup): Sub-Zero CO₂ / Dry Ice Fog (Dense & Stable)
Thermal Risk
Phase 1 (Proof of Concept): High Risk of Melting & Charring
Final Iteration (Current Setup): Zero Thermal Risk (Ambient to Cold Operating Temps)
Wind Speed Control
Phase 1 (Proof of Concept): Single-Speed Household Fan
Final Iteration (Current Setup): Variable High-Velocity Calibration Fan
Data Logging
Phase 1 (Proof of Concept): Manual Visual Observation
Final Iteration (Current Setup): Automated Arduino Sensor Logging