Embedded Files
Equinox was Penn Arial Robotics' (Penn AiR) 2025 SAE regular-class, high-lift electric aircraft, capable of carrying up to 55 lbs and featuring a 15-foot wingspan.
What I Worked On
Thrust Stand
Something I focused on this year was collecting thrust data to optimize propulsion system performance. To support this effort, we built a custom thrust stand inspired by Erwan Eko Prasetiyo’s “A Simple Brushless Motor and Propeller Test Stand for Experiment from Home.” You can find the original paper here.
The thrust stand evaluates the performance and efficiency of brushless motor–propeller combinations by measuring thrust, current, voltage, and power consumption in real time. This setup enables data-driven optimization of aircraft propulsion systems by identifying the most energy-efficient configurations.
The system is centered around an Arduino Uno, using a 10 kg load cell with an HX711 amplifier for thrust measurement and an INA219 sensor for electrical monitoring, creating a compact and reliable experimental platform.
Diagram of system architectur from Erwan Eko Prasetiyo’s paper.
This is a photo of the first iteration of my load cell mount. Initially, it worked well when connected to the electronics and calibrated, but over time we noticed a significant drift in the measurements. We suspected this was caused by the load cell shifting within the base, so we designed an improved version to address the issue.
Isometric View
Side View
Back View
You can find the CAD of the improved load cell mount here.
Even with the new mount, we achieved slightly better accuracy, but the error still increased over time. We believe this was due to zero drift, which occurs when the load cell’s “zero point” slowly shifts, causing readings to gradually deviate even under no load. This can be caused by factors like material creep, mechanical settling, or small electrical instabilities. We didn’t have time to fully address these issues before starting the data collection phase, so we opted to use the Mayatech M10 Motor Thrust Tester we already had. Nonetheless, attempting to build our own thrust stand was a valuable learning experience. In the future, we hope to try again, and we plan to mitigate zero drift by implementing periodic auto-zeroing in software, stabilizing the electronics’ power supply, and carefully controlling mechanical mounting.
Thrust Data Collection
Using ECalc with a quadratic regression model, we identified two promising propeller candidates for our Scorpion A-5025-215kv motor, those being 27x13 and 26x15. We then conducted both static and dynamic tests to determine the optimal choice. For the dynamic testing, we used a fan to simulate incoming airflow.
Above are graphs for Thrust(g) vs Power(W) from ECalc and from our collected data. As we can see our collected data validate what we got from ECalc.
From our test we concluded that propeller 27x13 would give us the best performace.
Video of Static Thrust Testing
PennAiR_Thrust_Data_Collection.mp4Video of LCD Display Showing Static Thrust Data
PennAiR_Thrust_Data_27x13.mp4Image of Dynamic Thrust Testing
Power Management Wiring
Our power management system followed the regulations listed in the rule book. A dedicated receiver battery powers the control electronics, maintaining communication even if the main power system fails. A power limiter sits between the receiver and the ESC to monitor and restrict total wattage. The motor battery delivers propulsion power through an arming plug, which acts as a safety switch to prevent accidental motor spin-up. The ESC then regulates power to the motor, while a BEC delete ensures the receiver battery and ESC regulator do not conflict.
Image is from SAE AeroDesign 2025 Rules
Electronics Mounting Plate
Another component I worked on was an electronics mounting plate designed for the front of the aircraft to shift the center of mass forward, compensating for a rear-mounted payload. We used Velcro to securely mount key electronics such as the battery, receiver, and main motor ESC.
CAD of Mounting Plate
Laser Cut Mounting Plate
What the Rest of the Team Built
I can’t go into everything the team built, but here are some other highlights.
Structural Optimization through FEA
Instead of relying on intuition alone, we used Finite Element Analysis (FEA) to evaluate wing spar concepts through a custom strength-to-weight index. Four designs were analyzed, ranging from a double-column plywood spar to a rectangular hollow beam with balsa webs and poplar plywood flanges. Ultimately, we found that a rectangular hollow beam with strategically placed cutouts offered the best efficiency, achieving an index of 4.20. This approach maximized structural strength while keeping the airframe light enough to support a projected 31 lb payload.
Aerodynamic Airfoil Selection
Selecting the right airfoil was a critical decision for balancing heavy-lift capability with real-world manufacturability. We analyzed 27 different high-lift airfoils, ultimately narrowing them down to five candidates. While some airfoils offered slightly higher lift, they were rejected because their thin trailing edges would be too fragile to build using balsa wood. We eventually selected the Eppler 420, which provided the highest predicted maximum take-off weight (50.68 lbs) while maintaining a generous 14° stall angle. This choice ensured that the plane could carry the maximum possible payload without the risk of a sudden stall during the steep climb-out phase of the flight.
Eppler 420 Airfoil
Static and Dynamic Stability Analysis
Our stability analysis was centered on achieving "Level 1" flying qualities, a standard that ensures the pilot can maintain control even in turbulent conditions. To maximize payload, we pushed the Center of Gravity (CG) as far aft as possible, limiting the static margin to 5%, which reduced the need for heavy nose ballast. This CG placement also directly dictated the tip-back angle, which we optimized to 16° to ensure "smooth rotation" during take-off. If the tip-back angle is too small, the plane could tip over on the ground; if it is too large, the pilot would struggle to lift the nose. By balancing the aft CG limit (14° tip-back) with the forward CG limit (18° tip-back), we ensured the aircraft remained stable on the runway while still being responsive enough to rotate and climb the moment it reached take-off speed.
Technical Drawing
Results
At SAE Aero Design West 2025 Competition we finished second among U.S. teams and nineth in the world. Here is a video of a successful flight from competition day.
PennAiR_Competition_Flight.mp4Page updated
Google Sites
Report abuse