Senior Project.
My two teammates and I are designing drones for a quadcopter speed-record attempt. We haven’t built them yet. I’m working on the CAD model and studying the airflow around the airframe.
Explore my 3D airflow study
draw it. test it. change it.
Follow the airflow.
Rotate the 3D model, follow the wakes behind its four propellers, and switch between airspeed and density.
Preparing the 3D airflow study…
Green marks incoming airGray marks Mach 1 or higher
Drag to rotate Scroll to zoom
Based on my CAD model. Converted from my STEP model.
Download the 3D modelWhat this simulation shows
The airframe uses three precomputed potential-flow solutions. Changing the airflow angles blends those solutions. The velocity view shows estimated local airspeed in meters per second.
The shortened propellers come from the STEP model. The original is an approximate 7-inch version reconstructed by extending the shortened blades. Both represent a nominal 15-inch pitch. The 7-inch swept circles overlap on this airframe, so the original is shown for comparison.
Each propeller adds a wake that accelerates, contracts, swirls with its direction of rotation, then spreads and fades downstream. The two diameters are compared at the same illustrative load of 20 N per propeller. That load is a model setting. Blade rotation and tracer motion are slowed for viewing.
The density view estimates local compression and expansion from airspeed using an isentropic relation with a simplified allowance for energy added by the propellers. Warm colors indicate denser air and cool colors indicate less dense air. “Density rise at a stop” is the ideal density increase if the incoming air slows to rest. Gray marks where the local estimate reaches Mach 1 and the subsonic approximation breaks down.
This study uses standard sea-level inlet conditions and leaves out blade-resolved flow, shocks, turbulence, separation, and cooling. Use it to explore flow patterns. Drag, lift, and achievable top speed require further analysis and testing.
NASA on potential-flow methods · NASA on propeller momentum theory · NASA on isentropic flow
My work on the airframe
I’m using CAD and AirShaper to work out the body shape, the path for cooling air, and how the printed parts will fit together. We plan to keep the motor setup the same when comparing airframe versions so we can see what changes with the shape.
Printing rigid and flexible parts together
We plan to use a dual-nozzle 3D printer to print a rigid carbon-fiber nylon tail section and a flexible TPU tip in the same build. The nylon holds the airframe’s shape, while the TPU can flex on touchdown. I’d model a joint with interlocking features, like dovetails, so the two materials are held together by their shape as well as whatever bond forms during printing.
A drop-shaped body
I’m shaping the shell like a drop, with a rounded nose, room for the hardware near the widest point, and a gradual taper toward the tail. The taper gives the air a better chance of staying attached as it flows toward the rear. A smaller low-pressure wake would mean less pressure drag. I’m comparing versions in AirShaper to see which shape gives us the best trade-off between drag and room for the hardware.
Cooling inspired by the P-51 Mustang
Our motor ducts borrow an idea from the Mustang’s radiator scoop. A forward-facing inlet captures air, the duct widens to slow it as it passes over the motor and picks up heat, and a narrower rear outlet can speed it up again.
The Meredith effect is the idea that this heated exhaust can offset some of the drag from bringing in cooling air. Our first goal is to keep the motors cool with as little added drag as possible. How much drag the outlet could recover is something we still need to test.
Parts we’re considering
- Motors
- AMAX 2826, AOS Supernova 3220, and T-Motor 3120. We’ll compare thrust, current draw, and temperature with the same propeller to see which motor gives us the most usable power.
- Propellers
- APC 7×15 propellers with a nominal 15-inch pitch for high forward speed. Shortening the diameter from 7 inches to about 6 inches lowers the tip speed at the same RPM, giving us more room before compressibility effects become significant.
- Electronic speed controllers
- Four APD controllers, one to drive each motor. We’re considering the 120F3 and 200F3 and will choose based on the pack voltage, current draw under load, and how much cooling each controller needs.
- Battery
- One option is two 7S, 1,600 mAh packs in series for a 14S setup. A higher voltage can deliver the same electrical power at lower current. We’ll weigh that against the pack weight and check that the motors and controllers can handle the voltage.
- Airframe and spinners
- The printed X-frame supports the motors, and the carbon-fiber nylon shell gives it a smooth outer shape. The TPU tail tip can flex on landing. Printed or machined aluminum spinners would smooth the airflow around the propeller hubs.
- Flight controller
- The flight controller adjusts motor speeds to keep the drone stable. Its blackbox logs let us track vibration and see whether changes to the frame or propellers affect the tune.
How we’ll build and test it
Before printing the final shell, I want to check how the electronics fit, leave a path for cooling air, and check the clearance between the propellers. I’ll estimate tip speed at our expected RPM and check the balance of the shortened blades.
On the bench, I want to compare thrust, current, temperature, and vibration before choosing a motor and propeller setup. I’ll compare motor temperatures with and without the duct at the same load and flex a sample of the printed tail joint to check how well it holds. I’d sand the shell seams flush and check the flight controller logs again after bonding the shell, since that can change how the frame vibrates.