The first real test of my late-night propeller design and analysis code
It started as a single line of code in September 2025.
Fast forward to today. On May 15th, 2026, have tested on a stand our first manufactured set of fixed-pitch convertible propellers for a lift + cruise aircraft.
They are real. Which I feel as a massive, unbelievable achievement for me - because in this convertible airplane project, with fixed-pitch propellers, we have to cheat physics at least twice:)
To achieve this ambitious goal, my code handles:
Fast forward to today. On May 15th, 2026, have tested on a stand our first manufactured set of fixed-pitch convertible propellers for a lift + cruise aircraft.
They are real. Which I feel as a massive, unbelievable achievement for me - because in this convertible airplane project, with fixed-pitch propellers, we have to cheat physics at least twice:)
To achieve this ambitious goal, my code handles:
- Automated aerodynamic design and optimization: it starts with flight modes, geometric constraints, and engine specs. It spits out optimal chord distribution, twist angles, and airfoil shapes.
- Lifting line theory for design: to minimize induced losses, the silent killer of efficiency.
- ML + CASADi for airfoil optimization at three sections.
- BEMT analysis core (Blade Element Momentum Theory).
- Stress analysis: because a beautiful prop that explodes at max RPMs is just shrapnel waiting to happen.
- NX CAD and CADFlo CFD automation: the code talks to Siemens NX and rebuilds the entire parametric model. Then it loads that model into CADFlo for CFD analysis.
- Full visualization and export: graphs, specs, 3D models, and stress heatmaps.
The "Compromise Propeller" Philosophy
Most props are optimized for one flight condition: takeoff, cruise, or max speed.
For our convertible airplane, we designed a multi-regime fixed-pitch propeller. The code was tasked with finding the geometry that delivers maximum efficiency across the entire flight envelope, for a prop manufactured from wood. It had to guarantee required thrust at every single mode - hover, transition, and cruise at various speeds and altitudes, with minimum possible torque and adequate deflection under load.
And now, here is the validation of my code at a static condition.
For the lift propeller, my BEMT analysis combined with the J-ref method for static extrapolation delivered the following median errors:
For the lift-cruise propeller, the median errors are:
Looking forward to dynamic wind-tunnel tests of the cruise-lift prop!
Most props are optimized for one flight condition: takeoff, cruise, or max speed.
For our convertible airplane, we designed a multi-regime fixed-pitch propeller. The code was tasked with finding the geometry that delivers maximum efficiency across the entire flight envelope, for a prop manufactured from wood. It had to guarantee required thrust at every single mode - hover, transition, and cruise at various speeds and altitudes, with minimum possible torque and adequate deflection under load.
And now, here is the validation of my code at a static condition.
For the lift propeller, my BEMT analysis combined with the J-ref method for static extrapolation delivered the following median errors:
- Thrust: 6.2%
- Torque: 9.8%
- Efficiency: 18.5%
For the lift-cruise propeller, the median errors are:
- Thrust: 24.7%
- Torque: 15.4%
- Efficiency: 10.6%
Looking forward to dynamic wind-tunnel tests of the cruise-lift prop!
