All projects
Aerospace / Research

Project Saturn

Project Saturn is a circular-wing VTOL that makes its own lift by blowing air across itself instead of moving through it. A central fan draws air from the bottom, through the body and throws it outward over a ring-shaped wing, and the saucer form falls straight out of that physics. It is a Master of Engineering thesis and a direct continuation of Hovering Through New Horizons.

Three-quarter render of Project Saturn: a dark saucer body with a blue domed canopy and curved ducts around the rim, on a light studio background. Front-on render of Project Saturn at eye level, showing the ring-shaped wing and the curved ducts beneath the rim. Underside render of Project Saturn, showing the central fan and the ring of curved stator vanes inside the circular wing.

01 Reverse the wing

An ordinary wing only works once the whole aircraft is moving fast enough. Project Saturn breaks that link: the wing stays still and a fan supplies the airflow, so lift no longer waits on forward speed. Feeding that flow evenly is what forces the wing into a full ring driven from the centre, and the round body follows from there.

Loading 3D model

02 The Air Force tried this

The disc was the result, not the target, and the resemblance to an old idea only became clear afterwards. In the 1950s the US military chased the same shape twice: Project 1794, a classified saucer meant to reach supersonic speed on the Coanda effect, and the VZ-9 Avrocar (link), which actually left the ground. Both fought the same instability and neither was tamed. Arriving at their shape from clean physics, decades later, reads as validation rather than influence.

Black-and-white photograph of the Avro VZ-9 Avrocar hovering just above a runway beside a hangar.
Colour photograph of the Avrocar from above, showing the central turborotor and the US Air Force and US Army markings on the disc.
Two technicians working on a preserved Avrocar inside a museum restoration hangar.
Labelled cutaway engineering diagram of the Avrocar, showing the central turborotor, three Continental J69 engines, the exhaust duct, and fuel tanks.

03 How it works

  • Draw. A centrifugal fan pulls air up into the centre of the body from below.
  • Spread. A radial duct turns the flow outward and feeds it evenly around the rim.
  • Accelerate. The air speeds up as it passes over the curved circular wing.
  • Lift. Faster air above the wing drops the pressure there, and that pressure difference is the lift.
  • Straighten. A ring of stators takes the spin out of the flow before it leaves.
Cutaway render of Project Saturn showing the airflow: air is drawn up into the central fan from below, then thrown outward across the circular wing, traced by red arrows.

04 Designing the wing

The wing section was chosen, not assumed. Twenty-five airfoils were screened at a Reynolds number of 70,000, the regime this scale model operates in, and ranked on lift-to-drag ratio. The SG6043 came out ahead of the common NACA 4412, and it set the profile carried through the rest of the design.

Airfoil-analysis sweep of 25 candidate sections at Re=70,000, showing lift, drag, lift-to-drag and moment plotted against angle of attack.
Lift-to-drag ratio against angle of attack for the SG6043 and the NACA 4412, showing the SG6043 peaking higher.

05 Proving it in CFD

Confidence was built one step at a time. The simulation worked up through six stages, from a 2D airfoil to the full ducted circular wing, each one checked for grid independence before moving on. The velocity contours below show the flow accelerating over the wing, which is where the lift comes from.

ANSYS mesh of the domain around the wing section: a fine structured grid hugging the surface that opens out into a coarser radial far-field.
ANSYS velocity contour over the wing section, with the flow accelerating to roughly 39 m/s across the upper surface, shown in red.

06 Driving the air

At the centre is a centrifugal fan, the part that does the real work. It pulls air up from below and throws it outward off curved blades, so the flow leaves faster and wider than it arrived. The blades are forward-curved, the shape best suited to a small fan moving a lot of air, and they were sized to fit the craft: set the body first, then shape the blades to feed it.

CAD render of the forward-curved centrifugal fan at the centre of the craft, which draws air up from below and throws it outward over the wing.
CAD render of the stator ring: seven curved vanes sweeping from the central hub, which catch the spinning air off the fan and straighten it.

The fan leaves one problem behind. The air comes off it spinning, and that spin is wasted energy that adds nothing to lift. A ring of stators corrects this, meeting the flow at the angle it leaves the fan and curving it back to straight, gently enough that it never separates from the surface. The vane count is a deliberate choice: 7 vanes against the fan's 24 blades, set so the two never align into a resonance that could shake the craft apart. Those same vanes also act as structure, tying the top and bottom halves of the body together.

Eye-level view of the assembly: the fan seated inside the stator ring in the lower half of the body.
Side cutaway of the full craft, transparent canopy over the body, showing the fan and stators in place with the curved vanes turning the flow.

07 Building and testing it

The design left the screen as a physical scale model. It sits on a wooden stand over a scale, so the lift it makes shows up as a drop in measured weight, with an anemometer reading the air it moves and a brushless motor spinning the fan. Lift and mass flow were recorded against fan speed, up to 5,800 rpm.

The craft's white duct body being worked on at the bench during the build.
The assembled scale model mounted on its wooden test stand, set up for a lift run.

08 The honest result

The numbers are reported as they came. Theory predicted 14.12 N of lift, CFD predicted 3.36 N, and the physical model delivered 0.31 N, measured at 5,800 rpm, past its design speed. The concept lifts, but it is still far from efficient. Stating that gap plainly is the point, and it marks where the next version starts: variable wing angle, control surfaces, and a tighter duct and stator design.

09 Previous prototypes

The final model was the last in a line of earlier builds. These are the prototypes that worked out the shape and the internals along the way.

Early CAD render: a domed saucer body with a rounded top cap and short curved blades, no outer ring yet.
A later render adding the full outer ring and a ducted fan stage with stators around the central dome.
A side-on render of the ringed saucer with the domed top cap and ducted blades visible underneath.
A low-angle render of the ringed saucer, showing the dome, the duct ring and the blade stage together.

The full thesis carries the sizing, the six-stage CFD validation, and the experiment in detail.

Read the full thesis (PDF)
Back to all projects Get in touch