Here a project of a Coanda Effect Saucer (CES) stabilized with a 9D0F IMU. The CES UAV, propelled by an electric engine, uses the Coanda effect to take off vertically, fly, hover and land vertically (VTOL). There is no big rotor like on an helicopter and the flight is very stable and safe for the surrounding.

The Coanda Effect has been discovered in1930 by the Romanian aerodynamicist Henri-Marie Coanda (1885-1972). He has observed that a steam of air (or a other fluid) emerging from a nozzle tends to follow a nearby curved surface, if the curvature of the surface or angle the surface makes with the stream is not too sharp.

I use my firmware AutoStab v4.0 installed the ArduIMU+ V2 flat with a special mixer for this device.

Stay tuned on this blog, more to come soon,


E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • hye :) I'm a student from south korea. I want to make your coandadrone for real. I have all of componets, but I couldn't find specific blueprints. so I wonder if I can get your specific blueprints. also whether I can use different components from yours. please answer ASAP

  • Hi, i have some questions:
    1. does it contribute to the airflow around the surface , if you would decrease the cap between rotor and main body ( or some kind of channel, to increase the velocity)
    2.if you change the shape of the body, so that the surface angle is Not so steep , would this some kind of influence the lift production?
    3. is there a better Solution as 3 gyros?
  • I know your original post is a little old now Jean-Louis, but it's good to see that, independently of you, I've been able to come to the same theoretical conclusions - I'm going to experiment in a slightly different direction, but the basics on show here will still apply in general.

    Did you ever do any further testing?



  • Bonjour M. Naudin,

    I would like to have your advice for indoor drone (no GPS) and low light end dusty environment
    Two options:
    1) Quadcopter – with optical Flow sensor
    2) Coanda saucer+ OptiFlow + with light and camera and within
    Best configuration for the control ... 9DOF or Gyro… lifting capacity ...your best  up-dated  configuration if possible please.
    Merci beaucoup ! Dan

  • With a winged aircraft, the wings do nothing except generate lift or hold fuel.  Therefore it's important to maximize wing surface area versus weight.   With the Coander Effect, the entire body generates lift, and payload can be put inside that body, with no need for wings.  That's where the extra efficiency come into play, I think.

  • The lift for the CES UAV is a function of the velocity of the air close to the contour surface. The JLN curve of lift v airspeed approximates to lift = k x airspeed squared in the range 10m/s to 35m/s (for a given contour surface area).

    This doesn't answer the thrust-lift query, but it does show that obtaining higher airspeeds, not necessarily higher thrust, is the key to obtaining higher lifts.

    The airspeed at the outer rim of the CES UAV is about 0.30 to 0.35 x the airspeed at the propellor duct outlet. There is thus residual airspeed (thrust -lift?) from the air moving downward below and off the contour surface.  It is not known whether this contributes to lift.  It does show however that part of the energy of the moving air is dissipated over the contour surface producing lift, and part of the energy is dissipated off the contour surface, which may or may not be contributing to lift.  This may be part of the explanation of why thrust does not equate directly to lift. I have speculated on whether the formula for the lift generated by an aerofoil moving in free air, lift = k x airspeed squared x area of wing (viewed from above) can be used to calculate the lift (not forgetting temperature effects).   

    I am re-investigating the issue of stability/flap control, the subject of an earlier blog. I met someone at last who has successfully flown a replica of the original `600mm' JLN CES but with a 9x5 propellor (instead of the 10x6/10x7), different motor and different (HobbyKing) gyroscopes for flap control. I am now trying to replicate his success.

    Stability.  There were earlier discussion on stability, where it was suggested that as the centre of gravity was above any `centre of lift' on the JLN CES, it was basically unstable. I built a variation of the JLN CES where the centre of gravity is below any possible centre of lift (but with no flap control) and it is still unstable, flipping over on take off (experiments showed that the variation on the JLN CES has ample lift for flight). This suggests that the proposition that instability is caused by the CofG being above the CofL may not be the whole story.

    George B Taylor

  • hi. i am prefer CES UAV , i am study simulation the body CES UAV in ansys , you can help me ,please.

    i also build a CESV1.2 same as you ,the thrust of engine is 1200g ,but when i assembleengine on the UAV , i measure the lift is 900g , i don't know why? you can tell me why, thank you very much

  • I built a 1:1 replica of Mr Naudin's NO1A, the precurser to the CESv1.2.  I did not acheive stable flight, the craft tilted over immediately on take off. Questions for Mr Naudin.

    Should the craft be stable after lift off to 1.5m with fixed vanes fitted but without the pitch and roll flaps or yaw control vanes fitted? I built an 800mm model with fixed vanes without the flaps or yaw control fitted which flew to 1.5m before tilting over.  I found that the hull of both crafts did not rotate with only the fixed vanes fitted.

    Are the pitch and roll flaps capable of correcting for tilt?  I have an impression that they can't. Experiments on a 1/8th sector suggested that lift doesn't change much between flap IN and Neutral position, but falls between Neutral and flap OUT.  Thus if a sector with a flap drops, the flap goes from Neutral to IN  without producing additional lift to correct the drop, but if the sector rises, the lift drops when th flap goes from Neutral to OUT.  

    I have been trying for a long time to build a craft stable in flight and aim to continue, videos show it can fly ina stable mode.

  • Hey Jean-Louis, just wondering, what is the software that you're using to do the CFD analysis that you're showing for this post.  I'm interested in testing out a model that I've made in Blender.

  • Hi i m Puneet from India , i m doing my B.E in mechanical engineering , i would like to do a project on analyzing the Coanda effect over different curved surfaces and also would like to compare the Lift/Thrust generated in a Coanda UAV to that generated in an RC Helicopter. If anyone of you could please help me in going about this project it would be of great help. Thank you

This reply was deleted.