1.6m Nausicaa Glider (Mowe)

About a month ago, I decided to get into RC aircraft.  I had been a flight simulator "armchair pilot" for twenty years, and had built RC car kits in the past, so it seemed like a logical new hobby to investigate.  I built the FT Simple Cub and had a successful maiden flight, which I attribute to its excellent, stable design, tips I learned from Flite Test videos, and my previous simulator experience.  

Nausicaä's glider, as seen in the manga.

Then, while watching some videos of the amazing FT Sea Angel inspired by Porco Rosso, I thought it would be fun to use the construction techniques I had learned building the Cub to design my own Miyazaki-inspired model.  For me, there was really only one choice.  I had my heart set on a craft from Miyazaki's 1984 anime, Nausicaä of the Valley of the Wind.  In this movie, the title character pilots an ultralight, no-cockpit, jet-powered flying wing.  It is called the Möwe, which is German for "seagull," due to the resemblance of its anhedral wingtips to those of a soaring gull.  Although it looks fanciful, it has in fact been built as a full-scale, piloted plane: the OpenSky M-02.  Others have created RC versions using different build techniques.  I thought that it must therefore be possible to build one FT-style, from foam board.

OpenSky M-02.

I started with a few sketches of the general layout, using many images from the Nausicaä anime and manga for reference.  I also looked at images and video of the OpenSky M-02.  Several design points jumped out at me.  The original glider, as depicted in the anime, has almost no vertical stabilizer.  On the M-02, two vertical stabilizers hang under the middle of each wing, replacing two elliptical pod-like structures in the original.  I knew my model would need some yaw stability, but I wanted to keep a look more faithful to the anime.  So I copied the general design of the stabilizers on the M-02, but gave them a more elliptical shape.The final working sketch, from my notebook.

Both the original and the M-02 use a small jet engine for propulsion.  I decided that in the interest of saving a few bucks, I would design my fuselage to fit the tractor prop power pack from my Cub.  I reasoned that later on, if the prop-powered version was successful, I could design a different fuselage around a small EDF unit.  All the electronics were from the Flite Test Power Pack B "Radial," using a 3-cell LiPo for power.

Structurally, I took a similar approach to many FT planes: a boxy fuselage with a hatch on the top, and a wing held on with rubber bands.  The inner span of the wing would use a box spar design, with a general airfoil shape scaled up from the Cub, but featuring a slightly swept leading edge and increased dihedral.  For simplicity, a stepped airfoil was used for the wingtips, with flat pieces of foam serving as vertical stabilizers.  

Cross section mock-up of center span showing spar structure and placement of vertical stabilizer.

I decided to use whatever nearby available craft foam I could lay my hands on to build the prototype model.  I ended up building it from Elmer's brand foam.  This was a learning experience.  I discovered that Elmer's foam, though durable, is about twice as heavy as the Adams foam that is used in Flite Test models.  While this made the model unexpectetdly heavy, it also meant that the intial tests and inevitable crashes would be less damaging.

Drawing parts with a ruler, triangle, and compass.  The book is "Model Aircraft Aerodynamics," by Martin Simons, from which I learned all I know of the subject.

I designed the plans using FreeCAD from https://www.freecadweb.org/.  Then I used hand drawing tools to transfer them to foam board as accurately as I could, and cut out my pieces.  The airframe went together in a weekend's worth of effort, requiring only three sheets of foam for the first model, with a wingspan of 1.1 meters.

Prototype airframe.  This fuselage was destroyed in a crash.

To determine roughly where the center of gravity of the plane should be, I used the online calculator at https://www.ecalc.ch/cgcalc.php.  Based on the results, I chose 75mm from the leading edge of the root of the wing as a conservative starting point.  This happened to be right underneath the front of the main spar.  Achieving this balance was more difficult than I had anticipated.  Because of the shape of the wing, the natural center of mass of the airframe was actually several centimeters behind the necessary location.  I ended up using two batteries placed side-by-side to get enough weight far enough forward in the nose.  These I wired in parallel, so at least I would benefit with a longer flight time.  I was careful to place a piece of scrap foam between the batteries and the firewall, in order to prevent the motor mount screws from piercing the batteries.  After several chuck attempts and adjustments, I took my rebalanced design out back and threw it into the wind.  It seemed to hang in the air as if it wanted to fly.  I decided it was ready for a powered flight.

Original wing with redesigned fuselage, showing parallel battery setup and nose reinforcement.

Under power, the craft flew remarkably well, although faster than I would have liked.  It required more throttle than my Cub to stay airborne, around 70 percent of full power.  Achieving a smooth landing was difficult due to the steep glide slope.  I attributed these characteristics to the low aspect of the wings.  I knew of the principle that a longer, narrower wing provides a better lift-to-drag ratio, and therefore glides better.  Unfortunately, in my efforts to make the model resemble the anime, I had given it short, stubby wings, which required more speed to generate enough lift.

Video showing 1.1m prototype in flight.

So I loaded up FreeCAD once more and stretched the wings out to 1.6 meters, without changing their chord measurements.  This turned out to be almost exactly 1/6 the span of the real OpenSky M-02, which is about 9.6 meters.  The wingtips were re-shaped for a sleeker, more bird-like and anime-accurate look.  The new wing was constructed from Flite Test foam.  Despite the increased wingspan, the use of the lighter foam resulted in a lighter aircraft overall than the smaller model.  Each wing used a sheet and a half of foam, meaning that the complete 1.6m model used four sheets.  

1.6 meter wing with 1.1 meter prototype.  Note the experimental pilot figure.

The fuselage also had taken quite a beating in testing.  In one crash, the rubber bands holding the wing on had ripped their wooden supports out, along with part of the fuselage.  I changed the design with a reinforced nose area that could better absorb rough landings while also helping with airframe balance.  My power pod had also gotten crumpled in a crash and would no longer hold the motor at a neutral angle, so I built a new shorter power pod from the heavier foam.  This eliminated any extra foam located behind the CG.  

Video of 1.6m Mowe in flight, demonstrating improved glide and landing ability.

The results in the air were exactly what I had hoped for - a slow moving glider instead of a rocket.  After getting a feel for the gentler dynamics, I found that the longer wing would cruise comfortably at only 60% throttle.  This dropped to 50% or less when flying into some light wind.  The only drawbacks I found were a greater tendency to slip sideways while in a turn and a general decrease in maneuverability.  Belly landings in my field became a cinch.  Later, I added a folding prop and spinner combination to curb its tendency to break the prop when landing.  I can now walk out to the field, hand launch my Möwe, and glide it back to my feet again and again for fifteen minutes before recharging the batteries.  It's great fun.  Everyone who sees it remarks on how much it looks like a bird in flight.  A version of the plane powered by a 50mm EDF is currently in development, and I am looking forward to adding an FPV system.

Experimental EDF version of the 1.6m Mowe, shown atop the first prop-driven model.

I wrote this article not just to share the plans for my Nausicaa glider design with other fans, but also my journey from simulator jockey to successful model aircraft designer.  I hope you've enjoyed it, and that some of you may build and fly my design, or be inspired to create your own.  Please contact me with questions or comments.

Thank you,

-Dok Cosmac

Experimental EDF Mowe in flight.  Note decreased efficiency compared to the prop version.

Below are my blueprints for the 1.6m prop-driven version of the Mowe, as well as the experimental fuselage to fit a 50mm EDF which uses the same wing.  You will need four sheets of foam, preferably Flite Test or Adams brand.  These images should be printed at 750x500mm for the correct scale.  The right wing image is mirrored to make the left.  This model uses Flite Test standard control horns and firewall, and works well with the Power Pack B.  The EDF version uses an FMS brand 50mm, 11-blade unit.  Both versions run on a pair of 1300 mAh 3S batteries wired in parallel.  There are no instructions, so please don't hesitate to shoot me a message if you are trying to build this and find yourself puzzled by my drawing.  I may create a build video at some point in the future if there is enough interest.