Suggestions for improving the DC-3
and How to fold the amazing OmniWing

by Dave Coffin

Refinements for the DC-3 Paper Airplane

Tuning a DC-3 for straight flight does not require cutting ailerons into the wings. It does require a clean, straight fold along the leading wing edge. You can achieve this in Step 20 by creasing around a thin, rigid ruler. Or you can make a light crease in Step 20 and later, at Step 30, pull it forward and press it flat. Don't flatten the entire wing, because perfectly flat wings tend to stall in the slightest turbulence.

Cut Step 26 at a slant to create a pointed rudder in front of the horizontal stabilizer. It's a tiny rudder, but the long tail will give it a lot of leverage.

Don't use the rudder just yet. Toss the plane and notice how it behaves at minimum speed. If it turns right, fold the left wing down more to shift the ballast to the left. Only when the plane "stalls straight" should you adjust the rudder for straight flight at higher speed.

Get this right, and you may be walking very long distances to retrieve your plane!

To reduce drag, do not crease and rip in Steps 22-24. Instead, measure and cut 1 1/4 inches from the edge. Taking inspiration from the OmniWing, you may tape over gaps in the wing. This also makes the wings stiffer, which is good.

Thin paper (less than 24lb) is not stiff enough for a full-size plane. Cut a sheet into four equal pieces and make mini-DC3s. Owing to the extreme stability of this design, minis fly almost the same as larger planes.

DC-3s are great fun at children's parties. Fold them ahead of time and mark them with names, letters, or numbers. Stack bodies and tails separately in a cardboard box. Tails are easily damaged by hard landings and people using them as handles, so pack plenty of spare tails.

Finding the optimal Angle of Attack

Step 28 says to make the leading edge of the wing "slightly higher" than the trailing edge. The precise angle of this fold has a huge effect on flight performance. The following experiments all used 24-lb inkjet paper 8.53 by 11.00 inches.

Measured from the centerline crease, the wing is 22mm high at the leading edge and T mm high at the trailing edge. The horizontal distance between these two points is 88mm, so the resulting angle of attack is:

A = arctan((22mm-T)/88mm)

Leaving room to insert the tail, the effective range of T is 16mm (A = 3.9°) to 22mm (A = zero). I've built and test-flown otherwise identical planes at T = 16.0 18.0 19.0 19.5 20.0 20.5 21.0 and 22.0 mm.

Low-T planes pitch up sharply at high speed, recover more quickly from stalls, glide more slowly, and crash less often than high-T planes. When thrown hard, low-T planes perform inside loops, so they cannot be thrown very high. T = 19mm (almost exactly 3/4 inch) is a good setting for children and novices.

Planes with T > 19mm will unfold along the centerline crease when thrown very hard. The wings briefly turn to a negative angle of attack, and the plane dives into the ground. This won't happen if you secure the centerline crease with a small drop of glue or flap of paper.

For the longest flights, set T between 20.0 and 20.5 mm. Facing into the wind, throw the plane hard, straight up. If the plane spins during the climb, install a fresh tail section, re-flatten the wings (paying special attention to the leading edges), and try to throw the plane straight.

The amazing OmniWing paper airplane.

After about six failed attempts, I finally built a a working OmniWing. Wow. It really soars. The shape is just like the wing of a high-performance hang glider. All gaps are covered with clear tape for a smooth, low-drag surface. This is, as I mentioned before, the only paper airplane _without_ a stepped wing surface.

Tape also makes possible a rigid wing that twists to a lower angle of attack at the swept-back tips. These wingtips provide continuous lift through a nose-up stall and recovery.

It's more difficult to launch than a DC-3, though not entirely beyond the ability of a 5-year-old. It can be launched gently for a simple glide, or power-launched for stunts, but the throw must not impart any yaw motion. The plane can absorb a small yaw by turning with it, but big yaws (e.g. after stalling out of a failed loop) are catastrophic.

Dave Coffin 7/5/2011

Inventor Mike Kelsey explains how to build one here:

How to fly the OmniWing:

Thank you Dave for your valuable suggestions. Dave Coffin's web site

| Return to Home Page |

Revised: July 31, 2011