[Table of Contents] People old and young enjoy
waxing nostalgic about and learning some of the history of early electronics. Popular Electronics was
published from October 1954 through April 1985. As time permits, I will be glad to scan articles for
you. All copyrights (if any) are hereby acknowledged.
Popular Electronics article archives.
|
I purchased a couple batches of vintage Popular Electronics magazines off of eBay
for use on my engineering website, RF Cafe; however,
upon scanning through the pages I was pleasantly surprised to find that many articles
on radio controlled airplanes were included. The 1950s and 1960s were relatively
early in the R/C sport, and such things were still considered a novelty. Of course,
today the toy shelves of even Walmart are full of R/C products. Kids today take
them for granted... as I suppose my generation took for granted
Erector Sets and Lincoln Logs.
Anyway, I have begun scanning and OCRing (Optical Character Recognition) some
of the articles and posting them here on AirplanesAndRockets.com.
This first Popular Electronics, from the December 1954 edition, was written by none other than
Bill Winter. Bill was the editor for
American
Modeler and
American Aircraft Modeler, and
Model Airplane News for many
years.
Flying the R/C Plane
So you've built a model plane for radio control-will it fly true and handle
easily? This is how you find out.
By William Winter
OF THE many important factors involved in successful radio control flying, none
is more important than the airplane and the builder's ability to balance and adjust
the model (i.e., trim) and fly it properly. A good airplane is as necessary as a
good transmitter or receiver; good flying technique as important as ability to tune
the electronic equipment. Much has been said about radios. Let's, for a change,
take a close look at the plane.
It is not necessary to be an aerodynamicist,
or to bone up on aero text books. But let's face it - an understanding of what makes
an airplane stable and controllable is as necessary as knowing plus from minus.
For the first time in a magazine of this type, we will outline these "trade secrets"
in as painless a manner as possible.
In this enlightened air age it is well known that an airplane's wing lifts because
of the partial vacuum created above its curved upper surface as the craft travels
forward through the air. More properly put, there is a pressure differential between
the air over the wing and the air under it, with the result that the wing tends
to lift, or move upward. It is also widely understood that the lift offsets the
weight of the machine, and that the thrust of the propeller offsets the resistance
(drag) of the plane created by its forward passage through the air.
An airplane has movement about three axes (see Fig. 1): one is the longitudinal
axis (fore and aft) similar to the barbecue spit, around which the plane rolls;
two is the directional axis, like the pivot in a weathervane, around which the plane
turns; three is the spanwise axis, like the fulcrum of a child's seesaw, around
which the plane noses up or down.
The plane is kept stable, or level, about these three axes, by certain arrangements
of its surfaces. See Fig. 4. The dihedral angle, or upward tilt of the wings, is
primarily responsible for the plane's resistance to rolling and spiraling into the
ground. Therefore, do not depart from the kit maker's plans and specifications.
Stability about the directional axis is governed by the proper size of the vertical
fin; stability on the fore-and-aft axis is achieved by the placement of the wing
relative to the center of gravity (CG), or balance point of the machine, and the
placement and area of the horizontal tail surface. This is perhaps an oversimplification,
but it serves our purpose adequately.
Directional and lateral (rolling) stability does not concern us here because,
if the plane is accurately built from a kit, carefully lined up, and free of warps
(warps, incidentally, are taboo, as they make the plane ungovernable) it is automatically
stable in these two respects.
Our big concern is the spanwise axis (the business of the seesaw). Consider that
the lift of the wing, if placed forward of the center of gravity, would tilt upward
its end of the seesaw (the nose) if not counter-balanced by some other force. This
balancing force is provided by the lifting action of the stabilizer, or horizontal
tail. Thus, the greater force of the lifting wing, acting through a small distance
or moment arm, about the center of gravity, is offset by the smaller lift of the
horizontal tail acting through a greater distance.
Let's assume that the plane is tail heavy - the nose tends to rise toward the
point at which normal air flow over the wing breaks down with sudden loss of lift
resulting in a stall. Our seesaw can be restored to balance in several ways. The
lifting force of the wing may be decreased by slightly reducing the angle at which
the wing meets the air, the wing may be moved back slightly (reducing its moment
arm), the center of gravity may be moved forward with the same effect; or, tackling
the problem from the other end, the lift of the stabilizer may be increased by adding
to the angle at which it meets the air stream.
Naturally, all of these possibilities may not be available - if the wing rests
upon a cabin top, it can hardly be moved forward or backward, but changes in CG
position may be made by relocating heavy bat-teries, and the angle of either wing
or tail may be altered by thin shims, such as thin balsa wood, matchbook covers,
etc., placed between the wing or tail and the fuselage. In Fig. 5, a shim is shown
being inserted between the tail and fuselage. Permanent shims may be cemented in
place after adjustments have been made. Of course, if the plane is nose heavy, tending
to dive, the corrective measures are reversed. So much for aerodynamics.
Before launching into the actual trimming and adjusting of that new model, it
would be wise to explain one very important stability factor which even model plane
fans very often do not appreciate. Regardless of changes and adjustments to the
plane's surfaces, always maintain a slight angular difference between the setting
of the wing and stabilizer. The wing should be set at a slightly greater positive
angle of incidence, at least one degree, but not more than 3 degrees greater than
that of the tail (work this out to a fractional part of an inch to facilitate measuring).
Because the tail works at a lesser angle, it
will reach the stalling point after the wing has stalled, continuing to lift, and
will always exert an increasing leverage as the plane noses up, to restore it to
level flight. Should the tail be set at the greater angle, or sometimes even at
the same angle, the plane will stall more severely and completely (because the stabilizing
effects of the tail suddenly vanish whim it stalls). Moreover, at high speeds, as
in a dive or spiral, the extra lift from the tail may overbalance the seesaw and
put the ship into a death dive from which recovery is impossible. Many modelers
forget this when they balance a tail-heavy plane by excessively increasing the positive
angle of the stabilizer - within reason it is safe to so adjust the stabilizer,
but never forget the need for angular difference.
The Hand Glide
The first test for any new plane is the hand glide. Perform this test in calm
weather or in the slightest possible breeze. Remove the propeller to avoid breakage.
If the model is a simple one with simple equipment, it is safe to leave the radio
and associated equipment in place. If it is a complicated job with vulnerable, heavy
radio, simply remove the radio provided it is located close to the center of gravity
of the plane; otherwise replace it with an equal weight.
Pick a grassy area where landings and bumps will be cushioned. Hold the ship
two or three inches behind the CG position, level it with the nose pointing at the
hori-zon (see Fig. 3), and run slowly without releasing the plane. Repeat this times,
running faster arid faster, until you can feel the point at which lift is exerted
as shown in Fig. 2. Now point the nose of the plane slightly below the horizon,
at a spot about 50 feet away. Run at the same speed as before, and as you release
the plane, impart the very slightest push. You may have to repeat this step a few
times to approximate the gliding speed of the ship.
Never throw the plane; unless it is very heavy, as it will rear upward, irrespective
of trim, to stall and 'then dive into the ground. It is better to launch the ship
too slowly at first which, in grass, means a hard landing at the worst.
Observe closely the manner in which the plane glides. Is the plane tail heavy
(stalls) or nose heavy (dives)? Usually, all such tendencies can be ironed out by
increasing or decreasing, respectively, the angle of incidence of the stabilizer;
if not, the CG position should also be altered, forward if tail heavy; backward,
if nose heavy. When the plane glides correctly, it should follow a straight flight
path from the hand until it lands two-points upon the wheels. It should not float
or appear to hover or describe a pretty arc toward the ground, whereby the plane
seems to flare out nicely for a three-point landing. This looks good, but the plane
would stall badly in flight, especially with the engine running.
You want a fairly fast and flat glide. If there is a breeze, the model should
evidence "penetration," boring straight ahead to a landing and not be buffeted aside
by the wind. If the glide path curves to right or left, adjust the rudder slightly
to remove the circling tendency but be sure that a warp isn't present in the wing,
stabilizer, or fin, and that all the surfaces are accurately lined up. If the wing
is slightly askew it will make the plane turn. If one tip is higher than the other,
or one stabilizer tip higher than the other, the results will be the same.
Now that we've gone through all the motions of the first trial glide, we will
have to check the plane on powered flight. Of course, if the model is a glider,
you will merely repeat the hand launched glide tests until the ship is completely
checked out and then, you're ready for a glide aloft. Next month, the first power-on
test flight will be discussed, and many of the points mentioned will apply equally
Well to the glider, which uses the same type of flight controls as the gas model.
END
Posted May 24, 2021 (updated from original post on 7/3/2011)
|