Sparrow Article & Plans
September 1973 American Aircraft Modeler
visitor David M. wrote to request this article Sparrow RPV, from
the September 1973 edition of American Aircraft Modeler. The Sparrow
was the forerunner to many of the world's modern marvels of technology
in the RPV - now called Unmanned Aerial Vehicles (UAV) - world.
Why "unmanned" rather than "remotely piloted"? Simple, it is because
many of the aircraft now fly autonomously for at least a segment
of their missions; therefore, it is its own pilot. Author Dave Scully
could not have known at the time he was describing the future of
everything from mass produced large, prefabricated aircraft, high
displacement engines, 14"-plus propellers, and the installation
of wireless sensors in aircraft. A large (80" and 160" wingspans),
heavy (25 lbs. & 200 lbs.), with 2 hp & 12 hp engines, and
16" props was a novelty back in the day, and was mostly the realm
of government-financed research projects and wealthy private modelers.
Today, dozens of manufacturers offer affordable (not by me, though)
giant scale models and a complete line of accessories. Use of more
than 5 servos was virtually unheard of, but now even some sailplanes
have 6 or more. For less than $200, you can now buy a complete wireless
video system, radios include remote telemetry data on system health
as well as airspeed, altitude, engine or battery temperature, and
much more. Hobby Lobby was selling a complete wireless system (PilotView
FPV 2400) fro a few hundred dollars until the FCC filed a violation
order. Any day now I expect to see a model scale synthetic aperture
radar (SAR) system for 3-D terrain mapping and remote infrared (IR)
night vision. The military already has it.
Say, why didn't
I get to work on this kind of a cool project when I was in the USAF,
like MSgt. Scully here?
Principal members of the Teleplane Team are (left to right):
Dave Scully, Don Lowe, and Jim Cline.
Sparrow flies with 24 lb. of video equipment
and both Ross 4 and Ross 6 cylinder engines. Here Jim Cline
puts the starter to it while Dave Scully holds on.
This colorful plane waits patiently for flight duty. Construction
utilizes mostly fiberglass and foam/plywood flying surfaces.
Top view shows that a great deal of flap area is available for
slow flight work.
Underside, Ross 4 installed at this time. Spreader bar at tail
is streamlined K&S tubing. Rudder linkage passes through
There's plenty of room in here for equipment installation. Remember,
it must have a payload for flight to keep the CG forward.
One servo pictured here is just for steering the nose gear.
A parallel servo handles the rudders.
The pod hatch removed. That Sony TV camera, in its present location,
could handle weather sampling equipment, radar, a laser, etc.
Lighting all the plugs at once is easy with the springs seen
here. Engine runs both ways, so a pusher prop is not required.
Note sump tank at firewall.
Tail boom at wing saddle shows location of elevator servo here,
rudder servo in other boom. This part of boom is balsa; rear
part is fiberglass.
Bottom of wing shows receiver location, flap servo and aileron
This RPV is more than an interesting project designed for carrying
movie cameras. It is a multi-purpose research or practical application
industrial tool for carrying substantial payloads in a pusher configuration.
by Dave Scully
Sparrow is a special purpose, remote-controlled
model being used by the Air Force at Wright-Patterson Air Force
Base to develop RPV (Remote Piloted Vehicle) technology. This design
represents the groundwork of a concept which may someday see the
RPV emerge as an operational weapons system performing missions
which are presently relegated to manned aircraft.
of such a concept are numerous. Aircraft no longer limited by man's
physical tolerances would be more maneuverable in air combat situations.
Hazardous missions such as low level recon, weapons delivery, and
target marking could be performed by expendable RPVs. Other advantages
include a high degree of mobility, lower development, production,
and maintenance costs.
What role does the RPV play in civilian
aviation? Remote-controlled models have been used for some time
as an aid in the development of full-scale aircraft, and more recently
as a means' of obtaining special aerial sequences for the film industry.
Universities and research agencies have utilized radio-controlled
models for various projects, among which the development of the
Hill autopilot is probably the best known.
for special purpose RPVs may exist in such areas as weather sampling,
high altitude research, and environmental control. We will undoubtedly
see a growth in engine development, control systems, and RPV design
to meet these future requirements. Sparrow is one design which could
be adapted for use where requirements dictate a model capable of
carrying a substantial payload or a model suitable for aerial photography.
Sparrow was designed by a team headed by Aeronautical Engineer
Raymond Fredette of the Flight Dynamics Laboratory at Wright-Patterson
Air Force Base, Ohio. Mission requirements for the full-size vehicle
called for an aircraft which would carry up to 100 lb. of payload
and cruise in the range of 80 to 100 mph. Design considerations
included an unobstructed forward view for video experiments, and
the capability of accepting a variety of payload configurations.
As a result, the design presented evolved as the most desirable
configuration for planned remote piloting experiments.
model presented in this article is actually a half-size version
of the Fredette design, which was built as an engineering aid, and
proved to be extremely useful as a low-cost method of elevating
performance, and developing fabrication techniques relative to its
fullscale counterpart. A Sony video system was later installed,
and basic piloting experiments via video were performed with excellent
Sparrow exceeded our expectations in its flight
characteristics, and its suitability as a test bed for future experiments.
Apart from a lack of readily available engines for models of this
size, I can honestly say that we encountered no major problems,
either in constructing or in flying Sparrow. Our plane has flown
well with a Ross 4 and even better with the Ross 6.
A model of this design as published, or possibly as an enlarged
version, might be useful to a modeler interested in performing experiments
of his own, or in taking in-flight movies. As I mentioned before,
inexpensive available engines pose a problem and possible alternatives
might consist of using two engines (60s or 80s) mounted on the booms
as in the P-38. Little would be gained by shrinking the design,
as the wing loading would increase with a corresponding decrease
in payload capability. Generally speaking, the larger the model
(wing area), the more efficient the design will be in terms of its
Fig. 1 shows a comparison of the
half-scale to the full-scale that may be helpful to those who would
want to determine their own requirements for an enlarged model.
||7.5 sq. ft.
||30.0 sq. ft.
||3 lbs. sq. ft.
||7 lb. sq. ft.
If a five-lb. movie camera was the desired payload, two 60s
on the halfscale would probably be sufficient to achieve good takeoff
and flight performance.
A very general description of the techniques used in fabricating
this model are all that will be described, as it is anticipated
that the experienced builder will adapt this design to his own proven
methods of construction. Suggested changes for a simplified method
of building this model would be to use a balsa wood box construction
for the pod and the booms, and the more conventional method of control
surface found on most RC models. The flaps could be eliminated on
the half-size version, as we found we could land quite slowly without
them. However, for larger models or flying fields that require steep
approaches and the slowest possible touchdown, by all means include
Pod - Fiberglass construction was utilized
to gain maximum internal space to accommodate payload requirements
consistent with minimum cross sectional area. As I mentioned, balsa
wood construction could easily be substituted and past construction
articles on scale models of the OV-10A would be a good source of
ideas, however I leave this to the ingenuity of the readers. Fiberglass
construction is not really difficult; the following is a brief description
of the steps we used in constructing the fiberglass pod. Experiment
with small parts such as wheel pants, cowls, etc. before tackling
a complete fuselage. Once you learn the process, it's hard to go
back to the glue and pin method of construction in terms of the
savings in cost and time.
The original pattern was carved
from urethane foam (Pro Foam would be suitable for this purpose)
and sealed with K& B coating resin. The resin is sanded and
reapplied as necessary to obtain a smooth surface. The pattern was
designed to split along the thrust line and the two halves were
mounted on plywood boards. The pattern is then coated with a good
paste wax and buffed thoroughly. A coat of vinyl separator is applied
by brush or spray and allowed to dry. The mold is constructed by
brushing on a gel coat (commercially available) to the prepared
pattern and allowing it to set up. The mold is then built up to
the required thickness (1/8" is sufficient) with layers of fiberglass
mat saturated with polyester resin. The mold is allowed to cure
at least 48 hours before removing it from the pattern and is then
washed with water to remove the separator. The steps involved in
making the finished pod are the same as for the mold except that
fiberglass cloth is used instead of mat. The pod on Sparrow is laid
up with one layer of seven-oz. cloth and a doubler of two-oz. cloth.
The rear of the pod is reinforced with seven-oz. cloth to handle
engine vibration. The excess cloth above the parting line of the
mold can be trimmed off with a sharp knife just after the resin
sets up (about 30 min.). The pod halves are left in the molds at
least 24 hours and then removed, washed and held together with masking
tape. They are then joined with a one-in. strip of cloth applied
to the inside with polyester resin. Plywood formers are installed
using a putty made of polyester resin and silica powder as an adhesive.
The forward hatch is cut out last with a razor saw and a flange
installed to accept hold-down screws.
Wing - The wing sections
used are NACA 23012 at the root and NACA 4412 at the tip. The leading
edge radius has been reduced and the forward coordinates have been
modified to allow the wing to operate within a Reynolds Number range
of .88 to 1.5.
Standard foam wing construction techniques
are used with the following exceptions. Access tunnels are cut through
the cores using the guides provided on the wing templates. The
wing is covered with 1/64th plywood using Bestine rubber cement
as an adhesive on foam-to-wood bonds and Formula 2 epoxy on wood-to-wood
If the wing templates are lined up on the datum line,
you will notice that the tip has effective wash out (e.q., the tip
has negative incidence as compared to the root section and insures
that the tip will stall last). This factor contributes to the outstanding
slow flight characteristics of Sparrow and it is important to insure
that the wing is built accurately. We found it advantageous to cover
the wing in the same manner as a Formula I racer. Each wing panel
was cut from a separate foam block which was parted to form an upper
and a lower mold; the wing core was then placed in these forms while
wrapping. Since the wing has no dihedral, the wing panels were joined
together using the lower foam blocks as an alignment guide. The
flaps are connected via a fiberglass arrow shaft pivoting on plywood
bearings. The flaps are glued to the shaft with epoxy and pinned
with toothpicks. The ailerons are connected using 90° bellcranks
and 3/16" dowel pushrods running through the access tunnels provided
in the wing. The extension harness for the boom servos are also
routed through these tunnels. The wing center section is reinforced
with fiberglass cloth and epoxy after joining.
booms are fabricated from Styrofoam and covered with 1/64th plywood
with the exception of the section at the wing saddle, which is built
up and sheeted with 1/8" balsa plank. A 1/16" balsa spline is glued
down the top center of the boom and a 1/16" plywood plate is installed
below the vertical stabilizer to add rigidity and serve as a mounting
plate for the elevator bellcrank. The 1/64th plywood covering is
dampened with a mixture of ammonia and water to prevent splitting
while wrapping and is bonded to the foam with rubber cement. The
plywood is lapped on top of the boom and the seam covered with a
1/2" strip of 1/64th plywood secured with epoxy. The forward section
of the boom is reinforced with fiberglass cloth and epoxy.
Tail Surfaces - The foam and balsa surfaces are constructed in much
the same manner as the wing. Note that the left vertical stabilizer
requires an access tunnel for the elevator linkage. The rudders
are mechanically interconnected using a nylon linkage housed In
a strut made from K&S streamline tubing.
and Miscellaneous - Sparrow has been flown with both the Northfield
Ross 4 cylinder and 6 cylinder engines. The Ross was chosen primarily
to reduce vibration effect on the video equipment and proved to
be satisfactory, although a bit expensive for the Sunday flier.
The Ross was not designed to be flown as a pusher but since it is
a reed valve engine it can be operated in either direction. To date
our Ross engines have been operated in reverse rotation using 14/6
and 16/6 tractor props respectively without suffering any ill effects.
Woodcraft Mfg. was the source of our props and I believe
they can supply the larger props in either the tractor or pusher
The fuel systems we use on our project
aircraft are strictly homemade (Jim Cline specials) and the normal
procedure is to find a container that will fill the available space.
In this case we used a plastic aspirin bottle which was modified
for fuel lines by using grommets (similar to the Tatone stick-a-tube
method) to seal the tubing where they enter the tank. Sparrow's
fuel system is not designed for prolonged inverted flight and it
proved satisfactory to run a feed line from the lower rear of the
tank and a vent line from the top of the tank to complete the system.
With the Ross engines it proved necessary to run a larger than normal
feed line (about 1/4" I D) to the engine and cobb up a brass tubing
manifold to feed the individual carburetors.
The main landing
gear strut was bent from 1/4" drill rod and retempered, however,
two pieces of 5/32" music wire bound together with brass wire and
soldered would do just as well. The tandem wheels (three-in. Du-Bro
slicks) on the main gear have worked well in the absence of larger
commercially available wheels.
Finishing - The model was
completed using K&B products. All wood was sealed with coating
resin and the model primed and ·finished with SuperPoxy paint.
Checkout - Our ground checks prior to flying consisted of
the following. CG located between 21.5" and 24" aft of the nose.
Control deflections: ailerons 150, rudder 300, elevator 250 up,
300 down, and flaps 00-450 range. The radio was given a thorough
range check both with the engine running and static. It was found
that a capacitor (Erie Red Cap PN 8101-050-651-10ZM) had to be added
on the Pro Line receiver to eliminate problems caused by the long
leads to the boom servos. One cap was added on each channel between
the decoder output and ground and the receiver was then retuned.
We then went one step further and secured the aircraft to the top
of a station wagon and charged up and down the runway checking for
flutter and the odd chance that we could get a 4000-lb. payload
off the ground. Finally, deciding that there was nothing left to
tighten, adjust, or paint, the batteries were charged and we headed
for the flight line.
with a gross weight in excess of 25 lb. (ten-lb. payload) requires
a long ground roll. With 100 of flaps lowered, the model accelerates
approximately 150 ft. before the nose wheel is lifted off, about
15 feet later the plane is airborne and climbs out quite rapidly.
In flight, Sparrow is stable and relatively easy to fly. I'm not
going to "put you on" and say that the design will fly the entire
FAI pattern with a sick engine. In fact, with a three-lb. wing loading
on a model of this size it should fly like a nervous brick, but
the point is, it doesn't. Sparrow's flying and handling characteristics
compare more closely to a light aircraft than to a model, although
forgiving in most circumstances, she won't tolerate rough handling
and is definitely a model for the experienced flier. The control
response is positive without being touchy and stalls are gentle
and predictable. (You can almost feel the controls get mushy.) With
full flaps and down trim the plane can be slowed to a crawl with
no indications of a stall.
can't give much of a comparison as to the flying qualities at different
weights since Sparrow has to carry payload or ballast to balance
properly. The lightest weight the model was flown at was 22.5 lb.
on the first test flight.
The only aerobatics that have
been performed to date have been rolls and one short field landing.
That is to say, I landed about 50 ft. short of a long field when
I found you can't stretch a glide with a dead engine. The airplane
suffered only minor damage but my pride suffered more by missing
a 7000-ft. runway with a seven-ft. model.
On landing, Sparrow's
sink rate is somewhat higher than one is used to (as you might have
gathered), but touchdown is quite slow and, as I stated before,
flaps have not been found necessary as a landing aid on this size
I would be interested in hearing from groups or individuals
who would benefit from a special purpose model such as described
in this article. Let me know the payload requirements and purpose
of its use. Sufficient interest in this area might warrant production
by a kit manufacturer of fiberglass and foam components that would
satisfy a variety of needs.
<click for larger version>
<click for larger version>
The AMA Plans Service offers a full-size
version of many of the plans show here at a very reasonable cost. They will scale the plans any size for you. It is always
best to buy printed plans because my scanner versions often have distortions that can cause parts to fit poorly. Purchasing
plans also help to support the operation of the Academy of Model
Aeronautics - the #1 advocate for model aviation throughout the world. If the AMA no longer has this plan on file, I
will be glad to send you my higher resolution version.
Try my Scale Calculator for Model Airplane Plans.
Posted April 25, 2011