by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often
Edition: Model Aviation - 2006/07
Page Numbers: 33,34,35,36,37,38,39,40,42,44
by Bill Werwage
This bubble-canopy version was built at the same time as the
featured razorback version. It has a blunter LE and forward TE
sweep.
Another bubble-canopy version, with a slightly longer tailmoment
arm, slightly shorter nose-moment arm, and slightly
wider fuselage.
High-performance CL Stunt design with stunning scalelike
appearance captured the F2B world crown in 2004
I’VE ALWAYS LIKED the looks of warbirds. They have a powerful,
sleek, dangerous appearance. I was impressed by the semiscale warbird
CL Precision Aerobatics (Stunt) models of the early to mid-1950s. Don
Still had his Stuka, Donald Hoag was flying his Wicked Witch F-86-
styled model, and Bob Elliot had the P-40 Warhawk.
All of those designs were successful fliers and looked
outstanding, with a substantial element of scale appeal. There were
many others, but those three stand out in my mind.
Wesley Dick—one of the better fliers in the country at that
time—built a version of Don’s Stuka, placing high at the Nats and
winning many local meets with it. Wes was from my area, so I had
the chance to watch him fly on an almost weekly basis.
The stabilizer and elevator are made from sheeted balsa cores
with support ribs. Text has details on this procedure.
The P-47’s engine bay. A PA .61 fitted with a Smith and
Werwage carbon-fiber tuned pipe supplies power. Notice plastic
uniflow clunk tank. There’s lots of room in this model.
Above: Bill begins a
pattern at the 2004
F2B World
Champs. He won
the Gold Medal for
the third time! Will
Hubin photo. Right:
The P-47 displays
solid, steady
inverted flight.
Barry McCool
photo.
Bob Hunt smoothly launches the Thunderbolt for its winning
flight at the F2B World Championships. Hubin photo.
Photos by Bob Hunt except as noted
The pipe-tunnel exit. Notice the consistent wall thickness, which
is a by-product of the molded-balsa technique.
Landing-gear struts are removable and are fitted with thin sheetplywood
fairings that are attached by retainer clips and screws.
This model’s exquisite detailing sets it apart. Notice the neat inked
lines and rivet details.
Bill used an airbrush to spray the simulated exhaust stains on the
sides of the lower fuselage. It gives a great scale effect.
He was an inspiration to me because of the
quality and consistency of his flying
performances, but I also was impressed by his
model’s aesthetics. The scale aspect of that
Stuka seemed to add measurably to the
appeal.
At that time I was learning the fine points
of I-Beam wing construction and flying the
Detroiter-style airplanes. I was obsessed with
these models’ sleek overall appearance.
Although I liked the scale looks of the
airplanes I have mentioned, I was not ready to
give up on the direction in which I was
headed with the I-Beam models. I knew I
wanted to build some semiscale Stunt models,
but that would have to wait awhile. And it
turned out to be quite awhile!
Years later at the 1968 Nats in Olathe,
Kansas, those present got a glimpse of just
how far the scale envelope in Stunt would
eventually be pushed. Al Rabe showed up
with a scale-looking P-51 Mustang.
The P-51’s fuselage was wider and more
scalelike than on any previous semiscale
Stunt models. It featured dihedral, a smallerthan-
usual stabilizer and elevator assembly,
almost-scale landing-gear length, and it flew
quite well.
That interest I had in scale a few years
earlier was reawakened. At that time I was in
the process of developing a new design that
eventually became the USA-1, so, again, my
urges had to wait.
During the time I was successfully
competing with the USA-1, Al Rabe
continued his development of the scale Stunt
model. After the initial Mustang in 1968 he
produced the first of several Bearcat designs,
which was built using the normal sheet and
hollowed block-type fuselage construction. It
flew well; he captured second place at the
1969 Nats with it.
The next year Al showed up at the Nats
with an even more advanced version of the
Bearcat. It featured a molded-balsa fuselage
and numerous other molded parts. It had
working shock-absorbing landing gear and his
now-famous movable rudder. Al placed
second again, proving that the scale Stunt
model was a viable weapon that the judges
would take seriously.
During this period I was fortunate enough
for Al to ask me to fly his models. While
impressed, I was not ready to trade the USA-1
for a scale Stunt aircraft for competition.
However, I did build a semiscale
Guardian featuring all of Al’s developments
to obtain a benchmark. When spring arrived
it stayed on the bench, where it remains
today and has left a mark.
My development of the USA-1’s
aerodynamics evolved into the Junar and the
Geo-XL series. I was beginning to build a
fairly large stable of great-flying airplanes
and felt that it was finally time to give scale
Stunt a serious try.
The Design Process: Of all the scale
warbirds I have admired throughout the
years, the P-47 Thunderbolt is one of my
favorites. It has perhaps the most powerful
look of all the famous World War II
fighters. It is also a subject that had not been
modeled frequently. Charles Parrott
designed a number of semiscale Stunt
models in the 1960s, and among them was a
P-47.
In 1994 I began drawing my first
Thunderbolt. I decided to use I-Beam
construction in a wing that featured airfoils
that are similar to those of my Super Ares.
The airfoils were also similar to those
Randy Smith used in his SV series of Stunt
designs. I went to a slightly higher aspect
ratio and used full-span flaps.
The actual I-Beam in that first model
was made from balsa and carbon. Paul
Walker supplied me with the materials for it
that were essentially the same as those he
used in his I-Beam Impact design.
The construction of this Thunderbolt
marked the beginning of my using molded
shells for the fuselage shapes. I used a
minimal fuselage crutch to hold the wing
and tail assembly in alignment and added
full-form molded top and bottom shells to
get the desired shapes.
At first the mass of this fuselage
concerned me; I wasn’t used to such a huge
appearance. My fears proved to be
unfounded, however, and future
Thunderbolt fuselages grew even larger and
wider.
I have heard that there is some sort of
aerodynamic principal for the aspect ratio of
a drop tank. The 6:1 ratio (length versus
width) is said to go through the air with the
least amount of drag. The thin fuselages
commonly used in classic Stunt-model design
may not be the ideal.
Each successive Thunderbolt I have built
has had a wider, deeper fuselage and has
come closer to the 6:1 principal. I have
noticed no discernible loss of performance
because of this, and these models may be
among the most stable in turbulence and
wind I have had.
The molded fuselage is certainly more
torsionally rigid, and that is always good. The
stability may simply be a by-product of the
extreme rigidity. At any rate, the larger
fuselage hasn’t hurt!
The Thunderbolt built in 1994 and flown
in 1995 weighed 48 ounces and was powered
by a PA .40 engine fitted with a Smith and
Werwage carbon-fiber pipe. I liked the way
this model flew from the first flight. It was
smooth, consistent, and had a great corner.
This model was painted in a fictitious airracer
scheme I liked very much. It won the
Concours trophy (voted on by the
competitors) at the 1995 Nats, so others must
have liked it too! The strange thing is, many
other modelers told me I should have painted
it in a military scheme.
For one reason or another I chose to use
my Geo-XL for competition in 1996.
However, I decided that I really wanted to
pursue the scale approach and started
building two more Thunderbolts for the next
year. I designed a razorback version (the
subject of this article) and a bubble-canopy
version.
As soon as construction began on those
models I realized how much more work there
was to them than to a standard Stunt model.
I’d say there is at least twice as much labor
required to build one scale airplane! In
addition, the two P-47s had different wing
designs and some different molds for the
parts.
I finished both at the same time and testflew
them. They flew exceptionally well,
but the razorback was decidedly better. It
required almost no trimming. The bubblecanopy
version, which now has more than
3,000 flights on it, is still being trimmed!
I believe the difference between the two
P-47s is the wing design. The razorback has
a straight-TE wing and a moderately blunt
LE. The bubble version has a slightly sweptforward
hinge line and a very blunt LE. I
don’t believe the TE sweep has anything to
do with the performance difference between
the two, but I do believe that the blunter LE
and the different airfoils used in the two
make a noticeable difference in flying
performance.
Understand that even the bubble version
is now an excellent-flying airplane. I would
not hesitate to use it at the highest
competition levels.
I have designed and built an additional
two Thunderbolts, one of which I flew at the
World Championships in France in 2000.
There may even be more in the future.
As you may have noticed from the
photographs and plans sheets, the razorback
has no dihedral. In fact, only one of my
Thunderbolts had any dihedral; it had one
inch per panel (two inches overall).
Although it flew as well as any through
round maneuvers and gentle square
maneuvers, it lost symmetry of trim and feel
when turned hard outside (down), as in an
Outside Square Loop. It also felt as if it lost
some lift on outside maneuvers but had
outstanding lift on inside maneuvers.
This problem could be trimmed out with
the use of over-and-under leadout
positioning and some experimentation with
biasing the flap ratio—perhaps something
akin to the differential flap arrangement Bob
Palmer used on the Thunderbird II. I really
like the way the razorback flies with no
dihedral, and I’ll leave it at that for now.
The razorback has been such a joy to fly
and compete with and is so reliable that I
didn’t want to wear it out. I have used the
other Thunderbolts for engine-system
development and trimming ideas and
Type: Semiscale competition CL Stunt
Wingspan: 62 inches
Engine: Precision Aero .61 or .65 with Smith and Werwage carbonfiber
tuned pipe
Flying weight: 54 ounces
Construction: Balsa and plywood with molded-balsa parts and
custom fixture-built wing
Covering/finish: Carbon-fiber mat, Japanese tissue, modeling dope
This shows the
generous wing area.
The model is light,
at 54 ounces, and
appears “buoyant”
in the air. It and its
creator are
Champions in everyswitched to the Razorback just before major
meets.
I have a “flyoff” between all my models
just to make certain I’m getting the best
combination at any given time. The winner is
usually the razorback. Because of this
process that model has fewer than 350 flights
on it in its almost 10-year existence!
CONSTRUCTION
I must caution readers that this is not a
good subject for a first built-up project or a
first Stunt model. It’s best to have at least a
couple “standard” Stunt models under your
belt before you attempt the more advanced
techniques that are required to build the P-47.
As I mentioned, this model features
molded-balsa shells for the fuselage shapes.
This may be a new technique for you and it
may sound difficult to do, but it is not that
hard to learn.
A description of the process would
require a long how-to article to fully explain,
but the idea is to mold a sheet of balsa over a
form to achieve the desired shape. That is
done by wetting a custom-sized balsa blank
in scorching water and then wrapping it down
tightly around a form that is the exact shape
you desire.
Once the balsa is dry, it will retain the
shape of the mold buck on which it was
formed. Then it can be trimmed and final-fit
on the fuselage formers and mated to the
fuselage sides. This is an oversimplification,
but it does explain the basic concept.
If you have never tried this method of
construction, obtain a copy of the Straight
Form and Compound Curve Balsa Molding
Techniques DVD/video that is available from
Robin’s View Productions. In that program
Bob Hunt explains and demonstrates how to
mold straight-form balsa parts, such as LEs
for wings, and I demonstrate the entire
compound-curve balsa-molding process.
Wing: The P-47’s wing is constructed with
the Lost-Foam Wing Building System that
Bob Hunt devised many years ago. I have
built several wings using this method,
including two of my Thunderbolts and the
second Geo-XL. It is the only wing-building
system I know of that keys on the outside
shape of the wing rather than a centerline.
This yields a much more accurate finished
wing.
With the Lost-Foam system a foam blank
of the desired wing panel is positioned atop a
drawing of the desired built-up wing. The rib
positions are lofted perpendicular to the
tabletop on the front and back of the blank.
The templates are attached and the core is cut
in the normal foam-wing manner.
The spar locations are transferred from
the template to the core, and the core is
sanded. Then the spar-location marks are
connected with a straightedge and a pen. This
is done on the top and bottom of the core.
The rib positions that were marked on the
front and back of the blank are connected
using a flexible straightedge, and the rib
locations are drawn on the top surface of the
core. Then the entire plan of the wing is
accurately drawn out in the lower cradle half,
including the spar location, the ribs, and the
TE position.
The core is accurately cut apart into rib
templates from which perfectly shaped ribs
can be generated. Then the wing is built in
the lower cradle half. The cradles (portions of
the original foam blank from which the core
was cut) are just as accurate negative
representations of the wing shape as the core
is a positive shape.
A complete how-to covering the Lost-
Foam technique would take many pages of
text and photos. A DVD/video set that takes
the viewer through the entire process step by
step is available from Robin’s View
Productions.
Also available from Robin’s View
Productions are the Lost-Foam Wing
Building System components for the P-47, to
allow you to build your own wings. The
company also provides a custom building
service that constructs the wing for you using
the Lost-Foam system. Contact information
is located at the end of this article.
We are beginning to see a shift in the old
accepted building paradigms, and each of
these new methods yields more accurate,
simpler-to-build, and repeatable models.
Things are improving.
Stabilizer and Elevators: The method I use
to build the stabilizer and elevators is perhapsstabilizer-and-elevator assembly. My
assemblies typically weigh 2.25-2.50
ounces, including the horn and hinges.
Compression Pieces: The only other item
to discuss is the use of vertical
compression pieces that transfer the
vibration and rotational torque effect of the
engine from the engine-mount assembly
out to the molded-balsa shells.
It is imperative to install these pieces
when assembling the formers to the
forward section of the fuselage. The plans,
and especially the accompanying isometric
cutaway drawing, depict these parts and
their location.
I install these pieces after the upper
fuselage formers have been positioned and
glued to their corresponding formers in the
fuselage crutch assembly. Note that the
grain runs vertically on these pieces,
effectively establishing them as shear
webs.
I install these pieces so that they stick
up slightly above the tops of the formers
and then carefully sand them to perfectly
fit the outside of the formers. This also
means that the taper of the fuselage at that
point is reflected in the angle of the webs.
Now the top molded sheeting piece will
make solid contact when installed. The
message is clear; take the time to fit these
pieces properly.
The idea to use these pieces came from
my good friend Al Rabe, who has done
more research and development on the
molded shell type of fuselage construction
than anyone else I’m aware of. For sure, Al
is the “moldiest” guy I know.
Finishing: I apply .02 carbon-fiber mat on
the fuselage and the flaps for strength and
rigidity. I obtain this material from
Aerospace Composite Products.
Some modelers prefer to cover the
entire airframe with the mat, but I do not
feel this is necessary with the inherently
strong wing and tail assemblies we are
using. My old buddy Bob Gieseke prefers
to cover the entire airframe when he
finishes his models, but he’s old enough to
be carbon dated himself.
I cover the wing and tail assemblies
with Japanese tissue and then proceed with
a standard-type dope finish. I do use Aero
Fill from Aero Products to mix the filler
coat. It is the lightest and easiest-to-sand
filler material I have found.
The multicolored panel effect was
achieved by mixing light-gray dope with a
bit of silver dope. I mix several batchesintend to write a complete article on this
technique in the future.
I use a clear-dope topcoat but spray the
nose area and a bit of the wing’s LE on
either side of the fuselage with catalyzed
clear polyurethane for protection against
fuel spillage.
Flying: My Thunderbolt is currently set up
to fly with a PA .61 or PA. 65 engine, but as
this is being written I am getting ready to
install a PA .75 and test the model with that.
I test-flew one of the .75s in Bob
McDonald’s P-47 and was extremely
impressed with the engine’s power and
controllability. If you are looking for
maximum performance, I strongly
recommend that you use the .75 in this
model.
Looking back at the past 12 years I’ve been
designing and building these P-47 semiscale
Stunt models, I believe I’ve found a formula
for performance that I’m happy to be able to
pass on to others. All these models have
flown wonderfully. A great deal of this can
be attributed to the inherent rigidity of the
molded-shell type construction coupled with
the wider fuselage cross-section. It’s almost
free strength!
However, there’s more to the formula
than one factor. Bob Hunt and I often