The new racers would be called the R series Super Sportsters and were to be the fastest land planes in the world.
Granville and Miller had no choice but to push as far out on the leading edge of technology as the modest resources of the Springfield Air Racing Association (SARA)and their little company would allow. Both ships were to be designed, built and test flown within the six months that remained before the 1932 National Air Races in Cleveland on September 1.
They were, for all practical purposes, forced to use the wire braced, wood, tube and fabric construction of earlier Gee Bees, So from the start they accepted the fact to simply live with those materials and seek an edge elsewhere.
They were also wedded to P&W air-cooled radial engines, so a blunt nose was another given that they would have to work around. It was a widely known engineering fact that the ideal shape for a body moving through a fluid was teardrop with a ratio between length and diameter of 3 to 3.5 to one, so Granville and Miller did a study to determine how close they could come to the ideal teardrop that had an engine, wings, landing gear, cockpit and a tail attached to spoil the airflow around it.
1/10 scale mahogany models of the R1 and R2 Supersportser designs were tested in the New York University wind tunnel by Pete Miller, Granny Granville, and the famous aeronautical engineering professor Alexander Klemin to determine the wing positioning and stability coefficients.
Demonstrating the scientific nature of their research, the Granvilles constructed 1/10 scale mahogany models of the R1 and R2 Supersportser and they were subjected to tests in the New York University wind tunnel. A year later, on May 18, 1933, Granny and his chief engineer, Howell W. Miller, presented a paper to the Society of Automotive Engineers in New York City describing the design and construction of their now-famous racers. Obviously they felt that they had two sound, viable contenders for the prizes offered in the upcoming races.
Wing location was very important, and tests were made of mid-wing and low-wing configurations, plus a position about half way between the two.
This three quarter position was found to create the least amount of drag, so it was selected.
Tests were also run on various engine cowling shapes, wing filet sizes and gear leg fairings. Probably the test that had the most impact on the design was the stability tests.
Engineers and pilots have notably always differed in opinion on the amount of stability fast airplanes should have. Engineers favor a lot of stability, maintaining that the less that a pilot has to control the airplane, the less drag will be created by the control surface and trim system displacement. Pilots on the other hand want a highly responsive airplane, even if that means cutting stability margins to the minimum. Granville and Miller decided to side with the pilots, knowing that a very sensitive airplane would result. They expected to gain performance by cutting down things such as tail volume to a bare minimum and depending on the pilots skill to make up for any handling shortcomings. This was perfect rational for a high performance racing plane.
The R-1 and R-2 were very soundly constructed aircraft. The structures were stress analyzed by Pete Miller, he stressed the entire aircraft for 12 Gs positive. After extensive research and testing the R-1 and R-2 were built. Only the best materials were used and workmanship was incredible. The wings were covered with mahogany plywood and then covered again in with "balloon-fabric" for more strength. The fuel caps were now enclosed inside the fuselage and the windscreen was constructed of 3-layered shatterproof glass.
Perhaps the most unusual feature was the pilot position just ahead of the vertical fin. With a fuselage of just 17 ft. 9in. in length an extreme aft location for the cockpit was necessary to balance the engine and Smith adjustable propeller up front. The R-1 and R-2 were considerably larger than the previous model Z: wing span, 25ft.length, 17ft.9in.wing area, 100 sq.ft. empty weight, 1,840 LB; with 160 gal of fuel and 18 gal for oil.
The R-1 and R-2 were nearly alike except that the R-1 had the larger, more powerful Pratt & Whitney R1340 "Wasp" engine. It has a diameter of 50-5/8" and a frontal area of 13.98 sq. ft. Nine cylinders, super charged with a 12:1 supercharger and devloping around 800hp. The R-1 was built specifically for the Thompson Trophy race (pylon racing).
The R-2 was built for the Bendix Trophy race (cross country racing), but it was thought that it would be a strong contender in pylon racing as well. The 1932 R-2 was powered by the very same Pratt & Whitney R985 that had powered the 1931 Model Z, rated at 535hp. The R985 engine had a diameter of 45-3/4" and a frontal area of 11.42 sq.ft. This allowed the R-2 to have a more streamlined cowling and come closer to the ideal teardrop shape.
The R-2 had an empty weight of 1,796 pounds and a gross weight of 3,883 pounds. With a wing area of 101.9 sq.ft, the wing loading was 38.1 pounds per sq. ft. The power loading with 50 gallons of fuel on board was 4.47 pounds per horsepower, but with full fuel, it was up to 7.32 pounds.
Other than the engine, the biggest difference between the two aircraft was that the R-2 had two fuel tanks totaling a 302 gallons versus a single 160 gallon tank in the R-1. The R-2 had a 20-gallon oil tank versus 18 gallons on the R-1. The R-2 had a fixed tail wheel while the R-1 had steerable tail wheel. The R-2 also had streamlined running lights in the wings and on the tail for night flights.
On August 10, painter George Agnoli finished the fancy scalloped red and white paint job on the R-1. On the 12th it was rolled out of its hangar to sit gleaming in the sunshine. Final adjustments postponed the first flight until the following day. On August 13, shortly after nine o'clock, Russ Boardman took off his coat, slipped on a parachute and flew the R-1 to the Bowles Agawam field across the Connecticut river. The performance figures were exhilarating. The R-1 had hit 240 mph without half trying and Boardman felt confident that 300 mph was well within reach.
Robert and Granny Granville were at Bowles when Boardman landed from the first test flight. As they opened the door of the plane, Boardman looked up at them, grinned and said, "You boys sure build airplanes." His only complaint was that the ship bobtailed during landing approach and apparently did not have enough fin area. Work was immediately begun to rectify that problem by adding two square feet of fin area and an increase in rudder area to match the added fin.
On Tuesday, August 16, Boardman was severely injured as he
spun a Model E Sportster into the trees on the Carew street side
of the Springfield airport as he was flying to Agawam to complete
the tests on the revamped R-1. With two weeks remaining before
the start of the races, neither aircraft had a pilot. Applications
flooded the Granvilles from every pilot in the country who had
any ambitions of appearing in the Cleveland races. Finally, on
August 22, Lee Gehlbach was chosen to fly the R-2. He would ferry
the R-2 to Burbank to fly the Bendix race from there to Cleveland.
Oil temperature problems were already starting to show up which
would plague the R-2 over the next few weeks. Earle Boardman,
Russ' brother, was prepared to fly SIark and Ed Granville to Los
Angeles and Kansas respectively to supervise the refueling of
the R-2 during the Bendix race.
A stroke of fate interjected a new name into the Gee Bee saga.
At Wichita, Kansas, Jimmy Doolittle was test flying his Laird
"Super Solution", which had been extensively modified
for the 1932 races. When he found that he couldn't get the wheels
down, he was forced to belly his aircraft in, eliminating it from
further competition, but emerging unhurt. On August 27, when it
was apparent that Russ Boardman would be unable to compete in
the National air races, Granny made telephone arrangements with
Doolittle for him to fly the R-1. On August 28, Doolittle arrived
at Springfield. While everyone expected him to take a turn or
two around the field to familiarize himself with the new aircraft,
dubbed by the press as "The Flying Silo", he simply
climbed in, headed west and never altered his course. Less than
two hours later the Granvilles received a telegram stating simply,
"Landed in Cleveland O.K., Jim." Lee Gehlbach indicated his confidence in Number 7 when he told
members of the press, "Number 7 is the most wonderful handling
ship I've ever flown. Doolittle added his praise of Number 11
by stating. She s got plenty of stuff. I gave her the gun for
just a few seconds and she hit 260 like a bullet without any change
for momentum and without diving for speed and she had plenty of
reserve miles in her when I shut her down. Only five planes were entered in the 2,369 mile Bendix race from Burbank
to Cleveland and one of them, Clare Vance in
his twin boom flying gas tank, dropped out of the race, because of a fuel leak.
That left Gehlbach in the R-2 and three Wedell-Wiliams racers piloted by Jimmy
Wedell, Jim Haizlip and Roscoe Turner. Making good time, Gehlbach's first fuel
stop was Amarillo, TX and he got underway with minimal time on the ground.
Shorly afterwards, however he noticed oil on his windshield. By the time he
reached Illinois it was completely covered with it, so he made an unscheduled
stop at Chanute Field in Rantoul, IL, hoping to fix the leak.
It turned out to be a cracked oil line that could not be repaired in time to
meet the race arrival deadline, however, so he filled the oil tank, removed the
canopy and flew on to Cleveland open cockpit.
The time on the ground at Chanute Field ruined his chance to win or finish
high in the Bendix and he ended up last of four racers that made it to Cleveland.
The prize money was quite generous, however, and Gehlbach made $1,500
(equivalent to about $30,000 today).
At Cleveland the oil leak was repaired and the canopy reinstalled and the R-2
was entered in the Thompson race and qualified at 247.339 mph.
Gehlbach ran in the 1,000 cu. in free-for-all but cut the first pylon, had to
recircle it, and ended up third behind the Wedell-Williams of Haizlip and Wedell
at 183.731 mph.He won $262($5,240 today).
Doolittle fared better in the Thompson trophy race. On September
5, nursing a hay fever attack, he blazed around the pylons at
a winning speed of 252.686 mph. Among those who witnessed his
victory were his tu-o sons, ten year old John and James Jr. Lee
Gehlbach finished fifth, and Bob Hall, the former Granville engineer,
was sixth in the field of eight in his Springfield "Bulldog"
racer. While at Cleveland, Doolittle set a world's land plane speed
mark over the regulation F.A.I. threekilometer course that had
been set up for a series of speed dashes sponsored by the Shell
Oil Co. On August 31, the 35year-old Doolittle averaged 293.193
mph on four runs over the speed course but this did not qualify
as an official record since he didn't have a barograph in the
plane to confirm that he was below the required 162 feet (50 meters)
during his runs. He subsequently made four more runs on September
1, averaging 282.672 mph, just .77 mph short of that required
to claim a new record. On his final run it appeared to the horrified
spectators that he was about to brush some trees just north of
the field. Later Jimmy said, "I was nowhere near them. I
must have been at least four feet over them." At the Eastern States Exposition in September of 1932, Jimmy,
in speaking of Z. D. Granville. said, "He builds a most excellent
airplane and it was the airplane that did the job." Finally,
in a letter dated September 7, 1932, and addressed to Granville
Brothers Aircraft. Doolittle commented, "Just a note to tell
you that the big Gee Bee functioned perfectly in both the Thompson
trophy race and the Shell speed dashes. With sincere wishes for
your continued success, I am, as ever, Jim.
In the Thompson the R-2 and the rest of the field of
racers were blown away by Jimmy Doolittle in the Gee Bee R1, once again Gehlbach
ended up behind the three Wedell-Williams. He finished in fifth place at an
average of 222.1 mph and won $500 (about $10,000 in today’s dollars) Doolittle
won $4,500 by winning the Thompson.
S.A.R.A. president Russell Boardman had recovered from his accident of a year
earlier and was going to fly the R-1 in the National Air Races out in Burbank, CA.
Russell Thaw was picked to pilot the R-2. Jimmy Doolittle had retired from air
racing as a champion and world-famous celebrity. Lee Gehlbach, in 1933, was
flying the No. 92 Wedell-Williams.
The 1933 National air races were to be held in Los Angeles
from July 1 through July 4, with the finish of the transcontinental
Bendix race from New York as one of the highlights. After being
received at City Hall on June 6 by Mayor O'Brien of New York,
Boardman, now recovered from his earlier injuries, and 22-year-old
Russell Thaw, in the re-engined R-1, and the modified R-2, left
Floyd Bennett field on the morning of July 1, 1933 along with
Roscoe Turner, Lee Gehlbach, Jim Wedeli and Amelia Earhart. Thaw
took off at 5:52 a.m. and Boardman followed, being the last to
depart. Thaw used almost the entire length of the field, dragging
his tail wheel as he struggled to get his heavily laden plane
into the air. Boardman, with a lighter load and higher horsepower,
made a perfect take-off and streaked westward. Boardman and Turner had announced that they would refuel in
Indianapolis while the others would let their fuel consumption
govern their landing places. Preliminary plans called for Thaw
to land at St. Louis and Amarillo, but his high rate of fuel consumption
caused him to land at Indianapolis. Turner arrived at Indianapolis
at 6:06 a.m. and within ten minutes he was once more winging his
way westward. Thaw was the next to land. Contrary to many published
reports, he made a perfect landing. On all earlier Gee Bees the
Granvilles had manufactured their own shock struts. Now their
racers were equipped with a commercially manufactured strut. In
making a rapid 180-degree turn to get back to the refueling area,
one of these struts collapsed and the left wing tip was damaged
near the outer aileron hinge as it struck the runway. Since it looked as if the damage could be readily repaired,
the plane was wheeled into a hangar and work was begun to restore
it to flying condition. Boardman was the next to arrive and he
chatted with Thaw as his plane was refueled. Then he took off
with 200 gallons of fuel on board. At about 40 feet in the air,
he lost control of Number 11, and it flipped on its back and crashed,
fatally injuring Boardman, who died on the morning of the 3rd,
leaving a wife and four-year-old daughter, Jane. Thaw was so shaken
that he withdrew from the race at that point. The other entries were also plagued with misfortune. Lee Gehlbach,
flying a Wedell-Williams racer, was forced to land near New Bethel,
Indiana, with a clogged fuel line, crashing through a fence but
emerging unhurt, and Amelia Earhart, in a Lockheed Vega, was forced
down in Kansas. Roscoe Turner won the race in 11 hours and 30
minutes, picking up the $5,050 first prize plus $1,000 for setting
a new East to West record. Wedell, who finished second, won $2,250. It was a horrible blow for the Granville brothers. In a few
minutes they had lost both planes and one pilot. Robert Granville
recalls, "I guess it was the point where our luck started
to go bad." Boardman's brother, Earle, also a pilot, was with him when
he died. On July 4, he flew Russ' body east to Hartford, Connecticut,
stopping to refuel at Syracuse, New York. On July 6, 1933, Russell
Boardman was laid to rest in the Miner cemetery in Middletown,
Connecticut, while six planes circled overhead and dropped flowers.
Among those present was John Polando of Lynn Massachusetts, with
whom Boardman had made his long distance flight to Turkey. The Gee Bee R-2 was repaired at Springfield in 1933 and while being flown by
James Haizlip. Making his third landing, Haizlip found himself floating too far
down the 200 ft. runway, he kicked in the right rudder to side slip and kill the
airspeed, but this caused one wing to stall, and the Gee Bee R-2 cartwheeled
down the runway, rolling itself into a ball. Haizlip escaped with only bruises
from the saftey harness.
During the fall of 1933, Z. D. Granville, Howell Sliller, and
Donald Delackner opened a consulting engineering office in New
York City in the hopes of continuing development of certain commercial
projects such as four, six and eight place airplanes. As far as
racers went, they were left with the shattered remains of the
R-1 and R-2. From the remains of the R-1 and the R-2 the Granvilles
built another racer, christening the resultant plane "Intestinal
Fortitude". It was known as "International Supersportster"
Model R-3. The plan was for Roy Minor to fly it in the Chicago
races of 1933. After "Intestinal Fortitude" was assembled. Granny
planned to fly to San Antonio to deliver the prototype Gee Bee Model E Senior Sportster that he
owed to a customer in that Texas City. En route. he planned to
visit Florida and the Mardi Gras in New Orleans. Approaching a
landing at Spartenburg. South Carolina, on February 12, Granny
suddenly noticed that there was construction in progress and his
only safe landing area was blocked by two workers, As he pulled up, his engine coughed and died and
he spun in from 75 feet. Zantford (Granny) Granville unfortunately died en route
to the hospital. Shortly after Zantford's death, the Gee Bee organization was
sold at a sheriffs sale.
But it was Delmar Benjamin and Steve Wolfe who finally put an R-series Gee
Bee back in the air.
Delmar Benjamin came up with the funding and Steve Wolf had the
building/restoration shop and the expertise to build the Gee Bee R2 replica.
After careful study and visits with Pete Miller the project began on January
1, 1991 and was test flown December 23, 1991.
Delmar's airshow performances in the Gee Bee R-2 replica proves that this
aircraft is fully controllable, capable of hard aerobatics and straight-line
speed.
He has completely destroyed the myth that the R-series airplanes were the
worst airplanes ever built. And Delmar Benjamin is probably one of the most
superbly skilled pilots of our time. So buy his book and read about building and
flying the Gee Bee R-2 replica.
The Granville Gee Bee Transport Model C-8 is the
result of experience in aerodynamic and structural design and practice that
has proved itself on the greatest of all proving ground - the race course.
In the years that we have been connected with aviation as pilots, mechanics
and engineers, we have been able to collect an invaluable store of
experience and information that enabled us to build the 1932 Gee Bee
Supersportster in which Major Doolittle succeeded in bringing the world’s
land-plane speed record of 294.38 miles per hour back to America. The
lessons learned in this racing game are of value to the commercial
builder. The Engine The engine selected for the racing job was
the Pratt & Whitney "Wasp," Model T3D1, supercharged to 730
horsepower at 2,300 rpm. The ability of this power plant to develop a
tremendous overload of horsepower over an extended period of time, together
with its horsepower weight ratio, gave this engine a great many
advantages. The Fuselage For minimum profile drag coefficient, a
streamline body should have a fineness ratio between 2 and 4. Also the
maximum diameter should be located at about one-third of the length of the
body. With these points in mind the Supersportster fuselage was laid out.
The fineness ratio was set at 3.2 and the maximum body diameter was
located at 34% of the fuselage length. The fuselage, of course, was not a
body of revolution, but approached this. The sections to about 30% of the
length were true circles. From this point aft, the sections became
ellipses, with the major axis vertical. The disposition of the cockpit near
the tail necessitated this. Having plotted a curve of the drag coefficient
against the fineness ratio for a series of geometrically similar streamline
bodies we found drag coefficient Dc (lbs./sq. ft./mph) reaches a minimum
value at a fineness ratio of nearly 2 and increases above and below this
fineness ratio. With a low drag coefficient, the fuselage cross-section
area could be increased greatly over other type fuselages before the
fuselage drags are equal. This allows a large, comfortable cabin for the
pilot, and with an eye towards commercial aircraft, permits the use of a
roomy interior for airplanes in use on airlines without sacrifice in
aerodynamic efficiency. Therefore, passengers can now travel in greater
personal comfort without increased cost to them or the airline operator.
The fuselage drag may be expressed in the form of a simple equation: Wings and Tail Surfaces For the selection of an airfoil for the
wing, it was of course desirable to use one with a low minimum drag
coefficient and suitable moment characteristics with a small total center
of pressure travel. A high value of maximum lift coefficient was necessary
to obtain a reasonably safe landing speed and quick take-off. In the racing
airplane low drag is probably the most important factor, followed by
suitable moment characteristics; while in the commercial job low drag is
important from the standpoint of high speed performance, but high maximum
lift with consequent good landing and take-off characteristics is most
essential. Substituting these values in the equation: Seven of the M-type airfoils may be
classified as the "stable series" of that series. The center of
pressure travels towards the trailing edge as the angle of attack is
increased from low to high angle of attack. The M-6 section has the
smallest total center of pressure travel, namely 9%. For the Supersportster
wing, sufficient spar depth was possible with ordinates of M-6 reduced to
65%. It is believed that the center of pressure travel is not altered by
this change, but the minimum profile drag of the wing is reduced. In
flight, the center of pressure travel is small, for the adjustable
stabilizer as installed on the Supersportster is not entirely necessary.
The elevators were found to be sufficient to maintain trim alone
throughout the entire speed range and with power on or off. It is therefore
believed that a small center of pressure travel is necessary
characteristic in racing airplanes. The tail surfaces of a racing plane
should be of sufficient area to permit adequate control at stalling speed
and also no larger than necessary because of parasitic drag. Furthermore,
large surfaces are apt to be oversensitive at high speeds. Engine Cowling The engine cowling regardless of type or
characteristics should permit sufficient cooling of the engine cylinder
heads and barrels. It should also be constructed as to allow smooth air
flow around the fuselage and have as large an anti-drag component as
possible. At this point, it might be said that a large number of tests to
determine the best chord length, angle to thrust line, and fore-and-aft
position could be made for the best combination for this particular model.
Time, however, was he limiting factor and experience dictated the general
lines to be followed for the design of the engine cowling. However, three
types were tested on the wind-tunnel model and drag values were obtained
for each type - single and double chamber cowlings and one designed to give
the nose of the fuselage a streamline appearance; this type was not
constructed to function on the anti-drag principle. The Landing Gear The landing gear of the racer was a
rather conventional type of fork and shock absorber, having an Aerol strut
with five inches travel, working on air and oil. The Cockpit The main reason for situating the
cockpit of the racer just ahead of the stabilizer was visibility, inasmuch
as the fuselage at this point had changed from the huge circle 61 inches
in diameter at the wing butts, to an ellipse 56 inches deep, and slightly
wider than the pilots shoulders at the top longeron. This allowed the use
of a small cabin which streamlines well into the fin, and at the same time
allowed the pilot to look down at the side and around the fuselage at the
best angle. The windshield was three piece "shatterproof" glass.
The covering from the windshield back to the head rest was Fibreloid,
secured through catches making this cover removable by the movement of a
small lever on the left side. This allowed the pilot to release the cover
in an emergency and exit through the top, or in the case of oiling up the
windshield, allowed him to release this cover, and with the windshield
fixed in place, fly the ship like an ordinary open cockpit job. General In the racer, exhaust gases were
deflected to the proper points within the N. A. C. A.
cowling by special stacks and allowed to mix with the air passing through
the engine and out the rear edge, thus eliminating the possibility of
disturbing the air flow around the outside of the cowl. Note:The picture above is a composite of a model of the Gee Bee C-8.The C-8 was never completed due to lack of funding.
© 1998 dgraves549@aol.comGee Bee Model "R1"(click picture for a larger
view).
The R1 and R2 at the 1932 Cleveland Air races.
Lee Gehlbach and the Gee Bee R-2 #7 after an uneventful cross country flight to Burbank to start the Bendix race. (click picture for a larger
view).
The total amount of money brought back to Springfield, MA made the SARA members
prosperous and happy.Z. D. "Grannie" Granville with the R-1 at Bowles field shortly before Jimmy Doolittle flew it to Cleveland.(click picture for
a larger view).
The R-2 (#7) left. and the R-1 (#11) in front of the Skyways hangar at Cleveland, Ohio in 1932.You can very clearly see the difference between the two aircraft in this photo.(click picture for
a larger view).
Jimmy Doolittle and the Gee Bee-R-1(click picture for
a larger view).
In 1933 the Gee Bee R-1 had a new 1,860 cid (around 1,000
h.p.) Hornet engine and the Gee Bee R-2 received the 1932 Gee Bee R-l's record
setting 740 h.p. Wasp engine and cowling. Also in 1933 Zantford (Grannie)
Granville designed a new, slightly thicker wing incorporating flaps and having
an increase in wing area to 132 sq. ft. from 100 sq. ft. All of which lowered
its landing speed to 65 mph. from over 100 mph. The old 1932 wing was put in
storage. This 1932 wing was the one that was used on the 1933 R-1/R-2 hybrid since
it was identical to the Rl's damaged wing.
Russell Boardman with the modified Gee Bee Model
"R-1", A Hornet engine was installed and the rudder had an extension added to
it,these modifications were for the 1933 race season. (click picture for a
larger view). New England Air Museum's: Gee Bee Model R1 replica.(New England Air Museum Photo) Gee Bee Model "R-2"Modified for the 1933
Cleveland Air Races. With a new wing with flaps,as well as the Wasp engine and
cowling from the R1 and an extension to the rudder.(click picture for a larger
view).
The dissapointed Granville brothers tried once more
with a hybrid model called the "Long Tail Racer". It used the "Hornet" engine,
the R1 fuselage with another section added and the wings from the 1932 R2. This
aircraft was built for the 1934 races. It was longer, had a new logo and was
expected to easily fly 300 mph.The letters "I F" on the nose stood for
"Intestinal Fortitude". Race pilot Roy Minor was chosen to test fly this aircraft,
and had a good first flight. But on the second flight he had a landing accident
and flipped over a fence. Roy was not hurt, but the same could not be said for the
aircraft. Although the aircraft was repairable, the Springfield Air Racing
Association saw fit to pull their funding. This ended the Granville Brothers
series of racing planes.
After another rebuild, the Long Tail Racer ended up
with Cecil Allen. Despite warnings from Pete Miller and Granville, Allen
installed a larger fuel tank well aft of the center of gravity (cg). Granville
and Miller feared the cg would be moved be to far aft, making the plane
impossible to fly. Ignoring these warnings, Allen took off with the fuel tank
full, lost control and was killed. This ended the R-1 and R-2 racers. The Long
Tail Racer was never rebuilt."Long Tail Racer" (click picture for a larger
view).
The Gee Bee R-2 Replica
The Gee Bee racers of the early 1930's have always fascinated pilots and
enthusiast, and a number of people have started R-1 projects (Ron Coleman of
Springfield, IL and Jess Shryack of Justin, TX).
Gee Bee R-2 Replica.(click picture
for a larger view).
The following is a transcript of an article from the July,
1933 Aero Digest.
I would like to thank Scott Brener for taking the time to transcribe and
provide this very interesting article.
Aero Digest magazine
July 1933The Influence of Racing Types on
By Z.D. Granville
Commercial Aircraft Design
The two greatest requirements in commercial
craft are speed and durability. The streamline shape that proves superior
in a fast racing ship is obviously better for commercial use. Structural
design that will withstand the terrific abuse of a race course will, in
all probability, prove durable when used in a commercial craft. For either
racing or commercial craft, durability is a fundamental essential, and
with this point in view it is obvious that the engine plays a very
important part.
The most dependable high powered aircraft
engines in the United States are of the radial air-cooled type, which
to the layman’s (and also many designers’) point of view are highly
undesirable from the point of streamlining and high speed. With the
necessity of using a radial air-cooled engine because of its durability
features, we began some years ago to investigate the possibilities of
streamlining such an engine. Experiments led us to believe that it was
possible to streamline an engine of this type in such a manner that it
would have no more drag than any other type of engine of equal horsepower.
We discovered that it was not nearly so much the frontal area and cross
section of an object as the shape that counted for the desired low drag.
Our 1931 Gee Bee Supersportster was the first
serious attempt to put into practice our theory in this respect. This job
with its large frontal area, huge fuselage, presented an unusual appearance
and brought hails of criticism from designers and pilots from all over
the country. Some predicted that the ship would never fly; others predicted
that if it did fly it had no possible chance of winning anything. In spite
of this the 1931 job brought home the Thompson trophy and established a new
American landplane speed record. The Gee Bee Supersportster of the
following year was designed solely for racing purposes and to establish
a new landplane speed record.
The appearance of this airplane with its large
frontal area and huge fuselage,
61 inches in diameter, represented so radically a departure from the usual
racing airplane that the design again evoked criticism from the
aeronautical fraternity, in spite of the fact that we had proved our point
on the 1931 ship.
To check our calculations on this job a scale
model was built and tested in the New York University wind tunnel. This
test proved our theory correct and the ship was built along the lines first
intended.
Experience with racing craft in the past
taught us that a propeller efficient at high speed is extremely poor on
the get-away and take-off, which counts seriously in a race from a standing
start and is particularly objectionable in a commercial job taking off with
a heavy load. With this in mind we selected a Smith controllable-pitch
propeller as the ideal installation which would give us the desired
take-off and still hold the motor down to the proper rpm at the proper
speed.
Df = Dc x A
Where
Df = total fuselage drag in
pounds
Dc = fuselage drag coefficient in
(lbs./sq. ft.)/ mph2
A = maximum cross-section area of fuselage
(sq. ft.)
From data compiled for the comparison of the
Supersportster fuselage with the fuselage of other well known types, it was
obvious that by careful attention to the particular shape of the fuselage
it was possible to get a lower drag coefficient per square foot of area
with consequently low drag for the entire fuselage. This was important as
far as the racing ship was concerned, but when taken into consideration
from a commercial standpoint the beneficial results are two fold:
First, a low drag fuselage on our commercial
job will result in better speed which means greater range, lower power
required and also lowering of flying time on the ship with consequent
reduction in operating cost.
Second, the possibility of using the large
fuselage without added drag also allows us to build a cabin which is roomy
and comfortable.
For the racing airplane the M-series of
airfoils recommended themselves because of their low drag and small center
of pressure travel. The drag of several of the M-series of airfoils at a
constant value of lift coefficient was plotted against thickness in per
cent chord at 30% chord. For this comparison, a value of lift coefficient
equivalent to a speed of 300 mph of the full scale airplane was
selected.
The lift coefficient equivalent to a speed of
300 mph is
Where:
CL =
L
P AV2
2
L = Gross weight of the
airplane (2,500 lbs.)
A = Wing area in square feet
(75 sq. ft.)
P = Density of air (.002378)
V = Speed in ft./sec/300
x 1,467 = 440 ft./sec
CL =
2500
.002378 x 75 x 3002
x 1.4672
2
= .1458
The tail surfaces were cantilever type
of all wood construction, similar to the wing. They were designed for 75
lbs. per sq. ft. with margins of over 100%, which together with the method
of attachment on the fuselage at widely separated points gives the desired
rigidity. This construction was much heavier than necessary, but we wished
to make sure to eliminate vibration due to flexibility.
The stabilizer area was 7.5 square feet.
A fillet of 2-inch constant radius was fastened to the stabilizer and held
against the side of the fuselage by spring tension and travels up and down
with the stabilizer during adjustment.
The stabilizer control was screw-type,
operated by a wheel directly in front of and at right angles to the seat,
within easy reach of the pilot, with his eyes still on the course. The
stabilizer has 15º of travel, being hinged about the front spar. In actual
practice, we found that no adjustment was necessary for landing, take-off,
or flying, after the proper adjustment was once made. The elevators were
controlled by cables from the stick to a bell crank mounted directly below
the stabilizer front spar, thence to the elevators by a push-pull tube with
ball-bearing joints.
The control stick was unusually long and
geared down in such a way as to make a small movement of the control
surfaces with a large movement of the stick. The control proved to be light
and sensitive.
The rudder of the racer was somewhat
different from the usual in rudder construction and design. It was merely a
movable end on the streamline fuselage. From previous experience it was
found that a thin rudder acting between the converging streams of air on
each side of the fuselage was too sensitive at high speeds if of sufficient
size for control at stalling speed. The rudder is nearly 12 inches thick
at the hinge-line and the sides of the rudder are therefore nearly parallel
to the flow of the air stream on each side. This makes the rudder action
soft at high speeds while of sufficient area for complete control at
stalling speeds. The actual operation proved sensitivity to be more
pronounced at stalling speeds than at high speeds.
Results of these tests showed the single
chamber cowl to be better than the others as far as drag is concerned. It
was believed that cooling would be sufficient under all conditions in which
the airplane was to be flown. This cowl was adopted and, it might be added,
the cooling was found adequate and the flow of air behind the cowl and
along the fuselage was excellent. This fact is attested to by the exhaust
smudge on the body.
From past experience we expected trouble
from the N. A. C. A. cowling, and governed the design
accordingly, but in spite of our precaution the cowling persisted in
sliding forward. It was held by its leading edge (supposedly the strongest
part of the cowling) by turnbuckles. This leading edge was reinforced by a
half inch tube rolled into the cowling. In spite of the extra strength of
this arrangement, the tendency of the cowling to move forward was so great
as to roll the entire leading edge inward, tending to run the entire
cowling wrong way out. We found it necessary to reinforce the nine points
of attachment as well as to build up the ribs inside the cowling at this
point, to hold them from falling back.
The chassis structure of this landing gear
forms the flying structure to which the streamline wires were attached.
The maximum loads in this structure from landing and the maximum loads in
flying were, of course, not applied at the same time, therefore, this
structure did double duty.
Warner Aircraft wheels and brakes were
actuated by the system of full back on the stick with directional control
by the rudder pedals. The 6.50-10 Goodrich tires were enclosed in boots
which were attached to the wheel forks in such a way that they travel
straight up and down with the wheel, keeping in line of flight and keeping
the wheel and tire covered to a maximum in all positions. The Goodrich
sponge rubber tail wheel was used with a three-inch oleo type shock
absorber, the whole unit being contained within the rudder, giving positive
ground control when taxiing or landing. The tail wheel was held in line of
the rudder by a cam action which was releasable, allowing the wheel to
swivel 360º on its vertical axis. This made the ship easy to handle in
the hanger.
This type of landing gear proved not only
to be of very low drag, but had fine shock absorbing and ground handling
qualities, as well as the ability to withstand a terrific load in rough
landings at high speed on rough fields.
Access to and from the cockpit was by means
of a door in the side of the fuselage. By means of a lever just ahead of
the door, the hinge pins could be withdrawn, and the door itself instantly
released. By this means, the pilot, in case of emergency could put his head
between his knees, release the belt, and roll over, instantly being clear
of everything with the exception of the stabilizer, which, having no
bracing of any kind and being right at the door, could not catch or injure
the pilot. This gives the pilot two means of exit in case of an emergency.
The pilot’s seat was adjustable, by means of a wheel on the left hand side,
allowing the pilot to adjust himself to the most comfortable position for
visibility.
The instrument board was fully equipped
with navigation and flight instruments. A Lewis Engineering company
thermocouple with selector switch was installed, with the Fahrenheit gauge
as near as possible in line with the pilot’s vision during a flight, where
he can easily note the head temperatures.
Ventilation of the cockpit was provided by
a three-inch flexible tube, bringing fresh air from a scoop just forward of
the cylinders on the engine. This stream of fresh air at high velocity is
controlled by a knob on the instrument board which forces a deflector plate
against the tube outlet, closing off or on, and deflects the air at will.
The ventilator behind the pilot’s head allowed exit of the air and brings
circulation of fresh air to the pilot’s face without drafts. The air is
exhausted through the joint of the fuselage and rudder.
Two gasoline tanks of eighty gallons each
were provided, with a shut-off valve at the extreme lower left of the
instrument board, allowing either or both tanks, or the twenty-five-gallon
reserve to be used. Oil was carried in a twenty-gallon flat corrugated tank
installed in the engine compartment.
A specially designed throttle, spark, and
altitude control, having large rigid knobs and individual friction
adjustments was used. This control unit consisted of a lever controlling
the Hamilton Standard controllable pitch propeller and also the control for
the Smith controllable propeller. Thus, engine controls were in one strong,
neat appearing and accessible unit, requiring the use of the left hand only
and not necessitating the shift of hands on the stick for any operation.
The throttle was of such rigid construction
as to allow the pilot to use it as a hand to steady himself while flying
along at high speed in rough air. From the gas tank section back, the
fuselage turtle deck is of plywood construction, the sides being built up
of fairing and covered with fabric.
Disappearing handles were provided for use
in lifting the tail for any purpose, as was also a rest contained within
the fuselage, under which a "horse" could be placed without
damaging the fairing.
Two hand holes with spring covers opening
inward were provided on the sides of the windshield, allowing the pilot to
clean the windshield in flight. These holes were ordinarily kept covered to
exclude any exhaust gases.
Owing to the bulkiness of the fuselage on
both the racing ship and transport, the first impression gives one the idea
that the ship is short coupled, as well as having extremely small tail
surfaces, which, however, is an optical illusion, as actually the tail area
and the distance from the leading edge of the wing to the elevator hinges
is greater than used on many of our previous models, including the 1931
racer. The tail surface area in relation to the wing area is greater than
standard practice, being 20.75% in the racer, while on the transport job
this amounts to 18.25%.
The racing ship was designed with a load
factor of 12 at high angle of attack with the exception of the tail
surfaces, whose load factors were considerably higher.
On the commercial job the load factors are
7 at high angle of attack with the exception of the tail surfaces, which
are slightly higher. The reason for the difference in load factors is
because of the difference in the weight horsepower ratio, as well as the
maneuverability. Also, the size of the racer allows for quicker
maneuverability and the ship is liable to be subjected to terrific loads
due to these maneuvers, or experienced when running into the wash of
another ship during a race or flying at high speeds through rough air.
Proof of the soundness of this design by
actual performance is ample reason for using such features of aerodynamic
and structural design in our commercial jobs, and the game of airplane
racing today has the same bearing on the commercial aircraft industry of
tomorrow that automobile racing has always had on the following automotive
industry. Race courses are laboratories and proving grounds of new and
better ideas, and pave the way for efficiency and safety in the airplane
and automobile of the future.
Specifications
Supersportster
R-1Transport
Model C-8
Wing span
25 feet
45 feet
Length
17 feet 8 inches
33 feet 9inches
Height
8 feet 2 inches
11 feet 6 inches
Net wing area
75 square feet
Gross wing area
335 square feet
Root chord
53 inches
132 inches
Wheel tread
76 inches
120 inches
Max. fuselage diameter
61 inches
Max. fuselage cross-sec
41 sq.feet
Aspect ratio
6.1 to 1
5.2 to 1
Incidence
2.5 degrees
2.5 degrees
Dihedral
4.5 degrees
4.5 degrees
Weight empty
1840 pounds
3925 pounds
Racing gross wt. (50 gals.)
2415 pounds
Maximum gross weight
3075 pounds
7000 pounds
Total fuel capacity
160 gallons
200 gallons
Performance
(Estimated)Supersportster
R-1Transport
Model C-8
High speed
294.38 miles per hour
225 miles per hour
Cruising speed
260 miles per hour
190 miles per hour
Landing speed
90 miles per hour
Landing speed (with flaps)
50 miles per hour
Rate of climb
6100 feet per minute
1100 feet per minute
Endurance, full throttle
2.14 hours
3.08 hours
Endurance, cruising
3.65 hours
4.5 hours
Range, full throttle
630 miles
695 miles
Range, cruising
925 miles
850 miles