Beech
Starship History Continued...
In
fact, from the time the program was put on a realistic
schedule in early 1984 it has experienced only two delays:
the first was announced in mid-1986 when we chose to redesign
the aircraft to take advantage of the new FAA regulations
to certify at 14000 pounds. The second came at the beginning
of 1988, and was necessary to correct a pitch damping
problem and to develop a stall warning system that would
adequately define a stall on an aircraft inherently designed
not to stall. This was the only unplanned delay necessary
to resolve technical problems in the program's five year
history.
All to production standard
The new schedule called for six prototypes, including
three flight test articles and the equivalent of three
more for static, environmental and damage tolerance testing.
As work got under way it began to divide itself into three
broad categories: development, certification and production.
In a more conventional program, production would have
taken a back seat to development and certification, but
the nature of composite construction -- making parts in
moulds -- virtually dictated that Beech build the Starship
prototypes with production tooling, and that gave production
an equal priority. To accommodate production we added
242,000 sq ft of manufacturing space.
Having been active in metal bonding technology for nearly
30 years, the company had seven operating autoclaves,
the largest of which was 12ft in diameter by 30ft long.
This was not large enough to support Starship production
activity, but with minor design compromises would work
to build the prototypes.
To support full-scale Starship production and to handle
the composite subcontract work that was hoped would follow,
a huge new autoclave, 60x25ft was installed. When completed,
it would be the second largest in North America.
Much of 1984 was taken up with building the tools and
manufacturing facilities required, and by 1985 a start
was made on building parts for Starship prototypes and
assembling the aircraft.
Before tooling could begin, Beech had to accomplish the
loads analysis and verification work required to validate
the design of the structure, because of Starship's configuration,
the FAA required the generation of substantially more
aerodynamic loads data than would have been usual for
a conventional design. It had to be proved that classical
loads analysis techniques would conservatively apply to
a tandem wing design.
Thorough wind tunnel testing established the pressure
distributions, which were corroborated with computer-generated
analysis. The computer's findings were confirmed with
flight tests on the Proof of Concept using pressure taps.
With the loads confirmed, the complex and time-consuming
process of developing a materials data base for composite
structure began, because none existed. To arrive at this
base, Beech installed a materials test laboratory and
began experimenting with the lamina properties of the
raw materials -- the tapes, fabrics and resins.
Starting with individual plies, we identified the properties
of the various materials, established statistically reliable
minimum values, and ultimately produced more than 8000
data points from which we could predict how an element
made of a specific material would react to various loads
and environments.
The
next step was element testing, building small test panels
that simulated the full range of structure. These were
subjected to wide-ranging conditions of temperature, moisture,
and static loads in shear, compression and tension. Then
they were subjected to cyclic loading to demonstrate damage
tolerance capability. Ultimately, an in-house software
package was developed to prove we could successfully predict
failure loads and modes.
Mistakes
-- catch 'em young
The purpose of all the testing was to reduce the risk
level of the overall program. If an article was inadequately
designed, we wanted to know before we reached the full-scale
test stage. It is much less expensive to redesign early
in a program. The goal was to make full-scale testing
a validation program, proving what we already knew was
going to work. We were very successful in this area, for
the vast majority of static test articles performed flawlessly.
In the few instances where problems were experienced,
minor redesign was sufficient to correct the situation.
More than 128 static load conditions were tested on the
various Starship static test certification articles, both
at room temperature and at elevated temperatures with
moisture conditions to simulate more extreme environmental
conditions than the aircraft could ever be expected to
encounter.
In a conventional metal airplane the next step would have
been fatigue testing. Composites do not fatigue in the
way metal does, so cycling composite structure does not
cause it to lose strength or crack. It was necessary,
however, to prove that Starship's structure could carry
design loads even with inflicted damage. To do this we
applied more than 1.6 million test cycles to various critical
assemblies.
The FAA did not have established design-life criteria
for composite structures, and it was through the materials
data base developed in the test program that the standard
for future designs was developed, a cycle test structure
for one lifetime (20000 hours), inspecting for damage
every 5000 hours; inflict damage, and cycle test it through
a second lifetime. If it will carry limit load at the
end of the test, the structure is approved.
Not all structures can be made as a single piece, some
must be attached together, and this is typically done
with either film or thicker paste adhesives; the work
accomplished in certifying Starship has become the basis
for industry standards on adhesively-bonded structures.
Multiple
redundancy
Large primary structural assemblies must be designed so
that in the event of any one bonded joint failing, the
remaining structure will still be able to carry the design
limit load and retain sufficient stiffness to resist flutter
with a safe margin above maximum operating speed.
Beech
uses ultrasonic testing to ensure the quality of every
structural part that goes into a Starship. Sound waves,
passed through water, measure density and detect flaws
or voids.
Crashworthiness and occupant safety has also been an important
consideration in the design, so fuselage drop tests at
increasing energies were made until visible damage, then
a second article was dropped at the required energy level
to see what effect it would have on the occupants. The
goal was to contain maximum lumbar loads below 1500 pounds,
the level at which crippling spinal injuries are likely
to occur; we hoped to stay below this level with a 10
ft/s drop.
