Message-ID: <4175DEB9.4060000@Logan.com>
Date: Tue, 19 Oct 2004 23:42:49 -0400
From: Rob Logan <Rob@Logan.com>
User-Agent: Mozilla Thunderbird 0.8 (Windows/20040913)
MIME-Version: 1.0
To:  lml@lancair.net
Subject: Trials of SS1
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

reprinted without permission from AW&ST Oct 18 2004 page 36

*Trials of SS1
*Winning the X Prize took courage.
Next craft will have electronic assist.
by Michael A Dornheim

     Rocket motors that either wouldn't ignite or stop burning, sticky
dampers that gummed up the controls and bleak choices after being
dropped from the mother ship --these are some of the trials that faced
designers and pilots of SpaceShipOne on its way to being the first
private spaceship and claiming the $10-million Ansari X Prize on Oct. 4.
     Scaled Composites test pilots Peter Siebold, Michael Melvill and
Brian Binnie gave their colleagues first-hand insight into SpaceShipOne
(SS1) development at the recent 48th annual Society of Experimental Test
Pilots (SETP) symposium held in Los Angeles.  A Discovery Channel/Vulcan
Productions television documentary that aired this month also added
fresh information.
     *THERE WERE SEVERAL *test failures in the competition between
SpaceDev and Environmental Aerosciences (eAc) to design and build parts
of the hybrid rocket engine.  SpaceDev's first 15-sec. run resulted in a
post-shutdown fire that destroyed part of the test stand.  The oxidizer
valve wouldn't fully close and kept feeding the fire.  Later, the eAc
configuration failed to ignite four times in a row.
     The two companies came up with significantly different thrust
profiles.  Scaled was looking for a slow ramp-up to give the pilot time
to pull the nose to a vertical ascent before full thrust started, so
that impulse was not wasted in the horizontal direction.
They wanted a high level of maximum thrust to get acceleration done
quickly while still in the atmosphere so aerodynamic controls would be
effective to counter thrust asymmetries.  Scaled also wanted a tailoff
of thrust that matched the craft slipping out of impulse with thrust low
enough to be countered by declining aerodynamics.  With the ideal
profile, this tailoff would start at 140,000 ft. at 80 kt. equivalent
airspeed (KEAS).
     Neither SpaceDev or eAc met Scaled's wishes.  The SpaceDev design,
which has four longitudinal ports in the rubber fuel for enough burning
area for high thrust, comes on with a bang, producing maximum thrust at
the start - not the smooth ramp-up envisioned to turn the corner.  The
peak thrust is only about 85% of the desired plateau, and declines
steadily from there, according to the Sept. 18 SETP presentation.
Despite the early start, this still means that to get enough total
impulse to loft SpaceShipOne above 100 km., the motor has to run longer
than desired, in the very thin atmosphere where control is tenuous.
     The eAc motor didn't ignite until five sec. after the switch was
thrown, and then also came on with a bang, but the initial combustion
instabilities were less.  It has a single port and compensates for the
lower burning area with fuel additives to increase burning rate.  But
not enough, because it only made about 65% of the desired thrust.  That
required the burn time to be even longer for sufficient run farther out
of the atmosphere.
     Longer burn time of the eAc motor was considered the more serious
problem, and the contract went to SpaceDev.  Engineers decided the could
run the engine down to 40 KEAS at about 180,000 ft. in tailoff thrust.
     The webs of rubber between the four SpaceDev ports thin out and come
apart toward the end of the run.  The chunks extrude through the nozzle,
causing frightening shaking and explosion noises in the cockpit.  It
happened at least three times on one flight and Melvill thought the tail
had blown off.  After minutes by himself in zero-g and entry, he was
relieved when chase aircraft said the spaceship appeared alright.
     SpaceShipOne may be the highest performance aircraft without
stability augmentation.  A lot was learned from the first 15-sec.
powered flight on Dec. 17, 2003.  The jolt at ignition pushed Binnie 6
in. back into the seat, taking the stick with him.  This increased the
angle of attack (AOA) beyond the angle for maximum lift for the first 10
sec., making the roll and yaw handling poor and causing roll and Dutch
roll on the way up.  Also, the partial load of oxidizer went to the back
of the tank, shifting the center of gravity aft, further compromising
handling.
     Melvill added, "It's really difficult to handle in the first 10-15
sec.  We jacked up the early oxidizer flow for more thrust but it gives
unstable flow - in the video you can see your head jerk back and forth.
You need instruments to round the corner, it's too disorienting.
     Binnie found that after initially flying with the sensitive manual
stick and rudder, it was hard to transition to using the electric trim
as primary flight control in supersonic part of boost.  That was
considered necessary because manual controls become very stiff at high
speed.
     He reports his situational awareness was narrow due to the high
workload.  Melvill echoed that comment.  "On my first powered flight
there wasn't enough mental bandwidth," he says.  "I didn't hear or feel
anything, I just focused on the display.  By the third time I noticed a
lot more.  The rocket made a weird howling noise at high altitude; I
didn't notice that on the first flight."
     A control surface damper was added for the first powered flight but
became so stiff at cold temperatures it caused the heavy landing that
collapsed the left landing gear.  This was first reported by /Aviation
Week & Space Technology, /but now Scaled has added that the damper
raised the control stick breakout force to 12 lb. - a large amount that
made the sensitive, lightly damped SpaceShipOne difficult to control
(/AW&ST /Dec. 22, 2003, p. 37).
     After the Dec. 17 flight, a heater was added to the damper to
eliminate the cold stiffness, making pilot Binnie the only one to have
experienced the high breakout forces.  "It was our fault to not put a
heater on it," Rutan says.  Binnie said of the first powered flight to
Rutan:  "Just when you think you've got it under control, something
different would happen."  Binnie's well executed flight on Oct. 4 was
his first on the spaceship since the hard landing.  "I'm happy to get it
behind me," he says.
     *SCALED'S FIXED-BASE *simulator became a comfortable training tool
due to the low flight rates, but it's "hard to remember that the sim
doesn't fly exactly like the real aircraft," Siebold says.  "It's harder
when the sim teaches you techniques that just don't work in the aircraft."
     On the Apr. 8 second powered flight, flown by Siebold, the oxidizer
slosh was addressed by completely filling the tank.  Even though it was
going to be a half-duration 40 sec. firing, the rocket also got a full
rubber fuel load to prevent case burnthrough if the oxidizer valve
didn't shut, as happened on the ground test.  This made it the heaviest
launch yet.
     The flight became tense immediately after the drop when Siebold
sensed  "lots of main wing (flow) separation" as well as poor lateral -
directional control, and discussed with the control room whether the
craft might go out of control if the rocket were lit.  This drama was
revealed in the Oct. 3 TV documentary.  The alternative of landing with
a heavy propellant load was poor, as was the idea of dumping the liquid
oxidizer, which would move the center of gravity aft due to unburned
fuel.  Melvill explained that would make the aircraft so tail-heavy it
would be landing with the stick full forward and going 150-160 kt. -
about 50% faster than normal.  The team decided it was better to ignite
the rocket.  The craft was gliding toward Earth for about 2 min. during
this time and the motor was ignited at the lowest altitude yet, 38,300 ft.
     Changes to reduce pilot motion in the seat included a ratcheting
seatbelt, stiffer cushions and an elbow restraint to prevent jerking the
stick back.  The rudders were canted out 3 deg. to get them out of an
aerodynamic deadband.
     Controllability seemed acceptable during boost, with two roll
excursions of about 45 deg. "Due to high wind shear the Dutch roll was
still excited, but Pete showed that with patience it would damp out by
itself, so little would be gained by trying to get in the loop with
conventional controls," Melvill told the SETP audience.  There was only
about 8 deg. of pitch overshoot on the 90-deg. vertical climb target,
and it took Siebold less than 4 sec. to roll 180 deg. so he could
correct in a positive pitch direction.
     Roll oscillations during the feather descent were within about + or
- 20 deg.  Large side forces moved the pilot's hand on the stick and
drove roll oscillations, but "Pete showed he could damp with aggressive
closed-loop lateral stick inputs," Melvill said.
     On Melvill's first powered flight on May 13, he was able to counter
his own misleading simulator training.  "The workload during pullup
after ignition is very high," he said.  "When it goes transonic the
Dutch roll mode kicks in and was exacerbated by wind shear on all
powered flights.  I found myself working hard to control roll
oscillations, using yaw trim as trained in the sim.  But my DNA
instincts took over and I resorted  to using the rudder pedals.  It
worked for me.  Instincts won over the simulator."
     Melvill found the May 13 first supersonic feathered entry to be
smooth but the loud airflow noise to be "very scary."  When it went
supersonic the ride became "very rough," with large amplitude vibrations
in the tail booms.
     The June 21 flight was the first to use a higher expansion ratio
nozzle and it produced more thrust asymmetry than before.  "The 0.4 deg.
of thrust asymmetry is a big deal, it caused me to fly southwest of the
course," Melvill says.  "Because I was concentrating on the height
predictor, I failed to put in enough right yaw to maintain the
trajectory and found myself 20 mi. south of Mojave.  We can glide 50 mi.
even in adverse wind  so there was no danger, but once defeathered I
headed back to the airport as soon as possible."
     *IN THE DESIGN PHASE*, the pilots wanted stability augmentation, but
Rutan wanted simple, reliable manual cable controls.  SpaceShipOne has
achieved its goal of being the first private spaceship (/AW&ST /Oct. 11,
p. 34), but only through a high level of pilot skill, courage, and training.
     It will be a different matter for a Virgin Galactic craft carrying
paying passengers (/AW&ST /Oct. 4, p. 30).  "It has to be very
straightforward to handle," said Alex Tai, a senior Virgin Galactic
official who is rated to fly airliners and business jets.  And the
Scaled pilots are sure that means the next craft will have stability
augmentation.


--