Dynamic Flight Simulator Lets Swedish Pilots Pull Gs
By William B. Scott
January 4, 2004
The Swedish air force is "flying" a combined flight
simulator/centrifuge that doubles as a JAS 39 Gripen fighter-pilot
trainer and medical laboratory for researching the stresses of
high-g combat. The system allows line pilots to pull up to 9g and
command any pitch and roll attitude, ensuring their experience will
differ little from actual flight.
Designed for exceptional flexibility and realism, Sweden's Dynamic
Flight Simulator (DFS) in Linkoping, Sweden, is believed to be the
first fourth-generation device--a ground-based system capable of
pulling gs and replicating modern tactical-aircraft flight
conditions--deployed by an air force.
Such simulators are a fairly recent advancement, although several
companies have been working on centrifuge-based systems for years.
A handful of fourth-generation units are being marketed globally,
and by all accounts, competition is fierce, with myriad claims and
counterclaims (AW&ST Dec. 1, 2003, p. 70). However, one Swedish
test pilot, who has evaluated most of the systems now available,
said realism and fidelity vary considerably.
I had an opportunity to "fly" the Swedish air force's DFS at the
Defense Materiel Administration (FMV) test and evaluation center
here, and was impressed by how closely the system approximates a
high-performance aircraft. There are unavoidable physical
limitations, of course, but it was definitely the most realistic
flight simulation I'd ever experienced.
Basically a fighter cockpit mock-up mounted on the end of a
centrifuge arm, the Swedish DFS represents a quantum leap in
fighter pilot training and flight physiology research capabilities.
It was designed and built by California-based Wyle Laboratories,
and boasts a number of features critical to emulating high-g
flight, such as:
* A high-torque direct-current (DC) motor that drives the 30-ft.
centrifuge arm.
* A large spherical gondola, suspended by gimbals and configured
with Gripen cockpit displays and controls, and a modest
three-screen visual array.
* A control system based on the Gripen--or other
aircraft--aerodynamic model that accurately simulates a fighter's
dynamic-response characteristics.
* Perceptual algorithms in the control system that "trick" a
person's physiological sensors, providing a degree of realism that
"simply is not possible in conventional centrifuges," according to
a Wyle official.
* A closed-loop control mode, which allows a pilot to fly the DFS
and fully command its motion. Preprogrammed modes are also
available.
* The Swedish air force/Wyle Labor-atories Dynamic Flight Simulator
features a 30-ft. arm and fully gimbaled gondola. Minimal lag and
delay times, ensuring the simulator's g-response closely
approximates an actual aircraft.
Wyle engineers set out to develop a centrifuge-based simulator
tailored for training and evaluating pilots in an elevated-g
environment, yet flexible enough to also conduct aeromedical
research. Consequently, they focused on features that would ensure
rapid response times.
"We wanted to give pilots the ability to 'fly' and interact with
the environment rather than just be a passive [centrifuge] rider,"
said Will Roberts, program manager for Wyle Laboratories' DFS
programs. "We've come a long way in being able to translate the six
degrees-of-freedom you get in an aircraft into the three
degrees-of-freedom that we can control in a centrifuge. It's not
perfect, but we think it's pretty good. There's room for more
research to make it even better."
Swedish air force (SAF) officials were prompted to invest in a
high-g simulator when they acquired the JAS 39 Gripen, a modern 9g
fighter capable of rapid g-onset (going from 1g to high-g loads in
a very short time) and sustaining those levels. They recognized
that aircraft and pilots could be lost to the "G-Loc" (g-induced
loss of consciousness) phenomenon, where humans can quickly pass
out when subjected to high gs. Incidents early in the Gripen
program led SAF leaders to conclude that not enough research had
been done in the area of high-g physiology. Consequently, they
decided to develop world-class expertise in this arena--while still
training SAF pilots to properly handle g-caused stresses inflight,
noted Kent Engstrom, FMV program manager for DFS.
Research is needed to better understand physiological phenomena
"the moment before G-Loc and the moment after [recovering] from
G-Loc--how you regain psychomotor and cognitive abilities with
various delays," said Leif Pettersson, an FMV test center operation
engineer. "These are possibilities for research . . . that I don't
think anybody has done before."
The system I "flew" was "mechanically unique," Roberts said. "The
arm's structure is like a wing--hollow aluminum with a stressed
skin. Other centrifuges use a box-beam construction. The
stressed-skin [design] gives us a much better natural frequency,
which allows us to do a lot of things without being disturbed by
structural effects."
The lighter, stronger arm simplifies controlling g-onset rates, and
keeps start/stop stresses lower, which enables the use of a smaller
drive motor. Wyle chose a 1,900-kw. Westinghouse DC motor
originally designed for steel mill applications, connected directly
to the centrifuge-arm drive system. It delivers about 7 megawatts
of peak power in approximately 100 millisec. when a pilot pulls
sharply on the simulator's cockpit control stick, demanding a high
g-onset rate.
"WE CAN GO FROM [ZERO] to maximum g-onset in about 120 millisec.
with this motor," Roberts said. But because the high power demand
is very brief, the motor doesn't overheat.
Before my DFS mission, a flight surgeon attached electrocardiogram
sensors to my chest, allowing real-time monitoring of heart
parameters. He assured me that was standard procedure. I also was
fitted with a Swedish-type g-suit, then strapped into a Gripen
ejection seat mounted in the DFS' gondola--a ball-type structure
suspended at the end of the centrifuge arm.
The Swedish AF simulator/centrifuge gondola emulates a Gripen
fighter cockpit with a three-screen out-the-window visual display.
Pilots are routinely "wired" with sensors that monitor heart
functions.
Up to 400 electrical power, video, computer, medical and other
signals are transferred from the cockpit to a control room via
numerous slip rings that also allow unrestricted rotation of the
gondola and arm. Five channels of gases--such as oxygen--also are
passed through a sophisticated slip-ring system.
Like all Swedish pilots who go through training in the DFS, I did
not wear a helmet or mask, because controllers need to monitor
facial expressions during high-g maneuvers. After a thorough
briefing and cockpit preflight by Capt. Christopher Gruv, I was
closed into the gondola, and controllers prepared for the "flight."
Maj. P.A. Klingstrom, an aerospace physiologist pilot and training
specialist, served as the mission controller, talking to me via
intercom.
Initially, the centrifuge arm turned at a slow, steady speed,
producing a 1.5g "steady state" or baseline. Inertial restrictions
require that the arm be in motion before a pilot starts pulling
high-onset gs. Rapid g-onset would demand almost infinite power to
go from a dead stop to a 10g/sec. rate the system is specified to
deliver. Roberts said the DFS has demonstrated a 14.5g/sec. rate
with a full load, plus an additional 100 kg. (220 lb.) of
weight.
I had been told to fly the simulator however I wanted, but with a
couple of cautions--no rapid negative-g pushovers, and don't be too
aggressive when rolling immediately from a right-bank high-g
condition to a left-bank high-g maneuver. The system would respond,
but these inputs would cause unneeded stress on its
components.
I made a few gentle turns to get a feel for the Gripen flight
controls, noting that the three-screen visual scene produced a
comfortable perception that one was flying contact. Then I rolled
into a steep turn and pulled on the stick, advancing the throttle a
bit. The DFS responded instantly, smashing me into the seat. Over
the next few minutes, I maneuvered in the low-g arena, eventually
pulling up to 4.7g. I've pulled many more in actual aircraft, but
never particularly enjoyed what some call the "practice bleeding"
of high-g flight.
I was quickly convinced the DFS would respond rapidly to control
inputs, and could easily deliver the requisite normal acceleration
to produce tunnel vision and make me gray-out. I had no desire--or
need--to demonstrate my limited "g-tolerance."
Flying daily in a tactical aircraft, a crewmember develops a
capacity to handle high-g conditions. He instinctively tightens his
legs and abdominal muscles, assisting the anaconda-like g-suit's
inflation, which squeezes legs and abdomen. Pilots are also trained
to perform a specific anti-g "straining maneuver" that involves
short-pulse breathing, while keeping the lower body muscles
tensed.
All this effort is essential to keep one's blood flowing to the
head as g-loads increase. If blood drains from the cranium while
pulling high gs, vision seems to close in from the edges, until it
appears you're looking through a small tube. That's called tunnel
vision. Higher gs--or relaxing at the wrong time--can close that
tunnel completely, and the eyes-wide-open pilot sees nothing but a
gray curtain. Objects are no longer visible, but he can still hear
the radio and engine, and is still flying the aircraft. He's
functioning, but just can't see anything.
At this point, a pilot or backseater is only a g or so away from
having his world go completely dark, a condition known as
"blackout." Then, he's no longer able to help himself or fly his
aircraft. Recovery requires reducing g-loads on the body, but after
blacking out, human beings are typically slow to respond. Many
seconds or even a minute can pass before a pilot is really capable
of controlling his aircraft properly.
The DFS is an effective means of training pilots to resist graying-
or blacking out by properly performing the anti-g straining
maneuver under controlled conditions. Some catch on faster than
others, but all can be trained, according to Klingstrom, who has
flown hundreds of high-g sorties in centrifuges. He's known here as
"The G-Monster."
As mission controller, he next suggested I fly a tail-chase
profile, presenting another aircraft in my simulated HUD. I was at
the aircraft's 6 o'clock position, and my job was to stay on his
tail as he maneuvered. I noted that the simulator's mechanization
was so good that I felt as if I really were flying, pulling
whatever gs were necessary to stay with the target as it
accelerated, climbed, dived and turned sharply. The profile was an
actual mission Klingstrom had flown in the DFS, where data were
captured and "canned" as a target for training missions.
As long as I kept my head motionless, pressed against the seat
headrest, the simulator closely replicated tactical flight, I felt.
Experimenting a bit, I turned my head slightly to the left and
right, and tilted it up and down. Immediately, my vestibular system
sensed I was rotating. However, pilots are told that will
happen--so don't do it.
However, rolling the "aircraft" and pitching up and down produced
very realistic feelings of flight. The fully gimbaled gondola is
unrestricted in pitch and roll, but sophisticated, well-tuned
control algorithms are required to produce such realism.
"WE HAVE TO DO some strange things to . . . fool your vestibular
system," Roberts said. For example, "when the arm speeds up, we
compensate for the increased gs by pitching the seat forward,
because the [human] vestibular system is sensitive to rotational
acceleration. You stay aligned with the force vector.
"How fast we accelerate the arm, then accelerate the pitch and roll
axes [simultaneously] is a big factor in the design," he added.
"This is a simulator mounted on a centrifuge, not a centrifuge that
happened to have a simulator in it, so that [required] a different
approach. Getting the simulation right was [a priority]."
Maj. Richard Ljungberg, a Swedish test pilot, has flown many
tedious hours in the DFS, trying to tweak and optimize those
control algorithms to ensure the simulator closely approximates
actual Gripen flight. "I've changed a lot of numbers [related to]
movement in roll, pitch and delays. The biggest thing is g-onset
response time, and the ability to control the centrifuge from low
gs to high gs. If you have a long delay [between] pulling the stick
and getting the gs, you learn the [wrong] response coordination. If
you're at high gs, and you feel like you're starting to gray-out,
and you unload [relax stick pressure], you want the machine to
respond. If you unload and nothing happens, it's ugly." That can
lead to negative training, doing more harm than good.
My flight was only about 30 min., but long enough to convince me
that the DFS emulated high-g tactical flight quite well. Through
clever algorithms and persistent refinement, the Wyle engineers and
Swedish pilots assigned here have done a good job of simulating
six-degrees-of-freedom flight.
Probably the most telling validation of DFS fidelity comes from SAF
pilots who go through training here. After flying a tail-chase
scenario and trying to shoot down the simulated target aircraft,
pilots are informed their time is up. They often respond, "Just one
more, OK?"
Normally, pilots hate simulator training. "I think we've come a
long way, creating a centrifuge/simulator that pilots can actually
enjoy flying," Ljungberg said. "Normally, when you go to a
centrifuge, you're like a dog--just sitting there, trying to strain
and survive the g-forces. But in this one, you're in control. Here,
you pull gs the same way [you do] when you go out and fly. It's a
good training tool."
The DFS also ensures pilots develop desirable habits in
coordinating stick movement with anti-g straining maneuvers. About
60 pilots have been trained in the DFS, but the FMV/Wyle team here
is continuing to refine both the initial and refresher
programs.
However, they also see tremendous potential for serious
physiological research. Ljungberg mentioned experiments to
determine how a pilot's perception of color and his hearing degrade
at high-g loads, for example. Combat scenarios involving target
acquisition and tracking, missile-firing and other tasks can be
explored to determine how pilots perceive an objective's location
under high-g loads. Similarly, the position of cockpit controls, or
even the color of head-up and head-down display symbology, might be
altered as a result of research findings.
"The sky's the limit," Roberts said. "We're just barely getting
started.
