Professional Documents
Culture Documents
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
Karl E. Modin
Ulf Clareus
Saab Aircraft Division
Linkoping, Sweden
Baltimore, Maryland
September 23-25, 1991
~ ~~
For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics
370 LEnfant Promenade, S.W., Washington, D.C. 20024
Page 1
* Head of Aerodynamics
* * Project manager, JAS 39 Aerodynamics,
AlAA member
Page 2
Preliminary design studies for the new aircraft started be used as landing strips, making the Air Force less
at Saab in 1979. A proposal was presented to the vulnerable to enemy attack. This strategy places spe-
Swedish Defence Materiel Administration in the middle cial demands on the aircrafts ability to opera?efrom
of 1981 and a contract for the development of five pro- short and extremely narrow runways. This is made
totypes and delivery of an initial batch of thirty Gripen possible by low take-off and landing speeds, excel-
aircraft, including support systems, was signed on lent handling qualities and with tailor-made electrical
June 30,1982. flight control system modes. The Gripen is designed
to operate independently of complex ground support
Four Swedish companies through the common compa-
systems. A built in auxiliary power unit (APU) cont-
ny IG JAS, are cooperating in the development of the
ributes to its high degree of autonomy. Flight line ser-
aircraft. Saab Aircraft Division is responsible for the
vice is carried out by conscripts.
development and production of the aircraft and sys-
tems integration. International Marketing is also hand- With a limited number of aircraft operating from many
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
These requirements mainly concern handling qualities Independently of the airplane layout, it was concluded
and flight performance. Other requirements not direcily that the FBW-system was superior compared to a
related to flight performance did, however, affect or conventional system in the following aspects:
influence the aerodynamic design of the aircraft. 0 Control functions and performance
These requirements included maintenance (easy ac- 0 Damage tolerance
cess to systems hardware for inspection and service), 0 Reliability
radar performance (nose cone diameter and blunt- 0 Maintenance and test
ness), pilot vision in combat and in landing approach 0 Growth potential.
on a nanow base 90 road (large cockpit) etc.
Thus, an early decision was made in favour of a digital
The requirements considered most important are sum- FBW-system. This also made it possible to enhance
marized below flight performance by making the studied configura-
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
I
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
the air intake position. The philosophy of operating from dispersed road
bases leads to requirements on low take-off and land-
A comparison of some major characteristics for the ing speeds, but also on good forward visibility for the
two remaining configurations as they were at the time pilot. Thus, the angle of attack in the approach for a
when the layout selection was made in December no-flare landing is geometry limited to enable the pilot
1980 is given below. Futher optimization would have to see the touchdown point. As a consequence, the lift
improved both layouts, but, of course, refinements potential of the delta canard 2105 layout is not fully
were only carried out on the winner. utilized during landing approach. Both configurations
were designed for the same approach speed resulting
in approximately 7 percent less wing area required on
the 2102.
Instability level
High speed performance
Slightly different inherent instability levels were chosen
The total cross sectional areas of the two configura-
for the canard and the aft tail configuration, typically
10 percent M A C at cruise conditions for the 2105 and
tions in cuts perpendicular to the length-axis, i. e. cor-
responding to M = 1.O, are shown in Figure 5. In Fig-
5 percent for the 2102. We consider, that the margin for
ure 6, the mean cross sectional area distributions of
an aft tail configuration, where overloading and stall-
ing of the trim surface would aggravate the situation cuts at angles corresponding to M = 1.1 are given. At
M = 1, canard configurations tend to obtain an unfavor-
by increasing the instability should be bigger than for
able saddle-form. Howewer, by careful local fuselage
a canard configuration, where trim surface stall tends
to stabilize the airplane. This is a consequence of a
design, this tendancy is not pronounced for the 2105
basic difference between the two layout principles and has disappeared completely at M = 1.l . Max
cross sectional area is some 9 percent lower for 2105
compared.
in comparison with the 2102.
On a canard aircraft a significant contribution to the
desired instability level is obtained from the movable CROSS-SFCTIONAL
canard. By removing this contribution, the airplane I I
AREA 1
turns slightly stable or neutral in pitch. On an aft-tail
configuration the opposite effect is obtained. These
characteristics of a canard airplane are used on the
Gripen in a flight control system back-up mode, where
the canard surfaces are feathered and allowed to free-
float in the air stream. The resulting increase in pitch
stability relaxes the demand on control surface rate
and, consequently, on the hydraulic system capacity.
In the extremely remote event of total loss of hydraulic
power, when the fully unstable aircraft would diverge
too rapidly for pilot safety, this feature will stabilize the
aircraft long enough for the pilot to eject. LENGTH
SV39 867
Also, the gain in low speed lift at approach angle of Figure 5. Cross sectional area distribution less air
attack and in lift dependent drag is less pronounced intake at M = 1.0.
on an aft tail aircraft than it is on a delta canard for an
increase in pitch instability.
Page 6
Final selection
In the layout selection several characteristics, some of
which are not directly related to aerodynamics, were
0 Structural weight/cost
0 Pilot vision in combat and landing
LENGTH 0 Store carrying capability
sv39 ma
Gun installation with respect to local environment
Figure 6. Cross sectional area distribution less air 0 Radar cross section
intake at M = 1.1. 0 Development risk
Of particular importance to supersonic wave drag is 0 Life cycle cost.
the slope of the area distribution towards the aft end In the technical evaluation, no single virtue alone was
of the aircraft. The absence of an aft tail and the for- decisive in the selection of the delta canard configura-
ward position of the wing on the fuselage, necessary tion. However, weighing all the pros and cons when
for the desired instability in pitch, makes it possible to compared to the aft-tail contender, the canard airplane
obtain an aerodynamically clean aft end on the canard was found to be the best candidate to meet the stipu-
configuration with a favorable area distribution. lated requirements from both technical and economi-
Zero lift drag for the two configurations is shown in cal points of view.
Figure 7. The canard configuration is slightly better at One additional important aspect was, that the delta-
subsonic speeds. At supersonic Mach numbers, the canard was considered to offer better possibilites for
difference is quite significant. optimization of flight characteristics with the fly by wire
CD@ system and to have better potential for the incorpora-
4 ZERO LIFT DRAG tion of unconventional control modes, e.g. futher de-
velopment of direct lift, direct sideforce for fuselage
aiming, gust alleviation etc. Another example is the
landing ground roll-out mode described in the next
chapter.
The final decision to select the delta-canard concept
for futher refinement and optimization was made in
December 1980. An extra bonus in this selection was
the benefits of earlier in-house experience with the
Viggen delta canard aircraft.
quirement for high wing sweep is in contrast to those edge flap position is a function of Mach number and
essential for good subsonic turn performance favour- angle of attack. At supersonic speeds the leading
ing low sweep and large wing span. A comprise edge flaps are positioned up, with a maximum deflec-
choice of planform had to be made, and a leading tion of 5 degrees. This in effect "decambers" the wing,
edge flap was incorporated in order to keep good flow reducing the profile drag at low angles of attack and,
quality over the wing. Leading edge flap deflection at consequently, improves supersonic acceleration.
maneuver conditions in conjunction with camber and
twist on the wing is not primarily intended to keep the The effects of the wing leading edge flaps were sub-
flow attached, but rather to delay vortex burst and to stantiated in wind tunnel tests. An example of a drag
direct suction forces in a more forward direction so as polar is given in Figure 8. A reduction of 10 percent,
to decrease drag. illustrating the effect on lift dependant drag, is
obtained at Mach number 0.5 when the flaps are
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
dLE = Oo
dLE = 30
p
bending moment.
- 02 0 02 04 06 08 1 ~ C Lsv39869
The leading edge flap system is fully automatic, being
actuated by one hydralic power unit through two me-
chanical rotary actuators on each side. The leading Figure 8. Effect of LE flaps on lift deoendant drag
Page 8
ci P
7 I
ALPHAo
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
0 5 10 15 20 ALPHAo For take-off and landing the leading edge flap is kept
sv39 a70 in neutral position for optimum vortex lift. Deflecting
the flaps during landing approach would result in a
Figure 9. Effect of LE flaps on the buffeting level at loss in trimmed lift.
different Mach numbers.
At an angle of attack approximately 5 degrees above
that used for take-off and landing, the leading edge
flaps are automatically deflected for improved direc-
tional and lateral stability.
Canard
The movable canard is an important improvement
compared to the Viggen, which uses a fixed canard
with trailing edge flaps for increased canard trim lift
during take-off and landning. A movable canard sur-
face in combination with the four elevons, rudder and
leading edge flaps, controlled by the EFCS, intro-
duces many possibilities to improve flight perform-
ance and handling qualities in addition to unconven-
tional control modes.
It is possible to obtain a choice of either maximum lift
to drag ratio .or maximum lift, depending on what is
needed at a specific flight condition, using the canard
and trailing edge control surfaces in combination. At
?
cruise and maneuver conditions the canard and ele-
sv39 87 t von deflections are optimized for low trim drag. At low
speed, high lift is desired for short field performance I
rather than low drag. The emphasis is then to carry a
Figure IO. Effect of LE flaps on directional stability at high load on the canard to allow a substantial lift in-
low speed. crease by deflecting the wing elevon trailing edge
down for trim. Available space on the sidemounted air
intake is limiting for the canard size. Thus, to obtain a
high trim load on the canard, good canard high lift
Page 9
Speed brakes
INLET FLOW
DISToRTloN UNSHIELDED
.* I
I
dc = -15'
- a,I = 00 !
n.
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
SV3Q 073
ALPHAo
Wind tunnel tests performed on a schematic forebody
with this type of intake at low speed revealed several Figure 17. Inlet flow distortion of shielded air intake
advantages. Significant angle of attack improvements with schematic fore body
relative to the unshielded inlet (no canard) was noted
in pressure recovery and distortion level for the unde- Both of these concepts, the belly and the shielded
flected canard, Figure 16 and 17. Aerodynamic un- inlets, also had disadvantages in the form of higher
loading of the canard to obtain a recovery in pitch re- supersonic drag, the need for a larger canard because
duced the shielding effect, as can be expected, but of the reduced carry-over loads as compared to the
also gave a serious reduction in yaw stability at high side-mounted intake concept, reduced available
angles of attack. At positive canard incidences flow space for external pylons on the fuselage, a compli-
separation on the canard and at the fuselage junction cated gun and landing gear installation and a higher
was observed especially at the lower intake mass flow radar cross sectional area.
ratios that occur at high speed. Thus, the side-mounted swept and staggered bifur-
cated inlets were retained as they were found to offer
the best allround performance, and also, because it
PTJPT~
INLET PRESSURE
RECOVERY
M = .18
Ag/A, = 2.9
b=o 1 reduced the development risks.
Drag reduction
Although the 2105 configuration in the evaluation was
superior to the aft tail 2102 in supersonic performance,
supersonic drag was still of major concern. An ambi-
tious goal of reducing supersonic wave drag by 20
percent was established. Tough requirements on re-
duced maximum cross-sectional area and an opti-
mized area distribution in the low supersonic Mach
I
ALPHA0 number region led to comprehensive fuselage rede-
sign. Configuration 2108, also comprising a new, more
Figure 16. Inlet pressure recovery of shielded air swept wing with a compound trailing edge sweep was
intake with schematic fore body. defined, Figure 18. The cross sectional area distribu-
tion for 2108 and 2105 are shown in Figure 19. The
fuselage cross section behind the canard is of a
marked blended-body type. The structural composi-
tion of this type of fuselage gives advantages both in
a large internal volume for easy installation of acces- "!
sories as well as low drag. f
i
Page 11
_.nnn /" , :i
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
I sv39 875
cnl I M = 0.95
I I I
I
L l u3
I
Figure 1%. Configuration 2108.
t SV39 876
Figure 2 1. Configuration 2 I 7 0.
C"
7
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
economy.
A project called 21 11 was defined in September 1981.
As a realistic starting point the fuselage and fin of proj-
ect 2108 was used with a new wing and canard incor-
porating the Rockwell design.
The wing design was optimized for subsonic turn per-
formance and supersonic acceleration in accordance
with the basic requirements.
In addition to aeroelastic tailoring, sophisticated flap
systems were needed in order to reconfigure from the
minimum cambered wing at supersonic speeds to the
highly cambered wing necessary to keep the flow at-
tached up to high lift coefficients in subsonic maneu- Figure 26. Join?Rockwell- Saab study configuration
vers. 21 17-4.
Optimization of the configuration geometry was ac- Comparatively complex structural design was applied
complished by a sequence of analytical design and on the 2111 in the effort to meet the design objectives.
wind tunnel test iterations performed jointly by Saab The wing flap system comprised three element full
and Rockwell. span leading and trailing edge flaps. The outboard
Four subvariants were derived in the design process two elements of the leading edge flap were double-
with slightly different geometric properties. The final hinged for optimal camber in subsonic maneuvers,
version, 2111-4, is shown in Figure 26. Compared to Figure 27. Spanwise blowing was introduced to in-
the reference delta-canard configuration 21 10, the crease canard authority in landing approach, but was
wing area could be kept some 13 percent smaller, the subsequently replaced by a canard flap. "Back-porch''
wing leading edge sweep was 10 degress less while type air brakes were also used as longitudinal trim
the span was 9 percent larger. devices and for pitch control at high angles of attack.
,,/'
0* -. - 2111-r
'-*
21 10 I
b 10 ALPHA' 20
SV39 881
Gripenwggen comparison
As previosly mentioned, the performance goals re-
n lated in a general way to the performance of
Sweden's current front line combat aircraft, the Saab
JA 37 Viggen. Some important predicted JAS Gripen
data are compared to actual JA 37 data in Figures
31 -33.
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
Different wind tunnels with suitable performance re- One important requirement was low life-cycle cost. An
garding Reynolds number, model scale and availabil- indication of in-service operational cost for the Gripen
ity have been used in the areodynamic development in comparison with the Viggen as a representative of
program. current generation combat aircraft is presented in Fig-
ure 33, where fuel consumption over distance versus
The low speed tunnels used include low and high Mach number is presented. Fuel economy is signifi-
speed tunnels in Sweden, Switzerland, United King- cantly better for the Gripen, with savings in the order of
dom, Netherlands, France, Canada and US, 30 - 50 percent.
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on February 18, 2016 | http://arc.aiaa.org | DOI: 10.2514/6.1991-3195
Figure 32.
-JAS 39
-
- - - JA 37
MAX POWE
'\
SV3Q 082
J
\
\ / numbers less than one, but the difference is within a
normal prediction accuracy band of 10 percent.
W
3
U
2
H=9KM
JAS 39
--- JA 37
MACH FLIGHT TEST
MACH
I Summary