Content uploaded by Giuliano Ranuzzi
Author content
All content in this area was uploaded by Giuliano Ranuzzi on Mar 12, 2014
Content may be subject to copyright.
FLPP PROGRAMME: IXV VEHICLE HYPERSONIC CFD SIMULATIONS
RONCIONI P. - CIRA, Italy
RANUZZI G. - CIRA, Italy
MARINI M. - CIRA, Italy
COSSON E. - AstriumST, France
Abstract. In this paper the main and relevant numerical results of the hypersonic CFD characterization of IXV vehicle, both
in flight and wind tunnel conditions, will be analysed and reported. The activities have been carried out by the
Aerothermodynamics and Space Propulsion laboratory of the Italian Aerospace Research Centre (CIRA) in the framework of
the ESA FLPP programme (Future Launchers Preparatory Programme). CIRA is involved, under the supervision of Astrium-
ST, in the activities aimed at the generation of the Intermediate eXperimental Vehicle (IXV) Aerodynamic and
Aerothermodynamic Data Base and the aerothermal environment characterisation, all of this mainly by means of CFD
computations in both flight and wind tunnel conditions. A detailed analysis of all hypersonic relevant phenomena is reported
in the paper, together with a description of configuration, flight condition and modelling effects.
Key words: Hypersonic Flow, Aerothermodynamics, Reentry Vehicles, Heat Transfer.
1. INTRODUCTION
The general objective of the IXV project (Intermediate eXperimental Vehicle), led by NGL Prime in the
framework of the ESA FLPP program, is to improve European capabilities in the strategic field of
atmospheric re-entry for space transportation, exploration, and scientific applications [1]. Several studies
on experimental vehicle concepts and improvements of critical reentry technologies have been
undertaken in recent years by ESA (ARD), France (Pre-X), Germany (Phoenix) and Italy (USV), in order
to consolidate their worldwide position in this strategic field [2]. Both on-ground and in-flight tests have
been scheduled and performed with the main goal to develop the future Reusable Launch Vehicle
(RLV) technology.
The aerothermodynamic studies can support and address both aeroshape consolidation and mission
analysis [3], whose goal is to minimize the heat fluxes to the windward part of the vehicle (nosecap) and
control surfaces (flaps). An enhanced flap configuration can allow reducing the necessary trimming
deflection angles and, consequently, the flap peak heating. Moreover, validation and update of on-
ground tools (wind tunnels and CFD codes) can be obtained by means of in-flight experimentation.
The flow field developing around IXV vehicle, a 4.4 m long slender body configuration which performs a
hypersonic guided re-entry controlled by means of a couple of active body flaps, has been characterized
through several detailed flow simulations, aimed at evaluating peculiar high speed phenomena such as
the effect of angles of attack and sideslip, symmetric and asymmetric flaps deflection, turbulence, non
equilibrium and Mach number. CFD numerical rebuilding of VKI Long-Shot wind tunnel tests have been
also performed for supporting the extrapolation-to-flight procedure.
2. CFD METHODOLOGY
The numerical code used to carry out the aerothermodynamic analysis of the IXV vehicle is the CIRA
code H3NS [4] that solves the Reynolds Averaged Navier Stokes equations in a density-based block-
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
structured finite volume approach with a cell centred, Flux Difference Splitting second order ENO-like
upwind scheme for the convective terms. Viscous fluxes are computed with a classical centred scheme.
Time integration is performed by employing an explicit multistage Runge-Kutta algorithm coupled with
an implicit evaluation of source terms. The H3NS code is available in both sequential and parallel
version.
Working gas is air that can be properly modelled in the hypothesis of ideal gas, equilibrium gas and
thermo-chemical non equilibrium gas. Different two-equation k-ε turbulence models are available in
H3NS for eddy viscosity calculation. Present simulations have been performed by using the k-ε
turbulence model with compressibility effects corrections [5], while laminar-to-turbulence transition is
imposed across surface lines (i.e. a transition front).
2.1. CFD Test Matrix
Astrium ST is responsible for the aerodynamic and aerothermodynamic database development of the
IXV vehicle and in order attain this target several (overlapping for cross-check purposes) work packages
have been assigned to various suppliers. Since October 2006, CIRA Aerothermodynamics and Space
Propulsion Laboratory has had in charge the task to conduct numerical simulations aimed to study the
IXV vehicle’s aerothermodynamics both in Flight (phases B
1
and B
2
at M
∞
=10.0, 15.0, 17.7) and Wind
Tunnel (VKI Long-Shot, M
∞
=14, phase B
2
) conditions. The main flow parameters are reported in Tab. 1,
where Eq.” stands for thermo-chemical equilibrium, “Neq.” for thermo-chemical non-equilibrium, “PG” for
perfect gas, “lam” for laminar regime, “transHL” for transition at body flap hinge line and “turb” for fully
turbulent regime.
M
∞
Re
∞,L
H (km) p
∞
(Pa) T
∞
(K) Condition
10.0 6.79·10
5
52.1 61.19 267.8 Flight, transHL, turb, Eq.
≈14.0 1.34÷2.67⋅10
6
- - - VKI-Long-Shot, lam, transHL, turb, PG
15.0 4.86·10
5
58.7 26.30 250.6 Flight, transHL, turb, Eq.
17.7 2.68·10
5
64.6 11.51 234.2 Flight, transHL, turb, Eq., Neq.
Tab. 1. IXV re-entry flight conditions (Reynolds number based on vehicle length L).
Note that Reynolds number is based on vehicle length L, that is equal to 4.4 m for flight configuration
and 0.2 m for wind tunnel model (that is scaled down by 1:22).
Three geometric configurations have been used to carry out numerical simulations: half body for
longitudinal symmetric cases, symmetric full body to account for sideslip effects and asymmetric full
body for longitudinal cases with different left and right flap deflection. For this last case, let us consider
that being δ
e
the deflection angle acting the body flap as an elevon, and δ
a
the deflection angle acting
the body flap as an aileron: the right body flap is deflected by δ
right
= δ
e
+ δ
a
and the left one deflected by
δ
left
= δ
e
- δ
a
.
The comparison of some subsets of flow simulations can therefore allow us to analyze different effects
as it will be shown in the next sections.
3. NUMERICAL RESULTS
The IXV re-entry vehicle is, from a general point of view, a 4.4 m long slender body configuration which
performs a hypersonic guided re-entry controlled by means of a couple of active body flaps (see Fig. 1).
The wind tunnel configuration used in VKI Long-Shot facility is simply a 1:22 scaled down model.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
Several grids have been generated by using the commercial software ANSYS ICEMCFD
®
. In particular,
for each simulation a different bow shock fitting has been necessary. An example of the block
decomposition is reported in the following Fig. 2, that shows clearly the grid surface topology of base
and flap region. A close-to-the body O-grid surrounding both the fuselage and the flap is used in order
to resolve the boundary layer with at least 25 grid points, while the number of grid points outside of the
boundary layer up to the far field is also around 25. The minimum normal grid spacing (at wall) is of the
order 10
-5
m. The number of computational cells for the half body configuration is about 2.3 millions,
distributed in 118 different blocks.
An initial assessment phase was conducted [6] aimed at evaluating the sensitivity of numerical results to
both the physical modeling (inviscid, laminar, turbulent) and grid refinement. The final selection
concerning the numerical prediction strategy to be adopted (physical model and number of cells) was a
compromise between accuracy of results and hardware availability.
A rapid grid convergence was observed for what concerns the global aerodynamic coefficients (and
pressure distributions also): low CPU time demanding numerical simulations (0.3÷0.4 million of cells)
can be used if interested to this goal only. However, especially for what concerns heat flux and
temperature distributions, the use of the finest grid (2.3 millions of cells for half body) was necessary to
provide reliable results for all the relevant aerothermodynamic parameters ([7]).
The complex phenomena characterizing the IXV vehicle flow field (strong bow shock, shock-shock
interaction, shock-wave boundary-layer interaction, base flow, flow spillage through the gap between
the flaps, huge expansions) make the numerical simulations very stiff, especially for the initial start-up
phase afforded with simplified numerical settings. Anyway, at the end of this start-up phase the actual
far-field and wall boundary conditions have been applied. A shock fitting-like grid adaptation and a
suitable distribution of points within the shock layer has been necessary in order to overcome the
numerical oscillations around the nose (i.e. the well known “carbuncle” phenomenon) ([8]).
The simulations foreseen in the test matrix of Tab. 1 can allow for the analysis of several phenomena:
variations of angle of attack and angle of sideslip, variation of flap deflection, effect of turbulence, effect
of Mach number (i.e. total enthalpy), effect of thermo-chemical non equilibrium, and effect of asymmetric
flaps configuration. In this section computed results are presented in terms of longitudinal distribution of
heat flux and pressure coefficient at the section Y=0.4 m, which corresponds to the flap half span.
The transition from laminar to turbulent boundary layer regime imposed at flap hinge line has been
considered as the best choice avoiding also the total disappearance of recirculation zone. In fact, the
transition imposed at separation front has shown a drastic reduction of separation zone extent. An
unsteady cyclic behavior could be supposed with the low Reynolds number (Re
∞,L
=2.68·10
5
)
characterizing these simulations: the turbulent transition is induced by the laminar separation, after that
Fig. 2: Grid surface topology of base and flap region.
Fig. 1: IXV_1.2. Front isometric view.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
a re-laminarization of the flow can be supposed due to the bubble reduction, and then a new laminar
separation with a new transition phenomena.
In the present computations the value of emissivity coefficient has been set to ε=0.80. Simulations with
different values of emissivity have been also carried out [6],[7], showing as a reduction of this parameter
causes higher values of temperature inside the boundary layer, and therefore an enlargement of the
separation pattern (increase of boundary layer thickness).
3.1. Effects of Configuration, Flight Condition and Modeling Parameters
The effect of the angle of attack in transitional flow hypothesis, for the M
∞
=15 test cases, seems to be
strong for the zone upstream the body flap and from α = 40 deg to α=45 deg (see Fig. 3). Smaller
differences are predicted between the α = 45 deg and α = 50 deg cases. For the body flap zone the
driving parameter does not seem to be the angle of attack, but the flap deflection as expected [9], [10]
and confirmed in Fig. 4. Increasing flap deflection causes a larger recirculation (i.e. separation front is
anticipated) and a higher peak heating on the body flap, that seems to increase about linearly with flap
deflection angle up to δ
flap
= 15 deg. However, a more complex vortex structure can be found for the
case M
∞
=15, α=45 deg, δ
flap
= 20 deg (see Fig. 5-Fig. 7) with respect to other cases. A nested vortex
and a saddle point (SP) appear inside the main separation zone around the flap hinge line, where the
transition from laminar to turbulent flow has been applied.
The interaction of three shock waves (bow shock, main separation shock and main reattachment shock,
see Fig. 5 on symmetry plane and Fig. 7 on flap surface) generates on over-load of both pressure (of
roughly 50%, see Fig. 6) and heat flux (of roughly 40%, see Fig. 4), that should be carefully taken into
account when defining the trimming conditions for the IXV vehicle, and a secondary (downstream of the
main reattachment line) separation zone.
Fig. 4. Flap deflection effect. M
∞
=15.0, section Y=0.4 m.
Fig. 3. Angle of attack effect. M
∞
=15.0, section Y=0.4 m.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
Fig. 4 shows the additional peak heating (about 900 kW/m
2
) over the flap due to the flow features
described above, while the average value (about 650 kW/m
2
) upstream of this additional peak is in trend
with the δ
flap
= 5, 10, 15 deg previous simulations with transition at flap hinge line. Figure 6 shows the
analogous behavior for what concerns the pressure coefficient distribution; there is an overload that
yields a maximum pressure coefficient value of 4.5 instead of a value that, extrapolating from the
upstream line trend, should be around 3.1.
Fig. 7: Skin friction lines distribution near the flap. M
∞
=15.0, α=45 deg, δ
flap
=20 deg.
Fig. 6: Flap deflection effect. M
∞
=15.0, Y=0.4 m.
Fig. 5: Interaction of three shock waves.
M
∞
=15.0, α=45 deg, δ
flap
=20 deg.
Second separation zone
Main separation line
Main reattachment line
Nested vortex
S
R
S
R
S
R
SP
SP
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
The fully turbulence hypothesis affects the nose-fuselage zone while the body flap region exhibits, for
the heat flux distribution, values of the same order of magnitude as for the transitional cases (see Fig. 8
where the δ
flap
= 10 deg transitional case is also reported as reference). A 75% higher heat flux is
predicted on the fuselage windside (i.e. section X=3 m) between the fully turbulent case and the
transitional case (locally the flow is laminar). The increase of flap peak heating with flap deflection angle
is kept. Fully turbulent simulations have been carried out also at M
∞
=10 (Re
∞,L
=6.79⋅10
5
). A linear trend
of heat flux peak with respect to the surface deflection can be observed over the flap zone in Fig. 9,
where the case M
∞
=15, δ
flap
= 5 deg has been reported for comparison purposes, clearly showing the
effect of Mach number (i.e. of the different level of total enthalpy).
Fig. 10 shows the heat flux distribution at the longitudinal sections Y=0.4 m (right) and Y=-0.4 m (left)
with asymmetric flap deflections, i.e. δ
right
=15 deg and δ
left
=5 deg. For M
∞
=15 a strong difference
(about 200 kW/m
2
) is predicted mainly in body flap peak heating and separation. A noticeable flow
spillage is predicted between the two flaps, as it can be seen from the skin-friction pattern reported in
Fig. 11 where the separation bubble of the right flap (δ
right
=15 deg) crosses the symmetry plane and
reaches the hinge line of the left flap (δ
left
=5 deg), where no separation is predicted. A view of the
streamtraces around the two flaps is shown in Fig. 12, where the flow spillage is clearly evidenced.
Pressure coefficient contours slices at longitudinal section Y=0.4 m and transversal sections X=3.5 m
and X=4.8 m are reported in Fig. 13, describing in detail the IXV vehicle’s windside bow shock, the
Fig. 11: Visualization of flow spillage between the two
asymmetric body flaps.
M
∞
=15.0, α=45 deg, δ
right
=15 deg, δ
left
=5 deg.
Fig. 10: Flap deflection effect for asymmetric flaps
configuration. M
∞
=15.0, section Y=0.4 m.
Fig. 8:Flap deflection effect. Fully turbulent case.
M
∞
=15.0, section Y=0.4 m.
Fig. 9: Flap deflection effect. Fully turbulent case.
M
∞
=10.0, section Y=0.4 m.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
shock-shock interaction around the right flap, the separation and reattachment shocks and the
asymmetric flow compression over the two asymmetrically deflected control surfaces.
A maximum difference of about 50 kW/m
2
is predicted, mainly at the end of the flap, for M
∞
=10.0 (see
Fig. 10).
The comparison with the results of half-body computations shows also that the asymmetric configuration
at zero angle of sideslip does not affect the heat flux distribution over the flaps. Since a fully turbulent
flow hypothesis has been assumed, the separation bubbles are reduced in size with respect to the
transitional cases. The right flap separated zone reaches but does not cross the vehicle’s symmetry
plane, and consequently does not affect the flow field around the left flap.
The main effect of the sideslip angle variation is on the maximum heat flux nose position. Fig. 14 and
Fig. 15 report the case at M
∞
=10 where the location of the maximum heat flux point is located quite
away from the stagnation point, where the value is 179 kW/m
2
(the maximum is located at Y=0.4 m and
is 188 kW/m
2
, while at Y=-0.4 m the value is 165 kW/m
2
). From Fig. 14 the difference in the flap region
between the two opposite sections (Y = ± 0.4 m) can be observed (about 20 kW/m
2
); the case at β=0
deg is also reported, being located in the middle position as expected. For M
∞
=15 the point of maximum
heat flux moves laterally from Y=0.260 m to Y=0.387 m with values that are of 445, 462 and 482 kW/m
2
,
respectively for β=0 deg, β=5 deg and β=8 deg [7], [11].
Fig. 15: Maximum Heat Flux and Stagnation Point location
for sideslip simulation. M
∞
=10, α=45 deg, β=-8 deg.
Fig. 14: M
∞
=10, α=45 deg, β=-8 deg. Heat flux
distribution and comparison with the symmetric case.
Fig. 13: Slices of pressure coefficient contours in the
windside region.
M
∞
=15.0, α=45 deg, δ
right
=15 deg, δ
left
=5 deg.
Fig. 12: Streamtraces around the two asymmetric body
flaps.
M
∞
=15.0, α=45 deg, δ
right
=15 deg, δ
left
=5 deg.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
The real gas non equilibrium effects are more evident in the nose-fuselage windside region (i.e. in the
shock layer) where the reduced relaxing time makes the equilibrium hypothesis less realistic, and the
CFD prediction results too much conservative (see Fig. 16). A small (and opposite) effect is also
observed for the body flap peak heating.
The comparison between the M
∞
=15.0 and M
∞
=17.7 transitional flow results (see Fig. 17) does not
show great differences in the predicted heat flux over the body flap region, whereas the effect of an
increasing Mach number (i.e. total enthalpy) is the reduction of the recirculation extent, as expected.
Fig. 18: Comparison of skin friction lines. Numerical simulations of flight (left) and wind tunnel (right) conditions.
Fig. 17: Comparison of heat fluxes. M
∞
=15.0 and
M
∞
=17.7, α=50 deg, section Y=0.4 m.
Fig. 16: Equilibrium vs Non Equilibrium. M
∞
=15.0, α=45
deg, section Y=0.4 m.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
For what concerns the wind tunnel test campaign conducted at VKI Long-Shot facility (no
measurements are still available at the time of writing) and the numerical rebuilding of these
experimental tests, a test matrix of thirteen CFD runs has been foreseen in the nominal test chamber
conditions provided by VKI, and some preliminary CFD results are reported in this work.
A significant comparison between one simulation in flight condition (M
∞
=15, Re
∞,L
=4.86·10
5
) and one
simulation in wind tunnel condition (M
∞
=14, Re
∞,L
=1.34·10
6
) is reported in Fig. 18-Fig. 20, respectively
in terms of skin friction lines pattern around the flap, heat flux and pressure coefficient longitudinal
distributions at section Y=0.4 m. For both simulations δ
flap
= 15 deg and laminar-to-turbulence transition
is imposed at flap hinge line. The reproducibility of flight conditions seems to be qualitatively good, both
in terms of flow features and surface properties: a similar separation bubble can be observed in Fig. 18.,
the larger pattern of wind tunnel case being due to a combined effect - on the transitional shock wave
boundary layer interaction - of ideal gas hypothesis (with respect to the equilibrium gas assumption in
flight) and greater Reynolds number. The higher value of heat flux in wind tunnel conditions (see Fig.
19) is mainly due to fixed wall temperature hypothesis with respect to the radiative equilibrium assumed
in flight, while the higher value of flight pressure coefficient (see Fig. 20) is due to a closer location of
the bow shock (chemical equilibrium against perfect gas).
3.2. Global and Distributed Aerodynamic Coefficients
The analysis of the functional dependence (with respect to the main parameters) of global aerodynamic
coefficients (C
L
, C
D
, C
S
, C
Mx
, C
My
, C
Mz
) are reported in this section. The predicted aerodynamic
coefficients in both flight and wind tunnel conditions are reported in this section. The reference
quantities used for forces and moments normalization are L
ref
=4.4 m and S
ref
=7.26 m
2
, whereas the
aerodynamic moments are calculated with respect to the Moment Reference Centre (MRC), i.e.
X
MRC
=2.552 m (from the nose), Y
MRC
=0.000 m, Z
MRC
=-0.110 m. For the wind tunnel model all the
reference quantities have been simply scaled by 1:22.
Lift (C
L
), drag (C
D
) and pitching moment (C
My
) coefficients in function of angle of attack and flap
deflections are reported, respectively, in Fig. 21, Fig. 22 and Fig. 23, the angle of attack variations on
the left, and the flap deflection variations on the right. Nearly linear trends of aerodynamic coefficients
with both angle of attack and (symmetric) flap deflection have been predicted, with the unique exception
of lift coefficient (Fig. 21, left), that shows a plateau moving from α=45 deg to α=50 deg (at both Mach
numbers), this being probably due to deteriorating effects of strong vortices detaching from vehicle’s
leeside (loss of vortex lift).
Fig. 20: Comparison of pressure coefficient distribution at
Y=0.4 m. Numerical simulations of flight and wind tunnel
conditions.
Fig. 19: Comparison of heat flux distribution at Y=0.4 m.
Numerical simulations of flight and wind tunnel
conditions.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
No significant effects of fully turbulent flow assumption (with respect to cases with transition at hinge
line) on aerodynamic coefficients are observed at the same Mach number (M
∞
=15), while a quite
remarkable effect of Mach number can be observed, in the fully turbulent flow hypothesis, from figures
reporting the coefficients versus the flap deflection angle (right of Fig. 21, Fig. 22 and Fig. 23).
Regarding non equilibrium flow assumption (with respect to the equilibrium flow result, M
∞
=15), a slight
loss of lift (Fig. 21, left) and a consequent reduction (around 10%) of pitching moment (see Fig. 23, left)
have been predicted, mainly due to the lower surface pressure on the IXV vehicle (i.e. the bow shock is
closer to the vehicle in equilibrium flow conditions).
Aerodynamic efficiency (E=C
L
/C
D
) decreases with angle of attack and flap deflection (although very
slightly), and it is not affected by Mach number, turbulence modeling and non equilibrium assumption.
Clearly linear trends of longitudinal aerodynamic coefficients (C
L
, C
D
, C
My
) with flap deflection have been
predicted, and no crisis of flap efficiency has been found out up to a deflection of 15 deg, as the
distribution of C
My
shows (see Fig. 23, right) and the behavior has been found not affected by Mach
number. This is due to the fact that the flow over the body flap is assumed fully turbulent for all flow
simulations, and the reattachment line is not significantly affected by flap deflection (see Fig. 23), thus
pressure recovery on the flap is conserved. The particular phenomenology predicted for the transitional
20-deg flap deflection M
∞
=15 case, and characterized by a localized overpressure and overheating over
the flap, does not seem to affect the regular trend of aerodynamic coefficients.
The effect of sideslip on rolling and yawing moment coefficients is shown in Fig. 24, whereas no effect is
predicted on longitudinal coefficients as expected. Linear trends of C
Mx
and C
Mz
with the sideslip angle
are observed (the same for the side force coefficient, C
s
), while a strong effect on these moments is
observed with the asymmetric flaps configuration (δ
right
=15 deg and δ
left
=5 deg) at β=0 deg and at both
M
∞
=15 and M
∞
=10, thus producing a rather strong lateral-directional control action.
β
ββ
β=0 deg, symmetric flaps (δ
δδ
δ
L
=δ
δδ
δ
R
=10 deg)
0.450
0.475
0.500
0.525
0.550
0.575
0.600
0.625
0.650
35 40 45 50 55
angle of attack [deg]
C
L
M=15.0, eq, transHL
M=17.7, eq, transHL
M=15.0, eq, turb
M=15.0, neq, transHL
M=10.0, eq, turb
M=14 (WT), pg, lam
M=14 (WT), pg, transHL
M=14 (WT), pg, turb
β
ββ
β=0 deg, α
αα
α=45 deg
0.450
0.475
0.500
0.525
0.550
0.575
0.600
0.625
0.650
0 5 10 15 20
flap def lection [deg ]
C
L
M=15.0, eq, transHL
M=15.0, eq, turb
M=17.7, eq, transHL
M=10.0, eq, turb
M=14 (WT), pg, lam
M=14 (WT), pg, transHL
M=14 (WT), pg, turb
Fig. 21: Lift coefficient in function of angle of attack (left) and flap deflection (right).
β
ββ
β=0 deg, symmetric fla ps (δ
δδ
δ
L
=δ
δδ
δ
R
=10 deg)
0.500
0.600
0.700
0.800
0.900
1.000
1.100
1.200
35 40 45 50 55
angle of attack [deg]
C
D
M=15.0, eq, transHL
M=17.7, eq, transHL
M=15.0, eq, turb
M=15.0, neq, transHL
M=10.0, eq, turb
M=14 (WT), pg, lam
M=14 (WT), pg, transHL
M=14 (WT), pg, turb
β
ββ
β=0 deg, α
αα
α=45 deg
0.500
0.600
0.700
0.800
0.900
1.000
1.100
1.200
0 5 10 15 20
flap def lection [de g]
C
D
M=15.0, eq, transHL
M=15.0, eq, turb
M=17.7, eq, transHL
M=10.0, eq, turb
M=14 (WT), pg, lam
M=14 (WT), pg, transHL
M=14 (WT), pg, turb
Fig. 22: Drag coefficient in function of angle of attack (left) and flap deflection (right).
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
β
ββ
β=0 deg, symmetric flaps (δ
δδ
δ
L
=δ
δδ
δ
R
=10 deg)
-0.140
-0.120
-0.100
-0.080
-0.060
-0.040
-0.020
0.000
35 40 45 50 55
angle of att ack [deg]
C
My
M=15.0, eq, transHL
M=17.7, eq, transHL
M=15.0, eq, turb
M=15.0, neq, transHL
M=10.0, eq, turb
M=14 (WT), pg, lam
M=14 (WT), pg, transHL
M=14 (WT), pg, turb
β
ββ
β=0 deg, α
αα
α=45 deg
-0.140
-0.120
-0.100
-0.080
-0.060
-0.040
-0.020
0.000
0 5 10 15 20
flap def lection [de g]
C
My
M=15.0, eq, transHL
M=15.0, eq, turb
M=17.7, eq, transHL
M=10.0, eq, turb
M=14 (WT), pg, lam
M=14 (WT), pg, transHL
M=14 (WT), pg, turb
Fig. 23: Pitching moment coefficient in function of angle of attack (left) and flap deflection (right).
α
αα
α=45 deg
-0.010
-0.009
-0.008
-0.007
-0.006
-0.005
-0.004
-0.003
-0.002
-0.001
0.000
0 2 4 6 8 10
side slip angle [de g]
C
Mx
dR=10°, dL=10°, eq, transHL, M=15.0
dR=15°, dL=5°, eq, transHL, M=15.0
dR=10°, dL=10°, eq, turb, M=10.0
dR=15°, dL=5°, eq, turb, M=10.0
dR=5°, dL=5°, pg, lam, M=14 (WT)
Lineare (dR=10°, dL=10°, eq, transHL, M=15.0)
α
αα
α=45 deg
-0.003
-0.002
-0.001
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0 2 4 6 8 10
side slip angle [deg]
C
Mz
dR=10°, dL=10°, eq, trans HL, M=15.0
dR=15°, dL=5°, eq, tr ansHL, M=15.0
dR=10°, dL=10°, eq, turb, M=10.0
dR=15°, dL=5°, eq, turb, M=10.0
dR=5°, dL=5°, pg, lam, M=14 (WT)
Lineare (dR=10°, dL=10°, eq, tra nsHL, M=15.0)
Fig. 24: Rolling moment (left) and yawing moment (right) coefficients in function of sideslip angle.
In the Fig. 21-Fig. 24 are reported also the global aerodynamic coefficients predicted in VKI Long-Shot
wind tunnel conditions which substantiate qualitatively all the trends found out in flight conditions, thus
confirming the good reproducibility of IXV vehicle’s aerodynamics in the Long-Shot facility.
Very important inputs required at system level are the distributed and local aerodynamic coefficients.
The generic “distributed aerodynamic coefficient” at a given location along the abscissa X is the
aerodynamic coefficient obtained by means of the integration of surface forces (pressure and friction)
from the nose to the local point X, and the “local aerodynamic coefficient” is the same type of coefficient
obtained with a local integration (from X to X+dX).
Fig. 26: Local aerodynamic coefficients.
Fig. 25: Distributed aerodynamic coefficients.
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
Fig. 25 shows a quite regular behavior for the distributed aerodynamic coefficients (the pressure
coefficient on the leeside and on the base is about zero), the change of the slope, for C
N
and C
A
,
between the station X=3 m and about X=4.4 m, and a quite strong increase over the flap (δ
flap
= 20 deg)
for all the coefficients. The slope of the pitching moment coefficient (see Fig. 25) changes sign at about
X=2.5 m (where the reference point is located) as well as the local contribution becomes negative (see
Fig. 26). Note as the main contribution to the axial coefficient (C
A
) is located mainly on the nose and flap
regions (see Fig. 26).
4. CONCLUSIONS
The aim of this work has been to report the CIRA Aerothermodynamics and Space Propulsion
Laboratory activities to support Intermediate eXperimental Vehicle (IXV) Aerodynamic and
Aerothermodynamic Data Base development in the framework of the ESA program “FLPP – NGL”, by
means of CFD numerical computations in both flight and VKI Long-Shot wind tunnel conditions.
The flow field developing around the IXV vehicle has been characterized by means of detailed flow
simulations aimed at evaluate several high speed phenomena such as the effect of angle of attack and
sideslip, flap deflection, turbulence, non equilibrium, Mach number (i.e. total enthalpy) and asymmetric
flaps configuration.
For the body flap zone the driving parameter is the flap deflection as expected and confirmed by results.
In general, an increasing flap deflection causes a larger recirculation (i.e. the separation front is
anticipated) and a higher peak heating on the body flap with a linear trend except for the particular case
M
∞
=15, α=45 deg, δ
flap
=20 deg, where a more complex vortex structure has been predicted for the
transitional case with respect to other cases. A nested vortex appears inside the main separation zone
around the flap hinge line. The interaction of three shock waves (bow shock, main separation shock and
main reattachment shock) generates on over-load over the flap of both pressure and heat flux, of
roughly 50% and 40% respectively that should be carefully taken into account when defining the
trimming conditions for the IXV vehicle, and a second separation zone.
In asymmetric flap conditions a difference of about 50 kW/m
2
is predicted in body flap peak heating at
M
∞
=10 and 200 kW/m
2
at M
∞
=15. No significant effect of sideslip angle on heat flux is predicted apart
from a lateral movement of the maximum heat flux nose position, and a slight increase of it;
The non equilibrium effects (only considered for M
∞
=17.7) are more evident inside the windside shock
layer where the reduced relaxing time makes the equilibrium hypothesis less realistic and the CFD
predictions too much conservative;
The fully turbulence hypothesis affects the nose-fuselage zone while the body flap region exhibits
values of the same order of magnitude as for the transitional cases, and the same trend of flap peak
heating with flap deflection.
Nearly linear trends of aerodynamic coefficients with both angle of attack and (symmetric) flap deflection
have been predicted. A plateau of lift coefficient is observed moving from α=45 deg to α=50 deg (at M
∞
=15, 17.7).
No significant effects of fully turbulent flow assumption (with respect to cases with transition at hinge
line) on aerodynamic coefficients are observed at the same Mach number (M
∞
=15), while a quite
remarkable effect of Mach number can be observed, in the fully turbulent flow hypothesis (from M
∞
=15
to M
∞
=10). The non equilibrium flow assumption exhibits a slight loss of lift and a consequent reduction
(around 10%) of pitching moment.
No crisis of flap efficiency has been predicted up to a deflection of 15 deg and beyond. The particular
phenomenology predicted for the transitional 20-deg flap deflection M
∞
=15 case, and characterized by a
P. RONCIONI et al./FLPP Programme: IXV Vehicle Hypersonic CFD Simulations
localized overpressure and overheating over the flap, does not seem to affect the regular trend of
aerodynamic coefficients.
Linear trends of C
Mx
, C
Mz
and C
s
with the sideslip angle are observed. A strong effect on these
coefficients is observed with the asymmetric flaps configuration (δ
right
=15 deg and δ
left
=5 deg) at β=0
deg, thus yielding strong lateral-directional control actions.
The global aerodynamic coefficients predicted in VKI Long-Shot wind tunnel conditions substantiate
qualitatively all the trends found out in flight conditions, thus confirming the good reproducibility of IXV
vehicle’s aerodynamics in the Long-Shot facility.
The comparison of CFD results with experimental measurements, either in terms of global aerodynamic
coefficients either in terms of surface properties (pressure, heat flux), will close the loop and support
strongly the extrapolation-to-flight procedure.
5. REFERENCES
[1] Baiocco P., Guedron S., Plotard P., Moulin J., “The Pre-X Atmospheric Re-entry Experimental
Lifting Body: Program Status and System Synthesis”, 57
th
IAC Congress, Valencia, Spain, 2-6
October 2006.
[2] Tumino G., Gerard Y., “Europe Among the World Players in Atmospheric Reentry”, ESA bulletin
128, November 2006.
[3] Oswald J. et al., “DLR-ONERA accurate CFD support to the Pre-X project”, 6
th
International
Symposium on Launchers Technologies, Munich, Germany, 8-11 November 2005.
[4] Ranuzzi G., Borreca S., “CLAE Project. H3NS: Code Development Verification and Validation”,
CIRA internal report. CIRA-CF-06-1017, September 2006.
[5] Grasso F., Falconi D., “Shock-Wave/Turbulent Boundary-Layer Interactions in Nonequilibrium
Flows”, AIAA Journal, Vol. 39, No. 11, 2131-2140, November 2001.
[6] Roncioni P., Ranuzzi G., Marini M., “FLPP-IXV Project – Phase B1.2 – Preliminary CFD Activities
Synthesis Report”, CIRA Internal Report, CIRA-CF-07-0223, March 2007.
[7] Roncioni P., Ranuzzi G., Marini M., “FLPP-IXV Project - Phase B1.2 CFD Activities Synthesis
Report”, CIRA Internal Report, CIRA-CF-07-1174, October 2007.
[8] Pandolfi M., D’Ambrosio D., “Numerical Instabilities in Upwind Methods: Analysis and Cures for the
“Carbuncle” Phenomenon”, Journal of Computational Physics 166, 271–301 (2001).
[9] Marini M., “FESTIP Technology Developments in Aerothermodynamics for Reusable Launch
Vehicles – Body-Flap Efficiency Studies for a Simplified Two-Dimensional FSSC-15-OAE Concept
Vehicle Configuration”, Final Technical Report CIRA-CR-AEP-99-206, December 1999.
[10] Marini M., “Body-Flap Efficiency Prediction of a FESTIP Concept Vehicle”, Second International
Symposium on Atmospheric Reentry Vehicles and Systems, Arcachon, France, March 26-29,
2001.
[11] Roncioni P., Ranuzzi G., Marini M., Cosson E., “Hypersonic
CFD Characterization of IXV Vehicle”,
WEHSFF-2007 Conference, 19-22 November 2007, Moscow, Russia.
