123

The computations are currently underway. A first result obtained with TAU is illustrated in Figure 3.4-23 for the detailed versions of the ATR-3 hub.

Figure 3.4-23 Instantaneous wake structure represented by constant kinematic vorticity surfaces on the detailed hub.

3.4.4.2 Flow Field around Complete Configurations

Simulation of the Unsteady Flowfield of the Complete BO 105 Helicopter

A final milestone of the French-German CHANCE project was to simulate the unsteady flow for a complete helicopter configuration. The test case chosen on the German side was the BO-105 wind tunnel model in high speed forward flight.

The flight conditions are M = 0.18, MTip =

0.65 and = -5.2°. Fuselage, main and tail rotors, skids and support strut were included in the flow simulation. The computational mesh with 12 million cells in total was composed of a set of overlapping grids. Pressure distributions on the fuselage, main and tail rotors compared well with experimental data from the HeliNOVI wind tunnel campaign. An overview of the flow field is given in Figure 3.4-24.

Blind Test Computations of GOAHEAD Configuration

An investigation was carried out using the flight mechanics tool HOST weakly coupled to the RANS solver FLOWer to simulate the flow around a helicopter configuration under different flight conditions. The computational model refers to the GOAHEAD wind tunnel model. It consists of a 4.1 m NH90 fuselage

model, powered ONERA 7AD main rotor and reduced scale BO-105 tail rotor, a simplified rotor hub, a strut and slip ring fairing and 8m x 6m wind tunnel test section of 20 m length.

Figure 3.4-24 BO-105 wind tunnel model with pressure contours and vortex cores.

Two forward flight conditions were considered: Cruise and high-speed flight at Mach numbers of 0.204 and 0.259, at -2o and -7o angle of attack, respectively. The selected cases represent two critical flight conditions where tail shake and dynamic stall phenomena are expected to occur. Both the main rotor and the fuselage of the complete configuration were simulated separately for the cruise case. Whenever the main rotor was considered, weak fluid-structure coupling was applied to generate predefined thrust and propulsive force, and to obtain the deformation of the blades.

Multi-block grids around the different elements were combined via the Chimera approach to build the three different computational configurations. The total number of grid points used was 15 million distributed in 169 blocks. Figure 3.4-25 depicts the surface grid for the complete helicopter configuration.

The three configurations were compared in terms of aerodynamic loads, surface pressure distribution and wake structure to analyze the aerodynamic interaction between the rotating and non-rotating components of the helicopter. A part of the computations represents some of the blind test activities of the EU project GOAHEAD. A sample of the results is shown in Figure 3.4-26. In the load distribution on the main rotor, noticeable differences were observed between the

124

isolated rotor and full helicopter cases, but only negligible differences in power consumption were found. Also major changes in the fuselage loads and surface pressure were found.

Figure 3.4-25 Surface grid on the model and wind tunnel walls.

Figure 3.4-26 Computed streamlines and surface pressure contours in high speed dynamic stall case.

Tilt Rotor Simulation

The tilt rotorcraft is an aircraft which combines the advantages of vertical takeoff and landing capabilities, specific to helicopters, with the forward speed and range of a turboprop airplane. Within the last years the EU funded a series of research projects to investigate the various technological aspects dealing with the tilt rotor aircraft.

The TILTAERO project (Tiltrotor interactional Aerodynamics) aims at the development of a common European database capable to cover the main aerodynamic interaction phenome-na. This database is employed to validate the prediction tools in order to assess their capa-bilities in capturing the interaction pheno-mena and to focus the research activity on those areas revealing lack of knowledge.

The Institute of Aerodynamics and Flow Technology carried out a series of time accurate RANS computations using the structured code FLOWer to predict the propeller performance and the aerodynamic characteristics of the tiltable TILTAERO nacelle-wing system (see Figure 3.4-27). Research on tilt rotor configurations, Figure 3.4-28, is currently continued in the NICETRIP EC project.

Figure 3.4-27 Computed surface pressure in low speed conversion flight.

Figure 3.4-28 NICETRIP configuration.

3.4.5 Low Noise Flight Procedures

Like for no other air vehicle the helicopter’s noise generation and radiation is extremely sensitive to changes in the flight attitude. A direct conclusion of this finding is to investigate low noise flight procedures for helicopters. Since basically no technical changes are needed for this noise reduction concept, well defined and flyable low noise procedures -once identified- may be applied immediately to modern but also to older, potentially louder helicopters to achieve noise reduction. The question to be answered was, how to arrive at complete low noise landing procedures and, even further, at optimized, i.e. minimum noise procedures. This topic was addressed in DLR projects PAVE and PAVE II as well as the EU project FRIENDCOPTER.

125

3.4.5.1 Design of Low Noise Flight Procedures

Aeroacoustic flight tests were performed in 2004 on the two highly instrumented DLR helicopters, the BO-105 and the EC135-FHS, in order to generate a noise data base for noise abatement flight procedure design and optimization. This campaign was performed in the framework of the PAVE project (Pilot Assistant in the Vicinity of helipads). The instrumentation of the helicopters covered all the noise relevant flight parameters including free stream velocity probes at the tip of nose booms. The helicopter position along time was measured with Differential GPS. The noise was measured on ground using 43 microphones operated using DLR’s wireless large area microphone system distributed on an 800 m diameter disk, see Figure 3.4-29.

Figure 3.4-29 Microphone layout for the 2004 PAVE campaign.

Instantaneous noise directivities from maneuver flights could as well be measured as steady flights directivities. The tests comprised steady flights at various velocities and aerodynamic slopes, up to 20 deg in tailwind descent flight. A series of maneuver flights was also measured: turns, turn roll-in, turn roll-out, begin and end of descent, etc. About 350 flights were performed including some entire landing and take-off procedures.

Flyover Noise Prediction with HEMISPHERE

A simulation code was developed for computing flyover ground noise for helicopters. This code HEMISPHERE2 is on the one hand able to extract source information

from measured flyover data. On the other hand HEMISPHERE2 is used to fly the helicopter as a numerically described sound source along arbitrary paths and to simulate the resulting sound field. For the source characterization typically a measured aero-acoustic database is post-processed by constructing spheres characterizing the noise directivity at a given emission instant, see Figure 3.4-30. This is done by back-propagating the ground microphones signals onto a fictive one-meter radius sphere centered at the rotor hub. HEMISPHERE2 takes into account outdoor propagation laws such as ground reflection depending on the type of substrate, geometrical spreading, atmospheric attenuation and Doppler Effect. The aerodynamic parameters relevant for the noise generation are also post-processed from the on-board data with the HOST tool and stored together with each acoustic emission sphere. DLR possesses now more than 4000 spheres characterizing the EC-135 FHS helicopter, and the same methodology is actually being applied to the BO-105.

Figure 3.4-30 Example of a constructed sphere.

Making the assumption that the helicopter emission can be characterized by a defined set of aerodynamic parameters (main rotor thrust coefficient, advance ratio and tip path plane angle) called noise relevant parameters, one can then use the spheres library to estimate the helicopter noise emission at each instant of a discretized arbitrary trajectory including maneuvers, and thus the corresponding noise footprint. A set of control points with prescribed velocities is used to generate any trajectory using a Bezier splines module. HOST

126

provides then the flight mechanics parameters including the noise relevant aerodynamic parameters. Avoidance of vortex ring state, autorotation and compliancy with the height-velocity diagram can be checked. Accelerations, pitch and slope angles can also be limited on demand for pilots comfort. Once the noise relevant flight parameters are known, a Delaunay interpolation is run to generate a simulated sphere from the closest ones in the library at each time step. The simulated spheres are then propagated on a defined ground microphone array to obtain levels, spectra and sound exposure level SEL.

BVI-Free Complete Landing of EC135

The analysis of the PAVE 2004 complete flight test data versus velocity, in combination with a basic helicopter flight dynamics model enabled a step by step flight procedure design. First, a procedure was designed manually by prescribing the time derivatives of the acceleration components for each time step. This procedure consists of a slight horizontal deceleration (vortices below main rotor), a descent close to autorotation (vortices above the main rotor) and a flare close to autorotation. This procedure was tested in flight in mid 2007 and was observed to be a BVI-free complete landing procedure.

Low Noise Flight Indicator to the Pilot

Considerations of how pilots act and how the available instruments in helicopter cockpits could help for flying quietly, led to considering the engine torque display in particular. BVI can be avoided either by the convection of the vortices underneath the rotor plane, which is performed with a large collective pitch, or by their convection above the rotor plane, using a small collective pitch value. As the collective pitch and the required engine torque are strongly linked, the engine torque was expected to be a BVI governing parameter. Indeed, the analysis of the PAVE 2004 flight tests showed a strong correlation between torque and noise, and in first approximation, the BVI noise occurrence torque range is independent of the velocity.

Therefore the torque display can judiciously be extended into a BVI noise indicator as shown in Figure 3.4-31. As the engine torque value and the collective stick position are

strongly linked, it is relatively easy and fast to control torque and to avoid BVI occurrence.

Figure 3.4-31 Coupled noise indicator and torque display (patent pending).

The loud area shown at very low torque corresponds to Fenestron (working in reversed flow) noise. The coupled torque-noise indicator not only tells the pilot if the helicopter is loud or not, but it helps the pilot to recover quickly and easily quiet flight conditions, as well in maneuvering flight (with small roll and pitch rates) as in steady flight.

Minimum Noise Procedures

Once the computation chain consisting of HEMISPHERE and HOST along with its associated library is available, it becomes possible to let a genetic algorithm optimizer adjust the splines control points to obtain the most silent trajectory possible for a given helicopter. The methodology is currently implemented and allows finding minimum noise flight procedures taking into account diverse parameters, like variations of wind in force and direction, or helicopter mass.

A reference landing procedure including a 6 degrees descent angle was also simulated (characterization of procedures, see top of Figure 3.4-32). By comparing the noise level contours of reference and optimized procedure (lower part of Figure 3.4-32) it is possible to quantify the theoretical gains in terms of dB SEL. For given ground positions a reduction of more than 10 dB SEL is observed. First data of minimum noise flight tests in mid 2008 confirmed these significant reductions in flyover noise.

127

Figure 3.4-32 Top: Reference (dashed) and optimized (solid) flight procedures, below: SEL of EC-135 FHS reference trajectory (center) and optimized trajectory (bottom).

3.4.6 Outlook

As a result of the intensive research work at DLR, first simulations for complete helicopter configurations have been presented during the last three years. However, many more simulations have to be performed in order to validate the methods for all flight conditions. A good validation base will be available at the end of the GOAHEAD project in 2009. The high requirements regarding computational hardware and the long run times of complete helicopter simulations prevent the use of CFD methods in helicopter industry today. Future

research will therefore be focused on reducing the computational times while increasing prediction accuracy. In addition, process chains for the design of helicopters will be set up.

Along with the increased capabilities of CFD prediction, research into the low drag design of helicopter hubs and fuselage has become feasible for the future. Multipoint optimum blade and planform design as well as the assessment of new rotors at the complete helicopter will therefore be addressed.

Based on a multidisciplinary optimization code and the acquired complete aeroacoustic data bases for BO 105 and EC135, exclusive to DLR, low noise flight procedures were successfully designed. Since currently this must rely on large and expensive measured flyover data bases of existing helicopters, the ground noise prediction for complete missions need to be based on a fully numerical description. This capability in turn will enable the virtual evaluation of the acoustic implications of novel rotorcraft technologies. Research efforts also need to be invested in studies on the integration of low noise procedures into air traffic systems. Experience shows that the optimum noise reduction of the main rotor is limited by the occurrence of the fenestron as dominant noise source. Flow control is seen as a useful technology to reduce fenestron noise significantly independent from the main rotor aerodynamics. Therefore appropriate flow control technologies must be developed and investigated in CFD simulations and aeroacoustic wind tunnel tests. The planned modification of the NWB in Braunschweig into a premier aeroacoustic wind tunnel as part of the aeroacoustic test center Braunschweig will enable such tests on models with relevant size.

128

3.5 Military Aircraft and Missiles

3.5.1 Introduction

Aerodynamic research for military aircraft and missiles has to satisfy specific requirements, often different from those for civil applications. One driving requirement for modern high performance fighter aircraft and missiles is high agility in the transonic velocity regime. Vortex-dominated flows over low-aspect wings or slender fuselages at high angles of attack are characteristic for these configurations. Separated flow and high maneuverability imply significant unsteady effects, representing severe challenges for both experimental and numerical methods. For missiles at high supersonic speed, structural and thermal loads on exposed components become limiting factors. Stealth requirements for future systems have a large impact on the aerodynamic shape, posing new challenges for stability and control. The institute utilizes its knowledge as well as the numerical and experimental tools developed for civil applications to address these problems. The experience gained from the further development and adaptation of numerical codes and experimental facilities again feeds back into the civil domain utilizing and generating synergy effects.

All current activities of the institute in these areas are bundled, together with contributions from other DLR institutes, in common DLR-projects. Their content is harmonized with requirements defined by the corresponding authorities at BWB/BMVg and industry and results are presented regularly at common meetings and workshops.

3.5.2 Objectives

General objective for military research work in the institute is to ensure the availability of knowledge and competence to provide an assessment and advisory function for the government agencies BWB/BMVg. This includes the operation and continuous development of the relevant experimental facilities.

The specific objective of the SikMa-project (“Simulation komplexer Manöver”) was to

establish an interactive simulation environment, based on the DLR-TAU code, for the simulation of a free-flying, fully configured, elastic fighter aircraft. For this purpose, the numerical aerodynamics, flight mechanics and aeroelasticity tools had to be coupled and solved in parallel. For the validation, a number of wind-tunnel experiments have been designed; among them highly advanced maneuver simulation experiments with an X-31 model with a complete set of dynamically remote-controlled flaps on the novel test rig with six degrees of freedom in the low-speed wind tunnel DNW-NWB.

As logical continuation, the current UCAV-2010 project aims at the development of methods and procedures to investigate and assess aerodynamic technologies for Unmanned Combat Aerial Vehicles (UCAV). Design tools adapted from civil projects and extended for UCAV-specific requirements are used to generate prototypes, which serve as test-bed for new control methods, new engine inlets, etc.

The main objective for the investigation of military transport aircraft is to ensure and increase mission success and safety when deploying personnel and material from the cargo ramp. This is to be achieved by better understanding and simulation of the flow field on the rear fuselage and the trajectory of deployed loads through the near field and wake of the aircraft. Time-accurate coupling of the CFD code with a 6 DOF flight mechanics module yields the prediction of the trajectories. Wind tunnel experiments using advanced optical imaging techniques are performed for validation.

For the missile-related activities, the objectives were further development of technologies and design methods for key components of advanced highly agile missiles. Special emphasis was put on new concepts for drag reduction at supersonic cruise Mach numbers, advanced control methods for enhanced maneuverability, the development of numerical capabilities and experimental facilities for maneuver and separation

129

simulation, design and qualification of inlets, aerothermodynamic loads on domes and thermal protection systems. Representative target configurations served as reference and test bed to assess the performance of the newly developed methods and components.

3.5.3 Maneuver Simulation of Fighter Aircraft

Within the DLR project SikMa a numerical tool has been developed and validated to simulate the unsteady aerodynamics of a free flying aeroelastic combat aircraft, by use of coupled aerodynamic, flight mechanics and aeroelastic computations. To achieve this, the unstruc-tured time-accurate flow-solver TAU is coupled with a computational module solving the flight mechanics equations of motion and a structural mechanics code determining the structural deformations. The numerical results are validated by experimental data from several specific wind tunnel experiments with different wind tunnel models.

3.5.3.1 Numerical Simulation of Complex Maneuvers

For the verification and validation of the simulation environment, the results of the numerical simulations are compared against data collected from various experimental simulations. For this purpose a validation strategy is defined, considering the model configurations and the test and validation scenarios. One of the model configurations is the delta-wing-configuration described in the next section. The model is used to show the capability of the TAU-Code to predict the unsteady aerodynamic behavior of configurations with vortex dominated flow fields. The final configuration is the X-31 model described in a later section.

Verification of the Coupling Procedure

In Figure 3.5-1 a CFD – flight-mechanics coupled simulation of the delta-wing with trailing edge flaps is depicted. The initial

attitudes are != 17"!and #$!= 0". The

trailing-edge flaps are deflected by %!=!&5" and the model is left free to roll once it has been released. The computational grid has

approximately 7'106 grid points and the area close to the surface is resolved by 20 prismatic layers. The simulations for the delta wing as

well as for the X-31 configuration (discussed later) are done on 32bit PC-Cluster in parallel on 32 and 48 processors, respectively. The

physical time step size is (t = 0.001s and 600 inner pseudo time steps are performed within each physical time step. The time for the chimera search within each physical time step can be neglected in comparison to the computational time for each physical time step.

Figure 3.5-1 1 DoF Free-to-Roll maneuver of delta-wing-flap-configuration through trailing-

edge flap deflection. M = 0.5, Re = 3.87'106,

! )!*+",!!#0!)!0",!%!)!&!5"-

The turbulence model used for this simulation

is the Wilcox k-.! turbulence model with the Kok-TNT-rotational correction approach, which was chosen from experience of other investigations simulating the flow around delta wings with sharp leading edges. Two calculations are done using this configuration. The first calculation is done without taking into account the effects of mechanical friction, while for the second calculation a friction moment of 3.5 Nm was assumed. (The exact value in the experiment could only be estimated to be within 2 Nm and 4 Nm.) The characteristic movement of the model, as well as the rolling moment, is well predicted by the second calculation. The analysis of the rolling moment shows:

- An asymmetric surface force distribution develops due to the asymmetric trailing-edge flap deflection, which in turn leads to a rotational acceleration around the longitudinal axis of the model.

- The maximum rolling moment is reached after a simulation time of 0.05s, where

the flaps are at %!)!2.5"!deflection. After this the rolling moment decreases and

130

reaches a temporary plateau at t = 0.1s, at which time the flaps are fully deflected

at %!)!/".

- The model reaches a trim-point at

#!)!0*", where the combined rolling moment is not large enough to overcome the aerodynamic damping due to the asymmetric loads on the wing together with the mechanical friction of the system.

The reason for the movement of the model is graphically explained in Figure 3.5-2. At the start of the simulation the wing is accelerated due to the asymmetric flap deflection, stage . The vortex on the luff side of the wing is strengthened with the increasing roll angle. The effective sweep angle on the luff side of the wing is decreasing, which in turn increases the normal component of the on-flow vector. This causes a stronger primary vortex on the luff side, which is located closer to the surface, thus leading to a higher local lift. On the lee side the opposite effect happens. The wing vortex gets weaker and the distance from the wing surface higher as the roll angle increases, which leads to a lower local lift on the lee side, see stage !. This effect causes the wing to decelerate, which in turn leads to the trim-point at

#!)!0*", stage " and stage #.

$

%

&

'

$

%

&

'

Figure 3.5-2 1 DoF Free-to-Roll maneuver of delta-wing-flap-conf. through trailing-edge flap deflection. Flow topology at four different stages.

M = 0.5, Re = 3.87'106, 1)*+",!#0)0",! %!)!&!5"-!

This example shows the capability of CFD – flight-mechanics coupling for the case of a delta-wing with trailing edge flaps. The main aerodynamic effects are qualitatively well predicted, but for a good quantitative agree-

ment with experimental results it is also necessary to take into account the exact starting conditions and all other parameters relevant for the maneuver scenario.

Validation for the X-31 Configuration

Having demonstrated the principal capabilities of the coupling procedures by means of the simple delta-wing-model, the next step is to determine the capability to accurately capture the steady and unsteady aerodynamics of the more complex X-31 configuration by comparing the numerical results with experimental data.

In Figure 3.5-3 the hybrid mesh topology is depicted. A pre-refined mesh is used for both steady and unsteady simulations. To cover a certain range of pitching and yawing angle the pre-refinement is adapted to the areas where the presence of vortices is assumed during the simulation. The steady state results are simulated with no sting, while the unsteady simulations are simulated with the belly sting support taken into account, as was the case in the experiments. The effect of the support is discussed later on.

Figure 3.5-3 Slices through pre-refined un-structured hybrid mesh of the X-31 configuration.

Figure 3.5-4 shows the numerically simulated 3D flow field over the X-31 configuration, and gives a good indication of the complexity of the vortex flow topology over the wing and fuselage. All control devices are taken into account. This means that all gaps between the trailing- and leading-edge flaps and the wing are represented by the computational model. In these simulations a computational

grid with approximately 15'106 grid points is used. The boundary layer region is resolved by 20 prismatic layers. For the X-31 simulations the one-equation Spalart-Allmaras turbulence model is used, which has been shown also by

131

other researchers to give good results on comparable configurations with blunt leading edges.

Figure 3.5-4 3D flow field over the X-31 configu-

ration at =18º.

In Figure 3.5-5 the lift and pitching moment coefficient versus the angle of attack is shown. For all calculations at 10°, 12°, 14° and 16° angle of attack a higher nose-up pitching moment is predicted by CFD, while the overall lift is fairly well predicted. This can be linked to an under prediction of the suction peak on the canard for these conditions. The overall pressure distribution on the wing is well predicted by CFD, only the formation of the vortex at the outer leading edge flap occurs at a somewhat higher angle of attack than in the experiment.

Figure 3.5-5 Lift and pitching moment coefficient over angle of attack. Comparison of TAU calcula-tions with experimental data.

Figure 3.5-6 shows the result of a guided yawing motion maneuver. The initial pitching

angle is 10!"and the yaw amplitude is #$=5!. For the unsteady simulations a physical time

step of #t = 0.02 is used with 1000 inner pseudo time steps. The maneuver simulations are resolved with 50 time steps per period.

Figure 3.5-6 Guided yawing motion of the X-31

configuration at = 10º, #$ = 4º, f = 1 Hz. Comparison of TAU calculations with experimental data.

The maneuver is done with a frequency of 1Hz. In this scenario the model is mounted on the belly sting support as described above. The overall aerodynamic behavior is captured by the numerical simulation, although the pitching moment is under-predicted. The gradient of the rolling moment is predicted to be higher than in the experiment and the hysteresis shown in the experiment is predicted to be lower by the calculation.

3.5.3.2 Experimental Simulation of Complex Maneuvers

For the validation of the numerical simulation, a new experimental capability for simulating the unsteady aerodynamic behavior of a fully configured fighter configuration has been developed, tested and installed. As target configuration the X-31 aircraft with movable control devices has been selected.

Two versions of the X-31 model have been built to a scale of 1:7.25. The so-called Lightweight (LW) model was designed for minimum weight and is fully made of carbon fiber reinforced plastic (CFRP) except for the leading edge flaps and strakes that are made of aluminum. All control surfaces can be set manually prior to a test.

The wing and the lower part of the fuselage of the second remote control (RC) model

132

were made of steel, most control surfaces of aluminum and the upper part of the fuselage and the rudder of CFRP.

The RC-model is equipped with eight servo motors for moving four leading edge flaps, the trailing edge flaps, canard and rudder via a remote control system during a test. The location of the motors and balance is depicted in Figure 3.5-7.

Figure 3.5-7 Exploded view of the X-31 remote control model mounted on a belly sting with connecting rod.

The mass of the RC-model equipped with motors was about 110kg, while the mass of the lightweight model could be kept just below 10kg.

Both models have been instrumented with miniature pressure sensors. Time signals with a bandwidth of at least 1 kHz were simultaneously obtained at up to 50 taps that are arranged along two sections in spanwise direction on both wings. Integral quantities, like forces and moments, were determined by an internal six component strain gauge balance. The models were mounted with a belly sting to the newly installed novel test rig “Model Positioning Mechanism” (MPM) in the DNW-NWB. This device, which was used for the first time for these experiments, provides six degrees of freedom (DOF) to move the model relatively to an arbitrary center of reference. Figure 3.5-8 shows the setup in the wind tunnel. A sophisticated data control and acquisition system has been developed and installed to ensure the synchronization of the motion of the model on the MPM with the corresponding actuation of its control surfaces and the data acquisition from the different sensors.

Figure 3.5-8 The X-31 remote-control model on the MPM support in the subsonic wind tunnel DNW-NWB. Note the three markers on the model used by the optical position tracking system.

First, as a basic test, pitching motions of high amplitude were considered. In order to

achieve amplitudes up to # = 15! an additional pitch module driven by a connecting rod was installed in the model. (cf. Figure 3.5-7). In all tests with large amplitude pitching oscillations the light-weight version of the X-31 model was used.

In Figure 3.5-9, representative pressure signals plotted versus the angle of attack from an

oscillation about o = 15! with an amplitude

of # = 15! are shown as an example. In this case, flows without vortices, with vortices and with vortex breakdown are encountered during one oscillation cycle. Both quasi-steady and unsteady cases were considered. The quasi-steady motion was realized by an oscillation at a small reduced frequency of k = 0.005 (f = 0.1Hz). In the unsteady cases, reduced frequencies up to k = 0.163 (f = 3 Hz) were applied. The pressure coefficient has been determined from unfiltered pressure data obtained at a pressure tap located at the outer part (spanwise direction) on the upper surface of the left wing. At this position the formation and breakdown of vortices can be well detected. In both cases the onset of vortex formation is observed at about the same

angle at 11!. However, for the dynamic case, the dependence of the pressure coefficient on the angle of attack develops a clearly defined hysteresis. In the quasi-steady case a decrease in lift caused by vortex

connecting rod

motors

balance

133

breakdown is well recognizable to begin and

stop at about 22!. In the unsteady case the dynamics cause the well-known opening of the curve. Here, the larger of the two pressure coefficients corresponds always with a decreasing angle of attack.

[ °]

Cp

0 5 10 15 20 25 30

-2

-1.5

-1

-0.5

Cp, dyn

Cp, st

Figure 3.5-9 Quasi-steady and dynamic pitching

oscillations of large amplitude ( o = 15! # = 15!, blue curve: k = 0.005, red curve: k = 0.163): pressure coefficient vs. angle of attack.

Figure 3.5-10 shows the results of the simulation of a steady-heading sideslip maneuver (SHSS) in the wind tunnel. The time histories of the motion of the model and its control surfaces were based on real flight tests. Since the original flight data are noisy and the geometry of the experimental X-31 aircraft is not necessarily as symmetric as that of the models, the data collected during flight tests have been smoothed, adjusted and artificially symmetrized by hand before being used for defining wind tunnel maneuvers. In Figure 3.5-10a well smoothed data of several angles of the aircraft (yaw) and its control surfaces are plotted versus time. Note that the time scale has already been adjusted according to the wind tunnel velocity and the scale of the model.

Slightly random oscillations of leading edge flaps and small changes in the angle of attack present in the real flight maneuvers are not significant. Therefore the corresponding angles have been held at constant values and the canard was permanently aligned into the flow direction. In the SHSS-maneuver, the rudder performs large motions and the

trailing-edge flaps are acting as ailerons opposing the rolling and yawing moment resulting from the motion of the rudder.

t [s]$%&

tel,&

ter,&

rud

[°]

0 5 10 15

-5

0

5$&

tel&

ter

&rud

(a)

t [s]

Cy,

Cl,

Cn

0 5 10 15-0.1

-0.05

0

0.05

Cy

Cl

Cn

Cl,0

Cn,0

(b) Figure 3.5-10 Simulation of a maneuver corresponding with steady-heading sideslip test points (a) top: given variation in time of the angles of

yaw ($), right and left trailing edge flap ( ter and tel, respectively), and rudder ( rud) (b) bottom: resulting variation in time of the coefficients of lateral force (Cy) and rolling and yawing moment (Cl and Cn).

Experimental results displayed in Figure 3.5-10b show that the coefficients of the lateral force and the rolling and yawing moment remain quite small during the whole maneuver. This shows that the manual

134

adjustments to the flight data of the maneuver have been successful. For comparison, for the same maneuver also coefficients of rolling and yawing moments obtained without flap motions (Cl,0 and Cn,0 all flaps in a fixed zero position) are included in Figure 3.5-10b.

3.5.4 Aerodynamics Technologies for Future Unmanned Combat Aerial Vehicles

In the future, it is foreseen that the role of manned combat aircraft will be taken over more and more by unmanned systems. Stealth requirements have a severe impact on the aerodynamic shape and require new solutions for maneuver control, inlets, etc. The DLR Project UCAV-2010 was set up to identify and assess UCAV relevant technologies. The investigations cover the pre-design process with fast, low fidelity methods as well as the detailed examination using high fidelity methods. The verification of the identified technologies and tools will be done by virtual and experimental prototypes.

3.5.4.1 The DLR UCAV-2010 Project - Overview

The following Figure 3.5-11 of the project structure shows the work packages within UCAV-2010, Figure 3.5-12 shows the overall workflow strategy of the project. The target configurations are characterized by delta wing planforms. One major target configuration will be presented in chapter 3.5.4.3.

Figure 3.5-11 UCAV-2010 work packages.

The contribution of the institute to the project consists in experimental and numerical investigations on the research of the flow physics of Delta and Lambda wing configurations with a vortex-dominated flow

field. Furthermore, the existing pre-design capabilities and tools will be extended for UCAV purposes which will be described in chapter 3.5.4.2. Experimental investigations will be performed related to control devices without deflected flaps, flow control, and integrated air intakes. In order to provide experimental data for numerical code validation, two different wind tunnel models will be available, one for the transonic regime for tests in the DNW-TWG and one for low speed tests in the DNW-NWB. The numerical investigations are done with the DLR TAU-Code. Objectives are the validation of the TAU-Code for Delta- and Lambda-wing configurations. A first result is shown in Figure 3.5-13.

Figure 3.5-12 Work flow strategy for the UCAV-2010 project.

Figure 3.5-13 Pressure distribution on the surface of the DLR-F17 UCAV configuration and the vortex topology at the wing apex at an AoA of 15°.

One major part of the project is the contribution within the RTO/AVT Task-Group AVT-161-“Assessment of Stability and Control Prediction Methods for NATO Air and Sea Vehicles”. The UCAV design data delivered by EADS-MAS was prepared to get consistent CAD data for the manufacturing of the wind tunnel model by NASA LaRC. A wide range of CFD calculations for load estimation of the model has been done prior to the start of the model design. Finally the institute will deliver

135

wind tunnel experiments with the UCAV model in the DNW-NWB as well as wind tunnel investigations and results using the second target configuration, the X-31. For both wind tunnel tests the focus lies on the dynamic aerodynamic behavior of the aircraft, i.e. the prediction of the dynamic derivatives and aerodynamic data for maneuver flight.

3.5.4.2 Design Methods for UCAV

As a part of DLR’s UCAV-2010 project, which has been briefly described in the previous section, a preliminary design and analysis system for UCAVs is currently being developed. Based on the tools and concepts of the TIVA project, an extended toolbox of analysis methods and process chains for UCAV design and analysis will be created. The scheme of the selected tool integration concept is shown in Figure 3.5-14. It consists of a central, XML-based database and an integration framework for connecting distributed analysis tools via network.

Figure 3.5-14 Scheme of the TIVA tool integration concept.

As UCAV-specific extensions to the TIVA-System, the institute provides tools for a fast and rough aerodynamic analysis of nearly arbitrary configurations, as well as methods to consider the aerodynamic effects of new technologies like flap- and rudderless steering.

Figure 3.5-15 shows a comparison of the calculated force and moment coefficients (under low-speed conditions) for an earlier stage of the DLR-F17 UCAV model. As can be seen, the used lifting-line model already fits quite well for the linear part of the TAU curves. Using a 2½-D approach to incorporate

2D airfoil polars into the lifting-line results allows for a prediction of the aerodynamic behavior when coming close to the maximum lift coefficient. The results of this approach, called POLINT, are displayed in the figure as a third set of curves. Due to highly three-dimensional flow conditions for this UCAV configuration (especially at higher angles of attack), the absolute pitching moment coefficients deviate significantly from the TAU results.

Angle of attack

Lift

forc

eco

eff

icie

nt

cL

Pitch

ing

mo

me

nt

co

eff

icie

nt

cM

y

0 5 10 15 20 250

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

-0.25

-0.2

-0.15

-0.1

-0.05

0

cL

LIFTING_LINE

cL

POLINT

cL

DLR-Tau

cMy

LIFTING_LINE

cMy

POLINT

cMy

DLR-Tau

Coefficients for lift force and pitching moment

C.o.G. location for15 % stability margin

Figure 3.5-15 Comparison of DLR-F17 UCAV (earlier stage) force and moment coefficients, coming from preliminary design (LIFTING_LINE and POLINT tools), as well as from high-fidelity methods (DLR-TAU code).

A second advantage of the described approach is the incorporation of airfoil drag into the result (where the lifting-line method only predicts induced drag). This leads to a quite good agreement with the TAU results up to an angle of attack of 10°, as can be seen in Figure 3.5-16. At higher incidence, this configuration produces very strong 3D separation effects which are beyond the scope of such preliminary design methods. Further contributions to the preliminary design tool are planned to cover buried engine intakes and a tool for semi-automatic mission based conceptual design.

3.5.4.3 Design of Generic UCAV Configuration

The design of the generic UCAV-configuration is driven by the requirements of upcoming missions for UCAV configurations, as for instance the capabilities of long endurance flights joined with a low observability.

136

Furthermore the capability of medium to high AoA maneuverability will be considered. This leads to Lambda wing configuration with a medium sweep of the leading edge. The DLR-F17 configuration is derived from the so called “Saccon” configuration developed by EADS-MAS for the RTO/AVT-161 task group. The geometry of the UCAV configuration is depicted in Figure 3.5-17 and above in Figure 3.5-13.

Drag force coefficient cD

Lift

forc

eco

eff

icie

nt

cL

0 0.1 0.2 0.3 0.40

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

LIFTING_LINE (only induced drag)

POLINT

DLR-Tau

Drag-polar

Figure 3.5-16 Comparison of DLR-F17 UCAV (earlier stage) drag polar coming from preliminary design (LIFTING_LINE and POLINT tools), as well as from high-fidelity methods (DLR-TAU code).

53°

1060.8mm

479mm

76

9m

m

53°

1060.8mm

479mm

76

9m

m

Figure 3.5-17 Planform of the Saccon and DLR-F17 configuration. Scaled for the DNW-NWB wind tunnel model.

The wing has a leading edge sweep of 53° and the leading edge has a sharp contour at the apex and is getting round just at the first wing stage. For low radar signature, the leading edge is parallel to the wing trailing edge and the outer wing section is parallel to the fuselage trailing edge.

Figure 3.5-18 shows the vortex topology over the wing at an angle of attack of 15° and a bank angle of 5°. The vortices start to occur at an angle of attack of 10° at the apex and the outer wing tip. First numerical investigations using the DLR-TAU code show this behavior. The outer wing tip vortex is moving upstream by enhancing the angle of attack. Further investigations show a strong sensitivity of the vortex topology when changing the Mach and Reynolds number.

= 15°

$ = 5°

= 15°

$ = 5°

Figure 3.5-18 Flow topology over the UCAV configuration. Total pressure distribution slices showing the vortex location over the wing.

For the determination of the aerodynamic behavior, dynamic wind tunnel tests will be done on the MPM system in the DNW-NWB, see Figure 3.5-19. The objective is the determination of the dynamic derivatives and the flow physics on the wing surface and in the field. Further investigations with the generic UCAV configuration will be done with a wind tunnel model for the DNW-TWG. Steady state and dynamic tests will be performed to extend the data base for higher Mach and Reynolds numbers in comparison to the low speed investigations in the DNW-NWB.

Figure 3.5-19 UCAV configuration on the MPM system in the DNW-NWB for the purpose of dynamic derivative measurements.

137

3.5.5 Military Transport Aircraft

Different questions associated with modern Military Transport Aircraft were investigated by several institutes of DLR since the 1990s within the framework of military technology research. Within the project MiTraPor (Military Transport Aircraft) all those activities have been integrated with the aim to achieve synergy effects and to look at the transport aircraft as a whole. To ensure safe and precise deployment of troops and material from the cargo ramp, a deep understanding of the flow physics on the rear fuselage and for the deployment process itself is necessary for the corresponding flight conditions and geometries.

3.5.5.1 Near Field Investigations on the Rear Fuselage

The flow in the aft section of a military transport aircraft is dominated by a complex system of vortices caused by the inclined underpart of the rear fuselage and the landing gear fairings, the so-called sponsons. With the tailgate down, regions with massively separated flow occur additionally, thus further increasing the complexity of these flow phenomena.

To assess the flow in the wake of military transport aircraft with open cargo ramp, CFD simulations of a generic Future Military Transport Aircraft (FMTA) were performed with the DLR’s hybrid flow solver TAU. These simulations were accompanied by wind tunnel tests at the DNW-NWB to obtain validation data to be compared with the simulation results. The wind tunnel model of the FMTA was equipped with 64 pressure sensors distributed on four cut planes at the rear fuselage and the open cargo ramp. In addition, in five rectangular areas in the wake of the cargo ramp and of the fuselage, PIV measurements of high accuracy were taken to obtain the three dimensional velocity fields in these areas.

As a first result of this work, the flow solver could be validated for the flow cases investigated with the wind tunnel model, since the evaluation of the results led to very good agreement between numerical and experimental data (see Figure 3.5-20 and Figure 3.5-21).

Figure 3.5-20 Velocity field in the wake of the cargo ramp, PIV measurements and TAU simulation.

Figure 3.5-21 Pressure distribution of two cut planes of the rear fuselage, experimental data (green symbols) and simulation results (solid line).

3.5.5.2 Precision Air Drop Investigations

Safe and pinpoint dropping of airborne troops and single military loads up to 20 tons as well as relief supplies has to be ensured under various conditions. For cargo drops this can range from high-altitude drops at high Mach- and Reynolds-numbers to low-speed low-altitude drops at full flaps with landing gears extended.

In contrast to the later stages of the drop trajectory, which are mostly governed by parachute characteristics and wind profile, especially the earlier stages with the loads still being near to the cargo ramp are crucial for the safety of the transport aircraft as well as for the safety of the deployed cargo. The focus of the investigations within MiTraPor therefore lies in a time-accurate, coupled simulation of aerodynamics and flight mechanics of the released cargo in the transport aircraft’s vortex-dominated flowfield near the open cargo bay. For the simulation of the trajectories, aerodynamic forces and

138

moments on the cargo are evaluated at every time step. Thereupon the new location, orientation and velocity of the cargo are calculated in a 6-DOF flight mechanics module and the aerodynamic analysis for the next time step will be done.

On the aerodynamic side, a method to evaluate the donor points for the cargo’s Chimera-grid (Figure 3.5-22(a), red) was developed, simultaneously blanking the points in the background mesh (blue) that need not to be calculated using hole-definition geometries (green) associated with the cargo. Preliminary simulations (Figure 3.5-22(b)) demonstrated the applicability for later high-fidelity multi-body simulations including (rigid) parachutes, high-lift systems and actuator disks.

(a)

(b) Figure 3.5-22 Coupled 6-DOF-Simulation of a cuboid behind the stretched wing-profile of a future military transport aircraft: (a) Chimera-grid; (b) distribution of x-component of geodesic velocity.

In order to validate the simulations, cargo-drop tests at DNW-NWB were conducted using cargos of different masses with stepwise increased complexity ranging from simple cuboids over cuboids with rigidly attached rigid parachute to cuboids with pin-jointed rigid parachute. Trajectories were

captured using a set of two cameras with a frequency of 300Hz and accuracy as good as 1% of the cargo’s length within the calibrated drop-zone (Figure 3.5-23(a) and (b)).

(a) (b) Figure 3.5-23 Measured trajectories: high-density store (a) and medium-density store (b) at 20m/s

and = 0!.

3.5.6 Missile Technologies

Aerodynamic and aerothermodynamic missile technologies are a well established research topic at the institute. During the last years, advanced CFD methods partially coupled with structural mechanics or flight mechanics modules have been more and more integrated into the research work. Missile-specific adaptations of the numerical tools as well as the continuous development of the relevant test facilities (TMK, TWG, VMK, H2K, HEG, RWG, and LBK) and advanced measurement and testing techniques remain important tasks.

From the large spectrum of actual needs, the institute focuses its research on a number of key technologies, identified in agreement with the BWB and BMVg to satisfy actual needs and embedded in the former DLR-project “Hochagiler Flugkörper” (HaFK) and the current project “Fortschrittliche Flugkörper Technologien” (FFT).

3.5.6.1 Configurational Aerodynamics

Design of Missile Reference Configuration

Within the DLR project FFT, three target configurations have been designed to serve as reference and demonstrator for the newly developed technologies. The design process of a missile configuration is a multidisciplinary task, which is quite complex because multiple disciplines like aerodynamics, structural

139

mechanics, propulsion, thermal protection, guidance and control and seeker technology are highly integrated leading to strong interactions. Because of the strong dependence between mission characteristics and design variables it is useful to design the reference configuration based on realistic assumptions rather than to investigate oversimplified generic models. The configurations shown in Figure 3.5-24, Figure 3.5-25, and Table 3.5-1 have been developed using engineering tools in simplified design loops. Each focuses on a different topic. Together they consider the current situation of missile development in Germany but may also be used with some adoptions as vehicles for long term research at DLR. The configurations LK1 and LK2 will be investigated in the wind tunnel TMK.

Figure 3.5-24 Guide Configuration LK1, wind tunnel model.

Figure 3.5-25 Guide Configuration LK2, wind tunnel model.

Separation Technology

The separation of missiles e.g. from a carrier aircraft is subject to aerodynamic interferences, which can significantly affect the trajectory of the missile and the launching platform. However, there is a lack of wind tunnel facilities with the ability to examine such separations. Therefore, an existing mechanism for the Trisonic Wind Tunnel Cologne (TMK) is reactivated and modernized

(Figure 3.5-26). The goal is to offer the ability to simulate the separation processes of arbitrary model configurations and to investigate these processes by force and pressure measurements as well as by optical methods.

LK1 LK2 LK3

Propulsion Rocket Ramjet Rocket

Mass [kg] 90 200 60

Length [m] 3 4 2

Diameter [m] 0.15 0.2 0.15

Mach number 2.5 4.0 5.0

Range [km] 20 100 20

Altitude [km] 0 to 25 0 to 25 0 to 5

Focus High agility, mechanical loads, drag reduction

Adjustable inlet, thrust control

Thermal loads, side jet control

Table 3.5-1 Overview of guide configuration.

Figure 3.5-26 Mechanism for simulation of separation processes.

The mechanism allows movements of four axes simultaneously, namely in vertical and horizontal directions for the upper stage and angle of attack of upper and lower stage. The next measurement position is always calculated based on the currently measured aerodynamic coefficients, applying the correct flight dynamic relations. Currently, the data

140

recording and the controller software is adapted in order to allow for a demonstration of the separation mechanism.

3.5.6.2 Flow Control for Drag Reduction

Highly agile missiles equipped with a radar or infrared seeker mostly have relatively blunt hemispherical nose shapes for optimal seeker functionality. This shape, however, leads to very high aerodynamic wave drag. Different flow control devices with the potential for reducing both drag and thermal loads are investigated.

Innovative Spike Technologies

The effectiveness of conventional aero-spike devices to protect the dome and to reduce wave drag is strongly dependent on the flight conditions, such as angle of attack, speed, and altitude. To overcome these drawbacks some innovative spike concepts were proposed and patented within the framework of the project FFT.

The first one, the pivoting-spike concept, is based on a self-aligning device, which uses the weathercock principle for passive alignment into the flow direction over wide ranges of angles of attack and yaw. Force and moment measurements in the DNW-RWG show that pivoting spikes yield much higher performance than conventional spikes. Not only drag is reduced substantially, but also the pitching moment and the lift-over-drag ratio is increased in a wide range of incidence-angles. This is demonstrated for the drag in Figure 3.5-27 in comparison with a conventional spike and the reference body without spike.

0.0

0.3

0.6

0.9

1.2

1.5

1.8

0 5 10 15 20 25 30 35

Angle of attack [°]

Fo

reb

od

y d

rag

co

eff

icie

nt

[-] without a spike

fixed spike-disc 1.0D

pivoting spike-disc 1.0D

M = 5

Re D = 1.5 x 106

Figure 3.5-27 Effect of fixed and pivoting spike-discs on the forebody drag at Mach 5.

Another novel device investigated is the permeable spike-disc, a spike with a permeable disk at the tip reducing the dependence of spike effectivity on flight speed and altitude. The disk provides a sufficient quantity of retarded flow required for a full-length flow separation on the spike. This was confirmed in proof-of-concept experiments conducted in the Ludwieg-Tube Facility DNW-RWG.

Local Energy Deposition

Heating the flow upstream of a blunt nose has an effect similar to a solid spike, which is why this method is also named hot-spike or air-spike. The most important mechanism of the underlying more complex physics is the decrease of total pressure and Mach number in the wake shear layer interacting with the bow shock due to the localized heating.

Since the most elegant way to realize a non-intrusive, steady energy deposition into the flow, namely a powerful continuous-wave laser, was not available at reasonable cost, the tests in the Ludwieg Tube Facility DNW-RWG were performed using a DC-arc discharge.

In Figure 3.5-28 the results of drag measurements on a hemisphere cylinder in the DNW-RWG are presented. The power effectiveness ratio is defined as ratio of the propulsive power reduction due to drag reduction to the electric power deposited into the flow. The impressive results show a maximal effectiveness of the energy deposition at arc-power levels shortly above 2.0 kW, where more than 13 times of the electric energy is saved in thrust.

0

4

8

12

16

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Arc power [kW]

Po

we

r e

ffe

cti

ve

ne

ss

ra

tio

[-]

Experiment

Hemisphere

Figure 3.5-28 Effect of a DC-arc discharge on the power effectiveness ratio for the hemisphere nosed model.

The effect of short-duration energy deposition was investigated using a Nd:YAG double-

141

pulse laser with a maximal output power of 420mJ / pulse at 10 Hz repetition rate. The time delay within the pulse-pair was varied from 100 s down to 0 s.

Single-pulse experiments at Mach 2 demonstrate the effect of the energy level and the distance of the discharge region from the nose of the model. The experimental results for energy levels in the range from 35mJ/pulse up to 666mJ/pulse show that some typical stages of the flow evolution are qualitatively independent from the energy level. This behavior is predicted very well also in the numerical calculations conducted with the DLR TAU-code. Figure 3.5-29 shows a sequence of density gradient images obtained numerically demonstrating the flow features at different stages during the hot-spot motion.

Figure 3.5-29 Evolution of the flow structure after single-pulse energy deposition at Mach 2 simula-ted numerically with the TAU-code.

Double-pulse experiments show the effects of hot-spot interaction at different time delays between two pulses of equivalent or different energy levels. Aero-optical or electro-magnetic effects, which were expected when focusing the second laser-pulse through the plasma cloud from the first pulse, have not been found so far in the experiment. Both, experimental and numerical investigations

show qualitatively a similar evolution of the hot spots. If this good agreement can be confirmed also by more detailed comparisons, the conclusion may be that aero-optical effects are not significant here and simplified models can be used to predict the sequence of interactions.

3.5.6.3 Design and Qualification of Intakes

The issue of design and qualification of RAMjet missile inlets includes the evaluation of the inlet performance, i.e. its pressure recovery, the captured air mass flow and the starting capability. Hence, different intake configurations featuring axially symmetric (Figure 3.5-30) and chin intakes with different entrance segments were designed and subsequently examined by comprehensive tests programs in the Trisonic Wind Tunnel (TMK) and the Vertical Wind Tunnel (VMK) at Cologne (Figure 3.5-31). Thereby the long test duration of these facilities allows the application of variable geometries to adapt the intake to different operation conditions.

Figure 3.5-30 Wind tunnel model of an axially symmetric front intake.

Figure 3.5-31 Axially symmetric front intake tested at Mach 3 in the VMK at Cologne.

142

Different kinds of intakes were analyzed at their starting Mach number and the design Mach number. Numerical and experimental studies were carried out with angle of attack

and yaw ranges of -30° !," 30°. High losses of pressure recovery and mass flow were measured, mainly caused by leeward vortices partly falling into the intake duct. By implementing a boundary layer bleed system, an improved pressure recovery and delayed buzzing could be achieved. Further analysis of the effects of leeward vortices and boundary layer separation included variations of the body length and intake orientation on the missile investigated at angles of attack up to 30° in the TMK. One common property of these kinds of intakes described above is that they have only one design Mach number. The present research is about the development of intakes with variable geometry, which are functioning over a wide range of (design) Mach numbers.

3.5.6.4 Maneuverability and Stability

Lattice Wings

Lattice wings as aerodynamic control elements show, especially in supersonic flows, some important advantages in comparison to conventional control surfaces. Small weight and volume as well as favorable aerodynamic effectiveness and yaw stability at high incidence angles make lattice wings very attractive for high-speed agile missiles. However, the lack of reliable tools for design and optimization of lattice wings still impedes application to realistic configurations.

Therefore, one target within the HaFK-project was the investigation of lattice control elements for the development of such tools. Wind tunnel experiments provided a database for different basic configurations of lattice wings in a broad range of flow conditions: Mach numbers from 0.3 up to 6.0, angles of attack from 0° to 90° and yaw angles up to 30°.

The results represent a comprehensive data base for force and moment coefficients, illustrating the body influence on drag as well as on the control effectiveness, cf. Figure 3.5-32. Based on this data, the superposition of coefficients for multiple fins was verified

and the half-empirical lattice wing theory was fitted and validated. The subsequently developed software tool “FastGRIDS” enables the design and the optimization of lattice wing configurations based on local flow parameters for a wide range of Mach numbers, Figure 3.5-33.

Figure 3.5-32 Influence of the body flow field on the tangential force coefficient of a single grid fin at Mach 5.3.

-1.00

-0.75

-0.50

-0.25

0.00

0.25

-20 0 20 40 60 80 100

!#$%&

cZ

a [

-]

Exp.TMK

Exp. RWG

FastGRIDS

Figure 3.5-33 Comparison of the experimental normal force coefficient vs. angle of attack for a basic grid fin configuration at Mach 3, obtained on two differently scaled test models in the TMK and RWG with results from “FastGRIDS”.

After successful implementation of this software module using the actuator disk interface of the Navier-Stokes solver TAU, a new fast numerical tool for aerodynamic design and optimization of realistic missile configurations with lattice wings was completed. Compared with a calculation of the fully resolved grid fin structure, this tool saves over 80% of the computing time. A subsequent validation with experimental data, cf. Figure 3.5-34, showed an uncertainty in the results of less than 10%.

To overcome the largest drawback of lattice wings, which is relatively high aerodynamic drag, an advanced lattice wing configuration with locally swept or meandered leading

143

edges for high speed range was proposed. Numerical flow simulations and wind-tunnel tests in the DNW-RWG were carried out at free-stream Mach numbers from 2 to 6 for angles of attack varied from 0 to 10°.

alpha

Cm

-5 0 5 10 15 20 25

CFD missile without wings

Exp missile without wings

CFD missile with wings

Exp missile with wings

Figure 3.5-34 Experimental and numerical pitching moment coefficients for missile configura-tion as a function of the angle of attack.

The results for the locally swept configurations show a reduction of the zero-lift total drag (up to 38%), as well as improvement of the lift-over-drag ratio (up to over 20%) over the unswept case. The beneficial effect of local sweep increases generally with free-stream Mach number, sweep angle, relative thickness of the blades, and the bluntness of their leading edges. It decreases with the incidence angle and for bigger relative tooth-sizes.

Side Jet Control on Generic Configurations

Side jets are a very effective control device especially at low dynamic pressures e.g. at high altitudes or during the starting phase, where conventional control surfaces become less effective. Also, the very short response time predestines this tool especially for highly agile missiles. But the side jet also produces a very complex interaction region in the vicinity of the jet nozzle, which is difficult to predict and strongly dependent not only on jet pressure ratio and stagnation pressure, but also on the jet configuration itself. This investigation was targeted on identifying a jet configuration, which provides a stable interaction over a wide range of angles of attack while maintaining a good control performance.

A variety of different jet configurations was investigated on a flat plate in the DNW-RWG, measuring and analyzing surface pressure, heat transfer, skin friction, and wall streamlines. For direct measurements of surface skin friction and heat transfer, two non-intrusive optical measurement techniques, the GISF (Global-Interferometry Skin-Friction technique) and QIRT (Qualitative Infra-Red Thermography) technique were further developed and applied. In addition, force measurements were carried out on a generic missile to gain the amplification factors, which describes the increase in side force (efficiency) due to the induced pressure distribution in the interaction regime.

The measurements were accompanied by numerical calculations conducted with the DLR TAU-code. Numerical and experimental data were used to confirm the results against each other. The general agreement is shown in Figure 3.5-35 for an example of the numerically simulated (left) and experimentally measured (right) heat flux of the single jet configuration on the flat plate.

Figure 3.5-35 Comparison of the heat flux data, numerically simulated (left) and experimentally measured by QIRT (right).

The experimental investigation of the generic missile resulted in a jet configuration which came closest to the requested target: four jets mounted side-by-side, see Figure 3.5-36.

Figure 3.5-36 Representation of the efficiency of different side jet control configurations, expressed as down-thrust.

144

It provides an amplification factor KF > 1 and the interaction remains very constant over a range of angles of attack from -10° to +10°. These findings were also confirmed by the numerical simulations.

Since on real missiles the side jet is provided by small cartridges of chemical propellant, another investigation focused on the different behavior of hot gas jets compared to cold gas, as in the previous experiments. A generic cone-cylinder-flare configuration was equipped with several pressure taps and tested at a free stream Mach number of 3 in the Vertical Test-Section facility (VMK), Figure 3.5-37. The results showed that for a given pressure ratio a seemingly larger amplification was observed for the hot gas jet, however, this effect might partially also be due to the presumably larger momentum of the hot gas jet.

Figure 3.5-37 Demonstration of the hot gas jet in the VMK.

To analyze the large-scale vortex structures of the hot jet, high-speed videos were recorded simultaneously with the pressure measurements. They show relatively periodical turbulent jet structures allowing manual evaluation of their spatial distribution. The calculated convection velocities Uc show values of the free-stream-velocity magnitude shortly behind the nozzle. Since the hot gas jet expands more intensely downstream, convection velocities of large-scale eddies finally exceed the free-stream-velocity level. In the far field, after the jet plume has been bent downstream, the convection angles rapidly decay towards the free-stream direction.

TAU numerical investigation of the flow field around a generic vehicle controlled by means

of a lateral jet in supersonic motion has been carried out using the DLR TAU-Code. Calculations were made for free stream Mach numbers between 2.8 < Ma < 17 for various parameters like angles of attack, jet pressure ratios, flight altitude and jet temperature. Several grids of varied density and structure, different turbulence models and the impact of jet chemistry have been investigated. Emphasis has been put on the turbulence model and generating mesh geometries with which correct results may be produced. The numerical outcomes were proved with test results from the Trisonic Wind Tunnel TMK of DLR and used to predict complex details of the jet/cross flow interaction. Particular interest was directed to illuminate high temperature effects in hypersonic flows, as for example excitation of internal degrees of freedom and energy relaxation effects, see Figure 3.5-38.

Figure 3.5-38 Calculated wall temperature distribution for perfect and real gas conditions.

Extensive parameter variations gave predictions on the efficiency of a side jet in super- and hypersonic flows at various atmospheric altitude conditions, concerning air pressure, air density and ambient temperature. Finally, by integrating the pressure around the missile surface, the efficiency of the side jet was characterized by the resulting thrust amplification factor, Figure 3.5-39.

Simulation of Maneuvers

The development of advanced highly agile missiles demands new and enhanced numerical and experimental tools for investigations of the complex, nonlinear and

145

unsteady flow phenomena during fast maneuvers. Numerical investigations are already well adequate for dynamic analysis of hypersonic vehicles, see Figure 3.5-40. The new multi-disciplinary approach, involving computational fluid dynamics and flight mechanics is based on the implementation of a 6-DOF motion model into TAU, realized within the DLR project IMPULSE. For the integration of moving control surfaces, the technique of overlaid grids (Chimera-Technique) is utilized. A novel extrapolation boundary condition on the corresponding hole boundaries was successfully investigated, after the standard technique has shown artificial shock reflections in hypersonic and supersonic flow. This method offers more precise dynamic analysis and thus more accurate prediction of the dynamic behavior of hypersonic vehicles.

Figure 3.5-39 Side-jet amplification factor vs. Mach number.

The first steps in experimental maneuver simulation were based on the technology developed earlier within the DLR-projects AeroSUM and SikMa for maneuvering fighter aircrafts. The aim of these demonstration tests in the transonic wind tunnel DNW-TWG was the simulation of free and guided roll motions of a generic missile equipped with grid fins to investigate unsteady forces and moments. For these tests a model of a generic missile with deflectable grid fins was developed, Figure 3.5-41. The model is allowed to rotate freely around the balance due to the grid-fin deflection. Force and moment were measured and Schlieren pictures taken for different combinations of incidence, roll, and fin angles. A clear dependence of the forces and

moments on the rolling rates was found. The forces in Figure 3.5-42 for the rolling missile consist of unsteady data from some hundred rotation cycles. The relatively good repeatability indicates their reliability.

Figure 3.5-40 Example of numerical predicted pressure loads on a generic missile.

Oktober 2004 Institut AS-HK 3

HaFKServomotor

Inkrementalgeber

6k-Piezo-Waage

Telemetrie

DLR-Gitterflügel

Figure 3.5-41 Sketch of the G1-model with the TWG-support for simulations of free and leading roll-motions in transonic flows.

A very remarkable result is that the phase shift of the measured forces due to the rotation is in opposite direction for the tangential and normal components see blue and red arrows in Figure 3.5-42. The second important finding is that the side-force, induced by asymmetric lee-side vortices (phantom-yaw), shows a clear reduction of the mean value for the rotating case, while the unsteady component increases distinctly.

Though providing valuable results, these investigations indicated some limitations of the technical concept. Therefore, a completely new wind-tunnel support for the maneuver simulation of realistic slender missiles in the DNW-TWG has been developed, Figure

146

3.5-43. The remarkable simulation range for fast pitching maneuvers includes single angle of attack sweeps from 0° to 45° and oscillating pitching motions up to 5Hz at ±22.5° and 10Hz at ±10° amplitude. The construction is planned to be finished in 2008, allowing much more realistic and appropriate conditions for the investigation of unsteady flow phenomena at high angles of attack during unsteady maneuvers than before.

-1.60

-1.20

-0.80

-0.40

0.00

0.40

-180 -135 -90 -45 0 45 90 135 180

Rollwinkel [°]

Tan

ge

nti

alk

raft

be

iwe

rt [

-]

Rotation 32 Hz

Stationär

! = 20° - 90°, 0°, 90° ... + Lage

- 45°, 45°… x Lage

-5.00

-4.00

-3.00

-2.00

-1.00

0.00

-180 -135 -90 -45 0 45 90 135 180

Rollwinkel [°]

No

rma

lkra

ftb

eiw

ert

[-]

Rotation 32 Hz

Stationär

! = 20° - 90°, 0°, 90° ... + Lage

- 45°, 45°… x Lage

Figure 3.5-42 Tangential and normal forces as function of the rolling angle with and without rotation.

Sting interference is a particular problem for wind tunnel testing with slender missiles at high angles of attack, since there is limited space available for a stand-off between model and support and the unsteady separated flow is sensitive to upstream sting interferences, cf. Figure 3.5-41. Therefore the flow field around the high angle of attack supports available in the DNW-TWG has been numerically characterized for Mach 0.5-1.6 and angles of attack between 0° and 55° using a generic representation of the model see Figure 3.5-44. Guidelines were created to position models on the TWG 15°, 25° and 45° rolling stings for minimum sting interference. The results were also used to optimize the design of the new maneuver simulator.

Figure 3.5-43 Side view of the new wind tunnel device for simulation of fast pitching motions in a wide angle-of-attack range.

Figure 3.5-44 Pressure coefficient contours for a TWG sting at 55° angle of attack (Mach 0.90).

To validate and examine the performance of the DLR-TAU code for missile specific applications at high angles of attack, the institute contributed to a test case of the GARTEUR AG42, chosen for comparison of different codes with existing wind tunnel data. The conditions at Mach 0.2 and angle

of attack ! = 45° provided all occurring difficulties like the previously discussed “phantom yaw”.

The quality of the TAU-code becomes visible in Figure 3.5-45 in the comparison of pressure distributions from different codes and procedures like RANS and DES, showing qualitatively a very good and partly also quantitatively a good agreement.

147

SA - DLR-TAU-codeSA - DLR-TAU-code

Figure 3.5-45 Pressure distribution coefficient in a lateral cut of the body at X/D = 1.67. Comparison of experimental data with different numerical results.

3.5.6.5 Aerothermodynamic Loads on Domes

During flight, missile domes protect internal infrared or radar sensor units against damages caused by heat, pressure loads, and possible environmental effects like rain, sand or ice erosion. At the same time, the radome structure has to allow microwave transmission at the lowest level of attenuation.

To qualify structures and new materials encountering these technological challenges and to support innovative developments in this key area, the aerothermodynamic research activities within the FFT project focuses on three particular items:

- prediction of external and internal thermal dome loads,

- evaluation of influences on these loads by flow conditions, geometries, material properties and internal structures,

- control of thermal balances in regard to given technical limitations.

To predict stagnation heating along the flight trajectory, a combined analytical/numerical tool has been developed, which provides the resulting temperature development on nose tips including radiation cooling and nonlinear material properties. Wind tunnel tests support the tool validation and allow extrapolation to flight conditions, see Figure 3.5-46. For detailed physical interpretations Figure 3.5-47 visualizes the effect of different angles of attack on windward and leeward surface

temperature distributions for a slender missile dome.

Figure 3.5-46 Stanton number distribution on a blunt nose at Mach 3 evaluated from IR data.

Figure 3.5-47 IR-measured temperature patterns on windward and leeward side of a dome.

One important design task is to keep the temperature level within the material limits and constraints defined by the electronic components integrated in the dome. Thermal interactions between the temperature on the dome surface and sensitive integrated elements are rather complex and have to be studied experimentally in the wind tunnel, preferably at real flow conditions.

For this purpose, the institute operates a set of adequate and unique facilities like the VMK wind tunnel, capable to simulate real flight conditions according to pressure and temperature levels up to Mach number 2.8 at ground conditions. As an example for test configurations Figure 3.5-48 shows a dummy of a radar sensor unit equipped with thermocouples. Such a thermal analysis of sensor components supports the development of innovative cooling concepts.

One possible option for the reduction of heat loads entering into a radome is the application of high emissivity surface coatings. Figure 3.5-49 proves a significant temperature reduction of more than 250K by coating a WHIPOX ceramic probe. The test was

148

performed at a high temperature level and over real flight durations in the LBK facility.

Figure 3.5-48 Dummy of a radar sensor unit equipped with thermocouples.

Figure 3.5-49 Verification of temperature reduc-tion by high emissivity coatings.

To duplicate the flow field of hypersonic vehicles in air from high altitudes down to sea level, i.e. to unit Reynolds numbers in excess of 100*106, a new Mach 6 nozzle with a core flow diameter of 300 mm was designed and constructed for the HEG facility. So far a condition with a nozzle reservoir temperature of 1600 K and a reservoir pressure of 39 MPa was established, representing Mach 6 flight in 11 km altitude with a unit Reynolds number of 42*106. The static conditions in the test section are 0.20 bar and 200 K.

A generic radome configuration of 80 mm diameter and a length of 90 mm was tested at these conditions in HEG. The contour consists of a tangential ogive and a spherical nose tip. In Figure 3.5-50, a Schlieren visualization of the flow past the radome is shown at 5° angle of attack. The shock wave around the model can be clearly identified. The shocks at the top of the flow field image are generated by permanent monitoring probes, which are located off-axis such that there is no interaction with the flow past the radome model on the centerline.

Figure 3.5-50 Schlieren visualization of the flow past a radome at Mach 6 in the HEG.

The surface pressure and the heat flux distributions measured by thermocouples along the windward center-line of the radome are given in Figure 3.5-51. The maximum heat load in the stagnation point is 6.3 MW/m2.

x[mm]

Q/Qstag

p[bar]

0 50 1000

0.2

0.4

0.6

0.8

1

0

4

8

heat flux HEGheat flux CFDwall pressure HEGwall pressure CFD

.

.

Figure 3.5-51 Measured and computed wall pressures and wall heat transfer distributions on a radome at Mach 6 flight in 11 km and 5° angle of attack.

The CFD data shown for comparison in this figure was obtained using the DLR TAU code. In order to model the transition from laminar to turbulent boundary layer flow, transition was forced at the curvature change between the spherical nose tip and the tangential give of the radome configuration. The numerical result using the k- turbulence model agrees well with measured data in HEG.

3.5.7 Outlook

Aerodynamics and thermo-aerodynamics of military vehicles are often characterized by

149

vortex flows, separation, unsteady effects and other difficult conditions such as high pressures and temperatures, driving even nowadays modern numerical and experimental simulation tools to the limits. For a better simulation quality, it is therefore essential to continuously adapt or actively develop modern simulation methods like multidisciplinary approaches, LES, DES, (and DNS) methods, 3-D unsteady optical test techniques, dynamic model supports, etc.

For UCAV configurations, it is planned to further include more disciplines into the design process and also move (partly) to higher fidelity methods. The impact of low signature requirements including integrated intakes will be addressed. The experimental simulation capability in the transonic regime will be extended to unsteady model supports.

For military transport aircraft, the complexity of precision air drop investigations will be raised by adding more details and components to the geometry and a stepwise approximation to real flight conditions. Methods for highly efficient time-accurate coupled high-fidelity simulations of multiple bodies in relative motion will be developed in order to better assess the safety of cargo-

dropping from the aircraft. In addition, the influence of ground effects at heights of down to 15 ft on flight conditions and dropping process are planned.

Future work on missile technologies will take advantage of the new experimental simulation capabilities for maneuver and separation investigations also for the numerical work. Drag reduction by energy deposition has to be further investigated and quantified. The development of intakes with variable geometry for a wide range of (design) Mach numbers will overcome the limitations of fixed geometry inlets.

Future activities on domes include the qualification of TBC’s (Thermal Barrier Coatings), erosion effects by rain and dust particles and its influence on heat loads. Wind tunnel tests on a dome equipped with an operating RADAR sensor unit will allow the evaluation of hot flow effects and wall heating on the microwave path for target detection.

Mission spectroscopy studies will be carried out to measure the IR-radiation of CO2 generated by the aerodynamic shock in front of the radome at Mach 6 in the HEG.

150

3.6 Spacecraft

3.6.1 Introduction

The Institute of Aerodynamics and Flow Technology was involved in all major German and European space technology programs during the last two decades. In particular, the institute participated in the aerothermodynamic design and the post flight analysis of the Atmospheric Re-entry Demonstrator (ARD) project. This first re-entry flight experiment of the European Space Agency (ESA), successfully performed in 1998, allowed the qualification of the institute’s numerical and experimental tools for the analysis of high enthalpy flows. The institute was member of the NASA X-38 project team. Regarding the development of this flight demonstrator of the International Space Station’s Crew Return Vehicle (CRV), the institute was involved in the compilation of the aerodynamic and aerothermodynamic databases. X-38 performed several subsonic demonstration flights between 1998 and 2000, permitting to assess the institute’s tools used to predict the aerodynamics of lifting entry vehicles. Further, the institute led the aerothermodynamics research activities in the framework of the German program TETRA (Technologies for future space Transportation systems). By means of this national program on re-entry technologies, which represents the largest one in this field since the eighties, the Institute developed its capabilities for the analysis of thermal protection systems.

At the beginning of the new century, the institute proposed to utilize sounding rockets to establish a German low cost hypersonic “free flight facility”. This was motivated by the urgent need to test new technologies under real hypersonic flight conditions at affordable cost. DLR met these demands with the definition of the SHEFEX program. The first project, SHEFEX-I, with its successful flight in October 2005 (Figure 3.6-1), was the proof of concept, i.e. the precursor experiment for a series of hypersonic flight experiments to be flown on sounding rockets.

Figure 3.6-1 October 27th, 2005 the Sharp Edge Flight Experiment-I (SHEFEX-I) performed its successful flight; view from the onboard camera.

Aerothermodynamics is the core topic of the institute related to spacecraft development. Numerical prediction methods and major ground based test capabilities were developed, applied and validated. The numerical prediction methods of the institute cover the whole spectrum of the continuum flow regime, from hypersonic to low subsonic speeds. These tools range from engineering pre-design methods and semi-empirical approaches to the DLR high-end computational fluid dynamics (CFD) tool TAU for the multidisciplinary design, analysis and optimization of aerospace vehicles, including the capability of modeling high enthalpy flows and combustion processes. The ground based test facilities of the institute cover a wide range of operating conditions. Due to the diversity of flow conditions and phenomena encountered in hypersonic and re-entry flight, this is of particular importance, because no single facility type can simulate all relevant flow parameters simultaneously. For Mach-Reynolds number simulations, the institute operates the following cold hypersonic ground based facilities: Trisonic Test Section Köln (TMK), Vertical Test Section Köln (VMK) and Hypersonic Wind Tunnel Köln (H2K). The operating range of the hypersonic Vacuum Wind Tunnels Göttingen (VXG) allows Mach-Reynolds simulations at high flight altitudes and rarefied high Mach number flow investigations. The verification and qualification of hot structures of space vehicles is performed

151

in the Arc-heated Wind Tunnels Köln (LBK). In the High Enthalpy Shock Tunnel Göttingen (HEG), the influence of chemically reacting flows on the aerodynamic behavior of re-entry vehicles as well as supersonic combustion in airbreathing engines is studied. In addition, the High Vacuum Plume Test Facility Göttingen (STG) provides world wide unique space vacuum conditions for the plume characterization of small and micro thrusters.

Aerothermodynamics is a field with an extremely wide spectrum of applications in aerospace engineering. In the present period of evaluation, the research activities of the institute concentrated on space transportation, rocket propulsion, hypersonic technology, health monitoring sensors and orbital technology.

3.6.2 Objectives

The major objectives of the institute’s space activities are the development and qualification of numerical and experimental tools for the complete aerodynamic and thermal design of aerospace vehicles, as well as the identification and evaluation of new technologies in order to advance hypersonic space transportation such that it becomes an affordable and reliable routine operation as the air transportation of today.

Essential improvements, far beyond current technologies are required in order to fulfill mission and safety constraints and to ensure economic viability of future hypersonic transportation systems. The growing importance of optimization strategies in the design process, involving the coupling of many disciplines, requires economical use of computer resources and validation of the applied numerical tools by means of carefully designed complex experiments in ground based facilities and in flight. These validation procedures also comprise the development and utilization of advanced measurement techniques. Thus, the three design tools, namely CFD, ground based testing and flight experiments, need to be continuously improved and linked in an optimal manner in order to accomplish the above mentioned objective.

3.6.3 Space Transportation

Due to its expertise, the institute is continuously involved in the aerodynamic and aerothermodynamic design and analysis of a large number of flight vehicles. These activities are supported through specific projects and research funding at national and European level. Examples are the programs ASTRA, SHEFEX, FLPP, AURORA and VEGA as well as the DLR research and development activities and ESA technology research projects. In this context, the three major tasks of the institute are the aerothermodynamic design of new systems, the assessment of industrial concepts and the qualification of thermal protection systems. Driven by these three tasks the institute pursues important complementary activities such as the physical modeling in CFD tools, the development and implementation of measurement techniques for ground based and flight facilities, and the multidisciplinary analysis (MDA).

3.6.3.1 Aerodynamic and Aerothermodynamic Design

The institute’s expertise related to system design is expressed through a series of innovative technology developments which originated from the institute. These DLR financed activities were carried out with the aim to significantly reduce the cost and to improve the performance of space transportation systems. Several of these studies received worldwide attention. Examples are the sharp edge flight experiment, SHEFEX, the liquid fly back booster, LFBB, the electromagnetic launcher for pico and nano satellites, EMAIL, and the electric thruster propelled Moon mission, “Mond-2016”. The institute participated also in the European Commission funded project FLACON which focused on the definition of requirements to be fulfilled by operational space tourism vehicles. These projects triggered the development and improvement of preliminary design tools and led to significant technological progress.

SHEFEX-I

In addition to the initialization and the project management, the major goal of the institute in the framework of the SHEFEX-I flight experiment was to exploit the enhancement of aerodynamic vehicle performance by using

152

sharp edge configurations. Simultaneously, the utilization of a facetted thermal protection system (TPS) will significantly reduce the manufacturing and maintenance cost compared to a conventional TPS system consisting of ten thousands of individually shaped tiles.

The successful flight of SHEFEX-I introduced the institute in the selected community of research laboratories with hypersonic flight capability. In particular it represents the first step towards the demonstration that sharp edge configurations are qualified for hypersonic atmospheric re-entry vehicles.

The aerodynamic layout of the complete sounding rocket launcher as well as the aerothermodynamic definition of the experiment was successfully performed based on Euler and Navier-Stokes computa-tions applying the DLR TAU code. In the framework of the post flight analysis, the potential of the numerical tools for configuration design could be demonstrated. As can be seen in Figure 3.6-2, the angle of attack obtained in trimmed flight attitude during the experiment is very close to that predicted numerically.

Figure 3.6-2 SHEFEX-I flight configuration (top). Comparison of the pre and post-flight assessment of the aerodynamic behavior and comparison with flight data at 20 km flight altitude (bottom).

The precursor SHEFEX-I demonstrated that hypersonic flight experiments are affordable and a valuable complement to ground testing regarding CFD code validation. Indeed, one of the goals of the flight experiment, related to the vehicle

aerothermodynamics, was the acquisition of flight-data for shock wave / boundary layer interaction. Here, turbulent boundary layer flows embedded in a cold hypersonic environment were of particular interest. For the post-flight analysis, numerical computations using a time accurate coupled fluid- / thermal-structure interaction approach were performed. The comparison with the recorded data is exemplarily shown in Figure 3.6-3 for a heat flux sensor located upstream of the compression corner on the lower vehicle surface. Currently, the database obtained during the SHEFEX-I flight for shock wave / boundary layer interaction in cold hypersonic flow constitutes the most complete and reliable one available to the NATO Research and Technology Organization (RTO) community.

Figure 3.6-3 SHEFEX-I flight data evaluation. Example of unsteady CFD (TAU) post-flight reconstruction of the surface heat flux measured by a sensor located upstream of the compression corner on the lower vehicle surface.

The post flight analysis of SHEFEX-I also included ground based testing in HEG and corresponding numerical investigations using TAU. As shown in Figure 3.6-4, excellent agreement between the data related to HEG and the flight data base was obtained regarding the considered pressure distribution on the windward side of the vehicle. The chosen HEG operating condition is related to a SHEFEX-I flight altitude of 21 km. According to the flight data base and the HEG post flight analysis, the shock wave / boundary layer interaction at the compression corner is turbulent at this flight point.

153

Based on SHEFEX-I, the institute expanded its know-how related to the definition, execution and analysis of hypersonic flight experiments including the reduction of the resulting flight data. Furthermore, the institute could for the first time close the loop consisting of aerodynamic and aerothermal ground testing, CFD, and flight testing with own data. The success of SHEFEX-I set up the basis for a long-term cooperation program with the USA Air Force Research Laboratory on hypersonic technologies, reinforced the interest on cooperative space transportation research with the French Space Agency CNES, and added momentum to the cooperation with the University of Queensland in Australia. Last but not least SHEFEX continues to be a strong motivation for young engineers and scientists.

Figure 3.6-4 Comparison of the pressure distribution on the windward side of SHEFEX-I resulting from the HEG post flight analysis (measurements and computations using TAU) with the flight data base (21 km flight altitude).

LFBB

In the framework of the German future space launcher technology research program ASTRA, a partially reusable space transportation system was considered as a potential means to reduce the operational cost of the Ariane launcher.

It consists of two winged reusable boosters, attached to the Ariane 5 core stage, which are able to return back to the launch site after stage separation. The institute was responsible for the aerodynamic design and analyses of these so called LFBBs, employing

the institute’s preliminary design tool HOTSOSE as well as results of flow field computations obtained with the TAU code. Subsequently, the design was verified in the cold hypersonic facilities TMK and H2K. Exemplary results of the study are presented in Figure 3.6-5 showing amongst others the excellent agreement between the predictions of the preliminary design tool HOTSOSE and the measurements obtained in the cold hypersonic ground based facility TMK.

Figure 3.6-5 Aerodynamic design of the liquid fly-back booster (LFBB). Inviscid flow field obtained with TAU for M=0.7 (top); Oil flow visualization in TMK (middle); Computed and measured supersonic pitching moment polars for different canard settings (bottom).

After finalization of the design, an aero-dynamic performance assessment of the complete transport system using a very detailed representation of the complex configuration including subsystems, struts and fairings was carried out (Figure 3.6-6). In the framework of a common study together with CNES in France, the potential of the LFBB concept regarding the development of a micro-launcher for payloads up to 500kg is currently analyzed.

154

Figure 3.6-6 Numerical aerodynamic perfor-mance assessment of a partially reusable space transportation system. Complete configuration (left); Mach number contours of the inviscid flow field resulting from TAU at M=5.99 during ascent (right).

EMAIL

The delivery of small payloads into low Earth orbit by means of propelled payload carriers launched from an electromagnetic accelerator is a concept that has the potential to realize access to space at extremely low cost. However, this Electro-MAgnetIc Launcher (EMAIL) concept requires that during the first 10 to 30 seconds after launch a hypersonic flight within the Earth’s dense atmosphere must be performed. Consequently, high mechanical and thermal loads occur and serious design requirements result for the payload and its carrier. Therefore, the institute’s preliminary design tools were extended to account for the coupling between the flight conditions along the corresponding ascent trajectory and the time dependant heating process of the vehicle structure (see Figure 3.6-7).

Currently, EMAIL is evaluated in the framework of a combined initiative of the French, Spanish and German Space Agencies with the goal to establish solutions for low cost pico- (up to 1kg) and nano-satellite (up to 50kg) launch capabilities. Further cooperative research activities related to EMAIL are pursued together with the University of Texas, USA, and the High Technology Institute of Ukraine.

Figure 3.6-7 Fluid- / thermal-structure coupling analysis of a generic vehicle along a low altitude M=6 flight. Unsteady stagnation point temperature evolution (upper). Instantaneous temperature distribution in the flow field and inside the TPS (1s after launch) for two different TPS materials (lower).

MOON-2016

The considered scenario of an unmanned, electric thruster propelled mission to Moon included a lander plus robotic Moon rover and a sample return capability to the Earth. The performed analysis included preliminary overall mass and energy considerations. After examination of the assets and drawbacks of different alternatives, a direct re-entry maneuver at super orbital speed was proposed, representing a big challenge for Europe. Preliminary investigations regarding the resulting aerothermodynamic loads during re-entry were carried out in cooperation with the University of Braunschweig. Infrared thermo-graphy was applied in the Ludwieg tube of the University of Braunschweig in order to determine the heat flux on the windward side of a potential Moon return vehicle configuration at M=6 (Figure 3.6-8).

C/CSiC

WHIPOX

155

Figure 3.6-8 Computed and measured surface heat flux distribution on the windward side of a potential Moon return vehicle. Experimental result (upper), computation assuming fully turbulent flow (middle) and computation assuming laminar fore body flow with transitional flow at the flaps (lower).

3.6.3.2 Assessment of Industrial Concepts

The institute fulfils a strategic role in Germany and Europe by devoting its numerical and experimental tools for the assessment of industrial concepts. Current activities are mainly related to the aerodynamic and aerothermodynamic characterization of the ESA Intermediate Experimental Vehicle (IXV) - the former French PRE-X re-entry configuration - in the framework of the European Future Launcher Preparatory Program (FLPP), and the characterization of the EXOMARS descent and landing module as part of the ESA AURORA program. In the context of the ESA Crew Space Transportation System (CSTS) project, the evaluation of several capsules and lifting body shapes was performed. Further activities included the assessment of aerodynamic and propulsion issues regarding the Phoenix landing demonstrator and the ESA launcher VEGA.

The main focus of the PRE-X / IXV investigations was the aerodynamic and aerothermodynamic characterization of the configuration and the study of shock / shock

and shock wave / boundary layer interactions at the control surfaces in cold and hot hypersonic flow. Experimental campaigns in two ground based facilities of the institute, HEG (Figure 3.6-9) and H2K (Figure 3.6-10), and ground - to - flight data extrapolation utilizing the CFD code TAU were carried out.

Figure 3.6-9 Pre-X / IXV model in the HEG test section (left) and computed (TAU) surface heat flux distribution for an HEG operating condition at a total specific enthalpy of 22 MJ/kg (right).

Figure 3.6-10 Shock / shock and shock / boundary layer interactions at the control surfaces of the PRE-X / IXV configuration obtained at M=8.7 in H2K. Combined infrared thermography and Schlieren image for low angle of attack showing a classical interaction topology (left) and interaction type occurring at high angle of attack (right).

One important goal which could be achieved in HEG is the identification of the range of control surface deflection angles at which interactions between the bow shock and the shock generated by the body flaps may occur. These must be avoided during operation because otherwise extremely high pressure and heat loads would be generated which could degrade the control surfaces.

In H2K it was observed that at high angle of attack and high flap deflection angle, the shock ahead of the control surfaces is completely detached and a strong shock / shock interaction is generated. The resulting

156

flow topology differs significantly from the one obtained at low angles of attack. Downstream of the interaction region, multiple jet like structures occur which may also lead to high local pressure and thermal loads.

Other activities in the framework of the IXV project were related to the numerical determination of the transonic dynamic derivatives by modeling the force oscillation technique used in the experimental facilities.

Current activities regarding the EXOMARS descent and landing module (DM) include the study of the influence of dust particles on the heat flux rate in laminar high enthalpy flow and the investigation of the drag caused by flow separation after front heat shield separation from the descent module (Figure 3.6-11). Subsequently, the aerodynamic and thermal characterization of the DM in laminar high enthalpy flow and the analysis of the dynamic derivatives as well as the turbulent heating in supersonic and cold hypersonic flow will be performed. In the framework of the EXOMARS project, the numerical aerothermodynamic study of the re-entry phase of a potential Mars return mission was also performed.

Figure 3.6-11 Computational and experimental study of the drag caused by flow separation after heat shield separation from the EXOMARS descent module; Computed (TAU) normalized pressure distribution and near surface stream lines (left);Oil flow visualization in TMK (right).

3.6.3.3 Qualification of Thermal Protection Systems

Arc heated wind tunnels are ground test facilities that allow long-duration tests at high enthalpy flow conditions. With maximum test durations of 120 and 30 minutes respectively, L2K and L3K are well-suited for the characterization and qualification of thermal protections systems (TPS) and hot structures for earth re-entry or planetary entries. While the facilities used to be operated solely with air, they were

upgraded to operate with different gases and gas mixtures, e.g. argon, helium and carbon dioxide. In order to simulate the Martian atmosphere in L2K, a mixture of 97% carbon-dioxide and 3% nitrogen is used. Figure 3.6-12 shows the flow field past a probe in a “clean” Martian atmosphere in comparison to a dust loaded flow during a material qualification test. During Martian entry, a space vehicle might be exposed to dust particles in higher atmospheric levels. Therefore, L2K was equipped with a dust injection system which allows a combined qualification concerning thermal and erosive loads. Depending on the particles’ size, particle velocities up to 2000 m/s were obtained.

Figure 3.6-12 Material qualification in different atmospheres in LBK; Mars atmosphere (CO2/N2 mixture) (upper); Dust loaded CO2/N2 flow (lower).

Material samples or complete structures may be placed into a hypersonic flow for testing. The setup of the facilities can be adapted to particular necessities of a certain test campaign and they can be operated dynamically in order to simulate complete entry trajectories. The objectives include material qualification, design qualification and flight qualification. The design of the PARES stabilizers was thermally qualified in L3K. Special attention was given to the influence of gaps on the heating of the cavity below the stabilizer panels. Other design qualifications were related to the ULTIMATE metallic TPS structure and an improved version of OHB’s aerothermodynamic measurement system (AMS) which allows the simultaneous measurement of pressure, temperature and heat transfer. The original AMS system was qualified in L2K for flight on X-38.

157

Recently, L2K was applied for the qualification of the TPS of an ascending system as well. The new insulation material for the Ariane 5 tanks was thermally qualified in cooperation with CRYOSPACE and Astrium. While in general tests are run in order to demonstrate a structure’s ability to withstand high heat loads, a different approach was pursued in cooperation with “Hyperschall Technologie Göttingen” (HTG) investigating the survivability of satellite structures during uncontrolled atmospheric re-entry.

3.6.3.4 Physical Modeling

An important complementary activity of the institute in support to the above described studies is the assessment of physical models to describe for example turbulent or chemically reacting flows during entry or re-entry. The importance of these studies is highlighted in the following subsections using selected examples.

Turbulence Models for Base Flows

The correct numerical modeling of highly separated flows occurring e.g. in the wake of capsules or winged re-entry configurations at high angles of attack is important for the determination of the aerodynamic behavior. For this type of flow, good results are obtained by hybrid turbulence models such as detached eddy simulation (DES). Therefore, the institute concentrates on the validation of such high order turbulence models for hypersonic flow problems. Figure 3.6-13 indicates that the selection of an inappropriate turbulence model has a strong impact on the base drag prediction. For the considered cylinder configuration, the pressure coefficient on the base would be underestimated by almost 50% when applying a classical RANS turbulence model.

Chemistry Models for Atmospheric Entry

For the validation of physico-chemical models used in the CFD TAU code, the flow in the shock layer of a cylinder placed with its axis transverse to the flow was studied in HEG. For this purpose, holographic interferometry to measure the phase shift distribution in the shock layer as well as

surface pressure and heat flux measurements were applied in conjunction with detailed CFD investigations (Figure 3.6-14). The experimental data which was obtained at total specific enthalpies of approximately 12 and 22 MJ/kg for air as test gas served as benchmark data for the validation of chemical reaction schemes used in CFD codes in the framework of the RTO AVT Working Group 10 “Technologies for Propelled Hypersonic Flight” and the RTO Task Group on ”Assessment of Aerothermodynamic Flight Prediction Tools through Ground and Flight Experimentation” AVT-136/RTG-043. The phase shift distribution can be correlated to the density difference between the free stream and the shock layer.

Figure 3.6-13 Computed (TAU) turbulent wake flow behind a cylinder in a M=2.46 flow resulting from DES; Time averaged turbulent kinetic energy, experimental (upper) and numerical (lower) (top right); Time averaged pressure coefficient across the cylinder base (lower).

The high enthalpy flow models were continuously improved, including the extension to consider other atmospheres than that of planet earth. In order to correctly predict the aerodynamic behavior and heat flux acting on a vehicle, e.g. during Mars entry, the numerical modeling of the CO2/N2 atmosphere has to be validated. For this purpose, different wind tunnel experiments, including the ones performed in LBK, were numerically investigated utilizing the DLR TAU code (Figure 3.6-15). The test cases comprised reacting and non-reacting flows with laminar as well as turbulent boundary layers. The comparison between numerical and experimental results showed that TAU is currently one of the few CFD codes capable to accurately predict 3D

158

flows past a capsule type Mars entry vehicle for a range of flow conditions.

Figure 3.6-14 Experimentally and numerically determined phase shift in the shock layer of a cylinder showing the influence of different chemical models; Cylinder model in the HEG test section including the grid used for the 3-D flow field computations (bottom right).

Figure 3.6-15 Computed mass fraction distribution of CO2 in the shock layer of the EXOMARS configuration (left) and comparison of measured (T5) and computed (TAU) wall heat flux distribution in the symmetry plane of the heat shield (right).

Currently the TAU code is extended to model the entry into the N2-CH4-Ar atmosphere of Titan. This activity also includes a coupling of the numerical tool with radiation and ablation models.

Shock / Shock and Shock / Boundary Layer Interactions

The modeling of turbulence together with the applied chemistry model may strongly influence the numerical prediction of shock / shock and shock wave / boundary layer

interactions at control surfaces in high enthalpy flows. Both effects can have a severe impact on the control surface performance of a re-entry vehicle and hence may jeopardize the complete mission. Due to continued research in this field, started in the framework of the development of the European Re-entry vehicle Hermes, the institute acquired an internationally recognized background. Currently it leads the related research activities of the RTO Task Group AVT-136/RTG-043.

In the framework of the ESA flight experiment project EXPERT, the institute focuses on the assessment of control surface efficiency and heating, particularly for high enthalpy flows. The studies comprise CFD computations using the TAU code and experiments in HEG. A generic flap configuration similar to the one of the EXPERT capsule was investigated (Figure 3.6-16) and it is aimed at comparing the obtained data with those that will be gathered in flight.

q' /q' 0

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.20.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

xz

y

x/L

q'/q

' 0

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.20.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

CFD; open

EXP; open

gap

gap

CFD; closed gap

EXP; closed gap

Figure 3.6-16 Comparison of the normalized computed (TAU, laminar flow) and measured (HEG) normalized heat flux distribution on a generic EXPERT flap configuration in high enthalpy flow; 20o body flap deflection angle with hinge line gap closed (upper) and open (lower).

3.6.3.5 Measurement Techniques

In order to extend the capabilities of the measurement techniques used in HEG, a new short duration internal force balance to measure lift, pitching moment and drag was designed, calibrated and tested (Figure 3.6-17). The balance is able to measure forces in timescales of a few milliseconds on instrumented models at angles of attack from -40 to 20 degrees.

159

Figure 3.6-17 Finite element representation of the internal stress wave force balance with cone model as used in the HEG shock tunnel.

In addition, several optical measurement techniques were implemented in HEG. In such an impulse facility the temporal development of the model flow must be known in order to assure that a steady flow field developed during the test time. Therefore, in addition to time resolved measurements of model surface properties, such as pressure and heat transfer, a high speed shadowgraph and Schlieren system was implemented at HEG. Phase step holographic interferometry was applied in HEG to obtain detailed information regarding the chemical relaxation process in a blunt body shock layer in high enthalpy flow (see chapter 3.6.3.4).

Complementary to the well established static stability measurements in the trisonic wind tunnel TMK, the free and forced oscillation techniques have been expanded to study the dynamic stability of both capsules and lifting body configurations.

3.6.3.6 Multidisciplinary Analysis

The continuous extension and assessment of the multidisciplinary analysis capability of TAU plays a key role in the institute’s aerospace research strategy.

Fluid / Thermal Structure Coupling

In the framework of the research project IMENS (Integrated Multi-disciplinary Design of Hot Structures) a software environment was developed, which is able to perform coupled simulations of hot space vehicle structures. Since validation is a basic requirement for coupled simulations to

become a common design tool, the validation experiments must be performed in well calibrated facilities using sophisticated measurement techniques. Coupled simulations including aerothermodynamics and structure modeling were performed for generic nose cap and flap geometries. The numerical investigation of a generic space vehicle control surface was performed utilizing the TAU code coupled to a structure mechanics solver via a surface interpolation routine. The study showed that due to strong coupling effects between the fluid and the structure, the temperature peaks occurring for example at contour edges in stand alone CFD solutions are not observed in the results of coupled computations. In Figure 3.6-18 this effect can for example be observed at the edge of the open hinge line gap of a generic flap configuration.

Figure 3.6-18 Measured (LBK) and computed (TAU) temperature distribution along the surface of a generic flap configuration with open hinge line gap.

The thermo mechanically coupled analysis of the flow past a generic nose model of a hypersonic vehicle was performed to investigate the effects of heat conduction inside the structure and deformations based on external pressure loads. The numerical results could reproduce the experimental data, showing good qualitative and satisfactory quantitative agreement with the temporal temperature development measured during the experiments as is shown in Figure 3.6-19.

Recently, the coupled simulation was extended to the interaction of high enthalpy flows with cooled structures. In particular, transpiration cooling was considered. Validation experiments were carried out utilizing a flat plate model with an integrated porous C/C sample, which was transpiration cooled using N2 gas as

160

coolant. Figure 3.6-20 shows the infrared thermography images of the model with and without cooling. While the temperatures on the non-cooled side remain unchanged as expected, there are significant changes on the cooled side. The temperature of the sample itself is considerably reduced. In the wake of the sample, the cover plate is cooled as well.

Figure 3.6-19 Schematic of the generic nose model experiment performed in LBK (upper) and comparison of measured and computed (TAU) deformations considering sole thermal and thermo-mechanical deformation (lower).

200.0°C

1200.0°C

Figure 3.6-20 Heat load reduction using transpiration cooling on a porous C/C sample; Non cooled (left) and cooled by 0.4 g/s N2 (right).

Fluid / Flight Mechanics coupling

Dynamic stability analysis based on the so-called free force oscillation technique is a long standing capability of the institute. It was applied experimentally and numerically for several configurations including the SOYUZ capsule and the lifting body configurations X-38 and Pre-X (Figure 3.6-21).

Figure 3.6-21 Dynamic stability analysis: Numerical free force oscillation simulation (TAU) for the PRE-X vehicle (left) and free force oscillation tests at transonic and low supersonic speeds in TMK for X-38 (right).

Due to the significant technical progress in computer technology and CFD, the direct coupling between fluid mechanics and flight mechanics comes into reach, allowing to perform an immediate analysis of a vehicle concept. Since currently the direct coupling of CFD and flight mechanics for the consideration of the free motion of a vehicle along a complete hypersonic trajectory is still too time consuming, a modular coupling procedure utilizing the trajectory optimization code REENT of the University of Stuttgart and the DLR surface inclination method HOTSOSE was developed as a first step (Figure 3.6-22).

Figure 3.6-22 Influence of the spin rate on the motion (angle of attack) of a biconic configuration in hypersonic flow as result of a coupled aerodynamic and flight mechanics analysis.

Currently, the modular coupling procedure is well suited for e.g. the layout of staged sounding rocket experiments such as SHEFEX-II. The chosen approach also allows the integration of the DLR Euler and Navier-Stokes code TAU in the future. An essential prerequisite of vehicle maneuver simulations is the consideration of movable control surfaces. Their integration in numerical computations represents a challenge due to the treatment of the mesh, which must move together with the

161

control surface. In order to avoid a new mesh generation for each surface deflection angle, the technique of overlaid grids (chimera technique) is utilized. Since the standard chimera technique showed artificial shock reflections in hypersonic and supersonic flow, a novel extrapolation boundary condition at the corresponding chimera hole boundaries had to be developed and was successfully applied.

Fluid / Magneto Gas Dynamic Coupling

A potential, novel approach to reduce heating during re-entry is the modification of the ionized flow in the vehicle’s shock layer by influencing the flow with a magnetic field. Experimental investigations of this magneto hydrodynamic (MHD) or magneto gas dynamic (MGD) effect were performed in L2K using argon as test gas.

The photographs in the upper part of Figure 3.6-23 show views of the flow fields obtained in L2K without (left) and with (right) application of a magnetic induction field. The visual appearance of the flow field changes significantly when the magnetic induction field is applied. The IR thermography images reveal the significant reduction of the surface temperature near the stagnation point. This effect corresponds to a heat flux reduction of about 80%. Further tests in HEG and LBK using air are in preparation. From the numerical point of view, the TAU code capabilities were extended to allow the solution of the Navier-Stokes equations, accounting for the presence of a magnetic field and assuming a low magnetic Reynolds number (Figure 3.6-24).

3.6.3.7 Future Activities

In addition to the continuation of the European programs FLPP and AURORA-EXOMARS, the next mission of the SHEFEX flight program, SHEFEX-II, is under development at national level after the successful flight and post flight analysis of SHEFEX-I. Also launched by a two staged sounding rocket system, SHEFEX-II will use an active aerodynamic control system during the re-entry phase. In addition, an actively transpiration cooled thermal protection system element and an advanced flight

sensor equipment will be part of the experimental payload.

Figure 3.6-23 MGD flow control investigations in the arc heated facility L2K. Photographs of the flow field and heat transfer measurements using IR thermography on the front face of the cylindrical model during tests without (left) and with (right) application of a magnetic induction field.

Figure 3.6-24 Numerical computation (TAU) of the influence of a magnetic field on a weakly ionized air shock layer flow in front of a sphere. Pressure contours without (left) and with magnetic field interaction (right). The configuration represents the setup of the planned HEG experiment.

One important objective of the IMENS-3C project is the modeling of ablation processes and its implementation into the fluid / structure coupling process. The validation will be performed with dedicated experiments using new generation ablators in the arc heated facilities L2K and L3K.

3.6.4 Rocket Propulsion

The research activities of the institute related to rocket propulsion were initiated in 2002 and comprise, among others, the aerothermodynamic design and assessment of launchers and their components. For example, nozzle flows and their interaction with the base flow of launchers and nozzle cooling systems are considered. The major goal of the institute’s activities in the field of rocket propulsion is to advance the numerical

162

investigation techniques for rocket thrust chambers. The studies are performed in the framework of the MoU “Propulsion 2010” between DLR and Astrium EADS, the DLR projects IMENS+, IMENS-3C and the European Flow Separation Control Device Working Group (FSCD). In general they are related to the European launcher Ariane 5 and its future developments.

3.6.4.1 Nozzle Flow

An important prerequisite for increasing the efficiency of future launchers is the aerodynamic optimization of the flow in thrust nozzles. This requires a reliable prediction of stationary and transient phenomena occurring inside the nozzle. In Figure 3.6-25, the numerical investigation of a dual-bell nozzle is exemplarily shown.

Figure 3.6-25 Snapshot resulting from the time accurate numerical computation (TAU) of the flow in a dual-bell rocket nozzle; Mach number contours, iso-surface of M=3 and streamlines in a cutting plane (p0 = nozzle reservoir pressure, pa = ambient pressure).

Detailed knowledge of the transition process between separated flow at the wall inflection point and fully attached flow is one requirement for its successful operation. A demand which has to be met by the numerical tool is the correct modeling of the interaction of the high Mach number core flow with the subsonic recirculation zones, turbulent shear layers and the shock system between the high speed core and the recirculation zones. The DLR TAU code was applied to study this nozzle concept. Experimental wall pressure distributions and locations of the transition point could be reproduced by CFD with good agreement for a wide range of chamber pressures. The hysteresis effect, which is important to prevent the nozzle from running into a flip-

flop regime, has been clearly reproduced in the CFD results.

3.6.4.2 Base Flow

For future rocket technologies a deeper insight into unsteady phenomena during the start phase of launchers is essential. In the framework of the HGF virtual institute RESPACE, experimental investigations of the hot plume (up to 1000 K) / external flow interaction was one of the main tasks (Figure 3.6-26). Experiments in H2K showed different frequencies of internal and external shock oscillations.

Figure 3.6-26 Pitot pressure contour and Schlieren image of a base flow field in H2K.

Unsteady interactions and resonances of flow separation inside the nozzle, the turbulent launcher wake, and the nozzle structure will play an important role in the design of future main stage propulsion systems. This so-called buffeting coupling phenomenon is one of the main challenges during ascent and was also a concern during the development and the first flight of the Ariane 5 launcher. Since the investigation of the complete scenario, including the external flow and the hot plume, is currently not possible in ground based facilities, numerical studies of this phenomenon are required. In this context, the application of coupled fluid / thermal-structure computations as well as advanced turbulence models is essential. Therefore, in order to obtain realistic descriptions of separated unsteady base flows, detailed studies of the applicability of modern turbulence models such as DES were carried