New Results in Numerical and Experimental Fluid Mechanics XI

On a swept vertical tailplane with infinite span tangential blowing over the shoulder of a deflected rudder is applied. For large rudder deflection angles the flow on the rudder is separated without blowing. A numerical study is conducted with the aim to increase the side force coefficient which might be required for a take-off condition if a one-sided engine failure occurs. With a continuous slot and sufficient mass flow rate the separation on the rudder can be reduced or avoided. It is shown that by using discrete slots this can be achieved for a similar side force coefficient gain with a smaller momentum coefficient. In addition the sweep angle of the incoming flow is varied showing a strong impact on the achievable side force coefficient. This is also true for the curvature of the rudder shoulder over which the jet is blown.

A broad range of numerical flow simulations are carried out during the design phase of a highly bent intake geometry. The main aim is to evaluate the aerodynamic characteristics of a projected wind tunnel model and an estimation of mechanical loads for the structural dimensioning. The numerical setup using the TRACE code is validated first against comprehensive experimental data of a NASA s-duct test case. Three different turbulence models are found to be capable of reproducing the main flow features that occur in bent intake ducts with an acceptable accuracy. The following steady simulations of the symmetric wind tunnel model show asymmetric flow solutions and convergence problems for two of the three turbulence models. URANS computations are therefore carried out including a sensitivity study towards time-step size and domain volume. The unsteady results using the three different turbulence models still exhibit significant deviations concerning mechanical loads and duct performance. A safety margin is thus estimated from the unsteady data to be used for the construction and testing of the wind tunnel model.

Aerodynamic characteristics of an engine side intake of a light helicopter are investigated experimentally and numerically with an emphasis on fast forward flight (Ma < 0.2). For this purpose, a novel full scale model of a helicopter fuselage section has been developed and tested. With a duct system, including a fan, engine mass flow rates are realized corresponding to real helicopter operating conditions. The wind tunnel data comprise aerodynamic interface plane (AIP) five-hole pressure probe and particle image velocimetry data which are compared to unsteady Reynolds-averaged Navier-Stokes simulations results. In order to assess the influence of the truncation of the fuselage, two numerical cases are compared. The first is a sectional fuselage case, incorporating all wind tunnel model components, and influential wind tunnel parts. The second is a full fuselage case. Only small differences between the sectional and full fuselage cases in flow direction upstream of the intake are present whereas the AIP total pressure distributions provide the same characteristics. The experimental data are also used to validate the numerical simulations while the complementary experimental and numerical data enables a detailed analysis of the flow distortion up to the AIP.

Results of numerical simulations on a high-lift configuration with an Ultra High Bypass Ratio (UHBR) engine are shown. In the area of the integrated engine, a complex vortex-system develops. Different longitudinal vortices proceed downstream on the suction side influencing the local flowfield. Steady and unsteady numerical simulations are performed at different angles of attack. As flow solver the DLR TAU Code is used and the Menter-SST eddy viscosity turbulence model and the JHh-v2 Reynolds-Stress-Model are applied. The predicted vortex system is analyzed and the effect on the flow field and the local stall behavior is shown. In particular the effect of the turbulence models of different types on the prediction of the vortex system and the flowfield is pesented.

Aerodynamic interference effects and potential benefits of an over-the-wing mounted (OWM) engine installation are examined with CFD methods. The reference configuration is a novel transport aircraft concept with STOL capabilities. The wing/body configuration is compared to the wing/body/engine configuration with power-off nacelle. Subsequently, jet-on computations are carried out to address the influence of the jet presence on the engine/airframe interference effects. The engine is initially considered in a reference position. Successively, position variations of the engine are performed to analyze the interaction with the supersonic region and shocks on the wing upper surface. To conclude the study, a preliminary evaluation of the installation drag is done, to investigate the potential of favorable interference effects and reduced drag with respect to more conventional under-the-wing mounted (UWM) engines installations.

The work summarizes recent efforts concerning the simulation of massively separated wake flows of commercial transport aircraft configurations via detached-eddy simulation (DES). Two levels of grid resolution and two time step sizes were defined for the unsteady simulation runs, with the aim of improving the existing best practices with respect to this type of simulation. The number of time steps required for statistics convergence of the results is evaluated. In the flow field, comparisons of the results between the runs show that lowering the physical time step increases the velocity deficit in the wake. By comparison, the time step’s impact on the generation of large-scale turbulent structures near the wing is smaller in magnitude. The reduction of grid resolution strongly impacts prediction of the leading edge shear layer, leading to significantly lower pressure fluctuations on the wing. The discretization of convective mean flow fluxes was observed to play a significant role in the context of shear layer separation.

Direct numerical simulations of turbulent pipe flow in a flow domain of length $$L=42R$$ and a friction Reynolds number of $$Re_{\tau } =180$$ and two different wall normal grid refinements were carried out and investigated in terms of turbulent high-order statistics. The phenomenology of large local wall-normal velocity fluctuations (velocity spikes), is discussed through time series and instantaneous flow field realizations. Due to their rare appearance both in space and time, statistical high-order moments take long time to converge. A convergence study is performed and for fully converged statistics the sensitivity of the wall-normal kurtosis component value on the wall is investigated. The fourth order statistical moment of the wall-normal velocity needs to be integrated at least $$\varDelta \tau ^+ = 2.4\cdot 10^{11}$$ units of an averaging coordinate in time and space to obtain a fully converged flatness value in the very vicinity of the wall. Furthermore, a near-wall grid spacing of at least $$\varDelta r^+=0.11$$ is needed to fully satisfy the no-slip boundary condition in terms of wall-normal flatness.

We present a turbulent boundary layer flow experiment at a significant adverse pressure gradient and at a high Reynolds number. We describe the design of the test case and the set-up in the wind-tunnel so that the flow is suitable for the validation of RANS turbulence models. We present RANS simulations using the SST k-$$\omega $$ model, and the SSG/LRR-$$\omega $$ and the JHh-v2 Reynolds stress model. We show that the predictive accuraccy in the adverse pressure gradient region is significantly effected by the predictive accuracy in the upstream located region of a favourable pressure gradient, where the mean flow follows a curved surface. The effects of flow history have to be taken into account when assessing turbulence models in the adverse pressure gradient region. We study in detail the role of the cross-diffusion term in the $$\omega $$-equation for favourable and adverse pressure gradients.

The present study is a part of a high-Reynolds-number experiment designed to establish a reliable data base helping to improve RANS turbulence models. In this work skin friction measurements were conducted in turbulent flows with significant pressure gradient using oil-film interferometry. The experiments were performed on a large-scale test model for three different flow velocities of 10, 23 and 36 m/s. Analysis of the skin friction distributions indicates that the state of the turbulent boundary layer remained out of equilibrium along the whole area investigated.

An active grid with individually controllable paddles was constructed to generate strong turbulence (Re_λ > 200). Measurements of turbulence quantities using the hot-wire technique were conducted in a wind tunnel along the flow on the centre line and on an off-centred line in the wake of this grid. The effect of two different motion sequences of the active grid paddles as well as the influence of the clock rate of the paddles were tested. Finally, measurements with a fixed paddle angle were carried out. The results confirmed that the active grid is suitable to generate the intended strong turbulence. Maximum Re_λ was obtained when the uniform probability density distribution was used. Turbulence generated with fixed paddle angle still delivered turbulence with Re_λ ~ 100.

Disturbance level measurements are carried out in the Hypersonic Ludwieg Tube Braunschweig (HLB) at $$\text{ Ma } = 5.9$$ considering three shapes of stagnation point probes (SPP) and two different types of high speed surface sensors. Navier Stokes mean-flow solutions are produced for the flow around the front shapes to increase the understanding of the dependence of the mean flow on the probe shape. Performed direct numerical simulation (DNS) for one SPP configuration delivers transfer functions as the ratio of incoming disturbance to surface data at the stagnation point. The transfer functions are used for computing the modal decomposition of representative disturbances in the freestream.

A supersonic film-cooling configuration with shock interaction is investigated experimentally by means of high-speed particle-image velocimetry. A laminar cooling film is injected at an injection Mach number of $$Ma_i=1.8$$ beneath a turbulent boundary layer at a freestream Mach number of $$Ma_\infty =2.45$$. A flow deflection of $$\beta =5^\circ $$ generates a shock wave which impinges upon the cooling film. The influence of the impingement position of the incident shock on the cooling film is analyzed by the time-averaged velocity fields and Reynolds shear stress distributions. The results show a significantly, shock-induced increase of the turbulent transport. Shock impingement close to the injection nozzle equally increases the Reynolds shear stress in the shear layer and in the boundary layer on the bottom wall. An impingement position further downstream drastically increases the turbulent transport in the boundary layer, however, in the shear layer the increase in turbulent transport is lower.

For a spacecraft traveling at hypersonic speeds, the design of the vehicle’s thermal protection system is crucial [2] and the accurate quantification of the surface heat flux is mandatory to aid this design. Traditional surface-mounted temperature sensors, such as thermocouples and thin film gauges, have been widely used as reported already over 40 years ago by [19]. These sensors have response times on the order of a few microseconds, and thus data can be sampled with high temporal resolution; typically 1 MHz. However, the spatial resolution is less satisfactory: the packing density is limited by the sensor size and a great amount of expertise is needed to mount sensors on narrow geometries (e.g. sharp leading edges or corners). In addition, the installation of large numbers of transducers results in complicated wiring and high costs. The clear advantages of temperature sensitive paints (TSPs) over the classical techniques are the non-intrusive measurement approach and the high spatial resolution. A feasibility application of fast-response TSPs in Europe’s major hypersonic facility, the High Enthalpy Shock Tunnel Göttingen (HEG), on transition measurements has already been published [18]. These measurements demonstrated the ability of the TSP method to be used for measurements of time resolved surface temperatures and consequently the determination of surface heat fluxes; the TSP used in the publication was based on the luminophore Ru(phen). This paper will discuss the prerequisites, the requirements and the calibration work which has to be accomplished to progress the development of ultra-fast temperature sensitive paints towards a more versatile application in hypersonic high enthalpy flows in the HEG. Additional information on the HEG is reported by Hannemann and Martinez Schramm [7]. The calibration procedure is shown for the laser dye 4-Methylumbelliferone (4MU) which has been determined to be the next promising candidate for the application in the HEG [15].

In this publication the redesigned DLR swept flat-plate experiment is introduced. It is based on the well-known crossflow reference experiment by Bippes et al. and consists essentially of a swept flat plate and a displacement body arranged above it. The differences between the old and the new setup and the reasons for the redesign will be described. The experiment is used for investigations on the mechanisms leading up to crossflow-induced laminar-turbulent transition in the flat-plate boundary layer and on different actuation methods for influencing them. The main features of the 3D boundary layer flow field such as the development of crossflow instabilities and the onset of high-frequency secondary instabilities during the laminar breakdown are presented.

The flow around delta wings is dominated by a leading-edge vortex system, which induces increased near wall velocities above the wing hence producing high suction peaks. These are responsible for the lift needed at high angle of attack aircraft maneuvering. In the flight regime beyond stall the flow separating from the leading-edge encounters a very steep adverse pressure gradient and consequently doesn’t roll up into a vortex-like structure. Rather, encloses a massive dead-water region over the entire wing. With unsteady jet blowing at the leading edge additional momentum is created leading to a reattachment of the flow at the wing surface thus increasing the lift significantly. The investigated flow control method can be applied for extending the flight envelope, enhancing maneuvering capability and flight stability. This flow manipulation technique is investigated on a generic half wing model at a very high angle of attack (α = 45°). The investigations comprise wind tunnel testing, using force measurements and stereoscopic particle image velocimetry, and complementary scale resolving numerical simulations, for a detailed analysis of the unsteady phenomena.

In this paper, spatial linear stability analyses are performed on three flow configurations where rotational effects are present. The first two configurations are the two-dimensional flows along a flat plate and along a curved plate, both with a rotation vector along the spanwise direction. The third configuration is the three-dimensional flow over a rotating disk with and without axial inflow. These flows are used as a verification of the extension of the stability analysis code NOLOT to rotating frames. For all these three flows, a perfect matching is observed in comparison with results from the literature for the neutral curves and the eigenfunctions. The present results show that rotation has a considerable effect on the stability of boundary layers. Depending on the intensity and orientation of the rotation vector, stabilization or destabilization of the boundary layer is observed.

This work investigates the flow in the near field of a patch of distributed roughness in the high-enthalpy boundary layer of a generic, three-dimensional Apollo capsule geometry at an angle of attack of $$24^\circ $$. Parallel Direct Numerical Simulations are performed for the laminar boundary layer on the windward side of the capsule for four different gas models. The effects of molecular dissociation as well as chemical and thermal non-equilibrium are presented for both the unperturbed flow and the flow behind the roughness. The development of the steady disturbances downstream of the roughness patch is analysed and quantified.

This paper presents a one-equation transition model for flows over airfoils and wings of transport aircraft. The transition model is based on a transport equation for an intermittency variable $$\gamma $$. It is a simplification of the local correlation based $$\gamma $$-$$\textit{Re}_\theta $$ transition model and is designated $$\gamma $$ transition model. The $$\gamma $$ transition model is built for external aerodynamic flows at high Reynolds numbers in a low turbulence environment. Special emphasis is put on the effect of favorable pressure gradients. The paper presents the $$\gamma $$ transport equation, two-dimensional validation test cases, and an example of a three-dimensional test case governed by Tollmien-Schlichting transition.

Dynamic forced impingement cooling is investigated experimentally. Due to the pulsation of impinging jets, strong ring vortices can be generated. Thereby, pulse parameters, e.g. frequency, amplitude, duty cycle and phase shift as well as geometrical parameters e.g. impingement distance, jet nozzle distance or jet nozzle arrangement, play an important role on maximizing the heat transfer on hot surfaces. In the present paper the local convective heat transfer on an electrically heated flat plate is being investigated, while under the influence of a 7 by 7 inline impinging jet array. The cooling efficiency of pulsed actuation with side wall induced crossflow is determined by comparing local Nusselt numbers to the steady blowing case.

Laminar–turbulent transition in (quasi–) three–dimensional boundary layers dominated by stationary crossflow vortices is studied for the setup of the DLR swept–flat plate experiment. The linear and subsequent nonlinear development of both the stationary crossflow vortices and their high–frequency secondary instabilities are modelled by nonlinear parabolized stability equations (PSE). In contrast to previous work, secondary instability theory is used only for initialization of secondary instabilities within the PSE analysis. Thereby, the nonlinear development of secondary instabilities including the generation of higher harmonics could be studied up to the stages where the feedback of the finite–amplitude secondary instability modes on the stationary crossflow vortices is no longer negligible and the skin–friction coefficient starts to deviate from that due to the mean flow distortion caused by the stationary crossflow modes alone. Moreover, the present approach allows rather efficient studies on the nonlinear development of secondary instabilities independent of their frequency.

In this work, a novel approach for flow manipulation by standing acoustic waves is presented. A phased array of ultrasonic transducers operating at 40 kHz is used to generate acoustic waves of various patterns. The interaction of these wave patterns with a thermally induced air flow (thermal convection) is evaluated. The Background Oriented Schlieren method (BOS) in combination with a thermal “contrast medium” allowed a clear visualization of the effects. The final images were generated from the raw data by applying an easy to implement pixel-by-pixel difference method.

Hybrid Laminar Flow Control (HLFC) is one of various promising future aircraft applications intended to reduce aerodynamic drag and hence also fuel consumption. Airbus investigated laminar flow control technology in previous decades with active suction at the leading edge of the vertical tail plane (VTP). Recent HLFC approach concerns not only laminar flow control but also system details such as internal compartment shape and air outlet design. One possible technical solution involves passively driving the system through suction at an air outlet. The current CFD investigation is intended to aid the definition of a robust system solution that will be implemented on a flight test aircraft in upcoming years. Here, HLFC will be applied at a specified VTP section. This investigation includes variations of outlet size, shape, opening angle and location at a typical passenger aircraft VTP. A detailed analysis of typical cruise flight conditions, i.e. variations of Mach number, flight level, AoA and sideslip angle, is performed while the focus is on outlet aerodynamic performance. One of the main objectives is achieved suction pressure and hence the mass flow driven through the system. There are large differences in system efficiency depending upon the outlet location and shape. A system mass flow requirement in order to apply HLFC at a specified VTP section is fulfilled by all three investigated air outlet locations: at the VTP tip, at a middle location and at a close-to-root location. However, the aerodynamic performance of these outlets is very different. Locations at the VTP tip and mid-VTP require large outlet areas and/or large flap opening angles. This generates high aerodynamic drag. The lower VTP location is found to be the most promising compromise with respect to the achieved suction pressures and associated aerodynamic drag.

In order to determine the critical cross flow N-factor for DNW-NWB wind tunnel, transition experiments are conducted on a 3D-wing model. Infrared thermography is used to detect spanwise the transition line at various angles of attack and Reynolds numbers. The obtained transition lines are then prescribed in RANS flow simulations of the experiment. Calculated boundary layer data of laminar regions serves directly as input for a subsequent analysis of boundary layer stability. Employing local linear stability analysis, critical N-factors are attained at the transition location for each inviscid streamline. A 2-N-factor method is used for individual treatment of Tollmien-Schlichting and cross flow instability. From the broad variety of observed transition scenarios, cross flow dominated cases are filtered. Those are used to calculate an uncertainty weighted mean value of the critical cross flow N-factor for NWB facility. The results show N_CFcrit to be in the range of 7.4–9.3. Remaining scatter and a noticeable dependency of N_CFcrit to the surface side of the model are likely to be related to methodological shortcomings of the eN method.

The influence of spanwise periodic steady blowing or suction on the development of stationary crossflow vortices is investigated experimentally in a three-dimensional boundary-layer flow. Different jet-outlet geometries were tested to control the naturally most amplified crossflow mode in this scenario. Although bypass transition at the actuator orifices is an issue in the blowing cases, the strongest receptivity was observed for steady blowing against the direction of the crossflow veloctiy component.

Within the following computational study, Tollmien-Schlichting waves will be simulated behind a step on a laminar profile by direct numerical simulations. The airfoil geometry is designed as a reference-configuration for wind tunnel experiments. The character of the dominating transition mechanism by Tollmien-Schlichting waves allows a 2D approach. A well-validated high-order numerical scheme is chosen, which is adapted for detailed simulations of transitional modes. All simulations of the instabilities were carried out on the original airfoil-geometry by using an entirely resolved step-region. The central part of the airfoil in the open test-section of the DNW-NWB contains an adjustable inset to allow settings of different forward and backward-facing steps, comparable with high-resolution CFD-results. Efficient time-accurate simulations of the TS-mode growth were carried out on extracted CFD-grids of the critical region. Determining growth rates of these modes allow comparison with Linear Stability Theory (LST) and the prediction of the laminar-turbulent transition.

The flow features developing in the presence of a two-dimensional forward-backward facing step of small width are visualized. The flow visualizations were conducted by means of Temperature-Sensitive Paint (TSP). The tests were conducted in the Laminarwasserkanal (laminar water channel) at the Institute for Aerodynamics and Gas dynamics (IAG) in Stuttgart. A step Reynolds number Re _k from 199 up to 1890 enables the investigation of laminar, transitional and turbulent reattachment of the flow. The variations of flow separation and impinging structures related to reattachment are well captured by the TSP technique. The TSP method requires a temperature difference between the model surface and the flow to visualize flow phenomena. This temperature difference alters the flow structures around the forward-backward facing step for low Reynolds numbers by inducing buoyancy forces in the boundary layer and recirculation zone.

In this experimental study, the effects of an isolated, cylindrical roughness element on laminar-turbulent transition is investigated and compared to three-dimensional, linear stability theory. Experiments have been done at both sub- and supercritical Reynolds numbers. In the subcritical case, controlled forcing was applied to show that the roughness amplifies unsteady disturbances which can grow large enough to trip the boundary layer. At a higher, but still subcritical Reynolds number, natural disturbances suffice to trip the boundary layer intermittently. Further increase of the Reynolds numbers leads to permanent transition. This transition Reynolds number and dominant frequencies are compared to the critical Reynolds number from theory to provide experimental evidence of a global instability.

The influence of Mach number ($$M = 0.35$$–0.65), chord Reynolds number ($$Re_c=6\times 10^6$$ to $$10\times 10^6$$) and pressure gradient ($$dc_p/dx = -0.6$$–0.07 m$$^{-1}$$) on laminar-turbulent boundary layer transition was experimentally investigated in the Cryogenic Ludwieg-Tube Göttingen (DNW-KRG). For this investigation the existing two-dimensional wind tunnel model, PaLASTra, which offers a quasi-uniform streamwise pressure gradient, was modified in order to reduce the size of the flow separation at its trailing edge. The streamwise temperature distribution and the location of laminar-turbulent transition were measured by means of temperature-sensitive paint (TSP). It was found that the transition Reynolds number exhibits a linear dependence on the pressure gradient, characterized by the Hartree parameter, and that an increasing Mach number leads to a linear decrease of the transition Reynolds number. The latter effect is likely due to an increase of the total pressure turbulence level with Mach number in DNW-KRG. The measured pressure and temperature distributions served as input for boundary layer calculations and linear-stability analysis. N-factors were calculated according to compressible and incompressible stability theory. At zero pressure gradient a critical N-factor of approximately 9.5 and 9.0 was found for incompressible and compressible calculations, respectively.

The wake flow behind a classical space launcher is highly unsteady and dominated by large-scale flow separation, especially under the influence of an under-expanded propulsive jet. Strong wall-pressure oscillations occur and can excite detrimental structural vibrations. This study investigates the applicability of convoluted trailing edges/lobe mixers to reduce the separation length and the dynamic loads on a generic axisymmetric launcher model. Experimental investigations were performed at a free-stream Mach number of $$M_{\infty } = 2.9$$ and a Reynolds number based on the model diameter of $$Re_D = 1.3 \times 10^6$$. A small reduction of the separation length due to the increased turbulent mixing effected by the lobes was observed, and a promising moderating influence on the effects of the jet plume in the low-frequency range could be achieved.

Shape optimization of an active and a passive drag-reducing device on a two-dimensional D-shaped bluff body is performed. The two devices are: Coanda actuator, and randomly-shaped trailing-edge flap. The optimization sequence is performed by coupling the genetic algorithm software DAKOTA to the mesh generator Pointwise and to the CFD solver OpenFOAM. For the active device the cost functional is the power ratio, whereas for the passive device it is the drag coefficient. The optimization leads to total power savings of $${\approx } 70\%$$ for the optimal Coanda actuator, and a 40% drag reduction for the optimal flap. This reduction is mainly achieved through streamlining the base flow and suppressing the vortex shedding. The addition of either an active or a passive device creates two additional smaller recirculation regions in the base cavity that shifts the larger recirculation region away from the body and increases the base pressure. The results are validated against more refined URANS simulations for selected cases.

We investigated a non-steady compressor stator flow of the type that would be expected in a pulsed detonation engine. The effects on the stator vane of the last stator stage were discussed and results from the two-dimensional compressor stator cascade in Berlin are presented. Static pressure measurements showed periodic flow separation phenomena on the measurement blade with respect to the working-phase of a choking-device mounted in the wake of the cascade imposing the non-steady outflow condition. By means of active flow control (AFC) it was possible to avoid flow separation on the measurement blade. Furthermore a figure of merit was defined in order to rate the increase of static pressure recovery with respect to the AFC input. Using an optimized AFC setup the flow-field could be stabilized.

Two-dimensional direct numerical simulations (DNS) are used to study the influence of active and passive suction at the juncture between two wing elements on laminar-turbulent transition. Suction is applied through the juncture’s gap in front of a forward-facing step. For first, fundamental investigations, the juncture is located in a compressible flat-plate boundary-layer flow without streamwise pressure gradient. To quantify the impact of suction on laminar-turbulent transition, N-factors are calculated. For this, the growth of introduced Tollmien-Schlichting (TS) waves is evaluated according to the $$e^N$$ method. Depending on the perilousness of juncture shapes, suction is capable or incapable to compensate their negative impact on laminar-turbulent transition. In addition to active suction, a method allowing passive suction is presented. The work concludes with a discussion about possible integrations of suction at aerodynamic elements, in particular at wings.

An experimental study of the effect of vortex ingestion into an S-duct diffuser on flow conditions at the engine face is presented. A simultaneous ingestion of a boundary layer is also considered. In order to form a boundary layer of well-defined thickness ahead of the intake the diffuser is arranged on a flat plate. The dependence of the circulation of the ingested vortex on the flow in the diffuser has been examined in a parametric study in a blow-down wind tunnel. The free-stream Mach number was varied between M = 0.30 and M = 0.65 at a fixed mass flow. In the introduction the importance of boundary layer ingesting highly integrated intakes is briefly discussed. This is followed by a description of the analytical design of the generic intake model, the cryogenic blowdown tunnel DNW-KRG in which the experimental tests have been performed, and the measurement techniques. Thereafter, experimental findings are discussed in detail and conclusions are drawn.

A method for the reconstruction of 3D density distributions of blade tip vortices is presented. The reconstruction is performed using a filtered back-projection on 2D density gradient data from a high-speed LS-BOS (laser-speckle-illuminated background-oriented Schlieren) setup, captured during the STAR (smart-twisting active rotor) project. Density profiles of blade tip vortices are presented, as well as comparisons of the size and development of the vortices. In addition the comparison to a PIV (particle image velocimetry) measurement is shown.

Several analysis algorithms were investigated for the detection of transition on the pitching airfoil DSA-9A. The state of the art method to detect unsteady boundary layer transition with experimental hot film data is a manual approach. Therefore, to reduce the investigation time and improve the data quality an automated detection has been created. In the most cases the computation time of the unsteady boundary layer detection is almost two orders of magnitude faster compared to manual extraction. Several algorithms will be discussed in this paper of which an algorithm utilizing the skewness produces the best results in a robust implementation.

The wake of wind turbines is modeled as a tip vortex helix with a vortex strength estimated from its rotor thrust. A Bo105 helicopter and an analytical isolated rotor model are subjected to the wake. In both cases the trim perturbations generated by the wake are computed and compared to each other. It is found that the rotor control margins are sufficient to compensate the wake disturbance.

Effects and influences of wing sweep on flutter mechanisms are investigated. For sweep angles up to $$40^\circ $$ and various positions of the elastic axis, the aeroelastic stability and in particular the shift of the flutter velocity is presented. The results are acquired by wind tunnel tests at subsonic flow conditions using a 2.5D configuration. The low Reynolds number airfoil Eppler E171 has been mounted with two degrees-of-freedom (2-DOF) in a flutter test rig, whereby measurements with a specifically limited parameter range are performed. The flutter behavior is changed due to different positions of the elastic axis, but is almost independent of sweep. The flutter boundary respectively the flutter velocity shows a significant dependency on sweep angle and the position of the elastic axes, which has not been reported in theory and experiments so far.

This work presents a model of an elasto-flexible membrane airfoil. As the deformation of the configuration is significant to change the fluid flow itself, so called 2-ways Fluid Structure Interaction (FSI) simulations are performed to reproduce the aerodynamics of the profile: the solution is achieved in different iterative loops where both models are mapped to each other until convergence is found or the process is stopped manually. A partitioned approach is used between the FEM solver CARAT++ and the U-RANS solver TAU to model the configuration. In order to validate the model, aerodynamic data, the deformation and the flow field resulting from the coupling are compared with experimental and numerical results generated with ANSYS. Some discordance appears between the two approaches: on the one hand, 3D effects during the experiments are very significant and on the other hand, the absence of the contact modeling in CARAT++ results affects the comparison.

A novel nonlinear system identification approach is presented based on the coupling of a neuro-fuzzy model (NFM) with a multilayer perceptron (MLP) neural network. Therefore, the recurrent NFM is employed for multi-step ahead predictions, whereas the MLP is subsequently used to perform a nonlinear quasi-static correction of the obtained time-series output. In the present work, the proposed method is applied as a reduced-order modeling (ROM) technique to lower the effort of unsteady motion-induced computational fluid dynamics (CFD) simulations, although it could be utilized generally for any nonlinear system identification task. For demonstration purposes, the NLR 7301 airfoil is investigated at transonic flow conditions, while the pitch and plunge degrees of freedom are simultaneously excited. In addition, the sequential model training process as well as the model application is presented. It is shown that the essential aerodynamic characteristics are accurately reproduced by the novel ROM in comparison to the full-order CFD reference solution. Moreover, by examining the results of the NFM without MLP correction it is indicated that the new approach leads to an increased fidelity regarding nonlinear ROM-based simulations.

Mixed convection in a differentially heated channel with an additional concentration gradient is investigated via direct numerical simulation (DNS) using a finite volume method based on fourth order accurate central differences in space and an explicit time integration scheme. The fluid is treated as a homogeneous fluid characterized by its temperature and mixing ratio. The Boussinesq approximation is employed to calculate the influence of the buoyant force resulting from density differences due to both quantities in preparation of future investigations of humid air flows in various ventilation applications. The simulations at different values for the modified Grashof number show a distinct effect of the mixing ratio on the flow and temperature profiles for values approximating the conditions found in humid air, in line with those found for purely thermal systems with comparable Grashof number.

This report compares two approaches for achieving a trimmed state of an aircraft configuration during an aerodynamic optimization. In the optimizer-based approach, balance equations are set as equality constraints to the optimizer. In the flow solver-based approach, balance equations are satisfied within the flow solver evaluation. These approaches are applied to a flying wing case, where blended trailing edge deflection is used to control the pitching moment. The wing is treated as rigid, and lift and pitching moment balance equations are taken into account for trimming. Tests are performed with varying numbers of shape design parameters and with varying numbers of flight points. It is concluded that the flow solver-based approach performs more robustly, and thus should be preferred in general, even though it may take more time than the optimizer-based approach.

We present recent developments within a high-performance discontinuous Galerkin solver for turbulent incompressible flow. The scheme is based on a high-order semi-explicit temporal approach and high-order spatial discretizations. The implementation is entirely matrix-free, including the global Poisson equation, which makes the solution time per time step essentially independent of the spatial polynomial degree. The algorithm is designed to yield high algorithmic intensities, which enables high efficiency on current and future CPU architectures. The method has previously been applied to DNS and ILES of turbulent channel flow and is in the present work used to compute flow past periodic hills at a hill Reynolds number of $$Re_H=10595$$. We also outline our on-going work regarding wall modeling via function enrichment within this framework.

A mathematical tool developed for calibrating model parameters of VRANS equations for modeling flows through porous medium is evaluated. A total of six parameters are introduced in a volume averaged RANS model to appropriately scale the impact of porous media on the overall flow. The calibration tool has been tested for a generic channel case and the results are compared with DNS simulations of the same. The results show a good agreement between the parameters obtained from the tool and a manual calibration documented previously.

This paper presents an assessment of a low-dissipation low-dispersion finite-volume scheme in the DLR-TAU code for scale-resolving simulations on unstructured grids. The scheme is tested in both wall-resolved and wall-modelled LES of the plane channel flow, in pure LES of the periodic 2-D hill flow on a family of grids, and in a hybrid RANS/LES of a generic 3-D Delta wing. Overall, the LD2 scheme is found to yield consistent results and sensitivities for structured and unstructured grids.

Turbulent mixed convection in a differentially heated vertical channel is simulated by means of direct numerical simulation and large eddy simulation. With regard to the LES different subgrid-scale models are compared and their ability to predict the flow is analysed. The comparison is based on statistical moments as well as two-point correlations that show the models’ abilities to represent the sizes of flow structures. A newly proposed local formulation for the classical Smagorinsky SGS model has been identified as the best choice for mixed convection flows, although it can still be improved. The importance of the accuracy of SGS heat flux models is shown to be negligible when the momentum transport is not predicted properly. Their influence can even lead to deviations in both momentum and heat transport as shown with the classical dynamic Smagorinsky model in combination with different SGS heat flux models.

The focus of this article is to describe the evolution of the spreading diameter and secondary droplets generated by splashing. High-speed visualization was used to study the time evolution of water droplets impacts with dry surfaces at Weber numbers between 3,500 and 10,000. Different prediction models of the maximal spreading diameter have been compared with each other and with the experimental data. A similarity between the spreading rates was observed in the last stage of the impact at high Weber numbers. The time evolution of the secondary droplets and the formation of the crown was observed and analyzed at the different Weber numbers.

The understanding and prediction of high-lift aerodynamics of a civil aircraft in landing configuration still lacks validated and comprehensively assessed databases involving numerical simulations, wind tunnel tests as well as flight tests. The joint research project HINVA (High lift INflight VAlidation) aims on closing this gap for a short to medium range transport aircraft together with its specific high-lift devices. The research aircraft, used to apply all three methodologies, is the Airbus A320 ATRA (Advanced Technology Research Aircraft). The present work comprised an airborne particle image velocimetry (PIV) measurement conducted as part of the second flight test campaign of the project HINVA. The specific task of the PIV system was the quantification of the outer strake vortex which was initiated by the outer vortex generator of the nacelle in the high-lift regime. PIV data at different Reynolds numbers and angles of attack was acquired and evaluated. This contribution summarizes the PIV flight test study by presenting its setup, realization and results.

Based on three experiments in wind and water tunnels, the challenges in quantifying the momentum coefficient are investigated and solutions to minimize the associated uncertainties are proposed. It is shown that including the viscous losses, determining the blowing slot height changes under loading conditions, as well as measuring the jet pressure is crucial for accurately quantifying the momentum coefficient.

The transonic buffet flow field around supercritical airfoils is dominated by self-sustained shock wave oscillations on the suction side of the wing. Current theories assume that this unsteadiness is driven by a feedback loop of disturbances in the flow field downstream of the shock wave of which the upstream propagating part is formed by acoustic waves. The present experimental study aims at analyzing the impact of acoustic disturbances on the shock wave movement during buffet. Therefore, the buffet flow field around a supercritical DRA 2303 airfoil model is influenced with artificial sound waves introduced downstream of the airfoil by a loudspeaker. The flow conditions are defined by the Mach number $$M_{\infty } = 0.73$$, the Reynolds number range based on the chord length of $$1.8 \times 10^6 \le Re_{c} \le 1.95 \times 10^6$$, and the angle of attack of $$\alpha = 3.5^{\circ }$$. High-speed particle-image velocimetry measurements are performed in the shock wave region. The results reveal that the unsteady flow field varies with the high-frequency pressure disturbances and their amplitude modulation. These data are considered as preliminary results which will serve for further studies concerning the investigation of buffet onset.

The microphone array method together with the beamforming technique, required for the post-processing of the recorded acoustic data, has been optimized for measurements on moving rail vehicles. In order to increase the spatial resolution along the moving vehicle a new type of microphone array was developed. This comprises a line array in combination with a two-dimensional elliptical acoustic mirror. Acoustic source were separated successfully using the properties of the elliptic mirror along the trajectory of the train by applying the beamforming methods for the vertical axis. The combination of both techniques allows a three-dimensional reconstruction of the source distribution. The design was successfully tested on a high speed train track in Germany, which is used by different type of Inter City Express and other passenger trains. It was possible to improve the quality of the results significantly and individual wheel sets and aeroacoustic noise caused by roof structures can be identified in the source maps. The resulting acoustic source data obtained with the microphone array technique, provides input data for noise prediction methods, which can fill gaps in existing noise source databases thus assisting in the goal for a deeper understanding of rail vehicle noise.

A low-cost setup for oil-film interferometry with white light was used to determine the flow topology and the wall shear stress on a delta wing at different angles of attack. Despite the simple setup, well reproducible and plausible comprehensive wall shear stress distributions on the upper side of the delta wing could be determined from experimental interference images. For the image evaluation, a new algorithm was proposed and developed, helping to determine automatically and quickly the local oil film thickness from a sequence of interference images.

The present paper describes an optical deformation metrology for scientific in-flight testing which was deployed to the Airbus A320 “ATRA” of the German Aerospace Center DLR. Within the national funded LuFo-IV/3 project “High-Lift In-Flight Validation” (HINVA) a multi-camera approach has been developed and tailored to the full-scale test bed in order to create a high fidelity wing deformation data set of dedicated dynamic and stabilized flight conditions up to maximum lift. Therefore, the wing configurations “cruise” and “landing” were measured with a variation of altitude and aircraft weight in the summer of 2012 at the Airbus facilities in Toulouse, France. The optical measurement technology is based on cross-correlation and photogrammetric principles. It enables to derive spatial 3D deformation and deflection from virtually reconstructed surfaces. After an introduction of the project and the metrology, the specific experimental set-up is described, followed by information about the test matrix. An exemplary analysis of the wing bending and twist is presented as well as measurement results of the trailing edge high-lift device with respective accuracy statements.

An overview is given of the novel Lagrangian particle tracking method ‘Shake-The-Box’, which allows particle tracking at high particle image density. The regularized interpolation of the discrete tracking information onto an Eulerian grid (FlowFit) shows an unprecedented spatial resolution. Application of the methods on two different experimental cases is briefly outlined.

An application of a hybrid RANS/LES method to trailing-edge noise, is presented in this contribution. The numerical framework of the CAA-code PIANO offers a basis for extending the Non-Linear Perturbation equations with viscous terms. As a result, a full Navier-Stokes equations perturbation analysis, denoted with “Overset”, can be performed on top of a steady background flow. With such a scale-resolving simulation tool, consisting of optimized higher-order numerical schemes for aeroacoustics, the investigation of sound source mechanism on first principals become feasible. An issue of such hybrid zonal approaches, namely the seeding of proper inflow turbulence, is solved with the Fast Random Particle-Mesh method in combination with the Eddy-Relaxation source term. The overall process chain of the intended Overset-LES usage is illustrated by means of NACA0012 trailing-edge noise computations at moderate Reynolds number and zero angle-of-attack.

Results from a numerical and experimental aeroacoustic assessment of 2D wind turbine blade sections are presented. CFD/CAA-based predictions using a synthetic turbulence method were conducted at a NACA 64-618 profile as well as at a new low-noise airfoil design RoH-W-18%c37. Validation experiments were performed in DLR’s Acoustic Wind-Tunnel Braunschweig (AWB) for varying transition locations. A trailing-edge noise reduction benefit of 2–4 dB in overall sound pressure level was predicted for the new airfoil under design conditions. A large laminar extent of the boundary layer significantly reduces the noise emission (by up to 8 dB) compared to equivalent cases with forced transition at the leading edge. An additional noise reduction (with realistic reductions of the peak levels by 4–6 dB) was accomplished by flow-permeable trailing-edge extensions which were successfully transferred to the two profiles from forerunner aerospace-related studies.

The paper presents results of unsteady Reynolds-averaged Navier-Stokes simulations (URANS), Delayed Detached Eddy Simulations (DDES) and Large Eddy Simulations (LES) of the flow and the noise propagation in a segment of the DLR’s cabin test facility Do728. Since the weakly compressible Navier-Stokes equations were solved in all cases, spectra of the sound pressure levels (SPL) were analyzed based on Fast Fourier Transforms of the predicted pressure fields. It was shown that LES on hexahedral meshes predict SPLs in good agreement with microphone measurements in the Do728 cabin. A comparison with predictions based on a hybrid approach involving the solution of a wave equation together with non-reflective boundary conditions on the one hand, and a Ffowcs Williams-Hawkings (FW-H) model on the other hand, revealed that the impact of the imperfect acoustic boundary condition in the LES on the sound pressure level at the receiver point is negligible. Thus, it was demonstrated that reliable predictions of SPLs in an aircraft cabin are possible if the corresponding LES can be performed.

This contribution presents a numerical study of jet noise installation effect for a generic configuration. The geometry definition consist of a single stream nozzle which is mounted closely under a NACA0012 wing. Such constellations produce significant interaction effects especially in terms of aeroacoustics. The applied methodology is a novel CAA approach which can be classified as a partly scale resolving simulation. In conclusion, the computed results are validated with measurements for a comparable constellation of a flat plate over a single stream nozzle with the same flow conditions. The validation of the CAA computation delivered a good agreement which indicates that all principal effects are numerically well reproduced.

Reduction of noise generated at geometric edges can be achieved by replacing solid material with porous inlays. The acoustic benefit for airfoil trailing edge noise was experimentally found to yield a reduction in sound pressure level of approximately 6 dB. Numerical methods are of interest to find optimal properties of the porosity. A successful method of modeling porous materials is the volume-averaging approach. Prior simulations have been enhanced by implementing a modified numerical treatment of the interface between the free fluid and the porous parts to model the interaction of the acoustic and flow quantities in these two regimes. Furthermore, numerical simulations have been run to show the influence of an anisotropic, respectively non-uniform porous trailing edge on the emitted sound.

Coaxial jets originating from a nozzle interacting with a two-element wing configuration consisting of a main wing and a flap are computed using large-eddy simulations (LES) to investigate in the long term the effect of the interaction on the sound field of this jet-wing-flap configuration. The secondary jet Mach number is $$M_s=0.37$$ and the Reynolds number based on the secondary velocity and diameter is $$Re_{s}=1.32 \times 10^6$$. The nozzle inlet boundary layers are modeled by Blasius velocity profiles with a boundary layer thicknesses of 20% of the nozzle channel half-width. The acoustic far-field is predicted by solving the linearized Euler equations (LEE) in a region attached to the LES domain. The jet streams interacts with the two-element wing configuration and lead to a significantly changed mean flow field in the wing-flap gap area and pressure at the wing surface. Additionally, a strong influence of the flap on the jet development is observed, especially in the case of a non-zero free stream velocity.

Helmholtz resonators play a key role as silencers in many technical applications. The aim of this work is to study the mechanism that governs the emission and reduction of noise. For the first time, we closely monitor the interaction between the acoustic field of a Helmholtz resonator’s geometry and a fully turbulent shearing flow by a Direct Numerical Simulation. The properties of a fully turbulent flat plate flow with and without a wall-mounted cavity are contrasted and compared to the Chase model.

The aerodynamic drag is the principal factor of the driving resistance of modern railway vehicles. To accurately predict the energy consumption, it is essential to estimate the drag that is fundamental for the development of new vehicles. One way to quantify the aerodynamic drag of a full-scale railway vehicle is to measure the wall shear stress since friction drag dominates. The goal of this study was to improve the prediction of the aerodynamic drag of railway vehicles by measuring the wall shear stress. Experiments were performed on a double-decker multiple-unit train KISS from Stadler to measure the wall shear stress under real operating conditions. Oil-film interferometry was used as a non-intrusive measurement technique with the setup installed inside the train to avoid flow disturbances induced by measuring probes. The obtained results are in good agreement with experimental full-scale data from the literature. The comparison to theoretical predictions based on pipe and flat plate experiments reveal clear deviations.

This paper presents the results of force measurements on a tandem configuration (platoon) of two train models (Type: Next Generation Train) at varying distances in heading direction in the water towing tank at the German Aerospace Center Göttingen (DLR). The trains had no mechanical connection to represent the occurring situation of virtual coupling. The experiments were conducted at a Reynolds number of 240,000 at 4 m/s. A ground plate was used to account for under floor effects. The gap distance was varied between 1.2% and 59.5% of the single train length. In comparison to a single vehicle, drag reductions of up to 30.4% were observed for the trailing vehicle at all distances. For the leading train reduced resistances at distances ≤12% were found. The combined drag of two vehicles is up to 16% less than the resistance of two single trains. With Particle Image Velocimetry (PIV) a strong interaction between both trains is observed at closest distance (1.2%). An area of reduced relative velocity and thus reduced stagnation pressure in front of the second train is found to be the reason for a smaller drag. At wider gaps (35.7% and 47.6%) a beneficial development of vorticity behind the first train is found to be the cause of a reduced drag value on the trailing vehicle.

Numerical computations using the Unsteady Reynolds Averaged Navier-Stokes (URANS) and Delayed Detached-Eddy Simulations (DDES) approaches are carried out to investigate the complex three-dimensional flow in the root region of a generic 10 MW wind turbine rotor. Preliminary studies regarding the time step size and the number of rotor revolution required for the time averaging procedure are conducted. In the blade outer region, URANS is sufficient to predict the general flow characteristics, but small discrepancies are observed in the blade root area where the flow is massively separated.

In the present study, a model turbine with 1.5 m rotor radius is investigated numerically by means of CFD (Computational Fluid Dynamics). A $$120^{\circ }$$-model of the turbine under free stream condition is created and a grid convergence study is performed. As the real model turbine will operate in a wind tunnel with high blockage ratio, the influence caused by limited space is analysed. Therefore, sectional loads along the blade radius, global loads as well as the axial induction and the effective angle of attack are investigated.

Based on the wind tunnel experiments of the Mexnext-III project, this paper presents TAU RANS simulations of the isolated MEXICO wind turbine rotor. Three test cases are considered featuring axial flow conditions at normal and stalled rotor operation. The RANS results are compared to PIV measurements along axial and radial traverses of the flow field and to pressure measurements at the rotor blade surface. The calculated rotor thrust and torque are compared to the experimental data as well. Laminar-turbulent boundary-layer transition near the rotor tip is accounted for in the RANS simulations. The influence of the predicted laminar flow on the rotor torque and power is studied by comparison to fully turbulent results.

The present study investigates the impact of different trailing edge flap parameters with regard to the application on wind turbine rotor blades. For this purpose 2D airfoil and 3D rotor simulations have been performed using the CFD code FLOWer. Trailing edge flaps are realized based on grid deformation. At first the flap shape meaning rigid or elastic types and the flap length is analyzed in 2D for a representative wind turbine airfoil. The 3D study investigates the effect of the radial position and extension along the blade span. For this purpose the DTU 10 MW reference wind turbine has been chosen. The present work provides an overview of the different aspects of the aerodynamic flap design.

This paper deals with a comparison of measurement and simulation data collected for a complex terrain in southern Germany. Moreover, a Delayed-Detached-Eddy-Simulations (DES) of a wind turbine in the same terrain is presented. The turbine chosen is a generic 2.4 MW wind turbine and a turbulent atmospheric boundary layer (ABL) is applied. The ABL is generated based on measurement data. The influence of complex terrain on the turbine performance is shown as well as its impact on the general flow field. An increase of loads can be noticed caused by the speed-up in the terrain. In addition, the upward flow causes non-axial inflow angles and leads to a deflection of the turbine wake.