ABSTRACT. This talk will present kinetic energy and entropy preserving (KEEP) schemes, non-dissipative and stable numerical schemes developed for high-fidelity computations of shock-free compressible flows. Our research group recently succeeded in conducting LES of the flow around whole aircraft configurations, using the KEEP schemes and an LES wall model. In general, flow computations become unstable when non-dissipative numerical schemes are used. The KEEP schemes, however, have performed stable computations of compressible flows without the aid of artificial viscosity, thanks to their superior kinetic-energy and entropy preserving property. In addition to the KEEP schemes, recent research works from our research group will also be briefly presented.

ABSTRACT. High aspect ratio wings are essential for solar powered unmanned aerial vehicles as it satisfies the requirements of long endurance and high lift-to-drag ratio. These wings are structurally very flexible leading to large deformations and associated aeroelastic problems. In the present work, fluid-structure-interaction (FSI) analysis of a high aspect ratio wing is conducted at low flow velocities using coupled Computational Structural Dynamics (CSD) - Computational Fluid Dynamics (CFD) approach. Both tip displacements and twist angles of the wing computed from the present approach are compared with the experimental and numerical results available in the literature. Further, the effect of wing deformation on the aerodynamic coefficients of the high aspect ratio wing is also discussed. It is observed that the wing deformations computed using the present CSD-CFD approach are in good agreement with the available experimental/numerical results. It is also observed that wing deformation has significant effect on the lift coefficient of the high aspect ratio wing at higher flow velocity.

ABSTRACT. In order to decrease the cost per kg for rocket launches, it is essential to recover the rocket stage, including strapons and boosters. There are several methods to achieve this recovery, one of which involves the use of a parachute to decelerate from higher to lower Mach numbers. Subsequently, a retro thruster is employed to reduce the velocity of the touchdown to near 0 m/s. Finally, floats are used to facilitate a gentle landing on water (PRF) [1]. This study focuses primarily on the final phase of recovery, namely the use of floats. Computational simulations were conducted to examine the impact of various stage-float systems on calm water from a height of 0.1 m and at a velocity of 10 m/s. A transient analysis was performed using the Volume of Fluid (VOF) method and a deforming mesh algorithm was utilized to move the model within the computational domain. The methodology was validated using experimental data from a wedge-drop test [2], and a strong correlation was established between the experimental and CFD data. The results include displacement, deceleration of the system, and impact pressure experienced by the float for all the models and are presented and compared in this paper.

ABSTRACT. Numerical simulations of flapping flight in this paper are inspired by birds and insects. Wings of birds and insects are flexible in nature, which leads to interaction of surrounding fluid and the wing structure. Numerical simulation of rigid flapping airfoil/wing has been carried out by various researchers. However, very limited studies are only available considering the flexible nature of the wing. Flexible wing flapping flight problems can be studied using Fluid Structure Interaction (FSI) simulations with suitable coupling of CFD and structural dynamics. Understanding the role of flapping wing kinematic parameters and deformation may provide the means to reduce the power requirement and improved maneuverability of flapping flights. Numerical simulations are performed using commercially available ANSYS CFD and Mechanical Enterprise software. Objective of this paper is to study the flow physics associated with the flapping wing and the effect of flexibility of flapping wing on aerodynamics forces. Towards this a 2D rigid flapping airfoil simulation was carried out and the results are presented. Preliminary results of FSI simulation of flow past a circular cylinder with flexible flap is carried out. Numerical simulation of flexible flapping wing is under progress, results will be presented in final paper

ABSTRACT. Transonic buffet is an unsteady flow phenomenon characterized by the presence of shock oscillations and flow separation. The cause of many of the launch vehicle failures is attributed to buffet phenomena in the transonic flow regime. In the present work, a computational procedure is established to estimate the buffet forcing function on a rigid launch vehicle using Detached Eddy Simulation (DES) approach. A hammerhead launch vehicle configuration reported in the literature is considered for the study. The time average (mean) and unsteady pressure fluctuations (RMS) are measured at various stations along the length of the vehicle. The mean and RMS values of Cp are also compared with the experimental and numerical results available in the literature. Further, the buffet forcing functions in terms of mean and RMS of side and normal forces acting along the length of the rigid launch vehicle are also discussed.

ABSTRACT. The present work focuses on computational study of two-way Fluid Structure Interaction (FSI) on a N1G missile fin. At high speeds, the missile experiences high thermal and pressure loading due to the existence of shocks, the interaction of the shocks causes large variation in the pressure loads and vibration of the fin. This inturn, affects the fluid flow around the fin. This fluid flow and structural interaction leads to cyclic change of stresses on the fin causes fatigue failure and also affects the stability of the missile body. Interaction between the aerodynamic forces and elastic forces in a structure causes aero-elastic problems. This can be predicted by solving Fluid Structure Interaction (FSI) analysis. The computations are carried out using a commercially available ANSYS solver. Aerodynamic forces acting on the fin are determined by solving a Reynolds-Averaged Navier-Stokes (RANS) based solver and the effect of aerodynamics of the fin on the structure is determined by the use of a Static Structural solver. Structural and fluid flow computational analysis shows good agreement with the experimental data. Further, to determine the effects of variable conditions, two free stream conditions are considered.

ABSTRACT. Effects of the swirl in the self-preservation region of turbulent round jets were studied using large eddy simulation for swirl numbers S = 0.3, 0.5, 0.7 (defined as a ratio of axial fluxes of azimuthal and axial momentum at the inflow plane) at a Reynolds number of 11,000, based on inflow velocity and jet diameter. Flow quickly achieved self-similarity for the mean axial velocity, while radial and azimuthal mean velocities reached it further downstream. With increasing swirl, the jet spreading angle increases, and correspondingly, the decay of the maximum velocity along the length of the jet is faster. The reciprocal of the centerline velocity and jet width grows linearly with streamwise distance with and without swirl, as noted in experiments; axial fluxes of linear and angular momentum were nearly invariant. Surprisingly, radial profiles of mean velocity and fluctuations, scaled with local centerline velocity, were the same independent of swirl.

ABSTRACT. In this work, we characterize the dissipation structures in a turbulent channel flow using ‘Minkowski Functionals’. These are geometrical descriptors of the structures. The structures are educed using a bandpass-filter-based multiscale method. For a three-dimensional structure, there are four Minkowski functionals, from which two ratios, ‘Planarity’ and ’Filamentarity’, are derived, which are then used to classify these structures. In the buffer region, the dissipation structures have a pancake-like structure. In the log-region and wake-region, the structures observed are blobs, short tubes, and long tubes. As per the proposed classification, no sheet or ribbon-like structures are observed for dissipation. Such characterization of dissipation structures can help us get a better understanding of dissipation range phenomena such as small-scale intermittency.

ABSTRACT. The study involves a direct numerical simulation (DNS) of uniform flow past a thin normal flat plate. The investigation focuses on the influence of the aspect ratio Ly/h (where Ly is the spanwise length of the plate and h is the plate width), on the wake transition to three-dimensionality that occurs behind the plate. To achieve three-dimensionality, the flow Reynolds number Reh (based on uniform inflow velocity, U0 and h) was varied from 130 to 200. The study includes results from simulations conducted with two spanwise extensions, specifically Ly/h = 6 and 12. By analyzing the force coefficients obtained at different Reynolds numbers and visualizing the three-dimensional (3D) flow structures in the wake, it was possible to compare the effects of these two spanwise extensions. For Reh = 175, a spanwise length of 12h is necessary to eliminate numerical effects observed at a spanwise length of 6h. A lower aspect ratio was able to capture the flow features correctly at Reh = 200. However, at Reh = 130 and 150, the aspect ratios used were insufficient to capture the mean quantities, which implies that a higher spanwise length is needed.

ABSTRACT. Direct numerical simulations (DNSs) have been carried out for fully developed turbulent flows through ribbed channels, where square ribs were arranged transversely to the flow on one of the channel walls. Based on the spanwise extent of the ribs, two configurations are considered ---two-dimensional (2D) configuration resulting from full-span-width ribs and a three-dimensional (3D) configuration, where the ribs extend only up to half the span-width of the channel, leaving the other half smooth. The 3D configuration thus produced a unique problem of coflowing rough and smooth channel flows. A striking phenomenon has been observed of the secondary roll-cells, exhibiting a strong updraft in the smooth half of the channel. Comparisons were also drawn with DNS of a smooth channel at the same friction Reynolds number 400 and it was found that the roll-cells on the smooth half not only affect the bulk flow negatively, but also attenuate the turbulence significantly. In spite of having a higher bulk velocity, the rough half of the 3D configuration has been found to be more turbulent than the 2D configuration; this has been attributed to the momentum transfer from the smooth half to the rough half.

ABSTRACT. Numerical simulation of the Natural Laminar Flow~(NLF) wing at $Re=6 \times 10^6$ is carried out using the incompressible flow solution code 3D-PURLES developed in-house. The RANS based $\gamma$-$Re_\theta$ transition model is effectively used to understand the transition and aerodynamic behaviour of this wing by comparing with the fully turbulent SST model. The transition model has captured the laminar region and predicted a lower drag coefficient compared to the SST model. The transition behaviour of the wing is also compared with root section airfoil of this wing and it was found that the extent of laminar region and the reduction in the drag count for the wing was found to be much lesser than that obtained for the airfoil.This analysis indicates that the three-dimensional flow significantly effects the transition characteristics and also justifies the need of modelling the transition phenomenon appropriately

ABSTRACT. Numerical analysis using Direct Simulation Monte Carlo (DSMC) simulations is used to study the flow characteristics related to hypersonic flow over a cone-shaped body during atmospheric entry. The cone is created by varying the angle of extrusion ($\alpha$) of a flat-nosed-faced cylinder. The results of the distribution of surface heat transfer and drag coefficient on each negative $\alpha$ are contrasted against the results obtained for zero and positive $\alpha$ for which compressible flow physics are well defined. This study has helped to improve the understanding of aerothermodynamic load on a cone-shaped space object in a rarefied flow environment.

ABSTRACT. Aerodynamic coefficients of a launch vehicle are a necessary input to trajectory design as well as simulations and control system design. There are various ways to generate this data and they include empirical/ engineering methods, CFD methods and wind tunnel tests. With the advances in grid generation techniques, computing facility and post processing capabilities, CFD codes are used routinely for aerodynamic characterization of a launch vehicle. In this study, a commercial CFD solver CFD++ has been utilized for flow field simulations over quad dominated body fitted mesh using Pointwise of an ISRO launch vehicle without strapons. Comparison of the results with other CFD codes as well as validation with NAL wind tunnel test results have been carried out for Mach numbers ranging from subsonic to supersonic Mach numbers. The results indicate good confidence in using CFD++ solver as a design tool for aerodynamic characterization of expendable launch vehicle.

ABSTRACT. This paper investigates the aerodynamics of combusted plume gases in RCS thruster nozzles with extensions in a typical crew module firing at three identified altitudes (at Za = 30km, 14km and 7km) using CFD. Two thruster configurations of different nozzle extension lengths (Le = De and 3De) are examined for flow separation and effective thrust, and inferences are drawn based on flow dynamics. Whereas nozzles with Le = De were able to generate over 95% of the expected thrust at 30 and 14km, for Le = 3De, it drops to about 92% at 30km and 90% at 14km. At 7km, the thruster with Le = De showed a thrust reduction of around 12%, and it dropped by a remarkable 45% for Le = 3De. Such decrease is attributed to scarfing of the nozzles, as well as boundary layer separation within nozzle extensions, which depends on the ratio of thruster combustion chamber pressure to ambient pressure. Indeed, predicting flow separation characteristics is highly sensitive to the grid cell size used to resolve sharp gradients in the flow domain, and the turbulence modeling strategy. This paper will present a grid independence study using AMR, and comparisons of BL separation location and the size of recirculation bubble from using RNG k-epsilon, k-omega SST and the SBES turbulence models.

ABSTRACT. This study focuses on the cooling mechanisms employed in liquid rocket engines to prevent damage from high temperatures. Spray cooling and the complexities involved in droplet impact on heated surfaces are explored. A two-dimensional computational model for analysing the behaviour of liquid methane droplets on a heated steel plate is presented. Liquid methane's hydrodynamic and thermodynamic properties are investigated at different droplet velocities and surface temperatures. The spreading ratio and Weber number are calculated to characterise droplet behaviour. Further, the wall surface temperature and droplet-breaking effects are also investigated through the volume of fluid (VOF) method.

ABSTRACT. The current work is an exploratory study on the Partially Rigidized parachute, aiming to build Aerodynamic and Aerothermal models of PRP. Initially we determined a domain of geometric parameters where all PRP configurations exhibit a fairly study drag. CFD studies are carried out to identify significant design pressure coefficients and the peak heat flux experienced by the PRP. These drag and thermal characteristics are majorly influenced by the geometric parameters (X_geo), as well as the Mach number (M) and Reynolds number (Re). The six-dimensional geometric design space is established based on steady configurations, while the flow variables M and Re (or ’h’) are selected in accordance with the potential trajectory corridor for a given forebody. Subsequently, this eight-variable input design space is sampled and a dataset is constructed using computational fluid dynamics (CFD). Finally, Artificial Neural Networks (ANN)-based surrogate models are developed for the aerodynamic and aerothermal modules using this dataset.

ABSTRACT. Large-eddy simulations of fully compressible turbulent mixing layers are carried out at convective Mach numbers up to 2.0. The focus is on studying the effects of a localised artificial diffusivity (LAD) shock-capturing scheme and three state-of-the-art, constant-coefficient and dynamic subgrid-scale (SGS) models. High-order compact schemes, suitable for resolving small-scale turbulent motions, are used as the underlying numerical method. The LES results are compared to previously reported direct numerical simulation (DNS) results of the same configurations. The results demonstrate that it is critical to include the LAD scheme and the SGS models to stabilise the simulations at high convective Mach numbers and that incorporating dynamic SGS models leads to better agreement with the DNS results compared to the constant-coefficient versions of the SGS models.

ABSTRACT. This paper presents numerical simulations of transonic flow over Sandia axisymmetric hump model using the code HiFUN. Present configuration consists of transonic flow past an axi-symmetric hump for a reference Mach number, M∞ = 0.875 and a chord-based Reynolds number, Rec ≈ 1 million. The oncoming turbulent flow becomes locally supersonic on the hump, leads to formation of weak shock that imposes an adverse pressure gradient onto the incoming boundary layer. This shock-boundary layer interaction causes it to separate, forming a separation bubble downstream of the hump which later undergoes reattachment. This shock-separated flow phenomenon is crucial for flow over wings with transonic flow and many other industrial applications and present a particular challenge to numerically captures it accurately. Transonic flows are especially difficult to predict because of strong viscous-inviscid interaction at such speeds. To begin with, detailed comparative study of flow behavior is studied by performing steady simulations using standard turbulence models, SA and k − ω SST. Present computations capture the main flow features such as pressure and skin friction coefficient, shock location, separation-reattachment points and are in reasonably good agreement with experimental data. However, the pressure and skin friction coefficient distribution on model surface were found to show azimuthal asymmetry in the flow. This may be attributed to three-dimensional effects if any or due to flow unsteadiness which might be revealed with Unsteady RANS and DES calculations and will be added in the final paper.

ABSTRACT. While RANS calculations are performed routinely as part of aircraft design, for a few conditions, a higher-fidelity simulations would be desirable. Performing Hybrid RANS-LES over a full aircraft con- figuration remains a challenge for CFD teams. However, some progress has been made in this regard and in this paper we show the efforts done in this direction and speculate on the time-scale when RANS-LES can be used for design support. We show the suitability of RANS-LES for flows of increasing level of complexity, starting with aerofoil, rectangular wing, Rudimentary landing gear (RLG) and finally some numerical experiments on full Saras aircraft and make projections on the way forward.

ABSTRACT. Numerical simulations are carried to assess the efficacy of optimal perturbation in inducing instabilities in the wake of a finite aspect ratio wing. Optimal perturbation is periodically injected into the flow and the subsequent flow evolution is monitored. A perturbation model for the wing-tip vortex is proposed based on the deformation of vortex in numerical simulation. The model is then used to investigate the long time evolution of vortex pair. The effect of frequency of injection, gap between the vortex pair, Reynolds number and initial energy of the perturbation is studied. It is observed that for large gap, perturbation decays with time. On the other hand low gap results in significantly slower decay. Increase in circulation(Γ) promotes the growth of instability.

ABSTRACT. The impingement of an initial vortex dipole with a T-shaped cavity in a bounded domain is studied numerically using the lattice Boltzmann method with the Bhatnagar-Gross-Krook collision model. The effect of various cavity widths ($w$) and Reynolds numbers (Re) are considered for the analysis. For the case of $w=1.0 d$, $d$ being the dipole diameter, the initial dipole descends into the cavity almost intact till it hits the bottom when a strong interaction with induced vorticity results in formation of a secondary dipole which ascends out of the cavity. When $w=0.75d$, the initial dipole also manages to enter the cavity, except that it loses some of the vorticity to secondary vortices that are formed at the convex corners of the cavity entry. When $w=0.625d$, however, the initial dipole does not descend into the cavity. Instead, strong interaction with the secondary vortices at the cavity entry forms two new dipoles with a pair exchange mechanism. These dipoles ascend and soon collide owing to their axes inclined towards each other. Larger ingestion of vorticity in the case of $w=0.5d$. When $w=0.5d$ as well, the initial dipole strongly interacts with the secondary vortices at the cavity entry and forms two new dipoles. However, these dipoles move away from each other and reach the top boundary through a parabolic trajectory. An essential feature being noticed is a continual reorganization of vorticity, forming new dipoles that continue to interact with solid boundaries.

Normalized enstrophy ($\Omega$) and palinstrophy ($P$) show distinct peaks during the dipole interactions with corners and flat boundaries, influencing enhanced kinetic energy decay. Flow evolution satisfies energy and enstrophy relation valid for no-slip boundaries. The peaks that appear during the corner interactions suggest a possible scaling with respect to the Re. The sharp cavity corners modify the scaling results of $\Omega_{max}$ and $P_{max}$ with relatively higher exponent values than dipole interactions with flat boundaries. Scaling exponents increases with decreasing cavity width. Overall, the positive and negative lobes of every dipole tend to retain distinct identities for $w=0.625d$ and $w=0.5d$ and can be traced back to their respective origins in the course of evolution.

ABSTRACT. The effect of compressibility on vortex breakdown topology in Vogel-Escudier (VE) flow is examined here. VE flows inside a cylinder created by one of the lid rotating. The numerical simulation of the Navier-Stokes equation using the finite volume method using OpenFOAM modules is done in this work. The flow is simulated for Reynolds numbers (Re) of 1000, 2200, 2700, 3400, and 5000. The effect of variation of Mach number (Ma) in the subsonic regime is discussed. The aspect ratio of the cylinder is fixed to Γ = 2.5. The compressibility is shown to quench the non-axisymmetric modes of perturbations. At higher Reynolds numbers, the progression from non-axisymmetric to axisymmetric flow with an increase in Ma is evident.

ABSTRACT. The vortex ring is usually modeled in computational studies involving vortex ring and shock wave interaction. However, the incompressibility assumption of the isolated vortex ring model usually limits its applicability to lower Mach numbers. To overcome this limitation, our approach is to generate a compressible vortex ring from the jet exiting the nozzle, and analyse its interaction with a shock wave. The near-field deformation and sound generation is simulated by solving the axisymmetric Navier-Stokes equations and compared with ring vortex-shock wave interaction using analytically defined vortex ring model.

ABSTRACT. The impact of Prandtl number on the evolution of a buoyant vortex dipole is examined in this paper. The numerical simulations are conducted using a pseudo-spectral approach and the dipole is modeled by a Lamb Dipole initialized similarly to previous studies. The time evolution of a buoyant vortex dipole is studied for various values of Pr keeping Re and Ra constant in the exercise. The results show that increasing the Prandtl number leads to a decrease in the propulsive velocity of the vortex dipole.