Collaborative Research Work

A regular cavity is mounted on a test model as shown in the left-side image. If the cavity inset is solid then it is called as baseline model. Otherwise with the microstructure inset, it is called as control model. High-speed schlieren images are taken for these two cases near a Mach number of 0.81. These images are decomposed using the SPOD technique and reconstructed only for a selected frequency of about 5.1 kHz as shown in the middle part of the image. The reconstructed schlieren images for 5.1 kHz across the spatial modes are played in the right-side image series showing a noticeable attenuation of acoustic wave radiation due to microstructure inset.

Passive control of transonic cavity oscillations using acoustic micro-structures | (IIT Madras, India)

Open-cavity flows at high-subsonic and transonic speeds generate strong pressure oscillations through coupling between shear-layer motion and the cavity’s internal acoustic field. In collaboration with Dr. R. Sriram and Dr. E. Hemaprabha at IIT Madras, this study investigates passive control using periodic acoustic micro-structures placed at the leading wall of a cavity with a length-to-depth ratio of two. The lattice inserts are designed to absorb acoustic waves near the dominant Rossiter frequencies, weakening wall reflections and the feedback that sustains oscillations. Wind-tunnel experiments spanning freestream Mach numbers of 0.60–0.86 combine unsteady pressure measurements, time-resolved schlieren imaging, and SPOD to assess the control mechanism, supported by preliminary numerical simulations. The experiments demonstrate suppression of the first and third Rossiter modes and a maximum reduction of 58.9% in the second-mode pressure spectral peak inside the cavity. Normalised RMS pressure fluctuations decrease by up to 47.6% at the cavity trailing wall. These findings establish acoustic micro-structures as a promising passive approach for mitigating cavity-induced pressure loading and noise in aerospace applications.

Keywords: transonic cavity flow, passive flow control, acoustic micro-structures, acoustic absorption

(a) schematic of the scramjet engine during the throttling stage, highlighting key aerodynamic features; (b) schematic of the model used for experiments; (c) High-speed time-resolved Z-type schlieren imagery (𝜕𝜌 ̅/𝜕𝑥) captured at 73,000 FPS showcasing the generation of unsteadiness over the cavity area; (d) High-speed time-resolved PLRS imaging captured at 73,000 FPS at the centreline of the model showcasing the generation of vortical structures and its subsequent breakdown downstream.

Hypersonic flow past complex cavity sub cavity configurations | (Technion IIT, Israel)

The hypersonic Ludwieg tunnel at the Technion Wind Tunnel Complex is used to investigate the unsteady flow physics of a cavity–sub-cavity configuration relevant to integrated scramjet–launch vehicle systems along with Prof. J Cohen at Technion. The model recreates the essential features of the actual interface, with a primary cavity representing the Single Expansion Ramp Nozzle (SERN) and an embedded deep sub-cavity representing the engine isolator. Under hypersonic flow, the shear layer separates at the cavity leading edge and interacts with shocks and the sub-cavity, driving high-frequency pressure oscillations and acoustic feedback loops. High-speed unsteady pressure transducers capture surface pressure fluctuations. Schlieren imaging resolves shock–shear layer interactions, and Planar Laser Rayleigh Scattering (PLRS) tracks the evolution of coherent vortical structures and shear-layer dynamics. Together, these diagnostics are used to identify the dominant oscillation modes, isolate the mechanisms behind self-sustained oscillations, and evaluate how effectively the model's geometry suppresses acoustic loads. The experiments also provide a reliable dataset for understanding how cavity and sub-cavity interactions change at hypersonic speeds. Such studies matter because cavity oscillations can cause severe vibrations, structural fatigue, and performance losses, and understanding them supports the design of safer, quieter hypersonic vehicles.

Keywords: hypersonic flow, cavity–sub-cavity flow, Ludwieg tunnel, schlieren imaging, PLRS imaging

Comparison of raw and cleaned schlieren images for non-stationary flows (left) and stationary flows (right), illustrating the removal of localized optical contamination that obscures experimental observations. The red circles identify examples of artifacts in the raw images, while the corresponding cleaned images reveal the flow field in these regions. The arrows indicate the flow direction. A central challenge is to distinguish contamination fixed in the imaging system from genuine physical features that may also remain stationary, such as standing shocks. This comparison motivates our collaborative research on physics-preserving AI restoration, which aims to combine information from neighbouring frames, motion estimation, and imaging calibration to recover obscured regions while retaining physically meaningful structures. The proposed framework will also assess reconstruction uncertainty and flag ambiguous regions for expert review, supporting reliable interpretation and quantitative analysis of high-speed scientific images.

Physics-preserving AI restoration of high-speed scientific images | (IIT Madras, India)

High-speed scientific imaging captures rapidly evolving phenomena in gas dynamics, ballistics, combustion, and impact experiments. However, contamination on optical windows, lenses, and mirrors, together with sensor artifacts, can obscure shock waves, interfaces, projectiles, and particle fields, limiting the reliability of experimental measurements. In collaboration with Dr. Ram Prabhakar from the Department of Data Science and AI at IIT Madras, this research aims to develop a physics-preserving, uncertainty-aware AI framework for restoring contaminated image sequences. The proposed approach combines information from neighbouring frames, motion estimation, imaging calibration, and application-specific physical constraints to identify artifacts and reconstruct obscured regions. Particular attention is given to distinguishing contamination from genuine stationary features, such as standing shocks, and preserving reliable image regions. Restoration uncertainty and physical-consistency checks will identify ambiguous cases requiring expert review. Validation through controlled experiments and quantitative measurements will target applications in schlieren imaging, ballistics, and particle image velocimetry (PIV), supporting more reliable interpretation of high-speed experimental data.

Keywords: high-speed scientific imaging, physics-preserving AI, optical contamination, video restoration, uncertainty quantification

The underlying fluid-dynamic mechanisms and wave patterns are clearly highlighted by the model schematic and experimental Schlieren visualization. In the baseline configuration, the incoming supersonic stream separates at the cavity leading edge, creating a shear layer that spans the cavity and generates a low-pressure recirculation bubble inside. Intense vortex shedding and continuous pressure wave propagation between the leading and trailing edges setup a strong fluid-resonant state, visible through pronounced pressure peaks and acoustic waves reflecting within the domain. Conversely, the introduction of ventilation holes at the sub-cavity base acts as a bleed mechanism, allowing a portion of the disturbance energy and mass to escape. This passive release alters the internal wave-propagation pattern, converting strong, coherent pressure waves into significantly weaker Mach waves and suppressing acoustic feedback to yield a substantially cleaner, more stable flow field.

Experimental study for supersonic cavity flow over a deep sub-cavity
and passive ventilation control | (Kyushu University, Japan)

Supersonic cavity flows are inherently characterized by complex self-sustaining pressure oscillations, which arise from the strong feedback mechanism between shear-layer instabilities and acoustic resonance. To address the resulting aeroacoustic loading along with Prof. Hideaki Ogawa at KU, this study investigates the flow dynamics over a deep sub-cavity and evaluates passive ventilation through a surface hole as an effective control strategy. Experiments were conducted in collaboration with JICA Friendship 2.0 at Kyushu University, Japan, using a planar convergent-divergent nozzle in a supersonic wind tunnel operating at a freestream Mach number of M_∞=2.5, total pressure p_0=450" kPa" , and total temperature T_0=290" K" . Wall-static pressure measurements were acquired across seven strategic locations covering the sub-cavity, main cavity, and aft wall to quantify variations in the flow-field dynamics. The findings demonstrate that passive sub-cavity ventilation effectively disrupts acoustic feedback and weakens wave coherence, offering a simple, passive solution for mitigation in scramjet flameholders, weapon bays, and launch vehicle interfaces.

Keywords: weapon bays, scramjet flameholders, launch vehicle interfaces

Page updated

Google Sites

Report abuse