Data-driven analysis of Intermittency in Compressible Turbulent Flows

Faculty: Prof. Konduri Aditya (CDS) and Prof. Rishita Das (AE)

Turbulence in high-speed compressible flows remains one of the central challenges in aerospace engineering. Intermittency characterized by the occurrence of rare and extreme events is a defining feature of compressible turbulence with critical consequences for high-speed aerodynamics, hypersonic boundary layers, supersonic combustion and autoignition. These events can strongly influence mixing, heat release, drag, and the stability of high-speed propulsion systems. Understanding and predicting such events is therefore crucial but remains a major challenge, particularly in compressible turbulence. Rishita Das’s research group at the Department of Aerospace Engineering, IISc, focuses on turbulence physics and modeling. Their recent work quantifies turbulence intermittency using data-driven information theoretic methods and forecasts extreme occurrences using a dynamical system approach. Konduri Aditya’s research lab at the Department of Computational and Data Sciences, IISc, develops scalable algorithms for direct numerical simulation of compressible flows on massively parallel architectures and formulate anomaly detection methods to identify extreme events in high-dimensional, multi-scale systems. Through a collaborative effort between these groups, this project aims to develop a comprehensive understanding of intermittency and model the mechanisms that govern the occurrence of extreme events in compressible turbulence.

The physics of compressible turbulent flows differs fundamentally from that of incompressible turbulence due to the presence of additional mechanisms such as density fluctuations, dilatational motion, and localized shock structures known as shocklets. These features affect the dynamics of the extreme events and may lead to enhanced intermittency. While intermittency and scaling laws in incompressible turbulence have been extensively studied over the past several decades, our understanding of compressible turbulence intermittency remains limited, particularly in regimes where compressibility effects become significant. In this project, we will investigate these phenomena using high-resolution direct numerical simulations of canonical flows representing compressible turbulence. The turbulent flow will be analysed using a range of data-driven approaches, including information-theoretic measures, modal decomposition, and moment-based statistical characterization of turbulent fluctuations. Through these analyses, the project aims to develop comprehensive understanding and models of intermittency in compressible turbulence, characterize shocklet-driven extreme events, and develop frameworks for detecting and predicting such extreme occurrences in the flow.

The PhD student should have a strong background in mathematics, statistics, and fluid dynamics, and an aptitude for data science and scientific computing.

Learning Outcomes G Skills from PhD project:

1. Strong foundation in data science
2. Rigorous training in high performance parallel computing
3. Expertise in information theory, statistical methods, machine learning
4. Expertise in fluid dynamics and turbulence

Relevant Publications:

1. Sarkar, S. and Das, R., 2025. Information-theoretic characterization of turbulence intermittency. arXiv preprint arXiv:2505.05304.
2. Aditya, K., Kolla, H., Kegelmeyer, W.P., Shead, T.M., Ling, J. and Davis IV, W.L., 2019. Anomaly detection in scientific data using joint statistical moments. Journal of Computational Physics, 387, pp.522-538.
3. Wang, J., Gotoh, T. and Watanabe, T., 2017. Scaling and intermittency in compressible isotropic turbulence. Physical Review Fluids, 2(5), p.053401.