Research

Coming soon...

Coming soon...

Extreme weather events bring huge losses to human and animal life, settlement and agriculture. With a global increase in carbon emissions and unsustainable agriculture practices, we increase the rate at which anthropogenic activities interact and influence the environment around us. For an improved understanding of the sensitivities of the environment and its predictability, many different groups in our department focus on the study of its dynamic parts -- the ocean and the atmosphere, two of its crucial components. Both of these systems are extremely complex to represent entirely in a computer simulation or a laboratory experiment. Hence, we try to understand simpler representative model problems that can include parameterizing many of its small-scale dynamics. For this, we use a combination of realistic or idealised numerical simulations, theoretical tools, laboratory-scale experiments, and observational data. 

Given the large Reynolds number, the atmosphere and the ocean are naturally turbulent. They are also stratified in temperature, density, and salinity which affect the way the turbulent mixing process disperses momentum and scalar. Different groups in our department are collectively working towards understanding stratified turbulence in the atmosphere and the ocean using stratified tank experiments, direct numerical simulations and Large-Eddy simulations of the Navier—Stokes equations with particles and droplets, simulations of the Reynolds-Averaged Navier—Stokes equations, and improving mixing parameterizations used in ocean and atmospheric operational models. Researchers are also working to understand the internal gravity wave spectra of stratified oceans and the instabilities associated with nonlinear gravity waves and their subsequent breakdown.

Some of us are interested in developing technologies for ocean exploration, like underwater remotely-operated vehicles. Studies are ongoing to understand the fluid dynamics of air curtains and their effectiveness in the presence of human movement and flash flood prediction using shallow water equations. Some recent works have looked at the spatiotemporal distribution of CO₂ in a human environment with implications for improving indoor air quality. Researchers in our department are also investigating droplet formation and growth in clouds which are a crucial component of the atmosphere and a big uncertainty in climate models; this holds a strong prospect of improved understanding of large-scale dynamical systems like the Indian monsoon. There are investigations on the turbulent dispersion of atmospheric pollutants and the thermodynamics of turbulent plumes. Some groups are working on the fundamental understanding of fluid-structure interactions which holds immense prospects in different fields, e.g., in improved design of dams and ocean-exploration systems like that used in deep-sea mining, in understanding wave interaction with off-shore structures, in designing efficient aerodynamic vehicles and economical and weather-proof structures, and in enhancing impact predictability of natural calamities etc.

Research groups focus into the understanding of fluid instabilities and turbulence. We investigate the transition in boundary layer flows and other shear flows, exploring how laminar flows evolve into turbulence under various conditions. Our studies also encompass the transition in elastic flows, examining both inertial and non-inertial effects to understand the intricate dynamics at play. We utilise both modal and non-modal instabilities for studying transition in above flows including viscosity-stratified flows. By analysing isotropic turbulence, we aim to uncover the universal characteristics of turbulent flows, providing insights into their chaotic nature. 

Additionally, we focus on the transition in thin film flows, where surface tension and viscosity play critical roles. Our mathematical modelling efforts are towards understanding the effects of surfactants and viscoelasticity on transitional flows, offering a theoretical foundation for practical applications. We also investigate turbulent pipe and channel flows, seeking effective control strategies to manage turbulence. Our work on the physics of turbulence drag reduction aims to develop methods to minimize energy losses, contributing to more efficient fluid systems. Together, these diverse research areas drive forward our understanding and control of fluid instabilities and turbulence.

Further, we combine mathematical modelling and computation for studying turbulent flows and heat transfer using LES, PANS and RANS employing various commercial CFD packages, OpenFOAM and developing in-house source codes in several flow configurations. We also use the averaged kinetic models to understand various fundamental phenomena such as thermal transport, mass transport, energy exchange among scales, nonlinear wave propagation, and applicative phenomena such as flow turbulence, nonlinear waves and flow stability. We leverage high-performance computing techniques to scale these models and understandings to aid technological development including high-speed vehicles and renewable energy harvesting devices. We also work on various turbulence models for modelling turbulent flows behind wind turbine, multiphase flows and clouds.

Researchers in the field of data-driven mechanics focus on development and application of novel scientific machine learning (Sci-ML) algorithms to enhance the analysis and design of mechanical systems. Key areas of study include predictive modeling of material behavior, optimization of mechanical structures, real-time monitoring and diagnostics, and the development of smart materials and systems. This research also explores the application of data-driven methods in complex systems such as aerospace structures and biomechanical systems. By leveraging vast datasets, physics models developed over hundreds of years, and advanced computational tools, data-driven mechanics aims to optimize performance, improve reliability, and foster innovation in all aspects of mechanics related problems.

Some Key research areas

Research Area Faculty

Research in the field multifunctional materials and structures is focused on design and analysis of engineered structures such as light weight composites, smart materials and intelligent devices that have the ability to carry out multiple functionalities with enhanced efficiency. These structures have utility in aerospace and naval engineering, robotics, biomedical devices, automobile engineering etc. The overarching goal is to enable lighter, stronger, and more intelligent structures with adaptability, enhanced durability and sustainability. Ongoing research in the department of Applied Mechanics can be broadly classified into the following categories

• Smart Materials and Structures: Smart materials such as piezoelectric materials, magnetostrictive materials, shape memory alloys etc. deform in response to external stimulus such as electromagnetic fields or temperature change. Research endeavors are focused on experimental, theoretical and computational analysis of a wide range of smart materials including (but not limited to) piezoelectric materials, magnetostrictive materials, electro- and magneto-active polymers, flexoelectric materials. 

• Lightweight and High-Strength Composite Structures: Lightweight, architected composite structures have wide range of applications in design of ship, aerospace and automobile structures. Active research areas in the department include design and modeling of lightweight, high-strength composite structures such as fiber reinforced composites, Honeycomb structures, Auxetic structures, Nanomaterials and Metamaterials. 

• Intelligent Device Design: Multifunctional materials have applications in the design of the next generation of actuators, sensors and transducer devices. Ongoing research efforts are focused on structural health monitoring, adaptive structures, force sensors, vibration energy harvesting, soft robotics, underwater acoustic transducers etc.

Multiscale modelling methods have emerged in the last few decades where empirical or phenomenological assumptions used in modelling macroscale behaviour is replaced with information obtained from microscale models. Models based on empirical assumptions account for the underlying physics of the problem in an approximate manner. These models often rely on extensive experimental testing and parameter fitting to make them more reliable, thus rendering them not entirely predictive. In multiscale models, these limitations are avoided by incorporating information from more reliable, physics-based models, thereby making these models more predictive and significantly reducing time and cost incurred in experimental testing. These models are also not limited by the restrictions of the length and time scale of application microscopic models and can be used to simulate macroscopic material behaviour. 

Several multiscale methods have been developed to model material behaviour spanning multiple length and time scales. They are broadly categorized as upscaling and concurrent methods. In upscaling methods, information from microscopic models are supplied to macroscale continuum models.  In concurrent methods, there is two-way coupling between the macroscale and the microscale models. The most common example of this involves passing macroscopic stress or strain from the continuum model to simulate the microscopic model within a representative volume element, which in turn returns quantities which otherwise would have to be modelled using constitutive assumptions. Some examples of multiscale methods include

• Coupling of Crystal Plasticity with inputs from microscopic Molecular Dynamics (MD) or Discrete Dislocation Dynamics (DDD) simultations to predict mechanical behaviour of metals and alloys. 

• Modelling of polymers, carbon nanotubes and nanosheets using continuum-atomistic coupling such as the Quasi-continuum (QC) method and other homogenization techniques. 

• Coarse-graining of Molecular Dynamics (MD) simulations to study phenomena at much larger length and time scales than what MD operates at. 

• Interatomic potentials from Density Functional Theory (DFT) which are used in MD simulations to predict material properties and thermomechanical response of various materials. 

• Mobility Laws developed from MD simulations which are used in DDD and Crystal Plasticity models. 

• Data-driven reduced order models from microscopic simulation data which significantly reduce computational cost of upscaled models.