Research

High-Re Turbulent Boundary Layer Physics

An important area where understanding turbulence is crucial is in turbulent boundary layers (TBL) – a thin region of highly complex fluid flow attached to a solid surface when there is a relative motion with its neighboring fluid. A high Reynolds number (Re) TBL occurs where there is a wide range of flow scales [for eg., $O\left(10\,\mu m - 500\,mm\right)$ ]. High-Re TBLs over both smooth– and rough-surfaces are present, for example, around ship and submarine hulls, airplane (and reentry vehicle) surfaces, turbine blades, automobiles etc. These regions are responsible for much of the drag and heat transfer between the fluid and surface. Despite this ubiquity of TBLs, the ability to accurately predict important engineering quantities for an arbitrary surface roughness geometry has been extremely challenging. Currently, we rely on costly experiments, simulations or conservative estimates to accurately know the engineering characteristics of flow over an arbitrary roughness profile. Small changes in geometry drastically alters the flow behavior and performance, limiting the utility and reliability of these ad-hoc prediction exercises.

To this end, CMAT lab investigates the following questions:

How do spanwise variations of wall-roughness geometry influence the boundary layer structure on the large scales? This is based on prior evidence that minute changes in wall roughness inherent to any realistic roughness-pattern can induce large scale secondary flow structures that are otherwise absent in idealized roughness.
How are passive- and active- scalars (such as pollutants, moisture, etc.) transported and mixed? This is crucial for many atmospheric phenomena, where the atmospheric boundary layer interacts with urban terrains in a highly complex manner. Understanding this interaction is crucial to accurately model transport of scalars (accidental or intentional releases, for eg.) and to model atmospheric chemistry accurately.
How can these observations be used towards predictive modeling? This is performed via synergetic collaborations with experts in computational fluid dynamics and turbulence modeling. See the list of Collaborators for more details.

See Publications annotated as ‘TBL‘ for details on the published results.

High Sc-Turbulent Jets and Complex Reactions

We use axisymmetric self-similar turbulent jets, which are a canonical configuration to study the fundamental aspects of turbulent mixing in liquids. For example, real world examples of turbulent jets occur in many combustion applications (where fuel is injected into air), chemical engineering systems (a chemical is injected into a quiescent or co-flowing chamber), natural phenomena (volcanoes, rivers discharging into oceans, etc.), aerospace applications (rockets, fuel injection in engines, etc.). High Schmidt number (Sc) turbulent jets predominantly occur where the injected fluid diffuses slowly into ambient fluid, such as in liquids (especially highly viscous ones), aerosols, etc. In such flows, the range of length scales associated with the fluid inhomogeneities is much larger [ $O(100-1000)$ ] times larger than the turbulent velocity fluctuations. Specifically, CMAT lab investigates the following questions:

How is the turbulence fundamentally different in variable-viscosity flows?
This is based on the observation that the turbulence behavior, when a low viscosity stream is injected into a high viscosity ambient fluid, is fundamentally different. The turbulence decay is a strong function of local viscosity, and the turbulence closure required modeling additional terms. We use high resolution PIV and PLIF techniques simultaneously to capture the turbulence transport mechanisms to better model these flows.
How does the variable-viscosity turbulence influence complex reactions?
It was previously noted that the asymmetric decay of turbulence in variable-viscosity flows influence the progress of competing reactions in a strong manner. Competing reactions have more than one reaction occurring competitively or in parallel. In variable-viscosity flows, local stoichiometry is a strong function of the turbulence evolution, and thus the reaction products produced.

Particle- and Aerosol-laden flows

The presence of particles in a fluid flow (gasses and liquids) can fundamentally alter the behavior of the turbulence. Specifically, we deal with small particles that are heavy and that constitute a significant fraction of the flowing mass (i.e. small volume fractions, $\phi_v$ ; high density ratios, $\rho_p/\rho_f$ ; and mass fractions of $O(1)$ ). In such flows, the turbulence evolves under a continuous exchange of turbulent kinetic energy between the fluid and particle phases, and the particles no longer fluid follow the flow faithfully. Additionally, under certain conditions, the particles tend to distribute inhomogeneously throughout the flow, forming particle-clusters and locally increased/decreased particle concentrations. These flow complexities reduce the reliability and accuracy of traditional turbulence predictive and flow modeling approaches, and require bespoke models.

This research at CMAT laboratory is performed using a combination of advanced experimental and numerical tools, in collaboration with Kasbaoui Research Group at ASU, to gain a holistic understanding of the full turbulence behavior. Experiments are capable of achieving higher Reynolds numbers in wall-bounded flows, while the numerical tools provide a more complete picture of the small-scale turbulence behavior. Specifically, the research at CMAT lab addresses the following research questions:

How can we experimentally capture the turbulence in particle and fluid phases simultaneously in wall-bounded flows?
We devise novel implementations of laser-based imaging to be able to measure the fluid turbulence [ $O(100\,\mu m)$ ], particle turbulence and particle clustering behavior simultaneously. This provides critical small-scale information of the energy exchange between the two media, that can then be used to develop generalized models for complex flow conditions.

Volumetric Absorption Tomography

When different gasses/liquids mix with each other in a highly turbulent environment, measuring the concentrations of the individual constituents everywhere is extremely useful, albeit equally challenging. The Volumetric Absorption Tomography leverages the differential absorption of light by the constituent species, calibrated backlighting, multiple imaging perspectives, and smart reconstructions to estimate the 3-dimensional flow field. The CMAT lab, in collaboration with Prof. Sam Grauer’s Research Group at Penn State University, extends these state-of-the-art approaches to reacting and non-reacting mixing environments to develop novel experimental approaches to study mixing phenomena.

The image on the left demonstrates the differential absorption of light by two jets of different colors, where they absorb different amounts of red, green and blue back lighting depending on the constituents.