Our goal is to understand and model turbulence. We combine large-scale simulations (i.e., computational fluid dynamics or CFD) and theoretical analysis for our research, with an emphasis on the multi-scale and non-equilibrium aspects of turbulence.
Big whorls have little whorls
That feed on their velocity,
And little whorls have lesser whorls
And so on to viscosity.
– Lewis F. Richardson, 1920
Figure courtesy of Allard Overmeen
High-Fidelity 'Experiments' of Turbulent Flows
Although we don't often rely on WD-40 and duct tape in our laboratory, our team considers ourselves experimentalists. We use high-fidelity simulations that offer a high-resolution and low-uncertainty approach to predicting fluid motion. These numerical `experiments’ act as powerful virtual microscopes, allowing us to observe and analyze small-scale motions that may be difficult to measure in reality. Additionally, they serve as virtual wind tunnels, enabling cost-efficient design and testing of large-scale systems that cannot be implemented in a physical lab. In essence, high-fidelity simulations are digital twins of physical experiments, providing unique capabilities beyond current measurement techniques. Our team is dedicated to pushing the boundaries of these virtual microscopes and wind tunnels to better understand turbulent flows in both natural and engineering applications.
Reduced optical access near stationary or moving objects prevents physical measurements. Simulations are capable of providing detailed 4D data and complete control of the objects. This figure shows the pore water flow that occurs when dermal scales bristle over shark skin, a possible cause of the distinguished hydrodynamic performance of sharks.
Eddies play an important role in mass and momentum transfer in turbulence. These fine vortices in a flow over a bump are predicted by simulation. They could be manipulated (i.e., flow control), leading to changes in the flow around the bump and the resultant drag (i.e., cost of operation).
Many instabilities in turbulence are difficult to be identified due to their 3D and time-varying nature. In this figure of a separating turbulent boundary layer, at least three types of instabilities: Görtler, Crow, and Kelvin-Holmholtz, can be identified from the well-resolved eddies.
Numerical experiments may also include coupled multiphysics such as particle transport, heat transfer, and combustion. What is shown here represents the airborne transmission of virus-laden buoyant respiratory droplets.
Googling "CFD" will give you many eye-catching images. However, appearances can be deceiving. For instance, the visualization of the evolution of a helicopter blade-tip vortex in this figure may look plausible based on current models (b-e), but they are actually far from accurate. Our team conducts high-fidelity simulations to inform the development of new models that can deliver the level of precision demonstrated in (f).
Shark dermal scales exhibit notable distinctions from those found in bony fish. While the ridges on their crown resemble the recognized drag-reducing riblets, researchers have not agreed upon the mechanisms of the hydrodynamic advantages of shark skin. Recently, there has been a growing interest in exploring the potential of denticles to reduce drag by controlling flow separation (detachment of flow from the surface) during sharks’ maneuvers. Our numerical simulation of the fluid flow over shark denticles revealed that they create micro channels between the stems underneath their crowns. The pore flow through these channels produces a thrust that gives the shark an extra boost (i.e., thrust) as it swims. This figure shows 1) the vortices in the flow that is separating from a bump covered by shark denticles (isosurfaces, colored by distance to the bottom wall); and 2) the flow on the surface of the bump (contour, colored by velocity, blue indicate reverse flow). The pore flow shown in the inset is under the crown of the denticles and pushes the denticles upstream