Abstract

Mixing and heat transfer enhancement due to chaotic advection in a converging-diverging channel flow is characterized and quantified by spectral element direct numerical simulations of the time-dependent Navier-Stokes, continuity and energy equations. The computational simulations are performed in a two-dimensional converging-diverging channel model composed of sinusoidal walls. Results are presented for laminar and transitional flow regimes, in the Reynolds number range 110<Re<600. Eulerian and Lagrangian flow characteristics, such as vortex dynamics and Lagrangian Lyapunov exponents, are evaluated to characterize the flow pattern and to determine the chaotic mixing as a function of the Reynolds number. Heat transfer parameters, such as the temporal evolution of the temperature distributions and Nusselt numbers, are obtained to determine the heat transfer performance. Lagrangian Lyapunov exponent evaluations indicate that the onset of chaotic mixing occurs at Re approximately 130, as the flow evolves from a laminar to a transitional state. Results reveal that the heat transfer enhancement is due to chaotic mixing of fluid particles whose Lagrangian trajectories diverge exponentially fast. It is also found that chaotic mixing is stronger in the vortical regions and spreads to other flow regions as the Reynolds number increases.