Research


My research interest lies in employing numerical simulations to study particle-laden flows. My objective is to model them in a simple and fundamental way so as to understand the physics behind these flows. An important feature of these flows is the large separation of scales involved. In order to achieve this, I perform Particle Resolved - Direct Numerical Simulations (PR-DNS). This involves resolving all the length scales of the flow as well as the particle itself.

Lattice Boltzmann Method Simulations of Stokes Number Effects on Particle Trajectories in a Wall-Bounded Flow


Abstract:
Experimental studies of particle-laden flows in a pipe show that the spatial distribution of the particles across the radius of the pipe is dependent on the Stokes number [Timothy C. W. Lau \& Graham J. Nathan, J. Fluid Mech. 2014]. It has been suggested that the Saffman lift effect [Saffman, P. G., J. Fluid Mech. 1965] makes a significant contribution to this spatial distribution. The Saffman lift effect has been studied in prior works by several authors and the relative contribution of the lift force has been studied within the context of various forces acting on particles in a flow. The lift force depends on the particle size and the velocity of the particle relative to the gas phase. In this study, the lattice Boltzmann method is employed to study the mechanism of particle migration of an isolated particle moving in a wall-bounded flow. The boundary condition proposed by Bouzidi \textit{et al.} [Phys. Fluids, 2001], which involves the bounce-back scheme modified to account for fractional link distances between the wall and the fluid node, is used for the particles. The force acting on the particle is found by adding the momentum lost by all the fluid molecules as they bounce back from the particle surface along the link joining the particle and the fluid boundary nodes. This force is used to update the position of the particle after every streaming step. The torque acting on the particle is determined similarly and is used to update the angular velocity of the particle. It is found that at low Stokes number the particle behaves like a neutrally buoyant particle and exhibits the Segr{\'e}-Silberberg effect. With increasing Stokes number, the particle exhibits an oscillatory behavior about its mean position. For large Stokes number, the particle oscillations are significant. If the ratio of channel height to particle diameter is increased, the particle moves closer to the wall and the oscillatory behavior is evident at lower Stokes number.


View Publication

Comparison of the structure of computed and measured particle-laden jets for a wide range of Stokes numbers


Abstract:
Turbulent particle-laden jets have been the subject of interest for many years on account of their relevance to several practical devices like engines, combustors and gasifiers. While prior experimental studies have examined particle-laden flows at high Stokes number, experimental data on particle-laden jets with particle Stokes number of the order of one and lower have not been available until recently. This study presents results from computations of particle-laden jets for Stokes number ranging from 0.3 to 500 and their comparison with measured results. The mean gas-phase velocity is found by solving RANS equations with a $k-\epsilon$ model for turbulence. The particles are solved in a Lagrangian framework with the coupling between the carrier and dispersed phase modeled using a drag coefficient with a high-Reynolds number correction. Particle-turbulence interactions are modeled using a random-walk dispersion model. The influence of Stokes number on the spreading rate of the carrier and dispersed phase is examined. It is shown that for the range of Stokes numbers considered, the computed results agree with measured particle centerline velocities within about 20\%. The changes in particle velocities predicted as Stokes number varies are consistent with measured changes in these variables. While no specific trends can be identified in the differences between computed and measured results that would relate the differences to Stokes number, several parametric studies are carried out to investigate the effect of jet inlet gas phase turbulence intensity, fluctuating particle velocity at the jet inlet, turbulence modulation and the dispersion model employed on jet spreading and centerline velocities.


View Publication

Publications


A. S. Jebakumar, K. N. Premnath, and J. Abraham
Lattice Boltzmann Method Simulations of Stokes Number Effects on Particle Trajectories in a Wall-Bounded Flow
Computers and Fluids, 124:208-219, 2016.


A. S. Jebakumar and J. Abraham
Comparison of the Structure of Computed and Measured Particle-Laden Jets for a Wide Range of Stokes Numbers
International J. Heat and Mass Transfer, 97:779-786, 2016.