Numerical and Experimental Investigation of multiphase flows. Fundamental and Applied studies

Abstract

Multiphase flows can be met both in nature and in industrial processes. The particular case of multiphase flows is particle-laden flows, consisting of fluid phase and particle phase. Study is aimed at the investigation of the dynamics of particles in laminar and turbulent flows. The dynamics of particles depends on their physical properties. The study of the particle’s dynamics in turbulent flow consisted of three parts: simulation of homogeneous isotropic turbulence (HIT), simulation of particle-laden flows and investigation of particles’ pair dispersion. Initially, the turbulent flow was simulated in OpenFOAM via Large Eddy simulation (LES) and Direct Navier Stokes (DNS) methods. Then not decaying homogeneous isotropic turbulence was achieved by addition of forcing term. The effect of different values of forcing constant on turbulent flow was investigated. The mesh convergence study was performed for LES and DNS. Results showed convergence in LES simulations and the reduction of volume-averaged velocity fluctuations (U’) in DNS with an increase of the number of mesh elements. The particle phase simulation was performed in Matlab and CFDEM. The simulation in Matlab revealed several constraints. The simulation of particles with turbulence phase for a short period of time was performed in CFDEM. However, due to the decaying nature of the flow, the investigation of particles’ trajectories could not be performed. The pair dispersion of inertial particles in turbulent flow was investigated for available experimental data. The results were represented on graphs with normalized scales. The correlation between the velocity of flow and the normalized value of the initial separation was determined. The mean square separation does not change significantly until the first decade. The motion of clots in blood flow was numerically simulated using Ansys Fluent package. The mesh convergence study was performed and the velocity of the particles in blood flow was analyzed. The results obtained for blood flow showed distribution close to Gaussian. The preliminary experiment was conducted for small particles of smoke. The smoke cloud was detected via the image intensity analysis.