PSI - Issue 71

Abhijit Parate et al. / Procedia Structural Integrity 71 (2025) 256–262

258

(b) Mesh Fig.1. Details of geometry and mesh adopted for numerical simulations

2.2 Solution Setup To simulate the particulate flow in a gas stream the Eulerian-Lagrangian approach is employed. The gas is considered as a continuous medium and Reynolds averaged Navier-stokes (RANS) equations are used to obtain the solution in the Eulerian scheme. The two-equation k- ω model is used to simulate the turbulence. The flow in the near-wall region is modeled using standard wall function. The working gas in this study is air, modelled as an ideal gas with the properties: Density: 1.225 kg/m³ (at 298 K, 1 atm), dynamic viscosity: 1.81×10 −5 kg/(m·s), specific heat capacity: 1005 J/(kg·K) and thermal conductivity: 0.0257 W/(m·K). The solid particles of density of 2650 kg/m 3 and size 100 μ m are considered as discrete phase. Their motion in a Lagrangian framework is obtained using Newton’s second Law. Two-way turbulence is employed to determine the effect of the solid and gas phase on each other. The solid particles get reflected back into the fluid medium after colliding with the pipe walls. The path of the particle after the collision is defined using the coefficient of restitution of 0.9. The effect of turbulence eddies on particle motion is determined from the discrete random walk (DRW) model. SIMPLE algorithm is used for Pressure-Velocity coupling. Standard discretization is applied for pressure. Second-order upwind discretization is used for momentum and turbulent kinetic energy. The solution is assumed to be converged when residuals become smaller than 10 -4 . 2.3 Boundary Conditions Gas and solid particles flow into the JIT through the inlet at a specified velocity through ‘velocity inlet’ as boundary condition and exits through the outlet with boundary condition ‘outflow’. The pipe surface is assigned wall boundary conditions. The particle mass flow rate is set as 0.05 kg/s for all cases. Time step size of 0.001 seconds is set for the transient simulation. A run of total 1 second is performed before collecting the particle velocity data from the simulation. 3. Results and Discussions 3.1. Validation Fig. 2 presents the contour of fluid velocity at mid sectional plane of the JIT. It follows the similar trend as reported in literature (Fan et al. 2011). Further, to validate the accuracy of the numerical model used in this study, a comparison is made between the numerical results and available experiment data (Fan et al. 2011) of particle velocity from the nozzle exit. Fig. 3 shows the comparison of normalized particle velocity i.e., the ratio of the particle velocity to the maximum particle velocity, with the variation along the axial length. It shows that the predicted particle velocity profile closely matches with the experimental data.