Computational guidelines and an empirical model for particle deposition in curved pipes using an Eulerian-Lagrangian approach

Computational guidelines are provided for modeling turbulent particulate-laden flows in curved pipes using a Reynolds-Averaged Navier-Stokes (RANS) approach in ANSYS FLUENT. The standard k-ε model and the linear pressure-strain Reynolds Stress Model (RSM) are used to close the RANS equation. Different near-wall treatments associated with the selected turbulence closures are employed to study their impact on estimating pressure drop and particle deposition (grade efficiency). The bends studied had different bend angles, bend curvature ratios, and the flows had various Re numbers. Experimental results available in the literature are employed to validate the computational results. The observed turbulent flows exhibit complex secondary flow patterns at the bend and these patterns are influenced by the bend curvature ratio and the flow Re number. The pressure drop along the curved pipe is well estimated using either closure for all three near-wall treatments with the exception of pressure at the inner and outer walls of the 180° bend, where an error of about 7% is obtained. With respect to particle deposition however, the performance of the RSM outperforms that of the standard k-ε model. Using an Enhanced Wall Treatment (EWT) combined with an RSM can improve the accuracy by as much as 19% in a 90° bend and up to 30% in a 180° bend when compared to other near-wall treatments and closure models. The RSM with EWT should thus be employed when modeling particle deposition in flows with curved streamlines. In addition, a new empirical model is proposed to model particle deposition efficiency in curved pipes. The model extends the range of previous empirical models and accounts for the effects of the Stokes number, the bend angle, and the curvature ratio.

► Turbulent flows and particle deposition in curved pipes are modeled.
► Guidelines for modeling pressure drop and particle deposition are provided.
► The RSM with enhanced wall treatment best predicts particle deposition.
► Pressure drop is less sensitive to near-wall treatments and to the closure used.
► An empirical model is developed for estimating grade efficiency curves.