A new correlation for prediction of drag coefficient on particle agglomerates using CFD

This work focuses on studying the drag coefficient of particle agglomerates and, based on previous studies, presenting a new drag correlation that improves the accuracy on estimating it. The study was carried out in the numerical field, using computational fluid dynamics (CFD), with experimental validation, to evaluate the drag acting in agglomerates of spheres in three conformations, according to the number of spheres (from 3 to 5). Based on the advances in the literature concerning the Kaskas equation for drag coefficient in spheres, the present study proposes the inclusion of a natural logarithm function, allied with shape parameters to characterize the agglomerate geometry, to enhance the capability of estimating the drag coefficient for a wide range of Reynolds numbers (0.1–3500). Results showed that, the use of the logarithm function improved the equation for all the shape descriptors tested, compared to previous correlations in the literature. Best results were observed by fitting the correlation using the flatness of the agglomerate, presenting a mean relative deviation of 3.79%, with maximum, minimum and standard deviations of 10.78%, −7.62%, and 5.06%, respectively.