Development of drag, lift, and torque models for cylindrical biomass particles in suspension systems

The motion of non-spherical particles is widely encountered in suspension systems on heat and mass transfer for energy, chemical, and environmental engineering. However, the existing force models apply the assumptions of spherical or simplified particle geometries, making them insufficient for accurately describing the hydrodynamic behaviors of typical biomass particles like cylindrical straws in industrial co-firing settings. To address this limitation, this study employs an immersed boundary method particle-resolved direct numerical simulation (IBM-PR-DNS) to investigate the hydrodynamic forces acting on cylindrical particles with an aspect ratio of 2.5. A comprehensive analysis is conducted over a range of Reynolds numbers (10 ≤ Re ≤ 2000), solid volume fractions (0.05 ≤ φ ≤ 0.2), and particle incidence angles (0° ≤ θ ≤ 90°). The results reveal that increasing Reynolds number and solid volume fraction greately enhance the complexity and heterogeneity of local flow structures, resulting in pronounced non-uniformity in particle scale forces. The drag shows a strong dependence on the incident angle, particularly at high Reynolds numbers, with a reduction of up to 46.33 % as θ increases from 5° to 85°. The lift and torque exhibit peak values near θ = 45°, and their magnitude increases with Re and φ due to intensified wake asymmetry and flow disturbances. Based on extensive data, the correlations for the drag, lift, and torque are developed using a genetic algorithm optimization framework. These models explicitly incorporate the effects of Reynolds number, solid volume fraction, and incident angle, and well indicate agreement with the IBM-PR-DNS results, with root mean square errors below 0.5. Compared to the existing correlations for ellipsoids; disks; and cubes, the proposed models demonstrate better accuracy and robustness for randomly oriented cylindrical biomass particles in the dense suspensions. This work might provide reliable generalized force models for cylindrical particles in multiphase flows, and offer a high-fidelity reference for numerical simulations of biomass co-firing, fluidized bed combustion, and related gas-solid flow systems.