Abstract: Abstract: In traditional research method, the medium delivered by two-phase flow centrifugal pump contains only one kind of particles with the same diameter, while in actual working condition the medium contains various kinds of particles with much more than one diameter. In order to reveal the characteristics of the solid-liquid two-phase flow in centrifugal pump under actual operating condition, steady numerical calculation of the internal flow in a centrifugal pump was performed by using Fluent software with the Eulerian model and the standard k-ε turbulence model. The Phase Coupled SIMPLE algorithm was used to solve pressure-velocity coupled equations. The liquid-solid drag coefficient was calculated by the Gidaspow model, and solid-solid drag coefficient was calculated by the Syamlal-Obrien-Symmetric model. The external characteristic test was carried out with the centrifugal pump model. And the experimental results were compared with the numerical results. The results showed that the numerical results were in good agreement with the experimental results. In this paper, the numerical results were used to study the flow in the pump. Three groups of monodisperse particle swarms, whose diameters were respectively 0.02, 0.3 and 0.7 mm, were chosen to be the solid phase. The motion characteristics of monodisperse particle swarm in the pump were studied. It could be seen that the PVF (particle volume fraction) distribution became more nonuniform with the increase of particle diameter. As the particles with large diameter had large mass, the distribution of particles with larger diameter in the impeller passage was closer to the blade pressure surface and the phenomenon of solid-liquid separation became more obvious. The two-phase flow in the pump was researched when the solid phase contained 2 kinds of same monodisperse particles whose diameters were changed simultaneously. For the 2 particle swarm, the PVF difference between them is less than 7×10-6%, and the relative velocity between them is less than 5×10-4 m/s. With the increase of the impeller diameter, the slip velocity of monitoring points in the suction side of the blade increased first and then decreased, while in the pressure side it decreased first and then increased. The slip velocity on the blade surface decreased as the particle diameter decreased. The total pressure at monitoring points on the blade surface increased with the increase of impeller diameter. The head, hydraulic efficiency, and total pressure difference between the inlet and outlet decreased with the increase of particle diameter. When transporting 2 particle swarms with equal diameter, the external characteristics of the centrifugal pump were similar to that when transporting single particle swarm. When the medium contained 2 particle swarms with different diameters, on account of the interaction among big particle, small particle and water, the effect of the particles with different diameters on the characteristics of the pump was complicated. The distribution of big particles and small particles presented different regularity. The volume fraction and relative velocity of the big particles would change when the small particle diameter increased together with the big particle diameter unchanged. The influence law of the small particle in the 2 particle swarms with different diameters was similar to the one in 2 particle swarms with same diameters. The existence of small particles made the distribution of big particles more uniform in the impeller passage. The head and hydraulic efficiency reached the lowest value for the combination of particles with the diameter of 0.7 and 0.02 mm. With the increase of small particle diameter, the head and hydraulic efficiency increased rapidly first, and then decreased gradually. They reached the maximal value (80.12 m and 60.05%, respectively) when the small particle diameter was about 0.15 mm. The research is comprehensive and practical to reveal the characteristics of the two-phase flow centrifugal pump, and can be referenced to pump design and structural optimization.