Abstract: Abstract: In order to deeply understand the flow distribution of the torpedo screen filter, the Reynolds-averaged Navier-Stokes (RANS) equation and Re-Normalization Group (RNG) k-ε turbulence closure model were used to simulate the flow field of this filter system. To ensure the reliability of the numerical simulation, the physical experiment results were compared to the numerical simulation results, and the result showed that the maximum relative error between the head loss in the physical experiment and the numerical simulation was 5.99% when the filtering system was operated at a maximum flow rate of 360 m3/h, indicating reliability of the simulation method. The simulation results showed that the torpedo components and the boundary conditions of the outlet notably affected the distribution rules of both the velocity and the pressure fields for the filter, especially for the torpedo components. The distributions of the flow velocities of inside and outside of the screen along the X-axis were not uniform during the filtering process. And the flow velocity distributions inside and outside of the filter screen along the X-axis were divided into 3 stages: 1) a rapid increase; 2) a rapid decrease; and 3) a gradual decrease. The maximum velocities inside and outside of the screen and their spatial locations along the X-axis were different. For the upper part of the screen, the inside flow velocity increased in a larger amplitude than did the outside flow velocity, and the flow velocity achieved its maximum (5.53 m/s) at the X value of 0.29 m. The outside water flow velocity of the upper part of the screen achieved its maximum value (3.9 m/s) at the X value of 0.33 m. When the X value ranged from 0.2 to 0.6 m, the inside flow velocity of the upper part of the screen was larger than that of the outside; however, the difference in the flow velocity along the X-axis continually decreased from the maximum difference of 4.1 m/s. For the lower part of the screen, the maximum flow velocity and its position of lower part of the screen was 5.4 m/s and 0.42 m, respectively. When the X value was 0.41-0.51 m, the flow velocity of the water in the outside area of the lower part of the screen was larger than that of water in the inside area. Nonetheless, the flow velocities of water in the other inside of lower screen were higher than those of water in the outside, and the maximum difference in the flow velocity was 2.28 m/s. There was a notable difference in the pressure distributions along the inside and outside of the screen. The pressure of the inside along the X-axis first rapidly increased, then gradually decreased, and finally stabilized. In contrast, the pressure of the outside along the X-axis exhibited a greater variation; for example, the pressure rapidly decreases at the outlet, and the fluctuations along the X-axis are relatively large. The pressure of the inside of the screen was larger than that of the outside; however, the pressure differences along the X-axis continually decreased, and the maximum and minimum pressure differences between the two were approximately 23 and 0.5 kPa, respectively. This indicated that the screen clogging was not evenly distributed, and the screen clogging began at the downstream end of the screen and developed progressively upstream towards the outlet. This phenomenon significantly affected the cleaning process of the filter. Therefore, it is strongly suggested that in a practical application the optimum drainage differential pressure must be circumspectly considered.