AbstractThe distribution of lubricant flow field inside the two‐stage transmission reducer is very complicated during splash lubrication, which is difficult to be visually simulated and analysed using traditional finite element methods (FEM). There are many problems in model processing, algorithm selection, mesh division, and computational workload. Through sufficient investigation and experimental validation based on literature, the moving particle semi‐implicit (MPS) method is proposed to study the splash lubrication characteristics of a rail vehicle reducer. Through this method, the visual simulation calculation of multiple working conditions of reducer with complex structure is realised, and the unreasonable structure of the gearbox body is optimised. The model of oil supply demand in the Hertz contact zone of gear transmission is established, and the time‐domain variation law of oil particles number in the contact zone and oil supply state are analysed. It is found that the optimised reducer model can meet the demand of oil supply under the four typical working conditions of the rail vehicle. The higher the initial oil volume is, the more sufficient the oil supply is. Compared with the working temperature conditions of 20 and 80 °C, the lubrication effect is better at 40 and 60 °C. By analysing the power loss proportion of each gear, it is found that the sum of power loss of gear 1 and gear 2 is dominant under different working conditions, reaching 76%–99.5% of the total churning power losses.