Abstract: Pork is the second most consumed meat in the world. The scale and intensification of pig farming have been ever-increasing in recent years. It is required for the continuous development of Precision Livestock Farming technologies. Cleaning inside the house is one of the most important parts of precision pig farming. Among them, the water jet cleaning pig manure is assumed as a two-phase flow of water-pig manure particles. This study aims to investigate the process and effect of water jet cleaning pig manure. Numerical simulations were conducted to simulate the two-phase interaction between the water jet and pig manure using computational fluid dynamics (CFD) and discrete element method (DEM). Firstly, the water jet flow field model was established to complete the flow net independence verification for the grid partitioning using Fluent software. Then, the pig manure particle model was constructed using Rocky. The relative error between the simulation and the actual measurement was 1.78%, indicating the reliability of the pig manure physical properties parameters and the pig manure simulation model. A linear adhesion model was introduced to optimize the pig manure model using coarse-grained modeling. The total number of particles was also reduced for the less computational complexity during simulation. Finally, fluid-solid coupling experiments were conducted using Fluent-Rocky DEM. The experiments were carried out to simulate the vertical impact cleaning of a pig manure layer at a distance of 300mm from the nozzle outlet using a water jet. The results showed that: 1) The process of water jet cleaning pig manure was divided into three stages: free jet cleaning, submerged jet cleaning, and impact strengthening. The water jet flew from the nozzle inlet and stabilized on the surface of the pig manure layer after approximately 0.3 s. As the cleaning time increased, the pig manure layer showed bowl-shaped pits with a central peak forming at the center of the pit. 2) The highest cleaning efficiency occurred within 0.45 s after the continuous water jet reached the surface of the pig manure layer. A conical nozzle performed better cleaning depth and cleaning efficiency than a straight one under the same structural parameters and operating pressure. The cleaning amount of the conical nozzle increased by about 0.83% at a continuous cleaning time of 0.45 s, compared with the straight one. 3) The maximum horizontal velocity of the conical nozzle was obtained at the target cleaning depths of 10 and 30mm. The maximum cleaning speed was negatively correlated with the cleaning target depth and working pressure, whereas, there was no significant relationship with the nozzle outlet diameter. There was the greatest impact on the depth of the cleaning target. The larger the nozzle outlet diameter and working pressure were, the larger the cleaning width and the larger the cleaning area were within the same time. A coarse-grained model was used to simulate the granulation of pig manure particles. A summary was made on the correlation between the CGM values and the mechanical parameters of pig manure particles. The efficiency of CFD-DEM coupled simulation was improved for the water jet cleaning of pig manure. The effectiveness of the simulation was verified to test the cleaning effect of the nozzle using the cleaning test platform. The research results can provide valuable information and theoretical references for the design and optimization of cleaning equipment for livestock and poultry manure.