Abstract Gravity flow deposition system has become a key exploration area in the northern part of the South China Sea.Due to the technical bottleneck at the present stage, gravity flow exploration in deep water area faces many difficulties, indicating that the distribution rules of these deep water gravity sedimentary reservoirs are complex. In order to strengthen theoretical research, the distribution rules of deep water sand bodies and the deposition process of gravity flow are further explored. This experiment takes sedimentology, hydrodynamics and etc. as the theoretical basis and physical simulation as the means to pre-structure a bottom shape of gravity flow in the glass flume in advance.Then the crossover and contrast experiments were divided into 15 groups, observed, described, recorded the transition of flow patterns from collapse to sandy debris flow and turbidity flow under the condition of single variable (different slope and concentration of sand).Respectively simulated under different density gravity flow deposition conditions about distribution of slab-like debris or debris within the debris flow.It is found that under the steep slope condition, the shear resistance of the sedimentary block is weakened and it is easy to deform. With the continuous erosion of the fluid, the fluid property transforms into a debris flow and deposits into the deep water area. The characteristics of particle distribution in gravity flow deposition with different densities are gravel and coarse particles are mainly distributed at the bottom of the channel,and gravel is randomly scattered distributed at the channel and its side edge.Under the condition that the hydrodynamic conditions are relatively stable, the particles are arranged along the water channel, and the sequence of the lath-like particles from chaotic to bedding is also reflected in the lateral direction.Through the quantitative analysis of the fluid transformation process and control factors of different types of gravity flow, the fluid transport flow pattern and particle distribution state with different densities are obtained to predict the high-quality reservoir, thus providing a powerful theoretical basis for the exploration and development of deep water area.