Research on dynamic characteristics of granular flow based on the material point method

The dynamic characteristics of granular flow is very important to improve the understanding of disaster development. The material point method was used to study the dynamic characteristics of granular flow, and the analytical and experimental examples were used to verify the correctness of the numerical simulation. Numerical result shows that the material point method has high accuracy and resolution when simulating the dam-break problem with discontinuous characteristic. The numerical information provided can be used to study the dynamic characteristics of the granular flow, but the numerical information is not easy to be measured by experiment. The numerical result shows that with the change of Froude number, the granular flow mainly goes through several stages: initial start-up, accelerated development, full development and deceleration deposition. The velocity profile changes from the power function distribution in the initial start-up stage to the linear distribution in the accelerated development stage, and forms a more uniform velocity profile in the full development stage. In the deceleration deposition stage, the velocity approaches zero to complete the final deposition. In different development stages, the lateral pressure coefficient of the section is associated with the motion state, but the lateral pressure coefficient is between the limit active state and the limit passive state obtained by the Savage and Hutter theory (SH theory). Unlike SH theory, which assumes that the lateral pressure coefficient can only be selected from two limit values, the lateral pressure coefficient should be continuously changed, which is reflected in the material point method. For large granular flow moving at high speed, the traditional depth-integrated model is effective because the main motion is in the full development stage. However, for complex granular flow processes, the material point method has wider applicability with fewer assumptions.