Application Of Time-space Adaptive Smoothed Particle Hydrodynamics In Large Deformation Of Landslide

Due to the damage caused by landslide,much more attention has been paid on this problem.Numerical simulation is usually used as an important means to predict and protect the landslide.The traditional grid-based numerical methods often face the challenges of the mesh distortion,when simulating the large deformation problem such as landslide.As a typical Lagrangian particle method,Smoothed Particle Hydrodynamics(SPH)has the advantage of being independent of mesh and the capability of simulating the free surface of large deformation problem.Therefore,the SPH method has been widely applied for the landslide problem in recent years.However,it also has the disadvantages on the computational efficiency and accuracy.In order to improve the efficiency and accuracy of the conventional SPH method,two key creative techniques has be studied in this research.: 1.The relative potential vector has been introduced to avoid the unphysical numerical error occurred by the traditional adaptive particle refinement(APR)which was proposed by Fledman,this algorithm can ensure the conservation of momentum and angle momentum of child particle after split,which can improve the accuracy of APR algorithm significantly.2.The Chebyshev polynomial has been introduced to calculate the orders of Runge-Kutta time integration based on current time-step.Then,we established the time-space adaptive smoothed particle hydrodynamics scheme by combined the Runge-Kutta Chebyshev explicit time integration with the improved adaptive particle refinement pattern,and the corresponding analysis and research on the large deformation of the landslide is carried out by using T-S ASPH.The main contents of this article are as follows:(1)The feasibility of SPH in the field of landslide problem analysis is established by collecting and sorting out the currently research status of landslide problems;at the same time,the shortcomings of traditional SPH methods are analyzed and summarized,and the research focus of this paper is established.(2)The basic theories of SPH and the mainstream SPH numerical instability improvements has been introduced by the same time,the influence of integral nodes of integration scheme has been discussed by the research.(3)The concept of relative potential vector has been introduced to improve the traditional APR algorithm to ensure the symmetrical distribution of split particles in this article,with the construction of corresponding spatial adaptive algorithm;at the same time,the Chebyshev polynomial combined with the Runge-Kutta integration method is used to propose a RKC self-adaptive explicit time integration implements a spatial self-adaptive algorithm.And based on the established of time-space adaptive hydrodynamics(TS-ASPH),research of large deformation problem of elastic constitutive was represented.Results shows that T-S ASPH has better performance in computational accuracy and efficiency by comparison of Fine-SPH and T-S ASPH.(4)the elastic-to-plastic algorithm framework has been established based on Drucker-Prager constitutive law and T-S ASPH algorithm,the analyzation of nonassociated flowing granular soil and unstable slopes was represented by following simulations,and the feasibility and superiors of T-S ASPH in the domain of large deformation problem of elastic-to-plastic analyzation was verified by compare the simulation results of Fine-SPH and following works.(5)the performances of T-S ASPH in large deformation problem of elastic and elastic-to-plastic has been summarized,and the advantages of T-S ASPH was represented,the drawbacks of current work have been concluded and the overview of the future work was published.