A method to improve the accuracy of landslide hazard chain simulation using the smoothed particle method

Abstract

This invention relates to the field of geological disaster prevention and control, specifically to a method for improving the accuracy of landslide disaster chain simulation using the smooth particle method. The method comprises the following steps: Step 1: Collecting survey data on the target landslide and its impact range to be simulated, and based on this data, clarifying the geological conditions, topography, landslide body physical and mechanical parameters, and water body physical and mechanical parameters of the target landslide and its impact range; Step 2: Constructing a numerical analysis model for landslide disaster chain numerical simulation based on the smooth particle method; Step 3: Setting initial calculation parameters based on the numerical analysis model constructed in Step 2; Step 4: Starting the simulation calculation, updating the effective radius of any calculation time step according to a preset rule, and continuing the calculation for the next time step until all calculations are completed, obtaining the numerical simulation result of the landslide disaster chain. This invention can significantly improve the accuracy of landslide disaster chain simulation using the smooth particle method.

Description

Technical Field

[0001] This invention relates to the field of geological disaster prevention and control, and specifically to a method for improving the accuracy of simulating landslide disaster chains using the smooth particle method. Background Technology

[0002] With the construction of large-scale water conservancy and hydropower projects, tunnel projects in high-altitude and cold regions, and other infrastructure projects, landslide disaster chains, represented by landslide-wave surges, landslide-dammed lakes, and glacial landslides-debris flows, are becoming increasingly frequent, causing enormous losses to people's lives and property. Simulating, analyzing, and predicting the evolution of disaster chains at potential landslide sites in advance is an effective means to reduce losses from landslide disaster chains. Based on the fluid characteristics of the landslide disaster chain evolution process, traditional numerical simulation methods for solids, such as the finite element method and discrete element method, are no longer applicable. Therefore, the Lagrange particle method, which discretizes the fluid into freely moving point particles, has become the best solution for simulating the evolution process of landslide disaster chains.

[0003] Common Lagrange particle methods include the Smoothed Particle Method (SPH), the Semi-Implicit Moving Particle Method (MPS), and the Finite Volume Particle Method (FVP). Compared to MPS and FVP, the SPH method, due to the introduction of artificial viscosity and the use of explicit solution, offers better computational stability and faster calculation speed, making it more suitable for numerical simulation of fluid motion on a large engineering scale, such as landslide hazard chains. The essence of the SPH method is to interpolate the physical properties of a particle based on the properties of its neighboring particles. The number and range of neighboring particles involved in the calculation are crucial to the accuracy and efficiency of the SPH method. During the calculation, an effective radius is typically used to control the number and range of neighboring particles. Currently, when using the SPH method for numerical simulation of actual engineering projects, it is generally assumed that the fluid particles are uniformly distributed, and a fixed effective radius is given for calculation based on experience or trial and error.

[0004] However, due to the complexity of topography, the fluid movement in a landslide disaster chain involves numerous turbulent and splashing areas. These areas experience particle accumulation and separation, causing drastic changes in the number and range of neighboring particles involved in the calculation, leading to increased computational errors and significantly reduced simulation accuracy. Therefore, this invention proposes a method for adaptively adjusting the effective radius during the calculation process, greatly improving the accuracy of the smoothed particle method in simulating landslide disaster chains without compromising computational efficiency. Summary of the Invention

[0005] The present invention aims to provide a method to improve the accuracy of landslide disaster chain simulation using the smooth particle method, in order to solve the problems of unclear initial effective radius and drastic changes in the number and range of neighboring particles when particle accumulation and separation occur in complex flow regions due to the use of a fixed effective radius.

[0006] The method for improving the accuracy of landslide hazard chain simulation using the smoothed particle method in this scheme includes the following steps:

[0007] Step 1: Collect survey data on the impact range of the target landslide and disaster chain to be simulated, and based on the survey data, clarify the geological conditions, topography, test results of the physical and mechanical parameters of the landslide body, and physical and mechanical parameters of the water body in the impact range of the target landslide and disaster chain.

[0008] Step 2: Construct a numerical analysis model for landslide disaster chain numerical simulation based on the smooth particle method;

[0009] Also includes:

[0010] Step 3: Based on the numerical analysis model constructed in Step 2, set the initial calculation parameters, which include the smoothing kernel function, the calculation time step size, and the initial effective radius;

[0011] Step 4: Start the simulation calculation. After updating the effective radius of any calculation time step according to the preset rules, continue the calculation for the next time step until all calculations are completed, and obtain the numerical simulation results of the landslide disaster chain.

[0012] The beneficial effects of this plan are:

[0013] By scientifically selecting the initial effective radius and updating it in real time during the calculation process in the simulation of landslide disaster chains using the smoothed particle method, the number of neighboring particles involved in the calculation remains stable when particle accumulation and separation occur in complex flow regions. This prevents drastic changes in the number and range of neighboring particles when particle accumulation and separation occur in complex flow regions, and greatly improves the accuracy of landslide disaster chain simulation using the smoothed particle method without compromising computational efficiency.

[0014] Furthermore, in step 2, a geometric model is constructed based on the geological conditions and topography described in step 1;

[0015] The computational domains of the landslide body and water body are discretized into a set number of particles with different physical properties;

[0016] Based on step 1, the physical and mechanical parameters of the landslide body and the physical and mechanical parameters of the water body are assigned to the corresponding physical and mechanical parameters of each material particle.

[0017] The beneficial effects are: by setting different parameters based on the landslide body and the water body, the characteristics of different targets can be represented, providing accurate data support.

[0018] Furthermore, in step 2, the constructed geometric model includes model boundaries, landslide body boundaries, and water body boundaries.

[0019] The beneficial effects are: the geometric model contains complete content and is closer to the actual target landslide.

[0020] Furthermore, the boundary conditions of the geometric model are constructed by setting mirrored particles or virtual particles.

[0021] The beneficial effect is that the way boundary conditions are set more accurately distinguishes different regions.

[0022] Furthermore, the computational domains of the landslide body and water body are discretized into a predetermined number of particles with different physical properties, including:

[0023] Discretize the landslide area enclosed by the landslide boundary in the geometric model constructed in step 2, and uniformly insert landslide particles.

[0024] Discretize the water body region enclosed by the water body boundary in the geometric model constructed in step 2, and uniformly insert water body particles.

[0025] Furthermore, in step 2, the physical and mechanical parameters of the landslide body and the water body are assigned to the corresponding physical and mechanical parameters of each particle, including the particle density, static shear stress, gravitational acceleration, viscosity, and position of the landslide body particles and the water body particles.

[0026] The beneficial effect is that by assigning physical and mechanical parameters to the various parameters of the particles, the consistency between the model and the actual target landslide is improved.

[0027] Furthermore, in step 3, setting the initial calculation parameters, including the smoothing kernel function, the calculation time step size, and the initial effective radius, includes the following steps:

[0028] Step 3.1, Set the initial effective radius according to the dimensions. , is represented as:

[0029] ;

[0030] In the formula, d = 1, 2, 3;

[0031] Step 3.2, set the smoothing kernel function to a cubic spline function, which is expressed as:

[0032] ;

[0033] In the formula, , It represents the distance between two particles. For the effective radius, It is a normalization factor related to dimension d, expressed as: , , ;

[0034] Step 3.3: Based on computer computing resources and simulation timeliness requirements, set the calculation time step size, and the calculation time step size should meet the following requirements:

[0035] ;

[0036] ;

[0037] ;

[0038] In the formula, The initial distance between adjacent particles in the water body region. Vmax represents the initial distance between adjacent particles in the landslide area; Vmax represents the maximum velocity of particles in the water area. This represents the kinematic viscosity of the landslide mass.

[0039] The beneficial effects are as follows: In the research and implementation of this method, definite values ​​of the initial effective radius under different dimensions are given when simulating landslide disaster chains using the smooth particle method. These effective radii under different dimensions are not arbitrarily set, but are obtained through scientific calculation methods. The use of scientifically determined values ​​replaces the initial effective radius determined by experience or trial and error, ensuring the accuracy of numerical simulation results under landslide disaster chain scenarios and greatly improving the timeliness of calculation. The calculation time step is constrained and calculated separately for water body region and landslide body region. In the process of constraint and calculation, the initial distance between adjacent water body region and landslide body region, the maximum velocity of particles in water body region, and the kinetic viscosity of landslide body are taken into account. This can realistically reflect the internal particle movement of actual water body and landslide body, ensuring that the obtained calculation time step does not miss the information represented by detailed particles, making the calculation time step value more reasonable, and thus ensuring the accuracy of landslide disaster chain numerical simulation results.

[0040] Furthermore, in step 3.3, Vmax is taken as 4m / s in actual engineering.

[0041] The beneficial effects are as follows: In this method, a non-arbitrarily set and determined value of Vmax is provided for actual engineering projects. Compared with the existing method of directly trying out the corresponding parameter values, this greatly reduces the workload and time of trying out the parameter values ​​and improves efficiency. In terms of improving the accuracy of landslide disaster chain simulation, it can provide a scientific value of the maximum particle velocity in the water area when there are no reference cases, which can ensure that the calculation time step size is more reasonable, so as to ensure the accuracy of the numerical simulation results of landslide disaster chain.

[0042] Furthermore, in step S4, the effective radius of any computation time step is updated according to a preset rule, including the following steps:

[0043] Step 4.1: For any computation time step t, calculate the equivalent particle sparsity of the i-th particle. The calculation method is as follows:

[0044] ;

[0045] In the formula, there is , The kernel function, used to calculate the equivalent particle sparsity, is determined by the following formula:

[0046] ;

[0047] In the formula, To calculate the effective radius at time step t-1, iteratively calculate it starting from time t=0. At t=0, the effective radius is the initial value. ;

[0048] Step 4.2, calculate the effective radius update coefficient. The calculation method is as follows:

[0049] ;

[0050] In the formula, d represents the dimension, and d = 1, 2, or 3;

[0051] Step 4.3, Calculate the updated effective radius The calculation formula is:

[0052] ;

[0053] In the formula, To calculate the effective radius at time step t-1, iteratively calculate it starting from time t=0. At t=0, the effective radius is the initial value. .

[0054] The beneficial effects are as follows: By introducing an equivalent particle sparsity, the sparsity of particles at any computation time step t can be quantitatively characterized, preventing particle accumulation and separation phenomena. This ensures that the number of neighboring particles participating in the calculation remains stable when particle accumulation and separation occur in complex flow regions, thereby improving the accuracy of numerical simulation of landslide disaster chains. Furthermore, by introducing an equivalent particle sparsity calculation kernel function, the initial particle arrangement order and structure can be completely ignored, avoiding the problem of missing or distorted neighboring particles when particles are rearranged in complex flow states, which would prevent all neighboring particles from participating in the calculation. This achieves an accurate characterization of particle sparsity, thereby improving the accuracy of numerical simulation of landslide disaster chains. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating an embodiment of the method for improving the accuracy of landslide disaster chains using the smooth particle method according to the present invention.

[0056] Figure 2 This is a two-dimensional model diagram of the landslide disaster chain numerical analysis in an embodiment of the method for improving the accuracy of landslide disaster chain simulation using the smooth particle method of the present invention;

[0057] Figure 3 The figure shows the numerical simulation results of the landslide disaster chain in an embodiment of the method for improving the accuracy of landslide disaster chain simulation using the smooth particle method of the present invention. Detailed Implementation

[0058] The following detailed description provides further details on specific implementation methods.

[0059] Example

[0060] A method to improve the accuracy of smooth particle method simulation of landslide hazard chains, such as Figure 1 As shown, it includes the following steps:

[0061] Step 1: Collect survey data on the target landslide and the impact range of the disaster chain to be simulated, and based on the survey data, clarify the geological conditions, topography, test results of the physical and mechanical parameters of the landslide body, and physical and mechanical parameters of the water body in the target landslide and the impact range of the disaster chain.

[0062] Step 2: Based on existing software, such as Ansys_Autodyne_SPH, construct a numerical analysis model for landslide hazard chain numerical simulation using the smoothed particle method. Specifically:

[0063] Based on the geological conditions and topography described in step 1, a geometric model is constructed. This constructed geometric model includes model boundaries, landslide boundaries, and water body boundaries. The boundary conditions of the geometric model are constructed using mirrored particles or virtual particles. The numerical analysis model is constructed using existing software, resulting in the following... Figure 2The boundary conditions in the model shown in this embodiment are set using virtual particles.

[0064] Based on step 1, the physical and mechanical parameters of the landslide body and the water body are assigned to the corresponding physical and mechanical parameters of each material particle, including the particle density, static shear stress, gravitational acceleration, viscosity, and position of the landslide body particles and the water body particles. Each material particle refers to the particle corresponding to the water body and the landslide body.

[0065] In this embodiment, the density, static shear stress, gravitational acceleration, and viscosity of the water particles are all 1000 kg / m³. 3 0 Pa, 9.8 m / s 2 0.001 Ns / m 2 The density, static shear stress, gravitational acceleration, and viscosity of the landslide particles were 1950 kg / m³. 3 1000 Pa, 9.8 m / s 2 1.0 Ns / m 2 Boundary particles and virtual particles are not assigned parameters.

[0066] Step 3: Based on the numerical analysis model constructed in Step 2, set the initial calculation parameters, including the smoothing kernel function, the calculation time step size, and the initial effective radius, including the following steps:

[0067] Step 3.1, Set the initial effective radius according to the dimensions. The dimension is the spatial dimension, represented as:

[0068] ;

[0069] In the formula, d = 1, 2, 3;

[0070] In this embodiment, simulation is performed in two-dimensional space, i.e., d=2, with an initial effective radius. .

[0071] Step 3.2, set the smoothing kernel function to a cubic spline function, which is expressed as:

[0072] ;

[0073] In the formula, , and This represents two distinct particles where i ≠ j. It represents the distance between two particles. For the effective radius, It is a normalization factor related to dimension d, expressed as: , , ;

[0074] In this embodiment, .

[0075] Step 3.3: Based on computer computing resources and simulation timeliness requirements, set the calculation time step size, and the calculation time step size should meet the following requirements:

[0076] ;

[0077] ;

[0078] ;

[0079] In the formula, For the time step of the numerical simulation, others not mentioned , , Intermediate quantities introduced into the calculation process have no specific physical meaning. The initial distance between adjacent particles in the water body region. Vmax represents the initial distance between adjacent particles in the landslide area; Vmax is the maximum velocity of particles in the water area, which is taken as 4 m / s in practical engineering. This represents the kinematic viscosity of the landslide mass.

[0080] In this embodiment, =0.025 s, =243.75 s, take =0.025s, based on the computer's computing power resources (CPU I9-12900h), the simulated computation time is 6 hours, and the computation time step is taken as . =0.01s.

[0081] The computational domains of the landslide body and water body are discretized into a predetermined number of particles with different physical properties, including:

[0082] Discretize the landslide area enclosed by the landslide boundary in the geometric model constructed in step 2, and uniformly insert landslide particles.

[0083] Discretize the water body region enclosed by the water body boundary in the geometric model constructed in step 2, and uniformly insert water body particles.

[0084] Step 4: Begin the simulation calculation. After updating the effective radius of any calculation time step according to the preset rules, continue the calculation for the next time step until all calculations are completed, and obtain the numerical simulation results of the landslide hazard chain, such as... Figure 3 As shown.

[0085] In step S4, the effective radius of any computation time step is updated according to a preset rule, including the following steps:

[0086] Step 4.1: For any computation time step t, calculate the equivalent particle sparsity of the i-th particle. The calculation method is as follows:

[0087] ;

[0088] In the formula, there is , The kernel function, used to calculate the equivalent particle sparsity, is determined by the following formula:

[0089] ;

[0090] In the formula, To calculate the effective radius at time step t-1, iteratively calculate it starting from time t=0. At t=0, the effective radius is the initial value. ;

[0091] Step 4.2, calculate the effective radius update coefficient. The calculation method is as follows:

[0092] ;

[0093] In the formula, d represents the dimension, and d = 1, 2, or 3;

[0094] In this embodiment, d=2.

[0095] Step 4.3, Calculate the updated effective radius The calculation formula is:

[0096] ;

[0097] In the formula, To calculate the effective radius at time step t-1, iteratively calculate it starting from time t=0. At t=0, the effective radius is the initial value. .

[0098] Compared with existing technologies, this embodiment uses the smoothed particle method to simulate the effective radius in real time during the landslide disaster chain process. First, based on the established numerical analysis model, the smoothing kernel function, the calculation time step size, and the initial effective radius are calculated. Then, based on preset rules, the equivalent particle sparsity is calculated to distinguish the particle accumulation and separation phenomena in areas with a large amount of turbulence and splashes during the fluid movement of the landslide disaster chain. By introducing the equivalent particle sparsity calculation kernel function, the initial particle arrangement order and structure can be completely ignored, avoiding the problem that the absence or distortion of neighboring particles during particle rearrangement under complex flow conditions will prevent all neighboring particles from participating in the calculation. The effective radius update coefficient is defined and calculated based on the equivalent particle sparsity, and the effective radius is updated by the effective radius update coefficient. This prevents the calculation error from increasing due to drastic changes in the number and range of neighboring particles, thus achieving an accurate characterization of particle sparsity and improving the accuracy of numerical simulation of landslide disaster chains.

[0099] The above descriptions are merely embodiments of the present invention, and common knowledge regarding specific structures and characteristics is not elaborated upon here. It should be noted that those skilled in the art can make various modifications and improvements without departing from the structure of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for improving the accuracy of a smooth particle method simulation of a landslide disaster chain, comprising the following steps: Step 1: Collecting survey data of a target landslide and a disaster chain affected area to be simulated, and based on the survey data, determining the geological conditions, topography, test results of physical and mechanical parameters of the landslide body, and physical and mechanical parameters of the water body of the target landslide and the disaster chain affected area; Step 2: Constructing a numerical analysis model for numerical simulation of the landslide disaster chain based on the smooth particle method; characterized in that it further comprises: Step 3: Setting initial calculation parameters based on the numerical analysis model constructed in Step 2, the initial calculation parameters including a smoothing kernel function, a calculation time step size, and an initial effective radius; Step 4: Starting the simulation calculation, after updating the effective radius of any one calculation time step according to a predetermined rule, continuing the calculation of the next time step until the entire calculation is completed, and obtaining the numerical simulation results of the landslide disaster chain; updating the effective radius of any one calculation time step according to a predetermined rule comprises the following steps: Step 4.1, For any computational time step t, compute the equivalent particle sparsity of the ith particle , computed as follows: ； wherein , The kernel function for the equivalent particle dilution level calculation is determined by the following equation: ； In the formula, The effective radius at the time step t-1 is calculated by iteration from the time t=0, and t=0 is the initial effective radius ; Step 4.2, calculating the effective radius update coefficient The calculation is as follows: ； where d represents the dimension, d = 1, 2, 3; Step 4.3, Calculate the updated effective radius The formula is: ； In the formula, The effective radius at time step t-1 is calculated by iteration from t=0, and t=0 is the initial effective radius .

2. The method for improving the accuracy of simulating landslide disaster chain by the smooth particle hydrodynamics method according to claim 1, characterized in that: In Step 2, a geometric model is constructed based on the geological conditions and topography described in Step 1; the calculation area of the landslide body and the water body is discretized into a set number of particles with different material state attributes; and based on Step 1, the physical and mechanical parameters of the landslide body and the water body are assigned to the corresponding physical and mechanical parameters of each material state particle.

3. The method of claim 2, wherein the method is characterized by: In Step 2, the constructed geometric model includes model boundaries, landslide body boundaries, and water body boundaries.

4. The method of claim 3, wherein the method is characterized by: The boundary conditions of the geometric model are constructed by setting mirror particles or virtual particles.

5. The method for improving the accuracy of simulating landslide disaster chain by the smooth particle hydrodynamics method according to claim 4, characterized in that: Discretizing the calculation area of the landslide body and the water body into a set number of particles with different material state attributes comprises: internally discretizing the landslide body region enclosed by the landslide body boundaries in the geometric model constructed in Step 2, and uniformly inserting landslide body particles; internally discretizing the water body region enclosed by the water body boundaries in the geometric model constructed in Step 2, and uniformly inserting water body particles.

6. The method of claim 2, wherein the method is characterized by: In Step 2, assigning the physical and mechanical parameters of the landslide body and the water body to the corresponding physical and mechanical parameters of each material state particle includes the particle density, static shear stress, gravitational acceleration, viscosity, and position of the landslide body particles and the water body particles.

7. The method of claim 1, wherein the method is characterized by: In Step 3, setting the initial calculation parameters, including the smoothing kernel function, the calculation time step size, and the initial effective radius, comprises the following steps: Step 3.1, setting initial effective radius according to dimension is represented as: ； where d = 1, 2, 3; Step 3.2: Setting the smoothing kernel function to a cubic spline function, which is represented as: ； wherein , represents the distance of two particles, is the effective radius, is a normalization factor related to the dimension d, expressed as: , , ; Step 3.3: Setting the calculation time step size based on computer computing resources and simulation timeliness requirements, and the calculation time step size should satisfy the following requirements: ； ； ； wherein is the initial distance of adjacent particles in the water body region, is the initial distance of adjacent particles in the landslide body region; Vmaxis the maximum velocity of particles in the water body region, is the viscosity of the landslide body movement.

8. The method for improving the accuracy of simulating landslide disaster chain by the smooth particle hydrodynamics method according to claim 7, characterized in that: In Step 3.3, Vmax = 4 m / s is taken in actual engineering.

Images