A fast discrete element modeling method for large-capacity regularly arranged structures

Abstract

This application discloses a discrete element rapid modeling method suitable for large-capacity regularly arranged structures, which realizes the application of discrete element technology to large-capacity geotechnical structure modeling, especially multi-layer system structure model. In the model construction process, the large-capacity model is first divided into several sections by splitting and then merging, which improves the efficiency of skeleton structure construction; then, through the discrete element search area control algorithm, the traversal range of particle units is narrowed in a targeted manner, which greatly improves the efficiency of mixture model construction while ensuring the accuracy of model calculation. This method expands the application scope of discrete elements and fills the application gap of this technology in large-capacity models.

Description

Technical Field

[0001] The present application relates to the field of numerical simulation technology, in particular to the numerical modeling technology of composite materials such as asphalt concrete and cement concrete, and specifically to a discrete element rapid modeling method suitable for large-capacity regularly arranged structures. Background Art

[0002] Discrete element numerical simulation technology has been widely used in civil engineering and other related fields as an important means to study the mechanical behavior of discrete media. This method overcomes the macroscopic continuity assumption of traditional continuous medium mechanics, simulates the mechanical properties of heterogeneous materials from a microscopic level, and reveals the complex mechanical behavior between particles. It has unique advantages in dealing with large deformation, discontinuity and other problems of discontinuous media.

[0003] In the process of model construction, in order to truly simulate the microstructure of the material, rigid clusters clumps generated with aggregate morphology as a template are placed within the model according to the aggregate grading. The coarse aggregate clump skeleton structure is established through the mixture generation control algorithm, and then the unbreakable rigid clusters clumps are converted into breakable flexible clusters cluster, thus completing the core content of geometric modeling.

[0004] At present, in the field of civil engineering, the discrete element method is often used to simulate indoor tests, such as uniaxial compression test, indirect tension test, semicircular bending test, etc. The size of such models is mostly within a dozen centimeters, and the model capacity is often controlled at tens of thousands of particle units. The motion law of particles in discrete systems is analyzed by the discrete element method, aiming to explain the deformation behavior and failure mechanism of materials under load. However, the discrete element method has not yet been applied to structural analysis and large-capacity model construction, which limits its application potential in a wider range of practical engineering problems.

[0005] On the other hand, due to the performance limitations of mainstream computing devices, there are two prominent problems in constructing discrete element models of large particle capacity, which seriously restrict the application of discrete element method in large-capacity particle simulation. First, the skeleton generation process of the discrete element model will consume a lot of time for real-time monitoring and balancing the internal stress of all particles in the model. When a certain number of rigid clusters are put in, the process efficiency is very low, and even the equipment cannot calculate; second, in the process of converting rigid clusters into flexible clusters, the judgment logic is that each pebble cycle judges the distance with all balls in the model, and identifies the particles with a center distance (ball center distance) less than the pebble radius as aggregate particles. The whole process increases exponentially with the expansion of the model capacity, which seriously reduces the modeling efficiency. The above two problems make it difficult for the current discrete element to construct large-capacity models. It is necessary to establish a discrete element fast modeling method suitable for large-capacity particles. Summary of the invention

[0006] In order to solve the above technical problems, the embodiment of the present application provides a discrete element rapid modeling method suitable for large-capacity regularly arranged structures, so as to solve the problem that the existing discrete element modeling technology cannot realize the rapid construction of large-capacity models, and help to promote the application of discrete element simulation technology in large-capacity models.

[0007] In order to achieve the above object, the present invention provides a discrete element rapid modeling method suitable for large-capacity regularly arranged structures, comprising:

[0008] Step S1: Determine the basic size of the model according to research requirements, and divide the original model into several sections based on preset principles;

[0009] Step S2: In the sections divided by the original model, an unprocessed section is selected and a coarse aggregate skeleton structure is established in the section using rigid cluster clump units according to aggregate gradation;

[0010] Step S3: Generate a number of ball units with the same particle size and regular arrangement in the entire model range, retain only the ball units in the segment selected in step S2 and delete the ball units in other segments of the model;

[0011] Step S4: Based on the discrete element search area control algorithm, the relative position of each pebble unit and the ball unit in the adjacent range is cyclically determined, so that all ball units contained in the range of each clump unit obtain the same group information, and the clump unit is converted into a flexible cluster unit, and then the clump unit is immediately deleted; wherein the basic constituent unit of the rigid cluster clump unit is the pebble unit, and the basic constituent unit of the flexible cluster unit is the ball unit;

[0012] Step S5: Obtain valid information of all ball units in step S4, and then delete all ball units in the segment. The valid information at least includes coordinate information of each axis, ball radius, group information and ID information;

[0013] Step S6: Repeat steps S2 to S5 to traverse all the sections and record the valid information of the ball units in all the sections;

[0014] Step S7: Based on the valid information of the ball units in all the sections in step S6, regenerate the second ball units at the corresponding positions in the entire model range, each second ball unit corresponds to a ball unit, and each second ball unit ball contains at least ID data and group information, so as to construct a discrete element geometric model with particle grouping information;

[0015] Step S8: define the contact model and microscopic parameters for the discrete element geometric model to complete the construction of the discrete element model.

[0016] Furthermore, the preset principle in step S1 includes taking material mix ratio as a division standard or taking preset structural layer distribution as a division standard or taking user's will as a division standard to form a final segmentation model.

[0017] Further, step S2 includes:

[0018] Starting from any section, rigid clusters are placed according to aggregate grading, and the coarse aggregate skeleton structure needs to be generated through self-weight deposition algorithm, compaction molding algorithm or volume expansion algorithm.

[0019] Furthermore, the arrangement rule in step S3 includes a tetragonal arrangement structure or a hexagonal arrangement structure.

[0020] Furthermore, the discrete element search region control algorithm in step S4 includes:

[0021] Step S41: obtaining a coarse aggregate skeleton structure established by rigid cluster clump units and ball units generated from an origin position according to an arrangement rule; the origin position is (0, 0) in a 2D model and (0, 0, 0) in a 3D model;

[0022] Step S42: Calculate the ID data of each ball unit according to the arrangement rule;

[0023] Step S43: traverse and obtain the center coordinates of the pebble unit and the ball unit respectively;

[0024] In the 2D model, the center coordinates of the pebble unit (x p ,y p ), the coordinates of the center of the ball unit (x b ,y b );

[0025] In the 3D model, the center coordinates of the pebble unit (x p ,y p , z p ), the coordinates of the center of the ball unit (x b ,y b , z b );

[0026] Step S44: based on the geometric shape corresponding to the adopted arrangement rule, the thickness of the ball unit is obtained; wherein, the average thickness of each ball unit in each row corresponding to the x and y directions is obtained in the 2D model, and the average thickness of each ball unit in each row and each layer corresponding to the x, y and z directions is obtained in the 3D model;

[0027] Step S45: based on the coordinates of the center of the pebble unit and the thickness of the ball unit, the number of particles corresponding to the position of each pebble unit in each axial direction is obtained, and the position correspondence between the pebble unit and the adjacent ball unit is formed after rounding; wherein, rounding up, rounding down, rounding off, or random rounding is adopted when rounding;

[0028] In the 2D model, get the number of particles t corresponding to the x direction at each pebble unit location x , the number of rows corresponding to the y direction t y ;

[0029] In the 3D model, calculate the number of particles t corresponding to the x direction at each pebble unit location x , the number of rows corresponding to the y direction t y , the number of layers corresponding to the z direction t z ;

[0030] Step S46: Setting up the search area extension range, based on the calculated t x and t y Construct a search area for similar ball unit particles; in the 3D model, based on the calculated t x ,t y and t z Construct a search area for similar ball unit particles;

[0031] Step S47: Obtain the ID data of the ball unit within the constructed search area, and traverse and calculate the search area of ​​each pebble unit.

[0032] Furthermore, in step S42, the ID data for the quadrilateral arrangement structure is calculated as follows:

[0033] Get the number of balls per row in the x direction in the 2D model x , ID data b ID =(t y -1)×n x +t x ;

[0034] Get the number of balls per row in the x direction in the 3D model xThe number of balls per layer in the z direction n z , ID data b ID =(t z -1)×n z +(t y -1)×n x +t x ;

[0035] Where: t x is the number of balls in the row corresponding to the ball in the x direction; t y is the row number corresponding to the ball in the y direction; t z is the number of layers corresponding to the ball in the z direction; n x is the number of balls in each row in the x direction; n z is the number of balls per layer in the z direction; b ID The ID corresponding to the ball.

[0036] Furthermore, in step S42, the ID data for the hexagonal arrangement structure is calculated as follows:

[0037] ID data under 2D model:

[0038] If the ball is located in an odd row b ID =(t y -1) / 2×(n x1 +n x2 )+t x ;

[0039] If the ball is located in an even row b ID =(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ;

[0040] ID data under 3D model:

[0041] If the ball is located in an odd layer and an odd row b ID =(t z -1) / 2×(n z1 +n z2 )+(t y -1) / 2×(n x1 +n x2 )+t x ;

[0042] If the ball is located in an odd layer and an even row b ID =(t z -1) / 2×(nz1 +n z2 )+(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ;

[0043] If the ball is located in an even layer and an odd row b ID =(t z / 2-1)×(n z1 +n z2 )+n z1 +(t y -1) / 2×(n x1 +n x2 )+t x ;

[0044] If the ball is located in an even layer and an even row b ID =(t z / 2-1)×(n z1 +n z2 )+n z1 +(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ;

[0045] Where: t x is the number of balls in the row corresponding to the ball in the x direction; t y is the row number corresponding to the ball in the y direction; t z is the number of layers corresponding to the ball in the z direction; n x1 n is the number of balls in odd rows in the x direction; x2 is the number of balls in even rows in the x direction; n z1 is the number of balls in the odd layers in the z direction; n z2 is the number of balls in even layers in the z direction; b ID The ID corresponding to the ball.

[0046] Further, in step S44:

[0047] In the 2D model, the thickness dimension x of each ball unit in the tetragonal arrangement structure in the x direction is dis =2r, thickness dimension y in the y direction dis =2r;

[0048] The thickness dimension x of each ball unit in the hexagonal arrangement structure in the x direction dis=2r, thickness dimension y in the y direction dis = r;

[0049] In the 3D model, the thickness dimension x of each ball unit in the tetragonal arrangement structure in the x direction is dis =2r, thickness dimension y in the y direction dis =2r, thickness dimension z in z direction dis =2r;

[0050] The thickness dimension x of each ball unit in the hexagonal arrangement structure in the x direction dis =2r, thickness dimension y in the y direction dis = r, thickness dimension z in z direction dis = r;

[0051] Where r is the average radius of the ball unit.

[0052] Further, in step S46, t x ,t y ,t z Based on the above, at least two numbers are floated as the new search area for similar ball unit particles. x ,t y ,t z Reduce i x 、i y 、i z and increase u x 、u y 、u z， Among them, i x 、i y 、i z 、u x 、u y 、u z are all integers ≥ 2, and the x-axis search area range (t x -i x )~(t x +u x ), y-axis search area range (t y -i y )~(t y +u y ), z-axis search area range (t z -i z )~(t z +u z ).

[0053] Beneficial effects of the present invention:

[0054] 1. Provide a model that can establish a large-capacity regular arrangement structure to fill the gap in existing technology;

[0055] 2. In the process of model construction, the large-capacity model is first divided into several sections by using the method of splitting first and then merging, which improves the efficiency of skeleton structure construction;

[0056] 3. Through the discrete element search area control algorithm, the traversal range of the particle unit is targetedly narrowed, which greatly improves the efficiency of mixture model construction while ensuring the accuracy of model calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 It is a flowchart of the fast modeling method of regularly arranged discrete elements applicable to large capacity of the present application;

[0059] Figure 2 This is a schematic diagram of the research object of the embodiment of the present application (the size of the asphalt pavement is 600×240×740mm);

[0060] Figure 3 This is a schematic diagram of a coarse aggregate clump model in a 3.0% cement-stabilized crushed stone section of the subbase of an embodiment of the present application;

[0061] Figure 4 (a) is a schematic diagram of the tetragonal arrangement structure of the present application;

[0062] Figure 4 (b) is a schematic diagram of the hexagonal arrangement structure of the present application;

[0063] Figure 5 It is a schematic diagram of the process of generating balls within the range of the application;

[0064] Figure 6 It is a schematic diagram of the relationship between the clump unit and the pebble unit, and the cluster unit and the ball unit of the present application;

[0065] Figure 7 (1) is a schematic diagram of the enlargement 2 of the pebble search area of ​​the present application;

[0066] Figure 7 (2) is a schematic diagram of the pebble search area expansion 5 of the present application;

[0067] Figure 7 (3) is a schematic diagram of the pebble search area of ​​the present application being enlarged by 10;

[0068] Figure 8 It is a schematic diagram of the discrete element model of asphalt pavement in this application. DETAILED DESCRIPTION

[0069] In order to make the purpose, features, and advantages of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the embodiments described below are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0070] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments.

[0071] In the description of the present application, it should be understood that the terms "upper", "lower", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0072] The present invention provides a discrete element rapid modeling method suitable for large-capacity regularly arranged structures. The entire flow chart is shown in Figure 1 ,include:

[0073] Step S1: Determine the basic size of the model according to the research requirements, and divide the original model into several sections based on the preset principles. The preset principles in this step include using the material mix ratio as the division standard, using the preset structural layer distribution as the division standard, or using the user's wishes as the division standard to form the final segmented model. Among them, when the material mix ratio is used as the division standard, although there are many types, in specific application scenarios, in most cases, there are only one or several material mix ratios. By dividing it into several sections, the number of balls in each section is controlled within a reasonable range, so that the efficiency of skeleton model construction is improved.

[0074] Step S2: In the sections divided by the original model, an unprocessed section is selected and a coarse aggregate skeleton structure is established in the section using rigid cluster clump units according to aggregate gradation.

[0075] As a preferred solution, when establishing the coarse aggregate skeleton structure, rigid cluster clump units are placed from any section according to the aggregate grading, and the coarse aggregate skeleton structure needs to be generated by a self-weight deposition algorithm, a compaction molding algorithm, or a volume expansion algorithm. This is a prior art and will not be described in detail. For details, please refer to the following:

[0076] 1. “Self-weight deposition algorithm” - Jin Lei, Zeng Yawu, Ye Yang, et al. Improvement of three-dimensional discrete element modeling method for irregular particles and their aggregates [J]. Chinese Journal of Geotechnical Engineering, 2017, 39(07): 1273-1281.

[0077] 2. “Compaction Algorithm” - He Liang, Zhou Zidong, VAN DEN BERGH Wim, et al. Discrete Element Simulation of Blockage Law of Porous Asphalt Mixture [J]. Journal of Transportation Engineering, 2023, 23(02): 78-91.

[0078] 3. “Volume Expansion Algorithm” – Zhang Wei, Xiao Weijian, Yuan Chuanniu, et al. Effect of particle size distribution on force chain evolution mechanism in iron powder compaction using discrete element method [J]. Chinese Journal of Mechanics, 2022, 54(09): 2489-2500.

[0079] Step S3: Generate several ball units with the same particle size and regular arrangement in the entire model range, and only keep the ball units in the segment selected in step S2 and delete the ball units in other segments of the model. This ensures that the IDs of all ball units will not be repeated in subsequent calculations. If ball units are directly generated in each segment, there may be ball units with the same ID in different segments, affecting subsequent data analysis.

[0080] As a preferred solution, the arrangement rule includes a tetragonal arrangement structure or a hexagonal arrangement structure, such as Figure 4 As shown, in the tetragonal arrangement structure, the ball units are regularly arranged in a horizontal and vertical manner, and in the hexagonal arrangement structure, the ball units are arranged in a honeycomb structure, with six ball units evenly distributed around one ball unit.

[0081] Step S4: Based on the discrete element search area control algorithm, the relative position of each pebble unit and the ball unit is cyclically determined, so that all ball units contained in the range of each clump unit obtain the same group information, and the clump unit is converted into a flexible cluster unit, and then the clump unit is immediately deleted; wherein, the basic constituent unit of the rigid cluster clump unit is the pebble unit, and the basic constituent unit of the flexible cluster unit is the ball unit.

[0082] Taking into account the needs of different models, especially for personal research needs, while meeting the research needs, the use of 2D models can pursue larger model scales or faster calculation speeds. 3D models take up more computing resources than 2D models. If a "tens of millions" particle capacity model is built in 3D, then a "tens of millions" particle capacity model can be built in 2D. The following describes the two cases of 2D models and 3D models respectively:

[0083] In the case of a 2D model, the discrete element search region control algorithm in step S4 includes:

[0084] Step S41: obtaining a coarse aggregate skeleton structure established by a rigid cluster clump unit and a ball unit generated from the origin position (0, 0) according to an arrangement rule;

[0085] Step S42: Calculate the ID data of each ball unit according to the arrangement rule. The ID data of the square arrangement structure is calculated as follows:

[0086] Get the number of balls per row in the x direction n x , ID data b ID =(t y -1)×n x +t x ;

[0087] Where: t x is the number of balls in the row corresponding to the ball in the x direction; t y is the number of rows corresponding to the ball in the y direction; n x is the number of balls in each row in the x direction; b ID The ID corresponding to the ball.

[0088] The ID data for the hexagonal structure is calculated as follows:

[0089] If the ball is located in an odd row b ID =(t y -1) / 2×(n x1 +n x2 )+t x ;

[0090] If the ball is located in an even row b ID =(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ;

[0091] Where: t xis the number of balls in the row corresponding to the ball in the x direction; t y is the number of rows corresponding to the ball in the y direction; n x1 n is the number of balls in odd rows in the x direction; x2 b is the number of balls in even rows in the x direction; ID The ID corresponding to the ball.

[0092] Step S43: traverse and obtain the center coordinates of the pebble unit and the ball unit respectively. The center coordinates of the pebble unit (x p ,y p ), the coordinates of the center of the ball unit (x b ,y b );

[0093] Step S44: Based on the geometric shape corresponding to the adopted arrangement rule, the thickness of the ball unit is obtained; wherein, the average thickness of each ball unit in each row corresponding to the x and y directions in the 2D model is obtained, where r is the average radius of the ball unit in the 2D model.

[0094] The thickness dimension x of each ball unit in the x direction in the tetragonal arrangement structure is dis =2r, thickness dimension y in the y direction dis =2r;

[0095] The thickness dimension x of each ball unit in the x direction in the hexagonal arrangement structure is dis =2r, thickness dimension y in the y direction dis = r.

[0096] Step S45: based on the coordinates of the center of the pebble unit and the thickness of the ball unit, the number of particles corresponding to the position of each pebble unit in each axial direction is obtained, and the position correspondence between the pebble unit and the adjacent ball unit is formed after rounding; wherein, rounding up, rounding down, rounding off, or random rounding is adopted when rounding;

[0097] In the 2D model, get the number of particles t corresponding to the x direction at each pebble unit location x , the number of rows corresponding to the y direction t y ;

[0098] Step S46: Setting up the search area extension range, based on the calculated t x and t y Construct a search area for similar ball unit particles, with t x ,t yBased on at least two numbers floating up and down as the new similar ball unit particle search area, for example, in this embodiment, for t x ,t y Reduce i x 、i y and increase u x 、u y , where i x 、i y 、u x 、u y are all integers ≥ 2, and the x-axis search area range (t x -i x )-(t x +u x ), y-axis search area range (t y -i y )-(t y +u y ).

[0099] Step S47: Obtain the ID data of the ball unit within the constructed search area, and traverse and calculate the search area of ​​each pebble unit.

[0100] In the case of a 3D model, the discrete element search region control algorithm in step S4 includes:

[0101] Step S41: obtaining a coarse aggregate skeleton structure established by a rigid cluster clump unit and a ball unit generated from the origin position (0, 0, 0) according to an arrangement rule;

[0102] Step S42: Calculate the ID data of each ball unit according to the arrangement rule;

[0103] The ID data for the tetragonal arrangement is calculated as follows:

[0104] Get the number of balls per row in the x direction n x The number of balls per layer in the z direction n z , ID data b ID =(t z -1)×n z +(t y -1)×n x +t x ;

[0105] Where: t x is the number of balls in the row corresponding to the ball in the x direction; t y is the row number corresponding to the ball in the y direction; t z is the number of layers corresponding to the ball in the z direction; n xis the number of balls in each row in the x direction; n z is the number of balls per layer in the z direction; b ID The ID corresponding to the ball.

[0106] The ID data for the hexagonal structure is calculated as follows:

[0107] If the ball is located in an odd layer and an odd row, ID data b ID =(t z -1) / 2×(n z1 +n z2 )+(t y -1) / 2×(n x1 +n x2 )+t x ;

[0108] If the ball is located in an odd layer and an even row, ID data b ID =(t z -1) / 2×(n z1 +n z2 )+(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ;

[0109] If the ball is located in an even layer and an odd row, ID data b ID =(t z / 2-1)×(n z1 +n z2 )+n z2 +(t y -1) / 2×(n x1 +n x2 )+t x ;

[0110] If the ball is located in an even layer and an even row, ID data b ID =(t z / 2-1)×(n z1 +n z2 )+n z2 +(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ;

[0111] Where: t x is the number of balls in the row corresponding to the ball in the x direction; t yis the row number corresponding to the ball in the y direction; t z is the number of layers corresponding to the ball in the z direction; n x1 n is the number of balls in odd rows in the x direction; x2 is the number of balls in even rows in the x direction; n z1 is the number of balls in the odd layers in the z direction; n z2 is the number of balls in even layers in the z direction; b ID The ID corresponding to the ball.

[0112] Step S43: traverse and obtain the center coordinates of the pebble unit and the ball unit respectively. The center coordinates of the pebble unit (x p ,y p , z p ), the coordinates of the center of the ball unit (x b ,y b , z b );

[0113] Step S44: Based on the geometric shape corresponding to the adopted arrangement rule, obtain the thickness of the ball unit (the thickness is mainly for the arrangement to take into account the space occupied by the gaps between the ball units), and obtain the average thickness of each ball unit, each row, and each layer in the 3D model in the x, y, and z directions, where r is the average radius of the ball unit; dis is the diameter of the ball unit in the x direction, dis is the diameter of the ball unit in the y direction, z dis is the diameter of the ball unit in the z direction;

[0114] The thickness dimension x of each ball unit in the x direction in the tetragonal arrangement structure is dis =2r, thickness dimension y in the y direction dis =2r, thickness dimension z in z direction dis =2r;

[0115] The thickness dimension x of each ball unit in the x direction in the hexagonal arrangement structure is dis =2r, thickness dimension y in the y direction dis = r, thickness dimension z in z direction dis = r;

[0116] Step S45: Based on the coordinates of the center of the pebble unit and the thickness of the ball unit, the number of particles corresponding to each pebble unit in each axial direction is obtained, and the position correspondence between the pebble unit and the adjacent ball unit is formed after rounding; wherein, rounding up, rounding down, rounding off, or random rounding is adopted when rounding; the number of particles corresponding to each pebble unit in the x direction is calculated. x , the number of rows corresponding to the y direction t y , the number of layers corresponding to the z direction t z ;

[0117] Step S46: Establishing a search area expansion range based on the calculated t x ,t y and t z Construct a search area for similar ball unit particles, with t x ,t y ,t z As a basis, at least two numbers are floated up and down as the new search area for similar ball unit particles. x ,t y ,t z Reduce i x 、i y 、i z and increase u x 、u y 、u z， Among them, i x 、i y 、i z 、u x 、u y 、u z are all integers ≥ 2, and the x-axis search area range (t x -i x )~(t x +u x ), y-axis search area range (t y -i y )~(t y +u y ), z-axis search area range (t z -i z )~(t z +u z ).

[0118] Step S47: Obtain the ID data of the ball unit within the constructed search area, and traverse and calculate the search area of ​​each pebble unit.

[0119] Step S5: Obtain valid information of all ball units in step S4, and then delete all ball units in the segment. The valid information at least includes coordinate information of each axis, ball radius, group information and ID information;

[0120] Step S6: Repeat steps S2 to S5 to traverse all the sections and record the valid information of the ball units in all the sections;

[0121] Step S7: Based on the valid information of the ball units in all segments in step S6, the second ball units are regenerated at the corresponding positions within the entire model range. Each second ball unit corresponds to a ball unit, and each second ball unit ball contains at least ID data and group information to construct a discrete element geometric model with particle grouping information.

[0122] Here, because the data of other segments are deleted in the process of continuously traversing the segments in step S6 in order to improve the calculation efficiency, it is necessary to collect all the valid information retained in step S6 in this step, record them in the file in ID order, and generate ball unit data covering the entire model at one time for subsequent use.

[0123] Step S8: define the contact model and microscopic parameters for the discrete element geometric model (the definition of the model here can be completed with the help of software and will not be elaborated on), and output to form the required discrete element model.

[0124] A specific embodiment is provided in this application for comparative analysis with the prior art:

[0125] like Figure 2 As shown, the research object of this embodiment is a semi-rigid base asphalt pavement.

[0126] Step S1: Rationally divide the model into several sections. The semi-rigid base asphalt pavement is initially divided into five sections according to the mixture type of each structural layer. Considering the actual model size and the different material mix ratios from bottom to top during actual paving, and considering that the 4.3% cement stabilized gravel section is large in size, the coarse aggregate skeleton generation process will still be very difficult, and this section will be further divided into the upper base section and the lower base section with reference to the structural layer.

[0127] Therefore, in this embodiment, the discrete element model is proposed to be divided into 6 sections, namely the upper layer SMA-13 ​​asphalt mixture section, the middle layer SUP-20 asphalt mixture section, the lower layer SUP-25 asphalt mixture section, the upper base layer 4.3% cement stabilized gravel section, the lower base layer 4.3% cement stabilized gravel section, and the bottom base layer 3.0% cement stabilized gravel section.

[0128] The computation time consumed by this process is negligible.

[0129] Step S2: Use rigid cluster clump units to construct the coarse aggregate skeleton structure in any section according to the material gradation. Figure 3 As shown, the subsequent operation process is explained using the 3.0% cement stabilized gravel section of the subbase as an example.

[0130] Aggregates with a particle size greater than or equal to 4.75 mm are selected as coarse aggregates, and aggregates less than 4.75 mm and cement mortar are considered as cement mortar. The data file of the true three-dimensional morphology of the coarse aggregate is scanned by three-dimensional reconstruction technology. The volume proportion of coarse aggregates of various particle sizes is calculated according to the mix ratio of 3.0% cement-stabilized crushed stone. The discrete element PFC software is used to place the coarse aggregate clump unit in the upper layer range with the above data file as a template, and the volume expansion algorithm is used to finally form a 3.0% cement-stabilized crushed stone mixture for the base layer. The length, width and height of this section are 600×240×200mm, respectively, and a total of 3,029,670 coarse aggregate clump units are placed.

[0131] This step can realize the coarse aggregate skeleton generation process for each layer in the subsequent traversal. The calculation time comparison results are shown in Table 1 below.

[0132] Table 1 Comparison of calculation time of coarse aggregate clump skeleton

[0133]

[0134] Step S3: Generate ball units with the same particle size and regular arrangement within the model range, and only retain the ball units in the current section after the deletion operation.

[0135] In this embodiment, the radius of the ball unit is set to 1.25 mm, and the ball units are arranged in a hexagonal arrangement structure and are fully distributed within the model range. Figure 4 As shown in the figure, delete the upper layer, middle layer, etc., and only keep the ball unit in the base section, and define the default group information of the ball unit as cement mortar, such as Figure 5 shown.

[0136] In this way, only the ball units in the base segment are retained and other redundant data are deleted, which can greatly optimize the subsequent operation time without causing multiple calculations of data due to the existence of other segments.

[0137] The computation time consumed by this process is negligible.

[0138] Step S4: Circularly determine the relative positions of the basic pebble units in the clump unit (each clump unit is composed of several pebble units, and there are positions where the pebble units intersect with each other) and the ball units in the segment (the ball units are arranged regularly and there is no intersection) so that the ball units contained in the clump unit range obtain the same group information and are converted into flexible cluster units. That is, this step is equivalent to calculating the relative position relationship between the pebble units in each clump and the ball units arranged according to the established rules after generating the clump unit, identifying the ball units contained in the clump, and converting the clump unit into a flexible cluster unit for subsequent simulation.

[0139] In this embodiment, a 3D model is constructed, specifically:

[0140] Step S41: Arrange the ID assignment rules of the model ball units according to the rules, and record the number of balls in each row in the y direction n y is 240, the number of balls per layer in the z direction is n z is 23040;

[0141] Step S42: setting ID data for each ball unit according to the arrangement rule;

[0142] Step S43: traverse the center coordinates (x p ,y p , z p ) and (x b ,y b , z b );

[0143] Step S44: Calculate the thickness x of each layer of ball units in the x, y, and z directions. dis ,y dis 、z dis , that is, x dis =y dis =z dis =2*0.00125=0.0025m;

[0144] Step S45: based on the coordinates of the center of the pebble unit and the thickness of the ball unit, the number of particles corresponding to the position of each pebble unit in each axial direction is obtained, and the position correspondence relationship between the pebble unit and the adjacent ball unit is formed after rounding;

[0145] Taking the pebble unit ID of 6035562 as an example, the coordinates of the sphere center are (0.22375, 0.115687, 0.16435), and the number of particles corresponding to this pebble unit position in the x direction is t x =89, the number of rows corresponding to t in the y direction y =46, corresponding to the number of layers in the z direction t z =65.

[0146] Step S46: The search area ranges of the similar ball unit particles are set to i x =i y =i z =u x =u y =u z =2 or 5 or 10.

[0147] In this embodiment, the search area range is as follows: Figure 7 As shown, red indicates that the search range increases successively. When the search range is 2, the number of balls is 983, when the search range is 5, the number of balls is 4247, and when the search range is 10, the number of balls is 18327.

[0148] Step S47: The ID of the ball unit adjacent to the pebble unit is 65*23040+46*240+89=1508729. Similarly, the IDs of all ball units in the search area can be easily obtained. The calculation time comparison results are shown in Table 2.

[0149] Table 2 Comparison of calculation time for converting coarse aggregate clump unit to cluster unit (taking search area 2 as an example)

[0150]

[0151] Step S5: Output the coordinate information, group information, ID information, etc. of all balls.

[0152] The computation time of this process is negligible.

[0153] Step S6: Repeat the above operation to output the ball information of the upper SMA-13 ​​asphalt mixture section, the middle SUP-20 asphalt mixture section, the lower SUP-25 asphalt mixture section, the upper base 4.3% cement stabilized gravel section, and the lower base 4.3% cement stabilized gravel section.

[0154] The computation time of this process is negligible.

[0155] Step S7: regenerate a ball (i.e., the second ball unit mentioned above) at a corresponding position according to the above information of the ball, and together form a discrete element geometric model endowed with group information.

[0156] like Figure 8 As shown, import the previous information, use the imported radius r, spatial coordinates (x, y, z), and ID data information to define the ball unit properties within the model range, and then complete the construction of the discrete element geometry model.

[0157] The model has a length, width and height of 600×240×740 mm and contains a total of 6,819,840 balls.

[0158] The computation time of this process is negligible.

[0159] Step S8: Define the contact model and microscopic parameters for the model, thus completing the discrete element model.

[0160] Both the old and new methods need to perform this step, so this process does not participate in the calculation time comparison.

[0161] The calculation time comparison of the complete process is shown in Table 3 below.

[0162] Table 3 Comparison of calculation time between the new and old methods

[0163]

[0164] From the above case analysis, it can be seen that basically each step is much faster than the existing algorithm, so we can see the advantages of this application in the calculation of large models.

[0165] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations (such as quantity, shape, position, etc.) can be made to the technical scheme of the present invention, and these equivalent transformations are all protected by the present invention.

Claims

1. A discrete element rapid modeling method suitable for large-capacity regularly arranged structures, characterized in that: include: Step S1: Determine the basic size of the model according to research requirements, and divide the original model into several sections based on preset principles; Step S2: In the sections divided by the original model, an unprocessed section is selected and a coarse aggregate skeleton structure is established in the section using rigid cluster clump units according to aggregate gradation; Step S3: Generate a number of ball units with the same particle size and regular arrangement in the entire model range, retain only the ball units in the segment selected in step S2 and delete the ball units in other segments of the model; Step S4: Based on the discrete element search area control algorithm, the relative position of each pebble unit and the ball unit is cyclically determined, so that all ball units contained in the range of each clump unit obtain the same group information, and the clump unit is converted into a flexible cluster unit, and then the clump unit is immediately deleted; wherein, the basic constituent unit of the rigid cluster clump unit is the pebble unit, and the basic constituent unit of the flexible cluster unit is the ball unit; The discrete element search region control algorithm in step S4 includes: Step S41: obtaining a coarse aggregate skeleton structure established by rigid cluster clump units and ball units generated from an origin position according to an arrangement rule; the origin position is (0, 0) in a 2D model and (0, 0, 0) in a 3D model; Step S42: Calculate the ID data of each ball unit according to the arrangement rule; Step S43: traverse and obtain the center coordinates of the pebble unit and the ball unit respectively; In the 2D model, the center coordinates of the pebble unit (x p ,y p ), the coordinates of the center of the ball unit (x b ,y b ); In the 3D model, the center coordinates of the pebble unit (x p ,y p , z p ), the coordinates of the center of the ball unit (x b ,y b , z b ); Step S44: based on the geometric shape corresponding to the adopted arrangement rule, the thickness of the ball unit is obtained; wherein, the average thickness of each ball unit in each row corresponding to the x and y directions is obtained in the 2D model, and the average thickness of each ball unit in each row and each layer corresponding to the x, y and z directions is obtained in the 3D model; Step S45: based on the coordinates of the center of the pebble unit and the thickness of the ball unit, the number of particles corresponding to the position of each pebble unit in each axial direction is obtained, and the position correspondence between the pebble unit and the adjacent ball unit is formed after rounding; wherein, rounding up, rounding down, rounding off, or random rounding is adopted when rounding; In the 2D model, get the number of particles t' corresponding to the x direction at each pebble unit location x , the number of rows corresponding to the y direction t' y ; In the 3D model, calculate the number of particles t' corresponding to the x direction at each pebble unit location x , the number of rows corresponding to the y direction t' y , the number of layers corresponding to the z direction t' z ; Step S46: Establishing the search area extension range, based on the calculated t' in the 2D model x and t' y Construct the search area; in the 3D model, based on the calculated t' x , t' y and t' z Constructing a search area; Step S47: obtaining the ID data of the ball unit in the constructed search area, and traversing and calculating the search area of ​​each pebble unit; Step S5: Obtain valid information of all ball units in step S4, and then delete all ball units in the section, wherein the valid information at least includes coordinate information of each axis, ball radius, group information and ID data; Step S6: Repeat steps S2 to S5 to traverse all the sections and record the valid information of the ball units in all the sections; Step S7: Based on the valid information of the ball units in all the sections in step S6, regenerate the second ball units at the corresponding positions in the entire model range, each second ball unit corresponds to a ball unit, and each second ball unit contains at least ID data and group information, so as to construct a discrete element geometric model with particle grouping information; Step S8: defining the contact model and microscopic parameters for the discrete element geometric model to complete the construction of the discrete element model.

2. The method for rapid discrete element modeling of large-capacity regularly arranged structures according to claim 1 is characterized in that: The preset principle in step S1 includes taking material mix ratio as a division standard or taking preset structural layer distribution as a division standard or taking user's will as a division standard to form a final segmentation model.

3. The method for rapid discrete element modeling of large-capacity regularly arranged structures according to claim 1 is characterized in that: The step S2 comprises: Starting from any section, rigid clusters are placed according to aggregate grading, and the coarse aggregate skeleton structure needs to be generated through self-weight deposition algorithm, compaction molding algorithm or volume expansion algorithm.

4. The method for rapid discrete element modeling of large-capacity regularly arranged structures according to claim 1 is characterized in that: The arrangement rule in step S3 includes a tetragonal arrangement structure or a hexagonal arrangement structure.

5. The method for rapid discrete element modeling applicable to large-capacity regularly arranged structures according to claim 4 is characterized in that: In step S42, the ID data for the quadrilateral arrangement structure is calculated as follows: Get the number of balls per row in the x direction in the 2D model x , ID data b ID =(t y -1)×n x +t x ; Get the number of balls per row in the x direction in the 3D model x The number of balls per layer in the z direction n z , ID data b ID =(t z -1)×n z +(t y -1)×n x +t x ; Where: t x is the number of balls in the row corresponding to the ball in the x direction; t y is the row number corresponding to the ball in the y direction; t z is the number of layers corresponding to the ball in the z direction; n x is the number of balls in each row in the x direction; n z is the number of balls per layer in the z direction; b ID The ID corresponding to the ball.

6. The method for rapid discrete element modeling applicable to large-capacity regularly arranged structures according to claim 5, characterized in that: In step S42, the ID data for the hexagonal arrangement structure is calculated as follows: ID data under 2D model: If the ball is located in an odd row b ID =(t y -1) / 2×(n x1 +n x2 )+t x ; If the ball is located in an even row b ID =(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ; ID data under 3D model: If the ball is located in an odd layer and an odd row b ID =(t z -1) / 2×(n z1 +n z2 )+(t y -1) / 2×(n x1 +n x2 )+t x ; If the ball is located in an odd layer and an even row b ID =(t z -1) / 2×(n z1 +n z2 )+(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ; If the ball is located in an even layer and an odd row b ID =(t z / 2-1)×(n z1 +n z2 )+n z1 +(t y -1) / 2×(n x1 +n x2 )+t x ; If the ball is located in an even layer and an even row b ID =(t z / 2-1)×(n z1 +n z2 )+n z1 +(t y / 2-1)×(n x1 +n x2 )+n x1 +t x ; Where: t x is the number of balls in the row corresponding to the ball in the x direction; t y is the row number corresponding to the ball in the y direction; t z is the number of layers corresponding to the ball in the z direction; n x1 n is the number of balls in odd rows in the x direction; x2 is the number of balls in even rows in the x direction; n z1 is the number of balls in the odd layers in the z direction; n z2 is the number of balls in even layers in the z direction; b ID The ID corresponding to the ball.

7. The method for rapid discrete element modeling applicable to large-capacity regularly arranged structures according to claim 6, characterized in that: In step S44: In the 2D model, the thickness dimension x of each ball unit in the tetragonal arrangement structure in the x direction is dis =2r, thickness dimension y in the y direction dis =2r; The thickness dimension x of each ball unit in the hexagonal arrangement structure in the x direction dis =2r, thickness dimension y in the y direction dis = r; In the 3D model, the thickness dimension x of each ball unit in the tetragonal arrangement structure in the x direction is dis =2r, thickness dimension y in the y direction dis =2r, thickness dimension z in z direction dis =2r; The thickness dimension x of each ball unit in the hexagonal arrangement structure in the x direction dis =2r, thickness dimension y in the y direction dis = r, thickness dimension z in z direction dis = r; Where r is the average radius of the ball unit.

8. The method for rapid discrete element modeling applicable to large-capacity regularly arranged structures according to claim 1, characterized in that: In step S46, t' x , t' y , t' z Based on the floating of at least two numbers above and below as the new search area, for t' x , t' y , t' z Reduce i x 、i y 、i z and increase u x 、u y 、u z， Among them, i x 、i y 、i z 、u x 、u y 、u z are all integers ≥ 2, and the x-axis search area range (t' x -i x )~(t' x +u x ), y-axis search area range (t' y -i y )~(t' y +u y ), z-axis search area range (t' z -i z )~(t' z +u z ).

Images