Construction method of initial packing model of composite cement paste microstructure based on grid optimization packing algorithm

Abstract

The present invention discloses a construction method for an initial packing model of the microstructure of a composite cement paste based on a grid optimization packing algorithm, including: 1. Setting and calculating the basic parameters of the microstructure model of the composite cement paste; 2. Generating a cube of the microstructure model and dividing it into an ordered grid array in space; 3. Establishing a grid information matrix to record the information of each grid; 4. Randomly generating a grid index sequence according to the current particle size grade number, particle pore sieve size, and grid array; 5. Placing cement particles within the particle size range corresponding to the current particle size grade number; 6. Placing mineral powder particles within the particle size range corresponding to the current particle size grade number; 7. Placing fly ash particles within the particle size range corresponding to the current particle size grade number; The method of the present invention is simple and efficient, can achieve a lower water-cement ratio, helps to predict the macroscopic properties of the cement paste, and has important significance for constructing a microstructure model of a composite cement paste containing irregular particles.

Description

Technical Field

[0001] The present invention relates to the field of constructing a microstructure model of a composite cement paste, and particularly to a method for constructing an initial packing model of the microstructure of a composite cement paste based on a grid optimization packing algorithm. Background Art

[0002] Cement components are complex, and it is relatively difficult to optimize and control their performance. Some scholars have used computer simulation to construct cement hydration and microstructure models, establish the relationship between cement components and their mechanical, durability and other properties, and use them to guide the production and application of cement. These include internationally popular models such as HYMOSTRUC3D, CEMHYD3D, and μic. Among them, the simulation of the initial three-dimensional packing structure of cement particles is the basis of the cement hydration and microstructure model.

[0003] Models such as HYMOSTRUC3D and μic often simplify cement particles as spheres for subsequent scientific research, which is easy to construct. However, cement particles are irregular, and using a spherical cement particle model brings many errors. Currently, the use of spherical harmonic functions to reconstruct the three-dimensional irregular morphology of cement particles and apply them to simulate the initial three-dimensional packing structure of cement particles is gradually increasing. This is because the advantage of spherical harmonic functions is that the three-dimensional shape of star-shaped particles can be easily represented mathematically like spheres and ellipsoids. At the same time, any particle parameters, such as volume, surface area, moment of inertia tensor, or integral mean curvature, can be easily calculated. Therefore, the simulation of packing based on spherical harmonic functions is an important research direction. Qian combined spherical harmonic functions with the traditional random placement algorithm to develop the Anm model. However, in the case of a low water-cement ratio and a large number of particles, the construction efficiency of the Anm model is low and it is not suitable for practical use. A low water-cement ratio is the basis for preparing high-performance cement-based materials. Obviously, the existing models are not suitable for guiding the design of high-performance cement-based materials with the characteristics of a low water-cement ratio.

[0004] Up to now, there is no method for efficiently constructing an initial packing model of the microstructure of a composite cement paste with a low water-cement ratio based on the true morphology of cement particles. Summary of the Invention

[0005] In order to solve the limitations of previous studies, the present invention proposes a method for constructing an initial packing model of the microstructure of a composite cement paste based on a grid optimization packing algorithm, in order to be able to efficiently construct an initial packing model of the microstructure. At the same time, this model can accurately reproduce the true particle packing state in terms of specific surface area, spatial distribution, and geometric morphology, which is of great significance for predicting the macroscopic properties of cement paste.

[0006] The present invention adopts the following technical solutions to achieve the above invention objectives:

[0007] The construction method of the initial packing model of the composite cement paste microstructure based on the grid optimization packing algorithm of the present invention is characterized in that it includes the following steps:

[0008] Step 1: Set and calculate the basic parameters of the composite cement paste microstructure model:

[0009] Step 1.1: Set the size of the cube CUBE, including: length L, width W, and height H;

[0010] Set the water-cement ratio as wcr, the density of cement as ρ PC , the density of mineral powder as ρ BFS , and the density of fly ash as ρ FA ;

[0011] According to the volume V = L×W×H of the cube CUBE, the water-cement ratio wcr, and the density ρ of cement PC , calculate the total mass M of the cementitious materials composed of cement, mineral powder, and fly ash;

[0012] Step 1.2: Set the mass proportion coefficient of cement as K PC , the mass proportion coefficient of mineral powder as K BFS , and the mass proportion coefficient of fly ash as K FA ; Multiply the total mass M by the corresponding mass proportion coefficients K PC , K BFS , and K FA , and respectively obtain the total mass m of cement PC , the total mass m of mineral powder BFS , and the total mass m of fly ash FA ;

[0013] Divide m PC , m BFS , and m FA by the corresponding densities ρ PC , ρ BFS , and ρ FA , and respectively obtain the total feeding volume V of cement PC , the total feeding volume V of mineral powder BFS , and the total feeding volume V of fly ash FA ;

[0014] Step 1.3: Set the size of the pore sieve as [W 1 , W 2 , W 3 , …, W i-1 , W i , …, W max ; where W i represents the size of the i-th pore sieve; W max represents the maximum size of the pore sieve; max represents the total number of pore sieves;

[0015] Set the cement particle size distribution per unit volume [VPC 1 ,VPC 2 ,…,VPC i-1 ,VPC i ,…,VPC max-1 , the blast furnace slag powder particle size distribution per unit volume [VBFS 1 ,VBFS 2 ,…,VBFS i-1 ,VBFS i ,…,VBFS max-1 and the fly ash particle size distribution per unit volume [VFA 1 ,VFA 2 ,…,VFA i-1 ,VFA i ,…,VFA max-1 ; where VPC i , VBFS i and VFA i respectively represent the to-be-dropped volumes of cement, blast furnace slag powder, and fly ash per unit volume within the particle size range (W i , W i+1 );

[0016] Multiply the total dropped volumes V PC , V BFS and V FA by the cement particle size distribution per unit volume, the blast furnace slag powder particle size distribution per unit volume, and the fly ash particle size distribution per unit volume respectively, and correspondingly obtain the cement particle size distribution [SVPC 1 , SVPC 2 ,…, SVPC i-1 , SVPC i ,…, SVPC max-1 , the blast furnace slag powder particle size distribution [SVBFS 1 , SVBFS 2 ,…, SVBFS i-1 , SVBFS i ,…, SVBFS max-1 and the fly ash particle size distribution [SVFA 1 , SVFA 2 ,…, SVFA i-1 , SVFA i ,…, SVFA max-1 , where SVPC i , SVBFS i and SVFA i respectively represent within the particle size range (W i , W i+1) The volume of cement, mineral powder, and fly ash to be put in

[0017] Set the threshold of the particle size order number to O, denote the current particle size order number as o, and o = max - i, initialize o = 1;

[0018] Step 2: Generate a cube CUBE with length L, width W, and height H, and take the lower - left - hand vertex of the cube CUBE as the coordinate origin, and take the length, width, and height of the cube CUBE as the X - axis direction, Y - axis direction, and Z - axis direction respectively to establish a three - dimensional Cartesian coordinate system;

[0019] Divide the cube CUBE into an orderly arranged grid array in space, and set the side length of each grid in the grid array to piex;

[0020] Traverse the X - axis, Y - axis, and Z - axis directions, and sequentially number the grids in the grid array, so as to establish a grid information matrix data_A for recording the spatial position information, grid index number, and grid status of each grid; initialize the grid status of each grid to 0, and if the grid is occupied, set the grid status to 1;

[0021] Step 3: Define a grid index matrix data_B for recording the grid index numbers of grids with grid status 1; extract the matrix recording the grid index numbers from data_A, and take the difference set with data_B to obtain a grid index matrix data_C with grid status 0;

[0022] Step 4: If the current particle size order number o is less than O, randomly generate T grid index sequences indx of grids with grid status 0 in data_C, and execute Step 5; otherwise, according to the maximum pore sieve size within the current particle size range, generate T small cubes with the same side length but different spatial positions in the grid array, sum the grid statuses included in each small cube, and then sort all the small cubes in ascending order, so as to obtain a grid index sequence indx composed of the grid index numbers of the grids at the centers of the T small cubes according to the sorting order, and then execute Step 5; where the side length of the small cube is an odd multiple of piex and the lower - left point coincides with the grid vertex;

[0023] Step 5: Put cement particles with a particle size range of (W i ,W i+1 )

[0024] Step 5.1: Initialize the variable l = 1, and judge whether the sum of the volumes of the cement particles with a particle size range of (W i ,W i+1 ) that have been put in is greater than or equal to the volume to be put in SVPC i , if so, execute Step 6, otherwise, execute Step 5.2;

[0025] Step 5.2: Randomly select an irregular cement particle from the database of spherical harmonic coefficient tables of irregular particles, and control the size of the irregular particle by scaling the coefficients in the spherical harmonic coefficient table of the irregular particle, so that the size of the irregular particle is within the corresponding particle size range (W i , W i+1 ), and obtain a cement particle to be placed;

[0026] Step 5.3: Determine whether l is greater than the number of elements in indx. If so, return to Step 4. Otherwise, perform a translation transformation and a rotation transformation on the cement particle to be placed according to the spatial position information corresponding to the l-th grid index number in indx, in accordance with the periodic boundary conditions, so as to generate a model of the cement particle to be placed;

[0027] Step 5.4: Determine whether there is interference between the model of the cement particle to be placed and the models of the particles that have been placed according to the double-interference determination method. If there is interference, assign l + 1 to l and then return to Step 5.3; otherwise, place the model of the cement particle to be placed with a particle size range in (W i , W i+1 ) into the cube CUBE, so that the model of the cement particle to be placed becomes a model of the particle that has been placed and record the particle information. Finally, update the grid state, the grid index matrix data_B, and the grid index matrix data_C, and return to Step 4;

[0028] Step 6: Place mineral powder particles with a particle size range in (W i , W i+1 );

[0029] Step 6.1: Initialize the variable l = 1, and determine whether the sum of the volumes of the mineral powder particles that have been placed with a particle size range in (W i , W i+1 ) is greater than or equal to the volume to be placed SVBFS i . If so, execute Step 7. Otherwise, execute Step 6.2;

[0030] Step 6.2: Randomly select an irregular mineral powder particle from the database of spherical harmonic coefficient tables of irregular particles, and control the size of the irregular particle by scaling the coefficients in the spherical harmonic coefficient table of the irregular particle, so that the size of the irregular particle is within the corresponding particle size range (W i , W i+1 ), and obtain a mineral powder particle to be placed;

[0031] Step 6.3: Determine whether l is greater than the number of elements in indx. If so, return to Step 4. Otherwise, perform a translation transformation and a rotation transformation on the powdered ore particles to be placed according to the spatial position information corresponding to the l-th grid index number in indx under periodic boundary conditions, so as to generate a model of the powdered ore particles to be placed;

[0032] Step 6.4: Determine whether there is interference between the model of the powdered ore particles to be placed and the model of the already placed particles according to the double interference determination method. If there is interference, assign l + 1 to l and then return to Step 6.3; otherwise, the powdered ore particles to be placed with a particle size range of (W i , W i+1 ) are placed into the cube CUBE, so that the model of the powdered ore particles to be placed becomes the model of the already placed particles and the particle information is recorded. Finally, update the grid state, the grid index matrix data_B, and the grid index matrix data_C, and return to Step 4;

[0033] Step 7: Place fly ash particles with a particle size range of (W i , W i+1 );

[0034] Step 7.1: Initialize the variable l = 1, and determine whether the sum of the volumes of the already placed fly ash particles with a particle size range of (W i , W i+1 ) is greater than or equal to the volume to be placed SVFA i . If so, assign o + 1 to o, update i = max - o, and then execute Step 7.2; otherwise, directly execute Step 7.2;

[0035] Step 7.2: If i = 0, stop placing, save the information of the already placed particles, and complete the construction of the initial packing model of the composite cement paste microstructure; otherwise, randomly generate spherical particles with a particle size range of (W i , W i+1 ) and use them as the fly ash particles to be placed, and execute Step 7.3;

[0036] Step 7.3: Determine whether l is greater than the number of elements in indx. If so, return to Step 4. Otherwise, perform a translation transformation and a rotation transformation on the fly ash particles to be placed according to the spatial position information corresponding to the l-th grid index number in indx under periodic boundary conditions, so as to generate a model of the fly ash particles to be placed;

[0037] Step 7.4: Determine whether there is interference between the model of the fly ash particles to be placed and the model of the already placed particles according to the double interference determination method. If there is interference, assign l + 1 to l and then return to Step 7.3; otherwise, the fly ash particles to be placed with a particle size range of (W i , W i+1) The fly ash particle model to be placed is placed into the cube CUBE, so that the fly ash particle model to be placed becomes the placed particle model and the particle information is recorded. Finally, the grid state, the grid index matrix data_B, and the grid index matrix data_C are updated, and step 4 is returned.

[0038] The construction method of the initial packing model of the composite cement paste microstructure based on the grid optimization packing algorithm according to the present invention is also characterized in that the double interference determination method in the step 5.4 includes:

[0039] Step 5.4.1: Obtain the axis-aligned minimum bounding cube C of the particle model to be placed through a self-written function t , and select the placed particle models that overlap with C t to form a set particles, and obtain the grid index sequence indx t included in C 0 ;

[0040] Step 5.4.2: Traverse each grid index number in the grid index sequence indx 0 , and judge whether the space position corresponding to each corresponding grid index number in data_A is occupied by the particle model to be placed through the spherical harmonic function. If it is occupied, record the corresponding grid index number into the grid index sequence indx 1 ;

[0041] Step 5.4.3: Coarse determination of the particle model to be placed:

[0042] Judge whether the grid state of each index number in indx 1 is all 0. If so, it means that the particle model to be placed meets the particle interference condition, and there is no interference between the particle model to be placed and the placed particle models; otherwise, execute step 5.4.4;

[0043] Step 5.4.4: Fine determination of the particle model to be placed:

[0044] According to the recorded particle information, use the two-particle contact function to judge whether there is interference between the particle model to be placed and any one of the placed particle models in particles. If so, it means that there is interference between the particle model to be placed and the placed particle models; otherwise, it means that there is no interference between the particle model to be placed and the placed particle models.

[0045] An electronic device according to the present invention includes a memory and a processor, characterized in that the memory is used to store a program for supporting the processor to execute the construction method, and the processor is configured to execute the program stored in the memory.

[0046] A computer-readable storage medium of the present invention is characterized in that a computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, the steps of the construction method are executed.

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0048] 1. The present invention introduces periodic boundary conditions, combines parallel computing technology and placement algorithms, solves the problem of large water-cement ratio at the boundary of the microstructure model during parametric modeling, and improves the efficiency of constructing the microstructure model by more than several times.

[0049] 2. The present invention uses a grid optimization packing algorithm to judge particle interference for the particle model to be placed, which is simple and efficient. To a certain extent, it improves the upper limit of the particle placement rate, enables the microstructure model to reach a lower water-cement ratio, and improves the practicability of the model.

[0050] 3. Compared with the existing methods for modeling the initial packing model of the microstructure, the grid optimization packing algorithm proposed by the present invention calculates the state of the cubic grid and represents the particles by spherical harmonic functions, making this method have the advantages of high modeling efficiency and realistic particle morphology. The generated initial packing model of the microstructure can accurately reproduce the real particle packing state in terms of specific surface area and particle morphology. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flowchart for modeling the initial packing model of the microstructure of the method of the present invention;

[0052] Figure 2 is a diagram of the cubic grid array of the model of the present invention;

[0053] Figure 3 is a morphology diagram of cement, mineral powder, and fly ash particles in the model of the present invention;

[0054] Figure 4 is a diagram for judging the periodic boundary conditions of the method of the present invention;

[0055] Figure 5 is a two-dimensional schematic diagram of the particle grid state of the present invention;

[0056] Figure 6 is a two-dimensional schematic diagram for judging particle interference of the present invention;

[0057] Figure 7 is an example diagram of the initial packing model of the microstructure with a water-cement ratio of 0.25 of the present invention;

[0058] Figure 8 is a sectional view of the initial packing model of the microstructure of the present invention along the X-axis, Y-axis, and Z-axis. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] In this embodiment, as Figure 1 shown, a construction method of an initial packing model for the microstructure of a composite cement paste based on a grid optimization packing algorithm includes the following steps:

[0060] Step 1: Set and calculate the basic parameters of the composite cement paste microstructure model:

[0061] Step 1.1: Set the dimensions of the cube CUBE, including: length L, width W, and height H;

[0062] Set the water-cement ratio as wcr, the density of cement as ρ PC , the density of mineral powder as ρ BFS , and the density of fly ash as ρ FA ;

[0063] According to the volume V = L×W×H of the cube CUBE, the water-cement ratio wcr, and the density ρ PC of cement, calculate the total mass M of the cementitious materials composed of cement, mineral powder, and fly ash by Equation (1);

[0064]

[0065] Step 1.2: Set the mass proportion coefficient of cement as K PC , the mass proportion coefficient of mineral powder as K BFS , and the mass proportion coefficient of fly ash as K FA ; Multiply the total mass M by the corresponding mass proportion coefficients K PC , K BFS , and K FA respectively to obtain the total mass m PC of cement, the total mass m BFS of mineral powder, and the total mass m FA of fly ash;

[0066] Divide m PC , m BFS , and m FA by the corresponding densities ρ PC , ρ BFS , and ρ FA respectively to obtain the total casting volume V PC of cement, the total casting volume V BFS of mineral powder, and the total casting volume V FA of fly ash.

[0067] Step 1.3: Set the size of the pore sieve as [W 1 , W 2 , W 3 ,…, W i-1 , W i ,…, W max; where, W i represents the size of the i-th pore sieve; W max represents the maximum size of the pore sieve; max represents the total number of pore sieves;

[0068] Set the cement particle size distribution per unit volume [VPC 1 , VPC 2 , …, VPC i-1 , VPC i , …, VPC max-1 , the blast furnace slag powder particle size distribution per unit volume [VBFS 1 , VBFS 2 , …, VBFS i-1 , VBFS i , …, VBFS max-1 and the fly ash particle size distribution per unit volume [VFA 1 , VFA 2 , …, VFA i-1 , VFA i , …, VFA max-1 ; where, VPC i , VBFS i and VFA i respectively represent the to-be-discharged volumes of cement, blast furnace slag powder, and fly ash per unit volume within the particle size range (W i , W i+1 );

[0069] Through equations (2), (3), and (4), obtain the to-be-discharged volumes SVPC i , SVBFS i+1 and SVFA i of cement, blast furnace slag powder, and fly ash within the particle size range (W i , W i ); Multiply the total discharge volume V PC , V BFS and V FA by the cement particle size distribution per unit volume, the blast furnace slag powder particle size distribution per unit volume, and the fly ash particle size distribution per unit volume respectively, and correspondingly obtain the particle size distribution of cement [SVPC 1 , SVPC 2 , …, SVPC i-1 , SVPC i , …, SVPC max-1 , the particle size distribution of blast furnace slag powder [SVBFS 1 , SVBFS 2 , …, SVBFS i-1 , SVBFS i , …, SVBFS max-1 and the particle size distribution of fly ash [SVFA1 , SVFA 2 , …, SVFA i-1 , SVFA i , …, SVFA max-1 , where SVPC i , SVBFS i and SVFA i respectively represent the volumes of cement, slag powder, and fly ash to be placed within the particle size range of (W i , W i+1 );

[0070]

[0071]

[0072]

[0073] Set the threshold of the particle size sequence number to O, denote the current particle size sequence number as o, and o = max - i. Initialize o = 1.

[0074] In this example, the length L of the cube CUBE is 100 μm, the width W is 100 μm, and the height H is 100 μm; the side length of each grid in the grid array is piex = 0.4 μm; the water - cement ratio wcr = 0.25, the density of cement ρ PC = 3.14 g / cm 2 , the density of slag powder ρ BFS = 2.93 g / cm 2 , and the density of fly ash ρ FA = 2.56 g / cm 2 ; the mass proportion coefficients of cement, slag powder, and fly ash are K PC = 0.6, K BFS = 0.2, and K FA = 0.2; the total volumes of slag powder and fly ash to be placed are V PC = 336134.5 μm 3 , V BFS = 120075.3 μm 3 , and V FA = 137429.9 μm 3 ; the maximum size of the pore sieve W max = 45 μm, and the minimum size of the pore sieve W 1 = 2 μm; the threshold of the particle size sequence number is O = 17.

[0075] Step 2: Generate a cube CUBE with length L, width W, and height H, and establish a three - dimensional Cartesian coordinate system with the lower - left - hand vertex of the cube CUBE as the coordinate origin and the length, width, and height of the cube CUBE as the X - axis direction, Y - axis direction, and Z - axis direction;

[0076] Divide the cube CUBE into an ordered grid array in space, and set the side length of each grid in the grid array to piex;

[0077] Traverse the X-axis, Y-axis, and Z-axis directions, and sequentially number the grids in the grid array to establish a grid information matrix data_A for recording the spatial position information, grid index number, and grid status of each grid; initialize the grid status of each grid to 0, and if the grid is occupied, set the grid status to 1.

[0078] Step 3: Define a grid index matrix data_B for recording the grid index numbers of grids with a grid status of 1; extract the matrix recording the grid index numbers from data_A, and take the difference set with data_B using the setdiff function to obtain a grid index matrix data_C with a grid status of 0.

[0079] Step 4: If the current particle size sequence number o is less than O, randomly generate T grid index sequences indx with a grid status of 0 in data_C, and execute Step 5; otherwise, according to the maximum sieve size within the current particle size range, generate T small cubes with the same side length but different spatial positions in the grid array, sum the grid statuses included in each small cube, and then sort all the small cubes in ascending order. Thus, after obtaining a grid index sequence indx composed of the grid index numbers of the grids at the centers of the T small cubes according to the sorting order, execute Step 5; among them, the side length of the small cube is an odd multiple of piex, and the lower left corner point coincides with the grid vertex;

[0080] In this example, T = 10000; L / piex = W / piex = H / piex = 250, that is, 250 grids are divided in the length, width, and height directions respectively; the microstructural cube is as shown in the left figure of Figure 2 The middle part of Figure 2 is a schematic diagram of dividing the microstructural cube into a three-dimensional grid array, and the schematic diagram of dividing the two-dimensional grid array is as shown in the right figure of Figure 2

[0081] Step 5: Place cement particles with a particle size range in (W i , W i+1 );

[0082] Step 5.1: Initialize the variable l = 1, and judge whether the sum of the volumes of the cement particles already placed with a particle size range in (W i , W i+1 ) is greater than or equal to the volume to be placed SVPC i . If so, execute Step 6; otherwise, execute Step 5.2;

[0083] ​Step 5.2: Randomly select an irregular cement particle from the database of spherical harmonic coefficient tables of irregular particles, and control the size of the irregular particle by scaling the coefficients in the spherical harmonic coefficient table of the irregular particle so that the size of the irregular particle is within the corresponding particle size range (W i , W i+1 ), and obtain a cement particle to be placed;

[0084] Step 5.3: Determine whether l is greater than the number of elements in indx. If so, return to Step 4. Otherwise, perform translational transformation and rotational transformation on the cement particle to be placed according to the spatial position information corresponding to the l-th grid index number in indx under periodic boundary conditions, so as to generate a model of the cement particle to be placed.

[0085] Step 5.4: Determine whether there is interference between the model of the cement particle to be placed and the models of the placed particles according to the double interference determination method; the double interference determination method specifically includes the following steps:

[0086] Step 5.4.1: Obtain the axis-aligned minimum bounding cube C of the model of the particle to be placed through a self-written function t , and select the models of the placed particles that overlap with C t to form a set particles, and obtain the grid index sequence indx t included in C 0 ;

[0087] Step 5.4.2: Traverse each grid index number in the grid index sequence indx 0 , and determine whether the spatial position corresponding to each corresponding grid index number in data_A is occupied by the model of the particle to be placed through spherical harmonic functions. If it is occupied, record the corresponding grid index number into the grid index sequence indx 1 .

[0088] Step 5.4.3: Conduct a rough determination on the model of the particle to be placed:

[0089] Determine whether the grid status of each index number in indx 1 is all 0. If so, it means that the model of the particle to be placed meets the particle interference condition, and there is no interference between the model of the particle to be placed and the models of the placed particles; otherwise, execute Step 5.4.4;

[0090] Step 5.4.4: Conduct a fine determination on the model of the particle to be placed:

[0091] According to the recorded particle information, use the two-particle contact function to determine whether there is interference between the particle model to be placed and any of the already placed particle models in particles. If so, it means there is interference between the particle model to be placed and the already placed particle model; otherwise, it means there is no interference between the particle model to be placed and the already placed particle model.

[0092] According to the double interference determination method, determine whether there is interference between the cement particle model to be placed and the already placed particle models. If there is interference, after assigning l + 1 to l, return to step 5.3; otherwise, place the cement particle models to be placed with a particle size range in (W i , W i+1 ) into the cube CUBE, making the cement particle models to be placed become already placed particle models and record the particle information. Finally, update the grid state, the grid index matrix data_B, and the grid index matrix data_C, and return to step 4.

[0093] In this example, as Figure 3 shown, the morphologies of cement particles and mineral powder particles are irregular, and the morphology of fly ash particles is spherical; the four cases of periodic boundary conditions are as Figure 4 shown. As Figure 5 shown, in the two-dimensional case, the grid state occupied by particles is set to 1, and the grid is filled with gray; the unoccupied grid state is 0, and the grid is filled with white; as shown in part (a) of Figure 6 , a rough determination is made on the particle model to be placed, and there is no interference with the already placed particle models. As shown in part (b) of Figure 6 , a fine determination is made on the particle model to be placed, and there is no interference with the already placed particle models. As shown in part (c) of Figure 6 , a fine determination is made on the particle model to be placed, and there is interference with the already placed particle models.

[0094] Step 6: Place mineral powder particles with a particle size range in (W i , W i+1 );

[0095] Step 6.1: Initialize the variable l = 1, and determine whether the sum of the volumes of the already placed mineral powder particles with a particle size range in (W i , W i+1 ) is greater than or equal to the volume to be placed SVBFS i . If so, execute step 7; otherwise, execute step 6.2;

[0096] Step 6.2: Randomly select an irregular mineral powder particle from the database of spherical harmonic coefficient tables of irregular particles, and control the size of the irregular particle by scaling the coefficients in the spherical harmonic coefficient table of the irregular particle, so that the size of the irregular particle is within the corresponding particle size range (Wi ,W i+1 ) and obtain a mineral powder particle to be placed.

[0097] Step 6.3: Determine whether l is greater than the number of elements in indx. If so, return to Step 4. Otherwise, perform a translation transformation and a rotation transformation on the mineral powder particle to be placed according to the spatial position information corresponding to the l-th grid index number in indx, in accordance with the periodic boundary conditions, so as to generate a model of the mineral powder particle to be placed;

[0098] Step 6.4: Determine whether there is interference between the model of the mineral powder particle to be placed and the model of the placed particles according to the double interference determination method. If there is interference, assign l + 1 to l and then return to Step 6.3; otherwise, assign the particle size range of (W i ,W i+1 ) of the model of the mineral powder particle to be placed into the cube CUBE, so that the model of the mineral powder particle to be placed becomes the placed particle model and record the particle information. Finally, update the grid state, the grid index matrix data_B and the grid index matrix data_C, and return to Step 4.

[0099] Step 7: Place fly ash particles with a particle size range of (W i ,W i+1 );

[0100] Step 7.1: Initialize the variable l = 1, and determine whether the sum of the volumes of the placed fly ash particles with a particle size range of (W i ,W i+1 ) is greater than or equal to the volume to be placed SVFA i . If so, assign o + 1 to o, update i = max - o, and then execute Step 7.2; otherwise, directly execute Step 7.2;

[0101] Step 7.2: If i = 0, stop placing, save the information of the placed particles, and complete the construction of the initial packing model of the composite cement paste microstructure; otherwise, randomly generate spherical particles with a particle size range of (W i ,W i+1 ) and use them as the fly ash particles to be placed, and execute Step 7.3.

[0102] Step 7.3: Determine whether l is greater than the number of elements in indx. If so, return to Step 4. Otherwise, perform a translation transformation and a rotation transformation on the fly ash particle to be placed according to the spatial position information corresponding to the l-th grid index number in indx, in accordance with the periodic boundary conditions, so as to generate a model of the fly ash particle to be placed;

[0103] Step 7.4: Determine whether there is interference between the fly ash particle model to be placed and the placed particle model according to the double interference determination method. If there is interference, assign l + 1 to l and then return to Step 7.3; otherwise, place the fly ash particle model to be placed with a particle size range of (W i , W i+1 ) into the cube CUBE, making the fly ash particle model to be placed become the placed particle model and record the particle information. Finally, update the grid state, the grid index matrix data_B, and the grid index matrix data_C, and return to Step 4.

[0104] As Figure 7 shown, in this example, the above method is used to construct an initial packing model of the composite cement paste microstructure based on the grid optimization packing algorithm, where the water-cement ratio is 0.25, the total placed volume is 59.3643%, and the total number of particles is 10,399. The computer parameters for simulating this example are Intel(R) Xeon(R) Gold 5218 CPU @ 2.30 GHz and 128 GB RAM, and the total time used is 159,467 s; among them, the volume of cement is 33.6135%, the particle morphology is irregular, and the number of particles is 5,800; the volume of slag powder is 12.0078%, the particle morphology is irregular, and the number of particles is 2,314; the volume of fly ash is 13.743%, the particle morphology is spherical, and the number of particles is 2,285. As Figure 8 shown, by using the self-written slicing algorithm to slice the model along the X, Y, and Z axes, it is verified that there is no overlap in the initial packing of the model constructed by the above method.

[0105] In this embodiment, an electronic device includes a memory and a processor. The memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0106] In this embodiment, a computer-readable storage medium stores a computer program, and when the computer program is run by a processor, it executes the steps of the above method.

Claims

1. A method for constructing an initial stacking model of a composite cement paste microstructure based on a grid optimization stacking algorithm, characterized in that: The following steps are involved: Step 1: Set and calculate the basic parameters of the composite cement paste microstructure model: Step 1.1: Set the dimensions of the cube CUBE, including length L, width W and height H; Let the water-cement ratio be wcr and the density of cement be ρ PC , the density of mineral powder is ρ BFS , the density of fly ash is ρ FA ; According to the volume of the cube CUBE V = L × W × H, the water-cement ratio wcr and the density of cement ρ PC , calculate the total mass M of the cementitious material composed of cement, mineral powder and fly ash; Step 1.2: Set the mass ratio of cement to K PC , the mass ratio of mineral powder is K BFS , the mass ratio of fly ash is K FA ; Multiply the total mass M by the corresponding mass ratio K PC , K BFS and K FA , and the total mass of cement m is obtained respectively PC , the total mass of the mineral powder m BFS and the total mass of fly ash m FA ; M PC 、m BFS and m FA Divide by the corresponding density ρ PC , BFS and ρ FA , and the corresponding total cement volume V PC , the total volume of mineral powder V BFS and the total volume of fly ash V FA ; Step 1.3: Set the size of the hole screen to [W1, W2, W3, …, W i-1 ,W i ,…,W max ]; among them, W i represents the size of the i-th hole screen; W max Indicates the maximum size of the hole screen; max indicates the total number of hole screens; Set the cement particle size distribution per unit volume [VPC1, VPC2, …, VPC i-1 ,VPC i ,…,VPC max-1 ]、Mineral powder particle size distribution per unit volume [VBFS1, VBFS2,…, VBFS i-1 ,VBFS i ,…,VBFS max-1 ] and the fly ash particle size distribution per unit volume [VFA1, VFA2, …, VFA i-1 ,VFA i ,…,VFA max-1 ]; VPC i VBFS i and VFA i Respectively represent the particle size range of cement, mineral powder and fly ash per unit volume (W i ,W i+1 ) within the volume to be delivered; The total delivery volume V PC 、V BFS and V FA Multiply the cement particle size distribution per unit volume, the mineral powder particle size distribution per unit volume, and the fly ash particle size distribution per unit volume, respectively, and the corresponding cement particle size distribution [SVPC1, SVPC2, …, SVPC i-1 ,SVPC i ,…,SVPC max-1 ], particle size distribution of mineral powder [SVBFS1, SVBFS2,…, SVBFS i-1 ,SVBFS i ,…,SVBFS max-1 ] and the particle size distribution of fly ash [SVFA1, SVFA2,…, SVFA i-1 ,SVFA i ,…,SVFA max-1 ], where SVPC i 、SVBFS i and SVFA i Respectively represent the particle size range in (W i ,W i+1 ) of cement, mineral powder and fly ash to be put in; Set the threshold of the particle level number to O, the current particle level number is recorded as o, and o=max-i, initialize o=1; Step 2: Generate a cube CUBE with a length of L, a width of W, and a height of H, and establish a three-dimensional Cartesian coordinate system with the lower left corner vertex of the cube CUBE as the coordinate origin and the length, width, and height of the cube CUBE as the X-axis direction, the Y-axis direction, and the Z-axis direction; Divide the cube CUBE into a grid array arranged in order in space, and let the side length of each grid in the grid array be piex; Traverse the X-axis, Y-axis and Z-axis directions, and number the grids in the grid array in turn, so as to establish a grid information matrix data_A, which is used to record the spatial position information, grid index number and grid state of each grid; initialize the grid state of each grid to 0, and if the grid is occupied, set the grid state to 1; Step 3: define a grid index matrix data_B to record the grid index number of the grid state 1; extract the matrix recording the grid index number from data_A, and take the difference set with data_B to obtain the grid index matrix data_C of the grid state 0; Step 4: If the current particle size sequence number o is less than O, randomly generate T grid index sequences indx with grid states of 0 in data_C, and execute step 5; otherwise, according to the maximum hole screen size within the current particle size range, generate T small cubes with the same side length but different spatial positions in the grid array, and sum the grid states contained in each small cube, and then sort all the small cubes in ascending order, so as to obtain the grid index sequence indx composed of the index numbers of the grids at the centers of the T small cubes according to the sorting order, and then execute step 5; wherein, the side length of the small cube is an odd multiple of piex, and the lower left corner point coincides with the grid vertex; Step 5: The particle size range is (W i ,W i+1 ) of cement particles; Step 5.1: Initialize variable l = 1, determine the particle size range (W i ,W i+1 ) is the sum of the volume of cement particles already placed greater than or equal to the volume to be placed SVPC i , if yes, go to step 6, otherwise go to step 5.2; Step 5.2: Randomly select an irregular particle of cement from the database of spherical harmonic coefficients of irregular particles, and control the size of the irregular particle by scaling the coefficients in the spherical harmonic coefficients of irregular particles so that the irregular particle size is within the corresponding particle size range (W i ,W i+1 ) and obtain a cement particle to be placed; Step 5.3: Determine whether l is greater than the number of elements in indx. If so, return to step 4. Otherwise, perform translation and rotation transformations on the cement particles to be placed according to the periodic boundary conditions based on the spatial position information corresponding to the lth grid index number in indx, thereby generating a cement particle model to be placed. Step 5.4: Determine whether there is interference between the cement particle model to be placed and the particle model already placed according to the double interference judgment method. If so, assign l+1 to l and return to step 5.3; otherwise, assign the particle size range to (W i ,W i+1 ) is placed into the cube CUBE, so that the cement particle model to be placed becomes a placed particle model and records the particle information, and finally updates the grid state, grid index matrix data_B and grid index matrix data_C, and returns to step 4; Step 6: The particle size range is (W i ,W i+1 ) of mineral powder particles; Step 6.1: Initialize variable l = 1, determine the particle size range (W i ,W i+1 ) is the sum of the volume of the mineral powder particles that have been placed greater than or equal to the volume to be placed SVBFS i , if yes, go to step 7, otherwise go to step 6.2; Step 6.2: Randomly select an irregular particle of mineral powder from the database of spherical harmonic coefficients of irregular particles, and control the size of the irregular particle by scaling the coefficients in the spherical harmonic coefficients of irregular particles so that the size of the irregular particle is within the corresponding particle size range (W i ,W i+1 ) and obtain a mineral powder particle to be put in; Step 6.3: Determine whether l is greater than the number of elements in indx. If so, return to step 4. Otherwise, perform translation and rotation transformations on the mineral powder particles to be placed according to the periodic boundary conditions based on the spatial position information corresponding to the lth grid index number in indx, thereby generating a mineral powder particle model to be placed. Step 6.4: Determine whether there is interference between the particle model of the mineral powder to be placed and the particle model that has been placed according to the double interference judgment method. If so, assign l+1 to l and return to step 6.3; otherwise, assign the particle size range to (W i ,W i+1 ) is placed into the cube CUBE, so that the mineral powder particle model to be placed becomes the particle model that has been placed and the particle information is recorded, and finally the grid state, grid index matrix data_B and grid index matrix data_C are updated, and return to step 4; Step 7: The particle size range is (W i ,W i+1 ) of fly ash particles; Step 7.1: Initialize variable l = 1, determine the particle size range (W i ,W i+1 ) is the sum of the volume of fly ash particles that have been placed greater than or equal to the volume to be placed SVFA i , if yes, assign o+1 to o, update i=max-o, and then execute step 7.2; otherwise, directly execute step 7.2; Step 7.2: If i = 0, stop adding, save the information of the added particles, and complete the construction of the initial accumulation model of the composite cement paste microstructure; otherwise, randomly generate a particle size range in (W i ,W i+1 ) as the fly ash particles to be placed, and execute step 7.3; Step 7.3: Determine whether l is greater than the number of elements in indx. If so, return to step 4. Otherwise, perform translation and rotation transformations on the fly ash particles to be placed according to the periodic boundary conditions based on the spatial position information corresponding to the lth grid index number in indx, thereby generating a fly ash particle model to be placed. Step 7.4: Determine whether there is interference between the fly ash particle model to be placed and the particle model already placed according to the double interference judgment method. If so, assign l+1 to l and return to step 7.3; otherwise, assign the particle size range to (W i ,W i+1 ) is placed into the cube CUBE, so that the fly ash particle model to be placed becomes the particle model that has been placed and the particle information is recorded, and finally the grid state, grid index matrix data_B and grid index matrix data_C are updated, and return to step 4.

2. The method for constructing the initial stacking model of the composite cement paste microstructure based on the grid optimization stacking algorithm according to claim 1 is characterized in that: The double interference determination method in step 5.4 includes: Step 5.4.1: Use a self-written function to find the axis-aligned minimum circumscribed cube C of the particle model to be placed t , and select the t Overlapping particle models that have been placed form a set of particles, and find C t The grid index sequence indx0 is included; Step 5.4.2: traverse each grid index number in the grid index sequence indx0, and use the spherical harmonic function to determine whether the spatial position corresponding to each corresponding grid index number in data_A is occupied by the particle model to be placed. If occupied, record the corresponding grid index number in the grid index sequence indx1; Step 5.4.3: Make a rough judgment on the particle model to be released: Determine whether the grid status of each index number in indx1 is 0. If so, it means that the particle model to be placed meets the particle interference condition, and there is no interference between the particle model to be placed and the particle model that has been placed; otherwise, execute step 5.4.4; Step 5.4.4: Make a detailed judgment on the particle model to be released: According to the recorded particle information, the two-particle contact function is used to determine whether there is interference between the particle model to be placed and any particle model that has been placed in particles. If so, it means that there is interference between the particle model to be placed and the particle model that has been placed; otherwise, it means that there is no interference between the particle model to be placed and the particle model that has been placed.

3. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports a processor to execute the construction method according to claim 1 or 2, and the processor is configured to execute the program stored in the memory.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the construction method according to claim 1 or 2 are executed.

Images