Design method of mix proportion of self-compacting geopolymer concrete in arid and cold region

Abstract

The application discloses a kind of self-compacting geopolymer concrete mix design methods considering arid cold region, and the application relates to the technical field of concrete mix design.It includes the following steps: obtaining self-compacting geopolymer concrete samples with different mix proportions, testing their flow characteristic parameters including T500 flow time, emptying time, setting time and mechanical characteristic parameters after hardening.Map the mix proportion and the corresponding parameters to generate a training sample set, and based on the set, establish a neural network model, and train to obtain a feature prediction model.Generate different mix proportion combinations through the known material addition range, input the feature prediction model to predict the flow and mechanical characteristic parameters.Calculate the mechanical performance parameters and combine the flow characteristic parameters to construct a fitness function.Using genetic algorithm to optimize the mix proportion combination, get the scheme with the highest fitness value, provide reliable protection for building construction in arid and cold regions, improve the safety and durability of the project.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of concrete mix design, in particular to a self-compacting geopolymer concrete mix design method considering arid and cold regions. BACKGROUND

[0002] The concrete structure of power transmission and transformation projects in the northwest arid and cold region not only bears the safe load combination, but also is eroded by special natural environment, such as freeze-thaw damage, soil salt ion erosion, high mineralization degree irrigation return water, atmospheric carbonization, dry-wet cycle damage, and the effects of natural disasters such as landslides, mudslides, and floods. These factors lead to erosion and damage phenomena such as cracks, frost heaving, erosion, carbonization, and alkali aggregate reaction in the concrete structure of power transmission tower foundation, which seriously affects the safe operation of power transmission and transformation projects in this region and restricts the normal play of the benefits of power transmission and transformation projects.

[0003] Geopolymer is used to describe inorganic aluminosilicate polymers containing natural minerals such as metakaolin or industrial solid wastes such as fly ash and slag synthesized by using high-alkali activators. This material is formed by depolymerization and then polycondensation under the action of alkali activation at room temperature or slightly higher temperature. Due to its unique structure, it has many excellent properties that silicate cement-based materials cannot achieve, especially in mechanical properties and durability.

[0004] Therefore, the preparation of self-compacting geopolymer concrete is the starting point. According to the working performance of self-compacting concrete, the mechanical properties and durability of the concrete are analyzed to determine the optimal mix ratio, which is a problem to be solved to prevent the damage of concrete structure and ensure the safe operation of power transmission and transformation projects.

[0005] In the prior art, the publication number CN117131683A discloses a self-compacting concrete mix design method based on the rheological properties and strength of the paste. The specific steps include: initially determining the self-compacting concrete mix ratio and determining its rheological parameters; testing the properties of cementitious materials and calculating the rheological properties of cement paste; testing the properties of fine aggregates and calculating the rheological properties of mortar; testing the properties of coarse aggregates and calculating the rheological properties of self-compacting concrete; optimizing and adjusting the self-compacting concrete mix ratio; preparing self-compacting concrete according to the optimized mix ratio and testing its mechanical strength. However, this design method based on the rheological properties and strength of the paste focuses more on the rheological properties and mechanical properties of the concrete, and does not adequately consider environmental factors such as temperature and humidity. This may result in the performance of the concrete not meeting expectations in specific climatic conditions such as arid and cold regions. Without specifically optimizing these factors, it may lead to engineering quality problems. At the same time, the influence of geopolymer on concrete is not considered, which reduces the accuracy and effectiveness of the optimized design results.

[0006] The above information disclosed in the BACKGROUND section merely to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0007] The purpose of the present application is to provide a self-compacting geopolymer concrete mix proportion design method considering arid and cold regions to solve the problems raised in the background.

[0008] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0009] A self-compacting geopolymer concrete mix proportion design method considering arid and cold regions, the specific steps comprising:

[0010] Obtain the flow characteristic parameters and the mechanical characteristic parameters of the hardened self-compacting geopolymer concrete samples under a plurality of different material preparation ratios through experiments; the preparation materials include metakaolin, sodium silicate, sodium hydroxide, water reducing agent, coarse and fine aggregates, and water; the flow characteristic parameters include T500 flow time, emptying time and setting time;

[0011] Establish a mapping relationship between the material ratio and the flow characteristic parameters and the mechanical characteristic parameters, construct a training sample set, train a neural network model based on the training sample set, and obtain a prediction model for predicting the flow characteristic parameters and the mechanical characteristic parameters;

[0012] Generate a plurality of mix proportion combinations within the preset dosage range of each material, input them into the prediction model, obtain the corresponding flow characteristic parameter prediction value and mechanical characteristic parameter prediction value, and calculate the mechanical performance parameters based on the mechanical characteristic parameter prediction value; combine the flow characteristic parameter prediction value to calculate the fitness function value of each mix proportion combination;

[0013] Iteratively optimize the mix proportion combinations using a genetic algorithm, select the mix proportion combination with the highest fitness function value as the optimal ratio, and prepare the concrete accordingly.

[0014] Further, the mechanical characteristic parameters include compressive failure load, compressive bearing area, segregation limit load, segregation section width, segregation section height, and transverse fundamental frequency of concrete before and after thawing;

[0015] Obtain the flow characteristic parameters of the self-compacting geopolymer concrete sample through experiments, wherein the T500 flow time in the flow characteristic parameters specifically refers to the time from the start of lifting the slump cone to the flow to the diameter of 500mm of the slump cone, which is obtained through the slump test;

[0016] The emptying time is obtained through the V-box flow test, specifically referring to the time required for the concrete mixture to be completely emptied after the V-box is filled with concrete samples.

[0017] Furthermore, the neural network model is trained based on this training sample set. A neural network model is established using the Long Short-Term Memory (LSTM) network model as the foundation. For the LSTM network model, an activation function and optimization algorithm are selected, with the Tanh function chosen as the activation function and Adam as the optimization algorithm. The formula for the Tanh function is:

[0018] ;

[0019] In the formula, Represents the Tanh function, with the independent variable... This represents the weighted sum of the neuron's inputs, that is, the result of the weighted sum of the inputs received by the neuron from the previous layer;

[0020] Simultaneously, the hyperparameters of the LSTM model are set, including: number of network layers, number of iterations, learning rate, batch size, number of training iterations, number of batches, and number of hidden layer neurons.

[0021] The network is set to a 4-layer structure, the number of iterations is set to 300, the learning rate is set to 0.001, the batch size is set to 32, the number of training iterations is set to 100, the number of batches is set to 256, and the number of hidden layer neurons is set to 32.

[0022] The input to the trained feature prediction model is the mix proportion data of the prepared material, and the output is the predicted values ​​of the corresponding flow characteristic parameters and mechanical characteristic parameters.

[0023] Furthermore, based on the predicted values ​​of the obtained mechanical characteristic parameters, the mechanical performance parameters of the concrete with the corresponding mix proportions are calculated, wherein the formula used for calculating the relative dynamic elastic modulus is:

[0024] ;

[0025] In the formula, This represents the relative dynamic modulus of elasticity of self-compacting geopolymer concrete after freeze-thaw cycles. This represents the predicted transverse fundamental frequency of self-compacting geopolymer concrete after freeze-thaw cycles. The initial predicted value of the transverse fundamental frequency of self-compacting geopolymer concrete before freeze-thaw cycles;

[0026] The formula used to calculate the segregation resistance strength is as follows:

[0027] ;

[0028] In the formula, To determine the segregation resistance strength of self-compacting geopolymer concrete, is the anti-segregation limit load prediction value, is the anti-segregation test inter-support distance, is the segregation section width prediction value, is the segregation section height prediction value, is the test temperature, is the test humidity, and are the daily average temperature and daily average humidity of the arid and cold region, respectively;

[0029] wherein the formula for calculating the axial compressive strength is:

[0030] ;

[0031] wherein, is the axial compressive strength, is the anti-segregation limit load prediction value, is the anti-segregation bearing area prediction value.

[0032] Further, the fitness function of the corresponding prepared material mix proportion combined concrete is calculated according to the mechanical property parameters combined with the flow characteristic parameters, wherein the expression for calculating the fitness function is:

[0033] ;

[0034] wherein, is the fitness function value of the different prepared material mix proportion combined concrete, is the setting time prediction value, is the T500 flow time prediction value, is the emptying time prediction value, is the flow time target value, is the emptying time target value.

[0035] Further, the prepared material mix proportion combination with the highest fitness value is obtained by genetic algorithm, which is used as the optimal prepared material mix proportion combination, wherein the specific logic includes: a group of prepared material mix proportion combinations are used as an individual, and the addition amount of each prepared material in the combination is used as a gene, wherein the specific logic is: all data of the prepared material mix proportion in the combination are coded as corresponding genes, and the same type of prepared materials are allelic genes, i.e. there are 6 genes in the first level of the optimized variable combination, including the addition amount of metakaolin, sodium silicate, sodium hydroxide, water reducing agent, coarse and fine aggregates and water. All prepared material mix proportion combinations are coded to obtain a number of individuals, i.e. ; based on all the obtained individuals, an initial population is constructed, and the initial population is calibrated as , and , Index representing different individuals in the initial population, and u = 1, 2, D, D is the total number of individuals in the initial population, wherein each individual has 6 genes, and the 6 genes correspond to the parameter values of metakaolin, sodium silicate, sodium hydroxide, water reducing agent, coarse and fine aggregate and water addition amount respectively.

[0036] Further, the genetic algorithm is used to obtain the optimal preparation material mix proportion combination, and the logic for obtaining the optimal preparation material mix proportion combination is that: selection, crossover and mutation operations are performed on the initial population in cycles, the individual with the highest fitness value in the previous population is put into the next iteration population, and an iteration population containing multiple new individuals is generated, it is judged whether the maximum iteration number is reached, if the maximum iteration number is greater than the maximum iteration number, the preparation material mix proportion combination with the maximum fitness value appearing in the iteration process is selected as the optimal preparation material mix proportion combination, otherwise the generated iteration population is taken as the initial population for iteration operation until the iteration termination condition is met, wherein the iteration termination condition is the set maximum iteration number.

[0037] Compared with the prior art, the beneficial effects of the present application are:

[0038] Firstly, the method adopts the combination of experiment and calculation, analyzes the flow characteristic parameters and mechanical characteristic parameters of the concrete samples with different preparation material mix proportions in detail. The optimization of the mix proportion provides a rich data basis and also ensures a comprehensive understanding of the performance of the concrete. By establishing a neural network model, the relationship between the complex mix proportion and the performance can be expressed in the form of a mathematical model, further improving the scientificity of the mix proportion design. Secondly, based on the generated feature prediction model, various mix proportion combinations can be quickly generated and their performances can be predicted within the known material addition range. The number of tests and time cost are reduced, and by combining with the genetic algorithm, the method can also effectively optimize the mix proportion combination and find the scheme with the highest fitness value. Such an optimization process not only improves the efficiency of the mix proportion design, but also ensures that the selected scheme can meet the strict requirements of the concrete performance in the dry and cold regions, reducing the engineering quality problems caused by unsuitable mix proportions. Finally, through simulation and prediction, ideal concrete performance can be obtained under different conditions. The application of this method will provide more reliable material guarantee for building construction in dry and cold regions, and improve the safety and durability of the project. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 It is a whole method flowchart of the present application;

[0040] Figure 2 It is a column chart of flow characteristic parameters corresponding to different metakaolin proportions;

[0041] Figure 3Fitting curve graph of metakaolin ratio-flowing time;

[0042] Figure 4 Fitting curve graph of metakaolin ratio-setting time;

[0043] Figure 5 Fitting curve graph of emptying time-metakaolin ratio;

[0044] Figure 6 Distribution graph of axial compressive strength and anti-segregation strength of samples corresponding to different metakaolin ratios;

[0045] Figure 7 Fitting curve graph of metakaolin ratio-axial compressive strength;

[0046] Figure 8 Fitting curve graph of metakaolin ratio-anti-segregation strength;

[0047] Figure 9 Statistical graph of relative dynamic elastic modulus corresponding to freeze-thaw times;

[0048] Figure 10 Bar chart of axial compressive strength and anti-segregation strength of concrete under different environmental parameters. DETAILED DESCRIPTION

[0049] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with specific embodiments.

[0050] It should be noted that, unless otherwise defined, technical terms or scientific terms used in the present application should be understood as their common meanings by those skilled in the art to which the present application belongs. The terms "first", "second" and similar terms used in the present application do not represent any order, number or importance, but are only used to distinguish different components. The terms "include" or "contain" and similar terms mean that the elements or objects before the terms cover the elements or objects listed after the terms and their equivalents, without excluding other elements or objects. The terms "connect" or "connected" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms "up", "down", "left", "right" and the like only represent relative positional relationships, which can change accordingly when the absolute positions of the described objects change.

[0051] Embodiment:

[0052] Please refer to Figures 1-10 The present application provides a technical solution:

[0053] A self-compacting geopolymer concrete mix proportion design method considering arid and cold regions, specific steps comprising:

[0054] Step 1: Obtain the flow characteristic parameters and the mechanical characteristic parameters of the self-compacting geopolymer concrete samples under different preparation material ratios through experiments; the preparation materials include metakaolin, sodium silicate, sodium hydroxide, water reducing agent, coarse and fine aggregates, and water; the flow characteristic parameters include T500 flow time, emptying time, and setting time.

[0055] The mechanical characteristic parameters include compressive failure load, compressive bearing area, anti-segregation limit load, segregation section width, segregation section height, and transverse fundamental frequency of the concrete before and after thawing;

[0056] Obtain the flow characteristic parameters of the self-compacting geopolymer concrete samples through experiments, wherein the T500 flow time in the flow characteristic parameters specifically refers to the time from the start of lifting the slump cone to the flow to the place with a diameter of 500 mm of the slump cone, which is obtained through the slump test;

[0057] The emptying time is obtained through the V-box flow test, specifically referring to the time required for the concrete sample to completely empty after filling the V-box.

[0058] Metakaolin is a high-activity mineral admixture with a layered silicate structure, which is amorphous aluminum silicate formed by calcining kaolin at a high temperature above 800°C. In this experiment, red metakaolin that has passed the detection of Henan Gongyi Wanda Detection Center is used as the cementitious material, with an effective ingredient of up to 92.7% and a silicon-aluminum ratio of 1.0, and high purity.

[0059] Based on the research of scholars at home and abroad, it is known that sodium water glass as an alkali activator has better effect. In this embodiment, flaky sodium hydroxide produced by Inner Mongolia Junzheng Chemical Industry Co., Ltd. is used, with a purity of up to 98.5%.

[0060] In this embodiment, powdered instant sodium silicate is used, with a molar ratio of SiO2 to Na2O of 2.86, i.e. a sodium silicate modulus of 2.86. A certain amount of sodium hydroxide can be added to adjust the modulus to prepare the required alkali activator modulus.

[0061] Aggregates are granular materials that play a skeleton and filling role in concrete and mortar, and are divided into coarse aggregates and fine aggregates. Fine aggregate has a particle size of 0.16 to 5 mm, and is generally natural sand; coarse aggregate has a particle size of more than 5 mm, and is generally gravel or pebble. To ensure that the clay content of the aggregate meets the requirements, the aggregate is washed before the test to reduce its clay content and oil stains, and is dried for use. The fineness modulus is 2.67 after screening calculation, which belongs to medium sand, and the grading interval is in zone II, with a clay content of ≤2%. The coarse aggregate uses continuously graded natural gravel from a sand and gravel plant, with a continuous particle size of 5 to 15 mm, a maximum particle size of not more than 15 mm, and a clay content of <1%.

[0062] For water, domestic water is used.

[0063] The viscosity, reactivity and alkalinity of the alkali activator have a great influence on the geopolymerization reaction, so it needs to be adjusted. In order to ensure the smooth progress of the geopolymer reaction, the PH of the water glass should be greater than 13.6, and the alkalinity can be improved by adding sodium hydroxide solution, but at the same time the modulus of the water glass will be reduced.

[0064] In order to adjust the modulus and concentration of the alkali activator to meet the requirements of concrete materials, sodium hydroxide solution is used to adjust the 40% concentration water glass solution prepared, and the concentration of sodium hydroxide solution and the mass ratio of sodium hydroxide solution to water glass solution are used to control the alkali activator. In order to improve the rapid dissolution of powdered instant sodium silicate in water, the water is heated before mixing, at the same time, the designed concentration of sodium hydroxide solution is prepared according to the mixing ratio, and finally the mixing is carried out according to the mass ratio of sodium hydroxide solution to water glass solution. Since sodium hydroxide solution releases heat, in order to ensure the preparation of concrete at room temperature, the mixed solution is sealed with a barrel cover, and then aged at room temperature for 24h for later use.

[0065] The T500 flow time is obtained by the following test steps: timing from the start of lifting the slump cone, and the time when the flow reaches a diameter of 500mm. Attention should be paid to the lifting speed of the cone and the measurement of time and other factors in the test. T500 should be controlled between 2s and 5s.

[0066] The emptying time is obtained by the V-shaped box flow test, and the specific steps include: after the V-shaped funnel is washed clean with water, it is placed on the rack with its top surface horizontal and the body side vertical. Ensure that the funnel is stable. Wipe the inner surface of the funnel with a damp cloth to keep it wet. Place a receiving container under the outlet of the funnel to receive the concrete mixture. Before filling the funnel with the concrete mixture sample, make sure that the bottom cover of the funnel outlet is closed. Fill the funnel with the concrete mixture sample until it overflows, and the concrete mixture sample is about 11L. Use a spatula to scrape the top surface of the concrete mixture along the upper end of the funnel.

[0067] Scrape the top surface of the concrete mixture, and let it stand for 10s 2s, quickly open the bottom cover of the discharge port of the funnel, and record the time from the moment of opening the cover to the complete emptying of the concrete mixture in the funnel. The moment when the light is observed to pass through from the top of the funnel is defined as the emptying time of the concrete. At the same time, observe and record whether the concrete mixture is blocked or not. It is advisable to conduct two tests on the sample within 5 minutes, and the arithmetic mean of the results of the two tests should be taken as the test result, with the result accurate to 0.1s. The test result should meet the following requirements: the emptying time of the concrete mixture should be accurate to 0.1s. The flow condition of the concrete mixture should be recorded in addition to the emptying time, such as whether there is a blocked condition or not during the flow process.

[0068] The setting time has a great influence on the use of emergency repair non-vibrating polymer concrete. In order to meet the requirements of construction, it is necessary to analyze the setting time of geopolymer. The setting time is analyzed by using a Vicat apparatus. When measuring the setting time, the mixed slurry is poured into the Vicat mold and the excess part is scraped off with a spatula, and then placed in a curing box for curing, and then the setting time is measured. When the initial setting test needle sinks to 4mm from the bottom plate, it is determined that it reaches the initial setting state; when the final setting test needle sinks into the test piece by 0.5mm, that is, the annular accessory cannot leave marks on the test piece, it is determined that it reaches the final setting state. Table 1 shows the changes of the flow characteristic parameters of the concrete samples corresponding to different proportions of metakaolin.

[0069] Table 1: Partial statistical table of flow characteristic parameter changes

[0070]

[0071] Through the analysis of sample data, it can be observed that there is a certain correlation between the proportion of metakaolin in cementitious materials and the test results. For example, as the sample number increases, that is, the proportion of metakaolin increases, the T500 flow time decreases significantly. From 6.5 seconds of sample number 1 to 3.6 seconds of sample number 10, the flow time gradually decreases, indicating that under a high proportion of metakaolin, the fluidity of the slurry is improved. In addition, the emptying time also shows a downward trend, from 12.5 seconds of sample number 1 to 8.4 seconds of sample number 10. This shows that as the proportion of metakaolin increases, the fluidity and workability of the slurry are improved.

[0072] At the same time, the setting time also shows a decreasing trend, the setting time of sample number 1 is 30 minutes, and the setting time of sample number 10 is reduced to 12 minutes. This shows that the increase of metakaolin has a significant effect on the setting performance, which may be due to the fine particle structure of metakaolin enhancing the reactivity of the cement slurry, thereby accelerating the setting process.

[0073] Step 2: Establish a mapping relationship between the material proportioning and the flow characteristic parameters and the mechanical characteristic parameters, construct a training sample set, train the neural network model based on the training sample set, and obtain a prediction model for predicting the flow characteristic parameters and the mechanical characteristic parameters.

[0074] The neural network model is trained based on the training sample set, and a long short-term memory network model (LSTM model) is used as the basis to establish the neural network model. The LSTM model selects an activation function and an optimization algorithm, wherein a Tanh function is selected as the activation function, and Adam is selected as the optimization algorithm of the LSTM model. The formula of the Tanh function is:

[0075] ;

[0076] In the formula, wherein f represents the Tanh function, and x represents the independent variable. wherein x represents the input weight sum of the neuron, that is, the result of the input from the previous layer after being weighted and summed by the neuron.

[0077] Meanwhile, the hyperparameters of the LSTM model are set, including the number of network layers, the number of iterations, the learning rate, the batch size, the number of training times, the batch processing quantity, and the number of hidden layer neurons.

[0078] The number of network layers is set to a 4-layer network structure, the number of iterations is set to 300, the learning rate is set to 0.001, the batch size is set to 32, the number of training times is set to 100, the batch processing quantity is set to 256, and the number of hidden layer neurons is set to 32.

[0079] The trained feature prediction model inputs the preparation material proportioning data and outputs the predicted values of the corresponding flow characteristic parameters and mechanical characteristic parameters.

[0080] In practical applications, the experimental data of concrete proportioning may not be uniformly distributed. LSTM can handle irregular time interval data, making it more effective in dynamic prediction. For example, at different test stages, the performance parameters of concrete may be recorded at different time points. LSTM can adapt to this irregularity and provide more reliable output. Traditional recurrent neural networks (RNN) are prone to gradient vanishing or gradient explosion when processing long sequence data, making model training difficult. LSTM effectively controls the inflow, outflow, and forgetting of information by introducing a gating mechanism, significantly improving the model's learning ability for long sequences. This feature makes LSTM more robust in handling complex relationships between concrete proportioning and performance.

[0081] Step 3: A plurality of mix proportion combinations are generated within the preset range of the mixing amount of each material, input into the prediction model, and the corresponding flow characteristic parameter prediction value and mechanical characteristic parameter prediction value are obtained, and the mechanical performance parameter is calculated based on the mechanical characteristic parameter prediction value; the fitness function value of each mix proportion combination is calculated in combination with the flow characteristic parameter prediction value.

[0082] The mechanical characteristic parameters include the compressive failure load, the compressive bearing area, the anti-segregation limit load, the segregation section width, the segregation section height, and the transverse fundamental frequency of the concrete before and after thawing.

[0083] The specific method for obtaining the mechanical characteristic parameters is as follows:

[0084] The fast freezing method is adopted, and the test is carried out on an HC-HDK9 / F type concrete rapid freezing and thawing test machine. The size of the test piece is 100mmx100mmx400mm, and three test pieces are tested at a time. The test piece is demolded and numbered and then cured in water for 28 days. When the test piece reaches the curing age, the freezing and thawing test is carried out. The test piece is taken out in time when the curing age reaches 28 days, and the initial value of the transverse fundamental frequency is measured. The test piece is placed in the center of the test piece box, and then the test piece box is placed in the test piece rack in the freezing and thawing box, and clean water is injected into the test piece box. During the entire test process, the water level in the box should always be at least 5mm higher than the top surface of the test piece. The freezing and thawing cycle process should meet the following provisions: each freezing and thawing cycle is completed within 2 to 4 hours, and the melting time should not be less than 1 / 4 of the entire freezing and thawing cycle time. During the freezing and thawing processes, the lowest and highest temperatures in the center of the test piece are controlled within (-18 2) ℃ and (5 2) ℃, respectively. At any time, the temperature in the center of the test piece is maintained within the range of -20℃ to 7℃. The time required for each test piece to drop from 3℃ to -16℃ should not be less than 1 / 2 of the freezing time; the time required for each test piece to rise from -16℃ to 3℃ should not be less than 1 / 2 of the entire melting time, and the temperature difference between the inside and outside of the test piece should not exceed 28℃; the conversion time between freezing and melting should not exceed 10min. The transverse fundamental frequency of the test piece is measured every 25 freezing and thawing cycles. Before measurement, the surface scum of the test piece is cleaned and the surface water is wiped off, and the initial value of the transverse fundamental frequency is measured. After measurement, the transverse fundamental frequency is measured again as the transverse fundamental frequency after the freezing and thawing cycle test.

[0085] The flexural strength test is performed according to the Standard Test Methods for Mechanical Properties of Ordinary Concrete, on a YAW-3000 electro-hydraulic servo pressure testing machine. The specimen size is 100mmx100mmx400mm, and three specimens are tested at a time. The specific test steps are as follows: the specimen is taken out from the curing box and tested in time, and the surface of the specimen and the pressure plate of the testing machine are wiped dry with a cloth; the positions of the two supports are adjusted, and the test block is placed on the supports, with the pressure surface being the side surface of the test block during molding. The contact surfaces of the supports and the pressure surface with the test block are smooth and uniform; the load is applied uniformly and continuously, the loading speed is controlled by force, and the uniform loading speed is 0.05MPa / s until the specimen is destroyed; the failure load of the specimen, the fracture position of the lower edge of the specimen, the width of the segregation section and the height of the segregation section are recorded.

[0086] The compressive strength test method is performed according to the Standard Test Methods for Mechanical Properties of Ordinary Concrete, including cube compressive and cylindrical axial compression tests. The cube compressive strength test of the test block is performed on a YAW-3000 electro-hydraulic servo pressure testing machine, and the specimen size is 100mmx100mmx100mm. Three specimens are tested at a time, and the compressive failure load and the compressive pressure area are recorded.

[0087] Based on the obtained predicted values of the mechanical characteristic parameters, the mechanical performance parameters of the corresponding prepared material proportioning combined concrete are calculated, and the formula for calculating the relative dynamic elastic modulus is:

[0088] ;

[0089] In the formula, is the relative dynamic elastic modulus of the self-compacting geopolymer concrete after n freeze-thaw cycles, is the transverse fundamental frequency predicted value of the self-compacting geopolymer concrete after n freeze-thaw cycles, is the transverse fundamental frequency initial value of the self-compacting geopolymer concrete before freeze-thaw cycles; wherein Table 2 shows the relative dynamic elastic modulus corresponding to different concrete samples under different freeze-thaw times.

[0090] Table 2: Relative dynamic elastic modulus statistical table

[0091]

[0092] The formula for calculating the anti-segregation strength is:

[0093] ;

[0094] In the formula, is the anti-segregation strength of the self-compacting geopolymer concrete, is the anti-segregation limit load predicted value, is the distance between the anti-segregation test supports, is the predicted value of the segregation cross-sectional width, is the predicted value of the segregation cross-sectional height, is the test temperature, is the test humidity, and are the average daily temperature and average daily humidity of the arid-cold region, respectively.

[0095] It is worth noting that the mechanical parameters measured in the test are corrected by the average daily temperature and average daily humidity of the arid-cold region, so that the data conforms to the performance in the actual environment, the distance between the supports in the segregation resistance test is the distance between the supports set in the segregation resistance test.

[0096] represents the segregation resistance strength. It reflects the tolerance of concrete under segregation conditions, usually expressed as the force per unit area that can be tolerated; the predicted value of the segregation resistance limit load represents the maximum load that the material can withstand before segregation occurs. The larger this value, the stronger the material's load-carrying capacity under segregation conditions; the distance between the supports in the segregation resistance test is a geometric property that affects material strength. The larger the distance between the supports, the greater the material needs to resist the effects of bending and segregation; the segregation resistance strength is inversely proportional to the width and height of the segregation cross-section, reflecting the influence of geometric shape on segregation resistance strength. As the width and height increase, the risk of segregation of the material may increase.

[0097] where the hydration reaction of cement is significantly reduced at low temperatures, resulting in slow development of concrete strength. This may affect the overall stability of the concrete, increasing the risk of segregation, low temperature usually weakens the fluidity of concrete and increases the viscosity, which may cause the aggregate to settle in the concrete, increasing the possibility of segregation, therefore the segregation resistance strength of self-compacting geopolymer concrete is proportional to the average daily temperature of the arid-cold region. In a low humidity environment, water evaporation speed is accelerated, which may cause the surface of the concrete to lose water too quickly, leading to surface drying, cracking and segregation, low humidity may cause changes in the consistency of the concrete paste, affecting the distribution of aggregate, and thus increasing the risk of segregation, therefore the segregation resistance strength of self-compacting geopolymer concrete is proportional to the average daily humidity of the arid-cold region.

[0098] where the formula for calculating the axial compressive strength is:

[0099] ;

[0100] where, is the axial compressive strength, is the predicted value of the compressive failure load, is the predicted value of the compressive bearing area.

[0101] Where the basic principle of the formula is the stress concept in mechanics. Stress is the force applied on a unit area, calculated by dividing the compressive load by the bearing area to calculate the compressive strength of the material. Table 3 shows the changes in axial compressive strength and segregation resistance under different mechanical properties of the corresponding sample concrete.

[0102] Table 3: Partial statistical table of axial compressive strength and segregation resistance data

[0103]

[0104] Through the analysis of the given sample data, it can be observed that there is a certain positive correlation between the proportion of metakaolin in the cementitious material and the mechanical properties. The relationship between the sample number and the axial compressive strength is very obvious. With the increase of the proportion of metakaolin, the compressive strength gradually increases from 20 MPa of sample number 1 to 35 MPa of sample number 10. This indicates that the increase of metakaolin may effectively improve the strength performance of concrete and enhance the bearing capacity of the structure.

[0105] Secondly, the segregation resistance also shows a gradually rising trend. The segregation resistance of sample number 1 is 3.5 MPa, while the segregation resistance of sample number 10 increases to 5.9 MPa. This result shows that increasing the proportion of metakaolin is beneficial to improve the segregation resistance of the mixture, reduce the delamination phenomenon of the material during construction and use, and thus improve the overall construction quality and long-term reliability.

[0106] Where the influence of daily average temperature and daily average humidity in arid and cold regions on the axial compressive strength is the same as the influence of self-compacting geopolymer concrete on the segregation resistance, which is not repeated here. Table 4 shows the influence of the axial compressive strength and the segregation resistance of concrete under the environmental conditions in arid and cold regions.

[0107] Table 4: Partial statistical table of actual values of axial compressive strength and segregation resistance under specific environmental conditions

[0108]

[0109] In terms of environmental temperature and humidity, with the increase of sample number, the environmental temperature gradually rises from -20°C to -11°C, and the environmental humidity remains relatively stable (20% to 22%). This environmental condition has a significant impact on the performance of concrete in cold climates, and lower temperature and humidity may affect the hydration reaction of cement and the strength development of concrete.

[0110] The results of the actual axial compressive strength and the actual anti-washing strength also reflect the relationship between the actual values and the theoretical values. The actual axial compressive strength gradually increases from 14 MPa of sample No. 1 to 26 MPa of sample No. 10, showing a certain improvement, but is lower than the theoretical value. Related to the influence of environmental temperature, humidity and other factors on the strength of concrete. Similarly, the actual anti-washing strength also shows a corresponding increase, from 2.0 MPa of sample No. 1 to 3.8 MPa of sample No. 10, although it is also lower than the theoretical value, but shows the adaptability of the material in extreme environment.

[0111] According to the mechanical performance parameters combined with the flow characteristic parameters, the fitness function of the corresponding prepared material mix proportion combined concrete is calculated, wherein the expression for calculating the fitness function is:

[0112]

[0113] In the formula, is the fitness function value of the different prepared material mix proportion combined concrete, represents different prepared material mix proportion combinations, is the setting time prediction value, is the T500 flow time prediction value, is the emptying time prediction value, is the flow time target value, is the emptying time target value.

[0114] Among them, the mechanical performance parameters of concrete directly affect its performance and whether it can be used, so the anti-washing strength, axial compressive strength and relative dynamic elastic modulus of self-compacting geopolymer concrete are taken as the positive optimization target, and the comprehensive influence is represented by the form . For the setting time, the fast setting concrete can reduce the construction time, can quickly enter the subsequent construction step, can reach a certain strength in a short time, and is suitable for emergency repair engineering occasions, so the setting time is inversely proportional to the fitness function value, and the logarithmic function is used to balance the influence of small setting time on fitness. Shorter setting time will improve the fitness and encourage the use of fast setting concrete.

[0115] and These two terms represent the difference between the predicted flow time and the target flow time, and the difference between the predicted emptying time and the target emptying time, respectively. The absolute value is used to ensure that whether the predicted value is higher or lower than the target value, it will have a corresponding effect on the fitness, and by adding the two error terms, the consideration of both flow performance and emptying performance is emphasized. Therefore, the difference between the emptying time and the flow time and the target value is inversely proportional to the fitness function value.​

[0116] Step 4: iteratively optimize the mixture ratio combination by using the genetic algorithm, select the mixture ratio combination with the highest fitness function value as the optimal mixture ratio, and prepare the concrete accordingly.

[0117] The highest fitness value of the preparation material mixture ratio combination is obtained by the genetic algorithm, which is used as the optimal preparation material mixture ratio combination, and the specific logic includes: a group of preparation material mixture ratio combinations is used as an individual, and the addition amount of each preparation material in the combination is used as a gene, and the specific logic is: all data of the preparation material mixture ratio in the combination is encoded as corresponding genes, and the same type of preparation material is allelic, that is, there are 6 genes in the first level of optimization variable combination, including the addition amount of metakaolin, sodium silicate, sodium hydroxide, water reducing agent, coarse and fine aggregate and water, all preparation material mixture ratio combinations are encoded, and a plurality of individuals are obtained, that is, ; based on all the obtained individuals, an initial population is constructed, and the initial population is calibrated as , and , represents the index of different individuals in the initial population, and u=1,2, , D, wherein each individual has 6 genes, and the 6 genes correspond to the parameter values of the addition amount of metakaolin, sodium silicate, sodium hydroxide, water reducing agent, coarse and fine aggregate and water.

[0118] The optimal preparation material mixture ratio combination is obtained by using a multi-objective genetic algorithm, and the specific logic is: the initial population is selected, crossed and mutated, the individual with the highest fitness in the previous population is put into the next iteration population, and a plurality of new individuals are generated, it is judged whether the maximum iteration number is reached, if the maximum iteration number is greater than the maximum iteration number, the preparation material mixture ratio combination with the largest fitness value in the iteration process is selected as the optimal preparation material mixture ratio combination, otherwise the generated iteration population is used as the initial population for iteration operation until the iteration termination condition is met, wherein the iteration termination condition is the set maximum iteration number.

[0119] The initial population is selected, crossed and mutated, the individual with the highest fitness in the previous population is put into the next iteration population, and a plurality of new individuals are generated, it is judged whether the maximum iteration number is reached, if the maximum iteration number is greater than the maximum iteration number, the preparation material mixture ratio combination with the largest fitness value in the iteration process is selected as the optimal preparation material mixture ratio combination, otherwise the generated iteration population is used as the initial population for iteration operation until the iteration termination condition is met, wherein the iteration termination condition is the set maximum iteration number.

[0120] ;

[0121] In the formula, represents the fitness of the th individual in the initial population, is the The probability of an individual being selected, where ,and Given a positive integer, an individual is selected using a roulette wheel method; the system randomly generates one. The random number within the interval is distributed within the selection interval of the corresponding individual based on the probability of being selected. This random number determines the individual to be selected in this round.

[0122] The logic for performing the crossover operation is as follows: Perform the crossover operation on two selected individuals, exchanging the two individuals at a certain gene locus and the chromosome segment after that gene locus to obtain two crossover individuals. The logic for performing the crossover operation on the two selected individuals is: using a single-point crossover method.

[0123] The logic for performing mutation operations is as follows: The mutation operation is performed on the two new individuals generated by the crossover operation, based on a set mutation probability. The selected gene is mutated to become one of its alleles. The mutated gene is then used to replace the gene at that allele, thus obtaining a new individual.

[0124] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0125] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0126] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0127] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A mix design method for self-compacting geopolymer concrete considering arid and cold regions, characterized in that, The specific steps include: The flow characteristics and mechanical characteristics after hardening of self-compacting geopolymer concrete samples with different material ratios were obtained through experiments. The materials included metakaolin, sodium silicate, sodium hydroxide, water-reducing agent, coarse and fine aggregates, and water. The flow characteristics included T500 flow time, evacuation time, and setting time. Establish a mapping relationship between the material ratio and the flow characteristic parameters and mechanical characteristic parameters, construct a training sample set, train the neural network model based on the training sample set, and obtain a prediction model for predicting the flow characteristic parameters and mechanical characteristic parameters. Multiple mix proportion combinations are generated within the preset dosage range of each material, and input into the prediction model to obtain the corresponding predicted values ​​of flow characteristic parameters and mechanical characteristic parameters. Mechanical performance parameters are calculated based on the predicted values ​​of mechanical characteristic parameters. The fitness function value of each mix proportion combination is calculated in combination with the predicted values ​​of flow characteristic parameters. Genetic algorithms are used to iteratively optimize mix proportion combinations, and the mix proportion combination with the highest fitness function value is selected as the optimal mix proportion, and concrete is prepared accordingly. The mechanical characteristic parameters include compressive failure load, compressive bearing area, segregation ultimate load, segregation section width, segregation section height, and transverse fundamental frequency of concrete before and after thawing. The flow characteristic parameters of self-compacting geopolymer concrete samples were obtained through experiments. Among them, the T500 flow time in the flow characteristic parameters specifically refers to the time from the start of the lifting of the slump cone to the point where the flow reaches a diameter of 500 mm in the slump cone, which was obtained through slump test. The emptying time was obtained through a V-box outflow test, specifically referring to the time required for the concrete mixture to be completely emptied after the concrete sample filled the V-box. Based on the predicted values ​​of the obtained mechanical characteristic parameters, the mechanical performance parameters of the concrete with the corresponding material mix proportions are calculated. The formula used to calculate the relative dynamic elastic modulus is as follows: In the formula, The relative dynamic modulus of elasticity of self-compacting geopolymer concrete after freeze-thaw cycles. The predicted transverse fundamental frequency of self-compacting geopolymer concrete after freeze-thaw cycles. The initial predicted value of the transverse fundamental frequency of the self-compacting geopolymer concrete before freeze-thaw cycles; The formula used to calculate the segregation resistance strength is as follows: In the formula, To determine the segregation resistance strength of self-compacting geopolymer concrete, The predicted value of the ultimate load against segregation. The distance between supports for the segregation resistance test. This is the predicted value for the segregation section width. This is the predicted value of the segregation section height. For the test temperature, To test humidity, and These represent the average daily temperature and average daily humidity in arid and cold regions, respectively. The formula used to calculate the axial compressive strength is as follows: In the formula, For axial compressive strength, This is the predicted value of the compressive failure load. This is the predicted value of the compressive bearing area.

2. The mix design method for self-compacting geopolymer concrete considering arid and cold regions according to claim 1, characterized in that: The neural network model is trained based on this training sample set. A neural network model is established using the Long Short-Term Memory (LSTM) network model as the foundation. For the LSTM network model, an activation function and optimization algorithm are selected, with the Tanh function chosen as the activation function and Adam as the optimization algorithm. The formula for the Tanh function is: In the formula, Represents the Tanh function, with the independent variable... This represents the weighted sum of the neuron's inputs, that is, the result of the weighted sum of the inputs received by the neuron from the previous layer; Simultaneously, the hyperparameters of the LSTM model are set, including: number of network layers, number of iterations, learning rate, batch size, number of training iterations, number of batches, and number of hidden layer neurons. The network is set to a 4-layer structure, the number of iterations is set to 300, the learning rate is set to 0.001, the batch size is set to 32, the number of training iterations is set to 100, the number of batches is set to 256, and the number of hidden layer neurons is set to 32. The input to the trained feature prediction model is the mix proportion data of the prepared material, and the output is the predicted values ​​of the corresponding flow characteristic parameters and mechanical characteristic parameters.

3. The mix design method for self-compacting geopolymer concrete considering arid and cold regions according to claim 2, characterized in that: The fitness function of the prepared concrete mix design is calculated based on the mechanical performance parameters and flow characteristic parameters. The expression used for the fitness function calculation is as follows: In the formula, To prepare the fitness function value of the material mix design for the concrete, This is the predicted value for setting time. This is the predicted flow time value for T500. This is the predicted value for the drainage time. The target value for the flow time. This is the target value for the emptying time.

4. The mix design method for self-compacting geopolymer concrete considering arid and cold regions according to claim 3, characterized in that: A genetic algorithm is used to obtain the optimal material mix ratio combination with the highest fitness value. The specific logic involves treating a set of material mix ratio combinations as an individual, and the amount of each material added within the combination as a gene. Specifically, all data related to the material mix ratios within a combination are encoded as corresponding genes. Materials of the same type are alleles. Therefore, the first level of optimization variables corresponds to six genes, including the amounts of metakaolin, sodium silicate, sodium hydroxide, water-reducing agent, coarse and fine aggregates, and water. Encoding all material mix ratio combinations yields several individuals, which are then... Based on all the obtained individuals, an initial population is constructed, and the initial population is labeled as... ,and , Represents the index of a distinct individual in the initial population, where u = 1, 2, D, where D is the initial total number of individuals in the population, and each individual Each has 6 genes, which correspond to a parameter value for metakaolin, sodium silicate, sodium hydroxide, water-reducing agent, coarse and fine aggregates, and water addition.

5. The mix design method for self-compacting geopolymer concrete considering arid and cold regions according to claim 4, characterized in that: The logic behind obtaining the optimal combination of preparation materials using a genetic algorithm is as follows: The initial population is repeatedly subjected to selection, crossover, and mutation operations. The individual with the highest fitness from the previous population is placed into the next iteration population, thus generating an iterative population containing multiple new individuals. It is then determined whether the maximum number of iterations has been reached. If it has, the combination of preparation materials with the highest fitness value during the iteration process is selected as the optimal combination. Otherwise, the generated iterative population is used as the initial population for iterative operations until the iteration termination condition is met, where the iteration termination condition is the set maximum number of iterations.

Images