Intelligent lateral confinement compression experiment device and method

Abstract

The application relates to the technical field of soil test, and discloses an intelligent side-limit compression experiment device and method.The device comprises an adaptive side-limit component, a loading module, a sensing module and a data processing module.The adaptive side-limit component is used for containing a to-be-tested aggregate, and comprises an arc-shaped steel plate, a servo push rod and a friction adjusting layer.The loading module is arranged above the adaptive side-limit component, and comprises a magneto-rheological damper and a loading plate.The sensing module is arranged in cooperation with the adaptive side-limit component, and is used for measuring corresponding multi-dimensional sensing data in the adaptive side-limit component.The data processing module is used for performing data processing on the multi-dimensional sensing data collected by the sensing module, and determining a test result corresponding to the to-be-tested aggregate based on a data processing result.The application realizes intelligent analysis and result output of test data, and improves test efficiency and accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geotechnical test, in particular to an intelligent side-limit compression experimental device and method. BACKGROUND

[0002] In geotechnical test, side-limit compression test is a key means to study the compression characteristics of granular aggregate, which is of great significance to the engineering application of recycled granular materials such as concrete solid waste aggregate. However, the existing side-limit compression device has many deficiencies. In terms of mechanical structure, the side-limit constraint is unreasonable. The traditional steel barrel is a rigid structure, the inner wall is smooth, and the friction between the barrel and the granular aggregate cannot be adjusted, which leads to uneven distribution of radial stress of the sample, deviating from the ideal side-limit condition. Moreover, the existing side-limit compression device has obvious defects in mechanical structure, data acquisition and processing, and test operation, and cannot meet the high-precision, high-efficiency and intelligent requirements of the side-limit compression test of granular aggregate. Therefore, a new type of intelligent side-limit compression device is needed to solve the above problems. SUMMARY

[0003] To solve the above technical problems, the embodiments of the present application provide an intelligent side-limit compression experimental device and method.

[0004] According to an aspect of the embodiments of the present application, an intelligent side-limit compression experimental device is provided, comprising: an adaptive side-limit component for containing a to-be-tested aggregate, the adaptive side-limit component comprising an arc-shaped steel plate, a servo push rod and a friction adjustment layer, the friction adjustment layer being used to simulate different roughness of side-limit friction coefficient, and the servo push rod being used to adjust the pushing force of the arc-shaped steel plate, so that the to-be-tested aggregate obtains uniform radial stress; a loading module, the loading module being arranged above the adaptive side-limit component, the loading module comprising a magneto-rheological damper and a loading plate, the loading plate being used to apply a vertical load to the to-be-tested aggregate in the adaptive side-limit component based on preset loading parameters, the magneto-rheological damper being arranged at the bottom of the loading plate, the magneto-rheological damper being used to start when the real-time vertical stress value on the to-be-tested aggregate is greater than a preset vertical stress threshold value, and adjust the damping coefficient through the magneto-rheological damper to adjust the impact energy of the loading plate; a sensing module, the sensing module being arranged in cooperation with the adaptive side-limit component, and being used to measure corresponding multi-dimensional sensing data in the adaptive side-limit component; a data processing module, the data processing module being used to process the multi-dimensional sensing data collected by the sensing module, and determine a test result corresponding to the to-be-tested aggregate based on the data processing result.

[0005] According to one aspect of the embodiments of this application, an intelligent lateral confinement compression test method is provided, applied to an intelligent lateral confinement compression test device. The method includes: acquiring an aggregate to be tested and determining the particle size distribution corresponding to the aggregate; determining the lateral confinement friction coefficient corresponding to an adaptive lateral confinement component based on the particle size distribution, determining the working parameters of a friction adjustment layer based on the lateral confinement friction coefficient, and loading the aggregate to be tested into the adaptive lateral confinement component according to preset loading parameters; after the aggregate to be tested is loaded, controlling a loading module to apply a vertical load to the aggregate to be tested based on preset loading parameters, and collecting real-time multi-dimensional sensing data corresponding to the adaptive lateral confinement component during the loading process; adjusting the preset loading parameters based on the real-time multi-dimensional sensing data, and determining target multi-dimensional sensing data based on the adjusted loading parameters, and determining the test result corresponding to the aggregate to be tested based on the target multi-dimensional sensing data.

[0006] According to one aspect of the embodiments of this application, adjusting the preset loading parameters based on the real-time multi-dimensional sensing data includes: determining the real-time state parameters corresponding to the aggregate to be tested based on the real-time multi-dimensional sensing data; inputting the real-time state parameters as input parameters to a machine learning module to determine the correction coefficient of the loading module through the machine learning module; and adjusting the preset loading parameters based on the correction coefficient.

[0007] According to one aspect of the embodiments of this application, the step of inputting the real-time state parameters as input parameters to a machine learning module to output correction parameters through the machine learning module includes: determining the real-time physical property parameters of the aggregate to be tested based on the real-time state parameters; inputting the real-time physical property parameters as input parameters to the machine learning module to output physical feature prediction values ​​through the machine learning module; determining the correction coefficient of the loading module based on the physical feature prediction values, and adjusting the preset loading parameters based on the correction coefficient.

[0008] According to one aspect of the present application, before the aggregate to be tested is placed into the adaptive lateral confinement component according to the preset loading parameters, the method further includes: obtaining the size information corresponding to the adaptive lateral confinement component, and determining the basic physical constraints corresponding to the aggregate to be tested based on the size information; determining the preset loading parameters corresponding to the aggregate to be tested based on the basic physical constraints, wherein the preset loading parameters include loading density, loading rate and single-layer loading thickness.

[0009] According to one aspect of the embodiments of this application, the method further includes: determining the sample loading type of the aggregate to be tested based on the particle size distribution, the sample loading type including continuous sample loading and layered sample loading; if the sample loading type is continuous sample loading, determining the cross-sectional area corresponding to the adaptive lateral confinement component based on the size information; determining the loose bulk density of the aggregate to be tested, and determining the sample loading rate of the aggregate to be tested based on the loose bulk density and the cross-sectional area.

[0010] According to one aspect of the embodiments of this application, the method further includes: during the process of placing the aggregate to be tested into the adaptive lateral confinement component according to preset sample loading parameters, determining the surface smoothness of the aggregate to be tested based on the real-time multi-dimensional sensing data; if the deviation between the surface smoothness and the standard smoothness is greater than a preset deviation threshold, adjusting the sample loading time of the aggregate to be tested to obtain an adjusted sample loading time; determining a target sample loading rate, a target sample loading density, and a target single-layer thickness based on the adjusted sample loading time, and correcting the preset sample loading parameters based on the target sample loading rate, the target sample loading density, and the target single-layer sample loading thickness to obtain the target sample loading parameters of the aggregate to be tested.

[0011] According to one aspect of the embodiments of this application, determining the test result corresponding to the aggregate under test based on the target multi-dimensional sensing data includes: determining the vertical stress, vertical strain, and real-time porosity of the aggregate under test based on the target multi-dimensional sensing data; plotting a first stress curve corresponding to the aggregate under test based on the vertical stress and vertical strain; taking the natural logarithm of the vertical stress to obtain ln stress, and plotting a second stress curve corresponding to the aggregate under test based on the real-time porosity and the ln stress; and determining the test result corresponding to the aggregate under test based on the first stress curve and the second stress curve.

[0012] According to one aspect of the embodiments of this application, the method further includes: determining real-time radial stress data within the adaptive side-limiting component based on the real-time multi-dimensional sensing data; determining a target thrust of the arc-shaped steel plate of the adaptive side-limiting component based on the real-time radial stress data; and determining a control strategy for the servo actuator of the adaptive side-limiting component based on the target thrust.

[0013] According to one aspect of the embodiments of this application, the method further includes: if the real-time multi-dimensional sensing data indicates that the real-time vertical stress on the aggregate to be tested is greater than a preset vertical stress threshold, then determining the target damping coefficient corresponding to the magnetorheological damper based on the real-time vertical stress; determining the adjustment strategy of the loading module based on the target damping coefficient, and controlling the loading module to execute the adjustment strategy.

[0014] In the technical solution provided in the embodiments of this application, the adaptive side-limiting component, through the synergistic effect of the arc-shaped steel plate, servo push rod, and friction adjustment layer, can simulate frictional environments with different roughness. It can also adjust the thrust of the arc-shaped steel plate with the help of the servo push rod, enabling the aggregate under test to obtain uniform radial stress, thus providing more realistic and stable stress conditions for the test. The loading module is positioned above the adaptive side-limiting component. The loading plate applies a vertical load according to preset loading parameters. The magnetorheological damper is activated when the real-time stress value of the aggregate under test exceeds a preset threshold. By adjusting the damping coefficient, it absorbs impact energy, effectively protecting the device and ensuring the stability of the loading process. The sensing module, in conjunction with the adaptive side-limiting component, can measure multi-dimensional sensing data, providing rich information for a comprehensive understanding of the aggregate state. The data processing module processes the collected multi-dimensional sensing data and determines the test results, realizing intelligent analysis and result output of the test data, improving test efficiency and accuracy.

[0015] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings:

[0017] Figure 1 This is a schematic diagram of the overall structure of an intelligent lateral compression experimental device, as shown in an exemplary embodiment of this application.

[0018] Figure 2 This is an exemplary embodiment shown in this application. Figure 1 A simplified block diagram of the adaptive lateral constraint component in the diagram.

[0019] Figure 3 This is an exemplary embodiment shown in this application. Figure 1 A simplified structural block diagram of the loading module and sensing module in the system.

[0020] Figure 4 This is a flowchart illustrating an intelligent side-limited compression experimental method according to an exemplary embodiment of this application.

[0021] Figure 5 The diagram illustrates a stress-strain curve and a porosity-stress relationship curve as shown in an exemplary embodiment of this application.

[0022] The reference numerals in the attached figures are explained as follows: 1. Adaptive lateral limiting component; 2. Loading module; 3. Sensing module; 4. Data processing module; 5. Machine learning module; 6. Control terminal; 11. Curved steel plate; 12. Servo push rod; 13. Friction adjustment layer; 14. Spiral sample loading device; 15. Rotary leveler; 16. Specific gravity measuring cell; 21. Servo motor; 22. Ball screw; 23. Magnetorheological damper; 24. Loading plate; 31. Gamma ray densitometer; 32. Pressure sensor; 33. Strain gauge; 34. Laser displacement sensor; 35. Humidity sensor. Detailed Implementation

[0023] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0024] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0025] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0026] In this application, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0027] First, it should be noted that in geotechnical testing, the lateral confinement compression test is a key method for studying the compressibility characteristics of granular aggregates, and it is of great significance for the engineering application of recycled granular materials such as concrete solid waste aggregates. Currently, steel drums are commonly used for lateral confinement compression tests on granular aggregates to obtain key parameters such as maximum and minimum dry density, specific gravity, maximum and minimum void ratio, and to plot stress-strain curves and the relationship between void ratio and ln stress. However, existing lateral confinement compression devices have many shortcomings. In terms of mechanical structure, the lateral confinement constraint is unreasonable. Traditional steel drums are rigid structures with smooth inner walls, and the friction between the drum and the granular aggregate cannot be adjusted, resulting in uneven radial stress distribution of the sample, deviating from ideal lateral confinement conditions, and affecting the accuracy of the measurement of maximum and minimum dry density. Furthermore, during compression, aggregate particles are prone to sliding along the drum wall, causing local density anomalies and affecting the accuracy of the aggregate test results.

[0028] To address the aforementioned issues, this application proposes an intelligent lateral compression experimental apparatus and an intelligent lateral compression experimental method.

[0029] Please see Figure 1 , Figure 1 This application illustrates an exemplary embodiment of an intelligent lateral constriction compression experimental apparatus, which is described in detail below. The intelligent lateral constriction compression experimental apparatus includes:

[0030] The adaptive side-confining component 1 is used to hold the aggregate to be tested. It includes an arc-shaped steel plate, a servo push rod, and a friction adjustment layer. The friction adjustment layer simulates the side-confining friction coefficients with different roughnesses, and the servo push rod adjusts the thrust of the arc-shaped steel plate to ensure uniform radial stress on the aggregate. The loading module 2 is positioned above the adaptive side-confining component. It includes a magnetorheological damper and a loading plate. The loading plate applies a vertical load to the aggregate within the adaptive side-confining component based on preset loading parameters. The magnetorheological damper is located at the bottom of the loading plate and activates when the real-time vertical stress value on the aggregate exceeds a preset vertical stress threshold. The damping coefficient is adjusted by the magnetorheological damper to regulate the impact energy of the loading plate. The sensing module 3 works in conjunction with the adaptive side-confining component to measure corresponding multi-dimensional sensing data within the component. The data processing module 4 processes the multi-dimensional sensing data collected by the sensing module and determines the test results corresponding to the aggregate based on the processing results.

[0031] For example, such as Figure 1As shown, in some feasible embodiments, the adaptive confinement component 1 can be a steel drum assembly, which can hold the aggregate to be tested. A loading module 2 is disposed above the adaptive confinement component 1, which can apply vertical stress to the aggregate within the assembly. A sensing module 3 is disposed within the adaptive confinement component 1, cooperating with it to measure multi-dimensional sensing data and transmitting this data to a data processing module 4. The data processing module 4 processes the multi-dimensional sensing data collected by the sensing module 3 to determine the test results corresponding to the aggregate, and then generates real-time test results including stress-strain curves and porosity-ln stress relationship curves based on the test results. Furthermore, in this embodiment, the intelligent side-limited compression experimental device also includes a machine learning module 5 and a control terminal 6. The machine learning module adjusts the loading parameters of the loading module 2 based on real-time multi-dimensional sensing data during the loading process and sends the adjustment strategy to the control terminal 6 so as to control and adjust each module through the control terminal 6.

[0032] Please see Figure 2 In some feasible embodiments, the adaptive lateral confinement steel barrel assembly is based on a steel barrel with a diameter of 150 mm and a height of 170 mm. Its structure includes a variable friction inner wall, into which a replaceable friction adjustment layer 13 is carefully embedded. This friction adjustment layer 13 can be made of polytetrafluoroethylene material with different roughnesses. By replacing the friction adjustment layer 13 with different roughnesses according to experimental requirements, various different lateral confinement friction coefficients can be accurately simulated. Simultaneously, miniature pressure sensors are cleverly arranged on the surface of the friction adjustment layer 13. These sensors can monitor the distribution of radial friction force in real time and accurately. Specifically, this friction force distribution can be expressed as:

[0033]

[0034] in, This is the frictional force (measured by a miniature force sensor on the inner wall). For normal pressure.

[0035] Optionally, the sidewall of the adaptive lateral confinement assembly is composed of four sections of arc-shaped steel plates 11, each connected to an independent servo push rod 12. Based on radial stress data fed back by a multi-parameter sensing system, the thrust of each arc-shaped steel plate 11 can be adjusted in real time to ensure uniform radial stress in the sample and meet the lateral confinement conditions. The adjustment amount ΔF is calculated according to the following formula:

[0036]

[0037] Where k is the proportionality coefficient. For the target radial stress, To measure radial stress, the automatic sample loading and leveling mechanism includes a spiral sampler 14 and a rotary leveler 15 mounted on the top of the steel drum. The spiral sampler controls the loading rate and thickness, while the rotary leveler 15 uses an ultrasonic sensor to detect the surface flatness of the aggregate and automatically adjusts the leveling force to ensure uniform loading. The formula for controlling the loading density of the aggregate under test can be expressed as:

[0038]

[0039] in, The sample mass is measured by a weighing sensor. The volume of the sample can be determined by the cross-sectional area of ​​the steel drum and the height of the sample.

[0040] Furthermore, a specific gravity measuring mechanism 16 is provided at the bottom of the adaptive lateral confinement component 1. This specific gravity measuring mechanism 16 can be a detachable specific gravity measuring tank, which can then measure the specific gravity of the bulk aggregate by the drainage method. The specific gravity calculation formula can be expressed as follows:

[0041]

[0042] in, For the quality of aggregate drying, This is to determine the mass of the aggregate in water, thereby enabling the determination of important physical parameters of the aggregate under test.

[0043] In one aspect of this embodiment, please continue to refer to Figure 3 The loading module 2 also includes a servo motor 21 and a ball screw 22. The servo motor 21 can drive the ball screw 22 to apply a vertical load. The loading rate is continuously adjustable within the range of 0.01-5 mm / min, and the loading force measurement range is 0-100 kN. The loading force control of the loading module can adopt a PID algorithm, and the specific control formula is as follows:

[0044]

[0045] in, To control the output, This is the proportionality coefficient. The integral coefficient is... These are the differential coefficients. This refers to the deviation between the target load and the actual load.

[0046] In one aspect of this embodiment, the loading module 2 also includes a loading plate 24 and a magnetorheological damper 23. The magnetorheological damper 23 and a pressure sensor 32 can be provided at the bottom of the loading plate 24. When the rate of change of the load change in the adaptive side confinement component is detected to be greater than the preset rate of change threshold based on the measurement result of the pressure sensor 32, the damping coefficient can be quickly adjusted by the magnetorheological damper 23 to absorb the impact energy and achieve protection of the aggregate to be tested and the adaptive side confinement component 1.

[0047] Optional, please continue reading Figure 3 The sensing module 3 also includes X-ray density meter 31, strain gauge 33, laser displacement sensor 34, and humidity sensor 35, wherein the latter is installed on the outside of the steel barrel of the adaptive side-confining assembly 1. The X-ray densitometer 31 allows for real-time measurement of the dry density of the aggregate under test. The formula for calculating the dry density of the aggregate under test is as follows:

[0048]

[0049] in, Wet density, Moisture content (can be measured by humidity sensor 35).

[0050] In some feasible embodiments, to achieve accurate stress and strain monitoring, a high-precision resistance strain gauge pressure sensor can be integrated into the bottom of the loading plate. Based on the Wheatstone bridge principle, this sensor acquires vertical loads in real time. The rigid coupling design between the sensor and the loading plate effectively suppresses stress diffusion effects, and the 100Hz high sampling frequency accurately captures the aggregate yield point. A radial strain field measurement system arranges six or more sets of foil strain gauges at equal intervals along the height of the inner wall of the steel drum, forming a 60° annular array covering the circumference of the drum wall. Combined with temperature self-compensation technology and a dynamic strain gauge, a radial strain distribution cloud map with a resolution of 1με is constructed, fully revealing the aggregate-drum wall interface friction effect. Vertical strain analysis employs dual laser displacement sensors (non-contact measurement; symmetrically arranged laser heads perform triangulation at a 45° measurement angle, achieving a displacement resolution of 0.001mm combined with an initial sample height of 150mm) to calculate the true axial strain in real time and eliminate mechanical deformation errors. The pore water pressure monitoring system embeds miniature piezoresistive sensors at the bottom center and at 1 / 3 and 2 / 3 of the height of the sidewalls of the steel drum. After Kalman filtering for noise reduction, the results are based on the Terzaghi effective stress principle (…). Corrected measured stress, where For effective stress, For the total stress, The pressure is pore water pressure. The displacement measurement system uses magnetostrictive displacement sensors arranged at the four corners of the loading plate, which, in conjunction with a dual-axis tilt sensor, reconstruct the three-dimensional displacement field. Then, a spatial geometric algorithm is used... (A=176.7cm) 2 The system calculates the volumetric compression in real time (using the cross-sectional area of ​​the barrel). It then works in conjunction with a gamma-ray densitometer to construct a volume and density model. ,in This enables dynamic mapping of aggregate density at the second level during the experiment.

[0051] In some embodiments of this application, the adaptive side-limiting component, through the synergistic action of the arc-shaped steel plate, servo push rod, and friction adjustment layer, can simulate friction environments with different roughness. It can also adjust the thrust of the arc-shaped steel plate with the servo push rod to ensure uniform radial stress in the tested aggregate, providing more realistic and stable stress conditions for the experiment. The loading module is positioned above the adaptive side-limiting component. The loading plate performs vertical loading according to preset loading parameters. The magnetorheological damper activates when the real-time stress value of the tested aggregate exceeds a preset threshold, absorbing impact energy by adjusting the damping coefficient, effectively protecting the device and ensuring the stability of the loading process. The sensing module, in conjunction with the adaptive side-limiting component, can measure multi-dimensional sensing data, providing rich information for a comprehensive understanding of the aggregate state. The data processing module processes the collected multi-dimensional sensing data and determines the experimental results, realizing intelligent analysis and output of experimental data, improving experimental efficiency and accuracy.

[0052] Furthermore, based on the above embodiments, please refer to... Figure 4 This application proposes an intelligent side-limited compression experimental method, which is applied to the intelligent side-limited compression experimental apparatus in any of the above embodiments.

[0053] like Figure 4 As shown, in one exemplary embodiment provided in this application, the specific implementation process of the above-mentioned intelligent side-limited compression experimental method includes at least steps S410 to S430, which are described in detail below:

[0054] Step S410: Obtain the aggregate to be tested and determine the corresponding particle size distribution of the aggregate.

[0055] Step S420: Determine the lateral friction coefficient corresponding to the adaptive lateral confinement component based on the particle size distribution, determine the working parameters of the friction adjustment layer based on the lateral friction coefficient, and put the aggregate to be tested into the adaptive lateral confinement component according to the preset sample loading parameters.

[0056] For example, when conducting aggregate-related tests, it is necessary to obtain the aggregate to be tested. This process requires precise collection of aggregate samples from appropriate sources based on the specific requirements and objectives of the test, ensuring that the obtained aggregate meets the test requirements in terms of properties and composition, and is representative and reliable. After collecting the aggregate, a detailed particle size distribution analysis is performed using professional sieving equipment and techniques. The aggregate is sieved according to different particle size ranges through a series of standard sieves, and the mass of the remaining aggregate on each sieve is accurately weighed. The percentage of aggregate in each particle size range is then calculated, ultimately determining the corresponding particle size distribution of the aggregate, providing crucial information for subsequent experimental operations.

[0057] After determining the particle size distribution of the aggregate, the next step is to determine the target friction conditioning layer corresponding to the adaptive side-confining component based on the particle size distribution. Since the friction characteristics between aggregates with different particle size distributions and the side-confining component differ during the test, in order to more accurately simulate the actual working conditions and ensure the accuracy of the test results, the target friction conditioning layer that matches the determined particle size distribution data can be selected from the pre-established correspondence model or empirical data table between different particle size distributions and friction conditioning layers. This ensures that the target friction conditioning layer can reproduce the friction environment of the aggregate in actual application to the greatest extent, making the test conditions closer to the real situation.

[0058] After determining the target friction conditioning layer, the aggregate to be tested is placed into the adaptive confinement assembly according to the preset loading parameters. These preset loading parameters cover various aspects such as the order of aggregate placement, the mass of each layer, the loading height, loading density, and loading rate. During the placement process, operators must strictly follow the preset parameters. For example, aggregates of different sizes or from different locations are placed layer by layer into the adaptive confinement assembly with the target friction conditioning layer installed, according to the placement order. After each layer is placed, it is weighed and confirmed according to the prescribed quality standards to ensure the accuracy of the mass of each layer. Simultaneously, the loading height must be controlled to avoid excessive impact force on the aggregate during placement, which could affect the original state of the aggregate and the test results. The aggregate can be compacted using a rotary leveler to eliminate voids between aggregates, ensuring a dense state within the adaptive confinement assembly. This guarantees that the entire loading process meets the test requirements and lays a solid foundation for the subsequent successful aggregate testing.

[0059] Step S430: After the aggregate to be tested is put into the test, the loading module is controlled to apply a vertical load to the aggregate to be tested based on the preset loading parameters, and real-time multi-dimensional sensing data corresponding to the adaptive lateral confinement component is collected during the loading process.

[0060] For example, after the aggregate to be tested is placed into the adaptive lateral confinement component, and it is ensured that the aggregate is accurately, uniformly, and densely arranged inside the component according to the preset loading parameters, the crucial loading and data acquisition stage begins. First, the loading module is meticulously configured according to the pre-set loading parameters for the experiment. These preset loading parameters are determined through rigorous theoretical analysis, preliminary simulations, and a comprehensive consideration of actual engineering needs, covering key elements such as loading rate, target loading value, and loading waveform. The loading rate determines the increase in load borne by the aggregate per unit time; different loading rates may cause different deformation and failure modes in the aggregate. The target loading value clarifies the maximum load expected to be borne by the aggregate in this experiment, which is related to the ultimate pressure that the aggregate may withstand in actual applications. The loading waveform simulates the change in load on the aggregate over time under actual working conditions, such as sine waves and square waves; different waveforms will have a unique impact on the mechanical response of the aggregate.

[0061] After the loading module completes precise configuration, the loading program is started, controlling the loading module to apply vertical loads to the aggregate under test according to preset loading parameters. As the vertical load is continuously applied, complex mechanical changes occur within the aggregate, with interactions such as rearrangement, compression, and friction between particles, and the aggregate as a whole deforms. Simultaneously, multimodal sensors arranged within the adaptive lateral confinement assembly begin real-time operation, comprehensively collecting real-time multi-dimensional sensing data during the loading process. As mentioned in the above embodiments, these multimodal sensors are diverse, including but not limited to strain gauges, laser displacement sensors, and... X-ray density meters, etc. Strain gauges can sensitively detect minute strain changes in aggregates under load, and by measuring strain, the stress state and deformation characteristics of aggregates can be further analyzed; laser displacement sensors accurately record the displacement of aggregates at different loading stages. Laser displacement sensors are crucial for studying the compressibility and resilience of aggregates.

[0062] Step S440: Adjust the preset loading parameters based on real-time multi-dimensional sensor data, and determine the target multi-dimensional sensor data based on the adjusted loading parameters, so as to determine the test results corresponding to the aggregate to be tested based on the target multi-dimensional sensor data.

[0063] For example, after acquiring real-time multi-dimensional sensor data during the loading process, this data must be immediately transmitted to a dedicated data processing module. This data processing module communicates with a machine learning model, which performs a comprehensive and in-depth analysis of the real-time multi-dimensional sensor data using preset judgment logic. Based on the analysis results, adjustment instructions are automatically generated according to preset rules to dynamically adjust the preset loading parameters. The direction and magnitude of the adjustment depend on the type and severity of the abnormal data. For example, if the pressure value is too low, the subsequent loading rate may be appropriately increased or the target loading value increased; if the strain increases too rapidly, the loading rate may be reduced or the increment of each loading step may be decreased to avoid premature aggregate failure. This dynamic adjustment process is real-time and continuous, ensuring that the loading process remains controllable and meets the experimental requirements. As the loading parameters are adjusted, the loading process continues, while the multi-modal sensors continuously collect data until the target multi-dimensional sensor data that meets the experimental requirements is obtained. This target data is obtained after multiple adjustments and optimizations of the loading parameters, and can more accurately and comprehensively reflect the true mechanical properties of the aggregate under test under specific loading conditions. Finally, the acquired multi-dimensional sensor data of the target is imported into professional test report generation software. The software further processes and analyzes the target data according to the preset report template and data analysis rules, and extracts key information such as the maximum bearing capacity of aggregate, elastic modulus, deformation characteristics, etc.

[0064] In some embodiments of this application, by accurately obtaining the particle size distribution of the aggregate to be tested and determining the target friction adjustment layer of the adaptive lateral confinement component accordingly, the actual stress environment of the aggregate can be better simulated, improving the accuracy of the test; the aggregate is placed according to the preset loading parameters to ensure the standardization and consistency of the loading; after placement, a vertical load is applied according to the preset loading parameters and real-time multi-dimensional sensor data is collected, which can comprehensively obtain various information of the aggregate during the loading process; the preset loading parameters are adjusted based on real-time data, which can achieve dynamic and precise control of the test process; finally, the target multi-dimensional sensor data is obtained and a test report is generated, which provides a reliable basis for in-depth research and evaluation of aggregate performance and helps to improve the scientificity and rationality of engineering design and construction.

[0065] Furthermore, based on the above embodiments, in one of the exemplary embodiments provided in this application, the specific implementation process of adjusting the preset loading parameters based on real-time multi-dimensional sensing data may further include steps S510 to S530, which are described in detail below.

[0066] Step S510: Determine the real-time state parameters of the aggregate to be tested based on real-time multi-dimensional sensor data.

[0067] Step S520: Input the real-time state parameters as input parameters to the machine learning module so that the machine learning module can determine the correction coefficient of the loading module.

[0068] Step S530: Adjust the preset loading parameters based on the correction coefficient.

[0069] For example, the machine learning dynamic adjustment analysis module 5 adjusts the loading rate through a reinforcement learning algorithm based on a comparison between real-time data and the prediction model. When the dry density increases slowly, the loading rate is increased to 1 mm / min; when it approaches the predicted maximum dry density, the loading rate is decreased to 0.2 mm / min. Simultaneously, based on feedback from the radial stress sensor, the thrust of the arc-shaped steel plate 11 is adjusted via the servo push rod 12 to ensure uniform radial stress.

[0070] Optionally, in some feasible embodiments, the prediction model determines the predicted state parameters corresponding to the aggregate to be tested, and then inputs the predicted state parameters and the real-time state parameters of the aggregate to be tested as input parameters into the machine learning module 5. Then, according to the corresponding correction parameters output by the machine learning module 5, the loading parameters of the loading module 2 are adjusted.

[0071] For example, the maximum dry density criterion can be used:

[0072]

[0073] in The coefficient of variation is the particle size distribution. To enable closed-loop optimization in machine learning, a parameter prediction error correction model is used.

[0074]

[0075] in, For preset loading parameters, These are real-time status parameters. To predict state parameters, To correct the coefficients, dynamic adjustments are made through reinforcement learning.

[0076] Therefore, the adaptive control coefficient for the loading rate can be expressed as:

[0077]

[0078] in, Ensures automatic deceleration when approaching maximum dry density.

[0079] In some embodiments of this application, real-time state parameters are determined based on real-time multi-dimensional sensor data and input into a machine learning module to obtain correction coefficients to adjust preset loading parameters. This enables intelligent, dynamic, and precise control of the loading process, making the experiment more closely resemble the actual situation and effectively improving the accuracy and reliability of the experimental results.

[0080] Furthermore, based on the above embodiments, in one exemplary embodiment provided in this application, the specific implementation process of inputting real-time state parameters as input parameters to the machine learning module so as to output corrected parameters through the machine learning module may further include steps S610 to S630, which are described in detail below:

[0081] Step S610: Determine the real-time physical property parameters of the aggregate to be tested based on the real-time state parameters.

[0082] Step S620: Input the real-time physical property parameters as input parameters to the machine learning module, so that the machine learning module can output the predicted value of physical features.

[0083] Step S630: Determine the correction coefficient of the loading module based on the predicted value of physical characteristics, and adjust the preset loading parameters based on the correction coefficient.

[0084] For example, the real-time physical property parameters of the aggregate under test can be determined based on the real-time state parameters collected by the sensing module. The real-time physical property parameters may include the real-time stress, real-time strain, real-time density change rate, real-time pore water pressure, and real-time gradation variation coefficient of the aggregate under test. Then, the real-time physical property parameters are input into the machine learning module, and the machine learning module outputs the corresponding physical feature prediction values. Then, the correction parameters of the loading module are determined based on the physical feature prediction values, and the preset loading parameters of the loading module are adjusted according to the correction parameters to obtain the target loading parameters of the loading module.

[0085] Optionally, the loading parameters of the loading module can be dynamically adjusted through reinforcement learning. For example, the experimental process optimization model uses a reinforcement learning algorithm to dynamically adjust the loading rate and compaction work with the goals of minimizing experimental time and maximizing parameter measurement accuracy. The reward function of reinforcement learning can be expressed as:

[0086]

[0087] in, For the test time, The prediction error is represented by α and β, which are weighting coefficients.

[0088] Furthermore, in some feasible embodiments, the machine learning module is also used for anomaly detection. Specifically, the anomaly detection model can utilize the Isolation Forest algorithm to identify anomalous data in the experiment. When the anomaly probability exceeds a threshold (e.g., 5%), an alert is issued and adjustments to the experimental parameters are suggested. The anomalous data detected by this anomaly detection model mainly includes anomalies in sensor data and their derived parameters collected in real time during the experiment. Specifically, firstly, there are anomalies in the raw sensor data, i.e., anomalies in the physical quantities collected in real time by the multi-parameter sensing system, such as the vertical stress monitored by the pressure sensor (…). Radial stress (e.g., stress value suddenly exceeds the normal range by 10%), radial strain measured by strain gauges, vertical strain measured by laser displacement sensors. Real-time dry density monitored by gamma ray densitometer ( The three-dimensional displacement field fed back by the magnetostrictive displacement sensor, the ultrapore water pressure (u) monitored by the miniature piezoresistive sensor, and the water content (w) measured by the humidity sensor, etc.; secondly, the abnormal calculation of derived parameters, that is, the abnormal engineering parameters calculated in real time by the data processing module, such as porosity. Density change rate strain hardening index Volume compression Coefficient of variation of gradation wait.

[0089] Optionally, the above-mentioned anomaly detection also includes anomalies in the experimental process state, i.e., logical contradictions between experimental operation and equipment state, such as fluctuations in sample density during the sample loading stage. ), Surface unevenness exceeds the limit ( During the loading phase, load control instability occurred (the deviation between the PID output and the measured value was too large), strain rate abruptly changed (e.g., the magnetorheological damper did not respond in time), and radial stress distribution was uneven (the stress difference between the four steel plate sections was extremely large). ), and the load mutation rate exceeding the threshold ( Security risks such as [missing information]. The model's decision-making logic is based on the input real-time five-dimensional state vector. The deviation of data points from the global distribution is assessed by dividing the path length using a tree structure, when the anomaly probability... The system will trigger an early warning and execute response measures, including automatically activating the magnetorheological damper to buffer the impact, adjusting the PID parameters or loading rate, and suggesting that the test be paused and the equipment (such as sensor calibration and sample uniformity) be checked.

[0090] In some embodiments of this application, real-time physical property parameters are determined by real-time state parameters and input into a machine learning module to obtain predicted values. Then, correction parameters are determined to adjust preset loading parameters, which can dynamically and accurately grasp changes in aggregate properties and optimize loading, effectively improving the accuracy and scientific nature of test results.

[0091] Furthermore, based on the above embodiments, in one of the exemplary embodiments provided in this application, before the aggregate to be tested is placed into the adaptive lateral confinement component according to the preset sample loading parameters, the specific implementation process of the above intelligent lateral confinement compression test method may also include steps S710 and S720, which are described in detail below.

[0092] Step S710: Obtain the size information corresponding to the adaptive lateral constraint component, and determine the basic physical constraints corresponding to the aggregate to be tested based on the size information.

[0093] Step S720: Determine the preset loading parameters corresponding to the aggregate to be tested based on the basic physical constraints. The preset loading parameters include loading density, loading rate and single-layer loading thickness.

[0094] For example, the first step is to obtain the dimensional information of the adaptive side-confining component. After accurately measuring the component and obtaining its key dimensional data, the basic physical constraints on the aggregate to be tested are determined based on this information. These constraints reflect the physical limitations imposed on the aggregate by the adaptive side-confining component, comprehensively considering the relationship between component size and aggregate properties. After clarifying the basic physical constraints, the preset loading parameters for the aggregate to be tested are further determined. These parameters cover multiple aspects, including loading density, loading rate, and loading thickness. Loading density determines the compactness of the aggregate during loading; loading rate is the amount of aggregate loaded per unit time, affecting loading efficiency and uniformity; and loading thickness specifies the range of aggregate loading thickness. These parameters collectively provide clear guidance for subsequent experimental operations.

[0095] Optionally, in some feasible embodiments, the loading rate and loading thickness are determined by basic physical constraints and real-time feedback control, while the loading height... Based on standard steel drum dimensions, for example, the diameter of the steel drum for the adaptive lateral confinement assembly. ,high The corresponding maximum sample loading height is set to Simultaneously reserve Compressed space. Single sample thickness. The maximum particle size of the aggregate to be tested Decision to ensure uniformity: Maximum particle size of the filling ( Coarse aggregate, such as the largest particle size Poor flowability can lead to particle jamming or segregation if the loading rate is too high; a lower loading rate of 20-50 g / s is recommended. Fine-grained aggregates have the greatest potential for this. Good fluidity allows for a suitable increase in the flow rate to 50-100 g / s. Secondly, uniform gradation is important. For aggregates with uneven gradation, the flow rate should be reduced to 30-60 g / s to prevent coarse particle concentration; for aggregates with uniform gradation, the flow rate can be increased to 60-100 g / s.

[0096] In some embodiments of this application, the basic physical constraints are determined by obtaining the size information of the adaptive lateral confinement component, and the preset loading parameters of the aggregate to be tested are determined accordingly. This ensures that the loading process is scientific and reasonable, and effectively improves the accuracy and repeatability of the test.

[0097] Furthermore, based on the above embodiments, in one of the exemplary embodiments provided in this application, the specific implementation process of the above-mentioned intelligent side-limited compression experimental method may further include steps S810 to S830, which are described in detail below.

[0098] Step S810: Determine the sample loading type of the aggregate to be tested based on the particle size distribution. The sample loading type includes continuous sample loading and layered sample loading.

[0099] Step S820: If the sample loading type is continuous sample loading, then determine the cross-sectional area corresponding to the adaptive lateral confinement component based on the size information.

[0100] Step S830: Determine the loose bulk density of the aggregate to be tested, and determine the loading rate of the aggregate to be tested based on the loose bulk density and cross-sectional area.

[0101] For example, before loading the aggregate to be tested, the loading type must be determined based on the known particle size distribution. Particle size distribution reflects the distribution of particles of different sizes within the aggregate; based on this characteristic, loading types can be divided into continuous loading and layered loading. Once continuous loading is determined, the cross-sectional area of ​​the steel drum of the adaptive side-confining component is accurately calculated using appropriate geometric calculation methods, combined with the dimensional information of the adaptive side-confining component. Simultaneously, the loose bulk density of the aggregate to be tested must be measured; this density reflects the mass per unit volume of the aggregate under natural packing conditions. Finally, by combining the obtained loose bulk density and the cross-sectional area of ​​the adaptive side-confining component, the loading rate of the aggregate during continuous loading is precisely determined according to a specific calculation formula, thereby ensuring that the loading process meets the experimental requirements.

[0102] Optionally, in some feasible embodiments, such as concrete solid waste aggregate... hour, In practice, continuous loading rather than stratified loading is used, with loading rates constrained by fundamental physical limitations. (Unit: g / s) is related to thickness control through a volumetric flow rate model, and its expression is as follows:

[0103]

[0104] in, The bulk density of aggregate (g / cm³) 3 (), can be determined by preliminary tests; The cross-sectional area of ​​the steel drum ( ); The rate of change of sample height (mm / s).

[0105] Then, according to the sample density formula (where m is the sample mass and V is the sample volume), the relationship between the sample loading rate (ṁ=m / t) and the target density can be derived as follows.

[0106]

[0107] in, Let A be the sample loading rate (mm / s), A be the cross-sectional area of ​​the steel drum, and the initial target density be ρ0 = 1.5 g / cm³. 3 Single sample loading thickness rate =0.5mm / s, then the sample loading rate However, due to the equipment's upper limit (100g / s), the speed needs to be reduced to 100g / s and the sample loading time extended.

[0108] Furthermore, based on the above embodiments, in one exemplary embodiment provided in this application, the above-mentioned intelligent side-limited compression experimental method further includes steps S910 to S930, which are described in detail below:

[0109] Step S910: During the process of loading the aggregate to be tested into the adaptive side confinement component according to the preset sample loading parameters, the surface flatness of the aggregate to be tested is determined based on real-time multi-dimensional sensor data.

[0110] Step S920: If the deviation between the surface flatness and the standard flatness is greater than the preset deviation threshold, the loading time of the aggregate to be tested is adjusted to obtain the adjusted loading time.

[0111] Step S930: Determine the target loading rate, target loading density, and target single-layer thickness based on the adjusted loading time, and correct the preset loading parameters based on the target loading rate, target loading density, and target single-layer loading thickness to obtain the target loading parameters of the aggregate to be tested.

[0112] For example, after obtaining the quantitative result of surface smoothness, the system will carefully compare it with the pre-set standard smoothness. The standard smoothness is determined comprehensively based on factors such as test requirements, aggregate characteristics, and the accuracy requirements of subsequent tests. If the deviation between the surface smoothness and the standard smoothness is found to be greater than the preset deviation threshold, it indicates that the surface smoothness after the current sample loading does not meet the expected standard, which may adversely affect the subsequent test results, such as affecting the stress distribution of the aggregate and leading to inaccurate test data. At this time, the system will automatically trigger an adjustment mechanism, using an intelligent optimization algorithm to adjust the loading time of the aggregate to be tested according to the magnitude and direction of the deviation. The intelligent optimization algorithm will comprehensively consider various factors, such as the flowability of the aggregate, the performance of the loading equipment, and historical loading data, and through continuous iteration and optimization, obtain an adjusted loading time that makes the surface smoothness closer to the standard smoothness.

[0113] After determining the adjusted loading time, the system further calculates and determines the target loading rate, target loading density, and target single-layer thickness based on this new time parameter, combined with the physical properties of the aggregate to be tested, the specifications of the loading equipment, and the specific requirements of the test, using relevant theories such as materials mechanics and process optimization. The target loading rate determines the amount of aggregate loaded per unit time, which affects the efficiency and uniformity of loading; the target loading density reflects the compactness of the aggregate after loading, which has an important impact on the mechanical properties of the aggregate and the test results; and the target single-layer thickness relates to the stacking of each layer of aggregate and the hierarchical structure of the overall loading. Thus, these three parameters are interrelated and mutually restrictive, jointly constituting the target loading parameters of the aggregate to be tested.

[0114] In some embodiments of this application, by accurately determining these target loading parameters, clear guidance can be provided for subsequent loading operations, ensuring that the surface flatness of the aggregate after loading meets the standard requirements, thereby improving the accuracy and reliability of the entire test.

[0115] Furthermore, based on the above embodiments, in one exemplary embodiment provided in this application, the specific implementation process of determining the test result corresponding to the aggregate to be tested based on the target multi-dimensional sensing data may further include steps S1010 to S1040, which are described in detail below:

[0116] Step S1010: Determine the vertical stress, vertical strain, and real-time porosity of the aggregate to be tested based on the multi-dimensional sensing data of the target.

[0117] Step S1020: Draw the first stress curve of the aggregate to be tested based on vertical stress and vertical strain.

[0118] In step S1030, the natural logarithm of the vertical stress is taken to obtain the ln stress, and the second stress curve corresponding to the aggregate under test is plotted based on the real-time void ratio and the ln stress.

[0119] Step S1040: Determine the test results corresponding to the aggregate to be tested based on the first stress curve and the second stress curve.

[0120] For example, based on these detailed and reliable multi-dimensional sensor data, specialized data processing algorithms and physical models can be used to accurately determine the key mechanical and physical parameters of the aggregate under test during the test process, namely vertical stress, vertical strain, and real-time void ratio. Vertical stress reflects the magnitude of the pressure exerted on the aggregate in the vertical direction, vertical strain reflects the degree of deformation of the aggregate under vertical stress, and real-time void ratio intuitively shows the ratio between the pore volume and the solid particle volume inside the aggregate. These parameters are crucial for a deeper understanding of the mechanical properties and deformation characteristics of the aggregate.

[0121] After obtaining the two important parameters, vertical stress and vertical strain, to more intuitively present their relationship, a first stress curve corresponding to the aggregate under test was plotted using professional plotting software, with vertical stress as the ordinate and vertical strain as the abscissa. This curve is like a "portrait" of the aggregate's mechanical properties. Through the curve's trend, slope, and other characteristics, the variation law of vertical stress in the aggregate under different vertical strains can be clearly observed, thus providing a preliminary judgment on the aggregate's elasticity, plasticity, and other mechanical properties.

[0122] Next, to further explore the intrinsic relationship between aggregate stress and pore structure, the natural logarithm of the vertical stress was taken to obtain the ln stress. Then, the second stress curve corresponding to the tested aggregate was plotted with the real-time void ratio as the ordinate and the ln stress as the abscissa. This curve reveals the changes in the pore structure of the aggregate during the stress process from another perspective, because the change in void ratio directly affects important properties of the aggregate such as strength and permeability. By analyzing the shape, inflection points, and other characteristics of the second stress curve, we can gain a deeper understanding of the evolution mechanism of the internal pores of the aggregate under stress, providing a strong basis for in-depth research on the mechanical behavior of aggregates.

[0123] Optionally, the maximum or minimum dry density criterion can be used, and its expression is as follows:

[0124]

[0125]

[0126] in, is the coefficient of variation of particle size distribution.

[0127] Furthermore, the curve feature extraction algorithm uses stress-strain curve yield point identification, and the second derivative mutation method can be used to determine the curve yield point, the expression of which is as follows:

[0128]

[0129] The slope of the strain hardening stage is piecewise fitted, and its expression is as follows:

[0130]

[0131] in, The initial elastic modulus, The hardening exponent can be optimized in real time by a multilayer feedforward neural network based on the error backpropagation algorithm.

[0132] Finally, based on the rich information contained in the first and second stress curves, and in accordance with the purpose and requirements of the experiment, professional analytical methods and theoretical knowledge were used to conduct a comprehensive and systematic evaluation and summary of the various performance indicators of the tested aggregate during the experiment. Furthermore, based on this, a detailed, accurate, and scientific test report for the tested aggregate was prepared. This report not only includes basic experimental information, the experimental process, and experimental data, but also provides an in-depth analysis and interpretation of the first and second stress curves. Ultimately, conclusions and suggestions were drawn regarding the mechanical properties, deformation characteristics, and pore structure of the tested aggregate, providing important reference for subsequent engineering applications and theoretical research.

[0133] Please see Figure 5 , Figure 5 This is a schematic diagram of strain curves and porosity versus ln stress curves shown in an exemplary embodiment of this application, which includes test results corresponding to three different groups of aggregates with different particle sizes.

[0134] An improved formula can be developed by introducing a stress history correction term into the dynamic evolution equation of porosity:

[0135]

[0136] in: The compression index (predicted in real time by a machine learning model). For reference stress, This is a step function (triggered during the loading phase). The change in void ratio caused by the i-th unloading rebound.

[0137] The relationship between porosity and ln stress can then be expressed using a piecewise linear regression model, as follows:

[0138]

[0139] Among them, critical stress It is determined by minimizing the least squares fitting error, and its expression is as follows:

[0140]

[0141] In some embodiments provided in this application, key parameters are determined based on multi-dimensional sensing data of the target and two types of stress curves are plotted to determine the test report. This can comprehensively and accurately present the mechanical properties of the aggregate under test, provide a reliable basis for in-depth analysis of its performance, and effectively improve the scientificity and practicality of the test results.

[0142] Furthermore, based on the above embodiments, in one exemplary embodiment provided in this application, the specific implementation process of the above-mentioned intelligent side-limited compression experimental method may further include steps S1110 to S1130, which are described in detail below:

[0143] Step S1110: Determine the real-time radial stress data within the adaptive lateral confinement component based on real-time multi-dimensional sensing data;

[0144] Step S1120: Determine the target thrust of the arc-shaped steel plate of the adaptive side-confining assembly based on real-time radial stress data;

[0145] Step S1130: Determine the control strategy of the servo push rod of the adaptive side limit component based on the target thrust.

[0146] For example, after successfully acquiring real-time radial stress data, the structural characteristics and mechanical transmission mechanism of the adaptive side-confining component are further analyzed in depth. The arc-shaped steel plate within the adaptive side-confining component, as a key structural component, has its stress condition closely related to the overall performance of the component. Then, based on the principles of mechanical equilibrium and relevant knowledge of materials mechanics, combined with real-time radial stress data, a series of precise mechanical calculations are used to determine the target thrust required by the arc-shaped steel plate under the current stress state. This target thrust is a key mechanical parameter that enables the arc-shaped steel plate to remain stably in the expected working position, effectively resist radial stress, and maintain the overall structural stability of the component. It requires comprehensive consideration of multiple factors, including the material properties, geometry, and real-time radial stress of the arc-shaped steel plate, to ensure that the calculated target thrust has high accuracy and reliability.

[0147] Finally, based on the determined target thrust of the curved steel plate 11, a control strategy for the servo actuator 12 of the adaptive side-limiting component 1 is formulated. As an actuator, the accuracy and timeliness of the servo actuator 12's movements directly affect whether the target thrust can be effectively applied to the curved steel plate 11. The control strategy fully considers the dynamic characteristics, response time, and collaborative requirements of the servo actuator with the entire system. Through advanced control algorithms, such as PID control algorithms or more complex intelligent control algorithms, the motion parameters of the servo actuator 12, such as speed and displacement, are adjusted in real time according to the deviation between the target thrust and the actual output thrust of the servo actuator 12. This enables the servo actuator to accurately and quickly output a thrust that matches the target thrust, thereby achieving precise control of the adaptive side-limiting component 1 and ensuring its stable and reliable operation under different working conditions, meeting the needs of experimental or engineering applications.

[0148] Optionally, in some feasible embodiments, the steel barrel sidewall of the radial stress adjustment device within the adaptive side-limiting component 1 is composed of four sections of arc-shaped steel plates 11. Each section of steel plate is connected to an independent servo push rod 12, which can adjust the thrust of each section of steel plate in real time based on the radial stress data fed back by the multi-parameter sensing system. The thrust of the steel plate and the radial stress data are directly related through a real-time closed-loop feedback adjustment mechanism. The core is to dynamically adjust the thrust of the steel plate, the expression of which is as follows:

[0149]

[0150] in, The thrust adjustment amount (unit: N) for a single-segment curved steel plate, i.e., the thrust that needs to be increased or decreased; This is a proportionality coefficient (unit: N / kPa), used to quantify the response intensity of thrust to radial stress deviation; The target radial stress (unit: kPa) is set according to the test requirements (it must be uniformly distributed); The radial stress (unit: kPa) was measured by a miniature pressure sensor on the inner wall of the steel drum. hour, For a positive value (increased thrust), the radial stress is increased by compressing the specimen with a steel plate; when hour, A negative value (reducing thrust) lowers the compression and reduces radial stress. Next is the dynamic calculation of real-time thrust. Specifically, the real-time thrust of a single steel plate segment is the sum of the initial thrust and the cumulative adjustment, expressed as follows:

[0151]

[0152] in, Initial thrust (preset base value based on aggregate type); multiple steel plates are independently adjustable, by comparing the corresponding values ​​of the four sections. Increase thrust in areas with low radial stress and decrease thrust in areas with high radial stress to ultimately achieve uniform radial stress across the entire barrel wall (requiring a very low radial stress variation). ).

[0153] In some embodiments of this application, real-time radial stress data is obtained by real-time multi-dimensional sensing data, thereby determining the target thrust of the arc-shaped steel plate and formulating a servo push rod control strategy. This enables the adaptive side-limiting component to achieve precise dynamic constraint on the aggregate, improving the accuracy and reliability of the test.

[0154] Furthermore, based on the above embodiments, in one exemplary embodiment provided in this application, the above-mentioned intelligent side-limited compression experimental method further includes steps S1210 and S1220, which are described in detail below:

[0155] Step S1210: If the real-time multi-dimensional sensing data indicates that the real-time vertical stress on the aggregate under test is greater than the preset vertical stress threshold, then the target damping coefficient corresponding to the magnetorheological damper is determined based on the real-time vertical stress.

[0156] Step S1220: Determine the adjustment strategy of the loading module based on the target damping coefficient, and control the loading module to execute the adjustment strategy.

[0157] For example, when real-time vertical stress detected based on real-time multi-dimensional sensor data exceeds a pre-set vertical stress threshold, this indicates that the vertical stress currently borne by the aggregate under test has exceeded the safe range or the preset test condition limits. If measures are not taken in time, it may lead to serious consequences such as aggregate damage, distorted test data, or even safety accidents. At this time, a response mechanism will be quickly activated. Based on the specific value of the real-time vertical stress, a pre-established complex mathematical model and algorithm will be used to determine the target damping coefficient corresponding to the magnetorheological damper. This mathematical model comprehensively considers the characteristic parameters of the magnetorheological damper, the material properties of the aggregate under test, and the nonlinear relationship between the real-time vertical stress and the damping coefficient, ensuring that the calculated target damping coefficient can accurately adapt to the current stress state and effectively adjust the mechanical performance of the system.

[0158] After determining the target damping coefficient of the magnetorheological damper, an adjustment strategy for the loading module is further formulated based on this target damping coefficient. As the core power output component of the entire experiment, the operating state of the loading module directly affects the progress and results of the experiment. The formulation of the adjustment strategy fully considers the dynamic characteristics of the loading module, the cooperative working relationship between its components, and its interaction mechanism with the magnetorheological damper. For example, parameters such as the loading speed, the magnitude of the loading force, or the loading frequency of the loading module may be adjusted according to the target damping coefficient to ensure that the loading module can operate stably and reliably under the new mechanical environment.

[0159] Finally, the formulated adjustment strategy is translated into specific control commands. Through advanced control algorithms and high-speed communication interfaces, the loading module is precisely controlled to execute the adjustment strategy. The control commands are transmitted to each execution component of the loading module in real time. Throughout the execution process, the operating status of the loading module and the stress changes of the aggregate under test are continuously monitored. Based on real-time feedback information, the adjustment strategy is dynamically optimized and adjusted to ensure that the real-time vertical stress on the aggregate under test is always kept within a safe and reasonable range, thereby ensuring the smooth progress of the test and the acquisition of accurate results.

[0160] In some embodiments of this application, the target damping coefficient of the magnetorheological damper is determined by judging when the vertical stress exceeds the threshold through real-time radial stress data, and an adjustment strategy for the loading module is formulated. This enables intelligent dynamic control of aggregate stress by the loading module, effectively avoiding damage to the aggregate due to excessive stress and ensuring the safety and accuracy of the test.

[0161] It should be noted that the intelligent side-confined compression experimental device and the intelligent side-confined compression experimental method provided in the above embodiments belong to the same concept. The specific operation methods of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the intelligent side-confined compression experimental device provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0162] The above description is merely a preferred exemplary embodiment of this application and is not intended to limit the implementation of this application. Those skilled in the art can easily make corresponding modifications or alterations based on the main concept and spirit of this application. Therefore, the scope of protection of this application should be determined by the scope of protection claimed in the claims.

Claims

1. A method for intelligent side-restricted compression experiment, characterized in that, The application is applied to an intelligent side limit compression experiment device, and the device comprises: An adaptive side limit component is used for containing the aggregate to be tested, and comprises an arc-shaped steel plate, a servo push rod and a friction adjusting layer. The friction adjusting layer is used for simulating different roughness of the side limit friction coefficient, and the servo push rod is used for adjusting the thrust of the arc-shaped steel plate, so that the aggregate to be tested is subjected to uniform radial stress. A loading module is arranged above the adaptive side limit component, and comprises a magneto-rheological damper and a loading plate. The loading plate is used for applying vertical load to the aggregate to be tested in the adaptive side limit component based on preset loading parameters. The magneto-rheological damper is arranged at the bottom of the loading plate. The magneto-rheological damper is used for starting when the real-time vertical stress value on the aggregate to be tested is greater than a preset vertical stress threshold value, and adjusting the damping coefficient through the magneto-rheological damper to adjust the impact energy of the loading plate. A sensing module is arranged in cooperation with the adaptive side limit component and is used for measuring corresponding multi-dimensional sensing data in the adaptive side limit component. A data processing module is used for data processing of the multi-dimensional sensing data collected by the sensing module, and determining the test result corresponding to the aggregate to be tested based on the data processing result. The intelligent side limit compression experiment method comprises: Obtaining the aggregate to be tested and determining the particle size distribution corresponding to the aggregate; Based on the particle size distribution, the side limit friction coefficient corresponding to the adaptive side limit component is determined, the working parameters of the friction adjusting layer are determined based on the side limit friction coefficient, and the aggregate to be tested is put into the adaptive side limit component according to the preset sample loading parameters; After the aggregate to be tested is put in, the loading module is controlled to apply vertical load to the aggregate to be tested based on the preset loading parameters, and real-time multi-dimensional sensing data in the adaptive side limit component during the loading process is collected; Based on the real-time multi-dimensional sensing data, the preset loading parameters are adjusted, and target multi-dimensional sensing data is determined based on the adjusted loading parameters, so as to determine the test result corresponding to the aggregate to be tested based on the target multi-dimensional sensing data.

2. The method of claim 1, wherein, The preset loading parameters are adjusted based on the real-time multi-dimensional sensing data, which comprises: Determine the real-time state parameters corresponding to the aggregate to be tested based on the real-time multi-dimensional sensing data; The real-time state parameters are input into the machine learning module as input parameters, so as to determine the correction coefficient of the loading module through the machine learning module; The preset loading parameters are adjusted based on the correction coefficient.

3. The method of claim 2, wherein, The real-time state parameters are input into the machine learning module as input parameters, so as to output the correction parameter through the machine learning module, which comprises: Determine the real-time physical property parameters of the aggregate to be tested based on the real-time state parameters; The real-time physical property parameters are input into the machine learning module as input parameters, so as to output the physical characteristic prediction value through the machine learning module; Determine a correction coefficient of the loading module based on the predicted value of the physical characteristic, so as to adjust the preset loading parameter based on the correction coefficient.

4. The method of claim 2, wherein, Before the method of pouring the to-be-tested aggregate into the self-adaptive side-limit assembly according to the preset sample loading parameter, the method further comprises: Obtaining size information corresponding to the self-adaptive side-limit assembly, and determining a basic physical constraint corresponding to the to-be-tested aggregate based on the size information; Determining a preset sample loading parameter corresponding to the to-be-tested aggregate based on the basic physical constraint, wherein the preset sample loading parameter comprises a sample loading density, a sample loading rate and a single-layer sample loading thickness.

5. The method of claim 4, wherein, The method further comprises: Determining a sample loading type of the to-be-tested aggregate based on the particle size distribution, wherein the sample loading type comprises continuous sample loading and layered sample loading; If the sample loading type is the continuous sample loading, determining a cross-sectional area corresponding to the self-adaptive side-limit assembly based on the size information; Determining a loose bulk density of the to-be-tested aggregate, and determining a sample loading rate of the to-be-tested aggregate based on the loose bulk density and the cross-sectional area.

6. The method of claim 4, wherein, The method further comprises: During the process of pouring the to-be-tested aggregate into the self-adaptive side-limit assembly according to the preset sample loading parameter, determining a surface flatness of the to-be-tested aggregate based on the real-time multi-dimensional sensing data; If a deviation value between the surface flatness and a standard flatness is greater than a preset deviation threshold, adjusting a sample loading time of the to-be-tested aggregate to obtain an adjusted sample loading time; Determining a target sample loading rate, a target sample loading density and a target single-layer thickness based on the adjusted sample loading time, so as to correct the preset sample loading parameter based on the target sample loading rate, the target sample loading density and the target single-layer sample loading thickness, and obtain a target sample loading parameter of the to-be-tested aggregate.

7. The method of claim 1, wherein, The method of determining a test result corresponding to the to-be-tested aggregate based on the target multi-dimensional sensing data comprises: Determining a vertical stress, a vertical strain and a real-time void ratio of the to-be-tested aggregate based on the target multi-dimensional sensing data; Drawing a first stress curve corresponding to the to-be-tested aggregate based on the vertical stress and the vertical strain; Taking a natural logarithm of the vertical stress to obtain an ln stress, and drawing a second stress curve corresponding to the to-be-tested aggregate based on the real-time void ratio and the ln stress; Determining a test result corresponding to the to-be-tested aggregate based on the first stress curve and the second stress curve.

8. The method of claim 1, wherein, The method further comprises: Determining real-time radial stress data in the self-adaptive side-limit assembly based on the real-time multi-dimensional sensing data; Determining a target thrust of an arc-shaped steel plate of the self-adaptive side-limit assembly based on the real-time radial stress data; Determining a control strategy of a servo push rod of the self-adaptive side-limit assembly based on the target thrust.

9. The method of claim 1, wherein, The method further comprises: If the real-time multi-dimensional sensing data represents that a real-time vertical stress on the to-be-tested aggregate is greater than a preset vertical stress threshold, determining a target damping coefficient corresponding to the magneto-rheological damper based on the real-time vertical stress; Determining an adjustment strategy of the loading module based on the target damping coefficient, and controlling the loading module to execute the adjustment strategy.

Images