Steel bridge deck pavement perception analysis system and method based on multi-dimensional data

Abstract

The invention relates to a steel bridge deck pavement sensing analysis system and method based on multi-dimensional data. The method comprises the following steps: testing raw materials to obtain construction retention time and interface parameters so as to guide mixture design, preparing a test piece according to the construction retention time and the interface parameters, and measuring a dynamic modulus main curve, a fatigue damage model and limit strain; establishing a reduced-scale test piece finite element model based on the main curve to predict bending-tensile strain, and performing a fatigue test in combination with limit strain to obtain reduced-scale life; obtaining a key stress range of the steel plate through full-scale model finite element analysis, and formulating an acceleration loading scheme according to the key stress range; loading is executed, multi-source monitoring data are collected, and the full-scale final fatigue life is determined; calibrating a fatigue damage model according to the service life; and finally evaluating the performance of the paving system based on the calibration model, the main curve and the full-scale life. According to the closed-loop verification process from the material to the full scale, the prediction model is iteratively calibrated through multi-scale test data, and the accuracy and reliability of long-term performance evaluation of the steel bridge deck pavement system are remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of engineering evaluation, and in particular relates to a perception and analysis system and method for steel bridge deck paving based on multi-dimensional data. Background Technology

[0002] As a key component of the traffic system of long-span bridges, the long-term durability of steel bridge deck pavement directly affects the safety, service life, and maintenance costs of the bridge structure. Due to the complex structure of orthotropic steel bridge decks, the pavement layer not only bears the direct action of vehicle loads but is also affected by multiple factors such as the deformation of the steel plate, changes in ambient temperature, and the bonding state of the interface. Its failure modes often manifest as a complex interplay of fatigue cracking, interlayer delamination, and rutting, which poses a significant challenge to the accurate assessment of its performance.

[0003] Currently, common evaluation methods in the industry often focus on a single scale or localized aspects. For example, at the material level, they emphasize testing the basic physical and mechanical properties of binders such as asphalt or epoxy resin; at the mixture level, they evaluate road performance through tests such as rutting and bending; or they use small-sized composite specimens for simple shear and pull-out tests. While these methods can reflect the characteristics of a specific level, they fail to construct a complete logical chain from material properties to the overall structural response. Due to the lack of a systematic multi-scale verification loop, there are often significant biases when predicting the long-term performance of actual bridge pavement systems based on material or scaled-down test results, leading to early damage to material and structural solutions that perform well in the laboratory in actual engineering projects. In addition, while traditional full-scale tests or observations of actual bridges can reflect the real situation, they have limitations such as high cost, long cycle, and uncontrollable variables, making them difficult to use for rapid comparison and optimization of solutions.

[0004] Therefore, there is an urgent need for a systematic evaluation method that can connect multiple scales, including materials, mixtures, composite structures, and full-scale models, and has clear parameter transfer and verification relationships between each scale, so as to achieve a more scientific, accurate, and efficient prediction and evaluation of the long-term service performance of steel bridge deck pavement systems, thereby providing a reliable basis for engineering design and maintenance decisions. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, system, equipment, and medium for sensing and analyzing steel bridge deck pavement based on multi-dimensional data to address the aforementioned technical problems.

[0006] Firstly, this application provides a method for perceptual analysis of steel bridge deck pavement based on multi-dimensional data, including:

[0007] S1. By conducting rheological viscosity tests on the key raw materials used in the pavement system of the target steel bridge deck, the construction allowance time range is obtained, and the interfacial bonding parameters between the key raw materials and different aggregates are obtained through contact angle measurement and pull-out test.

[0008] S2. Based on the construction allowance time range and interface bonding parameters, the mix proportion of the mixture is designed to prepare pavement mixture specimens; for the pavement mixture specimens, the dynamic modulus master curve is obtained through dynamic modulus testing, and the fatigue damage evolution model and fatigue limit strain based on cumulative dissipated energy are obtained through four-point bending fatigue test.

[0009] S3. Based on the master curve of dynamic modulus, a finite element model of a scaled-down composite structure specimen made of pavement mixture specimen and steel plate is established; the finite element model is used to predict the bending tensile strain response of the scaled-down composite structure specimen at the bottom of the pavement layer under bending load; based on the bending tensile strain response and fatigue limit strain at the bottom of the pavement layer, a three-point bending fatigue test is conducted on the scaled-down composite structure specimen to obtain the fatigue life of the scaled-down composite structure specimen.

[0010] S4. Based on the stress range of key parts of the steel plate under actual wheel load obtained from the finite element analysis of the full-scale segment model, determine the accelerated loading scheme for the accelerated loading test of the full-scale segment model; wherein, the full-scale segment model is made according to the actual bridge type.

[0011] S5. Perform accelerated loading cycles on the full-scale segment model according to the accelerated loading scheme, collect dynamic strain data and acoustic emission signals during the loading process, and obtain deflection data and radar detection data by periodically performing non-destructive testing on the full-scale segment model; determine the final fatigue life of the full-scale segment model under equivalent standard axle load based on the dynamic strain data, acoustic emission signals, deflection data and radar detection data.

[0012] S6. Compare the final fatigue life with the fatigue life of the scaled-down composite structure specimen and the predicted life extrapolated from the fatigue damage evolution model to obtain the comparison results; calibrate the fatigue damage evolution model based on the comparison results to obtain the calibrated fatigue damage evolution model.

[0013] S7. Based on the calibrated fatigue damage evolution model, dynamic modulus master curve, and final fatigue life, the performance of the pavement system is evaluated.

[0014] Secondly, this application also provides a steel bridge deck pavement perception and analysis system based on multi-dimensional data, used to implement the method described in the first aspect, the system comprising:

[0015] The raw material characteristic analysis module is used to obtain the construction allowance time range by conducting rheological viscosity tests on the key raw materials used in the pavement system of the target steel bridge deck, and to obtain the interfacial bonding parameters between the key raw materials and different aggregates through contact angle measurement and pull-out test.

[0016] The mixture performance modeling module is used to prepare pavement mixture specimens by designing the mixture mix proportion based on the construction allowance time range and interface bonding parameters; for the pavement mixture specimens, the dynamic modulus master curve is obtained through dynamic modulus testing, and the fatigue damage evolution model and fatigue limit strain based on cumulative dissipated energy are obtained through four-point bending fatigue test.

[0017] The fatigue prediction module for scaled-down structures is used to establish a finite element model of a scaled-down composite structure specimen made of pavement mixture and steel plate based on the master curve of dynamic modulus; predict the bending tensile strain response of the scaled-down composite structure specimen at the bottom of the pavement layer under bending load through the finite element model; and conduct a three-point bending fatigue test on the scaled-down composite structure specimen based on the bending tensile strain response at the bottom of the pavement layer and the fatigue limit strain to obtain the fatigue life of the scaled-down composite structure specimen.

[0018] The accelerated loading scheme planning module is used to determine the accelerated loading scheme for conducting accelerated loading tests on the full-scale segment model based on the stress range of key parts of the steel plate under actual wheel load obtained from finite element analysis of the full-scale segment model; wherein, the full-scale segment model is made according to the actual bridge type.

[0019] The structural response real-time monitoring module is used to perform accelerated loading cycles on the full-scale segment model according to the accelerated loading scheme, collect dynamic strain data and acoustic emission signals during the loading process, and obtain deflection data and radar detection data by periodically performing non-destructive testing on the full-scale segment model; based on the dynamic strain data, acoustic emission signals, deflection data and radar detection data, the final fatigue life of the full-scale segment model under equivalent standard axle load is determined.

[0020] The model dynamic calibration module is used to compare the final fatigue life with the fatigue life of the scaled-down composite structure specimen and the predicted life extrapolated from the fatigue damage evolution model to obtain the comparison results; based on the comparison results, the fatigue damage evolution model is calibrated to obtain the calibrated fatigue damage evolution model.

[0021] The comprehensive performance evaluation module is used to evaluate the performance of the pavement system based on the calibrated fatigue damage evolution model, the dynamic modulus master curve, and the final fatigue life.

[0022] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a steel bridge deck paving perception and analysis method based on multi-dimensional data as described in the first aspect.

[0023] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a steel bridge deck paving perception and analysis method based on multi-dimensional data as described in the first aspect.

[0024] The aforementioned multi-dimensional data-based perception and analysis system and method for steel bridge deck pavement obtains construction allowance time and interface parameters by testing raw materials to guide the design of the mixture. Based on this, specimens are prepared, and the dynamic modulus master curve, fatigue damage model, and ultimate strain are measured. A scaled-down specimen finite element model is established based on the master curve to predict flexural tensile strain, and fatigue tests are conducted using the ultimate strain to obtain the scaled-down life. The key stress range of the steel plate is obtained through full-scale model finite element analysis, and an accelerated loading scheme is formulated accordingly. Loading is executed, and multi-source monitoring data is collected to determine the final full-scale fatigue life. This life is used to calibrate the fatigue damage model. Finally, the performance of the pavement system is evaluated based on the calibrated model, master curve, and full-scale life. This closed-loop verification process from materials to full-scale, through iterative calibration of the prediction model using multi-scale experimental data, significantly improves the accuracy and reliability of the long-term performance evaluation of the steel bridge deck pavement system. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating a method for perceptual analysis of steel bridge deck pavement based on multi-dimensional data provided by the present invention;

[0027] Figure 2 This is a schematic diagram illustrating the process of determining the final fatigue life of a full-scale segment model under equivalent standard axle load in an optional embodiment of the present invention.

[0028] Figure 3 This invention provides a structural schematic diagram of a steel bridge deck paving perception and analysis system based on multi-dimensional data. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0030] refer to Figure 1 The document presents a flowchart illustrating a multi-dimensional data-based perception and analysis method for steel bridge deck pavement, as provided in this application. The method includes the following steps:

[0031] S1. By conducting rheological viscosity tests on the key raw materials used in the pavement system of the target steel bridge deck, the construction allowance time range is obtained, and the interfacial bonding parameters between the key raw materials and different aggregates are obtained through contact angle measurement and pull-out test.

[0032] Specifically, this step focuses on performance testing of key raw materials for the target steel bridge deck pavement system. The core objective is to clarify construction compatibility parameters and interfacial interaction characteristics, providing fundamental support for subsequent mix design. The technical principle is based on the viscosity evolution law in rheology and the wetting-bonding theory of interfacial chemistry. Quantitative testing establishes the correlation between the raw material's construction window and interfacial interaction parameters, ensuring that the subsequent mix exhibits good workability during construction and forms a stable interfacial bond with the aggregate.

[0033] Key raw materials mainly include binders and aggregates. The binder can be epoxy-based materials or modified asphalt, while the aggregates cover different lithological types. Rheological viscosity testing uses a rotational rheometer. Samples are prepared according to relevant standards, and multiple temperature gradients covering the construction temperature range are selected during testing. Viscosity changes over time are continuously monitored at a constant shear rate. The construction dwell time range is defined as the time interval between the initial viscosity of the raw material and the critical viscosity for workability. This range is determined by fitting a viscosity-time evolution curve. This range must meet the construction requirements of the entire process of mixing, paving, and compaction of the mixture, ensuring that the material does not lose workability due to excessively rapid viscosity increase, nor does it suffer from segregation due to excessively low viscosity.

[0034] Contact angle measurement was performed using a drop-method contact angle meter. First, the aggregate was processed into a smooth, flat sheet. After drying, the binder, heated to the application temperature, was dropped onto the surface of the aggregate sheet in the form of quantitative droplets. After the droplets were allowed to stabilize, images of their morphology were captured. The contact angle was calculated based on the Young's equation, the expression of which is: ,in Represents the surface free energy of solid aggregates. Represents the solid-liquid interfacial free energy between the binder and the aggregate. The liquid surface tension representing the binder, This represents the contact angle formed by the binder droplets on the aggregate surface. The size of the contact angle directly reflects the wetting effect of the binder on the aggregate; the smaller the contact angle, the better the wetting effect, which is more conducive to improving the subsequent interfacial bond strength.

[0035] Pull-out tests are used to supplement interfacial bond strength parameters. Aggregate-binder pull-out specimens are prepared according to standards, where aggregate particles are embedded in the binder matrix. After curing in a standard environment to the specified age, pull-out loading is performed using a universal testing machine. During loading, a constant tensile force is applied until the aggregate and binder separate, and the maximum pull-out force at this point is recorded. Interfacial bond strength is calculated as the ratio of pull-out force to the actual contact area. The actual contact area needs to be accurately measured using 3D scanning technology to ensure the accuracy of the bond strength calculation. Through contact angle and pull-out tests, an interfacial bond parameter database is ultimately established, clarifying the compatibility standards for different binder-aggregate combinations and providing a basis for subsequent aggregate selection in mixed materials.

[0036] S2. Based on the construction allowance time range and interface bonding parameters, the mix proportion of the mixture is designed to prepare pavement mixture specimens. For the pavement mixture specimens, the dynamic modulus master curve is obtained through dynamic modulus testing, and the fatigue damage evolution model and fatigue limit strain based on cumulative dissipated energy are obtained through four-point bending fatigue test.

[0037] Specifically, this step is based on the construction allowance time range and interface bonding parameters obtained in S1. The mix proportion of the pavement mixture is optimized through the balance design method, and system tests are carried out to obtain the dynamic mechanical properties and fatigue damage-related parameters of the mixture. The core technology is to construct a mixture system with good road performance, construction performance and fatigue durability, and to establish a quantitative correlation between performance and design parameters.

[0038] The mix design of the mixture follows relevant specifications. First, based on the interfacial bonding parameters, suitable aggregate types are selected to determine the aggregate gradation range. The proportion of aggregates with different particle sizes is adjusted using mineral aggregate gradation calculation methods to form a stable interlocking skeleton structure. Considering the construction allowance time requirements, Marshall specimens are prepared with multiple asphalt-aggregate ratio gradients. By measuring the porosity, stability, and flow value of the specimens, the optimal asphalt-aggregate ratio is comprehensively determined to ensure that the mixture has a dense structure while meeting the workability requirements during construction.

[0039] The preparation of pavement mixture specimens requires strict control of process parameters such as mixing temperature and time. After molding, they undergo standard curing to the specified age, and are then processed into specimens of corresponding sizes according to different test requirements. Dynamic modulus testing employs a material testing system, covering a wide temperature and frequency range. The temperature gradient must encompass both extreme and normal temperatures in actual service environments, and the frequency gradient must include the typical frequency range under traffic loads. Sine wave compressive loading is used, with stress levels controlled within a certain proportion of the mixture's compressive strength to avoid plastic deformation during loading that could affect the test results.

[0040] Based on the time-temperature equivalence principle, a standard temperature is selected as the reference temperature. The WLF equation is used to translate the dynamic modulus data at different temperatures and frequencies. The expression of the WLF equation is: ,in This is the time-temperature conversion factor, used to convert test data at different temperatures to equivalent data at a reference temperature, where T is the actual test temperature. The set reference temperature, and These are the fitting parameters related to material properties. A dynamic modulus master curve is constructed through data translation. The master curve can be fitted using a logarithmic function, the expression of which is: in, For dynamic modulus, The reference frequency is a, and the fitting coefficients are b. This master curve can intuitively reflect the stiffness characteristics of the mixture in a wide temperature range and a wide frequency range.

[0041] The four-point bending fatigue test was conducted using a fatigue testing machine. The specimen dimensions had to meet the requirements of the bending test. The support spacing and loading method were set according to the specifications. The loading frequency was selected from the typical frequency of traffic loads, and the temperature was controlled at room temperature. During the test, multiple strain level gradients were set, and multiple sets of parallel specimens were prepared for each strain level. A sinusoidal loading method was used for continuous loading until the specimen failed. The failure criterion was that the load decreased to a certain percentage of the maximum load.

[0042] Stress-strain hysteresis curves were acquired in real time during the experiment. The cumulative energy dissipation for each loading cycle was calculated based on the hysteresis curves. The expression for calculating the cumulative energy dissipation is as follows: ,in The energy dissipated in a single loading cycle, For stress, The strain is obtained by integrating the stress-strain curve for each cycle. The cumulative dissipated energy is the sum of the dissipated energy over all loading cycles, expressed as: Where W is the cumulative energy dissipation. Let be the energy dissipated during the i-th loading cycle. Based on fatigue damage theory, a fatigue damage evolution model is established, with the following expression: Where D represents the damage degree, used to characterize the degree of damage accumulation in the mixture during fatigue loading. This represents the cumulative dissipated energy at specimen failure. Model parameters were determined through nonlinear fitting to clarify the evolution of damage degree with the number of loading cycles. The fatigue limit strain was determined using the inflection point method. By plotting a double logarithmic curve of fatigue life versus strain, the strain corresponding to the point where the curve slope abruptly changes is the fatigue limit strain. This parameter is a key indicator for subsequent structural fatigue analysis.

[0043] S3. Based on the master curve of dynamic modulus, establish a finite element model of a scaled-down composite structure specimen made of pavement mixture specimen and steel plate; predict the bending tensile strain response of the scaled-down composite structure specimen at the bottom of the pavement layer under bending load using the finite element model; based on the bending tensile strain response at the bottom of the pavement layer and the fatigue limit strain, conduct a three-point bending fatigue test on the scaled-down composite structure specimen to obtain the fatigue life of the scaled-down composite structure specimen.

[0044] Specifically, this step explores the collaborative stress mechanism between the steel plate and the pavement layer by combining finite element simulation with experimental verification, and obtains the fatigue life of the scaled-down composite structure. The technical principle is to construct a scaled-down model based on the similarity principle, establish a finite element model using the dynamic mechanical parameters of the mixture, and predict key mechanical responses through simulation to provide a basis for the design of fatigue test parameters, ensuring that the test loading is consistent with the actual stress state of the structure.

[0045] The preparation of scaled-down composite structural specimens must follow the principle of similarity. The steel plates should be made of the same material as the actual steel bridge deck, and the specimen dimensions should be determined proportionally. The steel plate surface must be rust-removed and roughened, then coated with primer and adhesive. After curing for the specified time, the optimized pavement mixture should be laid. The pavement layer thickness should be set according to the actual project scale. The paving and compaction processes must simulate the actual construction process to ensure that the compaction degree of the specimen meets the design requirements and guarantees the consistency between the scaled-down model and the actual structure in terms of interface state and structural composition.

[0046] The finite element model was established using professional finite element analysis software. Shell elements suitable for thin-shell structures were selected for the steel plate, while solid elements supporting nonlinear material property definitions were used for the pavement layer. Element sizes were rationally divided according to the structural dimensions and stress characteristics to ensure a balance between computational accuracy and efficiency. Interface processing employed bound constraints to simulate interlayer bonding, assuming no relative slippage between layers. During material parameter input, the steel plate used elastic material properties, inputting the elastic modulus and Poisson's ratio. The pavement layer material properties were based on the dynamic modulus master curve obtained from S2. A user-defined material subroutine embedded the dynamic modulus's variation with temperature and frequency into the model, achieving accurate simulation of material properties.

[0047] Boundary conditions were set as simply supported constraints, limiting vertical displacement at both ends of the specimen while allowing horizontal displacement and rotation, simulating the support state of the actual structure. A three-point bending load was applied, with the load amplitude initially set based on the principle of equivalent stress in a real bridge. The load was gradually increased through finite element simulation, and the flexural strain response at the bottom of the pavement layer was extracted. A load-flexural strain curve was plotted to determine the load amplitude corresponding to when the flexural strain at the bottom of the pavement layer reaches the fatigue limit strain obtained in S2. This load amplitude is the loading amplitude for subsequent fatigue tests.

[0048] Three-point bending fatigue tests were conducted on a scaled-down composite structure based on finite element method (FEM) simulation results. A fatigue testing machine was used, and the specimen was loaded according to the simulated load amplitude, set loading frequency, and temperature conditions until failure. During the test, strain sensors were placed at key locations at the bottom of the pavement and key parts of the steel plate to monitor the strain response in real time, verifying the accuracy of the FEM simulation results. Simultaneously, an acoustic emission instrument was used to monitor crack initiation and propagation, and the damage development stage was determined by the count and energy of the acoustic emission signals.

[0049] Fatigue life is defined as the cumulative number of loading cycles until a specimen develops a significant crack or the load decreases to a certain percentage of the initial load. The final fatigue life of the scaled-down composite structure is obtained by averaging the test results of multiple sets of parallel specimens. During the test, fatigue failure modes must also be observed and recorded to identify the weak points and failure mechanisms of the composite structure, providing a reference for subsequent structural optimization.

[0050] S4. Based on the stress range of the key parts of the steel plate under actual wheel load obtained from the finite element analysis of the full-scale segment model, determine the accelerated loading scheme for the accelerated loading test of the full-scale segment model; wherein, the full-scale segment model is made according to the actual bridge type.

[0051] Specifically, the core of this step is to construct a full-scale segmental model consistent with the actual engineering project, clarify the accelerated loading parameters through finite element analysis, and achieve efficient simulation of the service state of the actual bridge. The technical principle is based on the principle of "axle load-stress synergistic equivalence", that is, by adjusting the test loading parameters, the stress range of key parts of the full-scale model steel plate is consistent with that of the actual bridge. At the same time, by increasing the loading frequency, the test cycle is shortened, solving the problems of long test cycles and high costs of traditional full-scale tests.

[0052] The full-scale segmental model is constructed based on the structural parameters of the actual bridge type. The transverse width is based on the standard lane width, and the longitudinal length on the single span. The material, thickness, U-rib dimensions, spacing, and diaphragm arrangement of the steel bridge deck are all consistent with the actual bridge, ensuring that the model's structural construction is identical to the real bridge. The pavement structure is the same as that used in the actual project, including pavement thickness and mixture type. The construction process strictly follows the actual project workflow, from steel bridge deck surface treatment, primer and adhesive application to pavement laying and compaction; each step is performed according to specifications to ensure that the model's construction quality is consistent with the real bridge, and the measured flatness, structural depth, and other indicators meet the design requirements.

[0053] The finite element analysis of the full-scale segmental model was conducted using professional finite element software to establish a full-scale numerical model. The element type selection and interface treatment method for the steel plate and pavement layer were consistent with those of S3, and the material parameters used the previously determined values. The actual wheel load parameters were determined based on the results of the actual bridge traffic survey, including the standard axle load size, wheel pressure, contact area, and average daily axle load frequency. The stress range of key parts of the steel plate was clarified through the stress monitoring data of the actual bridge.

[0054] In the finite element simulation, the actual wheel load is converted into an equivalent uniformly distributed load applied to the pavement surface. The load distribution area is consistent with the actual tire contact area. The stress response of key parts of the steel plate is extracted, and the relationship curve between axle load and stress is established. Its expression is: ,in denoted as the maximum stress at a critical point in the steel plate, P as the axial load, and k as the stress coefficient, which is obtained by fitting the finite element simulation results.

[0055] The accelerated loading scheme is determined based on stress equivalence. First, the stress range of key parts of the actual bridge steel plate is determined according to the relationship curve between axle load and stress, ensuring that the stress range of key parts of the steel plate is consistent with that of the actual bridge during the test. The loading frequency is set according to the capacity of the test equipment and the acceleration requirements. By increasing the loading frequency, the number of loadings per unit time is increased, thereby shortening the test cycle. The acceleration ratio is the ratio of the average number of loadings per day in the test to the average number of axle loads per day on the actual bridge.

[0056] The test temperature must cover the temperature range of the actual bridge's service environment, setting three gradients: low temperature, normal temperature, and high temperature, to simulate the impact of different temperature environments on structural performance. The cumulative number of loading cycles is determined based on the equivalent axle load cycles corresponding to the actual bridge's design reference period, ensuring that the test can simulate the performance state of the actual bridge after long-term service. Simultaneously, a detailed loading procedure is developed, with loading performed in segments according to different temperature gradients, and a fixed number of loading cycles set for each stage, simulating the service process of the actual bridge under different temperature environments.

[0057] S5. Perform accelerated loading cycles on the full-scale segment model according to the accelerated loading scheme, collect dynamic strain data and acoustic emission signals during the loading process, and obtain deflection data and radar detection data by periodically performing non-destructive testing on the full-scale segment model; determine the final fatigue life of the full-scale segment model under equivalent standard axle load based on the dynamic strain data, acoustic emission signals, deflection data and radar detection data.

[0058] Specifically, this step involves conducting accelerated loading tests, collecting multi-dimensional test data, comprehensively evaluating the fatigue performance of the full-scale segment model, and determining the final fatigue life. The technical principle is to use accelerated loading equipment to simulate repeated vehicle loading, and to capture changes in structural mechanical response and internal damage evolution through a combination of dynamic monitoring and non-destructive testing, thereby achieving accurate determination of fatigue life.

[0059] The accelerated loading test employs a professional accelerated loading system, with rubber pads simulating those of automobile tires used for the loading wheels to ensure that the contact pressure during loading is consistent with that of actual tires. The loading method is unidirectional continuous cyclic loading, strictly executed according to the loading plan established in S4, including load amplitude, loading frequency, temperature gradient, and loading procedure, to ensure the equivalence of the test process to the actual bridge service condition.

[0060] The data acquisition system consists of three parts: dynamic strain monitoring, acoustic emission monitoring, and non-destructive testing, forming a comprehensive monitoring system. Dynamic strain monitoring involves deploying strain sensors at key locations in the steel plate, the lower part of the paving layer, and interlayer interfaces. The sampling frequency of these sensors must meet the requirements for capturing dynamic strain responses, collecting strain data in real time during the loading process. The system focuses on analyzing parameters such as strain amplitude and strain rate of increase. Changes in these parameters are used to determine the attenuation of structural stiffness, which is a crucial indicator of accumulated structural damage.

[0061] Acoustic emission monitoring employs an acoustic emission instrument, with multiple sensors deployed at key locations on the model. Appropriate frequency ranges and thresholds are set to record the acoustic emission signals generated throughout the crack initiation and propagation process. By analyzing characteristic parameters such as the count, energy, and amplitude of the acoustic emission signals, damage evolution stages are defined. Sudden changes in signal characteristic parameters indicate that the internal damage of the structure has entered a rapid development stage.

[0062] Non-destructive testing is conducted periodically, with the testing frequency set according to the loading program, performed once after a certain number of loading cycles. Deflection testing uses a falling weight deflectometer, with sensors placed at different radii around the loading point to measure the deflection value of the pavement surface. The resilient modulus of the structure is calculated from the deflection value, and changes in the resilient modulus reflect changes in the structure's load-bearing capacity. Radar testing uses ground-penetrating radar, selecting an appropriate center frequency to perform a full-section scan of the pavement, generating radar images. By analyzing abnormal areas in the radar images, the location and size of defects such as voids, cracks, and delamination within the pavement are determined, achieving non-destructive identification of internal structural damage.

[0063] Visual inspection is carried out simultaneously. High-definition cameras and laser rangefinders are used at a set frequency to record the appearance defects of the pavement layer, including indicators such as crack length, width, rut depth, and pothole area, which directly reflect the damage status of the structural surface.

[0064] The final fatigue life is determined based on comprehensive multi-dimensional test data. Loading is stopped when any of the following conditions are met, and the cumulative number of loads at this point is the final fatigue life: 1) A through crack appears in the pavement layer, and the crack width reaches a set critical value; 2) The proportion of interlayer delamination area reaches a specified ratio, verified by radar detection and ultrasonic flaw detectors; 3) The rutting depth reaches a critical value; 4) The dynamic strain amplitude increases by a set ratio from the initial value, indicating severe structural stiffness degradation and inability to continue bearing loads. By comprehensively considering the results of multiple tests, the accuracy and reliability of the final fatigue life determination are ensured.

[0065] S6. Compare the final fatigue life with the fatigue life of the scaled-down composite structure specimen and the predicted life extrapolated from the fatigue damage evolution model to obtain the comparison results; calibrate the fatigue damage evolution model based on the comparison results to obtain the calibrated fatigue damage evolution model.

[0066] Specifically, this step optimizes the parameters of the fatigue damage evolution model established in S2 by comparing fatigue life data at different scales, ensuring that the model has cross-scale predictive capabilities. The technical principle is based on model validation and parameter optimization theory. By comparing and analyzing experimental data with model predictions, key model parameters are adjusted to reduce prediction bias, enabling the model to accurately reflect the fatigue damage development law of pavement systems at different scales.

[0067] First, we clarify the three sources of fatigue life data: the fatigue life of the scaled-down composite structure obtained from the S3 test, the predicted life extrapolated from the fatigue damage evolution model established in S2, and the final fatigue life of the full-scale segment obtained from the S5 test. The extrapolated predicted life is obtained by substituting the loading conditions of the scaled-down composite structure or the full-scale segment into the initial model, and then calculating the corresponding predicted fatigue life value through the model.

[0068] The comparative analysis uses relative error as the evaluation index. The formula for calculating relative error is: ,in The fatigue life was measured in the experiment. The fatigue life predicted by the model is calculated. The relative error between the test life of the scaled-down composite structure and the extrapolated predicted life, as well as the relative error between the test life of the full-scale segment and the extrapolated predicted life of the model, are calculated. These error values ​​are used to determine the prediction accuracy of the initial model.

[0069] The model calibration objective function is to minimize the combined relative error of the three lifetime data sets. The expression for the objective function is: The key parameters in the fatigue damage evolution model, mainly the cumulative energy dissipation coefficient and the damage evolution rate coefficient, were adjusted by optimizing the algorithm. Professional data analysis software was used for parameter optimization, and the parameter combination that minimizes the objective function was found through iterative calculations, resulting in the calibrated model parameters.

[0070] The validation of the calibrated model was carried out in two aspects. First, the calibrated parameters were substituted into the model to recalculate the predicted life of the scaled-down combined structure and the full-scale segment, and the new relative error was calculated to ensure that the error value was reduced to the allowable range in engineering. Second, the fatigue damage evolution curve of the calibrated model was plotted and compared with the acoustic emission damage data collected during the test. The degree of agreement between the model and the test data was judged by the curve fitting degree. The fitting degree should reach a high level, indicating that the calibrated model can accurately reflect the development process of fatigue damage.

[0071] By calibrating the model, the impact of differences in test conditions at different scales on the model's prediction accuracy is eliminated, enabling the model to predict fatigue life across scales and providing accurate model support for the subsequent performance evaluation of the pavement system.

[0072] S7. Based on the calibrated fatigue damage evolution model, dynamic modulus master curve, and final fatigue life, the performance of the pavement system is evaluated.

[0073] Specifically, this step integrates test data from three scales—material, scaled-down, and full-scale—with the calibrated model results to construct a multi-dimensional performance evaluation system. This system comprehensively evaluates the long-term service performance of the target steel bridge deck pavement system. The technical principle involves quantitative analysis from three core dimensions: mechanical properties, fatigue durability, and long-term stability. This comprehensive assessment determines whether the pavement system meets the actual engineering requirements, providing a scientific basis for engineering design optimization and maintenance decisions.

[0074] Mechanical performance evaluation centers on the master curve of the dynamic modulus, combining dynamic modulus data at different temperatures and frequencies to analyze the stiffness characteristics of the pavement system across a wide temperature and frequency range. Under normal temperature and typical traffic load frequencies, the load-bearing capacity of the pavement system is determined by the dynamic modulus value to ensure its ability to withstand the direct effects of vehicle loads. Under low-temperature conditions, the low-temperature crack resistance of the system is evaluated by combining the dynamic modulus and the low-temperature bending limit strain to prevent cracks caused by temperature stress in low-temperature environments. Under high-temperature conditions, the high-temperature rutting resistance of the system is evaluated by using the dynamic modulus and dynamic stability index to prevent permanent deformation in high-temperature environments.

[0075] Fatigue durability assessment is based on a calibrated fatigue damage evolution model and the final fatigue life of a full-scale segment. First, by comparing the final fatigue life of the full-scale segment with the equivalent axle load cycles corresponding to the design reference period of the actual bridge, it is determined whether the fatigue life of the pavement system meets the design requirements, ensuring that it will not experience fatigue failure within its design service life. Simultaneously, using the calibrated model, the remaining life of the pavement system under different traffic volume levels is predicted, providing a reference for engineering applications under different traffic load conditions and clarifying the applicable life of the pavement system in different scenarios such as extremely heavy traffic, heavy traffic, and moderate traffic.

[0076] Long-term stability assessment combines dynamic strain data during accelerated loading, non-destructive testing results, and the development of surface defects to analyze the performance degradation patterns of the structure under long-term loads and environmental effects. The overall load-bearing capacity stability of the structure is judged by the static load stiffness attenuation rate, which must be controlled within a reasonable range. The integrity and functional stability of the structure are evaluated using indicators such as the percentage of interlayer delamination area, crack width, and rut depth. Ensuring that these indicators meet relevant specifications demonstrates that the pavement system is not prone to early damage during long-term service and possesses good long-term stability.

[0077] Based on the evaluation results from the above three dimensions, the overall performance of the target steel bridge deck pavement system is assessed to determine whether it meets the usage requirements of the corresponding traffic level and is suitable for large-span steel bridge deck pavement projects. Simultaneously, based on the evaluation results, targeted maintenance recommendations are proposed, clarifying the maintenance priorities and timing at different service stages. When specific defects appear, corresponding maintenance measures should be taken promptly, such as crack sealing and partial milling and repaving, to ensure the safe operation of the bridge structure and extend its service life.

[0078] The aforementioned method for sensing and analyzing steel bridge deck pavement based on multi-dimensional data obtains construction allowance time and interface parameters by testing raw materials to guide the design of the mixture. Based on this, specimens are prepared, and the dynamic modulus master curve, fatigue damage model, and ultimate strain are measured. A finite element model of the scaled specimen is established based on the master curve to predict flexural tensile strain, and fatigue tests are conducted using the ultimate strain to obtain the scaled-down life. The key stress range of the steel plate is obtained through finite element analysis of the full-scale model, and an accelerated loading scheme is formulated accordingly. Loading is performed, and multi-source monitoring data is collected to determine the final full-scale fatigue life. This life is used to calibrate the fatigue damage model. Finally, the performance of the pavement system is evaluated based on the calibrated model, master curve, and full-scale life. This closed-loop verification process, from materials to full-scale, significantly improves the accuracy and reliability of long-term performance evaluation of steel bridge deck pavement systems through iterative calibration of the prediction model using multi-scale experimental data.

[0079] In one optional embodiment, obtaining the master curve of dynamic modulus through dynamic modulus testing includes the following steps:

[0080] S11. Under multiple preset temperatures and multiple preset loading frequencies, the uniaxial compression dynamic modulus of the pavement mixture specimens is tested to obtain complex modulus data.

[0081] Specifically, the core of this step is to obtain the complex modulus data of the pavement mixture specimens in a wide temperature and frequency range through uniaxial compression dynamic modulus testing. The technical principle is based on the dynamic mechanical properties of viscoelastic materials. The complex modulus can comprehensively reflect the elastic deformation and viscous deformation capacity of the material under dynamic load, providing basic data support for the subsequent construction of the dynamic modulus master curve.

[0082] The preparation of pavement mixture specimens must strictly adhere to the specifications for dynamic mechanical testing. The specimens should be cylindrical, and their dimensions must be designed to balance testing accuracy with uniform load transfer, ensuring uniform stress distribution within the specimen during loading and avoiding edge effects that could distort test results. After molding, the specimens must undergo standard curing to the specified age. During curing, the ambient temperature and humidity must be kept stable to prevent fluctuations in curing conditions from affecting the internal structure of the material. After curing, the specimen surface must be polished to ensure that the upper and lower end faces are parallel and flat, reducing stress concentration during loading.

[0083] The testing equipment utilizes a professional dynamic mechanical analysis system. This system must possess the ability to precisely control temperature, loading frequency, and load amplitude, and be capable of simultaneously acquiring sinusoidal dynamic loading and mechanical response signals. The selection of the preset temperature must be based on simulating the actual service environment of the target steel bridge deck pavement, covering a wide temperature range from low to high temperatures to ensure that the test data reflects the dynamic mechanical behavior of the material under different seasonal and regional temperature conditions. The determination of the preset loading frequency must consider the actual characteristics of traffic loads, covering the typical frequency range of load application during vehicle movement, including both low and high frequency bands, to capture the differences in the material's mechanical response under different load application rates.

[0084] During the test, the properly cured specimen is first installed between the upper and lower loading heads of the testing equipment. During installation, it is crucial to ensure that the specimen's axis coincides with the loading direction to avoid eccentric loading. Subsequently, the ambient temperature is gradually adjusted according to a preset temperature gradient. After reaching each preset temperature, sufficient isothermal time must be maintained to ensure that the internal temperature of the specimen matches the ambient temperature, eliminating the influence of the temperature gradient on the test results. The loading method employs sinusoidal dynamic compressive loading. During loading, the stress level must be strictly controlled, and the stress level must be set below the material's proportional limit to avoid plastic deformation during loading and ensure that the test data only reflects the material's viscoelastic response.

[0085] Under each preset temperature and loading frequency combination, after the loading stabilizes, stress and strain signals of the specimen are simultaneously acquired. Since the pavement mixture is a viscoelastic material, there is a phase difference between stress and strain under dynamic loading. By performing Fourier transform and phase analysis on the acquired stress-strain signals, the two key components of the complex modulus can be obtained: storage modulus and loss modulus. The expression for the complex modulus is: ,in Complex modulus is a comprehensive index characterizing the dynamic mechanical properties of materials; Storage modulus reflects the material's ability to store elastic deformation energy during loading, thus embodying the material's elastic properties; The loss modulus reflects the material's ability to lose energy due to viscous deformation during loading, embodying the material's viscous properties; i is the imaginary unit. Through the above testing procedure, tests are completed sequentially under all preset temperature and preset loading frequency combinations, ultimately forming a complete dataset containing temperature, frequency, and complex modulus (storage modulus and loss modulus), laying the foundation for the subsequent generation of the dynamic modulus master curve.

[0086] S12. Applying the time-temperature equivalence principle, the complex modulus data measured at different temperatures are shifted and superimposed to the selected reference temperature to generate the dynamic modulus master curve of the pavement mixture specimen at the reference temperature.

[0087] Specifically, this step is based on the time-temperature equivalence principle. The complex modulus data measured at different temperatures are shifted and superimposed to generate the dynamic modulus master curve at the reference temperature. The technical principle is that the mechanical response of viscoelastic materials has time-temperature equivalence characteristics. That is, the dynamic response of the material under high temperature and low frequency conditions has an inherent equivalent relationship with the dynamic response under low temperature and high frequency conditions. Through this characteristic, the test data with a wide temperature range and narrow frequency range can be transformed into a mechanical performance curve with a wide frequency range at room temperature, thereby comprehensively reflecting the long-term mechanical behavior of the material under actual service conditions.

[0088] The selection of a reference temperature must adhere to the principle of engineering practicality. Typically, a typical ambient temperature within the service environment of the target steel bridge deck pavement is chosen. This temperature should be in the middle of a preset temperature range to ensure that test data from different temperatures can be effectively shifted towards the reference temperature, avoiding data distortion due to an excessively extreme reference temperature. Once the reference temperature is determined, the standard test conditions at that temperature must be clearly defined as the benchmark for subsequent data overlay.

[0089] The core of the time-temperature equivalence principle is the calculation of the translation factor. The translation factor is used to convert the loading frequency at different temperatures into the equivalent frequency at the reference temperature, thereby achieving the superposition of complex modulus data at different temperatures on the same frequency axis. The translation factor is calculated using the WLF equation, a classic equation describing the time-temperature transformation relationship of viscoelastic materials, whose expression is: ,in This is the shift factor corresponding to temperature T; T is a preset test temperature. The selected reference temperature; and The material constants related to the characteristics of the pavement mixture need to be determined by fitting analysis of the test data. The fitting process aims to achieve the best superposition effect of the complex modulus data after translation at different temperatures.

[0090] The specific implementation of the data shifting operation involves, for each preset test temperature, performing an equivalent transformation on all preset loading frequencies at that temperature based on the calculated shift factor. The expression for calculating the equivalent frequency is as follows: ,in f is the equivalent frequency at the reference temperature; f is the actual applied frequency at the preset temperature. This is the translation factor corresponding to the preset temperature. Through this conversion, the "frequency-complex modulus" data pairs at different temperatures are uniformly converted into "equivalent frequency-complex modulus" data pairs at the reference temperature.

[0091] After completing the translation transformation of data at all temperatures, all "equivalent frequency-complex modulus" data points are plotted on a double logarithmic coordinate graph, where the horizontal axis represents the logarithm of the equivalent frequency and the vertical axis represents the logarithm of the complex modulus. Due to the effectiveness of the time-temperature equivalence principle, the translated data points at different temperatures will exhibit a continuous and smooth distribution trend. By fitting these data points with a suitable mathematical model, the master curve of the dynamic modulus at the reference temperature can be generated. The fitting model is usually a logarithmic polynomial or power function. During the fitting process, it is necessary to ensure that the model can accurately reflect the changing trend of the data points, and the goodness of fit must reach a high level to ensure the reliability of the master curve of the dynamic modulus.

[0092] The generation of the dynamic modulus master curve allows the test data that were originally scattered at different temperatures to be integrated into a continuous curve covering a wide frequency range. This curve can intuitively reflect the dynamic mechanical properties of the pavement mixture at different load frequencies at the reference temperature, providing accurate material mechanical parameter support for subsequent finite element modeling of scaled-down composite structures and fatigue damage analysis. It also provides a scientific basis for evaluating the evolution of the mechanical behavior of materials under long-term traffic loads.

[0093] In one optional embodiment, a fatigue damage evolution model based on accumulated dissipated energy and the fatigue limit strain are obtained through a four-point bending fatigue test, including the following steps:

[0094] S21. Under multiple preset strain levels, a four-point bending fatigue test with strain control was conducted on the pavement mixture specimen, and stress data, strain data and specimen stiffness modulus data were continuously recorded for each loading cycle.

[0095] Specifically, this step involves collecting dynamic mechanical response data of pavement mixture specimens at different strain levels through a four-point bending fatigue test in strain control mode. The technical principle is based on the core logic of "strain determines damage evolution" in fatigue damage mechanics. The strain control mode can accurately simulate the strain state of the actual steel bridge deck pavement layer under vehicle load, avoiding the problem of strain runaway caused by material stiffness decay in stress control mode. This ensures the consistency between the test process and the actual service conditions, providing reliable raw data for subsequent energy dissipation calculation and damage evolution analysis.

[0096] The preparation of pavement mixture specimens must adhere to the specifications for four-point bending fatigue testing. Rectangular beam structures are used, with their length, width, and height determined through mechanical calculations to ensure a pure bending segment in the mid-span region during loading, thus avoiding interference from shear forces on the test results. Specimens are formed using specialized molds. The uniformly mixed mixture is poured into the mold to the set thickness and then compacted using a vibratory compaction device. The compaction work and time must be controlled to ensure uniform compaction and meet design requirements. After molding, the specimens undergo standard curing to the specified age. During curing, stable ambient temperature and humidity must be maintained to prevent shrinkage cracks or humidity gradients within the specimen. After curing, the mid-span region and support contact area are ground to ensure a smooth surface, reducing stress concentration during loading. Simultaneously, the actual dimensions of the specimen are measured to provide accurate data for subsequent mechanical parameter calculations.

[0097] The testing equipment selected is a high-precision electro-hydraulic servo fatigue testing machine. This equipment must have strain closed-loop control function, capable of accurately outputting the preset strain level. The loading waveform is a sine wave, and the loading frequency is determined based on the actual characteristics of traffic load, typically selecting the typical frequency of load cycles during vehicle movement to ensure the equivalence of the test loading frequency with the actual load frequency. The test temperature is controlled at ambient temperature, and an environmental chamber is used to maintain a stable temperature in the test area to avoid the impact of temperature fluctuations on the mechanical properties of the materials.

[0098] The determination of the preset strain level needs to be based on the mechanical properties of the pavement mixture and the actual engineering strain range. Multiple strain levels with varying gradients should be selected, covering the range from micro-strain to ultimate strain, to ensure that the complete mechanical response of the material from the elastic stage, damage accumulation stage to failure stage can be captured. Multiple sets of parallel specimens need to be prepared at each strain level to eliminate the influence of individual specimen differences on the test results and ensure the statistical reliability of the data.

[0099] During the test, the cured specimen was first installed between the support and the loading head of the testing machine. The position of the specimen was adjusted to ensure accurate geometric relationship of the four-point loading, good contact between the loading point and the support, and no eccentric loading. Then, the strain control program was started, and the preset strain level, loading frequency, and upper limit of the number of cycles were set. After the test was started, the testing machine would continuously apply sinusoidal bending load according to the set parameters.

[0100] The data recording system works synchronously with the testing machine, acquiring stress, strain, and stiffness modulus data for each loading cycle in real time. Stress data is collected via force sensors mounted on the loading path, strain data is collected via strain gauges attached to the pure bending section at the mid-span of the specimen, and stiffness modulus data is automatically calculated by the testing machine system based on the real-time acquired stress and strain data. The data acquisition frequency must be at least 10 times higher than the loading frequency to ensure complete capture of the stress-strain response waveform for each loading cycle. The recorded data must include the peak stress, peak strain, stress-strain hysteresis curve, and corresponding stiffness modulus value for each cycle, forming a multi-dimensional dataset covering the entire loading process, laying the foundation for subsequent dissipation energy calculations and damage variable analysis.

[0101] S22. Calculate the dissipated energy for each loading cycle based on the recorded stress and strain data, and calculate the damage variable corresponding to each loading cycle based on the specimen stiffness modulus data.

[0102] Specifically, this step, based on the stress and strain data collected by S21, calculates the dissipated energy for each loading cycle. At the same time, it uses stiffness modulus data to derive damage variables. The technical principle is that as a viscoelastic material, the pavement mixture will dissipate some energy during fatigue loading through viscous deformation, internal crack initiation and propagation, etc. The accumulation of dissipated energy is directly related to the growth of damage variables. By quantifying the two, a quantitative relationship of fatigue damage evolution can be established.

[0103] The core of dissipated energy calculation lies in the integral operation of the area enclosed by the stress-strain hysteresis curve. Under sinusoidal cyclic loading, viscoelastic materials exhibit a closed hysteresis curve due to the phase difference between stress and strain. The physical meaning of the area enclosed by this curve represents the energy consumed by the material during a single loading cycle due to viscous loss and damage development. The expression for dissipated energy is: ,in The energy dissipated in the i-th loading cycle, Let the stress be the stress of the i-th cycle. Let be the strain of the i-th cycle. The integral symbol represents the closed integral of the stress-strain curve over a single loading cycle.

[0104] In the specific calculation process, the stress-strain discrete data points for each loading cycle are first extracted from the collected raw data and arranged according to the loading sequence to form a complete hysteresis curve. Numerical integration methods (such as the trapezoidal integration method) are then used to integrate the hysteresis curve. The trapezoidal integration method connects adjacent data points to form trapezoids, and the total area of ​​the hysteresis curve, i.e., the dissipated energy of a single cycle, is obtained by summing the areas of all trapezoids. During the calculation, it is necessary to ensure that the sampling density of the data points meets the integration accuracy requirements to avoid excessive errors in the dissipated energy calculation due to sparse data points. For each loading cycle, the dissipated energy calculation must be completed independently to form a dissipated energy sequence corresponding to each cycle, providing a basis for subsequent calculations of accumulated dissipated energy.

[0105] The calculation of the damage variable is based on the decay law of the stiffness modulus. The stiffness modulus is a direct characterization of a material's resistance to deformation. As the number of fatigue loading cycles increases, internal damage accumulates, and the stiffness modulus gradually decays. Therefore, the damage variable can be defined by the relative change in the stiffness modulus. The definition expression for the damage variable is: ,in For the damage variable corresponding to the i-th loading loop, Let be the instantaneous stiffness modulus of the i-th cycle. The initial stiffness modulus of the specimen.

[0106] initial stiffness modulus The value of is usually selected as the average value of the stiffness modulus during the first few loading cycles (e.g., the first 5 cycles) in the initial stage of the test. During this stage, material damage is relatively small, and the stiffness modulus is relatively stable, accurately reflecting the initial mechanical state of the material. Instantaneous stiffness modulus The stiffness modulus for each loading cycle is extracted directly from the test data and calculated by the testing machine system based on the peak stress and peak strain of that cycle. The calculation expression is as follows: ,in Let be the peak stress of the i-th cycle. Let be the peak strain of the i-th cycle.

[0107] Damage variables The value range is from 0 to 1, when When the time is right, it indicates that the material is undamaged and the stiffness modulus is in its initial state; when When the stress reaches zero, it indicates complete material failure, with the stiffness modulus decaying to 0. In actual calculations, the collected stiffness modulus data needs to be preprocessed to remove abnormal data caused by equipment noise or instantaneous fluctuations in the specimen, ensuring the accuracy of damage variable calculations. Using the above method, the damage variable sequence corresponding to each loading cycle can be obtained, establishing a correlation between damage variables and the number of loading cycles, providing data support for the subsequent construction of a damage evolution model.

[0108] S23. For each pavement mixture specimen, the cumulative dissipated energy is obtained by summing the dissipated energy from the first cycle to the current cycle. A power-law relationship model between the damage variable and the cumulative dissipated energy is established. The parameters of the fatigue damage evolution model were determined through fitting. and ;in, Represents damage variables, This represents the cumulative energy dissipation.

[0109] Specifically, this step obtains the cumulative dissipated energy by accumulating the dissipated energy of a single cycle, establishes a power-law relationship model between the damage variable and the cumulative dissipated energy, and determines the model parameters through data fitting. The technical principle is that the fatigue damage accumulation process of pavement mixture is essentially an energy dissipation process. The increase of the damage variable and the increase of the cumulative dissipated energy show a significant power-law correlation. The power-law model can accurately describe this nonlinear relationship and provide a mathematical tool for the quantitative prediction of fatigue damage.

[0110] The calculation of cumulative dissipated energy is based on the single-cycle dissipated energy sequence obtained in S22. The total dissipated energy from the first loading cycle to the current cycle is obtained through accumulation, and its expression is: ,in The cumulative energy dissipation corresponding to the i-th cycle is... Let be the dissipated energy of the k-th cycle, with the accumulation range from the 1st cycle to the 1st cycle. During the calculation, the energy is accumulated sequentially according to the loading cycle order to ensure the accuracy of the accumulated dissipated energy calculation for each cycle, forming a sequence of data showing the change of accumulated dissipated energy with the number of loading cycles. This data can intuitively reflect the total energy dissipation of the material during fatigue loading and is a direct manifestation of damage accumulation.

[0111] The power-law relationship model between damage variables and cumulative dissipated energy is a classic model summarized from a large amount of experimental data, and its expression is: , where D is the damage variable, used to characterize the degree of damage accumulation in the material; It represents the accumulated dissipated energy, which is an energy characterization parameter for damage accumulation; This is a proportionality coefficient, reflecting the degree of damage corresponding to a unit of cumulative energy dissipation. Its value is related to the energy dissipation efficiency of the material. This is the damage evolution index, reflecting the rate at which the damage variable increases with the cumulative dissipated energy. The larger the value, the more significant the promoting effect of the increase in cumulative dissipated energy on damage evolution.

[0112] The model parameters were fitted using the least squares method, which determines the optimal parameters by minimizing the sum of squared residuals between the measured and predicted values ​​of the damage variable. and This ensures the model fits the experimental data to the greatest extent possible. The specific steps of the fitting process are: first, all experimental data points... The data is processed to remove outlier data points caused by sudden specimen failure or equipment malfunction, ensuring data reliability. Then, a logarithmic transformation is performed on the power-law model to convert the nonlinear model into a linear model. The logarithmic transformation expression is as follows: ,make , , Then the linear model expression is By fitting the linear model using the least squares method, the coefficients a and are obtained. Then through The proportionality coefficient is obtained by reverse calculation. .

[0113] During the fitting process, a significance test is performed on the fitting results, and the coefficient of determination is calculated. Evaluate the goodness of fit of the model. The closer the value is to 1, the better the model fits the experimental data; generally, a value of 1 is required. To ensure model reliability, data from multiple parallel specimens should be fitted separately and the average of the parameters taken, or a global fitting method should be used to integrate all data to determine the optimal parameters, reducing the impact of individual specimen differences on model parameters. The resulting power-law model is a fatigue damage evolution model based on cumulative dissipated energy. This model can quantitatively predict the growth of damage variables through changes in cumulative dissipated energy, providing core mathematical model support for subsequent fatigue life assessment.

[0114] S24. Based on the fatigue life data of pavement mixture specimens at different strain levels, fit the stress-life curve equation; by extrapolating the stress-life curve equation to the preset number of cycles and combining it with the evolution characteristics of accumulated dissipated energy, determine the fatigue limit strain of the pavement mixture specimen.

[0115] Specifically, this step obtains the stress-life curve equation by fitting fatigue life data under different strain levels, and determines the fatigue limit strain by extrapolating the evolution characteristics of accumulated dissipated energy. The technical principle is that the fatigue limit strain is the maximum strain value of a material that does not fail within a specified number of cycles. By extrapolating the stress-life curve, the relationship between strain and fatigue life can be established. Combined with the evolution law of accumulated dissipated energy, the fatigue limit strain can be accurately defined, providing key indicators for structural fatigue design.

[0116] First, fatigue life data under different preset strain levels are compiled. Fatigue life is defined as the cumulative number of cycles a specimen undergoes under continuous loading at a corresponding strain level until failure. The failure criterion is the reduction of the specimen's stiffness modulus to a specified percentage (e.g., 50%), or the appearance of a through crack. For each strain level, the average fatigue life of multiple sets of parallel specimens is taken as the fatigue life corresponding to that strain level to ensure statistical representativeness of the data. Simultaneously, the peak stress data at specimen failure is recorded for each strain level, forming a three-dimensional dataset of "strain level - peak stress - fatigue life".

[0117] The stress-life curve equation was fitted based on the above dataset. A linear model in double logarithmic coordinates was selected as the fitting model. This model can well describe the relationship between the fatigue life and stress of the pavement mixture. Its expression is: ,in Fatigue life refers to the cumulative number of loading cycles at which the specimen fails. denoted as peak stress during loading; A is the intercept coefficient, which is related to the initial fatigue performance of the material; B is the slope coefficient, which reflects the degree of influence of stress change on fatigue life. The larger the value of B, the more significant the effect of stress increase on fatigue life attenuation.

[0118] The fitting process also employs the least squares method, converting the "peak stress-fatigue life" data points into linear data in a double logarithmic coordinate system. Coefficients A and B are determined by minimizing the sum of squared residuals. After fitting, the coefficients of determination are then used to determine the results. The goodness of fit is checked to ensure that the curve accurately reflects the intrinsic relationship between stress and fatigue life. The physical significance of the stress-life curve lies in its ability to predict the fatigue life corresponding to any peak stress, or the peak stress corresponding to any fatigue life, providing a basis for subsequent extrapolation analysis.

[0119] The selection of the preset number of iterations should refer to conventional standards in engineering practice, and is usually selected as follows: The number of cycles is used as a preset number of cycles corresponding to the fatigue limit. This value is a commonly used benchmark in engineering to determine the fatigue limit of materials, representing the number of cyclic loading cycles a material can withstand during long-term service. The extrapolation process calculates the corresponding preset number of cycles using the stress-life curve equation. peak stress That is, when hour, .

[0120] Verification was conducted by analyzing the evolution characteristics of accumulated dissipated energy. During fatigue loading, when the strain level is below the fatigue limit strain, the rate of increase of accumulated dissipated energy with the number of loading cycles gradually slows down and stabilizes, indicating that the material damage accumulation rate is extremely low and it can withstand long-term loads without failure. When the strain level is above the fatigue limit strain, the accumulated dissipated energy increases rapidly with the number of cycles until the specimen fails. Therefore, the peak stress corresponding to the preset number of cycles is determined. Then, by combining the evolution trend of the accumulated dissipated energy under the strain level corresponding to the stress, it is determined whether the strain is the fatigue limit strain.

[0121] Specifically, find the peak stress Corresponding strain level The change in cumulative dissipated energy with the number of cycles at this strain level was analyzed: if the cumulative dissipated energy increases slowly and the specimen shows no obvious damage when loaded to the preset number of cycles, then... This is the fatigue limit strain. If the accumulated dissipated energy continues to increase rapidly, the extrapolation strategy is adjusted, employing a nonlinear extrapolation method or incorporating experimental data from more strain levels for correction. The final determined fatigue limit strain is the maximum strain value at which the material does not experience fatigue failure within a specified number of cycles. This parameter provides a crucial basis for the load design of subsequent fatigue tests on scaled-down composite structures and the strain control of the actual bridge pavement, ensuring that the structure does not suffer premature failure due to fatigue damage during long-term service.

[0122] refer to Figure 2 In one optional embodiment, the final fatigue life of the full-scale segment model under equivalent standard axle load is determined based on dynamic strain data, acoustic emission signals, deflection data, and radar detection data, including the following steps:

[0123] S31. Based on dynamic strain data, strain amplitude evolution data is generated by calculating the moving average of strain amplitude at key points.

[0124] Specifically, the key points are selected from areas of structural stress concentration and weak points. First, the original dynamic strain data is denoised to remove interference signals. The strain amplitude for each loading cycle is calculated, i.e., the difference between the peak and trough strain in a single cycle. Then, an appropriate moving window size is selected, and the moving average of the strain amplitude is calculated according to the loading cycle sequence. Its expression is: ,in is the moving average strain amplitude of the k-th cycle; n is the number of cycles contained in the moving window; Let be the strain amplitude of the i-th cycle. The generated strain amplitude evolution data is presented as a time series curve, which intuitively reflects the long-term growth trend of strain at key points and is used to assess whether the strain continues to exceed the limit.

[0125] S32. By analyzing acoustic emission signals, the acoustic emission event rate and cumulative energy are extracted, and event characteristic data are generated based on the acoustic emission event rate and cumulative energy.

[0126] Specifically, high-sensitivity acoustic emission sensors are deployed in key structural areas, with the sensors closely fitted to the structural surface. Appropriate sampling frequencies, gains, and signal thresholds are set to collect acoustic emission signals during fatigue loading. The raw signals are then filtered to remove noise interference, and a peak detection algorithm is used to identify valid acoustic emission events.

[0127] The acoustic emission event rate is the number of effective events per unit time, reflecting the frequency of damage occurrence; the cumulative energy is the sum of the energy of all effective acoustic emission events, reflecting the total energy released due to damage accumulation. The event rate and cumulative energy are arranged by loading time or cycle number to form an event characteristic data sequence. By analyzing data abrupt changes and growth trends, it is assessed whether the interface damage has entered an active phase.

[0128] S33. Based on the deflection data, generate stiffness sequence data by back-calculating the overall equivalent stiffness of the full-scale segment model.

[0129] Specifically, a falling weight deflectometer was used to collect deflection data. Deflection sensors were placed at the center of the loading point and at different radii around it, and a test was conducted after a certain number of loading cycles. The collected deflection data was preprocessed to remove outliers and verify the rationality of the deflection basin data.

[0130] The core formula for calculating the overall equivalent stiffness based on the theory of elastic plates is as follows: ,in P is the overall equivalent stiffness; P is the loading amplitude; a is the loading contact radius; This represents the deflection value at the center of the loading point. The equivalent stiffness is correlated with the corresponding number of loading cycles according to the test sequence to form a stiffness sequence data. By analyzing the stiffness change trend, the overall performance of the structure is assessed to determine whether a sharp drop has occurred.

[0131] S34. By interpreting radar detection data, identify the morphology and extent of interlayer delamination and internal cracks, and generate damage image data.

[0132] Specifically, ground-penetrating radar is used to perform a full-coverage grid scan of the pavement layer, selecting an appropriate center frequency to ensure that the detection depth and resolution meet the requirements. The raw radar data is processed with range correction, filtering, and gain adjustment to highlight the reflection signal characteristics of the damaged area.

[0133] Based on the amplitude and phase changes of the reflected waves, interlayer delamination and internal cracks are identified. Delamination regions are characterized by continuous strong reflected waves in phase, while cracks are characterized by linear reflection bands formed by discrete strong reflection points. Damage contours are extracted using threshold segmentation and edge detection algorithms, and combined with scan coordinate information, a two-dimensional damage distribution map or a three-dimensional damage model is generated, marking the location, shape, and size of the damage. This is used to assess whether internal damage extends through critical sections.

[0134] S35. Based on strain amplitude evolution data, event characteristic data, stiffness sequence data, and damage image data, a failure assessment is performed on the full-scale segment model to obtain the assessment results. When the assessment results meet the preset failure criteria, the full-scale segment model is determined to have failed, and the cumulative number of equivalent standard axial loads is recorded as the final fatigue life. Among them, strain amplitude evolution data is used to assess whether the strain level at key points continues to exceed the limit, event characteristic data is used to assess whether the interface damage has entered the active period, stiffness sequence data is used to assess whether the overall performance of the structure has dropped sharply, and damage image data is used to assess whether the internal damage has penetrated the key section.

[0135] Specifically, a multi-dimensional failure criterion is preset: continuous exceeding of the moving average strain amplitude, abrupt increase in acoustic emission event rate and cumulative energy, a significant decrease in overall equivalent stiffness, and exceeding of the standard for internal damage area or crack penetration. During the assessment, the four types of data are first synchronized spatiotemporally according to the number of loading cycles, converted into unified quantitative indicators, and a weighted fusion algorithm is used to calculate the comprehensive evaluation value, the expression of which is: Where S is the comprehensive evaluation value; to Weights for the four types of data; to These are the corresponding quantitative indicators that have been normalized.

[0136] Structural failure is determined when the comprehensive evaluation value exceeds a preset threshold or any failure criterion is met. The cumulative number of loading cycles at failure is converted into the equivalent number of standard axle loads based on the fourth power law of axle load, which is the final fatigue life. The core conversion formula is as follows: ,in For the final fatigue life; The cumulative number of loading loops when failure occurs; This represents the test axle load amplitude. This is the standard axle load amplitude.

[0137] In one optional embodiment, the pavement system is evaluated for performance based on the calibrated fatigue damage evolution model, the master curve of dynamic modulus, and the final fatigue life, including the following steps:

[0138] S41. Input the design traffic load spectrum into the structural response analysis model that integrates the calibrated fatigue damage evolution model and the viscoelastic constitutive model of the pavement material defined by the master curve of the dynamic modulus, and calculate the cumulative damage development curve of the pavement system during service.

[0139] Specifically, the design traffic load spectrum is based on the actual traffic survey results of the target steel bridge deck, covering key parameters such as the distribution ratio of different axle load levels, the average daily axle load frequency, the load application frequency, and the temperature variation range of the service environment, comprehensively reflecting the actual load and environmental coupling effects faced by the pavement system. The structural response analysis model needs to integrate two core elements: first, a calibrated fatigue damage evolution model; and second, the viscoelastic constitutive model of the pavement material defined by the master dynamic modulus curve. The master dynamic modulus curve is constructed based on the time-temperature equivalence principle, and the specific formula is as follows: ,in The corresponding reference frequency at the reference temperature The dynamic modulus, where A is the fitting intercept coefficient and B is the fitting slope coefficient. This is the reference frequency.

[0140] The time-temperature conversion factor is calculated using the standard WLF equation, and the specific formula is as follows: ,in Temperature T is relative to a reference temperature. The time-temperature conversion factor, where T is the actual service temperature. For the selected reference temperature, and This is a constant related to the properties of the pavement material (determined by fitting experimental data). This formula is the core correlation of the viscoelastic constitutive model of the pavement material and can achieve equivalent conversion of the dynamic modulus at different temperatures.

[0141] Using the design traffic load spectrum as model input, the stress-strain hysteresis curves of the pavement system under each load cycle are obtained through numerical simulation. The dissipated energy of a single load cycle is calculated by integration, with the specific formula as follows: ,in Let be the energy dissipated during the i-th load cycle. The maximum strain in the i-th cycle, Let be the stress-strain function for the i-th cycle, with the integral range covering the complete hysteresis curve, corresponding to the energy change difference between the loading and unloading stages, respectively.

[0142] The cumulative energy dissipation in the i-th cycle is the sum of the energy dissipation in the previous i cycles, as shown in the following formula: ,in Let i be the cumulative energy dissipated in the i-th cycle. For the first The cumulative energy dissipation of each cycle, initial state .

[0143] Based on the calibrated fatigue damage evolution model, the specific formula for the damage increment in the i-th cycle is: ,in This represents the damage increment during the i-th load cycle. This is the proportionality coefficient. Both are damage evolution indices, and are constants obtained by fitting multi-scale experimental data.

[0144] The specific formula for the total damage variable after N cumulative axle loads is: ,in Let N be the total damage variable after N axle loads, where N is the cumulative number of axle loads. This represents the total cumulative energy dissipation after N axle loads.

[0145] By calculating the dissipated energy, cumulative dissipated energy, damage increment, and total damage for each load cycle using the above formula, a cumulative damage development curve is generated with the cumulative number of axle loads as the x-axis and the total damage variable as the y-axis. This curve visually presents the entire process of damage from its inception and slow accumulation to its rapid development, providing core data support for subsequent life prediction.

[0146] S42. Based on the final fatigue life, calibrate the performance degradation endpoint of the pavement system under accelerated loading test conditions.

[0147] Specifically, final fatigue life This refers to the equivalent standard axle load application count at which a full-scale segmental model reaches the failure criterion under accelerated loading test conditions. It directly reflects the ultimate bearing capacity of the pavement system under simulated real stress conditions. The calibration of the performance degradation endpoint is based on the damage state corresponding to the final fatigue life, determining the upper limit of damage to the pavement system from its initial intact state to its failure state. The calculation formula is as follows: ,in The upper limit of damage corresponding to the performance degradation endpoint. For final fatigue life The corresponding total cumulative energy dissipation, and The parameters of the calibrated fatigue damage evolution model are consistent with those in S41, namely the proportional coefficient and the damage evolution index.

[0148] Specifically, in the accelerated loading test, when the full-scale segment model is subjected to a cumulative number of standard axle loads, the test is conducted. At that time, the corresponding total cumulative energy dissipation is calculated using the structural response analysis model. Substituting these values ​​into the above formula yields the upper limit of damage. At this point, the internal damage to the pavement system has reached a critical level, manifested as a significant decrease in structural stiffness, through-cracks, or excessive interlayer delamination area. This is the performance degradation endpoint. The determination of this endpoint provides a clear reference for extrapolating the test results to actual service scenarios, ensuring the consistency between the predicted results and the actual service performance.

[0149] S43. Based on the mapping relationship between the cumulative damage development curve and the performance degradation endpoint, predict the cumulative number of standard axle loads that the pavement system will withstand when it reaches the predetermined service performance threshold under the design traffic load spectrum, and use the cumulative number of standard axle loads as the predicted service life assessment result of the pavement system.

[0150] Specifically, predetermined service performance thresholds Determined in conjunction with engineering design requirements and relevant specifications, this threshold is typically set as the critical value at which structural performance degrades to meet normal functional requirements. This threshold must be lower than the upper damage limit corresponding to the performance degradation endpoint. Reserve a certain amount of safety margin.

[0151] Cumulative damage development curve established based on S41 With performance degradation endpoint The mapping relationship is used to find the cumulative damage threshold that has been reached during service life by back-calculating the model. The corresponding cumulative number of standard axle loads First, the cumulative energy dissipation is calculated by using a predetermined service performance threshold. The core formula is: ,in Cumulative axle load count The corresponding total cumulative energy dissipation, To the predetermined service performance threshold, and The model parameters are consistent with those of S41.

[0152] because There is a one-to-one functional relationship between the cumulative axle load count N and the load (derived from the design traffic load spectrum and structural response analysis model, i.e.) ), which can be obtained through inverse function operations: ,in This refers to the cumulative number of standard axle loads applied when the pavement system reaches the predetermined service performance threshold. This represents the inverse function relationship between cumulative dissipated energy and cumulative axle load cycles.

[0153] Will Combined with the daily average axle load frequency in the design traffic load spectrum This can be further converted into a predicted service life. The formula is: ,in To predict service life, 365 is used as the average number of days per year. Ultimately, or This refers to the predicted service life assessment results of the paving system, which can provide direct quantitative basis for engineering design optimization and maintenance plan formulation.

[0154] The aforementioned method for sensing and analyzing steel bridge deck pavement based on multi-dimensional data obtains construction allowance time and interface parameters by testing raw materials to guide the design of the mixture. Based on this, specimens are prepared, and the dynamic modulus master curve, fatigue damage model, and ultimate strain are measured. A finite element model of the scaled specimen is established based on the master curve to predict flexural tensile strain, and fatigue tests are conducted using the ultimate strain to obtain the scaled-down life. The key stress range of the steel plate is obtained through finite element analysis of the full-scale model, and an accelerated loading scheme is formulated accordingly. Loading is performed, and multi-source monitoring data is collected to determine the final full-scale fatigue life. This life is used to calibrate the fatigue damage model. Finally, the performance of the pavement system is evaluated based on the calibrated model, master curve, and full-scale life. This closed-loop verification process, from materials to full-scale, significantly improves the accuracy and reliability of long-term performance evaluation of steel bridge deck pavement systems through iterative calibration of the prediction model using multi-scale experimental data.

[0155] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0156] Based on the same inventive concept, this application also provides a system for implementing the above-mentioned method for perceiving and analyzing steel bridge deck pavement based on multi-dimensional data. The solution provided by this system is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more embodiments of the steel bridge deck pavement perception and analysis system based on multi-dimensional data provided below can be found in the limitations of the steel bridge deck pavement perception and analysis method based on multi-dimensional data described above, and will not be repeated here.

[0157] In one exemplary embodiment, such as Figure 3As shown, a steel bridge deck pavement perception and analysis system 30 based on multi-dimensional data is provided to implement the methods in the above embodiments. The system includes:

[0158] The raw material characteristic analysis module 31 is used to obtain the construction allowance time range by conducting rheological viscosity tests on the key raw materials used in the pavement system of the target steel bridge deck, and to obtain the interfacial bonding parameters between the key raw materials and different aggregates by contact angle measurement and pull-out test.

[0159] The mixture performance modeling module 32 is used to prepare pavement mixture specimens by designing the mixture mix ratio based on the construction allowance time range and interface bonding parameters; for the pavement mixture specimens, the dynamic modulus master curve is obtained through dynamic modulus testing, and the fatigue damage evolution model and fatigue limit strain based on cumulative dissipated energy are obtained through four-point bending fatigue testing.

[0160] The fatigue prediction module 33 for scaled structures is used to establish a finite element model of a scaled composite structure specimen made of pavement mixture specimen and steel plate based on the master curve of dynamic modulus; predict the bending tensile strain response of the scaled composite structure specimen at the bottom of the pavement layer under bending load through the finite element model; and conduct a three-point bending fatigue test on the scaled composite structure specimen based on the bending tensile strain response at the bottom of the pavement layer and the fatigue limit strain to obtain the fatigue life of the scaled composite structure specimen.

[0161] The accelerated loading scheme planning module 34 is used to determine the accelerated loading scheme for conducting accelerated loading tests on the full-scale segment model based on the stress range of the key parts of the steel plate under actual wheel load obtained from the finite element analysis of the full-scale segment model; wherein, the full-scale segment model is made according to the actual bridge type.

[0162] The structural response real-time monitoring module 35 is used to perform accelerated loading cycles on the full-scale segment model according to the accelerated loading scheme, collect dynamic strain data and acoustic emission signals during the loading process, and obtain deflection data and radar detection data by periodically performing non-destructive testing on the full-scale segment model; based on the dynamic strain data, acoustic emission signals, deflection data and radar detection data, the final fatigue life of the full-scale segment model under equivalent standard axle load is determined.

[0163] The model dynamic calibration module 36 is used to compare the final fatigue life with the fatigue life of the scaled-down composite structure specimen and the predicted life extrapolated from the fatigue damage evolution model to obtain the comparison results; based on the comparison results, the fatigue damage evolution model is calibrated to obtain the calibrated fatigue damage evolution model.

[0164] The comprehensive performance evaluation module 37 is used to evaluate the performance of the pavement system based on the calibrated fatigue damage evolution model, the dynamic modulus master curve, and the final fatigue life.

[0165] Embodiments of this application also provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the aforementioned method embodiments.

[0166] Embodiments of this application also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0167] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0168] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for perceptual analysis of steel bridge deck pavement based on multi-dimensional data, characterized in that, The method includes: S1. By conducting rheological viscosity tests on the key raw materials used in the pavement system of the target steel bridge deck, the construction allowance time range is obtained, and the interfacial bonding parameters between the key raw materials and different aggregates are obtained through contact angle measurement and pull-out test. S2. Based on the construction allowance time range and the interface bonding parameters, the mix proportion of the mixture is designed to prepare pavement mixture specimens; for the pavement mixture specimens, the dynamic modulus master curve is obtained by dynamic modulus testing, and the fatigue damage evolution model and fatigue limit strain based on cumulative dissipated energy are obtained by four-point bending fatigue test. S3. Based on the dynamic modulus master curve, establish a finite element model of a scaled-down composite structure specimen made of the pavement mixture specimen and steel plate; predict the bending tensile strain response of the scaled-down composite structure specimen at the bottom of the pavement layer under bending load using the finite element model; based on the bending tensile strain response at the bottom of the pavement layer and the fatigue limit strain, conduct a three-point bending fatigue test on the scaled-down composite structure specimen to obtain the fatigue life of the scaled-down composite structure specimen. S4. Based on the stress range of key parts of the steel plate under actual wheel load obtained from the finite element analysis of the full-scale segment model, determine the accelerated loading scheme for the accelerated loading test of the full-scale segment model; wherein, the full-scale segment model is made according to the actual bridge type; S5. Perform accelerated loading cycles on the full-scale segment model according to the accelerated loading scheme, collect dynamic strain data and acoustic emission signals during the loading process, and obtain deflection data and radar detection data by periodically performing non-destructive testing on the full-scale segment model; determine the final fatigue life of the full-scale segment model under equivalent standard axle load based on the dynamic strain data, the acoustic emission signals, the deflection data and the radar detection data. S6. Compare the final fatigue life with the fatigue life of the scaled-down composite structure specimen and the predicted life extrapolated from the fatigue damage evolution model to obtain the comparison result; calibrate the fatigue damage evolution model based on the comparison result to obtain the calibrated fatigue damage evolution model. S7. Based on the calibrated fatigue damage evolution model, the dynamic modulus master curve, and the final fatigue life, the performance of the pavement system is evaluated.

2. The method according to claim 1, characterized in that, The process of obtaining the master curve of dynamic modulus through dynamic modulus testing includes: S11. Under multiple preset temperatures and multiple preset loading frequencies, the paving mixture specimen is subjected to uniaxial compression dynamic modulus test to obtain complex modulus data; S12. Applying the time-temperature equivalence principle, the complex modulus data measured at different temperatures are shifted and superimposed to the selected reference temperature to generate the dynamic modulus master curve of the paving mixture specimen at the reference temperature.

3. The method according to claim 1, characterized in that, The fatigue damage evolution model and fatigue limit strain obtained through four-point bending fatigue testing based on accumulated dissipated energy include: S21. Under multiple preset strain levels, a four-point bending fatigue test with strain control is performed on the pavement mixture specimen, and stress data, strain data and specimen stiffness modulus data are continuously recorded for each loading cycle. S22. Calculate the dissipated energy of each loading cycle based on the recorded stress data and strain data, and calculate the damage variable corresponding to each loading cycle based on the specimen stiffness modulus data; S23. For each pavement mixture specimen, the dissipated energy from the first cycle to the current cycle is summed to obtain the cumulative dissipated energy, and a power-law relationship model between the damage variable and the cumulative dissipated energy is established. The parameters of the fatigue damage evolution model were determined by fitting. and ;in, Represents the damage variable, This represents the cumulative energy dissipation; S24. Based on the fatigue life data of the pavement mixture specimens at different strain levels when they fail, fit the stress-life curve equation; by extrapolating the stress-life curve equation to a preset number of cycles and combining it with the evolution characteristics of the accumulated dissipated energy, determine the fatigue limit strain of the pavement mixture specimens.

4. The method according to claim 1, characterized in that, The step of determining the final fatigue life of the full-scale segment model under equivalent standard axle load based on the dynamic strain data, the acoustic emission signal, the deflection data, and the radar detection data includes: S31. Based on the dynamic strain data, strain amplitude evolution data is generated by calculating the moving average value of strain amplitude at key points; S32. By analyzing the acoustic emission signal, the acoustic emission event rate and cumulative energy are extracted, and event feature data are generated based on the acoustic emission event rate and the cumulative energy. S33. Based on the deflection data, generate stiffness sequence data by back-calculating the overall equivalent stiffness of the full-scale segment model; S34. By interpreting the radar detection data, identify the morphology and extent of interlayer delamination and internal cracks, and generate damage image data; S35. Based on the strain amplitude evolution data, the event characteristic data, the stiffness sequence data, and the damage image data, a failure assessment is performed on the full-scale segment model to obtain an assessment result. When the assessment result meets the preset failure criterion, the full-scale segment model is determined to have failed, and the cumulative number of equivalent standard axial loads is recorded as the final fatigue life. The strain amplitude evolution data is used to assess whether the strain level at key points continues to exceed the limit; the event characteristic data is used to assess whether interface damage has entered an active period; the stiffness sequence data is used to assess whether the overall structural performance has experienced a sharp decline; and the damage image data is used to assess whether internal damage has penetrated the key section.

5. The method according to any one of claims 1 to 4, characterized in that, The performance evaluation of the pavement system based on the calibrated fatigue damage evolution model, the dynamic modulus master curve, and the final fatigue life includes: S41. Input the design traffic load spectrum into the structural response analysis model that integrates the calibrated fatigue damage evolution model and the viscoelastic constitutive model of the pavement material defined by the master curve of the dynamic modulus, and calculate the cumulative damage development curve of the pavement system during service. S42. Based on the final fatigue life, determine the performance degradation endpoint of the pavement system under the conditions of the accelerated loading test; S43. Based on the mapping relationship between the cumulative damage development curve and the performance degradation endpoint, predict the cumulative number of standard axle loads that the pavement system will withstand when it reaches the predetermined service performance threshold under the design traffic load spectrum, and use the cumulative number of standard axle loads as the predicted service life assessment result of the pavement system.

6. A steel bridge deck pavement perception and analysis system based on multi-dimensional data, used to implement the method according to any one of claims 1 to 5, characterized in that, The system includes: The raw material characteristic analysis module is used to obtain the construction allowance time range by conducting rheological viscosity tests on the key raw materials used in the pavement system of the target steel bridge deck, and to obtain the interfacial bonding parameters between the key raw materials and different aggregates by contact angle measurement and pull-out test. The mixture performance modeling module is used to prepare pavement mixture specimens by designing the mixture mix ratio based on the construction allowance time range and the interface bonding parameters; for the pavement mixture specimens, the dynamic modulus master curve is obtained by dynamic modulus testing, and the fatigue damage evolution model and fatigue limit strain based on cumulative dissipated energy are obtained by four-point bending fatigue testing. The fatigue prediction module for scaled-down structures is used to establish a finite element model of a scaled-down composite structure specimen made of the pavement mixture specimen and steel plate based on the master curve of the dynamic modulus; predict the bending tensile strain response of the scaled-down composite structure specimen at the bottom of the pavement layer under bending load using the finite element model; and conduct a three-point bending fatigue test on the scaled-down composite structure specimen based on the bending tensile strain response at the bottom of the pavement layer and the fatigue limit strain to obtain the fatigue life of the scaled-down composite structure specimen. The accelerated loading scheme planning module is used to determine the accelerated loading scheme for conducting accelerated loading tests on the full-scale segment model based on the stress range of key parts of the steel plate under actual wheel load obtained from finite element analysis of the full-scale segment model; wherein the full-scale segment model is fabricated according to the actual bridge type. The structural response real-time monitoring module is used to perform accelerated loading cycles on the full-scale segment model according to the accelerated loading scheme, collect dynamic strain data and acoustic emission signals during the loading process, and obtain deflection data and radar detection data by periodically performing non-destructive testing on the full-scale segment model; and determine the final fatigue life of the full-scale segment model under equivalent standard axle load based on the dynamic strain data, the acoustic emission signals, the deflection data and the radar detection data. The model dynamic calibration module is used to compare the final fatigue life with the fatigue life of the scaled-down composite structure specimen and the predicted life extrapolated from the fatigue damage evolution model to obtain the comparison result; and to calibrate the fatigue damage evolution model based on the comparison result to obtain the calibrated fatigue damage evolution model. The comprehensive performance evaluation module is used to evaluate the performance of the pavement system based on the calibrated fatigue damage evolution model, the dynamic modulus master curve, and the final fatigue life.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

Images