Concrete life prediction method and system based on ardealite concrete erosion

Abstract

The invention discloses a concrete life prediction method and system based on ardealite concrete erosion, and belongs to the technical field of concrete structure life prediction, historical erosion data of an environment where a target structure is located are obtained, and the historical erosion data comprise a data sequence of acid concentration, water flow velocity, wet-dry cycle frequency and chloride ion concentration changing along with time; collecting material parameters of the ardealite concrete and performance attenuation coefficients of the ardealite concrete under the action of different erosion factors; constructing a multi-factor coupled life prediction model based on environmental data and material parameters, and dynamically iterating material performance by taking time as a main variable; inputting a service time interval, solving the model and outputting an evolution curve of compressive strength and porosity; setting a structure failure threshold value, judging whether a failure condition is reached or not, and outputting structure predicted failure time; according to the method, the performance evolution law of the phosphogypsum concrete in a complex environment can be effectively simulated, and high-precision and quantifiable life prediction is realized.

Description

Technical Field

[0001] This invention relates to the field of concrete structure life prediction technology, specifically to a method and system for predicting concrete life under phosphogypsum concrete erosion. Background Technology

[0002] With the development of resource utilization of construction waste, phosphogypsum, as an industrial byproduct, is widely used in the preparation of concrete. However, phosphogypsum concrete is susceptible to complex corrosive factors such as acidity, migration of hydration products, and salt corrosion during long-term service, leading to a decline in its mechanical properties, an increase in porosity, and structural deterioration, ultimately affecting the durability and service life of the structure.

[0003] Currently, there is a lack of methods for predicting the service life of phosphogypsum concrete under actual corrosive environments. Existing research mostly focuses on experimental testing, making it difficult to accurately quantify the dynamic relationship between different corrosion factors and the degradation of concrete performance, and also failing to achieve macroscopic service life assessment. Therefore, there is an urgent need for a method that can combine corrosion environment data, basic material parameters, and changes in service time to establish a predictive model to guide the design and maintenance strategies of phosphogypsum concrete in practical engineering applications. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for predicting the life of concrete under phosphogypsum concrete erosion, so as to overcome the shortcomings of the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for predicting the life of concrete under phosphogypsum concrete erosion, comprising: S100. Obtain historical erosion data of the environment in which the target structure is located, including changes in acid concentration, water flow velocity, wet-dry cycle frequency, and the time sequence T of chloride ion concentration. S200. Obtain the material parameters of phosphogypsum concrete, including initial porosity P0, compressive strength f0, ettringite content G0, water-cement ratio w / c, and the set of performance degradation coefficients S under the action of various erosion factors. S300. Based on T and S, establish a life prediction model W under the coupling effect of multiple factors. The life prediction model takes time as the main variable and dynamically iterates the evolution process of concrete parameters. S400, input the service time interval Δt of the target concrete structure, solve for W, and output the evolution curves of compressive strength f(t) and porosity P(t) within the service time interval; S500. Set the structural failure threshold fth. If f(t)≤fth, output the expected failure time tf of the structure as the prediction result of the concrete life.

[0006] Preferably, S100 includes: normalizing the changes in acid concentration, water flow velocity, wet-dry cycle frequency, and chloride ion concentration, and using the maximum-minimum normalization method to linearly transform the value of each factor within the target time period to the range of 0 to 1; and calculating the cumulative erosion impact value of each environmental factor within the analysis period based on the normalization results.

[0007] Preferably, the calculation steps of the cumulative erosion impact value include: setting the target analysis period as the period from the start time t0 to the end time t1, and calculating its integral value within the period based on the erosion impact index time function E(t), denoted as FE, wherein FE is used to quantitatively characterize the overall erosion intensity of the environment on the structural material; The calculation of FE is achieved by time discrete integration, which divides the time period into n time steps Δt, calculates the value of E(t) at each time step, and finally sums the product of the E(t) values ​​at all times and Δt to obtain the cumulative erosion load FE. The FE is input as a variable affecting the strength decay and pore evolution of concrete into the life prediction model W.

[0008] Preferably, S200 includes: The initial porosity was determined by the mercury pressure porosimeter method or nuclear magnetic resonance method, the compressive strength was obtained by the standard cubic specimen compression test, the ettringite content was determined by X-ray diffraction analysis, and the water-cement ratio was directly extracted from the mix design data. A material performance degradation response test system was constructed. Multiple combinations of different environmental erosion conditions were set up to periodically immerse phosphogypsum concrete samples and regularly measure the changes in compressive strength and porosity to form a dataset of measured curves showing the changes in material performance over time. Extract the slope or curve fitting parameters of the material properties' response to each erosion factor to form a set of parameters describing the performance degradation trend of concrete under different environmental effects. The set includes several attenuation coefficients, which are passed as input variables to the degradation model.

[0009] Preferably, the calculation steps for the attenuation coefficient include: taking environmental factors, including acidity, water flow velocity, wet-dry cycle frequency, and chloride ion concentration, as input variables, and taking the compressive strength or porosity measured at different time points as output variables to form a structured dataset; standardizing all input variables; and using the standardized data to establish a prediction model using the ridge regression method, extracting the regression weights corresponding to each input variable in the ridge regression model, and using them as coefficients characterizing the degree of influence of each erosion factor on the attenuation of concrete performance.

[0010] Preferably, a life prediction model W is established, using the erosion influence index time function E(t) and the material attenuation coefficient set as input parameters, with the concrete service time t as the main variable, to construct a set of state equations describing the evolution of material properties over time; the dynamic evolution functions of the material's compressive strength and porosity are defined and used as prediction indicators in model W, whose time change trends are determined by the combination of initial values, erosion index E(t), and attenuation coefficients, respectively; the model is solved using numerical methods, discretized using a fixed time step Δt, and the material properties at the current moment are updated based on E(t) and historical parameter states at each time step; a performance critical threshold is set to determine whether the material state at the current time step has reached the failure standard, and if so, the corresponding time t is output as the expected life of the structure.

[0011] Preferably, the construction of the lifetime prediction model W includes: Physical indicators highly correlated with concrete performance were selected as model output variables, including compressive strength, porosity, and damage variables. The erosion influence index time function E(t) was used as the driving variable, and combined with material decay parameters, an evolutionary differential expression of the state variables with respect to time was established, where the rate of change of the variables is controlled by the magnitude of E(t). The Euler explicit iterative method was used to discretely solve the evolution model, iteratively updating all state parameters at each time point. The sequence of state variables obtained by iterative calculation was stored as time-series data as a performance evolution trajectory, used to plot performance degradation curves and life trend prediction maps.

[0012] Preferably, in S400, the service time interval of the target concrete structure is set, the interval is discretized into several equally spaced time nodes, and the step size interval of each time node is determined for the stepwise solution of the subsequent model W. The erosion effect exponential time function E(t), initial material parameters, and attenuation coefficient are used as inputs. Model W is called, and at each time node, the compressive strength and porosity at the current time are calculated by a recursive algorithm based on the values ​​of the state variables at the previous time. During the model solving process, an adaptive step size control strategy is adopted. When the rate of change of the predicted value exceeds the set threshold, the time step is automatically reduced; otherwise, the time interval is increased. The calculated compressive strength f(t) and porosity P(t) at all time points are sorted and interpolated to output a complete performance evolution curve as a graphical result for life assessment.

[0013] Preferably, the process of solving model W includes: A numerical integration method based on the Runge-Kutta fourth-order method is used to solve for the changes of the state variables at each time point, and the dynamic update of the material properties is driven by the E(t) value at the current time. The evolution processes of strength and porosity are integrated into the same state space framework, and a multi-state variable coupled evolution path is constructed through joint solution. Output the time evolution curves of f(t) and P(t), including the curve slope, the maximum decay point, and the failure time node information.

[0014] This invention also provides a concrete life prediction system based on phosphogypsum concrete erosion, comprising: The environmental data acquisition module obtains historical erosion data of the environment in which the target structure is located, including changes in acid concentration, water flow velocity, wet-dry cycle frequency, and the time sequence T of chloride ion concentration changes. The material performance parameter modeling module obtains the material parameters of phosphogypsum concrete, including initial porosity P0, compressive strength f0, ettringite content G0, water-cement ratio w / c, and the set of performance degradation coefficients S under the action of various erosion factors. The life prediction model construction module establishes a life prediction model W under the coupling effect of multiple factors based on T and S. The life prediction model takes time as the main variable and dynamically iterates the evolution process of concrete parameters. The model solution and curve output module takes the service time interval Δt of the target concrete structure as input, solves for W, and outputs the evolution curves of compressive strength f(t) and porosity P(t) within the service time interval. The life determination and result output module sets the structural failure threshold fth. If f(t)≤fth, it outputs the expected failure time tf of the structure as the concrete life prediction result.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: 1. This invention constructs a dynamic life prediction model with engineering adaptability. By introducing the erosion influence index time function and the material attenuation coefficient set, it achieves coupled modeling of multiple environmental factors (such as acidity, water flow, chloride ions, etc.) and the concrete performance degradation process, making up for the shortcomings of existing methods in quantitative expression of material degradation behavior and multivariate dynamic simulation.

[0016] 2. This invention also utilizes high-precision numerical solution algorithms (such as the fourth-order Runge-Kutta method), a joint evolution mechanism of state variables, and a failure threshold determination model to output complete performance degradation curves and life prediction results, supporting differentiated evaluation needs for components of different functional levels in engineering projects. Compared to traditional empirical methods or single-factor life estimation methods, this invention significantly improves the scientific rigor, flexibility, and engineering applicability of the prediction results, and can be widely applied to the durability design and maintenance management of infrastructure, hydraulic structures, and industrial by-product concrete materials. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a flowchart of the method of the present invention.

[0019] Figure 2 This is a flowchart of the system modules of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1, please refer to Figure 1 As shown in this embodiment, a method for predicting the life of concrete based on phosphogypsum concrete erosion includes: S100. Obtain historical erosion data of the environment in which the target structure is located, including changes in acid concentration, water flow velocity, wet-dry cycle frequency, and the time sequence T of chloride ion concentration. S200. Obtain the material parameters of phosphogypsum concrete, including initial porosity P0, compressive strength f0, ettringite content G0, water-cement ratio w / c, and the set of performance degradation coefficients S under the action of various erosion factors. S300. Based on T and S, establish a life prediction model W under the coupling effect of multiple factors. The life prediction model takes time as the main variable and dynamically iterates the evolution process of concrete parameters. S400, input the service time interval Δt of the target concrete structure, solve for W, and output the evolution curves of compressive strength f(t) and porosity P(t) within the service time interval; S500. Set the structural failure threshold fth. If f(t)≤ fth, output the expected failure time tf of the structure as the prediction result of the concrete life.

[0022] In this invention, to achieve accurate life prediction of phosphogypsum concrete under complex environmental conditions, a time function model of erosion influence index based on multidimensional environmental factors is proposed. This model not only considers multiple corrosive parameters of the environment in which the concrete is located, but also performs time series analysis, normalization processing, and multi-factor weighted fusion. Furthermore, it obtains the cumulative erosion load during the long-term service of the structure through integral operations, providing key input variables for subsequent material performance degradation modeling and life prediction.

[0023] Specifically, this invention first acquires historical environmental data of the area where the target structure is located. This data includes, but is not limited to: changes in acid concentration (expressed as pH values), water flow velocity, wet-dry cycle frequency, and chloride ion concentration. This environmental data should cover the time period corresponding to the structural analysis cycle and be sampled at fixed time intervals, forming a multidimensional time series set of multiple environmental parameters changing over time. This step ensures that the collected data has temporal continuity and lateral integrity, providing a reliable foundation for subsequent model calculations.

[0024] Next, to address the differences in magnitude and physical dimensions of the aforementioned environmental factors, this invention employs the maximum-minimum normalization method to standardize the raw data. Specifically, the maximum and minimum values ​​of each parameter within the selected analysis period are used as benchmarks, and the value of that parameter at any given time is mapped to a range of 0 to 1 through a linear transformation. For example, if the acidity concentration at a certain time is x, the historical minimum is a, and the maximum is b, then the normalized acidity factor value is (xa) divided by (ba), and the result falls between 0 and 1. Other parameters such as water flow velocity, chloride ion concentration, and wet-dry cycle frequency are processed using the same method. Through this step, all factors are converted into dimensionless, comparable numerical functions, facilitating subsequent weighted analysis.

[0025] Subsequently, based on the normalized multi-factor results, this invention proposes a method for constructing the time function of the erosion influence index. In this method, the normalized acidity factor, water flow factor, chloride ion factor, and wet-dry cycle factor are denoted as A(t), V(t), C(t), and W(t), respectively. This invention assigns weights to each factor, denoted as α, β, γ, and δ, and the sum of these weight coefficients is 1. The weights can be obtained through regression fitting of existing experimental data, or can be set with reference to industry experience or structural exposure levels. By weighting and summing the values ​​of each factor at each time point according to their weights, a single-valued function that changes with time is formed, namely, the "erosion influence index time function E(t)". For example, if at a certain moment, A(t)=0.6, V(t)=0.4, C(t)=0.7, W(t)=0.5, and the weights are α=0.3, β=0.2, γ=0.3, δ=0.2 respectively, then the time function of the erosion influence index at that moment is 0.3×0.6 + 0.2×0.4 + 0.3×0.7 + 0.2×0.5, which is the comprehensive environmental corrosion intensity represented by that time point.

[0026] To more comprehensively characterize the total erosion intensity experienced by concrete during its service life, this invention proposes an integral form of the erosion influence index time function to calculate the "cumulative erosion load" FE. Specifically, the analysis period is set from the start time t0 to the end time t1, and the period is divided into several equal service time intervals Δt. At each time point, the corresponding erosion influence index time function E(t) is recorded, and the products of these index values ​​and the time step are added sequentially to form the cumulative erosion value for that time period. For example, if the values ​​of E(t1) to E(t5) are 0.3, 0.35, 0.32, 0.4, and 0.38 respectively between times t1 and t5, and Δt is 1 day, then the cumulative erosion load FE is 0.3×1 + 0.35×1 + 0.32×1 + 0.4×1 + 0.38×1, or 1.75, representing the total erosion influence intensity of the environment on concrete during this period. This index reflects the long-term cumulative erosion effect and can be used as input for strength and porosity evolution models.

[0027] In summary, this invention, by constructing a multi-factor environmental erosion index function and its integral accumulation mechanism, achieves an effective transformation from external environmental data to material degradation input variables, possessing a clear physical logic foundation and good engineering applicability. It is particularly suitable for the life assessment and maintenance strategy formulation of industrial by-product concrete structures (such as phosphogypsum concrete) in complex service environments.

[0028] To accurately predict the service life of phosphogypsum concrete under various corrosive environments, this invention proposes a systematic method for acquiring material parameters and a performance degradation modeling process. This method not only collects key initial material parameters but also constructs a multi-factor degradation response experimental system. Furthermore, it uses the ridge regression algorithm to quantitatively analyze the performance degradation impact of various environmental factors, ultimately forming a set of degradation coefficients for degradation modeling.

[0029] Specifically, the initial performance parameters of the target concrete sample were first determined, including but not limited to initial porosity, compressive strength, ettringite content, and water-cement ratio. To ensure the accuracy and scientific validity of the parameter acquisition, the following methods were employed: Initial porosity was measured using either the mercury pressure porosimeter method or nuclear magnetic resonance (NMR). The mercury pressure method is suitable for analyzing medium- to large-pore structures with a wide range of pore sizes, while NMR can provide images of the pore distribution under non-destructive conditions. The two methods can complement each other to improve the representativeness of the data.

[0030] The compressive strength was determined by standard cubic specimen compression test, in accordance with the "Standard for Test Methods of Mechanical Properties of Ordinary Concrete", ensuring that the test conditions, loading speed and specimen curing environment all meet industry standards.

[0031] The content of ettringite is an important indicator for evaluating the active mineral components in phosphogypsum concrete. This invention uses X-ray diffraction (XRD) analysis for phase analysis and semi-quantitative determination, and combines Rietveld full-spectrum fitting technology to improve data accuracy.

[0032] The water-cement ratio parameter is directly derived from the concrete mix design data. It is extracted by consulting the original construction records or material unit proportion tables and serves as one of the basic design parameters input into the model.

[0033] After obtaining the initial parameters of the aforementioned materials, this invention constructs an experimental system for the performance degradation response of phosphogypsum concrete. This system simulates the corrosion encountered by phosphogypsum concrete under complex actual service environments by designing multiple representative environmental conditions, such as different acidity levels (e.g., pH 4-7), different chloride ion concentrations (e.g., 1%, 3%, 5%), different wet-dry cycle frequencies (daily, weekly), and different water flow shear strengths. Under each environmental condition, concrete specimens are prepared and subjected to periodic immersion or spraying treatments. Fixed time points (e.g., 7 days, 28 days, 60 days, 90 days) are set for repeated measurements of compressive strength and porosity, thereby obtaining complete measured data curves showing the performance changes over time.

[0034] Based on the experimental data above, this invention further proposes a modeling strategy for calculating the performance degradation coefficient using the ridge regression method. First, environmental factor data (including acidity level, water flow velocity, wet-dry cycle frequency, and chloride ion concentration) under each experimental condition are used as input variables, and the concrete compressive strength or porosity corresponding to each time point is used as the output variable, constructing a structured dataset. Each row represents an experimental sample, and each column represents either the input factor or the output performance data, facilitating unified processing and modeling.

[0035] Next, all input variables are standardized. This step aims to eliminate the imbalance effect on the model results caused by scale inconsistencies between variables of different dimensions (such as pH value and flow rate units). The standardization method used is the "zero mean unit variance method," which involves subtracting the mean from the value of each variable and then dividing by its standard deviation to obtain a dimensionless input data matrix with a uniform scale.

[0036] After standardization, the ridge regression algorithm was used for model training. This algorithm introduces a regularization parameter into traditional multiple linear regression, which, by imposing a penalty term on the model coefficients, avoids model instability or abnormal results when there is high collinearity among variables (such as simultaneous changes in wet-dry frequency and chloride ion concentration). The ridge regression model obtains the regression weight of each input variable by minimizing the sum of the prediction error and the sum of the squares of the coefficients.

[0037] Finally, regression coefficients corresponding to each environmental factor are extracted from the trained ridge regression model. These coefficients can be directly regarded as the quantitative influence intensity of different environmental factors on the performance degradation of concrete under the current experimental settings, i.e., the "degradation coefficients". This set of coefficients can provide a parameter basis for the material degradation prediction model, supporting the subsequent construction of strength evolution functions, porosity growth models, or life prediction formulas.

[0038] To achieve accurate prediction of the lifespan of phosphogypsum concrete under complex environmental erosion, this invention designs and establishes a lifespan prediction model W in step S300, which uses time as the main variable and dynamically iterates concrete performance parameters. This model uses the previously obtained erosion influence index time function E(t) and the set of performance degradation coefficients S as core input variables. Combined with initial material state parameters, it constructs a performance evolution path and outputs the predicted failure time of the structure when a performance threshold is reached.

[0039] First, in the construction phase of the life prediction model W, this invention sets the structural service time t as the main variable of the model, and all material state parameters are described as functions of time. The main prediction objects include compressive strength and porosity, as important indicators reflecting the structural load-bearing capacity and durability.

[0040] The core of the model is based on the following logic: the impact of environmental factors on concrete performance is a cumulative process, and the intensity and duration of different environmental factors jointly determine the rate of performance degradation. Therefore, this invention uses the erosion influence index time function E(t) established above as a control variable, and introduces a performance decay coefficient extracted experimentally to quantify this influence. Taking compressive strength as an example, its evolution over time can be described as follows: the current strength value equals the initial strength value multiplied by a time-dependent decay function, and the form of this function depends on the environmental influence index and the material response parameters.

[0041] To construct this dynamic functional relationship, this invention sets compressive strength and porosity as state variables and introduces an evolutionary differential equation for them. This equation links the rate of change of the state variables to E(t) and the material sensitivity coefficient. For example, the rate of change of compressive strength can be set as follows: the strength decay rate is proportional to E(t) and controlled by a sensitivity coefficient fitted from an experiment. Similarly, porosity increases with a square root of time, and its growth rate is also jointly affected by E(t) and the intensity of environmental influences.

[0042] In terms of model solving, this invention employs a time discretization strategy, dividing the entire lifetime analysis cycle into equally spaced time steps Δt. The model starts from the initial time t0, and at each time step, based on the material parameter values ​​from the previous time step, the current E(t) value, and a known set of attenuation coefficients, iteratively calculates the compressive strength and porosity values ​​for the current time step. This iterative process uses the Euler explicit iterative method, that is, using the state variables from the previous time step plus the changes within that time step to predict the new state at the current time step.

[0043] Specifically, the following iterative operation is performed at each time step: Read the compressive strength and porosity at the previous moment; Based on the E(t) value and material parameters at the current time point, calculate the increase or decrease of the state variables within this time step; Update and store the current state value; Determine if the compressive strength is lower than the set failure threshold. If it is, terminate the iteration and output the current time as the predicted lifetime.

[0044] The failure threshold can be set according to the structural functional level, such as 30% of the initial compressive strength for load-bearing components and 50% for non-load-bearing components. Through this mechanism, the present invention can not only output complete strength-time curves and porosity-time curves, but also accurately predict the time node for performance failure, serving as a quantitative prediction result of the structural life.

[0045] Furthermore, to improve the model's practicality and accuracy, this invention supports the embedding of an environmental prediction module, which can introduce future climate conditions or groundwater chemical parameters into the E(t) function prediction, forming an integrated assessment tool that combines "prospective corrosion simulation + material response calculation." This lifetime prediction model is particularly suitable for phosphogypsum concrete structures significantly affected by acid corrosion, salt corrosion, and hydrodynamic circulation, such as infrastructure, hydraulic structures, and wet desulfurization related projects.

[0046] The service life prediction model W constructed by this invention can fully reflect the complex evolution process of the coupling of multiple environmental factors and material properties, and achieve high-precision prediction of the service life of concrete, providing a scientific basis for concrete material design, structural durability management and service performance evaluation.

[0047] In this invention, to achieve quantitative prediction of the service life of concrete structures, it is necessary to systematically solve the aforementioned completed service life prediction model W. Step S400 is the core computational stage of this invention, and its main task is to output the dynamic evolution curves of concrete compressive strength f(t) and porosity P(t) based on model W within a given service time interval. To ensure prediction accuracy and computational stability, this invention introduces several numerical analysis strategies, including high-order differential equation solving methods, adaptive step size control, joint evolution mechanism of state variables, and error correction modules, to ensure that the obtained curves are operable and reliable.

[0048] First, the service time interval of the concrete structure is defined as from the start time t0 to the end time t1. This invention discretizes this interval, dividing it into several time steps with equal or adaptive intervals, each step size denoted as Δt. The initial Δt can be set to 1 day or 1 week by default, determined based on the frequency of environmental changes and the sensitivity to performance evolution. This discretization transforms the continuous-time function into a discrete-time series within a numerical computation framework, facilitating the iterative solution of state variables.

[0049] Next, using the established model W, the compressive strength and porosity of the material are calculated at each discrete time point. Model W uses the erosion influence exponent time function E(t) as the external driving variable and the initial material parameters and attenuation coefficient set as internal parameters, iteratively updating the strength and porosity as time functions. To improve prediction accuracy, this invention uses the fourth-order Runge-Kutta method for numerical integration. This algorithm evaluates the rate of change of state variables at the current time point, the midpoint of the time step, and the end point within each time step, and obtains the accurate state prediction value for that step through a weighted average. Compared to the Euler method, the fourth-order Runge-Kutta method has significantly lower truncation error and stronger convergence, making it suitable for coupled models containing nonlinear dynamic evolution.

[0050] Furthermore, to balance computational efficiency and accuracy, this invention introduces an "adaptive step size control strategy." Specifically, during the calculation at each time step, if the rate of change of a state variable (such as compressive strength) exceeds a preset threshold (e.g., a decrease of more than 5% per step), the time step size Δt is automatically reduced to obtain a more refined ability to capture dynamic changes; conversely, if the rate of change is below a certain tolerance range, Δt can be increased to improve computational efficiency. This strategy achieves a balance between computational accuracy and efficiency through real-time judgment of the model's dynamic behavior.

[0051] To address the correlation between coupled variables during the solution process, this invention employs a "state-space joint evolution mechanism." That is, compressive strength f(t) and porosity P(t) are not solved independently, but are treated as two dimensions within the same state vector, and calculated simultaneously using a joint system of differential equations. This method considers the indirect impact of increased porosity on strength decay, as well as the synergistic effect of the two, consistent with the physical laws governing actual material degradation.

[0052] Furthermore, to enhance the model's adaptability and engineering application accuracy, this invention introduces a "dynamic error correction mechanism." This mechanism compares the predicted values ​​with historically known measured data (such as early field inspection intensity). If an error accumulation trend is observed, it automatically adjusts model parameters (such as local attenuation coefficients) or boundary conditions (such as the initial E(t) curve) to maintain a high correlation between the model's prediction results and the measured curve. This mechanism is particularly suitable for service structures with existing phased monitoring data and can significantly improve the realistic accuracy of lifespan prediction.

[0053] After model W completes calculations for the entire time interval, the compressive strength f(t) and porosity P(t) values ​​at all time points are stored as a time-series array. Subsequently, the data is smoothed using an interpolation algorithm (such as cubic spline interpolation) to plot a complete performance evolution curve. This curve not only shows the performance degradation trend of the material during its service life but also marks key nodes, such as: The steepest range of strength decline rate; the inflection point of porosity growth; the moment when the failure threshold is reached (e.g., the time point when the strength drops to 30% of the initial value).

[0054] Ultimately, this evolution curve can serve as a basis for formulating engineering maintenance and reinforcement strategies, as well as as fundamental data for material selection or service life prediction during the design phase.

[0055] In summary, by integrating high-precision numerical solutions, step size control, state-based solutions, and error correction mechanisms, this invention systematically completes the dynamic solution and visualization output of the life model of phosphogypsum concrete structures under complex corrosive environments, providing reliable technical support and implementation path for intelligent assessment of concrete life.

[0056] In the aforementioned steps, this invention constructs the concrete performance evolution curves f(t) and P(t), clearly demonstrating the dynamic degradation process of structural materials under multi-factor erosion. To transform the performance curves into quantifiable life prediction results, step S500 establishes a structural failure threshold mechanism and determines the expected failure time tf, i.e., the endpoint of the concrete life prediction, based on this threshold.

[0057] Specifically, the structural failure threshold fth refers to the point at which the compressive strength f(t) of concrete material decreases below a certain critical level, at which point it is considered to no longer possess the design performance or structural safety. This critical level can be flexibly set based on factors such as engineering standards, component functional levels, and safety factors, and has a certain degree of adaptability.

[0058] In the basic embodiment of this invention, fth is set by default as a percentage of the initial compressive strength f0 of the material, with a common setting range of 30% to 60%, depending on the specific structural application. For example: For major load-bearing components (such as beams, columns, and supports), the failure threshold can be set at 30% of the initial strength to ensure that it always meets the load-bearing requirements; for non-load-bearing components (such as infill walls and protective layers), the threshold can be appropriately relaxed to 50% or 60% to allow for a certain degree of material loss. For components requiring high durability in exposed environments, the threshold can also be increased based on standards and specifications, such as adopting the compressive strength limits specified in the "Code for Design of Durability of Concrete Structures".

[0059] In the specific solution process, this invention uses the following logic to determine the lifespan: During the solution process of model W, the current compressive strength f(t) is generated at each time point. The program compares the value of f(t) at each time point with the preset fth. If it is found that a certain time point ti satisfies f(ti)≤fth, it means that the structural performance has reached the failure standard. Then, this time point is identified as the "expected failure time of the structure" tf and is output as the lifespan prediction.

[0060] To improve the accuracy and stability of the judgment, this invention introduces a linear interpolation mechanism in the failure point identification process. Since time is simulated with an discrete step length Δt, the actual f(t) = fth may fall exactly between two time points. Therefore, after the two time points before and after the critical point are found to satisfy the conditions of "above the threshold" and "below the threshold" respectively, the program estimates a more accurate failure time point using linear interpolation. For example, if f(t30) = 12 MPa on day 30, which is above the threshold, and f(t31) = 9 MPa on day 31, which is below the threshold, and the threshold is set to 10 MPa, then the program calculates that the failure occurred between day 30 and day 31, and the estimated time point can be 30.5 days, which is output as the lifetime prediction result. Furthermore, to achieve broad adaptability and intelligent expansion of the model, this invention also introduces a "threshold adaptive setting module." This module supports the following functions: Automatic matching based on component functional level: Input the component type during the design phase, and the system will automatically retrieve the corresponding recommended threshold; Thresholds are inversely calculated based on the target lifespan: If the required lifespan of the project is 100 years, the system sets a threshold in reverse so that the predicted intensity curve just reaches the target at 100 years. Threshold adjustment based on on-site monitoring: If there are signs of damage at a certain stage after construction, the threshold can be appropriately lowered to reflect the expected actual performance.

[0061] The output format of lifetime prediction results in TensorFlow can be diverse, including but not limited to: Output lifetime time in numerical form; The graph is created by overlaying the f(t) curve with the horizontal line fth and marking the intersection points. In the structural digitization platform, the areas of components that are about to fail are displayed in a visually marked manner.

[0062] Finally, the predicted lifetime (tf) can serve as an important reference input for the formulation of operation and maintenance strategies, for example: Schedule time for structural maintenance and renovation; Establish regular testing intervals; Prioritize resource allocation for areas with high risk of degradation.

[0063] In summary, this invention constructs a structural failure threshold judgment mechanism and combines it with dynamic strength curve output to form a closed-loop calculation process from environmental erosion data to life end time prediction. It has good engineering adaptability and technical expansion potential, and is a key link in structural life modeling and prediction.

[0064] Example 2, please refer to Figure 2 As shown in this embodiment, a concrete life prediction system based on phosphogypsum concrete erosion includes: The environmental data acquisition module obtains historical erosion data of the environment in which the target structure is located, including changes in acid concentration, water flow velocity, wet-dry cycle frequency, and the time sequence T of chloride ion concentration changes. The material performance parameter modeling module obtains the material parameters of phosphogypsum concrete, including initial porosity P0, compressive strength f0, ettringite content G0, water-cement ratio w / c, and the set of performance degradation coefficients S under the action of various erosion factors. The life prediction model construction module establishes a life prediction model W under the coupling effect of multiple factors based on T and S. The life prediction model takes time as the main variable and dynamically iterates the evolution process of concrete parameters. The model solution and curve output module takes the service time interval Δt of the target concrete structure as input, solves for W, and outputs the evolution curves of compressive strength f(t) and porosity P(t) within the service time interval. The life determination and result output module sets the structural failure threshold fth. If f(t)≤fth, it outputs the expected failure time tf of the structure as the concrete life prediction result.

[0065] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for predicting the life of concrete based on the erosion of ardealite concrete, characterized by: Comprise: S100, obtain the historical erosion data of the environment where the target structure is located, including the change sequence T of the acid concentration, water flow speed, wet-dry cycle frequency and chloride ion concentration over time; S200, obtain the material parameters of phosphogypsum concrete, including initial porosity P0, compressive strength f0, ettringite content G0, water-binder ratio w / c and its performance decay relationship coefficient set S under the action of each erosion factor; S300, based on T and S, establish a life prediction model W under the coupling action of multiple factors, which takes time as the main variable and dynamically iterates the evolution process of concrete parameters; S400, input the service time interval Δt of the target concrete structure, solve W, and output the evolution curves of compressive strength f(t) and porosity P(t) in the service time interval; S500, set the structure failure threshold fth, if f(t)≤fth, output the predicted failure time tf of the structure as the concrete life prediction result.

2. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 1, characterized in that: S100 comprises: normalizing the acid concentration change, water flow speed, wet-dry cycle frequency and chloride ion concentration, using the maximum and minimum normalization method to linearly convert the value of each factor in the target time period to the interval of 0 to 1; according to the normalization result, calculate the cumulative erosion impact value of each environmental factor in the analysis period.

3. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 2, characterized in that: The calculation steps of the cumulative erosion impact value include: setting the target analysis period as the starting time t0 to the ending time t1, based on the erosion impact index time function E(t), calculating its integral value in the period, denoted as FE, FE is used to quantitatively represent the overall erosion effect of the environment on the structure material; FE is calculated by using time discrete integral method, dividing the time period into n time steps Δt, and calculating the value of E(t) at each time, finally adding up the product of all E(t) values and Δt to get the cumulative erosion load FE; FE is input into the life prediction model W as a variable affecting the strength decay and pore evolution of concrete.

4. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 1, characterized in that: Wherein S200 comprises: The initial porosity is measured by pressure mercury porosimeter method or nuclear magnetic resonance method, the compressive strength is obtained by standard cubic specimen compression experiment, the ettringite content is measured by X-ray diffraction analysis method, and the water-binder ratio is directly extracted from the mix proportion design data; A material performance decay response test system is constructed, a plurality of different environmental erosion working conditions are set, phosphogypsum concrete samples are periodically soaked, and the changes of their compressive strength and porosity are measured regularly to form a measured curve data set of material performance changing with time; Extract the response slope or curve fitting parameters of material performance to each erosion factor to form a set of parameters describing the performance degradation trend of concrete under the action of different environments, the set includes several decay coefficients, which are input into the degradation model as input variables.

5. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 4, characterized in that: The attenuation coefficient calculation step comprises: taking environmental factors including acidity, water flow speed, wet-dry cycle frequency and chloride ion concentration as input variables, and taking the compressive strength or porosity measured at different time points as output variables to form a structured data set; performing standardization processing on all input variables, establishing a prediction model based on the standardized data by using a ridge regression method, and extracting the regression weight of each input variable in the ridge regression model as a coefficient representing the influence degree of each erosion factor on the performance attenuation of the concrete.

6. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 1, characterized in that: A life prediction model W is established, which takes the erosion influence index time function E(t) and the material attenuation coefficient set as input parameters, and takes the concrete service time t as the main variable to construct a state equation set for describing the evolution of material performance over time; a dynamic evolution function of the compressive strength and porosity of the material is defined as a prediction index in the model W, and the initial value, the erosion index E(t) and the attenuation coefficient are combined to determine the time variation trend; the model is solved by a numerical method, a fixed time step Δt is used for discretization processing, and the material performance at the current time is updated according to E(t) and the historical parameter state at each time step; a performance critical threshold is set to determine whether the material state at the current time step reaches the failure standard, and if so, the corresponding time t is output as the predicted service life of the structure.

7. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 6, characterized in that: The construction of the life prediction model W comprises: selecting physical indexes highly related to the performance of the concrete as the output variables of the model, including the compressive strength, the porosity and the damage variable; taking the erosion influence index time function E(t) as a driving variable, combining the material attenuation parameters, and establishing a differential expression of the evolution of the state variable with respect to time, wherein the change rate of the variable is controlled by the size of E(t); the Euler explicit iteration method is used to discretely solve the evolution model, and all state parameters are iteratively updated at each time point; the state variable sequence obtained by iteration calculation is stored as time series data as the performance evolution trajectory, which is used to draw the performance degradation curve and the life trend prediction graph.

8. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 1, characterized in that: In S400, the service time interval of the target concrete structure is set, the interval is discretized into a plurality of equally spaced time nodes, and the step interval of each time node is determined for subsequent step-by-step solving of the model W; taking the erosion influence index time function E(t), the initial parameters of the material and the attenuation coefficient as inputs, calling the model W, and calculating the compressive strength and porosity at the current time according to the numerical value of the state variable at the last time point at each time node by using a recursive algorithm; In the model solving process, an adaptive step control strategy is used, when the change rate of the prediction value exceeds a set threshold, the time step is automatically reduced; otherwise, the time interval is increased; The calculated compressive strength f(t) and porosity P(t) at all time nodes are sorted and interpolated, and the complete performance evolution curve is output as the graphical result of the life assessment.

9. A method for predicting the life of concrete based on the erosion of phosphogypsum concrete according to claim 8, characterized in that: The process of solving the model W comprises: a numerical integration method based on the Runge-Kutta fourth-order method is used to solve the change of the state variable at each time point, and the E(t) value at the current time is used to drive the dynamic update of the material performance; The evolution of strength and porosity is unified in the same state space framework, and the coupled evolution path of multiple state variables is constructed by joint solution; The time evolution curves of f(t) and P(t) are output, including the slope of the curve, the maximum attenuation point and the failure time node information.

10. A system for predicting the life of concrete under attack by phosphogypsum concrete, for implementing a method for predicting the life of concrete under attack by phosphogypsum concrete according to any one of claims 1 to 9, characterized in that: It includes: The environmental data acquisition module acquires the historical erosion data of the target structure, including the change of acid concentration, water flow velocity, wet-dry cycle frequency and chloride ion concentration over time T; The material performance parameter modeling module acquires the material parameters of phosphogypsum concrete, including the initial porosity P0, compressive strength f0, ettringite content G0, water-cement ratio w / c and the performance attenuation relationship coefficient set S under the action of each erosion factor; The life prediction model construction module establishes the life prediction model W under the action of multiple factors based on T and S, which takes time as the main variable and dynamically iterates the evolution process of concrete parameters; The model solution and curve output module inputs the service time interval Δt of the target concrete structure, solves W, and outputs the evolution curves of compressive strength f(t) and porosity P(t) in the service time interval; The life determination and result output module sets the structure failure threshold fth, if f(t)≤fth, the predicted failure time tf of the structure is output as the concrete life prediction result.

Images