Non-parametric prediction method for fatigue life of steel-FRP composite bar

Abstract

The invention discloses a nonparametric prediction method for the fatigue life of a steel-FRP composite bar, and belongs to the field of fatigue performance evaluation of civil engineering materials and structures, and the nonparametric prediction method comprises the following steps: initializing component material attributes, load conditions and initial states of the steel-FRP composite bar, the component materials comprising an inner core steel bar and an FRP layer; according to an equivalent stiffness distribution principle, fatigue stresses of the component materials are calculated respectively; according to the fatigue stress of the component material, theoretical fatigue life of the component material is obtained in combination with the S-N curve relation, and cyclic jumping is executed; calculating accumulated fatigue damage of the component materials respectively based on results after cyclic jumping; and circularly updating the dynamic stiffness and the equivalent cycle number of the component material until the accumulated fatigue damage of any component material reaches a critical threshold value, judging that the component material is invalid, and outputting the fatigue life. According to the method, the fatigue life of the steel-FRP composite rib can be efficiently and accurately predicted, and the test cost is remarkably reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of fatigue performance evaluation of civil engineering materials and structures, and in particular relates to a non-parametric prediction method for fatigue life of steel-FRP composite bars. Background Art

[0002] Steel-FRP composite bars (SFCBs) are an emerging reinforcement material that leverages the complementary properties of steel and fiber-reinforced plastic (FRP). They combine the ductility of steel with the corrosion resistance of FRP, using steel as the inner core and FRP as the outer layer. In modern engineering structures, SFCBs are widely used in various load-bearing structures due to their excellent mechanical properties and durability. SFCBs, composed of an inner steel core and an FRP layer, exhibit unique fatigue behavior under cyclic loading. However, the complexity of their fatigue properties presents challenges for engineering design and application.

[0003] Experimental studies have shown that the fatigue behavior of SFCBs is highly dependent on the dynamic damage evolution of its component materials (core steel and FRP layer). Under cyclic loading, there is a significant difference in the stiffness degradation rate of the core steel and FRP layer, which triggers stress redistribution. Generally, the component with greater stiffness loss (such as the core steel) will transfer the fatigue load to the component with better stiffness retention (such as the FRP layer). However, this stress transfer is not absolute. For example, under high stress levels, the FRP layer may be subjected to stress levels above its fatigue limit, which accelerates damage and causes the load to transfer back from the core steel to the FRP layer. This dynamic coupling effect makes it difficult to accurately predict the fatigue life of SFCBs using models based on single-component materials.

[0004] In engineering practice, to meet design requirements for load-bearing capacity and durability, SFCBs of varying specifications, such as varying inner core diameters and FRP layer thicknesses, are often required. Predicting fatigue life using the traditional method recommended by current standards requires separate fatigue testing for each SFCB specification. This approach is not only time-consuming and labor-intensive, but also costly, severely restricting the large-scale application of SFCBs in real-world projects. Therefore, developing an efficient and accurate fatigue life prediction method is crucial for promoting the widespread application of SFCBs in engineering. Summary of the Invention

[0005] To address these technical issues, the present invention provides a nonparametric method for predicting the fatigue life of steel-FRP composite bars. This method calibrates the fatigue life (SN) curves of the component materials and combines dynamic stiffness degradation and stress redistribution mechanisms to construct a damage evolution framework for SFCBs. Furthermore, a cyclic jumping algorithm dynamically modifies the stress distribution direction and ratio based on real-time stiffness degradation and damage states, quantitatively characterizing the load transfer from the core steel to the FRP layer (or vice versa). Compared to traditional methods, this method significantly reduces the amount of testing, overcomes the limitation of traditional damage models' over-reliance on testing, and provides a viable solution for the fatigue design of SFCBs.

[0006] The present invention proposes a non-parametric prediction method for fatigue life of steel-FRP composite bars, comprising the following steps:

[0007] Initialize the component material properties, loading conditions, and initial states of the steel-FRP composite reinforcement, where the component materials include the core steel bar and the FRP layer;

[0008] According to the principle of equivalent stiffness distribution, the fatigue stress of component materials is calculated separately;

[0009] According to the fatigue stress of the component materials, combined with the SN curve relationship, the theoretical fatigue life of the component materials is obtained respectively, and the cycle jump is performed;

[0010] Based on the results after cyclic skipping, the cumulative fatigue damage of the component materials is calculated separately;

[0011] The dynamic stiffness and equivalent cycle number of the component materials are updated cyclically until the cumulative fatigue damage of any component material reaches a critical threshold. The component material is then judged to be failed and the fatigue life is output.

[0012] Optionally, the process of initializing the component material properties, load conditions, and initial state of the steel-FRP composite reinforcement is specifically as follows:

[0013] Initialize the component material properties based on the input elastic modulus and cross-sectional area of ​​the inner core steel bar, the elastic modulus and cross-sectional area of ​​the FRP layer, and the overall cross-sectional area of ​​the steel-FRP composite bar. Initialize the load conditions based on the input stress level. Set the initial number of cycles to zero and the cycle sequence to one to complete the initial state setting.

[0014] Optionally, the process of respectively calculating the fatigue stress of the component materials according to the equivalent stiffness distribution principle is specifically as follows:

[0015] Based on the component material properties of the inner core steel bar and FRP layer, the strains of both under peak fatigue load are calculated. Based on the strain synergy assumption, the fatigue modulus of the inner core steel bar and FRP layer are calculated. Based on the fatigue modulus and cross-sectional area, the fatigue stress borne by the inner core steel bar and FRP layer are calculated separately.

[0016] Optionally, the process of obtaining the theoretical fatigue life of the component materials respectively according to the fatigue stress of the component materials in combination with the SN curve relationship and performing the cycle jump is specifically as follows:

[0017] Based on the fatigue stress levels of the inner core steel bars and FRP layer, an improved SN curve relationship using the Batsoulas formula is substituted. The theoretical fatigue life of the inner core steel bars and FRP layer is calculated based on the Batsoulas formula. Based on the theoretical fatigue life, the step size of the cycle skipping is determined, and the cycle skipping is performed.

[0018] Optionally, the process of respectively calculating the cumulative fatigue damage of the component materials based on the results after the cycle jump is specifically as follows:

[0019] Based on the results after cyclic jumping, the degree of fatigue stiffness degradation of the inner core steel bar and FRP layer is calculated; the fatigue stiffness degradation degree includes a damage index based on fatigue stiffness degradation and a damage index based on residual strength decay; the damage index based on fatigue stiffness degradation is used to characterize the nonlinear cumulative damage of the component material in the early stage of fatigue loading, and the damage index based on residual strength decay is used to describe the damage mutation at the end of fatigue loading.

[0020] Optionally, the process of cyclically updating the dynamic stiffness and equivalent cycle number of the component materials until the cumulative fatigue damage of any component material reaches a critical threshold, then determining that the component material has failed and outputting the fatigue life is specifically as follows:

[0021] Based on the current cycle status, the dynamic stiffness of the inner core steel bar and FRP layer is updated; based on the updated dynamic stiffness, the equivalent number of cycles is calculated; the program status, including the number of cycles and the cycle sequence, is updated; and it is determined whether the cumulative fatigue damage of the inner core steel bar or FRP layer has reached the critical threshold. If so, the cycle is terminated and the fatigue life is output.

[0022] The present invention also provides a non-parametric prediction system for fatigue life of steel-FRP composite bars, which is used to implement the method described above, comprising:

[0023] A parameter initialization module is used to initialize the component material properties, load conditions, and initial states of the steel-FRP composite reinforcement, where the component materials include the inner core steel bar and the FRP layer;

[0024] Fatigue stress calculation module, used to calculate the fatigue stress of component materials separately according to the principle of equivalent stiffness distribution;

[0025] Fatigue life calculation module, used to obtain the theoretical fatigue life of component materials based on the fatigue stress of component materials and the SN curve relationship, and perform cycle skipping;

[0026] Fatigue damage calculation module, used to calculate the cumulative fatigue damage of component materials based on the results after cyclic jumping;

[0027] The cyclic update module is used to cyclically update the dynamic stiffness and equivalent cycle number of the component materials until the cumulative fatigue damage of any component material reaches a critical threshold, then the component material is judged to be failed and the fatigue life is output.

[0028] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method.

[0029] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method when executed by a processor.

[0030] The present invention also provides a computer program product, comprising a computer program, which implements the steps of the method when executed by a processor.

[0031] Compared with the prior art, the present invention has the following advantages and technical effects:

[0032] The present invention is based on initializing the component material properties, load conditions and initial state of the steel-FRP composite reinforcement to provide accurate initial parameters for the fatigue life prediction model, ensuring that the prediction results are accurate and reliable, and laying the foundation for subsequent predictions.

[0033] The present invention calculates the fatigue stress of component materials separately based on the principle of equivalent stiffness distribution, accurately distributes the stress of the inner core steel bar and the FRP layer, provides accurate stress input for fatigue life prediction, and truly reflects the actual stress state of the material.

[0034] The present invention obtains theoretical fatigue life and performs cycle skipping based on the fatigue stress of component materials in combination with an improved SN curve relationship, effectively predicts theoretical fatigue life, reduces the amount of calculation and improves efficiency through cycle skipping, and ensures calculation accuracy at the same time.

[0035] The present invention calculates the cumulative fatigue damage of component materials based on the results after cyclic jumps, dynamically tracks the fatigue damage accumulation process, provides a basis for fatigue failure judgment, and comprehensively reflects the development trend of material damage.

[0036] The present invention is based on cyclically updating the dynamic stiffness and equivalent cycle number of component materials until the cumulative fatigue damage reaches a critical threshold, updating material properties in real time, accurately determining the moment of fatigue failure, outputting accurate fatigue life, improving prediction accuracy, and enhancing model adaptability and practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0038] Figure 1 Schematic diagram of equal damage curves according to an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of a fatigue damage development curve according to an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of a loop skipping method according to an embodiment of the present invention;

[0041] Figure 4 Schematic diagram of the SFCBs fatigue life prediction process according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0043] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0044] Example 1

[0045] This embodiment provides a non-parametric prediction method for fatigue life of steel-FRP composite bars, comprising the following steps:

[0046] Initialize the component material properties, loading conditions, and initial states of the steel-FRP composite reinforcement, where the component materials include the core steel bar and the FRP layer;

[0047] According to the principle of equivalent stiffness distribution, the fatigue stress of component materials is calculated separately;

[0048] According to the fatigue stress of the component materials, combined with the SN curve relationship, the theoretical fatigue life of the component materials is obtained respectively, and the cycle jump is performed;

[0049] Based on the results after cyclic skipping, the cumulative fatigue damage of the component materials is calculated separately;

[0050] The dynamic stiffness and equivalent cycle number of the component materials are updated cyclically until the cumulative fatigue damage of any component material reaches a critical threshold. The component material is then judged to be failed and the fatigue life is output.

[0051] As a specific implementation method, the following steps are specifically included:

[0052] The basic framework of the damage model:

[0053] Improvements to the SN curve:

[0054] In fatigue design, the Basquin formula is generally used to express the fatigue life curve (SN curve). According to the Basquin formula, the SN curve can be expressed by formula (1):

[0055] (1)

[0056] Where, σ is fatigue stress; σ f ' is the fatigue strength coefficient; b is the fatigue strength index; N f is the number of cycles to failure, i.e. the theoretical fatigue life; A and B are constants. The Basquin formula is expressed in a double logarithmic coordinate system as the logarithmic stress level (logS) and the logarithmic fatigue life (logN f ) linear relationship.

[0057] However, for materials with obvious fatigue limits (such as steel bars and FRP bars), the Basquin formula has significant limitations: when the stress level is close to the fatigue limit of the material, its prediction curve is inconsistent with the horizontal axis (logN f ) intersect, causing its predicted value to continue to decline in the ultra-high cycle fatigue region, overestimating the degree of material damage and seriously deviating from the actual progressive damage behavior of the material. In addition, because the elastic modulus of the FRP layer is significantly lower than that of the inner core steel bar, its service stress level is usually lower than its own fatigue limit, further highlighting the applicability shortcomings of the traditional model. The current specification sets the fatigue life greater than 2×10 6 The range of cycles is assumed to be "infinite life". Although the artificial definition of fatigue limit simplifies the design process, it ignores the gradual damage accumulation of materials in the ultra-high cycle region and may underestimate the actual service risk. It is worth mentioning that low-cycle fatigue is usually defined as fatigue life less than 10 3 The fatigue life of the fatigue cycle is defined as the area of ​​​​1 million cycles, and the area between the two is defined as medium- and high-cycle fatigue. In response to the above problems, an improved SN curve characterization method is proposed, as shown in formula (2).

[0058] (2)

[0059] Where C is the material constant, σ0 is the fatigue limit, N e The minimum number of cycles required to cause fatigue damage. The Batsoulas formula uses a hyperbolic function, and its curve is in the middle and high cycle fatigue region (10 3 ≤N f ≤2×10 6 ) is consistent with the trend of Basquin formula, while in the ultra-high cycle fatigue region (N f ≥2×10 6 ) converges asymptotically to σ0, the material's theoretical fatigue limit. This property not only corrects the Basquin formula's overestimation of very high-cycle fatigue damage but also reveals, from a physical perspective, the durability boundary conditions of materials under high-cycle loading.

[0060] Damage accumulation under multi-level loading:

[0061] Different from the fatigue mechanism of pure steel bars / pure FRP bars, the fatigue damage of SFCBs under tensile fatigue is affected by the interaction between the inner core steel bars and the FRP layer. The influence of variable amplitude stress needs to be considered when predicting the damage of SFCBs. According to the fatigue damage accumulation criterion, the damage of the material before fatigue loading is D=0%, and its fatigue damage is D=100% when fatigue failure occurs. Therefore, in the Batsoulas curve, the SN curve and the coordinate axis together constitute the envelope of the fatigue damage value, where the SN curve represents the equal damage line when D=100%, and the two vertical coordinate axes represent the equal damage line when D=0%. The curve in between represents the equal damage line with damage degree in the range of 0~100%. According to the physical meaning of fatigue damage, the equal damage lines of different damage values ​​do not intersect. Therefore, in the Batsoulas model, the equal damage lines of fatigue damage in the range of 0~100% are represented by non-intersecting hyperbolas, such as Figure 1 shown.

[0062] Depend on Figure 1 It can be seen that any point on the equal damage line is exactly the vertex of its equivalent rectangle in the coordinate diagram. For example, point F on the equal damage line D1 is the vertex of the rectangle (OEFM), and point K on the SN curve is the vertex of the rectangle (OEKL). Therefore, the area ratio of the equal damage line D1 to the equivalent rectangle of D = 100% (SN curve) can be used to represent the fatigue damage when cyclically loaded n times at a given stress level, as shown in formula (3).

[0063] (3)

[0064] According to the concept of equal damage line, points with different stress levels on the same equal damage line have the same damage. Figure 1 When point F in moves to point G, we have:

[0065] (4)

[0066] Substituting formula (2) into formula (4), we can obtain:

[0067] (5)

[0068] Where n1 * is the equivalent number of cycles when the damage generated under the loading history (σ1, n1) is converted to the stress level σ2. Similarly, when Figure 1 When point H moves to point I, the equivalent cycle number n2 * for:

[0069] (6)

[0070] Where n2 * The damage generated under the loading history (σ1, n1) and (σ2, n2) is converted into the equivalent number of cycles when the stress level is σ3. Similarly, the cumulative damage calculation formula under multi-level loading is:

[0071] (7a)

[0072] (7b)

[0073] Where, subscript i represents the i-th cycle block in the loading process, N fi represents the fatigue life of the material under the i-th loading spectrum.

[0074] Phenomenological definition of injury:

[0075] The macroscopic phenomena of materials in a fatigue-damaged state are primarily fatigue stiffness degradation and a decrease in residual strength. Fatigue stiffness degradation reflects the attenuation of the material's overall load-bearing capacity due to accumulated internal damage under cyclic loading, and can be monitored in real time through non-destructive dynamic testing. Residual strength, on the other hand, represents the ultimate stress that a material can still withstand in a damaged state. However, its acquisition requires destructive loading tests (such as stretching to fracture), resulting in high testing costs, low efficiency, and the inability to reuse specimens. To address these issues, this embodiment proposes macroscopic definition parameters for fatigue damage in SFCBs based on the concept of isodamage lines.

[0076] Damage definition based on fatigue stiffness degradation:

[0077] In this embodiment, the fatigue modulus shown in formula (8) is used to calculate the dynamic stiffness of SFCBs.

[0078] (8)

[0079] Where n is the number of cycles; F is the fatigue modulus, F(n) is a method of expressing dynamic stiffness; σ max and ε max are the peak fatigue load and the corresponding strain value respectively; ε p is fatigue creep; E c is the elastic modulus under static loading. According to formula (8), when n=0, fatigue creep ε p (0)=0, F(0)=E c ; When n=N f When F(N f )=F u The fatigue damage defined by fatigue stiffness degradation is as follows:

[0080] (9)

[0081] Where D F It is fatigue damage defined based on stiffness degradation. Fatigue stiffness degradation under cyclic loading shows a typical three-stage trend. The first stage is a rapid development stage, which originates from the activation of internal defects in the material and the rapid accumulation of damage. For the steel component, dislocation slip induces microcracks to nucleate at grain boundaries / inclusions. For the FRP layer, it manifests as the nucleation and expansion of matrix microcracks and debonding of the fiber / matrix interface. Experiments show that this stage accounts for about 10% of the fatigue life. After that, it enters the second stage of constant rate development. After the initial damage reconstruction, the stiffness degradation rate stabilizes. This stage accounts for about 90% of the fatigue life. Finally, there is the destruction stage. When the damage accumulation exceeds the critical threshold, the stiffness decays drastically and the failure process presents sudden characteristics. The duration of this stage is usually less than 5% of the total life, often only a few hundred cycles.

[0082] Due to the unpredictability of the fatigue stiffness failure stage and its short duration, when establishing the fatigue stiffness prediction model, this embodiment focuses on the first and second stages, which account for the vast majority of the fatigue life. Therefore, the fatigue stiffness prediction model is established by combining formulas (3) and (9).

[0083] (10)

[0084] Damage definition based on residual strength:

[0085] According to damage mechanics theory, the decrease in the residual strength of the material can be attributed to the continuous decrease in the effective bearing surface area of ​​the cross section caused by the increase in fatigue damage. The residual strength after a certain number of cycles n can be expressed as:

[0086] (11)

[0087] Where σ R is the residual strength; σu is the static tensile strength; A(n) is the actual cross-sectional area after fatigue damage occurs. From the perspective of damage mechanics, under the condition that the fatigue load P remains unchanged, as the effective load-bearing cross-section continues to decrease, the effective stress σ * Gradually increases. Nominal stress σ and effective stress σ under fatigue load P * Calculated by formula (12) and formula (13) respectively:

[0088] (12)

[0089] (13)

[0090] Where A0 is the cross-sectional area when initially undamaged. Combining formulas (11), (12), and (13), the mathematical relationship between residual strength and effective stress can be derived as follows:

[0091] (14)

[0092] From formula (14), we can see that due to σ u and A0 are material property constants, so the residual strength under constant nominal stress σ is inversely proportional to the nominal stress. Then fatigue damage based on residual strength can be defined as follows:

[0093] (15)

[0094] Where D S A is fatigue damage defined based on residual strength; u is the critical cross-sectional area at fatigue failure. Substituting formulas (12) and (13) into formula (15), we can also obtain fatigue damage defined by effective stress:

[0095] (16)

[0096] Where σ u * is the critical value of effective stress failure. The evolution law of effective stress during the cycle is obtained. Different from the three-stage development trend of fatigue stiffness, effective stress increases slowly in the early stage of fatigue loading and then rises suddenly when fatigue failure is approaching. When fatigue damage D=0, effective stress σ * =σ; when fatigue damage D=1, the material undergoes fatigue failure, and the effective stress σ * =σ u * .

[0097] Given that the time-varying characteristics of the effective load-bearing section of a material during fatigue damage are difficult to measure directly through experiments, the traditional effective stress theory faces a significant application bottleneck. To this end, a fatigue effective stress definition method based on the residual life inversion mechanism is proposed: assuming that after a material is subjected to fatigue stress σ for n cycles, its effective stress σ * Corresponding to the remaining life N f -n corresponds to the critical stress value on the SN curve.

[0098] Therefore, combined with the SN curve equation of formula (2), the relationship between effective stress and remaining fatigue life can be expressed as:

[0099] (17)

[0100] By simple transformation, formula (17) can be transformed into:

[0101]

[0102] The curve represented by formula (18) is when n=0, σ * =σ, satisfying the initial condition of effective stress.

[0103] Binary failure criterion for fatigue damage:

[0104] like Figure 2 The figure shows a schematic diagram of the fatigue damage evolution process proposed in this embodiment based on a phenomenological perspective (fatigue stiffness degradation and residual strength attenuation). The curve reveals the co-evolution law of the two damage variables:

[0105] (1) Damage variable D F : The damage index based on fatigue stiffness degradation can effectively characterize the nonlinear accumulation process of the material in the early stage of fatigue loading (stage I and stage II);

[0106] (2) Damage variable D S : A damage index based on residual strength decay, suitable for describing the damage mutation characteristics caused by failure at the end of fatigue loading (stage III), and its evolution rate is significantly correlated with the change in effective stress.

[0107] Therefore, according to formula (9) and formula (16), this embodiment provides a binary damage criterion for fatigue failure of materials:

[0108] (19)

[0109] The physical meaning of this criterion is that when the fatigue cumulative damage D F or D SMaterial failure is determined when any variable reaches a critical threshold (D ≥ 1). Damage based on stiffness degradation provides the basis for stress redistribution between SFCB component materials, while the application of effective stress can capture sudden failures caused by local stress concentration. Together, these two factors form a progressive representation of fatigue damage.

[0110] The failure criterion of the material can be determined by the fatigue fracture strain and stress state. According to the strain equivalence assumption, the critical fracture strain and the nominal fatigue stress σ max , critical fracture strength σ u * The following equality relationship exists:

[0111] (20)

[0112] Where, ε u * is the critical fracture strain. Therefore, we only need to know the critical fracture strength σ u * The critical value of fatigue stiffness under a given stress level can be obtained:

[0113] (twenty one)

[0114] According to experimental observations, when the material damage reaches the critical threshold, it will be transiently destroyed, and the destruction duration is less than 5% of the total life. Therefore, this embodiment conservatively sets the failure condition to be the remaining life N f -n=5%N f and determine the critical fracture strength σ by formula (18) u * .

[0115] Stress redistribution mechanism of components in SFCBs:

[0116] Dynamic description of stress redistribution:

[0117] According to the strain synergy hypothesis, the following equation exists between SFCBs, their inner steel bars, and FRP layers:

[0118] (twenty two)

[0119] Where n is the number of cycles; ε max is the strain of the material under peak fatigue load; σ b,max , σ s,max and σ f,max are the stresses of SFCBs, inner steel bars and FRP layers under peak fatigue load; F b 、F s and F fare the fatigue moduli of SFCBs, inner core steel bars, and FRP layers, respectively. Therefore, according to the principle of equivalent stiffness distribution, the fatigue stresses of the inner core steel bars and FRP layers can be calculated using Equations (23a) and (23b).

[0120] (23a)

[0121] (23b)

[0122] Where A b 、A s and A f are the cross-sectional areas of SFCBs, inner core steel bars and FRP layers, respectively.

[0123] According to formulas (7a), (7b), (8) and (9), the dynamic fatigue modulus F of the inner core steel bar and FRP layer can be calculated respectively. s (n) and F f (n), and then substitute it into formulas (23a) and (23b) to calculate the fatigue stress borne by the component materials (core steel bar and FRP layer).

[0124] According to Equations (23a) and (23b), as the fatigue stiffness of the core steel and FRP layers gradually degrades, their stresses exhibit a continuous dynamic redistribution in each fatigue cycle. Specifically, the fatigue load shared by the two undergoes a non-steady-state evolution with the number of cycles. This stress evolution mechanism is equivalent to the core steel and FRP layers being subjected to multi-stage variable-amplitude fatigue loads with a time-dependent nature. Therefore, if the traditional prediction model based on the constant-amplitude fatigue stress assumption (i.e., assuming that the stress levels of the core steel and FRP layers are constant) is used to evaluate the fatigue life of SFCBs, significant errors will result from ignoring the stress redistribution effect. To accurately simulate this process, theoretically, it is necessary to calculate the stress redistribution state of each fatigue cycle one by one. However, this method requires a lot of computational time and carries the risk of error accumulation, making it unacceptable.

[0125] To address the above issues, this embodiment proposes a loop skipping method:

[0126] like Figure 3 As shown in the figure, the Cycle Jump Method divides the fatigue process into several groups of fatigue cycles. , by calculating each set of cycles one by one The damage development of the entire fatigue process can be obtained by calculating the damage within the fatigue period. This method will greatly reduce the calculation cost. The calculation accuracy of the cyclic jump method is better than related to, increase It can improve the calculation speed, but it will introduce errors due to the linear extrapolation assumption; reduce It can improve the accuracy, but it will significantly increase the consumption of computing resources. Therefore, it is necessary to choose a suitable , to balance the relationship between computational efficiency and computational accuracy.

[0127] To achieve a dynamic balance between computational efficiency and accuracy during fatigue damage evolution, this example proposes a damage mechanism-driven, piecewise adaptive step-size algorithm. Its core concept is to adapt to the staged nature of fatigue damage rates, using a small step size during the active damage phase to capture nonlinear cumulative effects and a larger step size during the steady-state phase to improve computational efficiency. The specific implementation strategy is as follows: In the range of 0 to 100 fatigue cycles, a fine step size Δn=10 is used to accurately characterize the stress redistribution effect caused by the initiation of initial microcracks; in the range of 100 to 1000 fatigue cycles, the step size is adjusted to Δn=100 to balance the nonlinear attenuation of damage accumulation and computational resource consumption; in the range of 1000 to 10,000 fatigue cycles, the step size is expanded to Δn=1000 to match the steady-state evolution rate of material stiffness degradation; in the range of 10,000 to 2,000,000 fatigue cycles, a step size Δn=10,000 is used, and the step size Δn is increased to 100,000 after 2,000,000 cycles to avoid over-iteration. Before using the cycle skipping method, the following assumptions must be met:

[0128] (1) Within the cycle step Δn, the component materials of SFCBs (core steel and FRP layer) maintain a constant stress level. The fatigue creep of the core steel and FRP layer increases with the increase of cycles, resulting in the degradation of their respective fatigue stiffness.

[0129] (2) After each jump, the fatigue stiffness of the component materials is updated according to the current damage state, and the distribution of the component stress state is completed before the next cyclic jump.

[0130] (3) The principle of irreversible damage must be satisfied during the cycle jump process, that is, the damage increases monotonically with the increase of cycles (dD / dn>0).

[0131] Based on the above assumptions, the fatigue life prediction process of SFCBs is as follows: Figure 4 The specific steps are as follows:

[0132] Step 1: Initialize parameters.

[0133] Input material properties (E s , E f , A s , A f , A b ), load conditions and initial state (σ, n=0, i=1);

[0134] Among them, E s and E fare the initial elastic moduli of the inner core steel bar and FRP layer, respectively, which are used to calculate their initial fatigue moduli.

[0135] Step 2: Stress distribution calculation.

[0136] According to the current material properties, the fatigue stresses borne by the inner core steel bar and the FRP layer are calculated using formulas (23a) and (23b) respectively.

[0137] Step 3: Fatigue life mapping and jumping.

[0138] Substituting the fatigue stress level of the component materials into the SN curve relationship shown in formula (2), the theoretical fatigue life N of the inner core steel bar and FRP layer is obtained. f,steel and N f,FRP , after which a loop jump Δn is performed.

[0139] Step 4: Damage criterion check.

[0140] The cumulative fatigue damage of the component materials is calculated based on formulas (9) and (16). At this time, the damage states of the inner core steel bar and FRP layer are obtained respectively: D steel ={D F,steel , D S,steel} and D FRP ={D F,FRP , D S,FRP The subscripts steel and FRP are used to distinguish the inner steel bar from the FRP. Failure is determined using the binary failure criterion formula (19).

[0141] (1) If D steel ≥1 and D FRP < 1, it is determined that the SFCBs have suffered fatigue failure due to the fracture of the inner core steel bars;

[0142] (2) If D FRP ≥1 and D steel < 1, it is determined that the SFCBs suffered fatigue failure due to the rupture of the FRP layer.

[0143] (3) If D FRP ≥1 and D steel ≥ 1, it is determined that the SFCBs have suffered fatigue failure due to the simultaneous rupture of the inner core steel bar and the FRP layer.

[0144] The above D steel and D FRP are the cumulative damage of the inner core steel bar and FRP layer respectively. When the SFCBs fail, the calculation is terminated and the calculated fatigue life M of the SFCBs is output. i =∑Δn, where ∑ is the summation symbol. Otherwise, proceed to step 5.

[0145] Step 5: Parameter update and iteration.

[0146] After the cycle jump, the properties and loading state of the component materials are changed. Therefore, before entering the next cycle jump, the dynamic stiffness of the material can be calculated by formula (10); the equivalent cycle number n * Calculated by formula (7a), the calculated value needs to be rounded down. After updating the material properties, the program status becomes: n=n * +Δn, i=i+1, then return to step 2.

[0147] Model Validation:

[0148] This example takes the equal damage line theory constructed by Batsoulas formula as the core, considers the stress redistribution mechanism between component materials, and proposes a non-parametric damage model for predicting the fatigue life of SFCBs based on the SN curves of the component materials. Since the prediction accuracy of the model directly depends on the calibration of the SN curves of the component materials, this example first verifies the applicability of Batsoulas formula to steel bars and FRP based on fatigue test data, and compares it with Basquin formula. ] Then, the fatigue life and dynamic stiffness degradation curves of three SFCBs specimens (SG12, SG16 and SG20) were predicted to systematically verify the prediction capability and physical rationality of the model.

[0149] Calibration of SN curve parameters of component materials:

[0150] To establish the input parameters for the fatigue life prediction model for SFCBs, this example calibrates the SN curves of the component materials (core steel and FRP layer) based on tensile fatigue test data. Here, SFCBs consisting of a plain round steel core and a glass fiber reinforced plastic (GFRP) protective layer are used as an example. The calibration specimens are selected to have completely consistent properties with the SFCB component, including the core steel reference specimen S8 and the GFRP layer reference specimen G12. The fatigue test data in Table 1 are fitted using the Basquin formula and the Batsoulas formula, respectively, to obtain the SN curve formulas for the core steel and GFRP layer, as shown in Equations (24) and (25), respectively.

[0151] Table 1

[0152]

[0153] The first item of the specimen number is the reinforcement type, and the second item is the stress level of fatigue loading. For example, "SG12-F96" means that a tensile fatigue test with a stress level of 0.96 is carried out on SFCBs with a model of SG12; max is the maximum fatigue load; σ max is the maximum fatigue stress; S is the stress level; σ ais the stress amplitude, σ a =0.5(σ max -σ min ), σ min is the minimum fatigue stress; N f is the fatigue life; in the failure mode, S means steel bar fracture, F means GFRP fracture, SF then FF means the steel bar first fatigue fractures and then the GFRP is broken, and RO means run out.

[0154] (twenty four)

[0155] (25)

[0156] Where S is the stress level. For the inner core steel bar, S=σ max / f y , for the FRP layer, S=σ max / f u .f y and f u Represent the yield strength of the inner steel bar and the tensile strength of FRP respectively. According to the comparison between the fitting curves of the two models and the experimental data, it can be seen that both the Batsoulas formula and the Basquin formula can fit the experimental data well, and the fitting coefficients R 2 The difference is less than 5%, indicating that both can effectively characterize the life attenuation law of the material in the medium and high cycle fatigue region. However, in the ultra-high cycle fatigue region (N f ≥2×10 6 ) The linear characteristics of the Basquin formula prediction curve cause the prediction curve to continue to decline, while the Batsoulas formula uses the hyperbolic characteristics to make the curve converge to infinity at S0=0.17 (inner core steel bar) or S0=0.09 (FRP layer), thereby accurately characterizing the fatigue limit of the material.

[0157] Based on the above analysis, the Batsoulas formula directly reflects the fatigue limit stress level of the material. Its asymptotic convergence avoids the very high-cycle prediction distortion associated with the Basquin formula, and it has clear physical significance. Therefore, this embodiment uses the Batsoulas formula as the model for calibrating the SN curves of component materials.

[0158] Fatigue life prediction:

[0159] To verify the proposed model's ability to predict the fatigue life of SFCBs, this example compares the model's calculation results with those of three SFCB specimens (SG12, SG16, and SG20). The specimen parameters are: a constant inner steel bar diameter (8 mm), and FRP layer thicknesses of 2 mm, 4 mm, and 6 mm, respectively. The model's prediction results are compared with the test data, and an equal damage line of 95% inner steel bar damage is introduced as the engineering safety threshold. To evaluate the model's prediction performance, this example uses the mean absolute error (MAE) to reflect the average deviation between the predicted and experimental values, and the root mean square error (RMSE) to measure the degree of dispersion of the prediction error, as shown in Equations (26) and (27), respectively.

[0160] (26)

[0161] (27)

[0162] Where m is the number of specimens; N pred,i and N exp,i are the predicted and experimental values ​​of fatigue life, respectively. However, both MAE and RMSE metrics are directly dependent on the magnitude of the data. For example, if the predicted life span is in the millions, then the MAE and RMSE values ​​will naturally be larger than those for data in the thousands, making it difficult to intuitively and uniformly describe the accuracy of the model. Therefore, this embodiment performs normalization processing, as shown in Formulas (28) and (29).

[0163] (28)

[0164] (29)

[0165] Where R MAE and R RMSE is the statistical index after normalization; N exp,ave It is the average value of the fatigue life test values ​​of the same group.

[0166] The results show that the model prediction curve in the double logarithmic coordinate system can well reflect the fatigue life development trend of different SFCBs and is in good agreement with the experimental fitting curve. The normalized statistical index R MAE =0.53 (overall), R RMSE =0.54 (overall). Among them, for specimen SG12, R MAE =0.67, R RMSE =0.69; for specimen SG16, R MAE =0.50, R RMSE =0.50; for specimen SG20, R MAE =0.41, R RMSE=0.44. As can be seen, the model prediction accuracy is higher for specimens with thicker FRP layers (SG16 and SG20). The predicted values ​​deviate from the experimental values ​​by about half an order of magnitude on average. This deviation may be due to the dispersion of fatigue data and the limitations of experimental data. The discreteness of material performance in the high-cycle fatigue region results in the model calibration parameters not fully covering extreme working conditions, and the lack of specimens in the ultra-high-cycle fatigue region (N f ≥ 2×10 6 ) of the test sample, affecting the extrapolation accuracy.

[0167] Furthermore, based on the damage evolution data output by the model, this example defines an iso-damage curve with a 95% damage level for the inner core reinforcement, which is used to quickly assess the fatigue life threshold of SFCBs. To quantitatively assess the conservatism of this iso-damage curve, this example uses the statistical index value shown in Equation (30) to quantify the conservative deviation of the predicted value.

[0168] (30)

[0169] Where n 保守 The number of specimens within the same group whose predicted values ​​were lower than the experimental values ​​was shown. The results show that the isodamage contour provides conservative life estimates for most specimens, with a conservative rate (the percentage of specimens with conservative predictions compared to the total number of specimens) reaching 78%. The overall conservative bias (CBP) is 42%, indicating that the isodamage contour, on average, is 42% lower than the conservative sample life. Specimens with overestimated predictions are distributed in the low-stress level region, primarily due to the high dispersion of fatigue data in this low-life region. Despite this dispersion, the model still captures the dominant trends in the fatigue life of SFCBs, demonstrating its ability to provide a reliable theoretical basis for engineering safety design.

[0170] Verification of fatigue stiffness degradation and stress redistribution mechanism:

[0171] To reveal the model's ability to characterize the dynamic evolution of internal damage in composite reinforcement, this example predicts the stiffness degradation curve of SFCBs specimens based on the proposed non-parametric damage model and compares it with the experimental monitoring data. The comparison shows that the model prediction curve reproduces the typical two-stage characteristics of stiffness degradation: initial loading (n≤10%N f ) Fast decay and late stage (n >10%N f ) develops steadily, which is consistent with the experimental data trend. However, the model overestimates the stiffness degradation of SFCBs (average deviation 20%), which is mainly due to the critical fracture strength σ u * The model relies heavily on the exponential rise in effective stress before failure, leading to an overestimation of the fatigue stiffness failure threshold. It is recommended that the failure criterion parameters be calibrated to improve critical accuracy. Despite this, the model still effectively captures the governing laws of stiffness degradation, providing a reference for preliminary engineering assessments.

[0172] In order to quantitatively reveal the load transfer mechanism between the component materials (core steel bars and FRP layers), this embodiment outputs the dynamic stress level based on the non-parametric damage model. The analysis shows that the stress level of the core steel bars (S=σ s / f y ) At the initial stage of loading (n≤10%N f ) decreases significantly and tends to be stable in the later period. The stress level of FRP (S=σ f / f u ) remains stable or increases slightly during the entire loading process. The starting point of the ultra-high cycle fatigue zone is set at 2×10 6 The calculation of formula (24) shows that in 2×10 6 At the first cycle, the fatigue strength of the inner steel bar reached S = 0.74, while that of the FRP layer reached S = 0.21. Clearly, stress transfer within the inner steel bar slowed its local damage rate, thereby extending its fatigue life. However, because its stress level remained above its fatigue strength, it failed before the FRP layer. Although the stress in the FRP layer increased, it never exceeded its fatigue strength, primarily playing a cooperating role in load-bearing rather than being the primary driver of failure, consistent with previous assumptions.

[0173] A reduction in the initial stress level will increase the degree of stress redistribution. Taking SG16 as an example, when the initial stress level decreases from S=1 to S=0.9, the stress level of the inner core steel bar decreases from approximately 0.02 to 0.05. This is attributed to the increased fatigue life of the material at low stress levels, which in turn triggers more stress redistribution. In addition, increasing the thickness of the FRP layer also exacerbates stress redistribution. Under the condition of an initial stress level of S=0.9, the stress level of the inner core steel bar decreases from approximately 0.03 to 0.07 when the FRP layer thickness increases from 2mm to 6mm (i.e., from SG12 to SG20). This finding proves that increasing the thickness of the FRP layer can effectively share the load and delay the accumulation of damage to the steel bar.

[0174] The above analysis shows that the nonparametric damage model proposed in this example accurately reveals the stress redistribution mechanism of SFCBs under different stress levels and FRP layer thicknesses, and quantifies the sensitivity of the initial stress level and FRP thickness to damage evolution and life prediction, providing a theoretical basis for the fatigue optimization design of SFCBs.

[0175] Comparison with the fitted model:

[0176] To comprehensively evaluate the performance of the nonparametric damage model proposed in this example for fatigue life prediction, a detailed comparative analysis was conducted with the traditional multiparametric damage model. The fatigue life of SFCBs specimens (SG12, SG16, and SG20) was predicted using both models. The accuracy of the prediction results was quantitatively evaluated using the statistical indicators defined in Equations (28) and (29).

[0177] By comparing the prediction results of the two models, the prediction results of the multi-parameter model are generally conservative. After calculation, its conservative rate is 64%. In contrast, the non-parametric damage model proposed in this embodiment is more conservative in the medium and high cycle fatigue region (N f ≤10 6 ) is conservative in life prediction; while in the ultra-high cycle fatigue area (N f >10 6 ), its life predictions tend to be overestimated, with a conservative rate of 39%. Furthermore, the overall prediction accuracy of the non-parametric damage model is comparable to that of the multi-parametric damage model. However, the non-parametric model exhibited prediction errors exceeding 50% for some specimens (such as SG12-F76). Further investigation revealed that this was primarily due to missing data from run-out specimens (i.e., undamaged specimens) during the ultra-high cycle period, which resulted in an ineffective extrapolation process.

[0178] In engineering practice, to meet design requirements for bearing capacity and durability, SFCBs of varying specifications (e.g., varying inner core diameters and FRP layer thicknesses) are often required. The prediction accuracy of multi-parameter damage models is highly dependent on fitting test data. Using this method to predict the fatigue life of each SFCB specification requires separate fatigue testing, which is costly and labor-intensive. While maintaining comparable overall prediction accuracy, the non-parametric damage model of this embodiment offers significant advantages in reducing the amount of testing required. According to the current international standard ISO 10406, determining the SN curve for a reinforcement requires at least three different stress levels, with no fewer than three replicate specimens at each stress level. Therefore, constructing a damage model for a reinforcement requires at least nine specimens, assuming a one-month testing period. Using traditional methods, designing five SFCB types would require five groups of 45 specimens, with a cumulative testing period of five months. However, using the method of this embodiment, designing five SFCB types requires only two groups of 18 specimens (testing and calibrating only the inner core steel and FRP layer), shortening the testing period to two months. After calculation, this method saves 60% in the number of tests and cycles compared with traditional methods, and is more economical.

[0179] In summary, the overall prediction performance of the nonparametric model proposed in this example is comparable to that of traditional methods, with low experimental dependence (only two-component tests are required). It demonstrates higher engineering economy in fatigue design and lays a solid theoretical foundation for the widespread application of SFCBs in practical engineering.

[0180] Advancedness of the model:

[0181] This example, leveraging the concept of equal damage curves, proposes a nonparametric damage model for predicting SFCBs based on the SN curves of the component materials (core steel and FRP layer). The key advantage of this model is that it considers only the SN curves of the component materials, while maintaining prediction accuracy comparable to traditional models, eliminating the need for additional experimental fitting parameters. Compared to traditional fitting-based multi-parameter models, this model significantly reduces testing costs and improves the feasibility of engineering applications.

[0182] In addition, by adjusting the parameters of the cross-sectional ratio of the component materials (As / Af), the model can be adaptively converted into a damage model of pure steel bars or pure FRP bars, thereby achieving a unified characterization of the fatigue behavior of composite bars and single bars, expanding its scope of application.

[0183] Existing nonlinear fatigue damage models based on isodamage lines are limited to calculating the remaining life and cannot obtain the actual damage values ​​at different cycle stages. In contrast, this model establishes a quantitative mapping relationship between stiffness degradation and damage accumulation, while also considering the interaction between component materials, providing a direct quantitative basis for evaluating structural fatigue performance.

[0184] Existing nonlinear fatigue damage models based on equal damage lines are mostly limited to residual life calculations and cannot capture the actual damage evolution process at different cycle stages. The model constructed in this embodiment overcomes this limitation, establishing a quantitative mapping relationship between stiffness degradation and damage accumulation, while fully considering the interaction between the inner core steel bar and the FRP layer, making the fatigue damage calculation more refined and physically reasonable. This model can not only effectively predict the fatigue life of various types of reinforcement, but also depict the entire process of material damage evolution under multi-level loading conditions, providing a direct quantitative basis for structural fatigue performance assessment.

[0185] In summary, fatigue life prediction for steel-FRP composite bars (SFCBs) is a key issue in promoting their large-scale application in engineering. However, traditional methods often rely on large amounts of experimental data and struggle to accurately characterize dynamic stress redistribution effects. To address this challenge, this paper proposes a nonparametric fatigue damage prediction method for SFCB fatigue life based on the SN curves of the component materials. Through theoretical modeling and experimental verification, the damage evolution patterns and the mechanisms controlling the life of SFCBs are systematically revealed. The main conclusions are as follows:

[0186] (1) Based on the Batsoulas formula for the reasonable prediction of ultra-high cycle fatigue life and the concept of equal damage lines, a material cumulative damage criterion applicable to multi-stage loading conditions was established. Through the phenomenological definition of damage, a quantitative mapping relationship between fatigue stiffness degradation and cumulative damage was constructed. In addition, the model adopts an improved effective stress definition and proposes a binary criterion for fatigue failure of SFCBs. Through the principle of equivalent stiffness distribution, the stress redistribution mechanism between the component materials is reasonably considered. On this basis, a non-parametric damage model is constructed to predict the fatigue life of SFCBs based solely on the SN curves of the component materials.

[0187] (2) In the calibration process of the SN curve of the component material, the Batsoulas formula can well characterize the N f ≤2×10 6 The fatigue life is within the specified range. The core steel bar parameters are calibrated as S0 = 0.17 (relative yield strength) and C = 4.08, and the FRP layer parameters are S0 = 0.09 (relative ultimate strength) and C = 2.51. However, due to the lack of supporting ultra-high-cycle test data, the proposed non-parametric model has certain errors in predicting stiffness degradation.

[0188] (3) Verification of the dynamic stiffness degradation and stress redistribution mechanisms revealed a complex interaction between the core steel and the FRP layer. The proposed model revealed the influence of load transfer on damage accumulation. The results also showed that increasing the thickness of the FRP layer significantly enhanced the stress redistribution effect, thereby slowing the damage accumulation rate of the core steel. This provides a theoretical basis for optimizing the cross-sectional design of SFCBs (such as adjusting the proportion of the FRP layer) to achieve a comprehensive improvement in material performance.

[0189] (4) Compared with traditional multi-parameter models, the non-parametric damage model proposed in this example only requires the calibration of the SN curves of the component materials, eliminating the need for repeated fatigue testing for SFCBs of different specifications. Case analysis shows that when predicting five SFCB designs, the model can reduce the number of tests by 60% (the number of specimens is reduced from 45 to 18) and shorten the test cycle by 60% (from 5 months to 2 months). Normalized statistical indicators show that the model is comparable to traditional methods in terms of prediction accuracy, but has significant advantages in terms of test cost and engineering efficiency.

[0190] Example 2

[0191] The present invention also provides a non-parametric prediction system for fatigue life of steel-FRP composite bars, which is used to implement the method described above, comprising:

[0192] A parameter initialization module is used to initialize the component material properties, load conditions, and initial states of the steel-FRP composite reinforcement, where the component materials include the inner core steel bar and the FRP layer;

[0193] Fatigue stress calculation module, used to calculate the fatigue stress of component materials separately according to the principle of equivalent stiffness distribution;

[0194] Fatigue life calculation module, used to obtain the theoretical fatigue life of component materials based on the fatigue stress of component materials and the SN curve relationship, and perform cycle skipping;

[0195] Fatigue damage calculation module, used to calculate the cumulative fatigue damage of component materials based on the results after cyclic jumping;

[0196] The cyclic update module is used to cyclically update the dynamic stiffness and equivalent cycle number of the component materials until the cumulative fatigue damage of any component material reaches a critical threshold, then the component material is judged to be failed and the fatigue life is output.

[0197] Example 3

[0198] This embodiment further discloses a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first embodiment.

[0199] Example 4

[0200] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first embodiment are implemented.

[0201] Example 5

[0202] This embodiment further discloses a computer program product, including a computer program, which implements the steps of the method described in the first embodiment when executed by a processor.

[0203] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A nonparametric prediction method for fatigue life of steel-FRP composite bars, characterized in that: The following steps are involved: Initialize the component material properties, loading conditions, and initial states of the steel-FRP composite reinforcement, where the component materials include the core steel bar and the FRP layer; According to the principle of equivalent stiffness distribution, the fatigue stress of component materials is calculated separately; According to the fatigue stress of the component materials, combined with the improved SN curve relationship, the theoretical fatigue life of the component materials is obtained respectively, and the cycle jump is performed; Based on the results after cyclic skipping, the cumulative fatigue damage of the component materials is calculated separately; The dynamic stiffness and equivalent cycle number of the component materials are updated cyclically until the cumulative fatigue damage of any component material reaches a critical threshold. The component material is then judged to be failed and the fatigue life is output.

2. The method according to claim 1, characterized in that The process of initializing the component material properties, load conditions and initial state of the steel-FRP composite reinforcement is specifically as follows: Initialize the component material properties based on the input elastic modulus and cross-sectional area of ​​the inner core steel bar, the elastic modulus and cross-sectional area of ​​the FRP layer, and the overall cross-sectional area of ​​the steel-FRP composite bar; initialize the load conditions based on the input stress level; Set the initial loop count to zero and the loop sequence number to one to complete the initial state setting.

3. The method according to claim 1, characterized in that The process of calculating the fatigue stress of component materials respectively according to the equivalent stiffness distribution principle is specifically as follows: Based on the component material properties of the inner core steel bar and FRP layer, the strain of both under peak fatigue load is calculated. The fatigue modulus of the inner core steel bar and FRP layer is calculated based on the strain. Based on the principle of equivalent stiffness distribution, the fatigue stress borne by the inner core steel bar and FRP layer is calculated separately in combination with their fatigue modulus and cross-sectional area.

4. The method according to claim 1, wherein The process of calculating the cumulative fatigue damage of the component materials based on the results after the cycle jump is specifically as follows: Based on the results after cyclic jumping, the degree of fatigue stiffness degradation of the inner core steel bar and FRP layer is calculated; the fatigue stiffness degradation degree includes a damage index based on fatigue stiffness degradation and a damage index based on residual strength decay; the damage index based on fatigue stiffness degradation is used to characterize the nonlinear cumulative damage of the component material in the early stage of fatigue loading, and the damage index based on residual strength decay is used to describe the damage mutation at the end of fatigue loading.

5. The method according to claim 1, wherein The process of cyclically updating the dynamic stiffness and equivalent cycle number of the component materials until the cumulative fatigue damage of any component material reaches a critical threshold, then determining that the component material has failed and outputting the fatigue life is specifically as follows: Based on the current cycle status, the dynamic stiffness and equivalent cycle number of the inner core steel bar and FRP layer are updated; the program status, including the cycle number and cycle sequence, is updated; and it is determined whether the accumulated fatigue damage of the inner core steel bar or FRP layer has reached the critical threshold. If so, the cycle is terminated and the fatigue life is output.

6. A non-parametric prediction system for fatigue life of steel-FRP composite bars, characterized in that: The method for implementing any one of claims 1 to 5 comprises: A parameter initialization module is used to initialize the component material properties, load conditions, and initial states of the steel-FRP composite reinforcement, where the component materials include the inner core steel bar and the FRP layer; Fatigue stress calculation module, used to calculate the fatigue stress of component materials separately according to the principle of equivalent stiffness distribution; Fatigue life calculation module, used to obtain the theoretical fatigue life of component materials based on the fatigue stress of component materials and combine with the improved SN curve relationship, and perform cycle skipping; Fatigue damage calculation module, used to calculate the cumulative fatigue damage of component materials based on the results after cyclic jumping; The cyclic update module is used to cyclically update the dynamic stiffness and equivalent cycle number of the component materials until the cumulative fatigue damage of any component material reaches a critical threshold, then the component material is judged to be failed and the fatigue life is output.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory, wherein: The processor executes the computer program to implement the steps of the method according to 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, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images