Fatigue damage calculation method based on dynamic weight correction

Abstract

The invention relates to a fatigue damage calculation method based on dynamic weight correction, and the method comprises the following steps: S1, collecting stress-level load test data, and recording a load application sequence; s2, constructing a dynamic weight coefficient model formula; s3, constructing a fatigue damage accumulation formula; and S4, calculating the damage degree D under different load sequences according to the step S3, predicting the fatigue life of the material, when D is not less than 1, considering that the material has fatigue failure, and when D is less than 1, considering that the material does not have failure. The method has the advantages that an accumulated fatigue calculation formula is improved on the basis of a traditional linear fatigue accumulation algorithm, the nonlinear damage acceleration effect under high stress is accurately reflected through a dynamic coefficient, and errors caused by neglecting the load sequence and the stress amplitude difference of a traditional linear model are avoided; meanwhile, the historical load influence can be updated in real time, and it is ensured that a calculation result is closer to an actual result.

Description

Technical Field

[0001] This invention belongs to the technical field of improving linear cumulative fatigue damage, and specifically relates to a fatigue damage calculation method with dynamic weight correction. Background Technology

[0002] The linear cumulative fatigue algorithm usually refers to the Miner criterion, i.e. n i Let N be the number of cycles at stress level i. i This represents the total number of cycles required for the material to reach fatigue failure at a given stress level, where D is the degree of damage. Fatigue failure is considered to have occurred when D = 1. In practical calculations, it is first necessary to determine the (SN) curves (stress-cycle count curves) of the material at different stress levels. Then, based on the actual load spectrum, the number of cycles at different stress levels is statistically analyzed and substituted into the formula for calculation.

[0003] Fatigue failure is predicted by linearly superimposing the damage proportions under different stress levels. Due to its simple calculation and ease of engineering application, it is widely used in life design in fields such as metal testing. However, this model has significant drawbacks. It linearly superimposes the damage from each stress cycle, ignoring the nonlinear characteristics of actual fatigue damage. Furthermore, traditional linear cumulative fatigue algorithms assume that fatigue damage is only related to the stress level and the number of cycles, and not to the order of load application. This leads to a situation where, if a high-stress cycle is applied first (causing crack initiation), the actual rate of damage in subsequent low-stress cycles may be accelerated by crack propagation; conversely, if a low-stress cycle occurs first (without forming a significant crack), the sustained damage from subsequent high-stress cycles may be lower due to the lack of pre-damage to the material. However, the Miner criterion still calculates the damage based on linear superposition. This results in a significant discrepancy between the final fatigue damage and the actual value. Therefore, there is an urgent need to propose a crack-resistant large-volume concrete structure suitable for bridge piers and abutments, along with its construction method. Summary of the Invention

[0004] Purpose of the invention

[0005] This invention provides a fatigue damage calculation method with dynamic weight correction to solve the problem of excessively large discrepancies between fatigue damage and actual values ​​in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A fatigue damage calculation method with dynamic weight correction includes the following steps: S1: Collect stress level load test data and record the load application sequence; S2: Construct a dynamic weight coefficient model formula: In the formula: w i σ is the dynamic weighting coefficient. i For the stress applied in the i-th time, σ refσ is the fatigue limit stress of the material, k is the nonlinear damage exponent of the material, γ is the load history attenuation factor, and Δσ is the load history attenuation factor. prev This represents the maximum difference between the preceding load and the current stress.

[0008] S3: Constructing the fatigue damage accumulation formula: In the formula:

[0009] w i : represents the dynamic weighting coefficient, n i : N represents the actual number of iterations. i : represents the maximum number of iterations;

[0010] S4: Calculate the damage degree D under different load sequences according to step S3, and predict the fatigue life of the material. When D is not less than 1, the material is considered to have failed due to fatigue. When D is less than 1, the material is considered to have not failed.

[0011] As a further description of the above scheme, in step S1, two sets of stresses are selected, namely high stress σ1 and low stress σ2, and the corresponding limiting cycle numbers N1 and N2 are obtained through experiments, based on the Basquin equation σ k ·N=C, solve the following simultaneous equations: The nonlinear damage index k value of the material was obtained.

[0012] As a further description of the above scheme, in step S1, σ ref The value is calculated based on the SN curve of the material itself, corresponding to 1.95-2.05×10. 6 The stress corresponding to the number of cycles.

[0013] As a further description of the above scheme, in step S1, the γ value needs to be calculated using the improved cumulative damage formula D=1, and the calculation steps are as follows:

[0014] Step 1: Design at least two sets of variable amplitude load tests, including a high-stress to low-stress sequential test and a low-stress to high-stress sequential test;

[0015] Step 2: In the high stress to low stress sequential test, first apply overload stress σ1 for n1 cycles, then apply normal stress σ2 until failure, and record the number of cycles n2;

[0016] In the low-stress to high-stress sequential test, normal stress σ2 is applied first for n'2 cycles, then overload stress σ1 is applied until failure, and the number of cycles n'1 is recorded.

[0017] Solving the improved fatigue damage formula D = 1, we get: Where D1 represents the cumulative damage value during the overload phase, and the calculation formula is: D1=(σ1 / σ ref )k ·(n1 / N1).

[0018] As a further description of the above scheme, in step S1, Δσ prev The calculation formula is: Δσ prev =max(σ1, σ2, σ3, ..., σ i-1 )-σ i When calculating the maximum value of the difference between the stress applied before fatigue and the current stress, if the stress value applied before fatigue is not higher than the current calculated stress value, the difference between the stress applied before fatigue and the current stress is 0.

[0019] Advantages and effects of the present invention:

[0020] This invention improves the cumulative fatigue calculation formula based on the traditional linear fatigue accumulation algorithm.

[0021] Through dynamic coefficients It accurately reflects the nonlinear damage acceleration effect under high stress, avoiding errors caused by traditional linear models (such as the Miner criterion) that neglect load sequence and stress amplitude differences. Furthermore, the material nonlinear exponent k requires only two sets of experimental data at different stress levels (high stress σ1 and low stress σ2), and is obtained by combining these data using the Basquin formula. It can be quickly determined without complex fitting, and has clear engineering definition criteria, namely the preceding stress difference Δσ. prev =max(σ1, σ2, σ3, ..., σ i-1 )-σ i It updates the impact of historical loads in real time to ensure that the calculation results are closer to the actual results. Attached Figure Description

[0022] Figure 1 This is a flowchart of the fatigue damage calculation method with dynamic weight correction according to an embodiment of the present invention. Detailed Implementation

[0023] 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, and 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.

[0024] like Figure 1 As shown, a fatigue damage calculation method with dynamic weight correction includes the following steps: S1: Collect stress level load test data and record the load application sequence; S2: Construct a dynamic weight coefficient model formula: In the formula: w i σ is the dynamic weighting coefficient. i For the stress applied in the i-th time, σ ref σ is the fatigue limit stress of the material, k is the nonlinear damage exponent of the material, γ is the load history attenuation factor, and Δσ is the load history attenuation factor. prev This represents the maximum difference between the previous load and the current stress; (if the previous applied stress value is not higher than the current calculated stress value, this value is 0).

[0025] S3: Constructing the fatigue damage accumulation formula: In the formula:

[0026] w i : represents the dynamic weighting coefficient, Δ represents the actual number of iterations, N i : represents the maximum number of iterations;

[0027] S4: Calculate the damage degree D under different load sequences according to step S3, and predict the fatigue life of the material. When D is not less than 1, the material is considered to have failed due to fatigue. When D is less than 1, the material is considered to have not failed.

[0028] In step S1 of this embodiment of the invention, two sets of stresses are selected, namely high stress σ1 and low stress σ2, and the corresponding limiting cycle numbers N1 and N2 are obtained through experiments. Based on the Basquin equation σ k ·N=C, solve the following simultaneous equations: The nonlinear damage index k value of the material was obtained.

[0029] In step S1 of this embodiment of the invention, σ ref The value is calculated based on the SN curve of the material itself, corresponding to 1.95-2.05×10. 6 The stress corresponding to the number of cycles.

[0030] In step S1 of this embodiment of the invention, the γ value needs to be calculated using the improved cumulative damage formula D=1. The calculation steps are as follows:

[0031] Step 1: Design at least two sets of variable amplitude load tests, including a high-stress to low-stress sequential test and a low-stress to high-stress sequential test;

[0032] Step 2: In the high stress to low stress sequential test, first apply overload stress σ1 for n1 cycles, then apply normal stress σ2 until failure, and record the number of cycles n2;

[0033] In the low-stress to high-stress sequential test, normal stress σ2 is applied first for n'2 cycles, then overload stress σ1 is applied until failure, and the number of cycles n'1 is recorded.

[0034] Solving the improved fatigue damage formula D = 1, we get: Where D1 represents the cumulative damage value during the overload phase, and the calculation formula is: D1=(σ1 / σ ref ) k ·(n1 / N1).

[0035] In step S1 of this embodiment of the invention, Δσ prev The calculation formula is: Δσ prev =max(σ1, σ2, σ3, ..., σ i-1 )-σ i When calculating the maximum value of the difference between the stress applied before fatigue and the current stress, if the stress value applied before fatigue is not higher than the current calculated stress value, the difference between the stress applied before fatigue and the current stress is 0.

[0036] Comparison of specific calculation examples between the traditional linear fatigue accumulation algorithm (Miner's criterion) and the improved formula of this patent:

[0037] The specific calculation example is as follows:

[0038] Load spectrum:

[0039] σ1 = 250 MPa, n1 = 2,000 cycles, N1 = 89,500 cycles

[0040] σ² = 220 MPa, n² = 30,000 cycles, N² = 352,000 cycles

[0041] Traditional Miner's criterion calculation:

[0042] Substitute the data into the formula:

[0043]

[0044] Conclusion: D<1, the material has not failed, and the remaining lifetime is 1-0.1075=0.8925.

[0045] Patented improved algorithm calculation:

[0046] The material nonlinear exponent k is used to input the load spectrum data into the formula.

[0047] High stress: σ1 = 250 MPa, corresponding to a lifespan N1 = 89,500 cycles.

[0048] Low stress: σ2 = 220 MPa, corresponding to a lifespan of N2 = 352,000 cycles.

[0049] formula: The calculated value of k is 10.71.

[0050] The load history attenuation factor is calculated as follows:

[0051] Parameter settings:

[0052] High-low sequence test

[0053] Phase 1 (Overload Phase): Apply σ1 = 250 MPa, cycle number n1 = 2000 times.

[0054] Second stage (normal stage): Switch to σ2 = 220MPa, continue looping until failure, and record the actual number of loops n2 = 30000 times.

[0055] Low-high sequence test:

[0056] First stage (normal stage): Apply σ2 = 220 MPa, and repeat n'2 = 35000 times.

[0057] Second stage (overload stage): Switch to σ1 = 250MPa, continue cycling until failure, and record the actual number of cycles n'1 = X (to be determined through experimentation).

[0058] High-low order trials: n1 = 2,000 trials, n2 = 30,000 trials

[0059] Low-to-high sequence experiment: n'2 = 35000 times,

[0060] D1=(σ1 / σ ref ) k ·(n1 / N1)=(250 / 200) 10.71 ·(2000 / 89500)=0.244

[0061] Substituting the load history attenuation factor into the formula, we get:

[0062]

[0063] Calculate dynamic weight coefficients

[0064]

[0065] Substituting the above values ​​into the improved fatigue algorithm formula yields different damage accumulations:

[0066]

[0067] D = 0.244 + 0.845 = 1.089

[0068] Conclusion: D > 1, material failure.

[0069] Conclusion: Improved accuracy: The patented algorithm considers the load sequence (the effect of high stress to low stress) and the nonlinear effect of stress amplitude, and the damage value D is significantly higher than that of the traditional method (1.089 vs. 0.1075), which is closer to the actual damage.

[0070] Conservative prediction: The remaining lifetime of the patent result is lower (failure vs. 0.8925), avoiding the optimistic prediction caused by the neglect of load history in traditional methods, and is suitable for high reliability fields.

[0071] This invention uses dynamic weight correction to make the damage calculation results closer to the actual material failure threshold, avoiding the optimistic predictions caused by traditional methods that ignore load history. For example, in the example, this invention predicts material failure (D = 1.089), while the Miner criterion misjudges the remaining life as 0.8925. This conservative prediction is particularly suitable for working conditions with extremely high safety requirements. Through dynamic weight coefficients and load history compensation, the patented algorithm significantly improves the accuracy of fatigue life prediction while ensuring computational simplicity.

[0072] Despite the introduction of dynamic weight correction, the fatigue damage accumulation formula of this invention only adds the calculation of weight coefficients to the traditional Miner criterion, and the parameters in the weight coefficients can all be obtained through standardized tests. At the same time, the calculation process is not significantly complicated, taking into account both engineering practicality and technological advancement.

[0073] The traditional Miner criterion simplifies fatigue damage as a linear superposition, neglecting the nonlinear acceleration effect caused by differences in stress amplitude (such as high-stress-induced crack propagation accelerating subsequent low-stress damage). This invention, through a dynamic weighting coefficient model, for the first time couples the material's nonlinear damage exponent k with the load history attenuation factor γ, quantitatively characterizing the accelerating effect of high stress on damage accumulation.

[0074] Traditional nonlinear damage models often require complex fitting or a large amount of experimental data. However, this invention only requires two sets of stress levels (σ1, σ2) and the corresponding limiting cycle numbers (N1 and N2). The material nonlinear exponent k can be quickly determined through the Basquin equation without relying on complex curve fitting. At the same time, the load history attenuation factor γ can be solved by solving two sets of variable amplitude load tests simultaneously with D=1. The engineering definition standard is clear, which significantly reduces the experimental cost and computational complexity.

[0075] Existing technologies assume that damage is independent of the order of load application, but in reality, high-stress-low-stress sequences and low-stress-high-stress sequences can lead to different damage rates due to differences in crack initiation. This invention addresses this by using Δσ... prev (The maximum difference between the preceding load and the current stress) The influence of historical loads is updated in real time. When high stress acts first, Δσ prev Non-zero, dynamic weighting coefficient w iThe increased load makes the damage calculation more consistent with the physical mechanism of high stress inducing crack propagation and accelerating subsequent damage. In the example, the present invention predicts material failure under a high-stress-low-stress sequence, while traditional methods misjudge it as not failing, verifying the accurate capture of the sensitivity to load sequence.

[0076] Traditional Miner's criterion simplifies fatigue damage as a linear superposition, neglecting the nonlinear acceleration effect caused by differences in stress amplitude (e.g., high-stress-induced crack propagation accelerates subsequent low-stress damage). This invention, through a dynamic weighted coefficient model wi, for the first time couples the material's nonlinear damage exponent k with the load history attenuation factor γ, quantifying the accelerating effect of high stress on damage accumulation. For example, in the numerical example, the traditional method calculates the damage degree D = 0.1075, while this invention, considering nonlinear effects, calculates D = 1.089, which more closely approximates the actual failure state of the material.

[0077] The above description is merely a few embodiments of this application and is not intended to limit this application in any way. Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any changes or modifications made by those skilled in the art without departing from the scope of the technical solution of this application using the disclosed technical content are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A fatigue damage calculation method with dynamic weight correction, characterized in that, Includes the following steps: S1: Collect stress level load test data and record the order of load application; S2: Formula for constructing the dynamic weight coefficient model: In the formula: w i σ is the dynamic weighting coefficient. i For the stress applied in the i-th time, σ ref σ is the fatigue limit stress of the material, k is the nonlinear damage exponent of the material, γ is the load history attenuation factor, and Δσ is the load history attenuation factor. prev This represents the maximum difference between the preceding load and the current stress. S3: Constructing the fatigue damage accumulation formula: In the formula: w i : represents the dynamic weighting coefficient, n i : N represents the actual number of iterations. i : represents the maximum number of iterations; S4: Calculate the damage degree D under different load sequences according to step S3, and predict the fatigue life of the material. When D is not less than 1, the material is considered to have failed due to fatigue. When D is less than 1, the material is considered to have not failed.

2. The fatigue damage calculation method with dynamic weight correction according to claim 1, characterized in that: In step S1, two sets of stresses are selected, namely high stress σ1 and low stress σ2. The corresponding limit cycle numbers N1 and N2 are obtained through experiments, based on the Basquin equation σ k ·N=C, solve the following simultaneous problems: The nonlinear damage index k value of the material was obtained.

3. The fatigue damage calculation method with dynamic weight correction according to claim 2, characterized in that: In step S1, σ ref The value is calculated based on the SN curve of the material itself, corresponding to 1.95-2.05×10. 6 The stress corresponding to the number of cycles.

4. The fatigue damage calculation method with dynamic weight correction according to claim 1, characterized in that: In step S1, the γ value needs to be calculated using the improved cumulative damage formula D=1. The calculation steps are as follows: Step 1: Design at least two sets of variable amplitude load tests, including a high-stress to low-stress sequential test and a low-stress to high-stress sequential test; Step 2: In the high stress to low stress sequential test, first apply overload stress σ1 for n1 cycles, then apply normal stress σ2 until failure, and record the number of cycles n2; In the low-stress to high-stress sequential test, normal stress σ2 is applied first for n′2 cycles, then overload stress σ1 is applied until failure, and the number of cycles n′1 is recorded. Solving the improved fatigue damage formula D=1, we get: Where D1 represents the cumulative damage value during the overload phase, and the calculation formula is: D1=(σ1 / σ ref ) k ·(n1 / N1).

5. The fatigue damage calculation method with dynamic weight correction according to claim 1, characterized in that: In step S1, Δσ prev The calculation formula is: Δσ prev =max(σ1, σ2, σ3, ..., σ i-1 )-σ i When calculating the maximum value of the difference between the stress applied before fatigue and the current stress, if the stress value applied before fatigue is not higher than the current calculated stress value, the difference between the stress applied before fatigue and the current stress is 0.

Images