Multi-axis random load fatigue crack propagation life analysis method

Abstract

The invention discloses a multi-axis random load fatigue crack propagation life analysis method. The method comprises the following steps: converting a multi-axis random load at a structure loading position into an equivalent random load at a dangerous point; establishing association between the random load at the dangerous point and the damage parameter on the crack propagation plane, and determining the most dangerous load combination through a multi-axis load projection theory; utilizing a rain flow counting method which introduces a processing strategy of rearranging according to a cycle starting time sequence to obtain cycle information and load sequence information of a random load spectrum; and based on a rain flow counting result, performing crack propagation life analysis by combining an NASGRO model and a cycle jump method, and calculating the total crack propagation life by using a Miner damage accumulation criterion. According to the method, the influence of the phase difference and stress ratio on the crack propagation life before the multi-axis random load is considered, the accuracy of crack propagation life prediction under the multi-axis random load is improved, and a basis is provided for damage tolerance design analysis of engineering components.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical structure random multi-axial fatigue crack growth life calculation, in particular to a multi-axial random load fatigue crack growth life analysis method. Background Art

[0002] Currently, research on the prediction of multiaxial fatigue crack growth life primarily focuses on conditions under uniform-amplitude multiaxial loading. However, in actual engineering service environments, components are often subjected to multiaxial loads whose amplitudes vary randomly over time. Therefore, the prediction of fatigue crack growth life under random multiaxial loading has become an important research direction in this field, with important practical significance for the safety assessment of engineering structures. Although some progress has been made in predicting crack growth life under uniform-amplitude multiaxial loading, a systematic and reliable prediction method has not yet been established due to the highly nonlinear and complex fatigue behavior under random multiaxial loading.

[0003] To address these issues, the present invention first improves the rainflow counting method by incorporating the influence of load order, thereby enhancing the accuracy of load spectrum processing in crack growth problems. This improved method is then combined with a typical crack growth model and linear damage criterion to propose a fatigue crack growth life estimation method under a random spectrum. This estimation method is then evaluated using existing test data. Finally, using load equivalent synthesis technology, combined with the relationship between damage parameters and random loads, the effect of phase difference on crack growth life under multiaxial random loading is addressed. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes a multi-axial random load fatigue crack growth life analysis method, which can solve the influence of phase difference on crack growth life in multi-axial random load.

[0005] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0006] The present invention is a multi-axial random load fatigue crack growth life analysis method, which includes the following operations:

[0007] Convert the multiaxial random loads at the structural loading point into the random loads at the equivalent dangerous point;

[0008] Establish the correlation between random loads at the dangerous point and damage parameters on the crack propagation plane, and determine the most dangerous load combination through multi-axial load projection theory;

[0009] The cycle information and load sequence information of the random load spectrum are obtained by using the rain flow counting method which introduces the strategy of reordering the process according to the cycle start time.

[0010] Based on the rain flow counting results, the crack growth life analysis was carried out in combination with the NASGRO model and the cycle jumping method, and the total crack growth life was calculated using the Miner damage accumulation criterion.

[0011] A further improvement of the present invention is that the conversion method of the equivalent random load at the dangerous point includes:

[0012] Define the structure to be subjected to multiaxial random load L m (t) action, where m = 1, 2, 3, ..., N, N represents the number of random load spectra, N = 6 × N LP , N LP Indicates the number of load application points. Each application point has 6 degrees of freedom. The load at the dangerous point is expressed as:

[0013]

[0014] Where, the multiaxial random load L m (t) is the force load or torque load; σ ij (t) is the stress state at the dangerous point, that is, the random load at the dangerous point, i, j = x, y, z correspond to the three axes in the O-XYZ coordinate system respectively; c nm is the coefficient of multi-axial random load, which is obtained through simulation under the premise of determining the dangerous points of the structure.

[0015] A further improvement of the present invention is to establish a correlation between random loads and damage parameters on the crack propagation plane, specifically including:

[0016] Stress state σ at the dangerous point ij (t) Coordinate transformation is performed to obtain the normal stress on the crack propagation plane and shear stress Among them, normal stress and shear stress The expression is:

[0017]

[0018] Where: is the coordinate transformation matrix;

[0019] The relationship between the damage parameter on the crack propagation plane and the random load at the dangerous point is established, and the expression is:

[0020]

[0021] Where, is the damage parameter on the crack propagation plane, B and K are coefficients related to the crack propagation characteristics of the material, and C m is the coefficient to be determined.

[0022] A further improvement of the present invention is to determine the most dangerous load combination through multi-axis load projection theory, specifically including:

[0023] Normalizing the correlation between random load and damage parameters on the crack propagation plane, we obtain:

[0024]

[0025] Where C′ m is a unit vector along the set direction in the m-dimensional vector space, ||C|| is the conversion coefficient between the pseudo damage value obtained under multi-axial random load and the actual damage value of the structure;

[0026] Get the damage parameter on the crack propagation plane Maximize the random load combination coefficient γ, the random load combination coefficient is expressed as:

[0027]

[0028] A further improvement of the present invention is to obtain the cycle information and load sequence information of the random load spectrum by using a rain flow counting method that introduces a strategy of rearranging the processing sequence according to the cycle start time, specifically including:

[0029] Load preprocessing: preprocess the random load data to obtain a simplified random load spectrum, including extraction of peak and valley values ​​and elimination of small-amplitude loads;

[0030] Rainflow counting: After the load preprocessing is completed, the main channel load is sorted and re-preprocessed, including re-ordering the load sequence, re-compressing the re-ordered load sequence, removing non-peak valley points and residual load points with amplitudes less than the set threshold, and synchronously deleting the number information corresponding to the removed load points, keeping the remaining numbers unchanged, and obtaining a (nr) × 2 two-dimensional array, where r is the number of deleted load points; rainflow cycle counting is performed to identify all valid load cycles, record the corresponding cycle amplitude, mean and corresponding number information, and the cycle counting results are recorded.

[0031] Store in a new two-dimensional array IDs, which contains m load cycles, each row corresponds to a cycle, and the columns include: cycle mean, amplitude, number of cycles, start time and end time;

[0032] Load sorting: The new two-dimensional array IDs are further sorted, including extracting the start time and end time information from the new two-dimensional array IDs, and sorting them from early to late by start time. If there are multiple cycles with the same start time, they are further sorted in the order of the end time. After the sorting is completed, the rain flow counting result that fully retains the load sequence information is obtained.

[0033] A further improvement of the present invention is to combine the NASGRO model and the cycle jump method to perform crack growth life analysis, specifically including:

[0034] Based on the influence of stress ratio and plastic crack closure effect on crack growth life, the NASGRO model is used to analyze the crack growth life. The load segment where stable crack growth occurs is fully coupled and calculated. The crack growth front of the structure is updated and the crack growth life is calculated. The number of cycles is determined and the load segment that does not reach the set value of crack growth is cycle skipped, i.e., automatic cycle skipping is performed.

[0035] A further improvement of the present invention is that the operation of automatic cycle jumping includes:

[0036] After the crack propagation starts, the crack propagation model is used to predict the crack propagation. The expression of the crack propagation model is:

[0037]

[0038] In the formula: da(N k ) is the crack extension calculated from the crack extension rate in the kth stable cycle, △N k is the jump period, △a k is the crack extension length;

[0039] Define the crack extension length △a k The tolerance range of the limit is set, and an adaptive algorithm is used to adjust the number of cycles of the next cycle jump according to the previous cycle jump. The expression is:

[0040]

[0041] Where: △N k+1 is the next jump cycle, is the maximum crack extension length, △a max , △a min are the maximum and minimum crack growth tolerances;

[0042] The range of crack growth increments is controlled by defining the maximum and minimum crack growth tolerances, thereby adjusting the number of subsequent crack growth cycles.

[0043] The beneficial effects of the present invention are as follows: the multi-axis random spectrum equivalence method used in the present invention can simplify the load input to a limited extent, and by establishing a quantitative relationship between the damage parameters and the random load on the crack propagation plane, the driving mechanism of the load fluctuation on the local damage is clarified. Based on the multi-axis load projection theory, the reasonable mapping of the spatial load to the crack dominant plane is realized, and the phase difference problem caused by the multi-axis random load is effectively solved; the present invention adopts an improved rain flow counting method, which strictly retains the actual load sequence by sorting the start time and end time of the cycle, thereby improving the accuracy of fatigue crack propagation calculation; combining the improved cycle counting method and the NASGRO crack propagation formula, the present invention proposes a fatigue crack propagation life analysis method under random load, the improved cycle counting method can more accurately identify and process the cyclic load under random spectrum load, and the NASGRO formula can be used to correct the stress ratio in spectrum loading and the crack hysteresis effect after high load on the crack propagation life. In summary, the present invention considers the influence of phase difference and stress ratio before multiaxial random loading on crack growth life, improves the accuracy of crack growth life prediction under multiaxial random loading, and provides a basis for damage tolerance design analysis of engineering components. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of a method in an embodiment of the present invention;

[0045] Figure 2 Schematic diagram of structural damage calculation under multi-axial random load spectrum in an example of the present invention;

[0046] Figure 3 Schematic diagram of the multi-axis circular projection method in an example of the present invention;

[0047] Figure 4 Schematic diagram of the cycle skipping technology in an example of the present invention. DETAILED DESCRIPTION

[0048] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only intended to illustrate the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0049] like Figure 1 As shown, the multi-axial random load fatigue crack growth life analysis method of this embodiment includes the following operations:

[0050] Step 1: Convert the multiaxial random load at the structural loading point into the random load at the equivalent dangerous point.

[0051] First, the static strength analysis of the structure is carried out to determine the dangerous point location of the structure. m (t) (m=1,2,3,...,N, N represents the number of random load spectra, N=6×N LP , N LP represents the number of load application points, each application point has 6 degrees of freedom), the stress at the dangerous point of the structure can be obtained by linear superposition of multi-axial random loads at the load application points.

[0052] Under multiaxial random load L m Under the action of (t), the load at the fatigue crack growth danger point is expressed as:

[0053]

[0054] Where, the multiaxial random load L m (t) is the force load or torque load; σ ij (t) is the stress state at the dangerous point, that is, the random load at the dangerous point, i, j = x, y, z correspond to the three axes in the O-XYZ coordinate system respectively; c nm is the coefficient of multi-axial random load, which is obtained through simulation under the premise of determining the dangerous points of the structure.

[0055] Step 2: Establish the correlation between the random load at the critical point and the damage parameter on the crack propagation plane.

[0056] Under multiaxial loads, fatigue cracks will propagate in prefabricated crack structures. The direction of fatigue crack propagation is usually determined and calculated based on the damage parameters on the crack propagation plane. By combining fracture mechanics principles with crack propagation rate models and utilizing the distribution characteristics of equivalent stress on the crack propagation plane, the intrinsic relationship between the external load spectrum and the evolution of structural damage can be effectively characterized. The present invention uses the combination of the maximum principal stress and shear stress on the crack propagation plane as the damage parameter, which is expressed as:

[0057]

[0058] Where B and K are coefficients related to the crack growth characteristics of the material. When the damage parameter only considers the principal stress in the crack growth plane (i.e., the crack growth is pure mode I crack growth), B = 0 and K = 1. If only the shear stress is considered (i.e., the crack growth is pure mode II crack growth), B = 1 and K = 0. If the damage parameter is a combination of the two, B = 1, and K is a material constant.

[0059] like Figure 2 As shown, the spatial stress state σ at the fatigue danger point ij (t) coordinate transformation can be used to obtain the normal stress on the crack propagation plane and shear stress Normal vector to the crack propagation plane With tangent vector They are expressed as direction cosines respectively. (in, Represents the unit vector of each axis in the O-XYZ coordinate system. The normal stress after coordinate system conversion and shear stress It can be expressed as:

[0060]

[0061] Where: is the coordinate transformation matrix.

[0062] The relationship between the damage parameter on the crack propagation plane and the random load at the dangerous point is established, and the expression is:

[0063]

[0064] Where, is the damage parameter on the crack propagation plane, B and K are coefficients related to the crack propagation characteristics of the material, and C m is the coefficient to be determined.

[0065] Step 3: Determine the most dangerous load combination through multi-axial load projection theory.

[0066] Normalizing the correlation between random load and damage parameters on the crack propagation plane, we obtain:

[0067]

[0068] Where C′ m is the unit vector along the set direction in the m-dimensional vector space, ||C|| is the conversion coefficient between the pseudo damage value obtained under the multi-axial random load and the actual damage value of the structure; All combinations of correlation coefficients that meet the constraints can be traversed by enumeration, thereby providing a set of feasible solutions that adapt to different load paths for damage assessment.

[0069] From a geometric point of view, C′ m is a unit vector along a certain direction in the m-dimensional vector space, so It is the projection of the load vector in a certain direction in the m-dimensional vector space.

[0070] like Figure 3 As shown, the extreme value of each axial load spectrum is set to UG m and OG m , then calculate the circumscribed circle of this circulant matrix rectangle, the center of the circle M k and radius ρ can be calculated by the following formula:

[0071]

[0072] The limit after projection in any direction C' can be expressed as:

[0073]

[0074] Where (M, C') is the projection of circle M in any direction C', is the limit after projection in any direction C'.

[0075] By using the above method, the multiaxial random load and its corresponding circulant matrix can be projected onto any given direction C'. Combined with the calculation of the random load at the equivalent dangerous point, the damage parameter on the crack propagation plane can be obtained. Maximize the random load combination coefficient γ, the random load combination coefficient is expressed as:

[0076]

[0077] Step 4: The cycle information and load sequence information of the random load spectrum are obtained by using the rain flow counting method that introduces a reordering processing strategy based on the cycle start time sequence.

[0078] To more accurately capture the impact of this load sequence on fatigue behavior, this example proposes an improved rainflow counting method for random load spectra. This method introduces a processing strategy that reorders the data by cycle start time. Compared with traditional methods, this improved strategy not only retains the load information statistically obtained by traditional rainflow counting methods, but also effectively restores the temporal structure of the original load sequence by sorting the counting results by cycle start time, thereby more realistically reflecting the impact of the load history on crack growth.

[0079] First, load preprocessing is performed. To simplify the subsequent fatigue analysis steps, the raw random load data must be preprocessed before cycle counting. This process primarily involves two aspects: peak and valley extraction and the removal of small-amplitude loads. By identifying local extreme points in the load sequence, representative peaks and valleys are extracted, resulting in a simplified load spectrum. To further reduce the data volume and improve computational efficiency, loads with small amplitudes and limited impact on fatigue crack growth life are screened out.

[0080] Then, rainflow counting is performed. After load preprocessing is completed, the extracted peak and valley values ​​need to be stored in a two-dimensional array according to a specific format. The first column of the array stores the peak and valley values ​​arranged in time series, and the second column contains the corresponding numbers. Next, the single-axis three-point rainflow counting process is performed on the first column of the array, which specifically includes the following two sub-steps:

[0081] (1) Arrangement and re-preprocessing of the main channel load. In order to unify the starting point of the rainflow counting, it is necessary to reorder the original load sequence, usually with the maximum value as the first peak-valley point. Subsequently, the reordered data is compressed again to remove non-peak-valley points and residual load points with small amplitudes. In this process, the numbering information corresponding to the removed load points should be deleted synchronously, while the remaining numbers remain unchanged to maintain the consistency of the data with the subsequent analysis process. The load array is processed to obtain a (nr) × 2 two-dimensional array, where r is the number of deleted load points.

[0082]

[0083] (2) Standard rainflow cycle counting. Following the procedure specified in ASTM E1049-85, the rainflow cycle counting operation is performed on the main channel load in sequence. This process will identify all valid load cycles and record their cycle amplitude, mean, and corresponding number information. The cycle results will be stored in a new two-dimensional array IDs, which contains m load cycles, with each row corresponding to a cycle, and its columns include: cycle mean, amplitude, number of cycles, start time, and end time.

[0084]

[0085] Finally, load sorting. To ensure the accuracy of the load history sequence in subsequent analysis, the aforementioned IDs array needs to be further sorted. By extracting the start and end time information from the array, first sort them from earliest to latest by start time. If there are multiple cycles with the same start time, they are further refined and sorted according to the order of the end time. After completing this sorting step, a rain flow counting result that fully retains the load sequence information can be obtained, providing reliable data support for fatigue crack growth life assessment.

[0086] Step 5: Based on the rain flow counting results, the crack growth life is analyzed by combining the NASGRO model and the cycle jumping method, and the total crack growth life is calculated using the Miner damage accumulation criterion.

[0087] In order to consider the influence of stress ratio and plastic crack closure effect on crack growth life, this embodiment uses the NASGRO model, which has the ability to fully describe the entire fatigue crack growth process (including the threshold region, stable growth region, and failure acceleration region), to perform crack growth life analysis. Its general expression is:

[0088]

[0089] Among them, △K th is the crack growth threshold, below which the stress intensity factor amplitude △K is too low to cause crack growth; KC is the fracture toughness of the material; C, n, p and q are constants determined by experiments; R eff is the effective stress ratio; f is the ratio of the crack opening stress intensity factor to the maximum stress intensity factor; K A is the stress intensity factor of the applied load; K R is the residual stress intensity factor; △K T is the stress intensity factor range.

[0090] K A and K R Used to calculate crack growth based on superposition and obtained from two separate analyses. K A is caused by the external load, and K R is related to the initial residual stress field. Both can be determined by the weighted function stress intensity factor algorithm, which is expressed as follows:

[0091]

[0092] Where a h is the crack half-length, σ is the stress distribution when the region does not contain any cracks, F is the geometric correction factor of the crack body, W is the width of the crack body geometry, and L is the length of the crack body geometry.

[0093] The total stress intensity factor is obtained by adding the stress intensity factors of the external loads and residual stresses of the structure, which can be expressed as:

[0094]

[0095] Where K A,max , K A,min is the maximum and minimum value of the stress intensity factor of the applied load, K T,max and K T,min is the total stress intensity factor for the maximum and minimum loads in the cycle, including the contribution of the residual stress intensity factor to the stress intensity factor of the applied load. Then, the stress intensity factor range and effective stress ratio are calculated separately and can be expressed as:

[0096] △K T =K T,max -K T,min =(K A,max +K R )-(K A,min +K R )

[0097]

[0098] In order to avoid the huge amount of calculation caused by the traditional method of performing detailed calculations for each cyclic load of crack propagation, a rapid analysis is performed using the cycle skipping technology. As shown in the figure, the AB segment represents the i-th multi-axial fatigue load, and the CD segment represents the k-th multi-axial fatigue load. It is assumed that the crack propagates stably in load segments similar to AB and CD, and the expansion in the remaining cyclic segments such as BC and DE is small, and the impact on the fatigue crack propagation life can be ignored. In the AB, CD and other cycle stages, a full coupling calculation is performed to update the crack propagation front of the structure, and the crack propagation life calculation is performed to determine the number of cycles. In the BC, DE and other cycle stages, cycle skipping is performed. After the crack propagation starts, the crack propagation model is used to predict the propagation of the crack in subsequent cycles.

[0099]

[0100] Among them, da(N k ) is the crack growth amount calculated from the crack growth rate in the kth stable cycle.

[0101] like Figure 4 As shown, in order to realize the automatic cycle jump processing method, we can define △a k The tolerance range of the limit is set, and an adaptive algorithm is used to adjust the number of cycles of the next cycle jump based on the previous cycle jump. This process can be summarized as follows:

[0102]

[0103] First, calculate the previous jump period △N k The maximum crack extension length caused by △N used to estimate the next cycle jump k+1 Then, by defining the minimum and maximum crack growth tolerances ∆a min and △a max These tolerances are used to control the range of crack growth increments, thereby adjusting the number of subsequent crack growth cycles. By continuously updating the crack growth length increment through the above method, a more realistic reflection of the change in fatigue crack growth with cycle number can be achieved, thus achieving the purpose of "cycle skipping".

[0104] Among the various fatigue damage accumulation theories, the most influential is Miner's linear damage accumulation criterion. Due to its concise form, convenient calculation, and good applicability under linear load conditions, it is widely used in engineering practice.

[0105] The total crack growth life cycle number is obtained by accumulating the crack growth cycles calculated in each calculation sub-step, and the fatigue crack growth life is obtained, which can realize the multi-axial fatigue crack growth life prediction of complex structures.

[0106] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and, unless defined similarly as herein, will not be interpreted in an idealized or overly formal sense.

[0107] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. Multiaxial random loading fatigue crack growth life analysis method, characterized by: The following operations are included: Convert the multiaxial random loads at the structural loading point into the random loads at the equivalent dangerous point; Establish the correlation between random loads at the dangerous point and damage parameters on the crack propagation plane, and determine the most dangerous load combination through multi-axial load projection theory; The cycle information and load sequence information of the random load spectrum are obtained by using the rain flow counting method which introduces the strategy of reordering the process according to the cycle start time. Based on the rain flow counting results, the crack growth life analysis was carried out in combination with the NASGRO model and the cycle jumping method, and the total crack growth life was calculated using the Miner damage accumulation criterion.

2. The multi-axial random load fatigue crack growth life analysis method according to claim 1, characterized in that: The conversion methods of the equivalent random load at the dangerous point include: Define the structure to be subjected to multiaxial random load L m (t) action, where m = 1, 2, 3, ..., N, N represents the number of random load spectra, N = 6 × N LP , N LP Indicates the number of load application points. Each application point has 6 degrees of freedom. The load at the dangerous point is expressed as: Where, the multiaxial random load L m (t) is the force load or torque load; σ ij (t) is the stress state at the dangerous point, that is, the random load at the dangerous point, i, j = x, y, z correspond to the three axes in the O-XYZ coordinate system respectively; c nm is the coefficient of multi-axial random load, which is obtained through simulation under the premise of determining the dangerous points of the structure.

3. The multi-axial random load fatigue crack growth life analysis method according to claim 2, characterized in that: The establishment of the association between the random load and the damage parameter on the crack propagation plane specifically includes: Stress state σ at the dangerous point ij (t) Coordinate transformation is performed to obtain the normal stress on the crack propagation plane and shear stress Among them, normal stress and shear stress The expression is: Where: is the coordinate transformation matrix; The relationship between the damage parameter on the crack propagation plane and the random load at the dangerous point is established, and the expression is: Where, is the damage parameter on the crack propagation plane, B and K are coefficients related to the crack propagation characteristics of the material, and C m is the coefficient to be determined.

4. The multi-axial random load fatigue crack growth life analysis method according to claim 3, characterized in that: The most dangerous load combination is determined by multi-axis load projection theory, specifically including: Normalizing the correlation between random load and damage parameters on the crack propagation plane, we obtain: Where C′ m is a unit vector along the set direction in the m-dimensional vector space, ||C|| is the conversion coefficient between the pseudo damage value obtained under multi-axial random load and the actual damage value of the structure; Get the damage parameter on the crack propagation plane Maximize the random load combination coefficient γ, the random load combination coefficient is expressed as:

5. The multi-axial random load fatigue crack growth life analysis method according to claim 1, characterized in that: The cycle information and load sequence information of the random load spectrum are obtained by using the rain flow counting method that introduces a reordering processing strategy based on the cycle start time sequence, specifically including: Load preprocessing: preprocess the random load data to obtain a simplified random load spectrum, including extraction of peak and valley values ​​and elimination of small-amplitude loads; Rainflow counting: After the load preprocessing is completed, the main channel load is sorted and re-preprocessed, including re-ordering the load sequence, re-compressing the re-ordered load sequence, removing non-peak valley points and residual load points with amplitudes less than the set threshold, and synchronously deleting the number information corresponding to the removed load points, keeping the remaining numbers unchanged, and obtaining a (nr) × 2 two-dimensional array, where r is the number of deleted load points; rainflow cycle counting is performed to identify all valid load cycles, record the corresponding cycle amplitude, mean and corresponding number information, and the cycle counting results are recorded. Store in a new two-dimensional array IDs, which contains m load cycles, each row corresponds to a cycle, and the columns include: cycle mean, amplitude, number of cycles, start time and end time; Load sorting: The new two-dimensional array IDs are further sorted, including extracting the start time and end time information from the new two-dimensional array IDs, and sorting them from early to late by start time. If there are multiple cycles with the same start time, they are further sorted in the order of the end time. After the sorting is completed, the rain flow counting result that fully retains the load sequence information is obtained.

6. The multi-axial random load fatigue crack growth life analysis method according to claim 1, characterized in that: The crack growth life analysis is performed by combining the NASGRO model and the cycle jump method, specifically including: Based on the influence of stress ratio and plastic crack closure effect on crack growth life, the NASGRO model is used to analyze the crack growth life. The load segment where stable crack growth occurs is fully coupled and calculated. The crack growth front of the structure is updated and the crack growth life is calculated. The number of cycles is determined and the load segment that does not reach the set value of crack growth is cycle skipped, i.e., automatic cycle skipping is performed.

7. The multi-axial random load fatigue crack growth life analysis method according to claim 6, characterized in that: The operation of automatic cycle skipping includes: After the crack propagation starts, the crack propagation model is used to predict the crack propagation. The expression of the crack propagation model is: In the formula: da(N k ) is the crack extension calculated from the crack extension rate in the kth stable cycle, △N k is the jump period, △a k is the crack extension length; Define the crack extension length △a k The tolerance range of the limit is set, and an adaptive algorithm is used to adjust the number of cycles of the next cycle jump according to the previous cycle jump. The expression is: Where: △N k+1 is the next jump cycle, is the maximum crack extension length, △a max , △a min are the maximum and minimum crack growth tolerances; The range of crack growth increments is controlled by defining the maximum and minimum crack growth tolerances, thereby adjusting the number of subsequent crack growth cycles.

Images