1. Introduction
The fatigue behavior of metal materials is crucial. It has important applications in the main load-bearing structures of major equipment, including aircraft components, shield machine cutterheads, automotive parts, and other structures. These parts or structures will continue to experience fatigue load conditions. The mechanical properties of the components will continue to accumulate and fail during actual loading. Therefore, accurate prediction of fatigue life is crucial for the design optimization process. Fatigue failure is one of the most common modes of failure in major equipment, posing a significant threat to the safe operation of equipment, including aircraft, trains, ships, bridges, oil platforms, as well as aircraft engine turbine blades, steam turbine blades, etc., as shown in
Figure 1. Due to the complex structure and complex loads of major equipment, the main load-bearing structures typically undergo extremely complex loads during operation, and are also subject to the effects of environmental corrosion. As a result, cracks tend to initiate at stress concentration areas or structurally weaker locations, and microcracks slowly grow under the action of service loads until they form macroscopic cracks, which greatly threaten the safety of the structure. Incidents caused by fatigue failure of similar major equipment still occur from time to time, making people increasingly concerned about structural fatigue issues. In fact, more fatigue does not lead to accidents, as fatigue failure is cumulative and generally requires a certain operating time to occur. Moreover, the location of the failure is uncertain before it occurs, and there are no obvious signs, so fatigue failure is hidden and very dangerous, which must be given high attention. To ensure the safe operation of equipment, it is necessary to conduct real-time detection and timely maintenance of the fatigue of the structure.
Fatigue is an accumulation process of damage where the material properties continuously degrade under the action of cyclic loads. The analysis of damage accumulation plays a crucial role in preventing fatigue failure, and the mechanism of damage evolution is one of the important issues in the study of fatigue behavior. Continuous loading under cyclic loading leads to cumulative fatigue damage in structural materials [
1]. Local damage caused by cyclic loads is an accumulation process composed of crack initiation, propagation, and fracture. During the cyclic loading process, local plastic deformation occurs in the area with the highest stress, leading to crack initiation. As the number of loading cycles that the part undergoes continues to accumulate, the propagation of cracks also increases. When the load reaches a certain number of cycles, the cracks will lead to structural failure and fracture. It is necessary to monitor the load history in real-time to predict the fatigue life of equipment, thereby achieving a fatigue load assessment of in-service equipment [
2]. It is crucial to effectively and timely predict and evaluate the stress, strain, external impact, and fatigue damage of structures [
3]. The remaining fatigue life is calculated through future load prediction models and physical failure models. For fatigue monitoring of mechanical structures, it is usually based on cumulative damage theory to calculate the fatigue life. However, a large number of experimental studies and practical engineering applications have shown that fatigue life prediction based on Miner’s linear damage theory may have non-conservative results. Since then, a large number of damage models have been proposed [
4,
5,
6].
Since the loading sequence effect is an important factor affecting the fatigue damage accumulation rule, Jang et al. [
7] considered the effect of loading sequence and introduced specific influencing parameters into the fatigue damage accumulation model. Qin [
8] and Xie [
9] et al. introduced the concept of damage curves based on the fracture expansion concept theory, considering the amount of damage at each load level as a power function related to the load. Xia [
10] summarized the fatigue damage assessment methods under random loads and proposed a fatigue cumulative damage model based on the equal damage curve. Bjørheim [
11] et al. used the damage stress function of fatigue life to establish a nonlinear model for assessing cumulative fatigue damage under multi-level loading. Jiang [
12] et al. proposed a low-cycle fatigue model considering the interaction of loads to predict the fatigue life under random loads. Mohamed et al. [
13] considered the damage accumulation process based on the influence of material fatigue performance, considering the correlation between load paths. Lin et al. [
14] proposed a fatigue life model considering the evolution of structural fatigue damage through failure mode analysis of residual strength. Gao et al. [
15] proposed a nonlinear model considering the influence of load sequence by introducing load-related parameters. Zhou et al. [
16] proposed a Corten-Dolan cumulative damage model considering wear degree. Yuan et al. [
17] and others, based on the DCA model, established a fatigue damage model considering residual strength degradation caused by variable amplitude cycles under stress amplitude. Due to the importance and complexity of key structures in aerospace, there is still a lack of a complete and effective monitoring system for real-time prediction of the fatigue life of key structures [
18]. Intelligent node technology is used to monitor the damage state, external impact load, and usage environment of key parts of the structure in real-time, obtaining the fatigue load spectrum during use. Based on big data algorithms, the dominant load spectrum is predicted, and the remaining fatigue life is calculated. Ensure to reduce manufacturing costs and extend the service life of the aircraft. Predict fatigue life failure and provide timely maintenance warning for flight shutdown. Therefore, the intelligent node monitoring technology studied in this paper can obtain the damage state and evolution law of key structures in the service process in real-time. Timely warning provides important support for flight shutdown maintenance and service life prediction.
In summary, due to the complexity of the structure of major equipment, the complexity of the load, the complexity of manufacturing, and the complexity of the service environment, the law of force flow transfer in the main load-bearing structure of major equipment is complex, the structural weak position is difficult to locate, the structure is difficult to detect, and the whole life cycle is difficult to assess, which is prone to cause catastrophic accidents and poses a huge threat to the economy and people’s life safety. This article starts from the evolution mechanism of fatigue damage and comprehensively considers the influence of mixed cyclic loading sequence of high and low cycle loads on fatigue life performance. Based on the theory of nonlinear fatigue damage accumulation, a new nonlinear damage accumulation fatigue life model is established, and the influence factors of fatigue damage accumulation are introduced to improve the prediction accuracy of the model. Therefore, accurate fatigue life analysis, especially the life assessment of large and complex equipment in fields such as aerospace, rail transit, and ships, has important practical significance and engineering significance, and has become a bottleneck that needs to be urgently solved for the healthy and efficient operation of major equipment.
2. Theoretical Methods
The damage curve method is one of the most classic nonlinear cumulative damage theories, which has been optimized and improved by scholars, The model proposed by Manson-Halford [
19] was widely used in later research, and the calculation formula for the model is:
Among them, D is the amount of damage, n is the number of cycles, Nf is the fatigue life of load, NRef is the damage life.
The damage of the material under the
i level load cycle is:
In the formula, ni is the i level cyclic load, Nfi is the i level fatigue life, ai is the fitting parameter, β is a constant related to material properties. Similar to Miner’s law, when the critical damage value is 1, the material structure fails. After preliminary experiments and experience accumulation, the empirical constant of 0.4 is generally directly taken for ai in practical applications.
Under the two-stage variable amplitude load spectrum, The Manson Halford model formula is:
In the formula, n1 is the 1 level cyclic load, Nf1 is the 1 level fatigue life, n2 is the 2 level cyclic load, Nf2 is the 2 level fatigue life.
Due to the Manson-Halford theory of the damage curve method considering the order of load action and stress magnitude, it is more accurate to calculate the damage of two-stage loads. However, when extended to multi-level loads, due to the complex coupling effect between loads, there are also errors in the calculation results. Taking the cumulative damage theory diagram under secondary loading as an example. The cumulative processes of fatigue damage under high and low loads are shown in
Figure 2, respectively. The accumulation of fatigue damage is related to stress levels, and the higher the load level, the faster the rate of damage accumulation. a
1 and a
2 respectively represent different stress levels, and D
a and D
b respectively represent the cumulative fatigue damage at the corresponding stress levels of a
1 and a
2. Considering the influence of loading sequence on fatigue damage, the cumulative process of fatigue damage under variable amplitude load is shown in
Figure 2, with load level
σ1 >
σ2. When the order of load action is high low, the damage accumulation process is: o-a-b; When the load sequence is low high, the damage accumulation process is o-b-a. It can be observed that when the damage accumulates to the same level, the number of cycles with high low load sequence is less than the number of cycles with high low load sequence.
Under the loading of the first stage load, the number of cycles is n1, and under the continuous loading of the second stage load n2, it ultimately leads to fatigue failure of the material structure, with a number of cycles of Nf2. The fatigue life of the first level load is Nf1, and the fatigue life of the second level load is Nf2.
Since the fatigue damage values of
a and
b are the same at the load level, the relationship between the load cycle and the equivalent cycle ratio is:
In the formula, the number of cycles at the Nf2 life level is n2, and the equivalent damage corresponds to the initial cycle ratio.
The fatigue life problem of high and low cycle composite loading is one of the common fatigue failure modes under the coupling of high cycle stress and low cycle stress [
20]. Under high and low cycle composite fatigue loading, high stress levels under low cycle cyclic loading cause severe damage to the material; The stress level under high cycle loading is low, but the high frequency also has a serious impact on the material. There is currently a lack of sufficient and accurate research on the complex interaction coupling damage caused by high-low cycle composite loading. The current experiment lacks complete performance data and cannot predict the fatigue life of high and low cycle composite loading well. Therefore, in order to study the actual damage process and fatigue life of key structures during service, it is necessary to conduct in-depth research on them.
The load spectrum waveforms of low cycle fatigue and high cycle fatigue for service structures subjected to complex alternating loads in actual engineering are shown in
Figure 3. Low frequency and high amplitude low cycle cyclic loads (
Figure 3a) and high frequency and low amplitude high cycle cyclic loads (
Figure 3b) jointly composite and interact with key structures. However, in practical engineering, the load spectrum of service structures is a random load spectrum. To ensure that it can be studied under complex test conditions, it is necessary to convert it into a typical test load spectrum through the principle of equivalent damage. The specific load waveform is shown in
Figure 3c. Physical parameters such as high cycle stress
σH, high cycle stress frequency
fH, low cycle stress
σL, and low cycle stress frequency
fL are commonly used for research. At the same time, it is closely related to parameters such as the ratio of high cycle stress frequency to low cycle stress frequency.
The impact of different number of cycles under each level of load on the accumulation of subsequent fatigue damage is also different. Therefore, the relationship between the ratio of cycle loading times to fatigue life (loading cycle ratio) and fatigue damage can be used to represent the impact of load action sequence on fatigue damage, as shown in
Figure 3. It can be found that:
When the loading order is high low,
When the loading order is low high,
Based on the classic model mentioned above, this paper proposes a complex fatigue damage life prediction model for high and low cycle coupled loading, considering the influence of high and low cycle coupled loading and multi-level spectrum on material structure, which is more universal and accurate.
When the external stress applied by the actual service structure is lower than the yield strength
σz of the material’s inherent properties, there exists a relationship between the loading stress amplitude of the material and the fatigue life of the structure:
Among them, σa is the stress amplitude of the material, is the fatigue strength, b is the material parameter.
Due to the complex causes of structural fatigue, fatigue damage is a key factor affecting the lifespan. When the structure experiences complete failure, the use of damage models for calculation is not considered. The cumulative value of the ratio of cyclic load to load fatigue life is the fatigue damage value.
According to the equivalent theory of fatigue cumulative damage characteristics of materials, the expression for the cyclic ratio is:
In the formula, is the number of cycles required to increase the damage from Coordinate axis O along the damage curve a to b under stress σ2.
When the cumulative damage
D reaches the critical value after
n2 cycles of
σ2, the material will experience fatigue failure, with a critical damage value of D = 1. At this point, the criterion for fatigue failure under the secondary load spectrum can be expressed as:
Based on this, the expression for the three-level loading damage model is:
Due to the improved accuracy of the Manson Halford damage model’s three-level loading damage prediction compared to the Miner criterion, it cannot meet the requirements of practical engineering research. Therefore, this paper introduces a correction factor to improve and expand the multi-level loading damage model, as shown below:
Among them, ni is the i level cyclic load, Nfi is the i level fatigue life, is the cyclic stress ratio, Dcr is the fatigue damage correction factor. Dcr needs to be experimentally determined.
However, considering the influence of load loading sequence on fatigue life, the fatigue life is less than 1 under high low loading and greater than 1 under low high loading. Therefore, this paper introduces the fatigue damage correction factor Dcr and improves and extends it to the multi-level loading damage model.
This article proposes a universal formula. Suitable for most metal materials and other isotropic similar materials. Mainly researching commonly used materials for major equipment. For example: steel, iron, aluminum, etc.