Fatigue Life Prediction Model Research Articles

Prediction model of fatigue crack propagation life of corroded steel plate considering pit characteristics

Pitting corrosion notably increases stress concentration, thus accelerating fatigue crack initiation and propagation at the pit base. This research quantifies the relationship between pit dimensions and fatigue crack development by introducing crack extension ratios (γa​ and γc)​. Employing a series of finite element models, the influence of pit characteristics on the stress intensity factor (K) at the crack tip is assessed. Based on these assessments, a formula for computing the K-value is proposed. Furthermore, a two-stage fatigue life prediction model for both partial and complete penetration is developed using the Paris equation. Results indicate that pits substantially affect the K-value during partial penetration. Specifically, as γa approaches zero, the K-value approaches zero, and as γa approaches one, it aligns with the K-value of a semi-elliptical surface crack. Conversely, in the complete penetration phase, the influence of the pit on the K-value is negligible, and the K-value can be calculated according to the plate with a central penetrating crack. Experimental validation confirms that the model generally maintains a prediction error within 10%.

Deformation mechanism and life prediction model of titanium alloy laser-arc hybrid welded joint during fatigue

The deformation mechanism and life prediction model of titanium alloy laser-arc hybrid welded joint during fatigue were studied. At the relatively small maximum cyclic stresses σmax (310 MPa to 350 MPa), the fatigue cracks initiated at the interface of lamellar α′ and needle-like α′ around the pore. At σmax=370MPa, the elongated lamellar α′ around the pore promoted fatigue crack propagation, leading to the formation of secondary cracks. At σmax=390MPa, the formation of two fatigue crack initiation locations and the occurrence of secondary cracks led to the maximum fatigue damage and the minimum fatigue life. In addition, the plastic deformation mainly occurred in β at σmax=310MPa, and it transformed into the phase interface of secondary α-β and granular β-α′ at σmax=350MPa. At σmax=390MPa, the main deformation forms were the cross-slip in β and the dislocations entanglement in α′. Finally, the fatigue life prediction model was established based on the equivalent cyclic stress, and the predicted fatigue life fell within a 3-fold error band.

Critical physics informed fatigue life prediction of laser 3D printed AlSi10Mg alloys with mass internal defects

The significant scatter in high cycle fatigue life of additively manufactured metallic components presents an increasing challenge to structural integrity. This fatigue life variation is radically attributed to the differences in physical features of critical defects that lead to crack initiation. To address this issue, this paper proposes an integrated framework for identifying critical defects and predicting fatigue life using physics-informed machine learning, with a focus on the impact of 3D defect features. By employing X-ray tomography, high cycle fatigue tests, and fractography analyses on post-mortem specimens, a dataset associated with mass internal defects is first built up to correlate the spatial geometric features of critical defects with fatigue life. A kernel support vector machine is then used to formulate a critical defect identification model, aimed at identifying critical defects among numerous defects by evaluating their geometric attributes. Finally, a fatigue life prediction model is developed using a physics-informed neural network, which incorporates the influence of defect geometry on fatigue life as physical constraints in the loss function. The integrated framework demonstrates that fatigue life predictions from identified critical defects in each specimen exhibit small deviations, with the average prediction falling within twice the error bands. This study is expected to provide a valuable reference for fatigue assessment of additively manufactured components through sequential critical defect identification and fatigue life prediction.

Structural strength and fatigue analyses of large-scale underwater compressed hydrogen energy storage accumulator

Underwater compressed hydrogen energy storage (UWCHES) is a potential solution for offshore energy storage. By taking advantage of the hydrostatic pressure of deep seawater, the compressed hydrogen can be isobarically stored in underwater artificial energy storage accumulators. The accumulator should withstand high pressure and large buoyancy and possess reliable anchoring to the seabed. In this study, the structural strength analysis and fatigue life of the large-scale accumulator is conducted employing the finite element method (FEM). The dimensionless stress prediction model and dimensionless fatigue life prediction model are developed through dimensional analysis and multivariate regression analysis. The performance of the accumulator with operating water depth of 100∼300 m, gas storage volume of 1081∼10128 m³, and concrete wall thickness of 0.1∼0.63 m is investigated. The results show that with an operating water depth of 100 m, gas storage capacity of 10,128 m3, and concrete wall thickness of 0.63 m, the maximum compressive stress is 1.43 MPa (yield strength is 60 MPa) and the maximum tensile stress of the accumulator is 2.55 MPa (yield strength is 6 MPa). The design fatigue life is 106 cycles which is larger than the expected service life of 104 cycles. Therefore, the accumulator structure meets the static strength and fatigue life. As the operating water depth increases with a consistent gas storage capacity, a transition in the stress state shifts from primarily tensile stress to predominantly compressive stress. The accuracy of the dimensionless stress prediction model and the dimensionless fatigue life prediction model were verified, with maximum deviations of 10.3 % and 13.7 %, respectively. Furthermore, the anchoring factor of safety of 1.12 is achieved.

Multiscale fatigue life prediction model for CFRP laminates considering the mechanical degradation of its constituents and the local stress concentration of the matrix

In the present research, a novel approach for predicting the fatigue life of composites is proposed using a multiscale fatigue model. This model considers two critical physical phenomena that occur in CFRP laminates under tension-tension cyclic loading – the degradation of the physical properties of CFRP constituents and the localized stress concentrations within the matrix. To achieve this, Mori-Tanaka’s mean field theory, a well-known micromechanics model, is employed to consider the deterioration of material properties in various phases, which is dependent on the applied stress level. Localized stress concentration phenomena within the matrix of a composite material are common when CFRP laminate is subjected to cyclic loading. The localized stress concentration on a matrix with different numbers of fibers and orientations using representative volume elements (RVE) is investigated. Two other scales of failure criteria are defined to evaluate its constituent at the material level and ply level at the macroscale. A comparison with fatigue experiment data of AS4/3501-6 composite shows good reliability of the proposed multiscale fatigue model.

Modeling the thermo-oxidative aging effects on fatigue life of natural rubber using a continuum damage mechanics approach

Thermo-oxidative aging (TOA) significantly impacts the fatigue life of rubbers by reducing the fatigue strength and modifying the slope of the Wöhler curve, rendering traditional time-temperature superposition methods inadequate. In this work, we aim to develop a model capable of predicting the effect of TOA on the fatigue properties of rubbers. For this purpose, dumbbell specimens of natural rubber (NR) were subjected to aging in an air-vented oven at different temperatures prior to mechanical testing. Tensile tests were conducted to quantify the aging effects on a microscale network parameter known as elastically active chains (EAC) density, along with determining ultimate properties such as the stress and strain at break. By using a time-temperature equivalence and the Arrhenius shift factor, master curves were constructed for both EAC density and ultimate properties. Subsequently, the fracture properties of the aged material were predicted using an energy limiter approach whose evolution follows that of EAC density. Furthermore, a relationship between the parameters of the continuum damage mechanics (CDM) approach (specifically, the power law parameters of the Wöhler curve) and the fracture properties was proposed, resulting in accurate estimates of the fatigue life regardless of the aging conditions. Satisfactory agreement was observed between the model predictions and the data for nine aging conditions encompassing four different temperatures. The CDM approach also allows assessing the evolution of mechanical damage within the material influenced by TOA. The novelty of this approach lies in establishing a direct relationship between fatigue and fracture properties. Moreover, this is the first time to the best of our knowledge that TOA has been explicitly incorporated into a fatigue life prediction model via the evolution of the EAC density.

Study on fatigue properties with composite waveform and variable amplitude of TC21 titanium alloy pulsed laser-arc hybrid welded joints

TC21 titanium alloy is widely used in the manufacturing of key components such as aerospace industry, its welded structure in actual service is often subjected to the cyclic loading with composite waveform and variable amplitude and leads to fatigue failure. Therefore, in this work, the composite waveform and variable amplitude fatigue properties of TC21 titanium alloy pulsed laser-arc hybrid welded joints were investigated in this paper, and the micro-deformation mechanism and crack initiation mechanism of welded joints under different initial maximum cyclic stresses were revealed. The results indicated that the micro-deformation mechanism was the competition between twinning and dislocation slip in the α′ phase at low stress. When the initial maximum cyclic stress increased, the generation and slip of dislocation dominated, and the main distribution of dislocations gradually transitioned from the α′ phase to the β phase. This caused the phase transition in the β phase to the α phase, which increased the energy at the α/β phase interface and reduced the resistance of fatigue microcrack initiation accordingly, thus promoting fatigue failure. In addition, pores were the main factor of fatigue failure for welded joints. Therefore, in this work, the influence of defect size, location and shape on the fatigue life of welded joints was considered to develop a fatigue life prediction model based on the control parameter Z-parameter and equivalent stress amplitude. The predicted composite waveform and variable amplitude fatigue life of TC21 titanium alloy welded joints was within the triple error band.

Physics-informed transfer learning model for fatigue life prediction of IN718 alloy

To address the challenges posed by inadequate data and data utilization in multiple scenarios of fatigue loading, a Physics-informed Transfer Learning (PITL) model has been developed to predict the fatigue life of IN718 superalloy. Strain-controlled low-cycle fatigue tests were carried out at 400 °C with three distinct strain ratios, which were subsequently segmented for individual transfer learning tests. PITL models with significant engineering value were built by integrating transfer learning methodologies rooted in TrAdaBoost with a physics-based model that hinges on the principles of equivalent strain theory. The findings suggest that PITL models exhibit improved accuracy and greater robustness compared to both transfer learning and physics models.

Fatigue Life Prediction of a Rubber Mount Considering the Self-Heating by the Hysteresis Loss of the Rubber Material Subjected to Cyclic Loadings

Failure analysis and fatigue life prediction are important steps in the design procedure of industrial products to assure the safety and reliability of their components. A model to calculate the heat generation by the hysteresis loss in natural rubber subjected to cyclic loads in a uniaxial direction was introduced and the fatigue life prediction model of the rubber was established based on the theory of continuum damage mechanics (CDM). Dynamic mechanical analysis (DMA) was used to measure the viscoelasticity of the rubber material. The heat generation in the rubber material was calculated using a subroutine with the programming software C++ and the temperature rise in the rubber material was analyzed by a thermal finite element analysis (FEA) method. Then static mechanical FEA models of a rubber mount were established and the strain contours of the rubber mounts at various loads were calculated. The total principal strain was used as the fatigue parameter to predict the fatigue life of the rubber mount. Finally, the fatigue lives of the rubber mounts at various loads were measured on a test rig to validate the accuracy of the prediction method. The test results indicated that the fatigue lives predicted considering the temperature rise agreed better with the test results compared with those not considering the temperature rise and we suggest our fatigue prediction method should be applicable to both rubber and other types of components.

The Smith-Watson-Topper parameter and fracture surface topography relationship for additively manufactured 18Ni300 steel subjected to uniaxial variable-amplitude loading

In this paper, the association between Smith-Watson-Topper (SWT) parameter and fracture surface topography is studied in additively manufactured maraging steel exposed to variable-amplitude fatigue loading. The post-failure fracture surfaces were examined using a non-contact 3D surface topography measuring system and the entire fracture surface method. The focal point is on the correspondence between fatigue characteristics, articulate by the SWT parameter, and the fracture surface topography features, represented by areal, volume, and fractal dimension parameters. A fatigue life prediction model based on SWT and fracture surface topography factors is proposed. The presented model expresses good compliance with fatigue test results. This model can be useful for post-mortem analysis of engineering elements under variable-amplitude loading fatigue, especially for materials produced by additive manufacturing (AM).

An optimized multiaxial fatigue life prediction model based on the subcritical plane method

Since a more rigorous requirement is raised for the performance of aero-engine increases, the multiaxial fatigue problem of its components turns out to be increasingly severe. In this study, a multiaxial fatigue life prediction model is developed by using the subcritical plane method, where the effect of non-proportional additional hardening under non-proportional loading is considered by both the deflection angle of the subcritical plane and the combined parameters of strain on the subcritical plane. Moreover, the cyclic strain hardening index in the traditional model is substituted with an additional strengthening material coefficient that is characterized by the static strength of the material, which reduces the difficulty in calibrating the parameters of the multiaxial life model. The effectiveness of the proposed model is validated by comparing the predicted fatigue life with some fatigue test data, and the errors between the predicted and the testing data fall into a factor of two error bands. In general, the proposed multiaxial life prediction model simplifies the calibration of model parameters, and the prediction accuracy is still ensured, which promotes the model to facilitate engineering application.

Entropy‐based method considering nonlinear hardening effect to predict the crack propagation life of superalloy GH4169 at elevated temperature

AbstractThis paper aims to determine the relationship between thermodynamic entropy generation and fatigue crack propagation life of superalloy GH4169 at 300–650°C. The entire specimen was considered as the thermodynamic system. The plastic energy dissipation in the crack tip was obtained by finite element simulation utilizing the Chaboche nonlinear hardening model. Then the cyclic entropy generation rate (CEGR) and the accumulated entropy generation are calculated by combining simulation and experimental methods. Results show that the CEGR is a power function of the stress intensity factor range, and it is almost a constant at fatigue failure. The fatigue fracture entropy (FFE) increases as fatigue cycles at failure increase at constant temperature, but it first decreases and then increases when temperature increases from 300 to 650°C. A fatigue life prediction model based on the thermodynamic damage parameter is established and verified by comparison with experimental results and available data in the literature.

Fatigue Behaviour and Life Prediction of YSZ Thermal Barrier Coatings at Elevated Temperature under Cyclic Loads

The concentration of interfacial normal stress at the free edges of thermal barrier coatings (TBCs) can result in coating spallation. Fatigue cracking is one of the main reasons for creating free edges under complex loads. It is crucial to investigate the fatigue cracking of coatings under cyclic loads to assess potential coating failure. To address this issue, a novel model was proposed to predict the fatigue life of the YSZ topcoat under stress parallel to the interface. Firstly, this study conducted uniaxial and tensile-torsional fatigue tests at elevated temperatures on specimens with atmospheric plasma-sprayed TBCs. The test results revealed that fatigue cracks appeared in the topcoat under cyclic loads, but these cracks did not propagate into the bondcoat or substrate immediately. The number of cycles before the topcoat cracked was found to be associated with the magnitude of the cyclic load. Secondly, this study analyzed the test conditions using the finite element method. Simulations indicated that the crack direction in the topcoat under complex loading conditions was aligned with the first principal stress direction. Finally, the fatigue life prediction model of the topcoat was established based on experiments and simulations. The predicted results fell within a fourfold scatter band.

Prediction of fatigue life of laminated composites by integrating artificial neural network model and non-dominated sorting genetic algorithm

The present study introduces an artificial neural network (ANN) model coupled with the non-dominated sorting genetic algorithm (NSGA-II) to establish an advanced model for fatigue life prediction of laminated composites. The developed model was trained on a comprehensive collection of dependent experimental data of carbon/epoxy composites. This training encompassed 14 distinct cases, covering various stress concentration geometries, loading levels, and stacking sequences to ensure the model’s robustness and generality. A data augmentation technique was employed to generate similar data to enrich the datasets. The best-obtained accuracy from the model, as measured by the coefficient of determination (R2 score), was 88% and 90% for the test and validation datasets, respectively. While using the data augmentation method, these values improved to 97% and 98%, respectively. The results from the model were compared to those of more conventional regression machine learning algorithms, such as decision trees and gradient boosting. To emphasize the superior predictive accuracy of the developed model, it is essential to note that the best R2 score obtained for the test and validation datasets was 68% using conventional models.

Gear contact fatigue prediction under mixed elastohydrodynamic lubrication with ellipsoidal asperity rough surface: Experimental and numerical investigation

A three-dimensional(3D) line-contact mixed lubrication model and contact fatigue life prediction model for gears are established, considering the elastoplastic contact of ellipsoidal asperities and their interactions. The localized lubrication states across the actual 3D rough surfaces are visually depicted. The direct asperity contact pressures calculation exhibits greater accuracy and universality. The proposed mixed lubrication model can be applied accurately and efficiently to heavy-load conditions. The synergistic mechanisms of multi-scale surface integrity parameters, such as surface morphology-lubrication coupling, hardness gradient, and residual stress, on gear contact fatigue has been revealed. The results reveal that surface roughness significantly affects the competitive failure mechanism between surface-initiated micropitting and subsurface-initiated macropitting. The increase in surface compressive residual stress is more sensitive to the enhancement of near-surface and subsurface fatigue lives, compared to the increase of peak residual stress and the depth of peak residual stress. The proposed physics-based gear contact fatigue life prediction model has been validated through gear contact fatigue experiments under various operating conditions. The maximum deviation between the predicted and experimental fatigue lives is within 10%. This paper presents theoretical methodologies and data-driven support for optimizing the design and manufacturing processes of gears, specifically focusing on enhancing anti-fatigue properties.

Investigating the relationship between natural frequency and residual strength and stiffness of cross‐ply laminate under cyclic loading

AbstractPredicting the fatigue life of wind turbine rotor blades is a challenging and crucial engineering task. This study investigates the correlation between modal parameters and the degradation of residual strength and tensile modulus in the cross‐ply glass epoxy laminate [0/90]7 used in wind turbine blades under fatigue loading. The tensile and vibration characteristics were assessed, followed by constant amplitude fatigue tests at 35%, 43% and 55% of the ultimate tensile strength, with R = 0.1 and a frequency of 8 Hz. The modal analysis was performed on the cycled specimens at life fractions from 0.05 to 0.70 and residual modulus and strength were obtained. The results establish a well‐defined correlation between these residual mechanical properties and the natural frequency. Normalized residual strength, tensile modulus and natural frequency demonstrated similar behaviors during the fatigue life. An initial rapid decrease in the first tenth of the life fraction was observed, followed by minimal changes up to a life fraction of 0.7. The strong correlation between the first mode natural frequency and both the residual strength and the tensile E‐modulus provides a promising basis for developing accurate fatigue life prediction models for fiber‐reinforced composite structures. © 2024 Society of Chemical Industry.

Transfer-Learning Prediction Model for Low-Cycle Fatigue Life of Bimetallic Steel Bars

The prediction of the low-cycle fatigue life of bimetallic steel bars (BSBs) is essential to promote the engineering application of BSBs. However, research on the low-cycle fatigue properties of BSB is limited, and fatigue experiments are time-consuming. Moreover, considering that sufficient data are needed for model training, the lack of data hinders the leverage of typical data-driven machine learning, which is widely used in fatigue life prediction. To address this issue, a transfer learning framework was suggested to accurately predict the low-cycle fatigue life of BSBs with limited data. To achieve this goal, 54 data points obtained from low-cycle fatigue tests on BSBs and 264 data points of other metallic bars were collected. Source models based on artificial neural networks (ANNs) were first constructed using the collected source dataset. Then, the learned knowledge stored in the source models was transferred to the transfer models. After that, transfer models were further fine-tuned and then tested using the target dataset of BSBs. The ANN models, which were of the same structure as the transfer models but only trained with the target dataset without transferring deep features from the source models, were set as baseline models. Compared with baseline models, the constructed transfer models could be used to accurately predict the fatigue life of BSBs. Moreover, the influence of hidden layers of ANNs on accuracy was examined by comparing one-layer and two-layer transfer models. Furthermore, the influence of key parameters on fatigue life of metallic bars was evaluated by feature analysis.

Low-cycle fatigue behavior of coated and uncoated single crystal Ni-based superalloy at various temperatures

Within this work, the effect of different oxidation-resistant coatings on the low-cycle fatigue (LCF) behavior and damage mechanisms of third-generation single crystal (SC) Ni-based superalloy was investigated. Stress-controlled LCF tests of bare alloy, MCrAlY-coated and PtAl-coated thin-walled specimens were carried out at 760 °C, 850 °C and 980 °C. A unique clamping system was meticulously designed to overcome the clamping challenge in high-temperature furnaces. The cracking behaviors and failure modes were determined by advanced microstructure characterization. The results show that the coatings are detrimental to the LCF lifetime of thin-walled specimens, especially at low-medium temperatures of 760 °C and 850 °C. For bare alloy, the tip cracks nucleate from the surface oxide layer, whereas cracks in the coating-substrate system nucleate prematurely from the coating surface and propagate to the substrate. The oxygen diffusion and the frontward propagation of cracks are synergistically enhanced in the SC substrate, resulting in oxidation-assisted crack propagation and ultimate failure. Finally, A fatigue life prediction model is proposed by introducing coating-induced crack initiation and oxidation effects, where the predictive results fit well with experimental results. This study provides a comprehensive insight into the damage mechanisms and coating-induced LCF life degradation of the coating-substrate system.

Experimental study on creep-fatigue damage of hot central plant recycling asphalt mixture containing high RAP

The objective of this study is to delve into the nonlinear damage evolution mechanism of hot central plant recycling asphalt mixture under the influence of creep-fatigue interaction. A predictive model for lifespan, considering the interaction of temperature and load, was developed. Dynamic creep tests were conducted at different temperatures to assess the creep effect of the asphalt mixture under actual road temperature conditions. The investigation began by acknowledging the material's viscoelastic properties, thereby deriving the loss compliance value - a key indicator of temperature effects - using an equivalent conversion factor and a viscoelastic mechanics model. This allowed for the incorporation of creep behavior into the load system. Furthermore, to analyze the nonlinear damage mechanism and fatigue prediction mechanism, direct tensile fatigue tests were performed at different temperatures and different stress levels. The fatigue damage evolution law of hot central plant recycling asphalt mixture containing high RAP considering the creep effect was revealed, and the fatigue life evaluation method under the combined action of temperature and stress level was proposed. The results show that the fitting degree of the time-temperature conversion equation and the viscoelastic mechanics model to the test data is more than 0.9, and the derived loss compliance value better captures the creep effect. Based on the nonlinear creep-fatigue damage model, the damage law of high-RAP hot central plant recycling asphalt mixture was identified. The optimized fatigue life prediction model fully considers the combined effect of temperature and stress level and analyzes the influence parameters of the real state of pavement fatigue. The prediction accuracy is within the requirements of engineering applications. The model's predictive accuracy meets the criteria for engineering applications, enhancing the multifaceted damage coupling analysis approach for hot central plant recycling asphalt mixture and predict the service life of recycled pavements under complex conditions, establishing a groundwork for the prediction and assessment of long-lasting recycled pavements.

Experimental investigation on surface integrity and fatigue performance of Ti60 alloy under ultrasonic impact treatment

The surface integrity and fatigue properties of Ti60 alloy under different ultrasonic impact processes were investigated. Ultrasonic impact treatment (UIT) with various parameters enhanced the surface integrity of the turned specimen. The Ra values of the UIT specimens were all less than 0.380 μm, with the surface stress concentration factor (Kst) values ranging from 1.075 to 1.120. The compressive residual stress values of the UIT specimens were enhanced, leading to the localization of the maximum residual stress (σrsm) in the subsurface. Increases in surface residual stress (σs) values exceeded 21.9 % and σrsm values were improved by more than 74.4 % compared to the turned specimen. The microhardness values showed improvement after UIT, with surface microhardness (HVs) values ranging from 408.5 HV0.025 to 461.6 HV0.025. The orientation of the grains was deflected in UIT specimens, and the depths of the plastic deformation layer (hp) values experienced increases of more than 17 times. The fatigue life (Nf) value of the Comb. U2 with the best surface integrity was approximately 9328.4 % higher than that of turned and about 4362.1 % higher than that of Comb. U1 with the worst surface integrity among the UIT specimens. Observing the fatigue fractures, the crack initiation locations of the UIT specimens were relocated to the subsurface at a distance of more than 150 μm from the surface. A higher fatigue life of the specimen was achieved when there was better surface integrity and when the fatigue crack source was located further away from the surface. The crack source in the UIT specimen exhibited a large and irregular shape, and there were numerous small, evenly distributed dimples and a limited number of cleavage planes in fatigue instantaneous fracture region. Additionally, it was found that the subsurface fatigue life prediction model demonstrates superior accuracy for shorter fatigue life.