Fatigue Prediction Research Articles

The single machine scheduling problem with bio-mathematical fatigue constraints

Despite the adverse impacts of occupational fatigue such as accidents and injuries in the manufacturing industry, it has not been systematically examined in the literature on production scheduling. In this paper, we integrate the classic bio-mathematical fatigue prediction model from the brain science literature into the simple single machine scheduling problem with sequence-dependent setup times. Then, we formulate the problem as a mixed-integer linear programming model and propose an adaptive large neighborhood search heuristic. The effectiveness of the heuristic is numerically validated through real cases. Finally, we argue that considering bio-mathematical fatigue prediction can lead to safer production schedules, notably reducing the fatigue working hours in our real case.

Probabilistic Modelling of Fatigue Behaviour of 51CrV4 Steel for Railway Parabolic Leaf Springs

The longevity of railway vehicles is an important factor in their mechanical and structural design. Fatigue is a major issue that affects the durability of railway components, and therefore, knowledge of the fatigue resistance characteristics of critical components, such as the leaf springs, must be extensively investigated. This research covers the fatigue resistance of chromium–vanadium alloy steel, usually designated as 51CrV4, from the high-cycle regime (HCF) (103–104) up to very high-cycle fatigue (VHCF) (109) under the bending loading conditions typical of leaf springs and under uniaxial tension/compression loading, respectively, for a stress ratio, Rσ, of −1. Different test frequencies were considered (23, 150, and 20,000 Hz) despite the material not exhibiting a relatively significant frequency effect. In order to create a new fatigue prediction model, two prediction models, the Basquin SN linear regression model and the Castillo–Fernandez–Cantelli (CFC) model, were evaluated. According to the analysis carried out, the CFC model provided a better prediction of the fatigue failures than the SN model, especially when outlier failure data were considered. The investigation also examined the failure modes, observing multiple cracks for higher loads and single cracks that initiated on the surface or from internal inclusions at lower loading. The present investigation aims to provide a fatigue resistance prediction model encompassing the HCF and VHCF regions for the fatigue design of railway wagon leaf springs, or even for other components made of 51CrV4 with a tempered martensitic microstructure.

P0306 Application of 5 different machine-learning approaches to develop an artificial intelligence pathway to predict IBD-associated fatigue: Analysis of 1200 data points in 531 IBD patients.

Abstract Background Fatigue is commonly identified by IBD patients as one of the most important issues that affect their wellbeing1. Despite advanced immune therapies, 50% of patients experienced significant fatigue. The central mechanisms beyond gut inflammation that drives fatigue are not known. Methods We utilized 5 different machine-learning (ML) approaches in a combined prospective and cross-sectional multi-center cohort2 in UK (2020-24) to develop models and identify factors that can predict IBD-associated fatigue. Our ML-modelling involved 531 IBD (260 UC, 246 CD, 25 IBDU) patients, ~100 clinical variables (including CUCQ32-patient reported outcomes, seasonality, BMI, drug treatments, disease phenotype and all hospital-based blood/stool tests) and 1200 data points over time. ML-approaches were gradient boosting, random forests, logistic regression, support vector machines and deep neural networks. We incorporated various methods to identify the relative importance of clinical factors and our ML-performance. We trained our ML-fatigue on overall IBD and on patients in remission. Results Using an IBD CUCQ32 dataset involving 2500 patient responses (Figure 1), we used patient-reported data of significant fatigue experienced in 10 out of 14 days as IBD-fatiguehigh (Figure 1). Through train-validate-test steps, we found that traditional statistics and machine learning performed similarly (Random Forest AUC of 0.75; logistic regression AUC of 0.74 using R). AdaBoost, XGBoost, random forest and logistic regression models performed similarly while support vector classifiers and neural networks performed poorly (Table 1). SHapley Additive exPlanations (SHAP) analysis of the models3, based on game theory, revealed that each model prioritises different variables - the top 3 factors for XGBoost were has_active_symptoms, weight and platelets; whereas logistic regression ranked seasonality variables (autumn and winter) higher. The analysis indicates the potential of ML to integrate non-linear factors in a way that traditional statistics do not capture well. ML performance to predict IBD-fatiguehigh in patients in biochemical remission (CRP<5 and calprotectin <250) was poor, with AUCs of 0.61 for logistic regression and 0.66 for random forest. Conclusion We provide a comprehensive ML-pathway to predict IBD-associated fatigue. Despite integration of multiple clinical factors and use of interpretable ML-approaches, prediction of fatigue in IBD patients, remains suboptimal. Our data suggests a large ‘hidden’ pathobiological component and current work is in progress to integrate deep molecular data and build a clinical-scientific AI model to improve understanding of IBD-associated fatigue.

Immunotherapy-Resistant Neuropathic Pain and Fatigue Predict Quality-of-Life in Contactin-Associated Protein-Like 2 Antibody Disease.

The long-term clinical outcomes and associated prognostic factors in contactin-associated protein-like 2 (CASPR2)-antibody diseases are unknown. A total of 75 participants with CASPR2 antibodies were longitudinally assessed for disability, quality-of-life, and chronic pain. Although most symptoms improved within 6 months of treatment, neuropathic pain and fatigue were the most immunotherapy refractory, and persisted for up to 6 years. Furthermore, these two factors-but not CASPR2 antibody levels or subclasses-independently predicted worse disability and quality-of-life at 24 months. Quality-of-life varied widely for any given modified Rankin Scale score, indicating a divergence between patient and clinician assessed outcomes. Further work should study the relative importance of these measures, and the immunopathogenesis underlying intractable symptoms. ANN NEUROL 2025.

An Experimental Approach for Investigating Fatigue-Induced Debonding Propagation in Composite Stiffened Panels Using Thermographic Phase Mapping.

This work introduces an experimental approach focused on investigating fatigue-driven debonding in a composite structure designed to simulate the complexity of a typical aeronautical panel. The debonding is placed between the skin and the stringer, and the structure has been tested under fatigue compression conditions. Using lock-in thermography, the damage evolution during fatigue cycles has been detailed monitored. Indeed, thermographic phase maps obtained after a predetermined number of cycles during the whole fatigue test have been graphically analysed and have allowed us to obtain an accurate measurement of the delaminated area extent and shape. Our approach advances the understanding of damage propagation in composite materials, contributing to the development of damage-tolerant structural designs and supplying valuable data to validate numerical fatigue prediction models. Furthermore, the use of non-destructive testing techniques, such as thermography, has been found crucial for accurately quantifying the extent and the shape of the debonding after a given number of fatigue cycles.

A simplified strain energy density approach for multiaxial fatigue predictions

A simplified strain energy density approach for multiaxial fatigue predictions

C-Reactive Protein is the Main Factor for Prediction of Fatigue in Preeclampsia Women

Preeclampsia is associated with many neuropsychiatric symptoms including fatigue. Many biomarkers are correlated with fibro fatigue (FF) scores. The present study aims to examine the changes in the advanced-glycated end-products (AGEs) and its receptor (RAGE), Vascular endothelial growth factor (VEGF) and its receptor (VEGFR1), C-reactive protein (CRP) and lipid profile and atherogenic indices in preeclampsia as predictors for preeclampsia patients with higher FF score. Preeclampsia patients were divided according to the FF score into high-FF (FF≥25) and low-FF (FF<25). The biomarkers and FF scores were measured in the patient groups and compared with healthy pregnant controls. PE women had higher levels of AGEs, RAGE, VEGF, and VEGFR1 than control women, and the rise is greater in the high-FF PE group compared with the low-FF group. PE women also have dyslipidemia and low-grade inflammation (high serum CRP), which rise with FF score. The multivariate generalized linear model (GLM) analysis showed no significant effect of the covariates (age, gestational age, gravidity, age of onset, and duration of PE) on the measured biomarkers. The study found significant differences in all parameters for high-FF PE patients (Partial η² = 0.727). CRP is the only receiver-operating characteristics (ROC) variable distinguishing high-FF PE women from low-FF PE women. The high-FF scores in PE are related to the lipid profile and atherogenic indices as well as dyslipidemia and low-grade inflammation. PE patients with high-FF can be significantly predicted by high-CRP levels.

A modified damaged stress model for fatigue life prediction based on load interaction

This study proposes a modified Damage Stress Model (DSM) for fatigue life prediction, addressing the variability of ultimate stress during damage transfer under multi-stage stress. The model incorporates Generalized Ultimate Stress (GUS), influenced by the ultimate stress, the sum of the previous-stage fatigue damage, and the ratio of adjacent stress. Comparative analysis with the Miner, DSM, and Manson models, using experimental data from multiple materials and stages, shows that the proposed model outperforms others, rectifying limitations of the original DSM and providing more accurate fatigue predictions.

Structural analysis and fatigue prediction of harrow tines used in Canadian prairies

The Canadian prairies are renowned for their substantial agricultural contributions to the global food market. Harrow tines are indispensable in farming equipment, especially for soil preparation and weed control before planting crops. During operation, these tines are exposed to repetitive cyclic loading, which eventually causes fatigue failure. Commercially available three different harrow tines named 0.562HT, 0.625HT, and 0.500HT undergo an experimental fatigue evaluation and are validated through Finite Element Analysis (FEA). Fatigue life estimation for different deflections under various real-field deflections was carried out where 0.562HT showed groundbreaking life compared with others. The study results showed that the fatigue life is highly dependent on geometry, number of coils, pitch angle, leg length, and coil diameter. The 0.354HT model, developed to investigate the effect of wire diameter, closely resembles the 0.500HT model. The harrowing ability of the four different harrow tine models against identical deflections has been analyzed. Experimental fractured surfaces went through morphological investigation. This research has an impeccable impact on prairies’ agricultural acceleration by saving time and mitigating unpredictable fatigue failure often faced by farmers. Even the observed failure phenomena can serve as motivation to develop more reliable and durable harrow tines, which could increase agricultural efficiency.

Guest Editorial of Virtual Special Issue: Artificial Intelligence‐Assisted Fatigue Prediction of Engineering Materials and Structures

Guest Editorial of Virtual Special Issue: Artificial Intelligence‐Assisted Fatigue Prediction of Engineering Materials and Structures

Experimental facility dedicated to detection and prediction of penstock fatigue induced by pressure oscillations

Abstract Hydroelectric powerplants play a vital role in the electricity production mix, especially during the ongoing energy transition towards renewable sources. However, they face operational challenges due to harmful stress loading of various components. Ensuring safe operation remains essential for people’s safety, electricity supply, and costs avoidance. The study focuses on penstocks and pipelines material damaging caused by fatigue, crack initiation and propagation. These critical components are often several decades old and expensive to refurbish or replace. Cycling loading, induced by pressure oscillations during start-stop or transient operations, accelerates material fatigue. Steel lining corrodes over time, and welds may contain original defects. To address this, a new testing facility that generates controlled cyclic pressure oscillations using the water hammer effect has been built. This specific closed-loop circuit allows accelerated fatigue testing of material probes, providing insights into crack initiation and propagation. The test rig operating principle is described while its 1D numerical model is introduced and validated with measurement data. The prediction of crack incipience and rupture of a tubular specimen with a pre-machined longitudinal weakening is in the end compared with the results of a fatigue test conducted on the test bench.

Exploration on flutter mechanism of a damaged transonic rotor blade using high-fidelity fluid–solid coupling method

Structural damage in turbomachinery is a primary origin of aeronautic accidents, which is receiving increased attention. This study is thus focused on the aeroelastic analysis of damaged blades, including the onset of flutter and underlying mechanisms. First, a high-fidelity fluid–solid coupling system is established with computational fluid dynamics (CFD) and computational structural dynamics (CSD) technologies, via which the dynamic aeroelastic analysis is conducted based on static aeroelastic deformation. Second, a damaged rotor blade is parametrically modelled with variable damage levels, extents, and positions. Finally, the modal identification method of spectral proper orthogonal decomposition (SPOD) is applied to observe flow details and provide physical insight into the flutter mechanism for damaged blades. Numerical analysis finds that there is a critical damage level below which the aeroelastic stability is positively improved with increasing damage level; otherwise, a significant loss of stability is induced. The damage location and extent further affect this critical damage level and the change rate crossing the threshold. The simulation with CFD/CSD finds that the high pressure near the trailing edge induced from boundary layer separation suppresses vibrations in stable conditions, but motivates vibrations during flutter, which is because of the high-pressure spread to nearing blades. SPOD modes reveal that high-frequency disturbances with large scale are primary factors inducing flutter, which is further stimulated by the high-order disturbances with small scale. This study provides a crucial foundation for the fatigue prediction for rotor blades in service and the optimisation design for high-performance turbomachinery in the near future.

Visual fatigue prediction using classification model based on physiological responses of occupants under office lightings

Appropriate management of visual fatigue is crucial for eye health as it significantly impacts life quality throughout an individual's life cycle. The shift to digital tasks in modern office environments has increased occupants' exposure to visual fatigue, leading to various social problems. This study aimed to develop visual fatigue prediction models based on physiological responses and classification algorithms. Experiments were conducted to collect physiological responses and subjective visual fatigue under various lighting environments. Collected data was refined and reorganized as predictor variables by different time windows and target variables by scales. Then, visual fatigue prediction models were developed using supervised machine learning algorithms (i.e., artificial neural network, support vector machine, gradient boosting machine, and random forest). And improved by feature selection and adding subject-label variables. The resulting two-scale and three-scale visual fatigue prediction models demonstrated an average performance of 93.73 % and 94.64 % respectively. This research can contribute to reducing societal costs and enhancing productivity by proposing visual fatigue prediction models that help manage eye health in office environments.

A novel prediction method for rolling contact fatigue damage of the pearlite rail materials based on shakedown limits and rough set theory with cloud model

Evaluation and prediction of wheel-rail rolling contact fatigue (RCF) damage can provide important theoretical guarantees for the service safety of wheels and rails and help make maintenance easier to plan. This study aims to develop a novel method for evaluating and predicting RCF damage of the pearlite rail materials with various initial shear yield strengths (ke). Based on the rough set mathematical theory incorporated within the cloud model of the comprehensive evaluation index (P0/ke*μt), a novel evaluation and prediction method for RCF damage states of various pearlite rail materials was constructed using the shakedown limits for pearlite rail materials with various initial shear yield strengths. To develop this novel prediction method, different evaluation indices for RCF damage states were designed. A comprehensive certainty approach was introduced to quantitatively analyze the actual measured values of distinct evaluation indices that corresponds to different RCF damage states, wherein the maximum value rule was applied. Moreover, the prediction results were confirmed after further verifying using the actual measured value of the P0/ke*μt. The results indicated that the predicted results were consistent with the test outcomes. The key feature of this prediction method was that it involved both the intrinsic shear yield strength of evaluated pearlite rail materials and wheel-rail rolling contact variables. On the basis of the two-dimensional classical shakedown map, a three-dimensional shakedown limit diagram for rail materials with varying initial shear yield strengths was further constructed using this novel prediction method. The three-dimensional shakedown limit diagram featured an inclined curved surface. As the initial shear yield strength of the pearlite rail materials increased, the curved surface tilted downward, indicating that an increase in the initial ke value of the pearlite rail materials could result in a lower shakedown limit.

Dynamic assessment of visual fatigue during video watching: Validation of dynamic rating based on post-task ratings and video features

People watching video displays for long durations experience visual fatigue and other symptoms associated with visual discomfort. Fatigue-reduction techniques are often applied but may potentially degrade the immersive experience. To appropriately adjust fatigue-reduction techniques, the changes in visual fatigue over time should be analyzed which is crucial for the appropriate adjustment of fatigue-reduction techniques. However, conventional methods used for assessing visual fatigue are inadequate because they rely entirely on post-task surveys, which cannot easily determine dynamic changes. This study employed a dynamic assessment method for evaluating visual fatigue in real-time. Using a joystick, participants continuously evaluated subjective fatigue whenever they perceived changes. A Simulator Sickness Questionnaire (SSQ) validated the results, which indicated significant correlations between dynamic assessments and the SSQ across five items associated with symptoms associated with visual discomfort. Furthermore, we explored the potential relationship between dynamic visual fatigue and objective video features, e.g., optical flow and the V-values of the hue/saturation value (HSV) color space, which represent the motion and brightness of the video. The results revealed that dynamic visual fatigue significantly correlated with both the optical flow and the V-value. Moreover, based on machine learning models, we determined that the changes in visual fatigue can be predicted based on the optical flow and V-value. Overall, the results validate that dynamic assessment methods can form a reliable baseline for real-time prediction of visual fatigue.

Research on the structural performance and fatigue life analysis of commercial vehicle stabilizer bars

The stabilizer bar is a crucial component of the car suspension, and its fatigue failure can significantly compromise the safety of driving. To verify whether the stabilizer bar meets the target fatigue life. This paper initially constructs a finite element model of the lateral stabilizer bar, computes the maximum Mises stress at critical sections, and formulates the stress at these points using mechanical principles. The theoretical results validate the effectiveness of the finite element model. Subsequently, the lateral stiffness of the stabilizer bar is calculated, and its accuracy under vertical load is validated using stiffness bench tests. Based on this, the rain flow matrix for one test cycle within the mixed segment of the test site is converted into a single-amplitude sine wave with a constant frequency using the equal damage method. Subsequently, bench fatigue tests and fatigue simulation analyses were conducted separately to determine the test and simulation lifespans under sinusoidal loads. The accuracy of fatigue simulation analysis was verified by comparing the test and simulation life. Random loads from road tests were utilized as input for the fatigue simulation, predicting the fatigue life of the stabilizer bar. This paper employs a combination of fatigue finite element analysis and fatigue testing to predict the fatigue life of stabilizer bars. This fatigue prediction scheme that combines finite element analysis and experimentation can be extended to other component systems, thereby shortening the development cycle of new products.

Fatigue prediction of wind turbine main bearing based on field measurement and three-dimensional elastic drivetrain model

Fatigue of main bearing significantly affects the reliability of drivetrain system (DTS) and strongly depends on the DTS configuration. Main bearing failure rate is high for Three-Point Mount wind turbine within the designed life. In this study, a three-dimensional elastic DTS model is established and validated with field measurements. The predicted shaft moments and torque arm motions show good agreements with observations. A numerical pounding model of main bearing is then proposed to evaluate the load factor due to pounding. Finally, the formulas in ISO 281 for predicting main bearing rating life L10 are modified by introducing two new parameters: life ratio and load factor. The predicted rating life L10 matches well with the onsite records.

Wearable network for multilevel physical fatigue prediction in manufacturing workers.

Manufacturing workers face prolonged strenuous physical activities, impacting both financial aspects and their health due to work-related fatigue. Continuously monitoring physical fatigue and providing meaningful feedback is crucial to mitigating human and monetary losses in manufacturing workplaces. This study introduces a novel application of multimodal wearable sensors and machine learning techniques to quantify physical fatigue and tackle the challenges of real-time monitoring on the factory floor. Unlike past studies that view fatigue as a dichotomous variable, our central formulation revolves around the ability to predict multilevel fatigue, providing a more nuanced understanding of the subject's physical state. Our multimodal sensing framework is designed for continuous monitoring of vital signs, including heart rate, heart rate variability, skin temperature, and more, as well as locomotive signs by employing inertial motion units strategically placed at six locations on the upper body. This comprehensive sensor placement allows us to capture detailed data from both the torso and arms, surpassing the capabilities of single-point data collection methods. We developed an innovative asymmetric loss function for our machine learning model, which enhances prediction accuracy for numerical fatigue levels and supports real-time inference. We collected data on 43 subjects following an authentic manufacturing protocol and logged their self-reported fatigue. Based on the analysis, we provide insights into our multilevel fatigue monitoring system and discuss results from an in-the-wild evaluation of actual operators on the factory floor. This study demonstrates our system's practical applicability and contributes a valuable open-access database for future research.

Predicting fatigue life of additively manufactured lattice structures using the image-based Finite Cell Method and average strain energy density

A well-known Achilles' heel of laser powder bed fusion (L-PBF) additive manufactured lattice structures is the difficulty in predicting fatigue properties. The presence of manufacturing-induced defects significantly affects the fatigue resistance of the porous component and must be accurately captured by predictive models. To tackle this challenge, the as-built geometry of the lattice needs to be modeled, which introduces another challenge on the computational front. For the first time, a model based on the computer tomography (μ-CT) reconstruction of the as-built lattice geometry is simulated with the efficient finite cell method and combined with the average strain energy density (ASED) to obtain accurate fatigue predictions. This work presents a workflow for determining the fatigue resistance of lattice metamaterial, followed by a case study for method validation. The validation shows a good agreement between the predicted fatigue life and the experimental results.

Effort of damage parameter in assessment of low cycle fatigue

A low-cycle fatigue (LCF) analysis is one of the main design stages for highly loaded structural elements used in various applications. For this analysis, it is necessary to determine the values of local stresses and deformations, taking into account both elastic and plastic regions in the zones of stress concentration. This study presents and assesses the engineering methods used for prediction of low-cycle fatigue in structural elements. For zones of stress (strain) concentration, the Neuber-Makhutov method for LCF, taking into account the type of material stress-strain diagrams, is employed. The concept of distributed damage, based on the main ideas of the continuum damage mechanics of Kachanov-Rabotnov, was used. An approach employing the damage parameter for assessment of damage accumulation in LCF in highly loaded areas of structural elements is presented.