Failure Prediction Research Articles

Quantitative prediction of wear failure based on dissipative methodology under dry friction

The evaluation and prediction on failure of friction pair are important for improving equipment reliability and life span. In engineering practice, researchers often determine the severe-wear stage by monitoring changes in wear and the friction coefficient. However, existing criteria for evaluating the wear process are nebulous, leading to the ambiguous relationship between the measured parameters and the wear process which cannot be precisely quantified. In this study, based on dissipation theory, an exploration into the system entropy production rate variations throughout the wear process has been conducted. This investigation reveals that entropy production rate variations at different wear stages, providing a basis for segmenting the wear process. Through experimental methods, a clear transitional stage with substantial entropy production rate variations between the mild-wear and severe-wear stages has been discovered. These findings gave a novel method for predicting wear failure. The efficacy of this proposed approach has been validated by experiment, thus laying a theoretical groundwork for early failure prediction of critical components principally prone to wear failure. Furthermore, this work provides the foundation for the construction of visual data system in friction-wear digital twin systems.

Critical slowing down characteristics of acoustic emission for fracture instability of sandstone down-slope rock bridge under cyclic wetting and drying

Down-slope rock bridge in slopes and dangerous rock masses are often affected by cyclic wetting and drying, leading to severe geological disasters. In order to investigate the deterioration mechanism, deformation failure mode and instability precursor information of sandstone down-slope rock bridge under different cyclic wetting and drying, uniaxial compression tests combined with acoustic emission (AE) tests were conducted. The failure of the specimen experienced four stages. As the number of the cyclic wetting and drying increased, the specimen’s uniaxial compressive strength and elastic modulus decreased, and the failure mode transitioned from tensile to shear. Utilizing the critical slowing down theory, the variance and autocorrelation coefficient curves of AE multi-parameters were analyzed. The abrupt change of RA value, RA/AF value, and b value at the crack damage stress point and the abrupt change of the peak frequency lagging behind the crack damage stress point can be used as the precursors of instability. The instability precursor identified based on the peak frequency variance is closer to the specimen’s peak stress point and can be used for short-term failure prediction. The autocorrelation coefficient curve of AE multi-parameters aligned with the variance curve, indicating that with an increased number of cyclic wetting and drying, the instability precursor appeared earlier, leading to a larger failure prediction time interval.

Firm failure prediction for small and medium-sized enterprises and new ventures

AbstractNew ventures and small and medium-sized enterprises (SMEs) are the engines that drive the development of the economy, productivity, and business. However, they differ with respect to their natures, and that may affect their choices and success. This paper investigates the determinants of failure for SMEs and new ventures in Portugal by employing a logistic regression technique to develop the one-year prediction models individually over the period from 2010 to 2018. The results show that age and size always play significant roles in discriminating the failure risk of both types of firm, but the financial predictors selected in the final default prediction models for SMEs and new ventures vary. Moreover, based on financial, age, and size predictors, the SME model performs much better than that of the new venture in the classification accuracy reported. This indicates that separate treatment should be carried out while predicting the failure likelihood of SMEs and new ventures.

A Behavior Model of SiC DMOSFET Considering Thermal-Runaway Failures in Short-Circuit and Avalanche Breakdown Faults

Accurate fault simulation and failure prediction have long been challenges for SiC MOSFETs users. This paper presents a behavior model of Silicon Carbide (SiC) double-implanted MOSFET (DMOSFET), considering thermal-runaway failures in short-circuit and avalanche breakdown faults on the basis of cell-level physical processes. The proposed model can simulate the faults with extremely high accuracy and precisely predict SiC DMOSFET’s short-circuit withstand time and critical avalanche energy. By finite-element simulations, cell-level physical processes of short-circuit and avalanche breakdown faults are clarified. The mechanisms of thermal-runaway failures are deeply discussed with references to existing studies. Based on semiconductor and device physics mechanisms, the proposed model is constructed upon a traditional behavior model of SiC MOSFET with several parallel branches that are proposed to describe the thermal-runaway failures during both faults. The Cauer thermal network model is used for estimating junction temperature within it. The proposed model is constructed in Simulink, and it is validated using short-circuit and unclamped inductive switching (UIS) tests.

AutoML‐based predictive framework for predictive analysis in adsorption cooling and desalination systems

AbstractAdsorption cooling and desalination systems have a distinct advantage over other systems that use low‐grade waste heat near ambient temperature. Since improving their performance, including reliability and failure prediction, is challenging, developing an efficient diagnostic system is of great practical significance. The paper introduces artificial intelligence (AI) and an automated machine learning approach (AutoML) in a real‐life application for a computational diagnostic system of existing adsorption cooling and desalination facilities. A total of 1769 simulated data points containing data indicating a failure status are applied to develop a comprehensive AI‐based Diagnostic (AID) system covering a wide range of 42 input parameters. The paper introduces a conditional monitoring system for adsorption cooling and desalination systems. The novelty of the presented study mainly consists of two aspects. First, the intelligent system predicts the health or failure states of various components in a complex three‐bed adsorption chiller installation using the extensive input data sets of 42 different operating parameters. The developed AID expert tool, based on selecting the best from 42 models generated by the DataRobot platform, was validated on the complex, existing three‐bed adsorption chiller. The AID system correctly identified healthy and failure states in various installation components. The developed expert system is very efficient (AUC = 0.988, RMSE = 0.20, LogLoss = 0.14) in predicting emergency states. The proposed method constitutes a quick and easy technique for failure prediction and represents a complementary tool compared to the other condition monitoring methods.

ON THE PERFECTION OF THE APPROACH TO THE FORMATION OF A PROGRAM FOR INCREASING THE RELIABILITY OF FIELD PIPELINES IN OC ROSNEFT

The article describes an approach for ranking and selecting field pipelines based on the total cost of ownership of the facility for the purpose of timely repairs and minimization of accidents, as well as its software implementation. The principle of the proposed approach is to apply a technical and economic model of the total cost of ownership of a pipeline, which takes into account the costs of maintaining the pipeline over a long period of time, which include the costs of pipeline maintenance activities, pipeline failures and replacement of pipeline sections. This model allows determining which replacement year corresponds to the minimum accumulated maintenance costs, by varying the year of pipeline replacement from the start of operations to the planning horizon. The model also allows to take into account the impact of operational technological measures on the cost of pipeline ownership and the selection of the optimal year of repair.In order to take into account the level of pipeline accidents in the process of operation on the basis of operational data, a mathematical probabilistic model based on the Weibull distribution was developed. The mathematical model allows to take into account for each pipeline the influence of technical characteristics (diameter, length, wall thickness, material, coating, design features), operating conditions (type of pumped medium, technological mode of pipeline operation) and information on inspection and diagnostics. The accuracy of failure prediction using this model is at least 90 %.The approach to the formation of programs to improve the reliability of field pipelines with regard to optimal repair periods based on the total cost of ownership model is described, taking into account various objectives and constraints, among which the main ones are:- required achievable level of accidents;- acceptable amounts of financing for re-pairs;- permissible physical volumes of repairs along the length.The approach to forming reliability improvement programs described in this article has been implemented in Rosneft's corporate software system RN-SM. The implementation of this software package in production activities is being implemented in 2023–2024.

Development of a clinical prediction tool for extubation failure in pediatric cardiac intensive care unit.

Extubation failure in pediatric patients with congenital or acquired heart diseases increases morbidity and mortality. This study aimed to develop a clinical risk score for predicting extubation failure to guide proper clinical decision-making and management. We conducted a retrospective study. This clinical prediction score was developed using data from the Pediatric Cardiac Intensive Care Unit (PCICU) of the Faculty of Medicine, Chiang Mai University, Thailand, from July 2016 to May 2022. Extubation failure was defined as the requirement for re-intubation within 48 h after extubation. Multivariable logistic regression was used for modeling. The score was evaluated in terms of discrimination and calibration. A total of 352 extubation events from 270 patients were documented. Among these, 40 events (11.36%) were extubation failure. Factors associated with extubation failure included history of pneumonia (OR: 4.14, 95% CI: 1.83-9.37, p = 0.001), history of re-intubation (OR: 5.99, 95% CI: 2.12-16.98, p = 0.001), and high saturation in physiologic cyanosis (OR: 5.94, 95% CI: 1.87-18.84, p = 0.003). These three factors were utilized to develop the risk score. The score showed acceptable discrimination with an area under the curve (AUC) of 0.77 (95% CI: 0.69-0.86), and good calibration. The derived Pediatric CMU Extubation Failure Prediction Score (Ped-CMU ExFPS) could satisfactorily predict extubation failure in pediatric cardiac patients. Employing this score could promote proper personalized care. We suggest conducting further external validation studies before considering implementation in practice.

Modelling and fatigue reliability investigation on wear prediction of piston/cylinder pair based on friction fatigue mechanism

This paper aims to assess the wear reliability characteristics of the piston/cylinder pair based on the wear failure prediction using numerical and experimental techniques for the water hydraulic state. This is due to the complex forces and contact states acting in the piston, hence, making it difficult to establish a numerical model for wear failure prediction. The Hertzian contact and Elastic-Plastic mechanical equivalence are proposed in numerically modelling the friction fatigue mechanism based on asperity damage for piston/cylinder pair. An experimental analysis is conducted for the reciprocating ceramic peek-2507 steel tribopair to validate the effectiveness of the prediction model. The error between the numerical and experimental models is assessed to be less than 10% with a coefficient of determinant R2 of 0.99. The numerical wear data were scattered within the 10% confidence interval, indicating the high survivability of the predicted data. In addition, the reliability characteristics based on the volume of wear predicting in risk monitoring using the Weibull probability distribution function are proposed to model the reliability and probability of failure for the piston/cylinder pair. From the volume of wear prediction from the computed Mean-Volume-To-Failure (MVTF) = 17.8 mm3, the probability of failure based on the numerical modelling is estimated to be at 0.52. Hence, the proposed numerical model is suitable for predicting the wear volume in the water hydraulic state for various conditions.

Failure Prediction with Construction Machinery Oil Analysis Results

Failure Prediction with Construction Machinery Oil Analysis Results

Estimation of slope stability using ensemble-based hybrid machine learning approaches

Mining is one of the most daunting occupations gain the sector since it entails risk at any point in the operation. In its operation, the main focus is on slope stability. To avoid slope failures, work should be performed in line with both the regulations and the safety criteria. Slope stability is essential in mining activities owing to slope failure putting productivity and safety at risk. Prediction of slope failure is difficult because of the complexity of traditional engineering techniques. Through study, recent technologies have helped mining companies predict slope problems quickly and effectively. In this current research, an ensemble of machine learning intelligence algorithms was used to estimate and assess the Factor of Safety (FOS). In Ostapal Chromicte Mine, India, 79 experimental and failure slope occurrences were tracked to gather in-the-moment field data. The available data were split into training and testing sets at random to build algorithms. The five influenced factors such as the unit weight, the friction angle, the cohesiveness, the mining depth, as well as the slope angle used as input variables to estimate the FOS. Selected machine learning techniques such as Multiple Linear Regression (MLR), Decision Tree, Random Forest (RF), eXtreme Gradient Boosting (XGBoost) and ensemble hybrid model combining eXtreme Gradient Boosting and Random Forest (XGBoost-RF) were developed to evaluate the FOS. The validity and efficiency of created models can be evaluated using standard evaluation parameters such as coefficient of determination (R2), root mean square error (RMSE), mean square error (MSE), normalized root mean square error (NRMSE), mean absolute percentage error (MAPE) and mean absolute deviation (MAD). The most precise model to assess the FOS across all models was discovered to be the XGBOOST-RF ensemble model, which had a high R2 of 0.931, MSE of 0.009, NRMSE of 0.069, MAD of 0.037, MAPE of 3.581 and an RMSE of 0.098.

Industrial Internet of Things: A Review of Improvements Over Traditional SCADA Systems for Industrial Automation

This review article provides an overview of the potential of the Industrial Internet of Things (IIoT) to revolutionize industrial automation. The IIoT is the Internet of Things (IoT) but in an industrial context, i.e., IIoT is used more to connect machines and devices in industrial environments. The IIoT has the potential to benefit from advances in artificial intelligence, particularly machine learning and deep learning, to increase efficiency and productivity and reduce overhead costs. We provide an overview of the supervisory control and data acquisition system, a definition of IIoT, and how IIoT can offer industry greater potential for system integration to improve automation and optimization. In addition, five of the major IIoT protocols are discussed, namely, message queue telemetry transport, advanced messaging queuing protocol, constrained application protocol, data distribution service, and open platform communication unified architecture. We then identified key IIoT improvements for industrial automation. These are; efficient and low-cost systems, digital twin, machine failure prediction, real-time remote monitoring, and security. We then discussed the key research in the literature for each category. We presented some public IIoT datasets so that researchers can use them to develop new learning models to improve the security of IIoT systems. Finally, we discussed some of the limitations, recommendations, and future perspectives for developing IIoT-enabled systems.

Ductile fracture prediction of 7A62 high-strength aluminum alloy under a wide range of stress states

The modified Gurson-Tvergaard-Needleman (GTN) model is employed to predict the ductile fracture of 7A62 high-strength aluminum alloy under a wide range of stress states. Mechanical tests were conducted on specimens with different stress states within the range of −0.33 to 1.35 stress triaxiality, including tension, notched tension, compression, and shear. The results indicate that at high stress triaxialities (0.8 ∼ 1.35), the fracture mechanism is intergranular ductile fracture. Under moderate stress triaxialities (0.33 ∼ 0.8), the fracture mechanism involves a combination of intergranular ductile fracture, void growth, and shear fracture. At low and negative stress triaxialities (−0.33 ∼ 0.33), plastic instability occurs due to uneven stress distribution, leading to shear fracture. Fractography analysis reveals that the fractures occurring under tensile stress are associated with enriched Mn particles of approximately 200 nm. The modified GTN model accurately predicts the load-displacement response, and the fracture paths under various stress states exhibit good consistency with experimental results. This study provides reference for failure prediction in the engineering application of high-strength aluminum alloys.

Composite Plate Optimization: Integrating Analysis Techniques and Design Iterations for Enhanced Mechanical Performance

This essay delves into the comprehensive analysis of composite plate behavior under various loading conditions, focusing on uniaxial loading and the effects of layup design on safety factors. Through a comparative study between Classical Laminate Plate Theory (CLPT) calculations and Finite Element (FE) analysis, discrepancies in stress and strain predictions are identified, attributing them to assumptions made in each method. Furthermore, the essay explores failure prediction using Tsai-Wu criteria and Maximum Stress theory, determining the safety of laminates under different loading scenarios. Additionally, Finite Element analysis is employed to assess the impact of layup design variations on safety factors, leading to recommendations for optimized designs. The limitations of FE models compared to physical tests are discussed, along with alternative approaches such as advanced manufacturing techniques and novel fiber architectures to further enhance safety factors and alter failure modes.

Hydraulic fault prediction of integrated transmission system based on LSTM

Integrated drive-line fluids play a role in transmitting torque, providing control pressure, lubrication, and cleaning in the integrated drive-train. The increase of oil abrasive particles will accelerate the wear rate of rotating parts, while insufficient oil will lead to rapid ablation and gluing of rotating parts. According to statistics, more than 75% of hydraulic equipment failure is caused by oil pollution. Oil pollution leads to filter blockage which will cause fatal failure so that the entire transmission system cannot work. Therefore, it is extremely important to predict the fault and health management of the hydraulic system. This paper studies the oil fault of the integrated transmission system, proposes a data preprocessing method, and solves the problem of difficult extraction of deterioration index in the multichannel strong noise signal of the hydraulic system through the information fusion algorithm and the mean smoothing algorithm. According to the abnormal data judgment criterion based on the sliding window, the identification of the early fault starting point of the hydraulic system is realized, which provides a basis for the start-up fault prediction and the division of normal and degraded states. A multi-channel fusion fault prediction algorithm for hydraulic systems based on a Long-Short Term Memory (LSTM) deep network is proposed, which realizes long-term and high-precision fault prediction of hydraulic systems. Finally, the oil failure prediction algorithm of the integrated transmission system is verified by using real vehicle fault data.

An Explainable Transformer-Based Deep Learning Model for the Prediction of Incident Heart Failure

Abstract: Predicting the incidence of complex chronic conditions such as heart failure is challenging. Deep learning models applied to rich electronic health records may improve prediction but remain unexplainable hampering their wider use in medical practice. We aimed to develop a deep-learning framework for accurate and yet explainable prediction of 6-month incident heart failure (HF). Using 100,071 patients from longitudinal linked electronic health records. we applied a novel Transformer based risk model using all community and hospital diagnoses and medications contextualized within the age and calendar year for each patient’s clinical encounter. Feature importance was investigated with an ablation analysis to compare model performance when alternatively removing features and by comparing the variability of temporal representations. A post-hoc perturbation technique was conducted to propagate the changes in the input to the outcome for feature contribution analyses. Our model achieved 0.93 area under the receiver operator curve and 0.69 area under the precision-recall curve on internal 5-fold cross validation and outperformed existing deep learning models. Ablation analysis indicated medication is important for predicting HF risk, calendar year is more important than chronological age, which was further reinforced by temporal variability analysis. Contribution analyses identified risk factors that are closely related to HF. Many of them were consistent with existing knowledge from clinical and epidemiological research but several new associations were revealed which had not been considered in expert driven risk prediction models. In conclusion, the results highlight that our deep learning model, in addition high predictive performance, can inform data-driven risk factor identification.

Combined Serologic and Genetic Risk Score and Prognostication of Phospholipase A2 receptor-Associated Membranous Nephropathy.

The aim of this study was to test whether a combined risk score on the basis of genetic risk and serology can improve the prediction of kidney failure in phospholipase A2 receptor (PLA2R)-associated primary membranous nephropathy. We performed a retrospective analysis of 519 biopsy-proven PLA2R-associated primary membranous nephropathy patients with baseline eGFR ≥25 ml/min per 1.73 m 2 . The combined risk score was calculated by combining the genetic risk score with PLA2R ELISA antibody titers. The primary end point was kidney disease progression defined as a 50% reduction in eGFR or kidney failure. Cox proportional hazard regression analysis and C-statistics were applied to compare the performance of PLA2R antibody, genetic risk score, and combined risk score, as compared with clinical factors alone, in predicting primary outcomes. The median age was 56 years (range, 15-82 years); the male-to-female ratio was 1:0.6, the median eGFR at biopsy was 99 ml/min per 1.73 m 2 (range: 26-167 ml/min per 1.73 m 2 ), and the median proteinuria was 5.3 g/24 hours (range: 1.5-25.8 g/24 hours). During a median follow-up of 67 (5-200) months, 66 (13%) had kidney disease progression. In Cox proportional hazard regression models, PLA2R antibody titers, genetic risk score, and combined risk score were all individually associated with kidney disease progression with and without adjustments for age, sex, proteinuria, eGFR, and tubulointerstitial lesions. The best-performing clinical model to predict kidney disease progression included age, eGFR, proteinuria, serum albumin, diabetes, and tubulointerstitial lesions (C-statistic 0.76 [0.69-0.82], adjusted R 2 0.51). Although the addition of PLA2R antibody titer improved the performance of this model (C-statistic: 0.78 [0.72-0.84], adjusted R 2 0.61), replacing PLA2R antibody with the combined risk score improved the model further (C-statistic: 0.82 [0.77-0.87], adjusted R 2 0.69, difference of C-statistics with clinical model=0.06 [0.03-0.10], P < 0.001; difference of C-statistics with clinical-serologic model=0.04 [0.01-0.06], P < 0.001). In patients with PLA2R-associated membranous nephropathy, the combined risk score incorporating inherited risk alleles and PLA2R antibody enhanced the prediction of kidney disease progression compared with PLA2R serology and clinical factors alone.

Prediction of post-hepatectomy liver failure by preoperative gadoxetate disodium-enhanced magnetic resonance imaging

BackgroundApproximately 15% of patients experience post-hepatectomy liver failure after major hepatectomy. Poor hepatocyte uptake of gadoxetate disodium, a magnetic resonance imaging contrast agent, may be a predictor of post-hepatectomy liver failure. MethodsA retrospective cohort study of patients undergoing major hepatectomy (≥3 segments) with a preoperative gadoxetate disodium-enhanced magnetic resonance imaging was conducted. The liver signal intensity (standardized to the spleen) and the functional liver remnant was calculated to determine if this can predict post-hepatectomy liver failure after major hepatectomy. ResultsIn 134 patients, low signal intensity of the remnant liver standardized by signal intensity of the spleen in post-contrast images was associated with post-hepatectomy liver failure in multiple logistic regression analysis (Odds Ratio 0.112; 95% CI 0.023–0.551). In a subgroup of 33 patients with lower quartile of functional liver remnant, area under the curve analysis demonstrated a diagnostic accuracy of functional liver remnant to predict post-hepatectomy liver failure of 0.857 with a cut-off value for functional liver remnant of 1.4985 with 80.0% sensitivity and 89.3% specificity. ConclusionFunctional liver remnant determined by gadoxetate disodium-enhanced magnetic resonance imaging is a predictor of post-hepatectomy liver failure which may help identify patients for resection, reducing morbidity and mortality.

Technical language processing for Prognostics and Health Management: applying text similarity and topic modeling to maintenance work orders

AbstractModern manufacturing paradigms have incorporated Prognostics and Health Management (PHM) to implement data-driven methods for fault detection, failure prediction, and assessment of system health. The maintenance operation has similarly benefitted from these advancements, and predictive maintenance is now being used across the industry. Despite these developments, most of the approaches in maintenance rely on numerical data from sensors and field devices for any sort of analysis. Text data from Maintenance Work Orders (MWOs) contain some of the most crucial information pertaining to the functioning of systems and components, but are still regarded as ‘black holes’, i.e., they store valuable data without being used in decision-making. The analysis of this data can help save time and costs in maintenance. While Natural Language Processing (NLP) methods have been very successful in understanding and examining text data from non-technical sources, progress in the analysis of technical text data has been limited. Non-technical text data are usually structured and consist of standardized vocabularies allowing the use of out-of-the-box language processing methods in their analysis. On the other hand, records from MWOs are often semi-structured or unstructured; and consist of complicated terminologies, technical jargon, and industry-specific abbreviations. Deploying traditional NLP to such data can result in an imprecise and flawed analysis which can be very costly. Owing to these challenges, we propose a Technical Language Processing (TLP) framework for PHM. To illustrate its capabilities, we use text data from MWOs of aircraft to address two scenarios. First, we predict corrective actions for new maintenance problems by comparing them with existing problems using syntactic and semantic textual similarity matching and evaluate the results with cosine similarity scores. In the second scenario, we identify and extract the most dominant topics and salient terms from the data using Latent Dirichlet Allocation (LDA). Using the results, we are able to successfully link maintenance problems to standardized maintenance codes used in the aviation industry.

Topography characterization of friction pairs during the real-time wear process in dry condition

At present, the prediction of wear failure is often based on temperature, friction coefficient, vibration, and other single or a few macroscopic parameters to judge the contact condition of the friction pairs. However, there are still deviations between the measured results and the actual situation. In the digital twin system for studying friction and wear phenomena, in-situ measurement and characterization of topography of friction pairs is a challenge problem which does not exist in other procedural twin systems. This paper investigates the relationship between in-situ measured macroscopic parameters and topographical parameters. First, the dynamic relationship between friction coefficient and fractal parameters is derived from the view of a microscopic contact, and the wear periods are determined based on the dynamic changes of fractal parameters. The derived friction-fractal formula is based on mechanical deformation and the Majumdar and Bhushan rough surface contact model, avoiding the assumptions based on specific wear mechanisms, then having the universal property. Second, the experimental verification is carried on which demonstrates good correspondence between the obtained topographical parameters based on the friction-fractal formula with the experimental results. The obtained friction-fractal formula provides a theoretical basis for predicting and analyzing the wear status of friction pairs in the critical equipment and supports the real-time contact state characterization of the friction pairs in the digital twin system.

A model for failure-time data with two dependent failure modes and prediction of future failures

Often in reliability studies we observe failure-time data with two dependent failure modes. In this article, by using a bivariate Weibull model with distinct shape parameters, we present a model for reliability data with two dependent failure modes. Inferential methods for the proposed model are discussed. The efficacy of the methods of inference is assessed through a Monte Carlo simulation study, and it is observed that the model and methods perform satisfactorily. An issue of practical interest for reliability engineers is to predict field failures at a future time. Prediction methods are developed in this setting. For illustrative purposes, analysis of a real dataset on failure of a device is presented. In summary, the model and methods presented in this article provide a comprehensive treatment for analyzing reliability data with two dependent failure modes.