Failure Modes Research Articles

Engineered Tendon-Fibrocartilage-Bone Composite With Mechanical Stimulation for Augmentation of Rotator Cuff Repair: A Study Using an In Vivo Canine Model With a 6-Month Follow-up.

Rotator cuff repair augmentation using biological materials has become popular in clinical practice to reduce the high retear rates associated with traditional repair techniques. Tissue engineering approaches, such as engineered tendon-fibrocartilage-bone composite (TFBC), have shown promise in enhancing the biological healing of rotator cuff tears in animals. However, previous studies have provided limited long-term data on TFBC repair outcomes. The effect of mechanical stimulation on TFBC has not been explored extensively. To evaluate functional outcomes after rotator cuff repair with engineered TFBC subjected to mechanical stimulation in a 6-month follow-up using a canine in vivo model. Controlled laboratory study. A total of 40 canines with an acute infraspinatus (ISP) tendon transection model were randomly allocated to 4 groups (n =10): (1) unilateral ISP tendon undergoing suture repair only (control surgery); (2) augmentation with engineered TFBC alone (TFBC); (3) augmentation with engineered TFBC and bone marrow-derived stem cells (BMSCs) (TFBC+C); and (4) augmentation with engineered TFBC and BMSCs, as well as mechanical stimulation (TFBC+C+M). Outcome measures-including biomechanical evaluations such as failure strength, stiffness, failure mode, gross appearance, ISP tendon and muscle morphological assessment, and histological analysis-were performed 6 months after surgery. As shown in the mechanical test, the TFBC+C+M group exhibited higher failure strength compared with other repair techniques. The most common failure mode was avulsion fracture in the TFBC+C+M group, but tendon-bone junction rupture was observed predominantly in different groups. Engineered TFBC with mechanical stimulation showed over 70% relative failure strength compared with normal ISP, and the other groups showed about 50% relative failure strength. Histological analysis revealed less fat infiltration and closer-to-normal muscle fiber structure in the mechanical stimulation group. This study provides evidence that mechanical stimulation of engineered TFBC promotes rotator cuff regeneration, thus supporting its potential for rotator cuff repair augmentation. This study provides valuable evidence supporting the use of a novel tissue-engineered material (TFBC) in rotator cuff repair and paves the way for advancements in the field of rotator cuff regeneration.

Intelligent Fault Diagnosis in Industrial Machinery: Leveraging AI with LSTM Autoencoder for Enhanced Fault Detection

Machinery Fault Detection (MFD) is an important process in contemporary industrial systems, where it predicts possible physical failures before they lead to a serious problem. This uses multiple technologies to monitor machine statuses (algorithms, data gathering systems and sensors) Using a servo-motor driven actuator for deployment, the Locking Mechanism is pre-assembled into an OEM ATE and will enable predictive failure mode identification (via monitoring and warnings of operational parameters i.e., vibration, temperature or auditory signals in-built to MFD systems) leading to Prophylactic maintenance before critical bottlenecks can occur. The dataset we used in our study was collected from Kaggle and it is called the SpectraQuest Machinery Fault Simulator (MFS) Alignment-Balance-Vibration (ABVT). We used LSTM Autoencoder, KNN, SVM and DNN to analyzed the data. Our LSTM Autoencoder model was very accurate and achieved a precision, recall, accuracy and F-score of 99%. We worked on very large scale datasets. It will help the system detect faults and predict their evolution over time, so you save maintenance costs and increase production in your factory. More research on the practical efficiency of these models in real-time across different industrial settings can create a path towards improved and scalable MFD solutions.

Web-Based Maintenance Prediction of Machine Conditions and Failure Modes Using Machine Learning

Machine failure may result in unplanned downtime, production losses, safety risks, and increased costs. Hence, predicting machine failures is a significant challenge faced by many industries. Predictive Maintenance (PdM) techniques can help mitigate these risks by predicting machine failures before they occur. This research, therefore, develops two supervised Machine Learning (ML) algorithms to predict failure conditions and the associated failure modes for a milling machine. Six machine features were investigated. The ML models were then developed using a four-stage process including data pre-processing, training, evaluation, and deployment. Several ML algorithms were applied then the results were evaluated using accuracy, recall, precision, and [Formula: see text]1-score. The results showed that the Random Forest algorithm was the most effective in predicting machine conditions and failure modes of the milling machine. A web application for PdM was finally developed and tested for the PdM of the milling machine. In conclusion, the developed web application including ML algorithms can support the effective PdM for the machine, which leads to enhancing its availability and improving productivity. Future research considers adopting ML algorithms for predicting machine conditions and failure types of other machines.

A service failure assessment model for smart product consumption experience based on customer perception

Customer perception is an important consideration factor in evaluating the quality of human-computer interaction services. Sustainable user experiences and marketing strategies can be created by analyzing customer perception. By understanding consumer satisfaction with product services in the customer perception area, appropriate product service failure prevention strategies can be formulated. A service failure evaluation model is proposed in this study, which considers the customer tolerance area to accurately evaluate consumers’ behavioral experiences from purchasing to using products. The concept of tolerance area is introduced, and a combination of the fuzzy Failure Mode and Effect Analysis (FMEA) method and the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method is used to construct a human-computer interaction service failure evaluation model. Potential service failure factors of smart speakers are accurately evaluated by this model, and these service failure factors are ranked within the tolerance area. The research identifies voice misinterpretation and signal connectivity issues as the primary risk factors impacting the quality of human-computer interaction for smart speakers. The application of this method not only enhances the evaluation of smart speaker human-computer interaction services quality but also aids in the precise identification and prioritization of critical failure modes. The proposed service failure prevention strategies can reduce consumer dissatisfaction and provide innovative references for smart product design and marketing. The findings bolster empirical evidence for service failure prevention strategies in smart products and pave the way for novel perspectives on enhancing the quality of human-computer interaction services.

Transdermal Drug Delivery Systems (TDDS): Recent Advances and Failure Modes.

Transdermal drug delivery systems (TDDS), commonly refered to as "patches", present a nonintrusive technique to provide medication without the need for invasive procedures. These products adhere to the skin and gradually release a specific dosage of medicine at a defined rate into the bloodstream. Compared with other methods of drug delivery, TDDS offer benefits such as reduced invasiveness, convenience for patients, and avoidance of the metabolic processes that occur when drugs are orally consumed. Throughout time, TDDS have been used to provide medications for various medical conditions (such as nicotine, fentanyl, nitroglycerin, and clonidine), and their potential for delivering biologics is currently being explored. This review investigates the current literature on the drug delivery efficacy of medical TDDS through the transdermal route. Additionally, the review addresses potential risks and failure modes associated with TDDS design and development as well as strategies for mitigating such risks. A thorough understanding of failure modes provides a blueprint to mitigate failure and produce high-quality efficacious therapeutics.

Eccentric compression behaviors of iron tailings and recycled aggregate concrete-filled steel tube columns

This paper presents a sustainable method for using iron ore tailings (IOTs) sand and recycled coarse aggregate (RCA) in concrete and concrete-filled steel tube (CFT) columns. The study first examines the impact of RCA replacement ratios (0 %, 50 %, 100 %) on concrete's mechanical properties, finding that both compressive and tensile strengths decrease with higher RCA content due to the negative effects of old interfacial transition zones and adhered cement in RCA. Then, twelve iron tailing and recycled aggregate concrete-filled steel tube (ITRAC-CFT) columns were tested under eccentric compression to assess the effects of varying RCA replacement ratios, eccentricity ratios (0.3, 0.5), and slenderness ratios (12.12, 19.05). Results indicate that RCA ratios have minimal impact on failure modes, ultimate bearing capacity, and stiffness, although ductility increases with higher RCA content. As slenderness ratios increase, the ultimate bearing capacity decreases by 5.3 % and 6.8 % for eccentricities of 0.3 and 0.5, respectively. The study further validates these findings through the finite element method (FEM), showing strong agreement between experimental and predicted results, average value, standard deviation, and coefficient of variation of Nu/NFE were 1.02 %, 0.04, and 3.92 %, respectively. Finally, a comparison with existing design codes (EC4, AISC 360, and GB 50396) reveals that AISC 360 provides the most accurate predictions of the ultimate bearing capacity of ITRAC-CFT columns. The research concludes that the comprehensive use of IOTs sand and RCA in CFT structural concrete is a highly viable and eco-friendly solution, offering potential benefits for sustainable construction practices.

Embers of autoregression show how large language models are shaped by the problem they are trained to solve

The widespread adoption of large language models (LLMs) makes it important to recognize their strengths and limitations. We argue that to develop a holistic understanding of these systems, we must consider the problem that they were trained to solve: next-word prediction over Internet text. By recognizing the pressures that this task exerts, we can make predictions about the strategies that LLMs will adopt, allowing us to reason about when they will succeed or fail. Using this approach-which we call the teleological approach-we identify three factors that we hypothesize will influence LLM accuracy: the probability of the task to be performed, the probability of the target output, and the probability of the provided input. To test our predictions, we evaluate five LLMs (GPT-3.5, GPT-4, Claude 3, Llama 3, and Gemini 1.0) on 11 tasks, and we find robust evidence that LLMs are influenced by probability in the hypothesized ways. Many of the experiments reveal surprising failure modes. For instance, GPT-4's accuracy at decoding a simple cipher is 51% when the output is a high-probability sentence but only 13% when it is low-probability, even though this task is a deterministic one for which probability should not matter. These results show that AI practitioners should be careful about using LLMs in low-probability situations. More broadly, we conclude that we should not evaluate LLMs as if they are humans but should instead treat them as a distinct type of system-one that has been shaped by its own particular set of pressures.

Availability modeling of competing risk system with multiple failure modes incorporating hybrid hazard rate model

Availability modeling of competing risk system with multiple failure modes incorporating hybrid hazard rate model

Tough Monolayer Silver Nanowire-Reinforced Double-Layer Graphene.

Mixed-dimensional nanomaterials composed of one-dimensional (1D) and two-dimensional (2D) nanomaterials, such as graphene-silver nanowire (AgNW) composite sandwiched structures, are promising candidates as building blocks for multifunctional structures and materials. However, their mechanical behavior and failure mechanism have not yet been fully understood. In this work, we have performed integrated experimental, theoretical, and numerical studies to explore the performance and failure modes of graphene-AgNW composite under tensile and impact loading conditions. In situ tensile tests using a nanoindenter, implemented with a push-to-pull device and a laser-induced projectile impact test system, are used to shed light on load-bearing mechanisms in graphene-AgNW composites. Multiple failure modes have been observed in both experimental setups and analyzed with numerical and theoretical models. Results show that in the tensile loading the distribution of AgNW, as characterized by the effective free length, is the key parameter determining the failure mode. As for the impact failure scenarios, compared with failure modes observed in pure graphene cases, the mechanical reinforcing effect of AgNW will transform the failure mode from a scattered tensile fracture along radial directions to a shear failure that is constrained in a relatively local domain. Theoretical analysis using shear lag modeling, Timoshenko plate theory, molecular dynamics modeling, and finite element modeling approaches are adopted to further establish the failure modes.

Eccentric load capacity of ultra-high strength concrete-filled steel tubular lattice short columns

To explore the behavior of ultra-high strength concrete filled steel tubular (UHSC-FST) lattice short columns under eccentric compressive loads, experimental analyses on four specimens and 276 finite element models using ABAQUS were conducted. The study focused on the effects of steel tube wall thickness, concrete strength, steel tube strength, and load eccentricity on failure modes, load-displacement curves, and ultimate load-bearing capacity. Results showed that for columns with minor eccentricity-induced compressive failure, the reduction coefficient of ultimate load capacity has a weak correlation with the parameters. However, for columns with significant eccentricity-induced tensile failure, the reduction coefficient increases with steel tube thickness and strength and decreases with concrete strength. The center-to-center distance of the columns has a minor effect. Based on these findings, a comparative analysis of existing standards was conducted, leading to the development of an accurate method for calculating the ultimate load-bearing capacity of eccentric CFST lattice columns. This method is applicable to various cross-sectional dimensions, material strengths, and steel tube wall thicknesses, with calculation errors within 10 %, ensuring it meets engineering precision requirements.

Axial compression behaviour of square steel tubes encased by and filled with high-strength recycled aggregate concrete

The axial compression behaviour of square steel tube (SST) columns encased with recycled aggregate concrete (RAC) and filled with RAC (RAC-encased RACFSST) columns was investigated. Using high-strength concrete, five specimens were tested under axial compression until failure. The experimental parameters included the coarse aggregate type, stirrup spacing, wall thickness of the SST and the existence of cross-tie bars in the SST. The failure mode, load–displacement curves, load–strain curves and ductility of the composite columns were obtained. The impacts of the various parameters on the mechanical performance of the RAC-encased RACFSST columns were evaluated. The failure modes of the composite columns were comparable. The inclusion of RAC enhanced the load-bearing capacity (LBC) of the column, but negatively affects its ductility and deformation capacity. Increasing the wall thickness of the SST and reducing the stirrup spacing significantly improved the axial compressive performance of the columns, as mainly reflected in the axial compressive capacity and ductility, respectively. Cross-tie bars inside the SST improved the LBC and deformation capacity. The axial bearing capacities of the columns and those in other studies were determined using existing codes. The computed results were very consistent with the experimental data.

Experimental investigation on the frost resistance of RCC layers with various interface treatments

This research investigates the influence of various interface treatment methods on the frost resistance performance of Roller Compacted Concrete (RCC) layers. Freeze-thaw cycle tests were conducted on RCC with interfaces treated using cement mortar, expansion agent mortar, and nano-SiO2 mortar. The study evaluated shear failure modes, shear strength, crack expansion, pore structure, and liquid uptake of the RCC interface after freeze-thaw cycles under different treatment methods. Results reveal that expansion agent mortar and nano-SiO2 mortar improve the shear strength of RCC after freeze-thaw cycles more effectively than cement mortar. The freeze-thaw cycles increase the width and expansion rate of cracks at the RCC interface during the shear process, while mortar treatments help delay crack development. Additionally, the pore volume at the RCC interface gradually increases, and the uniformity of pore distribution decreases under freeze-thaw cycles. During liquid uptake, mortar treatments slow down the upward migration of liquids by improving the pore structure at the interface. The liquid uptake of RCC exhibits a distinct linear relationship with its shear strength and pore structure. As liquid uptake time increases, water first fills larger pores before gradually entering smaller ones. Freeze-thaw cycles cause an increase in pore volume and enlargement, which leads to greater liquid uptake. Based on the damage mechanism of the RCC interface under freeze-thaw conditions, a model of shear strength degradation due to freeze-thaw cycles was established, aligning well with experimental results. Mechanism analysis indicates that nano-SiO2 and expansion agents enhance frost resistance by filling interface pores and cracks through reactions with certain compounds, thereby improving the microstructure.

Reliability analysis of subsea manifold system using FMECA and FFTA

Subsea manifold system is a complex system that occupies a pivotal role in contemporary ocean engineering and has a significant impact on the safety of marine resource exploitation. Reliability technology plays a significant role in ensuring the safe operation of the subsea manifold system. To perform a comprehensive analysis of the reliability of complex systems, a combination method of FMECA-FFTA (Failure Modes, Effects and Criticality Analysis - Fuzzy Fault Tree Analysis) is introduced in this study. Firstly, FMECA is used to accomplish a qualitative analysis of system reliability considering multifactorial failure modes, which included analyzing potential failure modes, causes of system failure, and evaluating the degree of hazard to the system through a risk matrix diagram. Then, FFTA is applied to build a fault tree model to divide the system into “system-subsystem-component” and determine the minimal cut sets for quantitative analysis of system reliability. In addition, fuzzy set theory is incorporated to improve the accuracy of handling uncertainty in quantitative reliability analysis. Finally, a qualitative and quantitative reliability analysis is conducted by using FMECA-FFTA method for subsea manifold system. The failure modes of the subsea manifold system are clearly identified, including high-risk modes such as external leakage, medium-high-risk modes such as fail to close/lock on demand, and medium-risk modes such as leakage of critical location, plugged, and effective measures should be taken to focus on preventive protection and regular testing for the high risk, medium-high risk and medium risk modes.

Experimental and numerical investigation of low‐velocity impact behavior and failure mode of shear deficient RC beam strengthened with CFRP strips

AbstractIt is well‐known that shear failure is a collapse mechanism that is the riskiest, fastest, and catastrophic and occurs with no visible signs of damage or prior warning for RC beams. Therefore, to prevent brittle shear failure, the RC beams should include sufficient shear reinforcement, such as stirrups, and be designed to have sufficient shear capacity. However, the RC beams' shear capacity becomes inadequate for various reasons. One of these reasons may be the acting of the impulsive impact load, which is uncommon and disregarded in the design phase on the RC beams. An experimental program was conducted to examine the impact behavior and failure mode of shear‐deficient RC beams in the scope of the present study. Besides, it aims to investigate the effectiveness of the strengthening method using CFRP strips in improving the general behavior, failure mode, and performance of shear‐deficient RC beams exposed to impact load. The time histories of the accelerations, displacements, impact loads, and strains in the CFRP strips were measured. They were interpreted how they are affected by experimental variables examined in the experimental study. Furthermore, the finite element model of the specimens was generated in the LS‐DYNA software, and the experimental and numerical results were compared by performing finite element analysis in terms of failure modes and general behavior.

A-188 Modal Failure and Effects Analysis, FMEA methodology, as a strategy to increase patient safety

Abstract Background Patient safety during the clinical care process is considered a priority in healthcare. Failure Mode and Effects Analysis (FMEA) is a proactive and systematic method of process evaluation and is used to identifying where and how each phase of the analytical process could fail allowing the detection of the causes and the effects they produce on the quality of the results that are issued. A regular use of this tool makes possible to identify the most critical steps, propose improvement strategies and evaluate their impact. The aim of the study is to evaluate applicability of the FMEA tool to identify errors in the pre analytical, analytical, and post analytical phase of a clinical laboratory as well as the main causes and effects, to reduce them by proposing improvement strategies Methods For each phase into which the laboratory activity is divided, the frequency and type of errors that are made and their final consequences on the patient are evaluated. A risk index (NPR=OxDxS) is calculated for each error considering the frequency with which it occurs (O), the ease with which it is detected (D) and the severity of the consequences on the patient (S). Values goes from 1 (rarely) to 10 (often). Improvement actions are proposed for all errors with a risk index greater than 100 and this index is re-evaluated after the implementation of that improvement measures. This same analysis is performed for three times over the last 10 years (in 2014, 2019 and 2023). Results In FMEA analysis carried out in 2014, a total of 40 errors were identified in the pre-analytical phase (63.5% of the total), 12 in the analytical phase and 11 in the post-analytical phase. 50% of the total errors detected required improvement strategies because they had an NPR≥100 value. In the 2023 evaluation, a total of 27 errors were identified in the pre-analytical phase, 10 in the analytical phase, and 5 in the post-analytical phase but only 6 required improvement strategies, which represents an 80% reduction in the most critical errors for patient safety. However, the pre-analytical phase continues to be the most critical, as it continues to account for 64% of the total errors identified, including the lack of identification of the patient or sample and the inadequate prior preparation of the patient. In analytical and post-analytical phases, the NPR values obtained in 2023 decrease significantly over analysis carried out in 2014 and its not needed to apply improvement strategies. Automation and the IT tools significantly reduce the number of errors with high NPR values, but it is undoubtedly the involvement of service professionals that makes the difference over the years. Conclusions The use of the FMEA methodological tool allows the design of appropriate indicators to detect and reduce errors in the clinical laboratory. Its application as part of a good quality control allows a continuous improvement in the laboratory process and consequently a guarantee of quality for the patient

Research on concrete-filled square CFRP steel tube under compressive-torsional hysteresis loading

Concrete-filled CFRP steel tube structures are presently employed on extensively in various engineering practices. This study examines the mechanical performance of concrete-filled square CFRP steel tubes under compressive-torsional hysteresis loads. Through the analysis of nine specimens, the research focuses on failure modes, torque–angle (T–θ) behavior, triaxial strain, and the synergistic effects of the steel tube and CFRP during cyclic loading. A numerical simulation method is initially introduced based on the test data to predict the actions of concrete-filled CFRP steel tube compressive-torsional specimens under hysteresis loading. This method is subsequently verified by contrasting it with the outcomes from representative experiments. In addition, the stress distribution of each material during the entire loading process of the sample is examined. Finally, using data from experiments and finite elements, combined with the trilinear model, a hysteretic model under compressive-torsional hysteresis stresses has been established for square CFRP steel tube filled with concrete. The calculated outcomes of the developed finite element model closely align with those of the hysteretic model.

Turbopump Parametric Modelling and Reliability Assessment for Reusable Rocket Engine Applications

The development of modern reusable launchers, such as the Themis project with its LOX/LCH4 Prometheus engine, CALLISTO—a reusable VTVL-launcher first-stage demonstrator with a LOX/LH2 RSR2 engine, and SpaceX’s Falcon 9 with its Merlin 1D engine, underscores the need for advanced control algorithms to ensure reliable engine operation. The multi-restart capability of these engines imposes additional requirements for throttling, necessitating an extended controller-validity domain to safely achieve low thrust levels across various operating regimes. This capability also increases the risk of component failure, especially as engine parameters evolve with mission profiles. To address this, our study evaluates the dynamic reliability of reusable rocket engines (RREs) and their subcomponents under different failure modes using multi-physics system-level modelling and simulation, with a particular focus on turbopump components. Transient condition modelling and performance analysis, conducted using EcosimPro-ESPSS software (version 6.4.34), revealed that turbopump components maintain high reliability under nominal conditions, with turbine blades demonstrating significant fatigue life even under varying thermal and mechanical loads. Additionally, the proposed predictive model estimates the remaining useful life of critical components, offering valuable insights for improving the longevity and reliability of turbopumps in reusable rocket engines. This study employs deterministic, thermally dependent structural simulations, with key control objectives including end-state tracking of combustion chamber pressure and mixture ratios and the verification of operational constraints, exemplified by the LUMEN demonstrator engine and the LE-5B-2 engine class.

Stretchable printed circuit boards using a silicone substrate of variable stiffness and conventional PCB fabrication methods

Abstract Stretchable Printed Circuit Boards (S-PCBs) offer unique advantages over rigid PCBs such as enabling conformability to changing environments and ergonomic designs with increased lifetime under dynamic loads. This study introduces an innovative fabrication method and a cost-effective solution for rapid prototyping stretchable printed circuit boards using commercially available materials. By utilizing silicone substrates with different levels of stiffness and structured copper sheets for electrical connections, the S-PCBs feature "stiff islands" embedded in a flexible base material. This fabrication method helps alleviate mechanical strain on strain-sensitive components while allowing local deformation of the S-PCB. The proposed fabrication method does therefore enable to integrate surface mount device components into S-PCBs and facilitates complex circuit designs while maintaining stretchability and fatigue resistance. Through material characterization, video strain analysis, as well as quasi-static and cyclic loading tests, this article demonstrates the efficacy of our approach. Based on the experimental results, we provide insights into failure modes and suggest design principles to further enhance the durability of S-PCBs fabricated with our method. We then conclude by presenting a soft wearable S-PCB demonstrator. The S-PCBs fabricated with this method withstood mechanical strains up to 100% and cyclic loads with 30% strain up to 625 cycles. The results are very promising for applications in soft robotics, wearable devices, and soft sensors and actuators. Overall, our study offers a comprehensive toolkit for fast S-PCB prototyping, paving the way for advancements in stretchable electronics with a high degree of complexity and stretchability.

Bearing performance and failure mechanisms of HSCM piles in marine soft soil under lateral loads

Helix Stiffened Cement Mixing (HSCM) pile represents a novel type of grouted composite pile known for its excellent compressive and tensile performance. Based on field tests, this study introduces a laboratory-based test design methodology and piling technique for HSCM piles. A comprehensive analysis of their horizontal bearing capacity, pile-soil failure mode, optimal piling technique, and reinforcement mechanism is performed. SEM-EDS technology is utilized at a microscopic scale to examine the reinforcement mechanism of HSCM piles. Through ABAQUS finite element simulation, a full-scale model simulating the HSCM pile-soil interaction is developed. The analysis includes pile cross-sectional stress, displacement, and pile-soil failure mode. The p−y curve of the HSCM pile is then plotted, elucidating the reinforcement mechanism of cemented soil. The results indicate that compared to ordinary helical piles, the horizontal bearing capacity of HSCM piles increases by 125.9%. SEM-EDS images reveal that cement and kaolin form a mutual bond, enhancing the strength of the soil around the pile. Additionally, unlike other composite piles, the direction of soil flow around HSCM piles encircles the first helix of the pile core. This research offers a design basis for understanding the horizontal bearing performance and failure mechanisms of HSCM piles in marine soft soil.

Coupling Fault Diagnosis of Bearings Based on Hypergraph Neural Network

Coupling faults that simultaneously occur during the operation of mechanical equipment are widespread. These faults encompass a diverse range of high-order coupling relationships, involving multiple base fault types. Based on the advantages of hypergraphs for higher-order relationship descriptions, two coupling fault diagnosis architectures based on the hypergraph neural network are proposed in this paper: 1. In the coupling fault diagnosis framework based on feature generation, the base faults serve as the hypergraph nodes, and each hyperedge connects the base faults. The generator, which consists of the hypergraph neural network, generates coupling faults as negative samples to enforce regularization constraints for the discriminator training. 2. In the coupling fault diagnosis framework based on feature extraction, each node represents a fault mode, and each hyperedge connects nodes with common failure modes. The multi-head attention mechanism extracts the features of base faults, and the common fault features in a hyperedge are aggregated via the hypergraph neural network. The inner product correlation is used to diagnose the fault modes. The results show that the diagnostic accuracy for coupling faults with the two frameworks reaches 88.6% and 86.76%, respectively. Both frameworks can be used for the diagnosis and analysis of high-order coupling faults.