Complex Failure Modes Research Articles

Research On Bolt Bending Analysis Method Based On FEM

In practical engineering, for instance the aerospace aircraft and engine industry, fastener connection is a very common joint structures, the current engineering typical strength analysis for bolt connection structure doesn’t include bolt bending analysis. In the engineering research and design phase, if the bolt connection plate is too thick or it contains shims or spacers that cause the bolt itself to be too long, then bolt bending failure analysis is also necessary. The bolt bending analysis is a complex failure mode, there is no unified analysis method for effective assessment for it. This paper introduces an improved calculation method which is proposed to specifically calculate the bending load of the bolt according to the force characteristics of the bolt for bolt bending analysis.

A microstructure-sensitive and derivative-free continuum model for composite materials: Applications to concrete

Owing to high strength to weight ratio, durability, corrosion resistance and design flexibility, composite materials are extensively used in engineering applications. While these materials have several useful properties, they often exhibit complex failure modes that arise from their heterogeneous microstructure details. A continuum model for composites should accordingly exploit the microstructural information for an accurate prediction of the non-local and non-linear behavior preceding failure. In simulating discontinuities such as cracks, a derivative-free continuum theory – peridynamics, to wit – has an advantage in that the integro-differential balance laws work even with discontinuities. It does not need the special measures required with continuum models using partial differential equations (PDEs) for simulating cracks. In this work, we use a novel non-local variant of the deformation gradient and propose a constitutive model for composite materials within a derivative-free set-up whilst incorporating the microstructural information, which enables a faithful reproduction of the macroscopic response leading to failure. Since concrete is perhaps the most commonly used composite material in large scale engineering applications, we implement the model to study fracture/damage and size effect in concrete. Our results are in close conformity with the experimental data available in the literature. We also show that complex phenomena like crack propagation and branching are accurately simulated via the proposed model with considerably reduced computational overhead.

Modeling Progressive Damage and Failure of Single-Lap Thin-Ply-Laminated Composite-Bolted Joint Using LaRC Failure Criterion.

Thin-ply composite failure modes also significantly differ from conventional ply composite failure modes, with the final failure mechanism switching from irregular progressive failure to direct fracture characterized by a uniform fracture with the reduction of the ply thickness. When open holes and bolt joints are involved, thin-ply-laminated composites exhibit more complex stress states, damage evolution, and failure modes. Compared to the experimental study of thin-ply-laminated composite-bolted joints, there are few reports about numerical analysis. In order to understand the damage evolution and failure mechanism of thin-ply-laminated composites jointed by single-lap bolt, a progressive damage model based on three-dimensional (3D) LaRC failure criterion combined with cohesive element is constructed. Through an energy-based damage evolution method, this model can capture some significant mechanical characteristics in thin-ply-laminated structures, such as the in situ effect, delamination inhibition, and fiber compressive kinking failure. The comparisons between the numerical predictions and experimental observations are made to verify the accuracy of the proposed model. It is found that the predicted stress-displacement curves, failure modes, damage morphologies, etc., are consistent with the experimental results, indicating that the presented progressive damage analysis method displays excellent accuracy. The predicted stress at the onset of delamination is 50% higher than that of the conventional thick materials, which is also consistent with experimental results. Moreover, the numerical model provides evidence that the microstructure of thin-ply-laminated composite performs better in uniformity, which is more conducive to inhibiting the intra-layer damage and the expansion of delamination damage between layers. This study on the damage inhibition mechanism of thin-ply provides a potential analytical tool for evaluating damage tolerance and bearing capabilities in thin-ply-laminated composite-bolted joints.

Macro/mesofracture and instability behaviors of jointed rocks containing a cavity under uniaxial compression using AE and DIC techniques

To better understand the macro/mesofracture and instability behaviors of the jointed surrounding rock of a cavity, a series of uniaxial compression tests on jointed granite samples with openings were conducted using acoustic emission (AE) and digital image correlation (DIC) techniques. The experiments indicated that with increasing joint angle, the peak strength and elastic modulus first decreases and subsequently increases, reaching a minimum as the joint angle approaches 45°. The linear transition point of the AE cumulative count before the peak stress is related to crack initiation or interpenetration, while each instantaneous increase in the AE count after the peak stress is associated with the stress drop caused by the localization of high strain or the formation or further development of visible cracks. The accelerated release of AE energy probably has a better early warning ability for the prediction and forecast of rock instability and failure. Moreover, the prominent strain localization bands indicate crack initiation, and the crack extension behavior signifies the gradual evolution of the strain localization bands. The existence of joints changes the stress distribution patterns in the samples, and the high stress concentration areas formed at the joint tips and around the cavities interact and jointly dominate the fracturing process, eventually leading to complex failure modes. The unstable failure of the specimens is mainly induced by tensile cracks around the caverns and the joint tips, accompanied by local shear failure.

Hierarchical level fault detection and diagnosis of ship engine systems

As smart ships are developed, research into engine management through remote monitoring and support from data collected at onshore control centers is being conducted. However, ship engine data collected from various sensors have high dimensions, large measurement errors, few labeled data, and insufficient amounts and quality of data, making it difficult to determine the condition of engines with complex failure modes under various operating environments.This study proposes a hierarchical level fault detection and diagnosis (HL-FDD) method that combines domain knowledge of ship engines and advanced data analysis techniques. The developed method extracts key features of reduced dimensions from the original variables (sensors) through an optimal hierarchical clustering and dimension-reduction model, allowing a hierarchy divided into the top (the entire system combining all features), middle (subsystems and feature sets), and bottom (components and sensors) levels. The reduced key features are used to generate robust regression models and dynamic thresholds (prediction intervals) according to the engine load, and dynamic thresholds determine whether the engine’s condition is normal or abnormal. The dynamic thresholds are able to automatically label abnormal conditions of the engine. Once anomalies are detected at the top level, the proposed method can sequentially search for data on features belonging to the middle and bottom hierarchies for detailed fault diagnosis to determine which engine subsystem or component(s) caused the engine fault(s). Actual data collected by ship operators verify the proposed method’s efficiency, reliability, and accuracy.

Finite element analysis of damage mechanisms of composite wind turbine blade by considering fluid/solid interaction. Part I: Full-scale structure

This paper is the first one of two that studies damage mechanisms of composite wind turbine blade using numerical analysis. An important highlight is to consider fully complex failure modes of the skin and web as well as the adhesive failure between them. The main work is divided into three parts: (1) the damage model for composite laminates developed by our research group is applied to the skin and web, and the bilinear cohesive model is used to predict the adhesive failure; (2) finite element model is established by considering fluid/solid interaction to study damage mechanisms of a 59.5 m long blade; (3) the wind velocity model including the normal and extreme wind shear effects as well as the tower shadow effect is considered for fluid elements as boundary conditions. By performing fluid/solid coupling analysis using ABAQUS software, distributions of the stress, displacement and damage for the skin, web and adhesive layer as well as the fluid field information under different wind velocity are explored. In addition, numerical results are also compared with those using the blade element momentum (BEM) theory and other existing work. Some conclusions for predicting damage mechanisms and suggestions for improving damage tolerance and load-bearing ability of the blade are given.

Thermal damage and acoustic emission characteristics of granite based on Brazilian splitting test

Rapid and non-destructive evaluation of the thermal stability of surrounding rocks and rock strength is important in field construction. Physical properties and acoustic emission (AE) characteristics of granites can be changed at high temperatures. Thus, investigations on the relationships between rock strength and physical properties could help to evaluate the thermal stability via a physical measuring method. In this study, granite selected from a geothermal drilling project was heated at 25–600°C and different physical properties and tensile strength were measured. Moreover, the thermal damage degree was evaluated and the tensile failure pattern was quantitatively characterized. Notably, the tensile strength of granites could be indirectly acquired through P-wave velocity, hardness, colour, mass loss and volume expansion. The experimental results showed that the variations could be roughly divided into two stages. (1) At 25–400°C, the mass, volume, P-wave velocity, hardness, tensile strength, colour and AE parameters changed slightly. The failure mode and damaged section did not differ significantly within this temperature range. (2) At 400–600°C, these physical and mechanical properties and AE parameters changed significantly. The formation of micro-cracks was aggravated by the loss of structural water and phase transition of quartz, leading to complex failure modes and jagged morphology of damaged sections. The thermal damage factor and the compressive damage factor defined by the P-wave velocity at different conditions increased greatly. In summary, there was no major damage and the internal structure was still relatively intact at 25–400°C. However, the granite showed serious damage in the high-temperature environment (400–600°C), resulting in great changes in physical and mechanical properties and AE characteristics.

A hybrid multi-phase field model to describe cohesive failure in orthotropic materials, assessed by modeling failure mechanisms in wood

Fracture mechanics is crucial for many fields of engineering applications, as precisely predicting failure of structures and parts is required for efficient designs. The simulation of failure processes is, both from a mechanical and a numerical point of view, challenging, especially for inhomogeneous materials, where the microstructure influences crack initiation and propagation and might lead to very complex crack patterns. The phase field method for fracture is a promising approach to encounter such materials, since it is able to describe complex fracture phenomena like crack kinking, branching and coalescence. Moreover, it is a largely mesh independent approach, given that the mesh is homogeneous in the area of the crack. However, the original formulation of the phase field method is limited to isotropic materials and does not account for preferable fracture planes defined through the material’s microstructure. In this work, the method is expanded to take orthotropic constitutive behavior and preferable directions of crack propagation into account. We show that by using a stress-based split and multiple phase field variables with preferable fracture planes, in combination with a hybrid phase field approach, a general framework can be found for simulating anisotropic, inhomogeneous materials. The stress-based split is based on fictitious crack faces and is, herein, expanded to support anisotropic materials. Furthermore, a novel hybrid approach is used, where the degradation of the sound material is performed based on a smooth traction free crack boundary condition, which proves to be the main driving factor for recovering commonly observed crack patterns. This is shown by means of a detailed analysis of two examples: a wooden single edge notched plate and a wood board with a single knot and complex fiber directions. In both cases, the proposed novel hybrid phase field approach is able to realistically reproduce complex failure modes.

Precise damage shaping in self-sensing composites using electrical impedance tomography and genetic algorithms

Fiber-reinforced composites with nanofiller-modified polymer matrices have immense potential to improve the safety of high-risk engineering structures. These materials are intrinsically self-sensing because their electrical conductivity is affected by deformations and damage. This property, known as piezoresistivity, has been extensively leveraged for conductivity-based damage detection via electrical resistance change methods and tomographic imaging techniques such as electrical impedance tomography (EIT). Although these techniques are very effective at detecting the presence of damage, they suffer from an inability to provide precise information about damage shape, size, or mechanism. This is particularly detrimental for laminated composites which can suffer from complex failure modes, such as delaminations, that are difficult to detect. To that end, we herein propose a new technique for precisely determining damage shape and size in self-sensing composites. Our technique makes use of a genetic algorithm (GA) integrated with realistic physics-based damage models to recover precise damage shape from conductivity changes imaged via EIT. We experimentally validate this technique on carbon nanofiber (CNF)-modified glass fiber-reinforced polymer (GFRP) laminates by considering two specific damage mechanisms: through-holes (as a function of number, size, and location) and impact-induced delaminations (as a function of impact energy). Our results show that this novel technique can accurately reconstruct multiple through-holes with radii as small as 1.19 mm and delaminations caused by low velocity impacts. The reconstructed delamination shapes and sizes were shown to be in much better agreement with the actual delaminations observed using optical microscopy than is achievable by traditional EIT alone. These findings illustrate that coupling piezoresistivity with conductivity-based spatial imaging techniques and physics-based inversion strategies can enable damage shaping capabilities in self-sensing composite structures.

Dynamic tensile failure of layered sorptive rocks: Shale and coal

Shale and coal samples were selected to clarify the tensile failure of layered sorptive rocks under the co-influence of bedding planes and loading conditions. Results indicate that, for these two rocks, tensile strength exhibited anisotropy. As loading rates rises, inverted Trumpet-shaped trend lines is proposed to reflect bedding effects with loading effects. The tensile strength of shale increases linearly with the bedding angle rises, while the value of coal shows great dispersion. The former is greater than the latter. An ultimate loading rate exists for shale and coal. Empirical relations were established to describe tensile strength with varying loading rate and bedding angle. Under the co-influences of bedding plane and loading conditions, more complex failure modes happened for coal and shale samples. Hydraulic fracturing along the direction of 60° (for shale reservoirs) or 45° (for coal reservoirs) to the bedding plane is more likely to produce a complex fracture network.

Buckling and ultimate load prediction models for perforated steel beams using machine learning algorithms

Large web openings introduce complex structural behaviors and additional failure modes of steel cellular beams, which must be considered in the design using laborious calculations (e.g., exercising SCI P355). This paper presents seven machine learning (ML) models, including decision tree (DT), random forest (RF), k-nearest neighbor (KNN), gradient boosting regressor (GBR), extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), and gradient boosting with categorical features support (CatBoost), for predicting the elastic buckling and ultimate loads of steel cellular beams. Large datasets of finite element (FE) simulation results, validated against experimental data, were used to develop the models. The ML models were fine-tuned via an extensive hyperparameter search to obtain their best performance. The elastic buckling and ultimate loads predicted by the optimized ML models demonstrated excellent agreement with the numerical data. The accuracy of the ultimate load predictions by the ML models exceeded the accuracy provided by the existing design provisions for steel cellular beams published in SCI P355 and AISC Design Guide 31. The relative feature importance and feature dependence of the models were evaluated and discussed in the paper. An interactive Python-based notebook and a user-friendly web application for predicting the elastic buckling and ultimate loads of steel cellular beams using the developed optimized ML models were created and made publicly available. The web application deployed to the cloud allows for making predictions in any web browser on any device, including mobile. The source code of the application available on GitHub allows running the application locally and independently from the cloud service.

Investigation of three different cooling treatments on dynamic mechanical properties and fragmentation characteristics of granite subjected to thermal cycling

Underground thermal engineering such as geothermal exploitation usually involves the response of impact loads due to the drilling operation and blasting load. This study investigated the dynamic mechanical properties and fragmentation characteristics of granite subjected to thermal cycling under natural cooling, water cooling, and liquid nitrogen cooling. Firstly, split Hopkinson pressure bar tests were performed on granite under three different cooling methods to study the dynamic mechanical parameters with thermal cycles. Subsequently, the damage factors were calculated to evaluate the damage on granite and the relationship between damage and stress. Finally, the granite fragments after impact were sieved, and the fractal dimension and average fragment size were introduced to analyze the fragmentation characteristics. The results demonstrate that the degradation of dynamic mechanical properties mainly occurs in the first four thermal cycles; with the increase of damage factor, the stress decreases linearly under the three cooling methods, and the damage caused by liquid nitrogen cooling to granite is the most significant, followed by water cooling. Fragments of granite change from axial splitting failure to a more complex composite failure mode. In addition, with the decrease of the average fragment size, the fractal dimension of the granite gradually increases, resulting in the higher fragmentation degree and the better uniformity of fragmentation.

Highly Sodiophilic, Defect‐Rich, Lignin‐Derived Skeletal Carbon Nanofiber Host for Sodium Metal Batteries

AbstractThree‐dimensional host structures with superior sodiophilicity and low nucleation barriers can help combat the complex failure modes of Na metal anodes originating from accelerated dendrite formation, anodic corrosion, and electrolyte depletion. This work reports the fabrication of a unique super‐sodiophilic, defect‐rich and hierarchically porous skeletal carbon nanofiber (SCNF) host for SCNF@Na anodes using electrospinning of the low‐cost, renewable lignin biopolymer. The uniform nucleation and plating of Na effectuated by the hierarchically porous structure coupled with the defect‐induced formation of a resilient, F‐rich solid electrolyte interface (SEI) layer offers excellent protection to the metallic anode. The defect‐rich porous structure plays an important role in mediating dense Na nucleation, planar growth, and electrochemical stability according to the depth profiling experiments and density functional theory calculations. The SCNF@Na composite anode maintains high Coulombic efficiencies (CEs) and electrochemical reversibility in asymmetric and symmetric cells. The full cells prepared by interfacing the SCNF@Na anode with a Na3V2(PO4)2F3 cathode delivers exceptional capacity retention of 106 mAh g–1 for 350 cycles with an average CE of 99.2% at 1C, and 103 mAh g–1 after 200 cycles at 4C. Such rationally designed carbon hosts derived from biopolymers open a new avenue for safe and low‐cost metal batteries.

Experimental study on bond strength failure of GFRP lap joints

Bonding has many advantages such as high adhesive strength, peel resistance, good toughness, shrinkage and corrosion-resistance, compared with bolting, welding and other connection patterns, which is the basic connection of the composite materials. The mechanical properties of the bond depend on the type and geometric parameters of the bond. In this paper, experiments were conducted to study the bond strength failure of GFRP single and double lap joints and explore related influencing factors. The variable factors of test specimens include bond type, length and thickness. The quasi-static tensile test was carried out at room temperature to analyze multiple influencing factors of lap joints and characteristic of GFRP bond failure strength. The results showed that the failure of the bond usually occurred at the interface of bond, and bond types, length and thickness have significant effects on the failure of bond. Failure strength increased firstly and then decreased with the increase of adhesive thickness. With the increase of bonding length, the failure mode of the double-lap joint become complex failure mode, which is accompanied by GFRP delamination near the surface of fiber of the bonding.

Damage and failure analysis of a SiCf/SiC ceramic matrix composite using digital image correlation and acoustic emission

The analysis of failure behaviors of continuous fiber-reinforced ceramic matrix composites (CMCs) requires the characterization of the damage evolution process. In service environments, CMCs exhibit complex damage mechanisms and failure modes, which are affected by constituent materials, meso architecture, inherent defects, and loading conditions. In this paper, the in-plane tensile mechanical behavior of a plain woven SiC f /SiC CMC was investigated, and damage evolution and failure process were studied in detail by digital image correlation (DIC) and acoustic emission (AE) methods. The results show that: the initiation of macro-matrix cracks have obvious local characteristic, and the propagation paths are periodically distributed on the material surface; different damage modes (matrix cracking and fiber fracture) would affect the AE energy signal and can be observed in real-time; the significant increase of AE accumulated energy indicates that serious damage occurs inside the material, and the macroscopic mechanical behavior exhibits nonlinear characteristic, which corresponds to the proportional limit stress (PLS) of the material.

Finite element analysis of flexible pipes under compression: A study on the damaged high strength tape and its effects on tensile armor instability

To study the compressive behavior of flexible pipes and tensile armor instability, a nonlinear Finite Element model was developed. This fully tridimensional model recreates a five layer unbonded flexible pipe. When the pipe is submitted to compressive loads, tensile armor instability may occur in the form of lateral buckling or radial buckling, also known as “birdcaging”. This complex failure mode may be influenced by several parameters, including defects on the layers that are supposed to restrain and protect the tensile armor, more specifically, the high strength tape. This layer is usually made of high a performance textile material that is not directly exposed to the environment. However, damage may be sustained when the outermost layer is subjected to deep cuts and breached. This article focuses on the effects of a damaged high strength tape and its implications on the tensile armor instability failure modes. The study shows that damage in the high strength tape may compromise the cohesion between the layers, significantly reducing the critical instability load and the flexible pipe's stiffness, as well as altering the yielding for the layers.

Realtime Health Monitoring of Composite Structures Using FBG Sensors

Although composite structures popular in aerospace structures, due to complex failure modes, frequent inspection is required, increasing maintenance costs. Consequently, a structural health monitoring (SHM) system to monitor and pre-emptively detect damage is crucial. The Institute for Integrated and Intelligent Systems (IIIS) and the Advanced Design and Prototyping Technologies Institute (ADaPT) at Griffith University are planning to design and build a real-time SHM system utilizing Fiber Bragg Grating (FBG) sensors to obtain health data and use machine learning for real-time damage classification. We propose an experimental setup to generate ground truth neural network training. This includes methods to obtain data on material condition, an interrogation system for real-time monitoring, and machine learning for damage classification and remaining useful life (RUL) prediction.

Study of spontaneous adiabatic shear bands in expanding rings under explosion by thermo-elastic-plastic phase field model

Metallic cylindrical shells under explosion load may exhibit complex failure modes, which is a challenge issue in explosive and impact engineering. The thermo-elastic-plastic phase-field model is used in this paper to study the expanding cracks of cylindrical shell under internal explosion load, called expanding rings. The spontaneous adiabatic shear bands (ASBs) evolution is successfully captured by considering the mesh-independent gradient damage characteristics. Thus, the typical experimental phenomena are effectively revealed from the numerical simulations. The evolutionary behavior of the ASBs is analyzed, which is induced by defects and the self-similar structure of ASBs. It is found that the initial defects in the expanding rings dominate the self-organized ASBs. The number of ASBs tends to be saturated with the increase of defects. In addition, a formula for calculating the structural radius of laminar ASBs is proposed, which can well predict the simulation result.

A Redundancy-Guided Approach for the Hazard Analysis of Digital Instrumentation and Control Systems in Advanced Nuclear Power Plants

Digital instrumentation and control (I&C) upgrades are a vital research area for the nuclear industry. Despite their performance benefits, deployment of digital I&C in nuclear power plants (NPPs) has been limited. Digital I&C systems exhibit complex failure modes including common cause failures (CCFs), which can be difficult to identify. This paper describes the development of a redundancy-guided application of the Systems-Theoretic Process Analysis and fault tree analysis for the hazard analysis of digital I&C in advanced NPPs. The resulting Redundancy-Guided Systems-Theoretic Hazard Analysis (RESHA) is applied for the case study of a representative state-of-the-art digital reactor trip system. The analysis qualitatively and systematically identifies the most critical CCFs and other hazards of digital I&C systems. Ultimately, the RESHA can help researchers make informed decisions for how, and to what degree, defensive measures such as redundancy, diversity, and defense in depth can be used to mitigate or eliminate the potential hazards of digital I&C systems.

Designing production-friendly machine learning

Building production ML applications is difficult because of their resource cost and complex failure modes. I will discuss these challenges from two perspectives: the Stanford DAWN Lab and experience with large-scale commercial ML users at Databricks. I will then present two emerging ideas to help address these challenges. The first is "ML platforms", an emerging class of software systems that standardize the interfaces used in ML applications to make them easier to build and maintain. I will give a few examples, including the open-source MLflow system from Databricks [3]. The second idea is models that are more "production-friendly" by design. As a concrete example, I will discuss retrieval-based NLP models such as Stanford's ColBERT [1, 2] that query documents from an updateable corpus to perform tasks such as question-answering, which gives multiple practical advantages, including low computational cost, high interpretability, and very fast updates to the model's "knowledge". These models are an exciting alternative to large language models such as GPT-3.