Scale Cracking Research Articles

Detecting large-scale underwater cracks based on remote operated vehicle and graph convolutional neural network

It is of great significance to quickly detect underwater cracks as they can seriously threaten the safety of underwater structures. Research to date has mainly focused on the detection of above-water-level cracks and hasn’t considered the large scale cracks. In this paper, a large-scale underwater crack examination method is proposed based on image stitching and segmentation. In addition, a purpose of this paper is to design a new convolution method to segment underwater images. An improved As-Projective-As-Possible (APAP) algorithm was designed to extract and stitch keyframes from videos. The graph convolutional neural network (GCN) was used to segment the stitched image. The GCN’s m-IOU is 24.02% higher than Fully convolutional networks (FCN), proving that GCN has great potential of application in image segmentation and underwater image processing. The result shows that the improved APAP algorithm and GCN can adapt to complex underwater environments and perform well in different study areas.

Evaluation of Internal Cracks in Turbine Blade Thermal Barrier Coating Using Enhanced Multi-Scale Faster R-CNN Model

Thermal Barrier Coatings (TBCs) have good performance in heat insulation during service on turbine blades. However, the accumulated residual stress will form cracks, which can easily lead to coating failure. To ensure safe operation, it is necessary to find a method that can evaluate the health of the coating. In this paper, a non-destructive evaluation technique based on Multi-Scale Enhanced-Faster R-CNN (MSE-Faster R-CNN) is proposed. Firstly, the Visual Geometry Group Network19 layer (VGG-19) was adopted as the baseline network to find the candidate crack Region of Interest (ROI). Considering the influence of the crack on the surroundings, the ROI was expanded to obtain the context information. Secondly, a multi-scale Faster R-CNN detector was used to refine the candidate regions, and provided a comprehensive feature for better crack detection. Finally, a fusion lifetime prediction model was proposed to estimate the remaining lifetime of the TBC. Extensive experiments were conducted to evaluate the performance of the proposed method. The results demonstrated that the proposed method can accurately locate (0.898) and detect (0.806) the cracks in different scales, and the lifetime estimation result reached the best level (Root Mean Square Error (RMSE) = 2.7); there wasas also an acceptable time cost (1.63 s), and all detection conditions of the error rates were below 15%, achieving the best results among the state-of-art methods.

Crack interaction and fracturing of geomaterials with multiscale cracks

Geomaterials contain numerous cracks whose interaction affects the macroscopic mechanical properties of the materials significantly. We propose a mechanical model of cracks of 3 scale levels to investigate the pattern of crack propagation and possible failure modes produced. For modelling crack interaction between different scale levels, we model the unloading domain of influence of a single crack as the sphere with the diameter being the crack size drawn around the crack center. When the tip of a lower scale crack propagates into the influence domain of a crack of a higher scale level, the propagating crack of a lower scale will be unloaded and never grow from this tip. Propagation and coalescence of cracks of different scale levels for different initial crack concentration result in different patterns of cracks propagation and failure. For relatively lower initial crack concentrations, the interaction of cracks of the first scale level dominates, producing failure due to the coalescence of cracks of the first scale level. With the increase of the initial crack concentration, the coalescence of cracks of the third scale level dominates leading to failure at a higher initial concentration. With the further increase of the initial crack concentration, the failure is produced by the coalescence of cracks of the second scale level. At even higher initial crack concentrations the failure is again produced by the coalescence of cracks of the first scale level. The dynamic strength decreases with the increase in the initial crack concentration. • A model of interaction, growth and coalescence of self-similar cracks of 3 scale levels is proposed. • The unloading effect of higher scale level cracks on the propagation of lower scale level cracks is considered. • Degradation of the effective deformation modulus is considered in solving crack-tip equations of motion. • The discreteness of crack propagation at the accelerating stage of loading is discovered. • At certain intervals of the initial crack concentration cracks of the lowest scale level will coalesce first.

Study of wave propagation in discontinuous and heterogeneous media with the dynamic lattice method

The development of a new dynamic lattice element method (dynamicLEM) as well as its application in the simulation of the propagation of body waves in discontinuous and heterogeneous media is the focus of this research paper. The conventional static lattice models are efficient numerical methods to simulate crack initiation and propagation in cemented geomaterials. The advantages of the LEM and the developed dynamic solution, such as simulation of arbitrary crack initiation and propagation, illustration and simulation of existing inherent material heterogeneity as well as stress redistribution upon crack opening, opens a new engineering field and tool for material analysis. To realize the time dependency of the dynamic LEM, the equation of motion of forced vibration is solved while using the Newmark-beta method and implementing the non-linear Newton–Raphson Jacobian method. The method validation is done according to the results of a boundary element method (BEM) in the plane P-SV-wave propagation within a plane strain domain. Further tests comparing the generated wave types, simulation and study of crack discontinuities as well as inherent heterogeneities in the geomaterials are conducted to illustrate the accurate applicability of the new dynamic lattice method. The results indicate that with increasing heterogeneity within the material, the wave field becomes significantly scattered and further analysis of wave fields according to the wavelength/heterogeneity ratio become indispensable. Therefore, in a heterogeneous medium, the application of continuum methods in relation to structural health monitoring should be precisely investigated and improved. The developed dynamic lattice element method is an ideal simulation tool to consider particle scale irregularities, crack distributions and inherent material heterogeneities and can be easily implemented in various engineering applications.

Hot corrosion behaviour of constant and pulsed current welded Hastelloy X in Na2SO4, V2O5, and NaCl salt mixture at 900 °C

The high-temperature corrosion behavior of constant current gas tungsten arc (GTA) and pulsed current gas tungsten arc (PCGTA) welded Hastelloy X with different filler wires (C263 and ERNiCr-3) are studied for 50 cycles at 900 °C. Molten salt I (MS I) (75% Na2SO4 + 25% V2O5) and molten salt II (MS II) (75% Na2SO4 + 20% V2O5 + 5% NaCl) were coated on the welded specimens. MS II coated substrate shows the highest weight gain than MS I with a parabolic constant for GTA ERNiCr-3 as 21.440 × 10–6 mg2/(cm4.s). Whereas PCGTA C263 welded sample with MS I, revealed parabolic constant (lowest) of 0.008 × 10–6 mg2/ (cm4.s). Based on the results, an increasing pattern of hot corrosion resistance of substrates is arranged as GTA ERNiCr-3 < GTA C263 < PCGTA C263 < PCGTA ERNiCr-3. PCGTA shows more refined grains, higher grain boundary volume, better corrosion resistance, and more protective phases like Cr2O3, NiO, NiCr2O4, CoCr2O4, Al2O3, NiFe2O4, NbO than GTA weldment. But phases such as Fe2O3, MoO3, and Cr2S3 (non-protective phases) decrease corrosion resistance due to acid fluxing of alloying elements that promote the oxide scale exfoliation, spallation, chipping, and cracking. This study observed that PCGTA with C263 filler in MS I and MS II environment provides good corrosion resistance at high temperatures.

Revealing the effect of Al content on the oxidation of γ'-strengthened cobalt-based superalloys

The effect of Al content (6–9 at%) on the oxidation of multicomponent γ'-strengthened cobalt-based superalloys at 800 ℃–1000 ℃ was revealed by combining experimental and CALPHAD approaches. The results showed that increasing Al content can improve the oxidation resistance of the alloys by promoting the formation of a dense Al2O3 layer and reducing the activity of Ti at 800 ℃ and 900 ℃. However, at 1000 ℃, increasing Al content can promote the formation of Co(W,Mo)O4 and TaTiO4, as well as oxide volatilization, which aggravated the cracking and spallation of the oxide scales and deteriorated the oxidation resistance of the alloys.

Structure and Deformation Behavior of Human Dentin

The relationship between structure and stress accommodation mechanisms (deformation and fracture) of human dentin on macro-, micro- and nano- scales is discussed. Dentin is the hard basis of human teeth with complicated hierarchically organized structure, which is attested as a natural composite consisted of a bioorganic matrix armed by collagen fibers and apatite crystallites. Dentin exhibits the unique strength properties. On the macroscopic level, under tensile load, it behaves like a brittle solid, and like a viscoelastic one in the case of compression. At the same time, on the microscopic scale cracks in dentin grow in a viscoelastic manner under tensile loading. Structure, mechanical properties and crack growth of human dentin on macro-, micro- and nano- scales, including TEM study, are considered in detail. It was shown that a brittle response under tension is the macroscopic feature of dentin caused by dentin channels, while viscoelasticity is its intrinsic property.

Damage evaluation of rock salt under multilevel cyclic loading with constant stress intervals using AE monitoring and CT scanning

When salt caverns are used for the medium of underground energy storage, the surrounding rock experiences cyclic loading due to periodical injection-production. The cyclic load contains constant stress intervals and varied stress amplitudes, which threatens the stability of the salt caverns. In this study, multilevel cyclic loading with constant stress intervals on rock salt were conducted under uniaxial compression conditions. Acoustic emission (AE) monitoring and post-test computed tomography (CT) scanning were used to evaluate the damage evolution during the loading process. Results show that the density of the hysteresis loops in overall stress-strain curves present two-stage alternating sparse and dense loops during the loading. The cumulative AE count/energy increases with increasing the number of cycles and loading stage, and the growth rate for a single stage gradually becomes constant. The AE signals were divided into six kinds according to frequency-amplitude distribution. The presence of stress intervals increases the number of cracks and the median scale cracks account for about 90% of the total number of cracks. By analyzing the CT images, the crack morphology of the salt sample without hold time shows several inclined main cracks with numerous branch microcracks while that with the hold time of 20 s shows multiple branch cracks and numerous microcracks. To quantitatively evaluate damage evolution during the loading process, a damage variable is defined based on the relative variation of AE parameters. The damage accumulation for a single loading stage presents a two-phase pattern, and the damage degree during the accelerated deformation accounts for approximately half of the total damage. This study can provide guidance for designing the operation parameters when salt caverns are used for the storage of natural gas, air and CO 2 . • Multilevel cyclic loading tests with constant stress intervals are conducted on rock salt. • The AE parameters characteristics are analyzed. • The crack distribution and fracture behavior are visualized by CT images. • Damage evolution behavior is quantitatively evaluated.

On a new high order phase field model for brittle and cohesive fracture: numerical efficiency, length scale convergence and crack kinking

Being able to seamlessly deal with complex three dimensional crack patterns like branching and merging, phase field models (PFMs) are promising in the computational modelling of fracture of solids. Regarding the damage field, if only its second order derivative is present in the governing equation we have a second order PFM and if its fourth order derivative is also present we have a fourth order PFM. Previous studies have demonstrated that fourth order PFMs, with its smoother damage profile, are better in convergence and able to model strong anisotropic fracture energies. However, previous fourth order PFMs were developed only for brittle fracture and do not embed the material strength directly in the formulation. This paper presents a fourth order phase field regularised cohesive zone model (fourth order PF-CZM) for both brittle fracture and quasi-brittle fracture. We present a semi-analytical approach to compute the coefficients of the rational degradation function used in the fourth order PF-CZM. The proposed model is tested with multiple benchmark problems for mode I and mixed-mode fracture, and the results demonstrate that the fourth order PF-CZM: (1) provides results independent of the length scale and (2) is more efficient than its second order counterpart. The proposed model was then applied to modelling strong anisotropic fracture energies. A detailed study on sawtooth like crack patterns, which is a signature of strongly anisotropic fracture, is carried out with focus on periodicity and length scale sensitivity.

Formation of high-temperature inner oxide scale on low alloy steels: Segregation, partitioning and transformation reactions

High-temperature oxide scales on two low-alloy steels were carefully studied. For the first time, the FeO inner scale on the steel with low Cr and Si contents is observed to incur a eutectoid reaction in a divorced mode during cooling, resulting into a bimodal Ni distribution near the scale/substrate interface. In contrast, a combination reaction of oxygen and FeO layer occurs on the steel with high Cr and Si contents where cracks in scales provide pathway for oxygen transport, and a fine-grained nanocomposite oxide band develops near the interface. Dealloying, partition and segregation during the scale formation are comprehensively addressed.

Concrete crack segmentation based on UAV-enabled edge computing

In recent years, the rapid development of UAV technology has greatly improved the efficiency of the detection of concrete bridge cracks. With the increase in the number of bridge inspection UAVs, the number of tasks handled by cloud services has increased linearly, resulting in increased computational pressure on cloud services. In order to reduce the computational load of cloud servers, we proposed a crack segmentation network based on UAV-enabled edge computing. However, due to the limitation of computational capability of edge computing and the strength inhomogeneity and background complexity of cracks, crack detection is still a challenging task. Thus, we proposed an effective concrete crack segmentation network based on UAV-enabled edge computing, the network used feature map fusion to fuse different levels of feature map information into lower-level features for crack detection. The atrous spatial pyramid pooling network was used to increase the low-resolution feature map receptive field information for cracks and to enhance the detection accuracy for cracks of different scales. In addition, loss functions for crack datasets were proposed to solve the problem of imbalance due to positive and negative samples in the concrete crack images. Experiments demonstrated that the proposed methods are better than the state-of-the-art edge detection and semantic segmentation methods in terms of accuracy and generality.

Phase-field modeling of electromechanical fracture in piezoelectric solids: Analytical results and numerical simulations

The optimal design, reliable manufacture and durable utilization of electromechanical systems demand objective modeling of failure under coupled electric field and mechanical actions. This is still a challenging and largely an open issue despite the large volumes of contributions from the discontinuous approach and the phase-field community for brittle fracture. In this work we propose a length scale insensitive phase-field cohesive zone model (PF-CZM) for electromechanical fracture in linear piezoelectric solids, overcoming all the issues of mesh dependence, crack tracking complexity, length scale sensitivity and crack nucleation inconsistency. More specifically, in the electromechanically coupled phase-field constitutive relations, the evolution law for the crack phase-field is established from the geometric regularization of sharp cracks and the energetic criterion in global form. In order to capture the unsymmetric mechanical and electrical fracture behavior, the mechanical energy release rate contributed from the positive cone of the effective stress in energy norm is employed as the crack driving force, resulting in a hybrid formulation. With a set of phase-field characteristic functions optimal for cohesive fracture, the critical stress upon crack nucleation and the post-peak softening regime are derived in close-form for the first time, and the experimentally observed effects of the electric field on the fracture load are predicted qualitatively. The proposed electromechanical PF-CZM is numerically implemented in the multi-field finite element method and applied to several representative examples. In all cases, as those for purely mechanical problems, the predicted crack patterns and fracture loads are insensitive to the phase-field length scale and the experimental results are well reproduced. In particular, the effect of electric field on the deflection of crack paths can be captured. These merits make the proposed PF-CZM promising in the modeling of electromechanical fracture in piezoelectric solids.

Analytical STEM Investigations of Nozzle Vane Surfaces after Turbine Engine Operation

The scope of this work includes detailed microstructural investigations using high-resolution scanning transmission electron microscopy of the oxide scales and deposits formed on nozzle ring vanes made of IN713LC alloy during the operation of an auxiliary power unit under real conditions. The paper presents the differences of oxidation processes between the suction and the pressure sides of the vanes. It has been shown that in the pressure side of the vane, where a greater flow of exhaust gases from the combustion chamber is present, the oxidation process is accompanied by a deposition of among others (i.a.) Fe, Mg, Ca, Na, and P, while on the suction side of the vane, the thermal stresses and the mechanical loading most likely lead to a cracking of the oxide scale. A segregation of Cr, Ni, Ta, and Ti to the boundaries of α-Al2O3 grains is observed, which may indicate their diffusion from a metal to a gas atmosphere.

Resistance sintering of CoCrFeNiAlx (x = 0.7, 0.85, 1) high entropy alloys: Microstructural characterization, oxidation and corrosion properties

The cyclic oxidation and electrochemical corrosion behaviors of CoCrFeNiAlx (x = 0.7, 0.85 and 1 mol) high entropy alloys (HEAs) produced by resistance sintering were explored. In accordance with the purpose of this work, CoCrFeNiAlx HEAs with different Al content were cyclically oxidized at temperatures of 800, 875 and 950 °C for 120 h, and their corrosion performances in 0.5 M H2SO4 and 3.5 wt% NaCl solutions were assessed using the potentiodynamic polarization. X-ray diffraction analysis results revealed that BCC and FCC phases along with a minor B2 phase formed in all three HEAs. When the Al content increased to 1 mol from 0.7 mol, the phase fraction of BCC increased by 34.7%, accompanied by an increase in microhardness by 22.6%. The increased Al content had a significant effect in enhancing the cyclic oxidation resistance of CoCrFeNiAlx HEAs. The oxidation rate of all the oxidized HEAs increases with either increasing temperature or decreasing Al content. The best oxidation resistance at temperatures of 800 °C, 875 °C and 950 °C was observed with CoCrFeNiAl HEA, which had minimum weight change data during 120 h of exposure. XRD analysis results revealed that Al2O3 and Cr2O3 were the main oxides that formed on surfaces of the HEAs after oxidation for 120 h. Cross-sectional micrographs showed that the thickness and continuity of oxide scales that formed on HEAs were strongly dependent on the Al content. The improved oxidation behavior of HEAs was attributed to the rise in Al content, which causes the increasing Al2O3 content in oxide scale. Over the performed oxidation duration, neither cracking nor spalling of the oxide scales was witnessed. The effect of Al exhibited a contrary pattern in terms of the oxidation and corrosion behaviors of HEAs. In both 3.5 wt% NaCl and 0.5 M H2SO4 solutions the HEA with the highest Al showed worse corrosion resistance among the HEAs, reflected in the more negative corrosion potential and higher corrosion current density as well as higher corrosion rate. The worsening of the corrosion behavior has been attributed to the increased phase fraction of BCC due to the increased Al content.

From Chaos to Control: Programmable Crack Patterning with Molecular Order in Polymer Substrates.

Cracks are typically associated with the failure of materials. However, cracks can also be used to create periodic patterns on the surfaces of materials, as observed in the skin of crocodiles and elephants. In synthetic materials, surface patterns are critical to micro- and nanoscale fabrication processes. Here, a strategy is presented that enables freely programmable patterns of cracks on the surface of a polymer and then uses these cracks to pattern other materials. Cracks form during deposition of a thin film metal on a liquid crystal polymer network (LCN) and follow the spatially patterned molecular order of the polymer. These patterned sub-micrometer scale cracks have an order parameter of 0.98± 0.02 and form readily over centimeter-scale areas on the flexible substrates. The patterning of the LCN enables cracks that turn corners, spiral azimuthally, or radiate from a point. Conductive inks can be filled into these oriented cracks, resulting in flexible, anisotropic, and transparent conductors. This materials-based processing approach to patterning cracks enables unprecedented control of the orientation, length, width, and depth of the cracks without costly lithography methods. This approach promises new architectures of electronics, sensors, fluidics, optics, and other devices with micro- and nanoscale features.

A micromechanical hydro-mechanical-damage coupled model for layered rocks considering multi-scale structures

An anisotropic Hydro-Mechanical-Damage coupled model was established in the framework of thermodynamics for saturated layered rocks, which contain arbitrarily-distributed cracks of smaller scales and orientated bedding planes of much larger scales. The proposed model considers the anisotropic damage growth and frictional sliding of cracks and shear sliding and dilatancy of bedding planes, as well as the interaction between cracks and bedding planes. Numerical implementation of the nonlinear coupled model was developed based on TOUGHREACT, which is a well-established simulator for multiphase fluid flow and reactive transport analysis. The proposed model was first validated with laboratory water injection test results on rock blocks, and then used to study the effects of injection rates, in situ stresses and bedding planes on fluid injection processes at the field scale, demonstrating the importance to consider the variation of multi-scale structures for better understanding the coupled Hydro-Mechanical responses in layered rocks.

Cause analysis and prevention of fish scale crack defect for steel cord

In the production process of a steel cord, for the surface defect of steel wire with a diameter of 3.42mm in the heavy-duty-drawing process of a semi-finished product, macro and micro-analysis were carried out, and the defect causes were verified through production field tests. It has shown that the unprotected hard collision during the transportation and the shoveling after arriving at the factory results in damages to the surface of the wire rod, which cause high temperature from the friction with the drawing die during the drawing process leading to the appearance of drawing martensite, and it eventually leads to the surface defects of the steel wire. Based on the reasons above, comprehensive preventive measures were formulated to reduce the frequency of occurrence for the surface defects on the steel wire.

Cracking and exfoliation behavior of oxide scale on T91 steel under different tensile stresses in oxygen-controlled lead-bismuth eutectic at 550 °C

The effect of different tensile stresses (0−180 MPa) on cracking and exfoliation behavior of oxide scale for T91 steel was studied in lead-bismuth eutectic containing 10−6 wt% oxygen concentration at 550 °C. The results demonstrate that the exfoliation area of magnetite increased with stress. When tensile stress reached about 180 MPa, only a little magnetite remained; the outer sub-layer of spinel began to crack and locally exfoliated, exposing the inner sub-layer of spinel. The cracks with four different morphologies originated at the outer sub-layer of spinel and propagated inward. A model was proposed to explain these phenomena.

Mixed pooling and richer attention feature fusion for crack detection

• We propose a novel automatic crack detection network. • We design a mixed pooling module. • We adopt channel-wise and spatial attention in different levels. Automatic image crack detection is a critical task for ensuring the safety of various facilities. However, it remains a challenging topic due to the complex background from long and sharp crack topologies. Inspired by recent advances on computer vision applications in deep learning, we propose a novel network architecture with richer feature fusion and attention mechanism and mixed pooling module for crack detection. The proposed network uses the mixed pooling module to replace the conventional spatial pooling. Moreover, we first extract the richer convolutional features to better characterize cracks. Then, we use a spatial attention (SA) in low level feature maps to capture the spatial structure information of cracks. Besides, we use a channel-wise attention (CA) to capture the features of high-level context. Finally, we fuse them together for the final crack prediction. A large crack dataset is used for training and testing. We evaluate our method on a large scale crack dataset, and experimental results on the DeepCrack dataset have demonstrate the effectiveness of the proposed method against state-of-the-art crack detection methods, which achieves Precision( P ) 87.3%, Recall( R ) 88.5%, and F -score over 88%.

Understanding of failure mechanisms of the oxide scales formed on nanocrystalline coatings with different Al content during cyclic oxidation

Cyclic oxidation behaviors of three nanocrystalline coatings with phase constitutions of γ/γʹ, single γʹ, and γʹ/β, respectively, tailored by adjusting the Al content, are investigated. The oxide scale on γ/γʹ coating exhibits serious surface rumpling accompanied by frequently cracking at the crests of undulations after 500 cycles at 1100°C. The premature degradation of the coating in microstructure (i.e. γ′ phase dissolution and concurrent grain coarsening), arising from high oxidation rate and low original Al content, not only increases its CTE to introduce larger thermal stress, but also decreases its hardness and elastic modulus, making it susceptible to rumpling and further inducing cracking of oxide scale. Comparatively, benefited from lower oxidation rate and higher Al content, the degradation of γʹ and γʹ/β coatings is less significant. They are subjected to lower thermal stress but possess higher deformation resistance during cycling. However, local spallation of the oxide scale occurs on γʹ/β coating since it is insusceptible to rumpling and thus the thermal stress is difficult to release in time by plastic deformation. The γʹ coating outperforms its counterparts of γ/γʹ and γʹ/β coatings, presenting low oxidation rate, high resistance to scale cracking and spallation, meanwhile, limited elements-interdiffusion with the underlying substrate.