Tiny Cracks Research Articles

An ultra-lightweight network combining Mamba and frequency-domain feature extraction for pavement tiny-crack segmentation

Pixel-level detection and segmentation of pavement cracks is a crucial but challenging task. Current deep learning methods mainly target larger cracks, often overlooking tiny cracks. Additionally, improvements in these methods typically involve more complex network structures, leading to reduced efficiency and higher computational demands. To tackle these challenges, we propose an Ultra-lightweight Network (ULNet) designed for tiny crack segmentation. ULNet boasts a low parameter count and minimal FLOPs, comprising three main components: the Mamba feature extraction branch, the frequency domain feature extraction branch, and the feature decoder. The Mamba feature extraction branch utilizes the proposed cross-visual Mamba feature extraction module, which enhances the accuracy of tiny crack detection by extracting global features through layer-by-layer processing and cross-fusion of feature information, all without relying on an attention mechanism. The frequency domain feature extraction branch transforms input features into the frequency domain and applies a Gaussian high-pass filter to smooth the image and extract high-frequency crack details while suppressing noise. Finally, the feature decoder incorporates a multi-scale feature selection and fusion module that integrates both channel and spatial attention, enabling effective extraction of tiny crack features from adjacent scales. Inspired by the mechanisms of biological visual information processing, we integrate the high-frequency features from the frequency domain feature extraction branch with the features from the Mamba feature extraction branch. This integration follows the interactive patterns of biological vision, allowing for layer-wise fusion and interaction of features, thereby supplementing the information from the Mamba branch and reducing information loss. We tested and evaluated ULNet on several public datasets (BJN260, Rain365, Sun520, and DeepCrack). Experimental results indicate that our method achieves state-of-the-art performance among all lightweight approaches.

Local-Global Feature Adaptive Fusion Network for Building Crack Detection.

Cracks represent one of the most common types of damage in building structures and it is crucial to detect cracks in a timely manner to maintain the safety of the buildings. In general, tiny cracks require focusing on local detail information while complex long cracks and cracks similar to the background require more global features for detection. Therefore, it is necessary for crack detection to effectively integrate local and global information. Focusing on this, a local-global feature adaptive fusion network (LGFAF-Net) is proposed. Specifically, we introduce the VMamba encoder as the global feature extraction branch to capture global long-range dependencies. To enhance the ability of the network to acquire detailed information, the residual network is added as another local feature extraction branch, forming a dual-encoding network to enhance the performance of crack detection. In addition, a multi-feature adaptive fusion (MFAF) module is proposed to integrate local and global features from different branches and facilitate representative feature learning. Furthermore, we propose a building exterior wall crack dataset (BEWC) captured by unmanned aerial vehicles (UAVs) to evaluate the performance of the proposed method used to identify wall cracks. Other widely used public crack datasets are also utilized to verify the generalization of the method. Extensive experiments performed on three crack datasets demonstrate the effectiveness and superiority of the proposed method.

Investigations of ice impact damage and residual performance of 3D angle-interlock woven composites based on in-situ sampling

This paper investigates the ice impact damage and residual performance of three-dimensional (3D) angle-interlock woven composites. Clear ice projectiles (free of microbubbles and ice cracks) are produced using the unidirectional freezing method, and subsequently launched against composite panels at velocities ranging from 85 m/s to 145 m/s. The impact process and full-field deflection are recorded through a 3D high-speed digital image correlation (DIC) system, and post-impact analysis is conducted via C-scan and microscope observations. Unlike hard body impacts, almost all external damage caused by ice impacts initially appears on the front face rather than the back face. Moreover, to better characterize the impact resistance, an improved compression after impact (CAI) method based on in-situ sampling is proposed, and contour maps of 65%, 80%, and 95% are constructed to illustrate the quantitative relationship between impact energy and residual performance. In particular, the area enclosed by the 80% contour map closely aligns with the damaged area detected by C-scan imaging, reflecting the limitation of C-scan in detecting tiny cracks. Finally, a multi-scale model has also been established for mutual verification with ice impact tests, where the meso-scale model predicts necessary input parameters for the macro-scale model. The current work can provide valuable guidance for developing experimental standards and damage assessment systems for 3D woven composites under hail threats.

Characteristics of Deformation and Damage and Acoustic Properties of Sandstone in Circular Tunnel Morphology under Varying Inundation Depths

When water damage occurs in a mine, variations in the immersion levels of tunnels at different burial depths can be observed. There is a significant relationship between the stability of the surrounding rock and the depth of immersion. Therefore, studying the deformation and damage characteristics of sandstone with circular holes at varying immersion depths, along with their acoustic properties, plays a crucial role in maintaining the stability of water-rich roadways. The TAW-2000 press and static strain system were utilized to investigate the mechanical properties, crack evolution, and deformation field distribution of sandstone with circular holes at varying immersion depths. Additionally, this study analyzed the impact of immersion depth on the characteristic parameters of acoustic emission. The results indicate that immersion depth is negatively correlated with the compressive strength and modulus of elasticity of sandstone; as immersion depth increases, the duration of the compression and yield phases of the rock samples also increases, while the duration of the elastic phase remains relatively unaffected. Furthermore, greater immersion depths correspond to a decrease in the total number of cracks, although the proportion of tensile cracks increases, making the formation of secondary cracks less likely. The frequency of acoustic emission events (transient elastic waves generated by the formation, extension, or closure of tiny cracks within the rock) shows a closely correlated dynamic with the stress–time curve of the rock sample. The acoustic emission ringing counts generated by rock samples under submerged water conditions tend to stabilize with a slight increase before signs of rupture appear. Additionally, the cumulative total energy of acoustic emissions from the rock samples decreases as the water level rises. These research findings provide significant reference value for addressing issues related to water immersion and the extent of water saturation in roadways within rock engineering.

Simulation and Experimental Research of V-Crack Testing of Rail Surfaces Based on Laser Ultrasound

Rail surface cracks are widespread damage that can lead to uneven surfaces of railheads and affect traveling safety. Non-destructive testing is needed to inspect rails regularly to ensure the normal operation of railroads. This paper proposes a laser ultrasonic testing method combining variational mode decomposition and diffractive Rayleigh wave time-of-flight to detect tiny cracks on the rail surface quantitatively. The finite element method was combined with experiments to simulate and experimentally investigate cracks of different sizes numerically. In the numerical simulation, the location of the crack was determined by B-scan. Afterward, the interaction between various types of ultrasound and cracks was comparatively analyzed, and the crack size was quantitatively characterized using useful information from the ultrasound signals. The results show that the time-of-flight method can detect arbitrary cracks with low error. Therefore, the experimentally acquired ultrasound signals used the time difference between the diffracted Rayleigh wave and other ultrasound waves to detect the crack information quantitatively. The variational mode decomposition method was used to separate the ultrasonic signals and extract the best surface wave modes to improve the signal-to-noise ratio. The results show that the combination of variational mode decomposition and time-of-flight method can effectively detect the size of cracks.

Flexural behavior and numerical simulation of reinforced concrete beams with a UHPC stay-in-place formwork

To overcome the shortcomings of traditional concrete structures such as long construction cycle, high construction cost and poor durability, seven composite beams with a UHPC stay-in-place formwork (UCB) and one reinforced concrete (RC) beam were designed. An experimental study was conducted to investigate the flexural behavior of composite beams under the static load. The research parameters include formwork thickness, reinforcement rate, and formwork surface treatment. The results revealed that roughening the surface of UHPC formwork has a positive impact on enhancing the integrity of composite beams. The utilization of UHPC stay-in-place formwork can effectively improve the stiffness, cracking load, and peak load of members. Compared with that of the RC beam, the cracking load and peak load of composite beams increased by 63 % ∼ 103 % and 6 % ∼ 15 %, respectively, and the yielding stiffness increased by 23 % ∼ 41 %. Different from the RC beam, new cracks occurred on one side of the initial crack's end and continued to extend upwards rather than along the original crack. The existence of a significant quantity of tiny cracks in composite beams effectively slowed down the increase in the width of pre-existing cracks in the early stage of loading. Due to the multi-crack pattern exhibited by the UHPC formwork, the strain concentration of the steel bars was effectively mitigated. When subjected to the same load level, the strain of the longitudinal steel bars in composite beams was smaller than that of the RC beam. The formula for calculating the flexural bearing capacity of composite beams is established through theoretical analysis. In addition, the numerical model of the composite beam is established by the finite element method (FEM). The influence of the contact method of the UHPC-NC interface on the flexural performance of the composite beam is explored. The supplementary analysis of the parametric study of the reinforcement rate and the UHPC formwork thickness is carried out. When using ABAQUS for the numerical analysis of the composite beam, it is appropriate to adopt the cohesion model or the Tie model for the UHPC-NC interface, and it is not appropriate to use the Coulomb friction model.

DCNAM: Automatic detection of pixel level fine crack using a densely connected network with attention mechanism

Deep-learning-based crack identification has emerged as a prominent research area in structural health monitoring. Although the detection of common cracks has been the predominant focus in previous studies, the identification of tiny cracks has often been neglected. Efficiently managing thin cracks is vital, because they can threaten the overall structural integrity over time if left unaddressed. We address this gap by targeting thin cracks within a broad category of crack types. We introduce a fine-crack-detection algorithm that efficiently detects both common and tiny cracks. Owing to the limited availability of publicly accessible datasets specifically focused on thin cracks, we collect images of fine cracks to train and evaluate our algorithm. To validate the efficiency of our method, we conduct experiments on three publicly available crack datasets and our private dataset. Compared with the baseline neural network, our proposed approach demonstrates superior performance across all evaluation metrics. Furthermore, our model exhibits impressive generalization ability across the datasets, with the F1 score and mean intersection over union improving by 22.42% and 28.07%, respectively. Notably, our observations indicate that the advantages of the proposed method become more pronounced as the dataset size increases.

Effect of a uniform magnetic field on the nonlinear instability of double cylindrical interfaces concerning three magnetic liquids

The present manuscript relates to the movement of three immersed viscous magnetic liquids in porous media that are positioned between three concentric cylindrical interfaces. Every cylindrical layer has an axial extension that extends in both vertical directions to infinity. Moreover, the pressure gradients are the driving force behind the uniform motion of all fluids in a similar upward motion at varying velocities. Permanent magnetic fields are tangentially oriented exert stress on the system. The computations are rendered simpler by applying the viscous potential theory. The Maxwell equations are utilized for the magnetic field, while the Brinkman-Darcy equations define the fluid mobility. Moreover, the rise of the disturbance and the subsequent extraction of fuel from the tiny cracks of the reservoir stone can be managed through the implementation of engineering techniques such as electromagnetic field impacts and oil extraction engineering. Typically, the nonlinear strategy involves incorporating the applicable nonlinear boundary conditions and analyzing the linearized equations of motion. This research offers a framework for verifying theoretical models and simulations based on experimental observations. The inclusion of a homogeneous magnetic field introduces complexity to the system, rendering it a suitable candidate for validating and improving models in the field Magneto hydrodynamics. The originality of the problem lies in the dual nonlinear stability of cylindrical interfaces when subjected to uniform magnetic field. Accordingly, two nonlinear characteristic differential equations controlling the surface displacements are produced. The nonlinear stability prerequisite is met by applying the matrix concept and the multiple scale technique in conjunction with a theoretical analysis of stability. Additionally, the Routh-Hrutwitz criterion is encompassed to judge the stability cofiguration. A detailed examination of the associated nonlinear stability requirements is showed. Meanwhile, the estimated limited solutions of perturbed surfaces are achieved. For the cylindrical middle layer, it is found that the outer cylindrical interface has a more stabilizing effect than the inner one. The approximate solutions for displacements at the interface are calculated. The influence of Weber numeral of the problem on the stability profile is investigated.

Bottom Crack Detection with Real-Time Signal Amplitude Correction Using EMAT-PEC Composite Sensor.

During electromagnetic ultrasonic testing, it is difficult to recognize small-size bottom cracks by time of flight (ToF), and the lift-off fluctuation of the probe affects the accuracy and consistency of the inspection results. In order to overcome the difficulty, a novel composite sensor of an electromagnetic acoustic transducer (EMAT) and pulse eddy current (PEC) is designed. We use the amplitude of a bottom echo recorded by EMAT to identify the tiny bottom crack as well as the amplitude of PEC signals picked up by the integrated symmetric coils to measure the average lift-off of the probe in real time. Firstly, the effects of lift-off and bottom cracks on the amplitude of bottom echo are distinguished by combining the theoretical analysis and finite element method (FEM). And then an amplitude correction method based on the fusion of EMAT and PEC signals is proposed to reduce the impact of lift-off on the defect signal. The experimental results demonstrate that the designed composite sensor can effectively detect a bottom crack as small as 0.1 mm × 0.3 mm. The signal fusion method can accurately correct the amplitude of defect signals and the relative error is less than ±8%.

An Investigation of Seismic Behaviour of Prestressed Concrete Frame Structures using Pushover Analysis

Abstract: The prestressed concrete system is a prominent material in civil engineering, with applications ranging from buildings, bridges, foundations, piling, silos, stadiums, roadways, and other infrastructure. However, this work is primarily concerned with its applicability to building frame constructions subjected to seismic loading. The prestressed concrete system for frame constructions not only reduces the amount of structural components, but it is also a cost-effective technology that provides outstanding strength and stiffness, as well as quick and simple site erection. Hence, we will be able to use these prestressed concrete components to their full capacity, unlocking engineering and social benefits that were previously impossible. Prestressed concrete designs, as opposed to reinforced concrete designs, can give structures with substantially longer spans, no cracks, or fewer cracks with tiny crack widths. Nonetheless, research on the seismic behaviour of prestressed concrete frame structures is relatively restricted, with the majority of studies focusing on reinforced concrete structures. Finally, the provisional versions of modern building codes such as ACI 318, GB50010, and EC2 do not provide thorough processes for seismic design of prestressed concrete structures. In this work, the seismic design and behaviour of prestressed concrete building frames are explored using the most recent design code. Three model-building codes will be investigated and compared, including ACI 318- 14, Eurocode 2004-2, and Chinese GB500010-2010. The 10-story prestressed concrete frame system is subjected to gravity and seismic loading, and the strong-column weak beam mechanism is archived to conform with the newly implemented building design code. To examine the behaviour of prestressed concrete frame structures, nonlinear static pushover analysis is used to capture the reaction under earthquake loading. The stiffness and strength needed will be designed in compliance with the three major building codes indicated. In addition, the finite element model will be created using CSI Sap200 to capture its actions.

Hierarchical Supramolecular Aggregation of Molecular Nanoparticles for Granular Materials with Ultra High-Speed Impact-Resistance.

The high-speed impact-resistanct materials are of great significance while their development is hindered by the intrinsic tradeoff between mechanical strength and energy dissipation capability. Herein, the new chemical system of molecular granular material (MGM) is developed for the design of impact-resistant materials from the supramolecular complexation of sub-nm molecular clusters (MCs) and hyper-branched polyelectrolytes. Their hierarchical aggregation provides the origin of the decoupling of mechanical strengths and structural relaxation dynamics. The MCs' intrinsic fast dynamics afford excellent high-speed impact-resistance, up to 5600 s-1 impact in a typical split-Hopkinson pressure bar test while only tiny boundary cracks can be observed even under 7200 s-1 impact. The high loadings of MCs and their hierarchical aggregates provide high-density sacrificial bonding for the effective dissipation of the impact energy, enabling the protection of fragile devices from the direct impact of over 200 ms-1 bullet. Moreover, the MGMs can be conveniently processed into protective coatings or films with promising recyclability due to the supramolecular interaction feature. The research not only reveals the unique relaxation dynamics and mechanical properties of MGMs in comparison with polymers and colloids, but also develops new chemical systems for the fabrication of high-speed impact-resistant materials.

Micro-surface crack inspection with arc-shaped bi-grating laser ultrasound spectroscopy

This paper introduces a novel method for detecting and characterizing initial micro-surface crack in metal structures and materials, with a particular focus on initial fatigue cracks. The proposed technique utilizes an arc-shaped double-spaced grating mask to spatially modulate laser beams, forming an arc-shaped bi-grating laser source on the surface of the specimen. This configuration excites two focused surface waves that have the same focus point but differ in frequency and propagate in opposite directions. By monitoring surface wave signals as they pass through defects and analyzing the relative changes in the amplitude of the frequency components in the spectrum of the reflected and transmitted waves, the initial defects can be identified. Both simulation and experimental results validate the effectiveness of this method in detecting tiny artificial surface cracks, including those shorter than 0.3 mm. In addition, this method has been successfully applied to the detection of actual initial fatigue cracks. Experimental findings demonstrate that focused surface waves enable highly sensitive detection of sub-millimeter fatigue cracks.

Residual load-carrying performance of ECC-RC composite beam after drop hammer impact

Four types of specimens with various thicknesses of ECC layer in the tensile side of RC beam are fabricated. The influence of ECC layer on the dynamic response and residual capacity of RC beams is investigated through drop hammer impact test and post-impact static loading test. The results reveal that increasing the thickness of the ECC layer in the tensile side leads to less major cracks and more tiny cracks on the beam side under impact, and the damage of concrete in the impact zone alleviates. The incorporation of the ECC layer plays an insignificant role in displacement restitution and residual load-carrying capacity. However, ECC-RC composite beams effectively mitigate displacement ductility degradation. Compared to RC beam, ECC-RC composite beams also exhibit superior energy dissipation capacity at the post-impact loading stage. The parametric analysis indicates that a thicker ECC layer in composite beams with a low tensile reinforcement ratio effectively reduces maximum impact displacement but has a negligible influence on residual load-carrying capacity. In contrast, increasing the ECC layer in beams with a high tensile reinforcement ratio offers a limited reduction in maximum impact displacement but its residual load-carrying capacity significantly improves after impact.

Surface crack detection on selective laser melting printed Inconel 718 using a laser generated ultrasound technique and phase space reconstruction

Abstract In the field of metallic additive manufacturing, Selective Laser Melting (SLM) has become a predominant technology due to its advantages of short production cycles, high precision, and low cost. It is frequently employed in the production of complex parts. This paper proposes the use of a scanning laser line source, in conjunction with the singular value decomposition method, to reconstruct phase space and identify surface cracks in SLM specimens. The scanning laser line source addresses the limitations of a single line source, which is often unable to accurately detect tiny cracks. By comparing experimental and simulation data, the results demonstrate that the scanning laser line source can effectively compensate for some of the detection deficiencies of a single line source.

How Tiny Cracks Lead to Large-Scale Faults

Researchers could soon gain new insights into fault development in Earth’s brittle crust, thanks to a computational approach that harnesses experimental observations of microscale rock damage.

EAFNet: Extraction-amplification-fusion network for tiny cracks detection

Tiny cracks are often overlooked in the inspection process, causing huge economic losses and dangerous accidents. Therefore, tiny cracks should be detected in a timely and accurate manner to eliminate the disease at the initial stage. Inspired by the fact that humans are more likely to capture conspicuous information when observing objects, we propose a novel three-stage Extraction-Amplification-Fusion network (EAFNet). Specifically, in the extraction stage, we utilize an effective backbone network for feature extraction of tiny cracks. In the amplification stage, we design a Tiny Feature Amplification (TFA) module to amplify the extracted features. In the fusion stage, we propose a Two-Branch Fusion (TBF) module to fully fuse the feature maps at different resolutions. To make tiny crack information more ‘conspicuous’, we propose an activation function TinyReLU to enhance the contrast of the tiny cracks with the background. In addition, we construct a Tiny Crack (T-CRACK) dataset with six different backgrounds and a Cross-scale Crack (C-CRACK) dataset. On both datasets, EAFNet achieves an advantage over the existing 8 advanced networks. The two datasets are available at: https://github.com/EAFNet/EAFNet.

Pixel-Level Crack Identification for Bridge Concrete Structures Using Unmanned Aerial Vehicle Photography and Deep Learning

Traditional manual inspection technology has the problems of high risk, low efficiency, and being time-consuming in bridge safety management. The unmanned aerial vehicle (UAV)-based detection technology is widely used in bridge structure safety monitoring. However, the existing deep learning-based concrete crack identification method has great limitations in dealing with complex background and tiny cracks in bridge structures. To address these problems, this study designs a crack pixel-level high-performance segmentation model for bridge concrete cracks that is suitable for UAV detection scenarios using machine vision (MV) and deep learning (DL) algorithms. First, considering the high requirements for the computing performance of the MV-based model for UAV-based detection, the ResNet-18-based lightweight convolutional neural network is used to represent the traditional large-scale backbone network of the pyramid scene parsing network (PSPNet) to develop a high-performance crack automatic identification model. Then, considering that bridge concrete cracks have the characteristics of subtle shapes and complex backgrounds, the spatial position self-attention module is inserted into the PSPNet to improve its detection accuracy. A concrete bridge is used for the case study, and a dataset of cracks in bridge concrete structures collected by UAVs is constructed and used for model training. The experimental results show that the loss function of the developed method in the training process results in a smooth decline, and the developed algorithm achieves the evaluation indicators of 0.9008 precision, 0.8750 recall, 0.8820 accuracy, and 0.9012 IOU on the bridge concrete crack dataset, which are significantly higher than other state-of-the-art baseline methods. In addition, four common UAV bridge detection scenarios, including low light, complex crack forms, high background roughness, and complex background scenes, are used to further test the crack detection ability of the developed crack identification model. The experimental results show that the proposed crack identification method can effectively overcome interference and real-size pixel-level segmentation of crack morphology. In addition, it also achieved a detection efficiency of 35.04 FPS, which shows that the real-time detection ability of the method has good applicability in the UAV detection scene.

Using Sandwiched Silicon/Reduced Graphene Oxide Composites with Dual Hybridization for Their Stable Lithium Storage Properties.

Using silicon/reduced graphene oxide (Si/rGO) composites as lithium-ion battery (LIB) anodes can effectively buffer the volumetric expansion and shrinkage of Si. Herein, we designed and prepared Si/rGO-b with a sandwiched structure, formed by a duple combination of ammonia-modified silicon (m-Si) nanoparticles (NP) with graphene oxide (GO). In the first composite process of m-Si and GO, a core-shell structure of primal Si/rGO-b (p-Si/rGO-b) was formed. The amino groups on the m-Si surface can not only hybridize with the GO surface to fix the Si particles, but also form covalent chemical bonds with the remaining carboxyl groups of rGO to enhance the stability of the composite. During the electrochemical reaction, the oxygen on the m-Si surface reacts with lithium ions (Li+) to form Li2O, which is a component of the solid-electrolyte interphase (SEI) and is beneficial to buffering the volume expansion of Si. Then, the p-Si/rGO-b recombines with GO again to finally form a sandwiched structure of Si/rGO-b. Covalent chemical bonds are formed between the rGO layers to tightly fix the p-Si/rGO-b, and the conductive network formed by the reintroduced rGO improves the conductivity of the Si/rGO-b composite. When used as an electrode, the Si/rGO-b composite exhibits excellent cycling performance (operated stably for more than 800 cycles at a high-capacity retention rate of 82.4%) and a superior rate capability (300 mA h/g at 5 A/g). After cycling, tiny cracks formed in some areas of the electrode surface, with an expansion rate of only 27.4%. The duple combination of rGO and the unique sandwiched structure presented here demonstrate great effectiveness in improving the electrochemical performance of alloy-type anodes.

An efficient unstructured quadrilateral-dominated surface mesh generation method for arbitrary geometry with tiny features or crack defects in DiBFM

An efficient unstructured quadrilateral-dominated surface mesh generation algorithm is presented to generate high-quality surface elements for arbitrary geometry with tiny features or crack defects. For complex entities with proximity features, a novel type of meshing template has been proposed to solve the problem of entity boundary fitting of the outside-in grid-based method. The algorithm also works for regions with proximity features, as well as for arbitrary geometry with one or multiple cracks. The procedure incorporates aspects of well-known meshing techniques, but includes some original and efficient processes. The refinement field is constructed according to mesh size, boundary curvature, surface curvature and other geometry features. In order to effectively generate the core mesh, a fast intersection algorithm between topological elements and the solid boundary is established, which improves the efficiency of intersection calculation. The geometry-based techniques are employed for boundary matching of the core mesh. The method has been successfully employed in conjunction with the dual interpolation boundary face method (DiBFM). Several realistic geometries with cracks or proximity features are applied to demonstrate the accuracy and validity of the proposed approach.

Preparation of Low-Viscosity Epoxy Resin Sealing Agent and Evaluation of Injection, Plugging, and Degradation Properties.

The technology of water plugging and increasing production in high water cut reservoirs of low permeability is a common problem in the industry. Epoxy resin, displaying excellent mechanical properties and adherent performance, can easily inject a tiny crack, forming a long-term blocking barrier. This study aimed to investigate an easily injectable degradable epoxy resin sealing material. The injectable performance, long-term stability, and mechanical and plugging properties were comparatively analyzed in the fractured core, and the degradable performance was discussed in the degrading solution. The result showed that the range of R (R is the ratio of EOG and MHHPA) from 1 to 1.1 and the mass fraction range of EMI from 0.01 to 4 wt % are the optimal formulations (EOGM). The curing time from 1 to 12 h could be regulated by adjusting the dosage of EMI, as well as the strength being more than 60 MPa. The plugging agent's initial viscosity is lower than 100 MPa s at 20 °C and injecting pressure is lower than 0.1 MPa. After curing for 24 h, compressive strength was more than 72.76 MPa, 3.6 times higher than that of cement, and the adhesion strength was 4.41 MPa when the contact area was 75.93 cm3. Breakthrough pressures for sealing 1-5 mm fractures were all more than 10 MPa, and the breakthrough pressure for 1 mm crack even reached 29.4 MPa. Epoxy resin/acid anhydride system could be degraded in a mixed solution of phenol-potassium salt-heavy aromatics within 7 days at 60-100 °C, which reduced the plugging well risk of the epoxy resin plugging agent. These results suggest that an epoxy resin/acid anhydride plugging agent can be employed effectively and safely for the injection of tiny cracks, which is of great engineering significance.