Issue 02, 2026
Experimental study on the morphological characteristics of wildfire front driven by meteorological conditions
LIU Chang;JI Kunpeng;ZHOU Yiqi;LI Peng;HAN Jingshan;LI Junhui;[Objective] The morphology of large-scale wildfire fronts is a crucial factor in assessing wildfire spread and determining safe distances for transmission corridors. Understanding the dynamic evolution of these fronts is essential for effective real-time wildfire monitoring and prevention. Most existing research focuses on small-scale experimental scenarios that struggle to replicate the complex interactions among multiple factors, such as fuel types and meteorological parameters, during actual wildfires. Few studies have examined fire front morphology in large-scale scenarios that incorporate meteorological conditions, and current research often falls short of the accuracy needed for effective wildfire prevention and control. This study aims to investigate the effects of different fuel types and meteorological conditions on wildfire front morphology through large-scale experiments, providing experimental data and theoretical support for the development of fire front spread models applicable to actual wildfire situations. [Methods] A 50 m×40 m full-scale wildfire combustion experimental platform was constructed for this study, integrating UAV thermal infrared imaging, tower-based visual monitoring, and multi-source meteorological sensing systems. We conducted large-scale, systematic experiments on wildfire spread using two common surface fuels(wheat straw and pine needle litter) under varying meteorological conditions. The analysis focused on the effects of fuel type, wind direction, and wind speed on key parameters, including fire front morphology, temporal variation in stable fire front length, and fire front propagation angle. We systematically compared surface fire-front spread under various working conditions. [Results] The results revealed the following: 1) Fire front morphology is significantly affected by fuel type, where wheat straw produces a smooth arc-shaped front, and pine needles result in a sharp, multi-branched, and irregular morphology. The fire front angle increases continuously during combustion, with the temperature decay rate in the burned area of pine needles being significantly faster than that of wheat straw. Wind direction dictates the overall spread direction of the front, whereas wind speed primarily affects the size of the front angle. 2) The variation trend and fluctuation amplitude of stable fire front length are jointly influenced by fuel and meteorological conditions. The fire front length of wheat straw decreases steadily over time, whereas that of pine needles exhibits significant short-term oscillations. Greater differences in maximum and minimum wind directions lead to more intense fluctuations in fire front length. Under identical wind directions, higher average wind speeds correspond to greater extreme values of fire front length. 3) The fire front propagation angle gradually decreases during the spread process. The wheat straw fire front is generally smooth with minor fluctuations, whereas the pine needle fire front displays significant local curvature and irregular trajectories. Greater stability in wind direction and higher average wind speeds result in a smaller average fire front propagation angle, causing the front to approach a straighter line. [Conclusions] Through large-scale surface fire spread experiments, this study elucidates the influence of fuel type and meteorological conditions on key parameters such as fire front morphology, temporal variation of stable fire front length, and fire front propagation angle. It reveals the comprehensive influence mechanism between meteorological conditions and fuel properties regarding fire front morphology, offering a large-scale experimental basis and critical parameter support for developing wildfire spread prediction models and improving wildfire prevention and control strategies in transmission corridors. Future research will expand these large-scale wildfire experiments to include more complex scenarios, thereby enhancing our understanding of real wildfire behavior.
Experimental platform for fire dynamics in oil-filled electrical equipment
SUN Ruibang;ZHANG Xinwei;SHANG Fengju;LIU Chang;[Objective] Oil-filled electrical equipment is widely used in substations/converter stations and hydropower projects as key facilities for securing power supply. However, transformer oil leaks from the rupture, accumulates at the base of oil-filled equipment, and encounters an ignition source, forming an external heat source that leads to a fire in the oil-filled equipment. Notably, heat from this external source is transferred back to the equipment via conduction, convection, radiation, and other pathways. This accumulated internal heat causes oil from ruptured oil-filled equipment to be sprayed and ignited by an external heat source, forming a jet fire. The transition to a jet fire triggers nonlinear shifts in the system's original state, leading to further deterioration in the degree of fire hazard. As fire incidents involving oil-filled equipment pose a major safety hazard in the electric power industry, effective and reliable prevention and control strategies must rely on a precise understanding of their dynamics. Therefore, a comprehensive and profound understanding of the characteristics of oil-filled equipment jet fire dynamics under the influence of external heat sources is of practical importance for improving the fire prevention and control capabilities of the electric power industry and for developing major fire monitoring and early warning technologies. In essence, once internal transformer oil leaks and burns, fire development mainly experiences two typical stages: first, the instability of oil-filled equipment combustion under the influence of an external heat source to form a jet fire, and second, the formation of a jet fire, which considerably changes the typical characteristics of the fire parameters. An oil-filled-equipment fire is a combination of combustion phenomena of multiple fire modes, such as bottom pool fire, sidewall flow fire, and top jet fire. [Methods] Previous studies have focused on single-mode fire dynamics experiments. Little attention has been paid to key scientific issues, such as the evolution of typical fire characteristics and the prediction of fire behavior under combined combustion of multiple fire modes, which have posed considerable challenges for the prevention and control of fires in electric-power charging equipment and fire rescue missions. In view of oil-filled equipment jet fire accidents and their complex fire characteristics, issues in electric power fire prevention and control remain serious challenges. Fundamental scientific questions regarding these dynamics remain unresolved, necessitating further theoretical studies to address the safety challenges they pose. [Results] In this study, we designed and constructed an experimental platform to simulate the fire dynamics of oil-filled electrical equipment. By integrating the measurement systems for the mass-loss rate, temperature, radiation heat flux, and image acquisition, we clarified the basic combustion phenomena of these fires. Furthermore, we established the typical phases and morphological characteristics and revealed the evolution laws of the characteristic parameters, such as the flame height, flame temperature, and flame radiation. [Conclusions] We identified the catastrophic mechanism of oil-filled equipment jet fire at its essence, further enriching the theory of fire dynamics and providing strong scientific and technological support for enhancing fire prevention and control strategies within the power industry.
Design of an experimental platform for controlling fine particulate matter using high-voltage electrically charged liquid droplets
TENG Chenzi;ZHANG Yun;REN Sida;Beijing Academy of Science and Technology;[Objective] The average concentration of fine particulate matter(PM2.5) in China significantly exceeds the World Health Organization–recommended standards, thereby undermining the environmental benefits achieved through ecological improvements. Traditional ultra-low-emission particle control strategies lack systematic investigation of method synergy and multi-parameter coupling, resulting in a gap between current practices and actual flue gas control objectives. To meet the stringent requirements for ultra-low emission of fine particulate matter, the wet electrostatic precipitator, as a preferred option for construction and retrofitting, offers notable advantages. However, it often faces challenges, including high water consumption, secondary emission pollution, and elevated operating costs. [Methods] Previous studies have demonstrated that electrohydrodynamic atomization has considerable potential for enhancing fine particle removal. This technique promotes collision, interception, and coalescence of particulate matter at reduced water consumption, thereby improving water-based removal efficiency for fine particles. Guided by the principles of efficient and low-consumption fine particle control, this study integrates electrohydrodynamics, aerosol mechanics, and ventilation control theories. By analyzing electrostatic precipitation devices reported in recent years and referencing single-parameter models, engineering prototypes, and relevant parameters, an experimental platform was designed and constructed. The platform employs high-voltage electro-coupled charged liquid droplets generated via electrokinetic atomization in combination with an electrostatic field to control fine particles; it consists of a pollutant generation system, a high-voltage power supply system, a multi-parameter coupled dust removal system, and a measurement and analysis system. Key parameters, including corona electrode configuration, dust collection electrode design, electrokinetic atomization settings, and rapping ash cleaning mechanisms, are continuously adjustable, enabling multidimensional collaborative coupling control to optimize fine particle removal performance. [Results] Using this experimental platform, the effects of key parameters, including electric field strength, residence time, and flue gas concentration, on the motion characteristics, spatial distribution, and removal efficiency of fine particles were systematically investigated. The dynamic evolution mechanism of particle capture under coupled operational control parameters, airflow characteristics, and electrokinetic atomization was elucidated. These findings provide a theoretical basis and technical support for optimizing efficient fine particle capture and offer important implications for advancing collaborative aerosol control strategies. [Conclusions] The results demonstrate that fine particle control using high-voltage electro-coupled charged liquid droplets integrates the advantages of electrohydrodynamic atomization and electrostatic fields, effectively promoting coalescence and agglomeration of fine particles into larger ones. Under the synergistic action of the electrostatic field, charged liquid droplets enhance particle capture efficiency across all size ranges, significantly reducing fractional penetration compared with a dry electrostatic precipitator. Further increases in electric field strength amplify the effectiveness of charged liquid droplets in particle removal. Moreover, under long-term operation, the dust removal device maintains clean plate surfaces and consistently high particle capture efficiency.
Large-scale experiments on surface fire spread rate
LIU Chang;JI Kunpeng;ZHANG Sihang;LI Peng;HAN Jingshan;YANG Zhi;[Objective] Surface fire spread rate is a key parameter for characterizing surface fire behavior, and understanding its variation patterns is of great significance for wildfire prevention and control. Existing research predominantly relies on small-scale experiments, which limits the applicability of fire spread models built on these data to real wildfire scenarios. This study aims to investigate the effects of fuel type and fuel bed density on surface fire spread rate through large-scale experiments, thereby providing an experimental basis and theoretical support for the development of fire spread models applicable to actual wildfires. [Methods] Utilizing a large-scale power grid wildfire experimental platform, this study conducted surface fire spread experiments under different fuel types(shrubland surface litter and coniferous forest surface litter) and fuel bed densities(1.0, 1.5, and 2.0 kg/m2) within a 2 000 m2 combustion area. The experimental setup included an array of 99-K-type thermocouples to collect surface temperature data, unmanned aerial vehicles to record visible and infrared imagery of the fire spread process for extracting fireline morphology, and a small weather station to monitor real-time meteorological conditions such as wind speed and direction. By analyzing the flame front spread rate, fireline expansion rate, and temperature response characteristics, surface fire spread behavior under various working conditions was systematically compared. [Results] The experimental results demonstrate the following points. 1) The flame front spread rate is comprehensively regulated by fuel type, fuel bed density, and meteorological conditions. The acceleration phase of the flame front occurs earlier in shrubland surface litter than in coniferous forest surface litter. Increasing the fuel bed density reduces the peak flame front spread rate while enhancing combustion stability. Meteorological factors are the primary cause of the observed multipeak fluctuations in the spread rate. 2) The fireline expansion rate exhibits fluctuating characteristics, with peak values determined by fuel type. The peak fireline expansion rate of shrubland surface litter is greater than that of coniferous forest surface litter. An increase in fuel bed density promotes fireline expansion in shrubland surface litter but inhibits it in coniferous forest surface litter. 3) The temperature response characteristics reflect flame front spread and fireline expansion behaviors. Fuel type governs the continuity of fire head spread; the loose structure of shrubland surface litter facilitates uniform heat transfer, whereas the compact structure of coniferous forest surface litter leads to heat accumulation. Fuel bed density influences the speed and spatial direction of the temperature response by modifying internal oxygen supply and combustion completeness. [Conclusions] Through large-scale surface fire spread experiments, this study clarifies the influence of fuel type and fuel bed density on flame front spread rate, fireline expansion rate, and temperature response characteristics. The resulting dataset provides large-scale experimental support for developing predictive fire spread models for actual wildfires and offers valuable insights for wildfire prevention and control along power transmission lines. Future work will involve conducting large-scale wildfire experiments under different slope terrains to deepen the understanding of real wildfire spread behavior.
Rolling force prediction modeling for strip cold rolling based on mechanism-data fusion
SUN Youzhao;WANG Xiangchen;LI Jingdong;WANG Xiaochen;SUN Yamin;YANG Quan;[Objective] The precise estimation of rolling force during the process of cold continuous rolling is of paramount importance for ensuring product quality, enhancing automation levels, improving production efficiency, and optimizing process settings. However, the conventional cold rolling force mechanism model often relies solely on process parameters during the cold-rolling stage. It disregards the genetic effects of the hot rolling process on the material's structure and properties, and it cannot effectively capture the complex, nonlinear impact of cross-process parameters on the rolling force. This results in limited prediction accuracy and generalization ability. This study proposes a cold rolling force prediction model based on Bayesian optimization and an improved light gradient boosting machine(BO-LightGBM), aiming to comprehensively explore the process coupling between hot and cold rolling. The model aims to enhance adaptability and accuracy in predicting force during cold rolling across various steel grades and production scenarios. [Methods] The modeling process involves the development of a multi-source feature system, incorporating 7-dimensional hot rolling parameters(e.g., finish rolling temperature, coiling temperature, final thickness, etc.), 11-dimensional cold rolling parameters(e.g., strip width, rolling speed, deformation resistance coefficient, etc.), and the predicted output from the traditional mechanism-based model. This comprehensive feature set enables the model to represent a cross-process fusion of variables that collectively influence rolling force. It is acknowledged that there is variability and coupling across different rolling stands in a tandem cold rolling mill. Therefore, a stand-specific modeling strategy is employed. The development of independent prediction models tailored to each stand facilitates the capture of local process characteristics, nonlinear interactions, and contextual dependencies. Furthermore, a Bayesian optimization algorithm is employed to automatically fine-tune the hyperparameters of the BO-LightGBM model for each stand. This approach effectively reduces human intervention, avoids suboptimal manual tuning, and enhances the efficiency and robustness of the learning process. [Results] The impact of incorporating upstream process data was evaluated by training rolling force prediction models on two distinct datasets. The first dataset contained only cold rolling parameters, while the second dataset included both hot and cold rolling parameters. A series of comparative experiments have demonstrated that 1) following the implementation of the hot rolling process parameters, the mean absolute error in rolling force prediction for each stand has been shown to decrease by an average of 1.803 t, whereas the root mean square error has been demonstrated to decrease by an average of 2.573 t, thereby indicating a substantial enhancement in accuracy. 2) In a real industrial cold rolling setting system, the model has been found to enhance the rolling force setpoint accuracy for MR T-4 CA and MR T-5 CA steel grades by 2.024% and 1.962%, respectively, thereby underscoring its practical engineering value and operational significance. [Conclusions] The proposed BO-LightGBM rolling force prediction model demonstrates excellent performance in terms of accuracy, robustness, and generalization. The model effectively incorporates upstream hot rolling data and employs stand-specific learning with automated hyperparameter optimization, thereby capturing the hereditary influence of the hot rolling stage and overcoming the limitations of traditional mechanism models in cross-process modeling. The model offers a promising data-driven solution for intelligent process control in modern steel rolling operations and supports the advancement of smart manufacturing in the metallurgical industry.
Experimental design of traffic sign detection and recognition in complex scenes based on YOLO-BSNM
SONG Jun;CHU Zhihan;FAN Zihao;ZOU Ben;JIAO Wanguo;[Objective] This study aims to address the critical challenges associated with recognizing small traffic signs in autonomous driving scenarios. These challenges are particularly pronounced in highly dynamic and visually cluttered environments and adverse meteorological conditions, as well as when the signs are far away. [Methods] To achieve enhanced recognition performance, this study adopted an improved systematic detection method based on the YOLO v11n architecture. In particular, an optimized framework, referred to as YOLO-BSNM, was constructed for complex scenarios by integrating multichannel attention(MCA) with a channel–height–width three-branch dynamic fusion mechanism, a normalized weighted loss(NWDLoss), a 160×160-pixel high-resolution small-object detection head, and a Bi FPN feature fusion module. To enhance the discriminability of features of small objects, MCA employed a channel–height–width three-branch dynamic fusion mechanism. NWDLoss addressed the localization sensitivity issues of traditional IoU-based losses via probabilistic distribution matching. The small-object detection head incorporated a sampling–fusion–extraction layer to mitigate detail loss resulting from resolution decay in deep features. Finally, the BiFPN feature fusion module performed the weighted fusion of multiresolution feature maps to reduce the critical feature loss. [Results] The experimental results demonstrated that the improved YOLO-BSNM algorithm achieved a precision(P) of 81.7%, a recall(R) of 75.4%, and an mAP50 of 83.4% when using the custom dataset, indicating substantial advancements over the baseline algorithm. Concurrently, the model size was smaller by 0.62 million parameters through lightweight optimization, achieving higher detection accuracy with fewer computational resources. The results also indicated that the framework showed enhanced adaptability to challenging scenarios, such as environments with blurring and occlusions. The edge-device deployment capability of the algorithm offers a robust foundation for fulfilling the real-time detection needs of intelligent transportation systems. [Conclusions] This study successfully developed an efficient and accurate traffic sign identification technology based on the YOLO-BSNM model. This technology overcame the limitations of traditional recognition methods and showed improved efficiency and precision in recognizing traffic signs. It provides substantial support for advancing autonomous driving. Deployment and performance verification on an embedded Raspberry Pi 4B platform demonstrated that the optimized model achieved a good balance between architecture lightweightness and detection accuracy, effectively reducing the false alarm and missed detection rates. This advancement meets the objective of enhancing performance while minimizing parameter overhead. YOLO-BSNM accomplished simultaneous enhancements in micro-object feature discrimination and multiscale information integration through the optimized combination of an attention mechanism and a feature fusion strategy. The probabilistic distribution matching–based loss function enhanced the stability of bounding box regression. These technical innovations led to an efficient detection framework, providing a novel approach for detecting micro-objects in complex environments. The lightweight detection system represents a promising solution for deployment in resource-constrained scenarios, such as drone-based remote sensing and satellite image analysis. The integration of multimodal data from infrared sensors, LiDAR, and dynamic environment adaptation algorithms enables the model to overcome detection challenges under variable lighting and extreme weather conditions. This research provides a technical framework for autonomous driving applications and elicits new avenues for investigating universal micro-object detection solutions in computer vision.
Predicting the remaining useful life of ion etching systems using a spatiotemporal graph convolutional network
ZHENG Liyang;XU Jian;WU Jia;LONG Hao;ZHU Cuitao;[Objective] In semiconductor manufacturing, the remaining useful life(RUL) of equipment must be accurately predicted to ensure production efficiency and minimize economic losses. However, this task is fraught with substantial challenges. The heterogeneity of multisource sensor data, which encompass various signal types and measurement scales, poses a complex data integration problem. Meanwhile, the scarcity of key failure samples makes it arduous to train reliable prediction models. Traditional prediction methods, which are based on single-dimensional modeling, struggle to capture the intricate physical coupling relationships among different components of the equipment. Moreover, they cannot adequately reproduce the evolution laws of cross-temporal and spatial states during equipment degradation. Hence, these methods have limited prediction accuracy, lack interpretability, and cannot meet the high demands of modern semiconductor manufacturing processes. This study addresses these issues by developing a spatiotemporal joint modeling method that integrates a temporal convolutional network(TCN) with a graph convolutional network(GCN). The joint modeling aims to achieve an in-depth multiscale analysis of the dynamic degradation laws of equipment. [Methods] First, a learnable GCN is constructed. Based on the physical topology of sensors installed on the equipment, the GCN is designed to model the spatial relationships among different sensor nodes. Through a multi-order neighborhood information aggregation mechanism, the GCN effectively extracts the hierarchical spatial correlation features of the equipment. This process allows the model to understand the interactions among different components and their influences on each other in the spatial domain. Next, the TCN with a hierarchical dilated convolution architecture plays a vital role in handling time-series data. The dilated convolution layers capture the long-term trend features of equipment degradation without sacrificing the ability to detect local fluctuation patterns. By employing stacks of multiple layers with different dilation rates, the TCN analyzes the time-series data at various time scales to reveal hidden patterns and trends in equipment degradation. Finally, a spatiotemporal interaction module deeply fuses the spatial and temporal feature data extracted by the GCN and TCN. This integration enables the model to comprehensively examine the spatial and temporal aspects of the state of the equipment, facilitating accurate regression prediction of the RUL. [Results] The performance of the proposed model was evaluated via rigorous experiments on the dataset of an industrial ion etching system. The results demonstrated that the proposed model performed remarkably better than traditional models. Specifically, in terms of the root mean square error(RMSE) prediction index, the proposed model achieved a reduction of 20.6% compared to the long short-term memory(LSTM) model. The LSTM model, which is widely used in time-series prediction, often struggles in capturing complex spatial relationships. In contrast, the proposed model integrating the GCN effectively overcomes this limitation. Compared to that of a convolutional neural network(CNN), the RMSE of the proposed model was lower by 23.3%. CNNs are mainly designed for extracting spatial features in images, and their application in equipment RUL prediction without considering the unique spatiotemporal characteristics of equipment data leads to suboptimal performance. The specialized spatiotemporal architecture of the proposed model provides a more suitable solution. Compared to that of the multilayer perceptron(MLP), the RMSE of the proposed model was lower by 29.9%. The simple fully connected structure of the MLP is incapable of effectively modeling spatial and temporal dependencies, highlighting the advantage of the architecture of the proposed model. Additionally, when compared with those of the classical TCN and the Transformer network, the RMSE of the proposed model was lower by 11.0% and 13.7%, respectively, further validating the effectiveness of the joint spatiotemporal modeling approach. [Conclusions] This study not only provides an innovative and effective method for the predictive maintenance of industrial equipment but also has far-reaching implications for the semiconductor manufacturing industry. Enabling highly accurate RUL predictions, it helps enterprises optimize their maintenance strategies. Such optimization can substantially reduce unplanned downtime losses during wafer manufacturing, improving production efficiency and reducing costs. Furthermore, it paves the way for shifting the paradigm of intelligent manufacturing from a reactive post-failure response approach to a proactive pre-failure intervention strategy. This transition can enhance the overall reliability and competitiveness of the semiconductor manufacturing industry, driving it toward a more intelligent and sustainable future.
Theory and experimental research on subwavelength dielectric gratings
XU Canhua;MAO Mengyao;XUE Kongsong;DENG Yongsheng;XU Zhongwei;[Objective] This study undertakes a comprehensive investigation of the phase modulation properties of subwavelength dielectric gratings(SWDGs), emphasizing the role of key structural parameters—namely, grating period, duty cycle, substrate refractive index, and grating height—in determining the resulting phase retardation. This research aims to demonstrate that SWDGs function as highly tunable optical elements that enable precise two-dimensional light-field manipulation, thereby offering distinct advantages over conventional birefringent waveplates in terms of angular stability, broadband performance, and integration flexibility. Ultimately, this study seeks to establish theoretical and experimental frameworks to support the implementation of SWDGs in advanced photonic platforms, including metasurfaces, polarization control devices, and integrated optical systems. [Methods] A dual approach was employed, integrating numerical simulations and experimental measurements. The finite element method(FEM) was employed in COMSOL Multiphysics to develop a rigorous electromagnetic model of the SWDG, enabling the simulation of phase retardation under variable structural parameters and incident conditions. The behavior of TE-and TM-polarized waves, their transmission efficiency, and the resulting phase difference were the specific focus of the simulations. In the experimental phase, a quartz-based subwavelength grating with a period of approximately 826 nm and a height of 1 280 nm was fabricated via ultraviolet nanoimprint lithography. A bespoke optical configuration was engineered to assess the phase retardation characteristics at 1 550 nm under normal incidence. The system employed Mueller matrix polarimetry, using a laser source, linear polarizers, and a rotating analyzer, in conjunction with a power meter. The resulting intensity profiles as a function of analyzer angle were recorded and fitted using MATLAB to extract the exact phase delay and optical axis orientation. [Results] The phase retardation is principally dictated by the grating height and the substrate's refractive index. An approximate linear relationship with height is observed, with a monotonic increase at higher refractive indices. The impact of variations in grating period on phase delay was found to be minimal, whereas the duty cycle showed a nonlinear effect, with an optimal value of approximately 0.4. The experimental measurements corroborated the numerical predictions, yielding a phase retardation of 0.44 rad, which closely matched the simulated value of 0.43 rad, resulting in a minor error of only 2.3%. The optical axis orientation was approximately 1.71 rad. The minor discrepancies observed between the simulation and the experiment were attributed to three factors: fabrication imperfections, slight misalignments in the optical path, and non-ideal polarization elements. The SWDG exhibited high transmission and consistent performance across a range of incident angles, thereby underscoring its robustness and suitability for practical applications. [Conclusions] This study successfully illustrates that SWDGs can be designed and fabricated to achieve tailored phase retardation. This offers a versatile and efficient alternative to conventional waveplates. The strong correlation between simulation and experimental results validates the use of FEM-based modeling for the design and optimization of SWDGs. Notable advantages of these gratings include broad angular acceptance, wavelength flexibility, and compatibility with standard nanofabrication processes. These characteristics render them highly promising for applications in metasurfaces, adaptive optics, optical sensing, and on-chip photonic systems. Subsequent research endeavors should explore dynamic and reconfigurable grating designs, as well as their integration with other functional optical elements to further expand their utility in next-generation optical technologies.
Application of an integrated system for in situ sample preparation and hydraulic fracturing to unconsolidated sandstone
LI Shuqian;LIU Wei;DENG Jingen;TAN Qiang;XU Kaikai;[Objective] Recently, hydraulic fracturing has been widely applied in unconsolidated sandstone reservoirs, leading to the development of an integrated fracturing and sand control technique called fracture packing. Unconsolidated sandstone typically has high porosity and permeability, poor cementation effects, and low strength. Furthermore, the mechanisms of fracture initiation and propagation are complex, involving multiple rock deformation and failure modes, such as tensile, shear, and plastic compaction. These mechanisms are not yet fully understood, making it difficult to optimize fracturing fluid and process parameters and to control fracture morphology. A major reason for this knowledge gap lies in the experimental methods used. Conventional laboratory methods involve separate sample preparation and fracturing processes, which introduce stress disturbances and pressure changes during cementing. These issues are particularly severe in unconsolidated sandstone, potentially compromising the accuracy of the results. To address these challenges, this study developed a new experimental methodology to more accurately investigate the hydraulic fracturing behavior of unconsolidated sandstone. [Methods] We developed a large-scale experimental setup for the physical simulation of in situ sample preparation and the true triaxial hydraulic fracturing of unconsolidated sandstone. It allowed continuous operation from sample preparation to fracture propagation under stable stress conditions, eliminating stress disturbances caused by sample transfer. Our methodology involved the designing of an artificial unconsolidated sandstone formula based on natural core analysis. Chemical cementing agents were deliberately avoided to replicate realistic formation strength properties. Each experiment began by establishing a simulated wellbore model, followed by filling a sand mixture inside the load frame. After the target stresses stabilized, the fracturing simulation began immediately without sample movement. Using this setup, we conducted a series of comprehensive tests to investigate the effects of varying permeability on hydraulic fractures in unconsolidated sandstone reservoirs. Furthermore, by leveraging the setup's sample preparation efficiency, we explored the fracturing behavior of heterogeneous unconsolidated sandstone, including scenarios with near-wellbore damage zones and vertically stratified formations containing shale barriers. The fracturing fluid used in the experiments was an organoboron crosslinked gel, with variations in polymer concentration and injection rate as key parameters. Throughout fracturing, pressure data were recorded in real time, and the post-test fracture geometries were characterized through detailed layer-by-layer dissection. [Results] Experimental results revealed the substantial leaking of the fracturing fluid at the initiation site and along the fracture path in unconsolidated sandstone, representing a typical characteristic and the primary control of fracture propagation in these reservoirs. Furthermore, the results showed that active leak management is important for transforming the fracture mechanism from a complex shear failure to one involving the development of more effective tensile fractures. Direct visual evidence from our tests showed that hydraulic fractures successfully traversed the near-wellbore damage zone in the unconsolidated sandstone. Moreover, in vertically heterogeneous formations with negligible stress contrast, fractures traversed low-permeability shale layers, leading to significant vertical growth. [Conclusions] In summary, this study developed an innovative experimental framework for modeling true triaxial hydraulic fracturing in unconsolidated sandstone. Its pivotal achievement was the integrated apparatus that enabled sample preparation and subsequent fracturing without stress-induced disturbances, providing reliable, representative results. Our findings advance understanding of fracturing mechanics in unconsolidated sandstone formations, highlighting the important role of fluid leak control and revealing previously overlooked fracture height containment issues in heterogeneous reservoirs. These results provide a validated physical basis for optimizing key design parameters, such as the fracturing fluid composition and pump rate, in unconsolidated sandstone reservoirs.
Acoustic emission evolution characteristics of tensile failure in red shale controlled by bedding and roadway stability control
ZHU Yuankun;MA Zhenqian;ZHANG Jimin;[Objective] A critical issue with red shale roadways has been identified in Guizhou phosphate mines. Over 70% of roadways pass through red shale, with a 34% roof collapse rate and 5 200 yuan/m support cost. The deep red shale surrounding rock is particularly susceptible to high ground stress and dynamic disturbances(excavation, blasting), which can worsen its instability. Existing studies confirm the bedding effect of red shale, however, its correlation with the angle-fracture mode remains ambiguous. Conventional methodologies prove ineffective in capturing real-time microfractures, and the application of the acoustic emission(AE) rise angle–average frequency(RA–AF) criterion remains underdeveloped. This study explores the tensile properties, fracture mechanisms, and AE characteristics of red shale under different bedding angles to optimize roadway support. [Methods] Red shale from Guizhou phosphate mines was processed into specimens measuring Φ50 mm × 30 mm and characterized by 0°, 30°, 45°, 60°, and 90° bedding angles. The specimens were dried at 115 ℃ for 24 h and subsequently sealed to avoid moisture. An MTS81 electro-hydraulic servo machine with a loading rate of 0.1 mm/min and a PAC PCI-2 AE system were utilized in the experimental setup. Two AE probes(AB-glued) were utilized to monitor signals, with key parameters including ringing count, energy, RA, and AF. The RA–AF criterion(RA = 3 μs·d B–1 threshold) was employed to distinguish between tensile and shear fractures, with stress–AE curve coupling analyzed. [Results] The tensile strength of red shale decreased significantly as the bedding angle increased, dropping from 6.20 MPa at 0° to 3.71 MPa at 30°, 3.43 MPa at 45°, 3.30 MPa at 60°, and finally to 3.17 MPa at 90°, representing a total reduction of 48.9%. The fracture mode of red shale also showed a clear transformation law. Specimens with 0° and 90° bedding angles(RA values of 1.2 μs·d B–1 and 2.1 μs·d B–1, respectively) exhibited primarily tensile failure. Specimens with 30°–60° bedding angles(RA values ranging from 18.5 μs·d B–1 to 32.7 μs·d B–1) exhibited an increasing proportion of shear failure, with 45° identified as the critical bedding angle at which shear failure accounted for 75%. The AE signals of red shale exhibited three stages that were highly coupled with the stress curve: stable accumulation(linear stress increase and weak AE activity), sudden release(peak stress and sharp surge of AE signals), and dissipation(low stress level and stable AE signals). Furthermore, specimens with 30°–60° bedding angles demonstrated nonlinear stress fluctuations before reaching the peak stress, accompanied by multi-peak variations in AE signals. Additionally, an RA threshold of >8 μs·d B–1 was proposed to predict shear slip 20 min in advance. Industrial tests showed that this scheme reduced the roof collapse rate of roadways to 9% and decreased the per-meter support cost by 1 200–1 800 yuan. [Conclusions] The bedding angle exerts a predominant influence on the anisotropy and fracture mode of red shale. The three-stage AE and RA–AF criterion effectively characterizes failure. A differentiated support scheme is proposed: roadways at 0° are reinforced with Φ22 mm high-strength anchor cables(≥150 kN); those at 30°–60° are anchored with epoxy grouting(≥5 MPa) + U-steel(45° with 0.8 m grouting spacing and AE monitoring); and those at 90° are reinforced with anchor cables + 120 mm C25 shotcrete. This approach is instrumental in ensuring the safety and efficiency of mining operations in Guizhou phosphate mines.