最新刊期
卷 57 ， 期 6 ， 2025
-
Advances in Monitoring of Bedload Transport in Mountain Rivers with High-frequency Acoustic Signals AI导读
“山区河流推移质输移监测技术取得新进展，专家构建多源声震监测系统，为水沙灾害评估提供科学依据。”摘要：Significance Quantitative prediction of bedload transport in mountain rivers is a key component in assessing water-related hazards and advancing hydraulic engineering applications. However, bedload transport rates exhibit substantial spatio-temporal variability under given flow conditions, leading to a limited understanding of bedload transport processes due to measurement difficulties, particularly in steep mountain streams. This study aims to emphasize research progress in monitoring technology for bedload transport in mountain rivers to advance quantitative measurement of bedload rates.ProgressDirect and indirect methods were generally two typical approaches for measuring bedload transport rate. The traditional direct methods measured the transported bedload mass using installed physical samplers and traps on the riverbed for a short time duration. These direct methods for measuring bedload were generally of low efficiency and limited accessibility due to the complex mechanisms of the bedload transport process, which made it challenging to obtain continuous and long-term bedload rates under large flow discharges and steep mountain streams. In contrast, the indirect methods, particularly the acoustic-based monitoring devices, including piezoelectric sensors, hydrophones, microphones, geophones, and seismometers, provided long-term continuous data on the bedload transport process. Bedload transport triggered high-frequency acoustic vibrations, which offered new perspectives for bedload measurement and can be translated into bedload transport rates based on model inversion. Considerable efforts were made toward developing various acoustic bedload monitoring systems, which were constructed and utilized to record high-frequency acoustic vibrations generated by inter-collisions among particles and by bedload particles impacting the riverbed, with the goal of improving the understanding of the bedload transport process. These acoustic-based indirect bedload monitoring technologies included the use of the node seismometer system, the hydrophone or underwater microphone system, the acoustic Doppler current profiler, the Swiss plate geophone (SPG) system, the mini-plate accelerometer (MPA) system, the Japanese pipe microphone (JPM) system, the phased microphone array (PMA) system, and the square pipe system (SPS). The acoustic signals recorded by these systems were utilized to analyze the acoustic sources and ultimately served to derive the bedload transport fluxes and grain size distributions based on calibration relationships that linked sediment-generated signals with bedload characteristics such as particle size and transport mass. However, significant discrepancies between predictions and observations were found, possibly due to variations in the spatial impact locations of bedload particles, impact angles, particle shapes, streamflow velocities, grain size distributions, and particle transport modes. The key issues in monitoring the bedload transport process using high-frequency acoustic signals were summarized in the following four aspects: 1) the construction of a multimodal acoustic monitoring system to fully record the high-frequency signals generated by transported bedload mass impacting the system plate or riverbed; 2) the characteristic analysis and processing of bedload-generated signals through noise reduction, convolutional transformation, and characterization of vibration signals to obtain signal packets, numbers of signal impulses, maximum amplitudes, and characteristic frequencies including centroid and peak frequencies, power spectral de-nsity, and other important parameters; 3) the development of algorithms for interpreting bedload-generated high-frequency signals, including localization of the acoustic source, establishment of the physical model of the acoustic source, and construction of the signal propagation and attenuation model; and 4) the inversion of physical parameters of bedload transport, utilizing the vibration signals with established quantitative calibration models to calculate key parameters such as particle size, transport velocity, and transport rate.Conclusions and ProspectsIn this study, the research progress on high-frequency signals utilized to monitor the process of bedload transport in steep mountain rivers is summarized based on the four aspects mentioned above, in combination with the latest findings of the authors. The current discussion addresses several issues confronting this research field and provides recommendations to overcome these challenges. Of particular concern are the inversion models and calibration relationships between the generated acoustic signals and bedload transport, whose adaptability to different river environments remains open for further investigation. In addition, persistent issues such as the limited diversity of monitoring methodologies and the challenges related to the comprehensive recording of signal characteristics continue to present obstacles. In response to these challenges, the authors recommend developing a robust and comprehensive acoustic monitoring system based on multiple sensors. In addition, the study emphasizes the necessity of conducting more detailed investigations into the source mechanisms of bedload-generated signals to enhance the understanding of the complex interactions between acoustic signals and bedload transport dynamics. In addition, the systematic establishment of databases that correlate various signal response indicators with bedload transport characteristics is essential. Finally, integrating insights from machine learning theory can provide improved predictive capabilities, enhancing the accuracy of inferring bedload transport rates from generated signals. Accordingly, this study provides a scientific reference for advancing the monitoring of bedload transport in steep mountain rivers, quantitatively assessing hydrological hazards, and guiding the construction and long-term operation of hydraulic engineering str-uctures.关键词：mountain river;bedload;acoustic vibration signals;monitoring system;source mechanism;bedload transport rate520|467|0更新时间：2026-01-23 - “最新研究揭示了溃坝洪水波形和波高受上下游水深及坡度影响的规律，为溃坝洪水灾害学提供理论依据。”摘要：ObjectiveThe propagation of dam-break flow is significantly influenced by both upstream and downstream initial water depths, while the effect of slopes on flow propagation should not be overlooked. It is recognized that numerical simulations still have limitations in accurately modeling large-scale and large-deformation water surface phenomena. In addition, on-site observations of dam-break flow remain limited due to spatial and temporal constraints. Dam-break experiments allow visualization of the evolution of dam-break waves, wall impact pressure, and other related dynamics. Therefore, experimental studies of dam-break flows are of great importance compared to numerical simulations. However, few existing experimental studies simultaneously investigate the influence of water depths and slopes on the spatiotemporal evolution of dam-break wave regimes and heights.MethodsAccordingly, an experimental investigation was conducted to examine the evolution characteristics of wave regimes and wave heights resulting from a wet-bed dam break. A large-scale flume measuring 52 m in length, 1.2 m in width, and 1 m in height was utilized, which was divided into two sections by a 10 mm thick fiberglass board positioned 12 m from the inlet. The upstream water was released by lifting the fiberglass board connected to an electric motor and controlled by a computer program to minimize experimental errors. The flume bottom and sides were made of toughened glass, which facilitated the observation of the experimental process while maintaining the accuracy of the detected water surface. The slope of the flume bed was adjusted using 13 pairs of hydraulically operated lifting columns that were automatically controlled with high precision. Various combinations of upstream water depth hp (hp = 0.5, 0.6, 0.7, and 0.8 m), downstream water depth hd (hd = 0.1, 0.2, 0.3, and 0.4 m), and slopes i (i = 0, 2‰, 4‰, and 6‰) were considered to evaluate their effects on the dam-break waves, resulting in a total of 64 scenarios. The spatiotemporal evolution of the dam-break waves was measured using ten wave gauges positioned at different downstream locations.Results and DiscussionsThe experimental results showed that the regimes of the dam-break waves were strongly influenced by both the upstream and downstream water depths as well as by the slopes. The regimes of the dam-break waves were distinguished using the global Froude number (Fx). Specifically, when Fx was less than 1.2, the wave propagation resembled that of an undular wave, whereas when Fx was greater than 1.2, it exhibited characteristics similar to a bore wave. In addition, the rise pattern of the dam-break flood level was significantly affected by the global Froude number Fx. Based on the variation of the experimental data, a flood-level rise model for the increase of the dam-break flood level ΔH with time T for different values of Fx was proposed. The root mean square error between this flood-level rise model and the results of existing studies was 1.7%, indicating that the proposed model accurately predicted the water-level rise for the dam-break flood. In addition, the relationship between the dimensionless water-level rise velocity (C) and the global Froude number Fx was investigated for two different waveforms. The value of C increased as Fx increased when Fx was less than 1.2 and remained constant as Fx increased when Fx was greater than 1.2. This study also examined the applicability of the Stoker solution in predicting wave heights by comparing the experimental data with the Stoker theoretical results. The findings demonstrated that when the dam-break wave propagated as a bore wave, the maximum relative error between the Stoker theoretical solution and the experimental data was 5.97%, indicating that the Stoker solution exhibited minimal error in determining the wave height under the bore wave condition. However, when the dam-break wave propagated as an undular wave, the maximum relative error between the theoretical Stoker solution and the experimental data reached 23.63%, indicating that the Stoker solution has a larger error in estimating the wave height under the undular wave condition. Formulae for calculating the maximum wave height hmax and Fx under different upstream and downstream water depth ratios α (α=hd/hp) were established to enable more accurate prediction of the variation in maximum wave height, considering the influence of slopes and different upstream and downstream water depths. The calculation method presented above was compared to existing experimental data and showed satisfactory agreement.ConclusionsThe varying water depths upstream and downstream contribute to the propagation of dam-break waves in two distinct forms, distinguishable by the proposed global Froude number. The phenomenon is characterized as follows: when Fx is less than 1.2, the wave manifests as an undular wave, whereas when Fx exceeds 1.2, it transforms into a bore wave. In the case of undular waves, the rate of water level rise increases proportionally with Fx, while in bore waves, this rate remains relatively constant. Based on the influence of Fx on the dam-break wave surface, a functional relationship was established between the rise in flood levels after a dam break and Fx. It was observed that when the dam-break wave propagates as an undular wave, the initial wave height represents the peak height. The experimental data also indicate that the ratio (hmax‒h1)/hp decreases linearly with increasing Fx values (h1 is the wave height according to the Stoker theory solution). In addition, a formula for calculating the maximum wave height was derived by integrating this finding with the Stoker solution. The results contribute significantly to a comprehensive understanding of the formation and evolution of dam-break flood waves, providing a theoretical foundation for establishing dam-break flood hazard science.关键词：wet-bed;global froude number;wave regime;flood level rise;maximum wave height310|497|0更新时间：2026-01-23
- “最新研究发现，小浪底水库异重流排沙出库受河床纵比降和水沙动力影响，建立了异重流出库判别图，为多沙河流水库排沙提供依据。”摘要：ObjectiveThe Xiaolangdi Reservoir, as a representative sediment-loaded river reservoir, shows that turbidity current venting plays a crucial role in mitigating reservoir sedimentation during initial sediment retention operations. Understanding the mechanical conditions under which the turbidity current reaches the dam for effective sediment discharge or venting presents a critical operational challenge in managing sediment-loaded river reservoirs. The theoretical identification of turbidity current venting provides direct technical support for water and sediment regulation strategies, which form the objective of this study.MethodsTurbidity current venting and the turbidity current reaching the dam were considered equivalent situations. The turbidity current events were selected based on the conditions that, for the inflow, the daily sediment concentration exceeded 20 kg/m³ and the daily sediment transport rate exceeded 10 000 kg/s, which were used as the thresholds for turbidity current event occurrence. For practical significance, a daily sediment concentration greater than 1 kg/m³ for the outflow was employed as the standard to determine whether sediment discharge occurred. The concept of river work was introduced, as the movement of the turbidity current needs to overcome resistance to perform work, referred to as turbidity work. It was assumed that the energy dissipation of turbidity current movement from plunging to cessation was equal to turbidity work, and the turbidity work when it reached the dam was defined as critical work. Therefore, the theoretical discrimination of turbidity current venting was that the turbidity current work needed to exceed the critical work before reaching the dam. The turbidity current movement was controlled by its sediment content and the longitudinal gradient of the riverbed, by neglecting the water and sediment exchange and mixing processes. The ratio J0/Jc of the average longitudinal gradient J0 of the riverbed from the plunge point to the dam and the critical gradient Jc of the turbidity current is utilized to represent the relative longitudinal gradient of the riverbed, indicating the topographic conditions of turbidity current movement. The ratio uT/L, where u is the turbidity current equilibrium velocity, T is the duration, and L is the radial distance from the plunge point to the dam, is utilized to represent the relative distance of motion, indicating the dynamic factor of the turbidity current. The theoretical discrimination is expressed as J0/Jc · uT/L > K, where K is a parameter related to a specific reservoir. Therefore, the discriminant diagram of turbidity current venting, composed of the relative longitudinal gradient of the riverbed (J0/Jc) and its hydrodynamic factor (uT/L), was determined.Results and DiscussionsA total of 92 turbidity current events were identified, of which 55 resulted in sediment discharge. Statistically, flood events with higher sediment concentrations and larger water volumes produced turbidity currents with more significant sediment discharge. There was no clear correlation between the event-averaged flow discharge and sediment concentration, and these parameters cannot be directly employed to distinguish between turbidity currents that reached the dam and those that did not, i.e., for discrimination purposes. A clear positive correlation existed between the event water and sediment amounts; however, these two parameters failed to serve as effective discriminants. The sediment amount of turbidity current events that did not reach the dam was primarily concentrated below 80 million tons, whereas that of the events that reached the dam was mainly distributed in the range of 0.2～320.0 million tons, exhibiting a bell-shaped distribution. The water volume of turbidity current events that did not reach the dam was mainly concentrated in the range of 0.1～1.6 billion m3, while that of the reached events was mainly concentrated in the range of 0.4～1.6 billion m3, both showing bell-shaped distributions. The larger the water and sediment amounts in a turbidity current event, the higher the probability of its reaching the dam. The entire coordinate area of the discriminant diagram was divided into two regions, namely the dynamic region at the top left and the inertial region at the bottom right. As the parameters J0/Jc and uT/L increased, the density current was more likely to advance toward the dam and occupy the dynamic region. In the dynamic zone, the turbidity current quickly reached the dam by its potential energy. In the inertia zone, a significant deceleration of the turbidity current occurred due to insufficient potential energy. For most turbidity current events in the inertia zone, the head of the current was still able to reach the dam due to the energy supplied by the trunk of the turbidity current. The flood events in which the turbidity current did not reach the dam were mainly located in the lower left corner of this coordinate system. The critical condition for the sediment discharge of turbidity currents was represented by a hyperbolic curve with K=12 for the Xiaolangdi Reservoir. This discrimination chart effectively distinguished the outflow of density currents. Thirty-three of the 37 non-vented events were located at the lower left of the curve, and 45 of the 55 vented events were located at the upper right of the curve. The discrimination diagram effectively distinguished sediment discharge by turbidity currents.ConclusionMost of the turbidity current events since the establishment of the Xiaolangdi Reservoir were successfully identified using the discriminant diagram. The proposed theoretical sediment discharge conditions effectively predict the sediment discharge of multiple floods that have already occurred. The discriminant diagram also provides a dependable basis for achieving efficient sediment discharge of turbidity flow in the reservoir of the sediment-laden river.关键词：turbidity current;Xiaolangdi Reservoir;sediment discharge;discrimination407|289|0更新时间：2026-01-23
- “在气候变化影响下，青藏高原冰碛湖溃决风险增加，专家建立了冰碛湖稳定性评价模型，为评估冰湖溃决洪水风险提供技术支撑。”摘要：ObjectiveMoraine-dammed glacial lake outburst floods (GLOFs) occur with increasing frequency on the Qinghai‒Tibet Plateau in the context of global warming, posing serious threats to downstream settlements and hydraulic infrastructure. Accurately determining the critical conditions that trigger GLOFs and quantitatively evaluating the stability of moraine-dammed lakes remains an urgent scientific challenge.MethodsThis study developed a stability assessment model grounded in the principles of physics and soil mechanics, comprehensively incorporating key processes such as lake expansion, landslide-induced surge generation and propagation, overtopping erosion, and piping erosion. The framework enabled the analysis of complex, multi-scale, and multi-process interactions governing lake stability. Within this model, the temporal variation in lake water storage was simulated using a water balance equation that accounted for precipitation, runoff, glacier meltwater, snowmelt, and infiltration. Then, based on the principles of energy conservation and gravitational similarity, the model quantified the surge characteristics generated by rigid and granular landslides entering the lake and simulated the attenuation of surge amplitude during propagation using empirical relationships derived from three-dimensional flume experiments. Finally, the critical hydraulic heads associated with overtopping and piping failures were derived. The critical overtopping condition was determined by calculating the threshold shear stress required to initiate the motion of coarse particles on the downstream dam slope. At the same time, the Kenney‒Lau (K‒L) method was applied to evaluate the internal stability of the moraine’s granular structure and determine the critical hydraulic gradient and corresponding head for piping failure. A lake stability coefficient was defined as the ratio of the actual hydraulic head at the dam front to the critical failure head by integrating these components, enabling a quantitative assessment of lake stability and precise identification of the dominant failure mechanism.Results and DiscussionThe model was validated through its application to Jialong Co, a moraine-dammed lake in Tibet that experienced a historical outburst in 2002. The results showed that between 1988 and 2018, the stability coefficients for overtopping (
- “在特殊工程问题下，面板堆石坝安全监测技术取得新进展，为高坝高、高寒、高地震烈度、深厚覆盖层及狭窄河谷等复杂条件下的面板堆石坝筑坝水平提供理论支持和实践指导。”摘要：SignificanceThe construction of concrete face rockfill dams (CFRDs) in China presents both remarkable development opportunities and considerable technical challenges. The complex topographic and geological conditions, harsh natural environments at dam sites, and the transition from 200 m to 300 m in dam height impose greater demands on the advancement of safety monitoring technologies for CFRDs. Typical engineering challenges include high dam heights, alpine environments, strong seismic activity, deep overburden foundations, and narrow valleys. These demanding conditions inevitably result in issues such as significant deformations, damage to the face slab and joint waterstops, and subsequent dam leakage, all of which threaten the operational safety of the dams. Therefore, beyond precise design, standardized construction, and comprehensive management, systematic safety monitoring of CFRDs is essential based on their physical and mechanical characteristics to fully meet safety monitoring requirements under these specialized engineering conditions. This review aims to provide systematic theoretical support and practical guidance for the advancement of safety monitoring technologies for CFRDs, ensuring the long-term operational safety of these dams under the “three highs, one deep, and one narrow” engineering challenges.ProgressThis study focused on specific engineering challenges such as high dam height, alpine regions, high seismic intensity, deep overburden, and narrow valleys. It systematically reviewed the structural issues of CFRDs under the “three highs, one deep, and one narrow” conditions, emphasizing that the core of CFRD technology lies in deformation control and comprehensive deformation coordination. Research indicated that an increased dam height exacerbated CFRD deformation, which led to panel cracking and weakened the dam’s anti-seepage system. CFRDs in high seismic intensity zones exhibited significant permanent deformation, making them prone to local cracks and internal damage. CFRDs constructed in extremely low-temperature conditions demonstrated poor durability and a higher risk of waterstop structure failure. CFRDs built on deep overburden layers were susceptible to uneven settlement, long-term deformation, and panel cracking. CFRDs located in narrow valleys experienced large deformation gradients, which caused damage to waterstop structures and panels. The study also reviewed the development of safety monitoring technologies for CFRDs, summarizing key monitoring techniques under special engineering conditions. It identified critical monitoring areas, main monitoring items, and common monitoring methods, highlighting that deformation and seepage remained the primary monitoring focuses for CFRDs under “three highs, one deep, and one narrow” conditions. Specifically, high dams required attention to deformation and seepage safety; CFRDs in high seismic intensity zones needed enhanced strong-motion monitoring; CFRDs in alpine regions needed a focus on ice pressure, dam body, and panel deformation; CFRDs built on deep overburden layers required monitoring of foundation settlement, cutoff wall deformation, and anti-seepage effectiveness; CFRDs in narrow valleys demanded emphasis on surface and internal deformation, panel deformation, and rockfill stress. The study introduced self-developed monitoring equipment, such as a dual-prism device based on GNSS monitoring, to align monitoring technology with CFRD construction standards under special conditions. Based on dam safety monitoring work, it summarized improvements in construction methods and control indicators, analyzing engineering measures to enhance CFRD serviceability under “three highs, one deep, and one narrow” conditions. The control methods for comprehensive deformation coordination in CFRDs were also explored. Specifically, for high CFRDs, attention was paid to the overall three-dimensional spatial deformation coordination, the independent multi-layer sealing and self-healing capabilities of the waterstop structures, the repairability of the dam body, the addition of permanent horizontal joints to enhance deformation adaptability, and the strict control of hard rock dam construction and particle gradation to improve dam strength. For CFRDs in high seismic intensity zones, comprehensive seismic measures were recommended, such as reasonable dam crest elevation, optimized drainage systems, reinforced foundations, expanded cushion zones, high-quality construction materials, gentler slopes or reinforced structures, adaptive waterstop materials, increased panel thickness and reinforcement, and enhanced strong-motion monitoring. In alpine regions, thin-layer heavy vibratory compaction, improved panel frost resistance, enhanced thermal exchange, increased reinforcement, and better cushion zone drainage were indicated. For CFRDs built on deep overburden layers, full-depth cutoff walls, connecting slabs for improved plinth conditions, and stress-deformation analysis of cutoff walls were essential. In narrow valleys, medium-hard rockfill materials, reasonable pre-settlement periods, flexible fill materials, special compaction zones, high toe walls, or internal plinth designs were recommended to improve stress-deformation states and ensure stability. The core of CFRD technology lies in ensuring deformation coordination between the dam body and panels, which was achieved through numerical analysis, optimized dam zoning, higher downstream rockfill compaction standards, and balanced full-section filling to prevent panel separation and cracking, enhancing overall dam stability.Conclusions and ProspectsResearch indicates that as dam height increases, combined with complex topographic and geological conditions and harsh natural environments at dam sites, the physical and mechanical properties of CFRDs become increasingly complex. An in-depth investigation of the structural characteristics of CFRDs under the “three highs, one deep, and one narrow” conditions is essential not only for advancing dam construction and design feedback but also for enhancing dam safety monitoring. In this regard, monitoring projects and methods for CFRDs under high dam height, high seismic intensity, alpine regions, deep overburden, and narrow valleys should emphasize key indicators such as dam deformation, strong-motion monitoring, ice pressure, foundation settlement, and panel deformation. Developing new precision monitoring instruments with large ranges and high water pressure resistance is necessary to overcome the inherent limitations of conventional monitoring technologies. These instruments are vital for accurately detecting abnormal conditions under extreme climatic and harsh environmental circumstances, ensuring that the safety monitoring requirements of CFRDs under special engineering challenges are met. In the future, predictive technology for CFRD safety performance requires significant enhancement. Large-scale and high-stress experiments on dam construction materials should be prioritized to establish constitutive models that comprehensively incorporate multiple influencing factors. In addition, attention should be devoted to special engineering challenges such as high altitude, steep slopes, and karst landforms, which affect dam design and monitoring technologies. The integration of advanced technologies such as the Internet of Things (IoT) and artificial intelligence (AI) is also essential to raise the development of intelligent and precise safety monitoring systems, enabling real-time monitoring and early warning of dam operational conditions.关键词：concrete face rockfill dam;safety monitoring;high dam height;alpine region;high seismic intensity;deep overburden layer;narrow valley163|279|0更新时间：2026-01-23
- “在泥沙运动领域，专家建立了泥沙颗粒起动概率公式，为山洪治理、水力发电等提供解决方案。”摘要：ObjectiveThe incipient motion of sediment constitutes an integral component in the study of sediment transportation. It plays a significant role in the design of stable channels, sediment transport analysis, and hazard assessment related to channel degradation. Due to the stochastic nature of sediment movement in natural rivers, defining the incipient motion criteria precisely is challenging. Therefore, accurately assessing the pickup probability of natural sediments remains essential. The objective of the present study is to develop a theoretical pickup probability formula for uniform sediment by employing a force balance mechanism in conjunction with a stochastic method.MethodsThe forces acting on a sediment particle that are considered to influence the incipient motion included the drag force, lift force, submerged weight, and resistance force. Based on the rolling motion, a critical relation for the incipient motion is derived. Under the condition of incipient motion, the tangential and normal forces are related to the angle of repose of sediment particles. The incipient motion of a sediment particle occurs when the tangential drag force exceeds the tangential resistance forces. Hence, a theoretical pickup probability formula was developed based on the force balance model. Experimental results showed that the time-averaged velocity increases with increasing flow depth. It reached the maximum value at the water surface for wide open-channel flow. The time-averaged velocity was distributed based on a logarithmic law. The near-bed time-averaged velocity in the present pickup probability equation was derived by assuming the drag force acted at a position of 0.35 d distance from the theoretical bed level. For uniform flow, previous results showed that the turbulence intensity followed the exponential law. Also, the turbulence intensity in the water flow direction decreased with increasing water depth. With reference to the experimental data for a rough open-channel flow bed conducted by Kironoto, the parameter of near-bed turbulence intensity in the present pickup probability equation was obtained. The angle of repose is a fundamental property of sediment particles. Under static conditions, the angle of repose is the steepest incline beyond which the slope of the sediment particles collapses. This property is important in sediment transportation. Previous results showed that it increases with sediment size d. With reference to the experimental data conducted by Yang, the present study derived a theoretical formula to calculate the angle of repose in a water medium. Unlike previous studies, the angle of repose equation developed in this study is a function of dimensionless sediment size d*. The parameter of the tan value of the angle of repose was determined based on the angle of repose equation developed in this study. The drag and lift force coefficients were usually considered as unknown functions of the particle Reynolds number. Different values of both coefficients were reported in the literature. This study combined the two coefficients with an unknown function. The unknown function in the pickup probability equation was determined by comparison to the measured data.Results and DiscussionsThere are five parameters: the angle of repose of sediment, near-bed average velocity, near-bed turbulence intensity, drag force coefficient, and lift force coefficient. Once they are determined, the present pickup probability equation can be used in engineering applications. This study obtained all the parameters with reference to the experimental data in rough open-channel flow. The measured pickup probability data were utilized to test the proposed model. Also, the present pickup probability model was compared to the Einstein formula, Engelund and Fredsoe formula, and Cheng and Chiew formula. The pickup probability was computed using the present study formula, Einstein formula, Engelund and Fredsoe formula, and Cheng and Chiew formula with the measured input data to test the accuracy of the pickup probability equation developed in the present study. The pickup probability predictions by the four equations were listed in an error table to explain the results. The pickup probability formula developed in this study has an average relative error of 22.8% for Luque data, 66.9% for Guy ripple data, and 28.2% for Guy dune data. The results showed that the equation developed in this study provides the best representation of experimental measurements. Einstein's formula has an average relative error of 772.9% for Luque data, 2 666.4% for Guy ripple data, and 418.4% for Guy dune data. Einstein's pickup probability formula produces the highest errors, probably because his formula was not verified with experimental data in his original papers. For a given pickup probability, the pickup probability equation can be utilized to calculate the critical shear velocity of uniform sediments. Under threshold conditions, the present pickup probability P = 0.000 045 98, Einstein pickup probability P = 0.02, Engelund and Fredsoe formula P = 0.018 7, and Cheng and Chiew formula P = 0.013. The threshold value P of this study fell into the "weak movement" criterion, whereas the other three equations fell into the "middle movement" criterion. In addition, the present pickup probability model fitted the classic Shields diagram quite well under the incipient motion condition.ConclusionsThe integral expression of sediment particle pickup probability is derived. This expression fully reveals the force mechanism acting on sediment particles and compensates for the limitation of previous pickup probability formulas that ignored the drag force of water. The pickup probability formula developed in this study demonstrates the highest calculation accuracy. In addition, for sediment particles in the critical state, the pickup probability derived in this study tends to align with the weak moving pickup criteria, whereas other formulas correspond to the moderate moving pickup criteria. Finally, when compared to the classical Shields curve, the pickup probability formula proposed in this study shows good agreement, while other formulas yield a straight line that can only determine the incipient shear stress of large particles.关键词：uniform sediment;average flow velocity;flow lift force;flow drag force;incipient motion of sediment241|372|0更新时间：2026-01-23
WATER DISASTER MITIGATION AND WATER ENVIRONMENT REGULATION
- “在建筑隔震技术领域，专家通过试验研究了铅芯橡胶隔震支座在大变形加载下的滞回性能，为高烈度地区减轻地震灾害提供解决方案。”摘要：ObjectiveLead rubber bearings (LRBs) tend to exhibit significant stiffening and degradation behavior when subjected to large shear strain deformation. However, these two phenomena, attributed to the Scragging and Mullins effects, are not thoroughly examined or adequately integrated into seismic isolation design procedures. The objective of this study is to comprehensively investigate the large-strain properties of LRBs. In addition, a numerical comparison is presented to demonstrate variations in seismic response when different modeling techniques are applied during strong earthquake shaking. The findings of this study significantly contribute to developing a more precise design methodology for seismically isolated structures, particularly in regions with high seismic intensity.MethodsReversed cyclic tests were conducted on two different types of bearings. The dynamic compression-shear machine was employed for these tests. The reversed cycles were executed in a pseudo-static manner with an average velocity of 2 mm/s. A constant compression of 12 MPa was maintained throughout the test. The testing protocol consisted of four distinct loading sequences that covered both low-to-moderate and large strain loading amplitudes. The test sequences include equal amplitude, increasing amplitude, unsymmetric, and near-field loading scenarios, respectively. The maximum shear force in the first cycle was significantly greater than that of the subsequent cycles, primarily due to the Scragging effect. From the second to the eleventh cycles, the force gradually decreases, attributed to the temperature increase in the lead core. The maximum shear force decreases by 14.6% and 22.1% in total for LRB500 and LRB800, respectively. The variation in the equivalent damping ratio follows a similar trend. In addition, the results demonstrated that the traditional bilinear model adequately captures the hysteretic response of LRBs under design-level shear strain with satisfactory accuracy.Results and DiscussionsThe experimental results for large shear strain deformation were compared. A significant stiffening effect occurred when the shear strain exceeded 200%. The secant stiffness after the stiffening stage was found to be 1.6 times and 1.4 times greater than the design-level stiffness for LRB500 and LRB800, respectively. The loading history influenced the stiffening behavior, where a 20.2% additional increase in shear force occurred under the same shear amplitude level. In addition, the experiment revealed the occurrence of degradation, primarily during the initial large strain loading cycle. The traditional model failed to accurately capture the hysteretic response of LRBs under large shear strain deformation. A comprehensive evaluation of the seismic performance of an actual base-isolated building was conducted using the benchmark model of an isolated RC frame structure located in Luding, Sichuan Province. This building experienced significant shear failure in most of its isolators during the 2022 Luding earthquake. Regarding modeling techniques, a simplified multi-degree-of-freedom shear-type model was adopted to represent the isolated superstructure, while the isolation layer was modeled in detail. Both the traditional Bouc-Wen (BW) model and the Generalized Bouc-Wen (GBW) model were utilized to represent the horizontal hysteretic behavior of LRBs to investigate the seismic response of the building, considering the large strain nonlinearities of LRBs. The GBW model was specifically developed to account for the large-strain behavior of LRBs, building upon the basic framework of the BW model. It is demonstrated that the hysteretic curves generated by the GBW model closely matched the test results. The analysis employed field-recorded seismic waves from both far-field and near-field sources. The time history and response spectrum were given. The peak input acceleration was 527 cm/s2, and a significant velocity pulse was observed in the near-field record. This study examined the differences in seismic response when using the BW and GBW models. A comparison was made for the shear deformation and hysteretic curves of LRBs under far-field earthquakes. The results indicated that LRBs experienced moderate shear strain under far-field earthquakes; thus, the large strain nonlinearities were not significant. Therefore, the isolation displacement and hysteretic curves obtained using both models were similar. However, under near-field earthquakes, differences were observed. The maximum shear deformations calculated using the BW and GBW models were 428% and 363%, respectively. The significant hardening behavior observed during the first large strain cycle effectively limited the maximum isolation displacement of the LRB-isolated building; however, it raised concerns regarding the potential amplification of peak shear forces. The result demonstrated that incorporating the large strain behavior of LRBs in the modeling led to increased story drift in the isolated superstructure.ConclusionsThis study provides comprehensive insights into the large-strain nonlinearities of LRBs through experimental investigations. In addition, numerical simulations were performed to compare the dynamic seismic response of an isolated building while considering these large-strain nonlinearities. The results indicate that LRBs exhibit significant hardening and degradation behavior under considerable shear strain, which significantly influences the seismic response of both the isolation layer and the superstructure. These findings establish a foundation for developing a more accurate seismic isolation design methodology. Future research should further investigate the mechanical properties of seismic isolators under ultimate seismic conditions to enhance the seismic resilience of structures located in near-fault regions.关键词：seismically isolated structure;lead rubber bearing;large strain loading;bearing hardening;strength degradation408|458|0更新时间：2026-01-23
- “在地震工程领域，专家分析了隔震结构在长周期地震动作用下的动力响应，并提出了调谐减震设计方法，为隔震结构减震控制提供解决方案。”摘要：ObjectiveLong-period ground motions have a serious impact on long-period structures and may even cause damage. Currently, seismic isolation technology has demonstrated excellent technical advantages. However, it remains essential to evaluate whether isolated structures with long-period characteristics will be significantly affected by long-period ground motions. Furthermore, it is necessary to verify whether incorporating a tuned mass damper (TMD) can improve the seismic performance of isolated structures subjected to such unique ground motion characteristics.MethodsTo ensure the reliability of the analysis results, 50 near-fault pulse-type ground motions and 50 far-fault long-period ground motions were selected as input seismic loads based on relevant screening criteria. For the near-fault pulse-type ground motions, NGA-West2 is used as the initial screening database. The correlations between pulse period and response spectrum, as well as between PGV/PGA and response spectrum, are analyzed. Finally, a threshold of PGV/PGA > 0.2 is used as an additional screening criterion. For the far-fault long-period ground motions, the weighted square average of the periodicity of the ground motion amplification coefficient spectrum between 2 s and 6 s is considered as the boundary parameter of ground motions, with a threshold value greater than 0.4 defining long-period ground motions. The acceleration response spectra of the selected records are significantly higher than the standard design response spectrum, exhibiting obvious long-period characteristics. Based on the dynamic characteristics of isolated structures, 19 simplified single-degree-of-freedom isolated structure models are established, each with a yield displacement of 0.75 mm, post-yield stiffness ratio of 1∶13, and yield-weight ratios ranging from 0.01 to 0.10. The isolation bearing adopts a bilinear model. To efficiently process large amounts of data, a MATLAB program based on the Newmark-β method was self-developed, and its accuracy was verified through comparison with results obtained from commercial software. The amplitudes of the selected ground motion acceleration are adjusted to five intensity levels of 200, 300, 400, 510, and 600 cm/s2. Under the above calculation conditions, the maximum displacement response of the isolation layer and the maximum acceleration response of the superstructure were obtained through elastic-plastic time-history analyses. Subsequently, the correlations among the dynamic responses of the isolated structure, post-yield stiffness ratio, ground motion type, and ground motion intensity were analyzed. The tangent stiffness corresponding to the maximum displacement was used to define the equivalent stiffness, and the equivalent damping ratio was determined according to the principle of energy consumption. The parameters of the TMD were optimized by controlling the mean square value of the structural displacement response. Finally, the damping rate of the TMD on the peak displacement of the isolation layer and the acceleration of the superstructure was analyzed. The study considered both single-degree-of-freedom and multi-degree-of-freedom isolated structures.Results and DiscussionsThe results indicate that under long-period ground motions, the isolation layer exhibits large displacement responses. To effectively control the maximum displacement of the isolation layer, the yield-weight ratio of the isolation layer should be more than 0.06. However, considering the safety of non-structural components in the superstructure, the yield-weight ratio should not exceed 0.07 to avoid excessive acceleration responses in the superstructure. Under far-fault long-period ground motions, the control effect of the TMD on the displacement of the isolation layer and the acceleration response of the superstructure increases with an increase in the input ground motion intensity. The TMD is effective in controlling the dynamic response of the isolated structure subjected to rare and extremely rare ground motions. In contrast, under near-fault pulse-type ground motions, the hysteretic delay in the tuning response of the TMD causes a poor damping effect on the displacement of the isolation layer. As the yield-weight ratio increases, the damping effect worsens and may even become negative. Nevertheless, the acceleration response of the superstructure has a relatively obvious damping effect.ConclusionUnder different intensity levels of ground motion, the difference in the degree of nonlinearity of the isolation layer can cause detuning of the TMD, which might lead to significant differences in the damping effect. Therefore, the input level of the ground motion must be carefully considered in the design of TMDs.关键词：long-period ground motions;seismic isolated structures;tuned mass damper;near-fault pulse-like;far-fault long-period271|490|0更新时间：2026-01-23
- “在抗震设计领域，专家提出了新型塑性铰区扩大截面预应力混凝土摇摆自复位桥墩，具有良好自复位能力和可修复性，抗震性能显著提升。”摘要：ObjectiveA functionally recoverable rocking self-centering prestressed concrete (RSPC) pier with an expanded section in the plastic hinge zone is proposed to comply with the current requirements of the resilient seismic design concept and to enhance the seismic performance and rapid post-earthquake recovery of prefabricated assembled pier columns.MethodsBased on the OpenSees software, the numerical calculation models of the RSPC piers were established. After verifying the reliability of the models, the seismic performance of the RSPC piers was studied, and the effects of parameters were analyzed, including the reinforcement ratio of energy-consuming rebars and the initial prestress values for prestressed tendons of the upper and bottom sections. The evaluation method for the resistance resilience of the rocking self-centering piers was proposed from the energy perspective, and the RSPC piers were analyzed under different design parameters.Results and DiscussionsTwo-section prefabricated rocking self-centering piers, namely the prefabricated pier with unequal height (RSPC1) and the prefabricated pier with equal height (RSPC2), were designed. The hysteretic curves of the RSPC1 and RSPC2 piers exhibited an obvious flag shape, the pinching effect was significant, and both showed superior self-centering ability. The hysteresis curves of the energy-consuming steel bars of the RSPC piers were analyzed. When the energy-consuming steel bars entered the plastic stage, the hysteresis curves of the RSPC2 piers' energy-consuming steel bars were fuller than those of the RSPC1 piers. Before loading reached the ultimate displacement, the cumulative energy consumption of the monolithic reinforced concrete (RC) piers was always greater than that of the RSPC piers. However, the maximum cumulative energy consumption of the RSPC1 and RSPC2 piers was 57.1% and 76.6% higher than that of the RC piers, respectively. The energy dissipation capacity of the RSPC2 piers was greater than that of the RSPC1 piers, and the maximum cumulative energy dissipation was 12.4% higher than that of the RSPC1 piers. The equivalent viscous damping coefficient (ζeq) of the RC piers was significantly greater than that of the RSPC piers. After the piers entered the plastic stage, ζeq of the RSPC2 piers was greater than that of the RSPC1 piers, and the hysteretic curve fullness of the RSPC2 piers was higher than that of the RSPC1 piers. Compared to the RC piers, the peak bearing capacity of the RSPC piers increased by 83.7% and 94.9%, while the ductility coefficient increased by 26.5% and 20.6%, respectively. The peak bearing capacity of the RSPC2 piers was 6% higher than that of the RSPC1 piers, and the displacement ductility coefficient was 4% lower. The residual displacement of the RSPC piers was significantly smaller than that of the RC piers, with the residual displacement deviation ratio being less than 1%. The maximum residual displacement of the RSPC1 and RSPC2 piers was only about 1/12 and 1/9 of that of the RC piers, respectively. Taking the RSPC2 piers as the benchmark model, the reinforcement ratio of the energy-consuming rebars and the effective prestress values for prestressed tendons of the upper and bottom sections were selected as the main influencing parameters for the parameter analysis. The reinforcement ratio (ρED) of the energy-consuming steel bars was 0.34%, 0.47%, 0.57%, 0.61%, and 0.69%. With the increase of ρED, the bearing capacity, cumulative energy consumption, and residual displacement deviation ratio of the RSPC2 piers gradually increased. When ρED = 0.69%, the residual deviation ratio of the pier exceeded 1%. When ρED was less than 0.61%, the residual offset ratio was less than or equal to 1%, and the contribution rate of the lateral strength of the energy-consuming steel bars was less than 35%, which ensured that the piers maintained better self-centering ability and energy dissipation capacity. The effective prestress of the upper prestressed tendons (σST) was 162, 323, 484, and 581 MPa. When σST = 581 MPa, the loading displacement reached 36 mm, the bearing capacity of the pier decreased significantly, and the loading was stopped. Under other working conditions, as σST increased, the bearing capacity of the RSPC2 piers decreased, while the cumulative energy consumption and residual displacement increased. The maximum joint openings of the upper section were 1.17, 0.68, and 0.36 mm, and their ratios to the joint openings of the bottom section were 16.7%, 9.7%, and 4.9%, respectively. As σST increases, the joint opening of the upper section of the precast segment decreases significantly. When σST = 484 MPa, the ratio of the upper joint opening to the bottom joint opening was the smallest, and the prestressed tendons improved the integrity of the upper section, enhanced the rocking performance, and ensured that the RSPC2 piers maintained better self-centering ability and energy dissipation capacity. The effective prestress of the bottom prestressed tendons (σSB) was 215, 430, 646, and 861 MPa. As σSB increases, the bearing capacity of the RSPC2 piers rises, while the cumulative energy consumption and residual displacement decrease. The stiffness after yielding increases, and the displacement ductility coefficient first increases and then decreases. The displacement ductility of the RSPC2 piers reached the optimal value when σSB = 646 MPa. The maximum difference in the ratio of the upper joint opening to the bottom joint opening was only 3%. The seismic resilience of the RSPC2 piers was analyzed from an energy perspective. The initial resistance resilience ratio (Rs) was greater than 1.00, and the residual resistance resilience ratios (R1, R2, Ru) were greater than those of the reference piers. The Rs, R1, R2, and Ru of the RSPC2 piers all increased with the increase of ρED. When ρED = 0.53% and 0.61%, the comprehensive resistance resilience reached its optimal value. As σST increases, Rs decreases but remains greater than 2.0, while R1, R2, and Ru gradually increase. When σST = 484 MPa, the residual resistance resilience of the RSPC2 piers was superior. As σSB increases, Rs gradually improves, while R1, R2, and Ru decrease, but the changes remain within 0.004. When σSB = 646 MPa, the comprehensive resistance resilience reached its optimal state.ConclusionsThe hysteretic curves of the RSPC piers exhibited an evident flag-shaped pattern with a significant pinching effect, and the residual deflection rate was less than 1%, indicating good self-centering ability and repairability. The lateral bearing capacity and displacement ductility of the RSPC piers increased by more than 80% and 20%, respectively, compared to RC piers, demonstrating superior seismic performance. Considering the comprehensive lateral bearing capacity, energy dissipation capacity, and self-centering capacity, the reasonable range of the reinforcement ratio of the energy-consuming rebars in the RSPC piers was determined to be 0.5% ~ 0.6%. The seismic performance of the new piers was optimal when the effective prestress values for the prestressed tendons in the upper and bottom sections were 484 and 646 MPa, respectively. The proposed evaluation method for assessing the resistance resilience of the RSPC piers can accurately identify the resistance resilience of the piers at different stages.关键词：prestressed concrete bridge columns;rocking self-centering;restorable function;seismic resilience359|892|0更新时间：2026-01-23
SEISMIC ISOLATION AND ENERGY DISSIPATION FOR ENGINEERING STRUCTURES
- “在自动驾驶等领域，针对行人遮挡问题，提出了全局特征聚焦与信息增强网络GFFIE-Net，有效提升了检测精度。”摘要：ObjectivePedestrian detection is a crucial task in computer vision, particularly in applications such as autonomous driving, robot navigation, and intelligent surveillance. However, pedestrian occlusion in real-world scenarios remains a significant challenge. Occlusion causes a sharp reduction in the visible range of targets and a substantial loss of pedestrian features, making it difficult for detectors to effectively distinguish between targets and pedestrians. Existing methods, including post-processing optimization, specific model-based improvements, and body-part feature-based approaches, have limitations such as inaccurate handling of heavily occluded positive samples, high computational complexity, and susceptibility to background noise. Therefore, developing a more effective method to address pedestrian occlusion detection is essential to enhance the performance of pedestrian detectors.MethodsThe proposed global feature focusing and information enhancement network (GFFIE‒Net) employed HRNet‒W32 as the backbone network to generate multi-scale feature maps with different resolutions (1/4, 1/8, 1/16, and 1/32 of the input image). These feature maps captured both high-level semantic information and low-level spatial details, which were essential for detecting pedestrians in complex scenes. The convolutional block attention module (CBAM) was embedded after the feature maps to enhance the feature representation and reduce background noise interference. CBAM adjusted the importance of each channel and spatial location in the feature maps through operations such as global average pooling, maxpooling, and small fully connected neural networks in both channel and spatial attention dimensions. This process strengthened the feature information in key areas and suppressed background noise, enabling the network to focus on the target area. Then, considering the limitations of CNN-based methods in global information extraction, the Mamba module was cascaded after the CBAM. The Mamba module first flattened the feature maps into one-dimensional image patch vectors and then used linear layers for feature extraction and transformation. It captured global contextual information and long-range dependencies between feature vectors through forward and backward processing using the state space model (SSM). This process assisted in extracting contextual information around occluded pedestrians and inferring complete pedestrian features based on visible ones. Finally, a hierarchical feature fusion mechanism was designed. This mechanism first utilized the bilinear interpolation algorithm to adjust the spatial resolution of different-scale feature maps to be consistent. Then, it concatenated the three high-dimensional and low-resolution feature maps rich in semantic information along the channel dimension to enhance the deep semantic representation. After that, it combined the preliminarily fused feature map with the low-dimensional and high-resolution feature map containing more detailed location information along the channel dimension. This achieved a comprehensive fusion of high-level semantic and positional detail information, enabling the algorithm to capture multi-level semantic features. The final feature map was processed by a detection head, which generated center heatmaps, scale heatmaps, and offset maps to predict pedestrian bounding boxes.Results and DiscussionsAblation experiments were designed from four aspects to comprehensively verify the effectiveness of the proposed GFFIE-Net improvements. First, the effects of different global information extraction methods on the experimental results were investigated. Second, the effects of various modules on the network performance were analyzed. Third, the impact of different scales on network performance, sequential cascade structure, and the rationalization of hierarchical feature fusion were explored. Fourth, the robustness of the designed enhancement modules was verified by testing them on different backbone networks. Extensive experiments were conducted on three challenging pedestrian datasets: CityPersons, Caltech, and CrowdHuman. The experimental results showed that the R metric reached 43.7% on the heavily occluded subset of the CityPersons dataset, representing an improvement of 4.4 percentage points compared to the baseline method; 33.6% on the heavily occluded subset of the Caltech dataset; and 43.2% on the CrowdHuman dataset, outperforming several mainstream methods. Finally, a visualization analysis of the detection boxes and center heatmaps was conducted. Seven representative practical scene images were selected from the three datasets, including traffic, intersection video surveillance, nighttime, high-density traffic, strong light, small target, and crowded pedestrian scenes. The results showed that compared to the baseline network, GFFIE‒Net produced more significant central responses and more accurate detection box positioning for occluded pedestrians. In the high-density traffic scene, for example, when multiple pedestrians were occluded by one another, the baseline network failed to detect many pedestrians, and the central heatmap exhibited weak responses to occluded individuals. In contrast, GFFIE‒Net accurately identified and located occluded pedestrians. This indicated that GFFIE‒Net effectively handled occluded pedestrians in various scenarios, demonstrating strong adaptability and high detection performance.ConclusionsThe proposed GFFIE‒Net, integrating the CBAM module, Mamba module, and hierarchical feature fusion mechanisms, effectively addresses the challenges of feature loss and background noise in occluded scenarios. The experimental results from three benchmark datasets demonstrate the superiority of GFFIE‒Net compared to existing methods, particularly in managing heavily occluded pedestrians. Future research can explore semi-supervised or self-supervised learning using limited labeled data. This approach can reduce dependence on large-scale labeled datasets, enhance model generalization, and improve the method's applicability and accuracy across diverse scenarios.关键词：pedestrian detection;mamba;feature enhancement;CBAM144|226|0更新时间：2026-01-23
- “在民间文学文本修复领域，专家提出了基于知识蒸馏的修复方法，有效提升了修复质量和效率。”摘要：ObjectiveFolk literature serves as an important carrier for depicting the social life and cultural perspectives of the general public. Due to natural, historical, or human factors, the words in folk literature texts are often ambiguous, difficult to identify, or even completely missing. For effective research and dissemination, it is necessary to repair incomplete folk literature texts. A significant difference exists between folk literary text data and the pre-training data used during the pre-training phase of pre-trained language models. For example, differences occur in the form of specialized vocabularies and structural features. These differences lead to catastrophic forgetting when directly fine-tuning pre-trained language models, as the model must perform extensive parameter adjustments and can forget previously learned universal language knowledge. Avoiding catastrophic forgetting in pre-trained language models for this repair task and ensuring that the restored sentences align with the linguistic characteristics of folk literature are the two main challenges. A knowledge-distillation-based method for folk literature text restoration is proposed to address these issues.MethodsConsidering the characteristics of limited annotated data, the presence of specialized vocabularies, and the structural nature of folk literary texts, this study adopted a pre-trained language model to expand knowledge distillation and train the student network, enabling the automatic restoration of incomplete folk literary sentences. First, the pre-trained language models and student networks were utilized to extract the basic feature vectors of characters from folk literary texts. These basic feature vectors were then utilized to construct semantic feature matrices, which underwent intermediate feature knowledge distillation. This process involved computing the SmoothL1 loss between the semantic feature matrices of each layer in the pre-trained language model and the student network, ensuring minimal distribution differences between the output features of the student network and the teacher network. The student network’s comprehension of the overall semantic meaning of sentences was enhanced by leveraging the teacher network’s understanding of character-level general knowledge. Then, the structural relationships among the basic feature vectors in the semantic feature matrix were treated as the structural knowledge of folk literary text sentences. A structural feature matrix was constructed and subjected to structural feature knowledge distillation to reinforce the constraints of structural knowledge during the parameter update process of the student network, enhancing the structural regularity of the repaired sentences.Results and DiscussionsFor the three typical genres of folk literature, the corresponding datasets were constructed, and experimental studies were conducted. In the comparative experiments, BERT applied to the constructed folk literary text datasets showed improvements in average bilingual evaluation understudy (BLEU) values by 0.12%, 0.80%, and 0.29%, and reductions in PPL (perplexity) values by 146.07, 168.80, and 72.52, respectively. GPT applied to the constructed folk literary text datasets showed improvements in average BLEU values by 1.00%, 1.28%, and 0.66%, and reductions in PPL values by 233.25, 303.39, and 144.96, respectively. BART applied to the constructed folk literary text datasets demonstrated improvements in average BLEU values by 6.19%, 6.41%, and 11.67%, and reductions in PPL values by 38.75%, 7.48%, and 14.82%, proving the effectiveness of the proposed method. In the ablation experiments, the average BLEU of the w/S model was, on average, 0.3% higher than that of the w/F model, indicating that structural feature knowledge distillation has a better effect on improving the accuracy of folk literary text sentences compared to intermediate feature knowledge distillation. The PPL index was, on average, 22 higher, indicating that intermediate feature knowledge distillation has a better effect on improving the fluency of folk literary text sentences. The results of the ablation experiments also indicated that combining these two distillation methods further improved the average BLEU index and reduced the PPL index compared to the w/S model and w/F model. In the mask rate experiment, the combined knowledge distillation method showed improvements in average BLEU indices and reductions in PPL indices relative to traditional fine-tuning methods, demonstrating the robustness of the combined knowledge distillation method. In addition, when the mask ratio was set to 15% or 20%, the average BLEU and the PPL metrics typically demonstrated the most optimal improvement, indicating that the combined knowledge distillation method was more effective in capturing crucial linguistic information from incomplete folk literary text sentences when the number of missing characters was moderate, providing the student network with rich and accurate semantic and structural knowledge. The case study intuitively demonstrated the execution results of the combined knowledge distillation method, indicating that the method generated coherent, complete, and well-formatted sentences.ConclusionsConsidering the specific vocabulary and structural features of folk literature, the catastrophic forgetting phenomenon faced by existing controllable text generation methods, and the insufficient generalization when handling data from the vertical domain of folk literature, a combined knowledge distillation method is proposed. This method constructs semantic and structural feature matrices and conducts knowledge distillation on both. Experimental results demonstrate that the method effectively prevents catastrophic forgetting in pre-trained language models and generates sentences with more accurate semantics, comprehensive content, and improved alignment with the formatting requirements of folk literature texts.关键词：folk literature;text restoration;knowledge distillation;catastrophic forgetting;structural knowledge85|156|0更新时间：2026-01-23
ARTIFICIAL INTELLIGENCE
-
Experimental Study on Mechanical Properties of Coconut Fiber Reinforced Lime Improved Soil Mechanics AI导读
“据最新研究，椰壳纤维改性石灰土能显著提升工程适用性，纤维掺入改善了石灰土性质，最优掺量为0.75%，为实际工程应用提供参考。”摘要：ObjectiveThe application of lime soil in engineering is further enhanced through the introduction of new materials for improvement and optimization. Coconut fiber, a natural high-molecular material with superior mechanical properties, presents a promising solution. Utilizing coconut fiber to improve lime soil not only reduces construction costs but also increases soil strength and stability while mitigating resource waste and environmental pollution. Therefore, this study aims to enhance the mechanical properties of lime soil by employing coconut fiber as a reinforcing material and to examine its macroscopic mechanical behavior and microstructural characteristics.MethodsThe study analyzed the effects of different curing ages (7, 14, 28, and 60 days), coconut fiber content (0, 0.25%, 0.50%, 0.75%, and 1.00%), dry and wet cycles (0, 1, 2, 3, 4, 5, and 6 times), and compaction degrees (90%, 93%, 96%, and 99%) on the mechanical properties of lime soil by conducting the unconfined compressive strength test, dry and wet cycle test, nuclear magnetic resonance (NMR) test, and scanning electron microscope (SEM) test. These tests were performed to reveal the internal mechanism behind the improvement of the mechanical behavior of coconut fiber reinforced lime soil.Results and Discussions1) For soils of the same age, the addition of coconut fiber significantly increased soil strength compared to soils without coconut fiber. As the amount of coconut fiber increased, the strength of the soil showed a tendency to first increase and then decrease, while the destructive strain of the soil gradually increased, changing the soil behavior from brittle to plastic. This occurred because the random distribution of coconut fiber formed a spatial mesh structure that restrained soil deformation and particle movement, improving soil integrity. However, fiber doping exceeding 0.75% created weak planes, reducing strength. Under different maintenance ages, the compressive strength gradually increased, and the destructive strain became smaller, showing brittle damage. This was mainly due to ion exchange, volcanic ash reaction, and carbonation reaction of lime, generating cementitious substances that reinforced the soil. 2) Under the action of dry and wet cycles, the soil surface produced cracks and shedding, and its quality was obviously reduced. However, the mass loss of lime soil mixed with coconut fiber was significantly less, proving that coconut fiber inhibited crack formation and reduced surface detachment. When the compaction degree was the same, with the increase in the number of cycles, the soil mass showed a trend of increasing and then decreasing, reaching the maximum after the first cycle. This initial increase was attributed to the ongoing reactions of unreacted lime requiring moisture and the volcanic ash reaction during drying, generating gelling substances that filled pores and increased strength. However, with further cycles, the lime was consumed, cementitious material and quality gradually reduced, leading to a decrease in compressive strength. When the cycle times were the same, an increased compaction degree led to greater soil compressive strength. A higher compaction degree resulted in smaller soil pores and a larger contact area between coconut fiber and soil, producing greater interface friction and higher soil strength. This indicated that the addition of coconut fiber and increased compaction degree enabled the soil to effectively resist the effects of wet and dry cycles. 3) The microscopic test showed that the pore space of the soil exhibited a tendency to decrease and then increase with the number of wet and dry cycles, reaching the maximum and having the least cementitious material in the sixth cycle. The addition of fiber caused the pore space in the lime soil to increase significantly. Fiber guided the distribution of hydrate, producing more gelling material that bonded the fiber and soil together. The rough fiber surface generated greater friction with the soil, limiting soil deformation.ConclusionsCoconut fiber-lime-treated soil shows strong potential for application in road and slope engineering projects. The research findings provide valuable insights for practical implementation. The study demonstrates that incorporating coconut fiber effectively enhances the mechanical properties and durability of lime-treated soil, increases its strength, alters its failure mode from brittle to more plastic behavior, and improves its resistance to damaging wet-dry cycles.关键词：coconut fiber;lime;unconfined compressive strength;Dry and wet cycle;Microanalysis233|494|0更新时间：2026-01-23 - “在桥梁服役状态评估领域，专家通过实桥试验和数值模拟，揭示了服役27年RC空心板桥的抗弯性能，为桥梁运维决策提供参考。”摘要：ObjectiveAccurate assessment of the flexural bearing capacity of reinforced concrete (RC) voided slab bridges after service degradation is a critical aspect of bridge operation and maintenance management. It provides a reference for evaluating service conditions and making maintenance decisions for similar in-service bridges. This study focuses on a simply supported RC voided slab bridge with 27 years of service history located in the Shangqiao Interchange section of the Chengdu‒Chongqing Expressway. Based on inspection data, in-situ static load tests, bending failure tests in the laboratory, numerical simulations, and theoretical analysis, the research systematically investigates the flexural bearing capacity of the bridge structure and evaluates its operational safety based on current and outdated design codes.MethodsFirstly, the historical inspection reports of the bridge structure were reviewed, with the inspection results obtained in 2020 serving as the baseline for a comprehensive analysis of the bridge’s disease characteristics and distribution. Secondly, prior to the bridge’s demolition, in-situ static load tests were conducted using two tri-axle heavy-duty trucks, applying four loading conditions and eight-stage loading methods to measure and analyze the strain and deflection of each RC voided slab. Based on the measured deflection data, a calculation method for the measured load transverse distribution coefficient of RC voided slab bridges was proposed and compared to theoretical results. The working performance of the hinge joints between the RC voided slabs was then evaluated. Thirdly, three voided slabs were dismantled from the original bridge for bending failure tests in the laboratory, enabling experimental investigations on the flexural bearing capacity of individual slabs and revealing the failure process and mechanism of RC hollow slabs. Fourthly, a three-dimensional nonlinear solid model of the RC hollow slab was developed in ABAQUS, numerically reproducing the mechanical behavior throughout the failure process, which demonstrated high consistency with the experimental results. Finally, by considering different working states of the hinge joints, the flexural bearing capacity of the original bridge was comprehensively evaluated through theoretical calculations, and a comparative analysis of its service performance and structural safety was performed based on both current and historical bridge design codes in China.Results and DiscussionsAfter 27 years of service, due to prolonged full-load operation, transverse cracks, longitudinal cracks, mesh cracks, honeycombing, bursting, spalling, and other defects were observed almost at the bottom of each RC voided slab, with hinge joint deterioration identified as the most critical factor compromising the bridge’s flexural bearing capacity. The damaged hinge joints impeded effective load distribution across the transverse direction and, in severe cases, even caused single-slab load-bearing phenomena, significantly undermining the bridge’s structural integrity. In-situ tests revealed highly uneven mid-span deflection distributions among slabs, confirming weakened hinge joint connections, which aligned with the observed hinge joint deterioration. Under the four loading conditions, the measured load distribution coefficients across the RC voided slabs were generally similar. Under eccentric vehicle loading, the maximum distribution coefficient on the loaded side reached 0.440 7, while the minimum on the unloaded side was only 0.036 3, an 11.1‒fold difference, far exceeding the theoretical 1.9‒fold ratio. This finding demonstrated that after 27 years of service degradation, the original bridge’s hinge joints no longer functioned effectively as transverse connectors, severely impairing the overall flexural performance of the bridge structure. Bending failure tests in the laboratory showed that despite initial defects, the flexural capacity of RC voided slabs remained satisfactory. The loading behavior and failure processes of the three tested slabs were nearly identical, exhibiting typical under-reinforced beam bending failures, with residual mid-span displacements exceeding 200 mm after loading termination, yet without slab fracture. The load-mid-span deflection curves of all three slabs displayed distinct yield and failure thresholds, approximating three linear segments with varying slopes. The post-yield plateau segment exhibited stable and gradual deformation, indicating good ductility. The average yield and ultimate loads were 544.8 kN and 605.4 kN, respectively, with a mean ultimate deflection of 271.4 mm and a ductility coefficient of 8.51. Numerical simulations incorporating a plastic damage constitutive model successfully replicated crack propagation and slab failure processes, with simulated load-deflection curves closely matching the experimental data. Although the yield load simulation error was minimal (-2.3%), the yield deflection prediction exhibited larger discrepancies. The simulations more accurately captured the simultaneous yielding of tensile reinforcement and slab concrete. Therefore, numerical modeling is a valuable complementary tool to experimental studies, providing additional perspectives for assessing the in-service capacity of RC voided slab bridges. Theoretical calculations indicated a mere -0.3% error between the computed and measured flexural capacities. Comparative analyses of the original bridge's flexural performance under current and historical design codes (referred to as “old code”) were conducted by considering three hinge joint conditions: theoretical maximum distribution coefficient, measured maximum coefficient, and extreme single-slab loading. Results showed that the bending moment effects calculated under the old code were significantly lower than those under the new code, reaching only 68.2%, 65.8%, and 64.8% of the new code’s values, respectively. Even under the extreme single-slab condition, the old code's requirements were met with a 60.3% reserve capacity. In contrast, the current code yielded reserves of only 49.5%, 17.0%, and 3.8% for the three conditions. When based on the yield moment, the reserve capacity under the measured shear key condition was merely 5.3%. Further deterioration leading to single-slab loading will reduce the flexural capacity below the moment effect, pushing the slab into plastic deformation and posing severe safety risks under overload scenarios.ConclusionsThis study systematically analyzes post-service degradation characteristics employing multiple research methodologies on actual RC hollow slab bridges. The research provides reliable methodologies for assessing the service conditions of numerous in-service RC hollow slab bridges and offers essential decision-making references for the maintenance operations of similar structures.关键词：RC voided slab bridge;hinge joint disease;load tests;transverse distribution of load;bending tests;bearing capacity265|467|0更新时间：2026-01-23
- “在张拉整体结构轻量化设计领域，专家提出了两阶段质量优化方法，通过量子天牛须搜索算法等，有效提升了结构优化效果，为复杂结构设计提供新方案。”摘要：ObjectiveConsidering economic costs, structural applications, and convenient transportation, the lightweight design of tensegrities with multi-self-stress modes is increasingly popular. Existing approaches primarily concentrate on mass optimization of tensegrity structures with multi self-stress modes under specific loads, without addressing the initial forming conditions that ensure structural geometrical stability. Neglecting these conditions can increase the risk of collapse once the structural stiffness diminishes due to unpredictable variations in external loads. In addition, in the case of complex tensegrities with multi-self-stress modes, the influence of prestress distribution on structural mass has been largely ignored, as current methods fail to establish the relationship between structural mass and self-stress modes. Therefore, conducting research on the mass optimization of tensegrity structures with multi-self-stress modes based on heuristic optimization algorithms is of great significance.MethodsA multi-stage mass optimization method was proposed to achieve lightweight design of complex tensegrity structures with multi-self-stress modes. In the first stage, based on the form-finding theory of tensegrities, the L2 norm was applied to characterize the geometric symmetry of the structures, and the integrity-feasible prestressing was obtained. In the second stage, the Similar Transforming Strategy (STS) was utilized to expand the population search range once the judgment on similar prestress distribution was completed. Then, the constrained optimization model of structural mass minimization was established by setting the objective function as the minimum structural mass under full stress, and the constraint conditions, such as cable yield, bar yield, and yield buckling of bars, were considered. The effects of the Quantum Beetle Search Algorithm (QBSA), Quantum Evolution Algorithm (QEA), and Beetle Antennae Search Algorithm (BASA) were compared concerning the search of the combination coefficient for optimal multi-self-stress modes, and therefore, the objective of lightweight design of the structural system was achieved. The maximum node displacement of all comparison objects was adjusted to the same value through the insertion of the Adjusting Prestress Level (APL) strategy to accurately compare the impact of various prestress distributions on optimal structural mass. The performance of three different heuristic optimization algorithms in the process of mass optimization of the spatial four-way tensegrity plate before and after the implementation of the APL strategy was compared. In addition, the optimal algorithm for subsequent research was selected. The influence of various geometric parameters on the lightweight design of the tensegrity torus structure was analyzed. Comparative studies were conducted using the following three schemes: Scheme 1 (simultaneously implementing APL and STS), Scheme 2 (implementing APL without executing STS), and Scheme 3 (implementing STS without executing APL).Results and DiscussionsThe effectiveness of this method was comprehensively examined through three illustrative examples, and the following noteworthy results were obtained: 1) For the four-way tensegrity plate without the APL strategy, the elite individuals of QBSA, QEA, and BASA were 35, 33, and 6, with structural masses of 4 534, 4 808, and 5 039 kg, and optimization rates of 8.05%, 5.85%, and 3.24%. With the implementation of the APL strategy, the number of elite individuals in QBSA, QEA, and BASA was 36, 30, and 5, with structural masses of 3 790, 4 280, and 4 889 kg, and optimization rates of 58.95%, 51.21%, and 48.73%. QBSA has the highest number of elite individuals, which results in the lightest optimized mass and the greatest optimization rate, demonstrating its superior optimization performance. In addition, implementing the APL strategy effectively ensures the consistency of the maximum nodal displacement, while the structures optimized by all three algorithms achieve lighter masses and better results. 2) Changes in prestress distribution affect the loading conditions of tensegrity structures. As the structural mass reduces, the maximum nodal displacement exhibits fluctuations, and changes in prestress distribution also influence the nodal position with maximum displacement, while it remains unchanged after adjusting the prestress level. 3) The analysis of the tensegrity torus shows that when kn is determined, the optimized mass decreases with the increase of nx at first, and then rises when nx exceeds 4; when nx is determined, the optimized mass improves with the increase of kn. In addition, compared to Scheme 2 and Scheme 3, Scheme 1 achieves a lighter structural mass and a higher optimization rate. 4) The optimization results of the triangular prism tensegrity plate show that the variation of prestress distribution has a minor impact on structural mass optimization for plate B with Free Prestress (FP), while it has a significant impact on plate A with FP. In addition, the research results on plate A and plate B (with FP or Splicing Prestress (SP) applied) indicated that the connection bar plays an important role in structural lightweight design. In addition, the final optimized mass of plate A with FP and plate B is lower than the optimized mass corresponding to SP.ConclusionsThe establishment of multi-stage objective functions facilitates effective lightweight design while ensuring the integral feasible prestress of tensegrity structures. QBSA demonstrates superior local and global optimization capabilities, and by incorporating APL and STS strategies, the optimization outcomes are effectively enhanced. In addition, for complex assembled prestressed members, this method achieves a more favorable prestress distribution compared to that obtained through the conventional SP approach.关键词：tensegrity structures;lightweight design;multi-self-stress modes;prestress level;prestress distribution167|507|0更新时间：2026-01-23
- “在建筑领域，专家通过风洞试验分析了建筑深宽比和来流湍流特性对矩形高层建筑扭转向脉动风荷载空间相关性的影响，为建筑结构设计及荷载规范修订提供参考。”摘要：ObjectiveTall buildings in contemporary construction practice predominantly adopt slab-type designs with substantial side ratios. In these rectangular tall structures, the response to torsional wind vibrations becomes a critical issue. Accurate calculation of wind-induced vibrations requires a comprehensive understanding of the spatial correlation of fluctuating wind loads. Existing torsional coherence function models, derived from wind tunnel tests on specialized or small side ratio building models, often neglect the effect of diverse building section forms on the coherence function. Therefore, their applicability to common tall buildings with larger side ratios remains limited. This study addresses this limitation by proposing a series of spatial correlation mathematical models for torsional fluctuating wind loads on high-rise buildings, enhancing their practicality and applicability.MethodsFirstly, four types of wind fields (O1, S1, O2, and S2) were simulated based on data from the Engineering Sciences Data Unit (ESDU). The mean wind speed profile was generated using the logarithmic rate formula recommended by ESDU—85020. The theoretical turbulence intensity profile was produced based on the formula recommended by ESDU—82026. The turbulence integral scale was determined using the formula indicated by ESDU—74031, and the fluctuating wind speed spectrum was obtained using the von‒Karman spectrum recommended by ESDU—74031. Turbulence intensity in S‒type wind fields was greater than that in O‒type wind fields, and the turbulence integral scale in type‒1 wind fields was larger than that in type‒2 wind fields. Secondly, 21 test models with side ratios ranging from 1/9.0 to 9.0 were created by assembling 12 segments, using a length scale of 1∶200. The completed model measured 0.50 m in height, 0.06 m in width, and 0.06 to 0.54 m in length. Seven layers of measuring points were placed vertically at heights of 0.10H, 0.30H, 0.50H, 0.65H, 0.80H, 0.90H, and 0.98H. The measuring point layers were numbered 1 to 7 from bottom to top, maintaining a consistent arrangement across all layers. Finally, synchronous pressure measurement wind tunnel tests were conducted on the 21 models under four wind fields. Time history data of wind pressure coefficients at the measuring points on the models were collected using synchronous pressure scanning valves. The sampling frequency was 400 Hz, and the sampling duration was 90 s, producing a total of 36 000 data points.Results and DiscussionsBased on the experimental results, the vertical correlation coefficient and coherence function of buildings with various side ratios under different wind fields were calculated. The influences of side ratio, turbulence intensity, and turbulence integral scale on the vertical correlation coefficient and coherence function of torsional fluctuating wind loads were analyzed. Mathematical models of the vertical spatial correlation of torsional fluctuating wind loads for rectangular high-rise buildings with side ratios ranging from 1/9.0 to 9.0 were established using the least squares method, and the accuracy of these models was compared to the experimental data. The results showed that the correlation coefficient of torsional fluctuating wind load exponentially decreased with increasing separation distance, and the attenuation rate varied with side ratio. When D/B
- “在城市矿坑回填治理领域，专家验证了高聚物碎石桩加固尾泥地基的有效性，为矿区生态恢复和土地开发提供新方案。”摘要：ObjectiveThe backfill treatment of surrounding mining pits represents an emerging land-use strategy. Utilizing tailings stored in mining areas as backfill material reduces construction costs and addresses the challenge of tail mud disposal. Polymer gravel piles provide high strength, rapid formation, and maintenance-free performance after completion, meeting the requirements of complex foundation treatment. Model tests are conducted to investigate the reinforcement mechanism and bearing characteristics of polymer gravel piles in such foundations to verify the effectiveness and feasibility of polymer gravel piles in treating granite tail mud foundations. The influence of pile length, pile diameter, and polymer content on reinforcement performance is analyzed. The findings provide a potential technical solution for ecological restoration and land development of mine surfaces.MethodsThe materials of the model pile were gravel and polymer, and the maximum particle size of the gravel was controlled within 10 mm. The test gravel was screened with a standard sieve, and gravels with particle sizes of 5~10 mm were selected as the pile material. The polymer was a non-aqueous reaction-type two-component polymer, which was mixed with polyol and isocyanate at a mass ratio of 1:1. After a specific period, a foam-solidified substance was formed. The polymer crushed stone pile was formed by mixing the polymer with gravel. First, direct shear tests and consolidation tests were conducted on granite tail mud with different moisture contents to obtain the shear strength index, compression modulus, and compression coefficient of the granite tailings. Then, shear strength tests were performed on polymer gravel specimens with different polymer contents to determine their shear strength using a DZJ-300 type large direct shear instrument. Then, a total of nine sets of model test schemes for the unreinforced tail mud foundation, the tail mud foundation reinforced by gravel pile, and the tail mud foundation reinforced by polymer gravel pile were designed, and model tests were conducted to study the reinforcement of the granite tail mud foundation using different reinforcement methods. A jack and reaction device were employed to load the foundation model, displacement meters were utilized to record the foundation settlement, and earth pressure boxes were utilized to record the stresses at the pile top, pile bottom, and pile side. The effects of pile length, pile diameter, and polymer content of the polymer gravel piles were analyzed. Finally, the bearing characteristics of the reinforced foundation were tested, including the measurement and analysis of parameters such as settlement and pile bottom stress.Results and DiscussionsAfter the static load test, it was observed from the excavated piles that the polymer material reacted to form a well-connected and uniform pile with good integrity. The overall shape of the pile was cylindrical, and no defects or necking phenomena appeared on the pile surface, indicating a complete pile structure. During the reaction process, the polymer materials bonded with the surrounding soil, and their volume expanded, producing a compaction effect on the soil around the pile. This process formed a jagged structure on the pile surface, which increased the contact area between the pile and soil, improving the bearing capacity of the tail mud foundation. As the density of the polymer increases, the cementation and compaction effects also become stronger. Based on the test results, the final settlement of the tail mud foundation reinforced with the gravel pile was 98.62 mm, whereas the final settlement of the tail mud foundation reinforced with the polymer gravel pile was 42.78 mm. The reinforcement effect of the polymer gravel pile was significantly better than that of the traditional gravel pile, resulting in a 56.6% reduction in final settlement. It was evident that the bearing capacity of the granite tail mud foundation increased substantially after reinforcement with polymer gravel piles. As the load gradually increased, the stress ratio of the gravel pile to soil for the gravel pile exhibited a slower growth trend, eventually reaching 3.32, whereas the stress ratio for the polymer gravel pile showed a linear growth trend, eventually reaching 5.30, which was 59.6% higher than that of the gravel pile. With the increase in load, the pile bottom stress also increased gradually. When the load reached 300 kPa, the pile bottom stress for the polymer gravel pile and the gravel pile were 121 kPa and 40 kPa, respectively. It was evident that the pile bottom stress of the polymer gravel pile increased by 202.5% compared to the gravel pile. From the curve of soil vertical stress variation along the depth of the model foundation, it was observed that the vertical stress in the soil was mainly concentrated in the upper part of the foundation and rapidly attenuated with increasing foundation depth. The length, diameter, and polymer content of the piles have a significant impact on the reinforcement effect. As these parameters increase, the reinforcement effect also improves accordingly. The research findings provided a new technical approach for surface ecological restoration and land development of abandoned mine pits, demonstrating broad application po-tential.ConclusionsThis study confirms the effectiveness of polymer gravel piles in improving granite tail mud foundations through model testing. The results reveal that polymer gravel piles significantly enhance the bearing capacity of the foundation and effectively reduce its settlement. In addition, pile length, pile diameter, and polymer content are critical factors influencing the reinforcement performance. Therefore, in practical engineering applications, appropriate parameter combinations should be selected based on site-specific conditions to achieve optimal reinforcement outcomes.关键词：polymer gravel piles;granite tail mud;foundation treatment;static load model test202|351|0更新时间：2026-01-23
- “在建筑领域，研究人员开发了新型高连接效率锚具，有效提升纤维增强复合材料筋的锚固性能，为解决预应力钢筋问题提供新方案。”摘要：ObjectiveThe fiber-reinforced polymer (FRP) tendon possesses high specific strength, good corrosion resistance, and high creep rupture stress, making it an ideal material to resolve the critical issues of excessive weight and corrosion in prestressed steel bars or strands. However, developing a novel anchorage with high connection efficiency is essential due to the anchoring challenges associated with anisotropic FRP tendons. The second-generation variable-stiffness wedge (VSW) anchorage, referred to as the second-generation anchorage, is developed to address the limitations of the first-generation VSW anchorage, including size, assembly complexity, cost, and wedge follow-up, while maintaining comparable anchoring performance. The development cycle of new anchorages can be significantly shortened by verifying the universal applicability of the second-generation VSW anchorage to various FRP tendons, enhancing engineering practicality and cost-effectiveness.MethodsFirstly, three types of epoxy resin cast bodies modified with quartz sand, basalt fiber, and carbon fiber, respectively, were prepared to evaluate their compressive behavior. The failure mode and load-displacement curve of the cast bodies were analyzed, and the compressive str-ength and elastic modulus were calculated based on the testing standards. Secondly, the VSWs were manufactured using the molding method. Composite layers consisting of chopped fibers impregnated with vinyl resin and bidirectional fiber cloth were alternately stacked to enhance structural integrity. The layered materials were compression-molded at 120 °C for 20 minutes using a hydraulic press. Post-machining involved angular cutting of high-modulus segments followed by surface roughening to improve interfacial bonding. A dual-material system (quartz-modified resin and epoxy resin) was cast in a split mold for ambient curing. A 0.03 mm stainless steel foil was epoxy-bonded to the collet surface to reduce friction and enhance wedging synchronization with FRP tendons. Thirdly, the end-splitting and surface sandblasting anchoring methods were utilized to investigate the tensile properties of BFRP and CFRP tendons. Fourthly, an auxiliary set of steel wedges was installed at the end of the VSW anchorages to achieve synchronous wedging action of multiple wedges. Three types of anchorages were compared: the first-generation anchorage, the second-generation anchorage, and the stainless steel wedge anchorage. The static anchoring performance of these anchorages was evaluated through static tensile tests.Results and DiscussionsThe findings indicated that the displacement corresponding to the maximum load of the quartz sand-modified resin decreased with the increasing mass fraction of quartz sand compared to pure resin. This behavior was attributed to the resin matrix's dominant contribution to deformation, where reduced resin content diminished the matrix's deformation capacity. In contrast to both pure resin and quartz sand-modified resin, the chopped fiber-modified resin exhibited superior load-bearing capacity and enhanced post-peak load retention. This improvement primarily resulted from the inherent high strength of the fibers and the effective constraint on transverse deformation and crack propagation achieved through the randomly distributed chopped fibers, which collectively improved the modified resins' strength and modulus. The compressive strength and elastic modulus of the pure resin were measured as 106.6 MPa and 3.1 GPa, respectively. For the quartz sand-modified resin, the compressive strength initially decreased and then increased with rising sand content, rather than exhibiting a monotonic growth trend. When maintaining identical preparation processes and reinforcement materials, the accuracy of the compressive strength was influenced by void defects and size effects. Subsequent optimization was achieved through vacuum degassing to mitigate these interfacial imperfections. The processes of end splitting and surface sandblasting were beneficial for enhancing the interfacial bonding capacity between the FRP tendons and resin. The small steel wedge anchorage with a limiting action effectively addressed the issue of asynchronous sliding among multiple VSWs, and the stainless steel sheet attached to the outer surface of the VSWs contributed to improving their sliding capability with the FRP tendons. The variable-stiffness design was advantageous in mitigating stress concentration in the FRP tendons within the anchoring area. The anchoring efficiency (ηa) for round ϕ7 BFRP tendons, round ϕ10 BFRP tendons, and round ϕ10 CFRP tendons using the second-generation anchorages was recorded as 88%, 86%, and 99%, respectively. When BFRP tendons were anchored, the anchoring efficiency of the second-generation anchorages significantly surpassed that of the steel wedge anchorages (ηa = 78%) and approached that of the first-generation anchorages with larger anchorage size (ηa = 91%). When CFRP tendons were anchored, the anchoring efficiency achieved by the second-generation anchorages remained comparable to that attained by the stainless steel wedge anchorage (ηa =107%).ConclusionsFor the second-generation anchorage, the VSW consists of three segments with varying elastic moduli: a low-elastic-modulus segment made of epoxy resin, a medium-elastic-modulus segment composed of quartz sand-modified epoxy resin (with a quartz sand to resin mass ratio of 100%), and a high-elastic-modulus segment incorporating vinyl resin-impregnated chopped fibers and bidirectional fiber fabric. The end splitting combined with surface sandblasting effectively enhances the bond strength between smooth FRP tendons and resin. The auxiliary steel wedge anchorage resolves the asynchronous wedging problem of multiple VSWs, while externally bonded stainless-steel thin sheets improve the synchronous wedging capability of VSWs with FRP tendons. The variable-stiffness design significantly mitigates stress concentration in FRP bars within the anchorage zone. For BFRP tendons, the second-generation anchorage achieves substantially higher efficiency than stainless-steel wedge anchors while approaching the performance of larger first-generation anchors. For CFRP tendons, the efficiency not only exceeds the code-specified requirement of 90% but also matches that of stainless-steel wedge anchorages. Future work will focus on continuous optimization of variable-stiffness wedge dimensions and materials, as well as refining manufacturing processes to reduce production costs. Further investigation of the impact performance of FRP tendon-VSW anchorages will also enhance their engineering applicability.关键词：bridge engineering;BFRP tendon;CFRP tendon;variable-stiffness wedge anchorage;anchoring efficiency246|420|0更新时间：2026-01-23
- “在建筑施工领域，专家提出了“L”形不出筋叠合板端部设计新方法，通过静载试验验证了其可行性，为提高施工效率提供新方案。”摘要：ObjectiveTraditional composite slabs require protruding reinforcement bars in the prefabricated base plate during production to ensure the integrity of the interface between joints and beam-panel connections. However, these protruding bars hinder the handling and stacking of components, adversely affecting on-site construction and component quality. Therefore, this study proposes a construction method for composite slabs with an L-shaped rabbet that eliminates the need for protruding reinforcement bars.MethodsThis study conducted static loading tests on six full-scale floor slab specimens with end beams to investigate the effect of the L-shaped rabbet end design method on the integrity of beam-panel connections and the bearing capacity of composite slabs. The specimens consisted of three composite slabs with L-shaped rabbets (design method one), two traditional composite slabs, and one cast-in-place solid slab with the same dimensions. The L-shaped rabbet composite slabs served as the experimental group, while the remaining slabs were designed traditionally and served as control groups 1 and 2, respectively. The study compared the load-bearing capacity and crack resistance of the new slabs. The loading was applied by stacking weights, with each weight block weighing 1 ton (900 kg). Each floor slab accommodated up to nine weight blocks in a single layer. The loading sequence followed a centrally symmetrical pattern to ensure even stress distribution during testing. The measurement parameters during the test included: 1) the number of weight blocks applied; 2) vertical displacement at the mid-span and on the upper part of beams, and horizontal displacement at the lower part of beams; 3) strain of reinforcing bars under stress; and 4) crack width at the mid-span and at the interface between the slab and beams. Strain gauges were attached to the reinforcing bars on the upper and lower surfaces of the composite slabs. Displacement meters were installed at the slab ends, mid-span, along the beam edges, and at the top. Crack depth and width were observed using a crack depth-width gauge under various loading conditions. Instruments such as strain acquisition devices, high-precision displacement meters, and crack observation tools were utilized to collect data from critical areas of the specimens. These included strain on L-shaped reinforcement at the ends, strain on protruding reinforcement at the ends of traditional composite slabs, deflection at the mid-span of composite slabs, and crack widths and spacings at the mid-span. These data provided a detailed basis for analyzing the overall performance of the floor slabs and the local performance of steel and concrete components.Results and DiscussionsUpon analyzing the experimental results, the study evaluated the number, spacing, and width of mid-span cracks, as well as the width of end cracks, as observation indicators. The crack development process in each group of slabs was similar. Cracks initially appeared at the locations of maximum bending moments at the slab ends and mid-span. As the load increased, both the number and width of mid-span cracks increased continuously, with smaller crack spacing and transversely extending cracks, exhibiting typical flexural failure characteristics. Regarding the number of cracks, the new slabs had an average of 13.3 mid-span cracks, close to the 14 cracks observed in traditional slabs. In terms of crack spacing, the average spacing for the new slabs was 10.96 cm compared to 9.275 cm for the traditional slabs, indicating a small difference. This finding indicated that, compared to traditional slabs, the new end construction did not reduce the bearing capacity at the mid-span. Regarding crack width at the slab ends, under a uniformly distributed load of 27.9 kN/m2, the end crack width at each sampling point indicated that the new slabs exhibited similar crack widths to traditional slabs and performed significantly better than the cast-in-place slabs. This similarity was evident from the crack width data and the consistent changes in the curve slopes between the two slab types, confirming that the L-shaped rabbet end design provided comparable overall integrity at the beam-panel connections to that of the traditional protruding reinforcement method in composite slabs. Regarding deflection, the mid-span deflection was considered the representative value. The development of deflection in the new slabs was similar to that of the traditional slabs. The rate of deflection increase in the new slabs was slightly lower than that in traditional slabs, and the deflection at the limit state of bearing capacity was also lower, while the cast-in-place slab demonstrated the lowest stiffness. Based on a comprehensive analysis of deflection and cracking behavior, the end construction with L-shaped reinforcement did not significantly affect the bending stiffness at the mid-span of composite slabs. Regarding strain, similar to traditional reinforcement, the L-shaped reinforcement experienced minimal compressive stress during normal use, particularly before concrete cracking, and contributed little to the bending capacity at the member ends. As the height of the compressed zone decreased, the L-shaped reinforcement at the ends began to bear tensile stress. In the later stages of loading, both the L-shaped reinforcement in the new slabs and the protruding reinforcement in traditional slabs yielded, fully utilizing the load-bearing capacity of the reinforcement. The design of the L-shaped rabbet compensated for the potential reduction in bond strength caused by the shorter length of the L-shaped reinforcement.ConclusionsThe width and distribution of end cracks in the new slabs are highly similar to those in traditional slabs and significantly better than those in cast-in-place slabs using the L-shaped rabbet during the loading process. This finding indicates that the L-shaped end design method provides sufficient reliability in beam-panel connections. The L-shaped bend restricts relative slippage between the reinforcement and the concrete, ensuring a stable connection between the slab end and the main beam. It compensates for the potentially low bond strength caused by the shorter length of the L-shaped reinforcement, enhancing structural safety. The load-bearing capacity of the new slabs did not significantly decrease and was equivalent or nearly equivalent to that of traditional protruding reinforcement composite slabs, satisfying the requirements for normal service limit state and ultimate limit state conditions. This preliminary validation confirms the feasibility of the L-shaped rabbet design for composite slabs without protruding reinforcement at the ends.关键词：composite slab without extended bars;full-scale test;L-shaped groove;mechanical performance;beam-slab interface201|465|0更新时间：2026-01-23
- “在地质工程领域，专家通过结构面剪切蠕变试验，揭示了结构面与剪切面夹角、间距对岩体蠕变特性的影响，为高陡边坡稳定性分析提供新思路。”摘要：ObjectiveLong-term creep deformation of a slope is significant for major engineering practice and the natural environment. A high and steep rock slope located in a high tectonic stress area can creep due to the elevated stress level of the slope. After long-term creep deformation accumulates, it can lead to the instability of the slope along the sliding surface. At this stage, the creep on the sliding surface controls the overall deformation of the slope. In general, the slope sliding surface is not entirely consistent with the rock mass structural plane, while the shear creep of the sliding surface is governed by the creep characteristics of the structural plane. Generally, the slope sliding surface is not entirely consistent with the rock mass structural plane, yet the shear creep of the sliding surface remains governed by the creep characteristics of the structural plane.MethodsThis study examined the effects of the angle and spacing between different structural planes and shear planes on the directional shear creep characteristics of the rock mass to thoroughly understand the creep failure mechanism of the slope along the slip surface, particularly the influence of the structural plane on the shear creep mechanical properties of the slip surface. The shear creep tests of the structural plane were designed, and the direct shear test and direct shear creep test under a certain normal force were conducted on the second (JRC=3) structural plane of the Barton standard joint profiles to determine its mechanical and creep mechanical properties. The creep curve obtained from the laboratory test was analyzed and fitted to establish the creep constitutive model and preliminary parameters of the structural plane. Based on numerical analysis, the direct shear creep numerical test was conducted on the horizontal structural plane specimen model, and the final constitutive parameters of the structural plane were obtained. Using these parameters, the numerical analysis of rock mass models under different angles between structural planes and the shear plane, and different spacings of structural planes was conducted.Results and DiscussionThrough these numerical experimentations, the creep mechanical properties, such as failure load and steady-state creep rate before failure, were analyzed. The results indicated that, for the rock mass with a single structural plane, different angles between structural planes and the shear plane significantly influenced the rock mass. When the angle between structural planes and the shear plane was less than 14°, the creep curve of the shear plane was controlled by the structural plane, and the shape of the creep curve resembled that of the structural plane. When the angle between structural planes and the shear plane was greater than 14°, the creep behavior was strongly influenced by the matrix, so the creep curve changed considerably, resembling that of the matrix. Under lower stress, creep did not essentially occur. The steady-state creep rate of the first stage before failure has great significance for practical engineering applications. The steady-state creep rate of the first stage before the failure load was related to the characteristics of the structural plane and the failure load. Under the same failure load, the larger the knot-shear angle, the smaller the shear creep rate. When the knot-shear angle increases, the failure load rises, and the shear creep rate also improves. Compared to the characteristics of the structural plane, the failure load has a greater influence on the shear creep rate. For the rock mass with different structural plane spacings, at the same knot-shear angle, a larger structural plane spacing causes the creep failure load to occur closer to the single structural plane creep curve.ConclusionsWhen the spacing of structural planes was reduced to a certain extent, such as when the spacing was 50 mm, the shear displacement was similar, and the creep curve of the rock mass was closer to that of a single structural plane. When the spacing of structural planes was greater than 50 mm, the shear creep mechanical properties of the rock mass were similar to those of structural planes, and the influence of structural plane spacing on the shear plane can be ignored. The fitting formula between the density of the structural plane on the shear plane and the failure load was obtained. The critical structural density of the overall strength of the rock mass was 40.99%.关键词：Direct shear creep test;numerical analysis;Failure loads;steady creep rate of the previous stage148|422|0更新时间：2026-01-23
-
Acoustic Emission Localization Model of Concrete Material Based on Convolutional Neural Network AI导读
“在水工混凝土健康监测领域，专家构建了基于卷积神经网络的声发射定位模型，实现了快速准确定位损伤源位置，为无损检测提供参考。”摘要：ObjectiveHydraulic concrete produces acoustic emission phenomena due to cracking and other damage. It is important to quickly and accurately locate the damage source based on acoustic emission signals for real-time monitoring of the health status of hydraulic buildings. The traditional iterative localization method is greatly affected by the initial value of iteration and the iteration process. An inappropriate initial value of iteration often leads to unstable or divergent iterations, ultimately resulting in poor localization accuracy. Therefore, the selection of the initial value of iteration is critical in the traditional iterative localization method. In addition, traditional iterative localization methods are heavily influenced by the number of sensors and environmental noise, which reduces localization stability and efficiency. The rapid development of deep learning in recent years provides new approaches for acoustic emission localization. Deep learning demonstrates strong feature extraction capability and generalization ability. In response to the limitations of traditional iterative localization methods, a convolutional neural network-based acoustic emission localization model is constructed, which improves the efficiency, stability, and accuracy of acoustic emission localization to a certain extent.MethodsThe absolute propagation time of acoustic emission was generally difficult to obtain, but the arrival time difference between the sensors contained sufficient information that was utilized to localize the acoustic emission source position. This study constructed a convolutional neural network-based acoustic emission localization model using a cylindrical concrete specimen as the experimental object, with the 3D coordinates of eight sensors and the propagation time difference as input, and the 3D coordinate position of the acoustic emission source as output. In addition, the effect of the number of convolutional layers and the size of the convolutional kernel on the localization accuracy was analyzed, and the optimal convolutional neural network structure was obtained. At the same time, the traditional iterative localization method was compared, and the advantages of the constructed localization model in terms of localization accuracy and efficiency were analyzed to verify the localization performance of the constructed localization model.Results and DiscussionsFor the case of 8 sensors, the optimal number of convolutional layers was 4, and the optimal convolutional kernel size was 3×1. In the X, Y, and Z directions, the root mean square errors (RMSE) of the localization model were 0.865 2, 0.826 6, and 0.722 1 mm respectively, the mean absolute errors (MAE) were 0.532 2, 0.617 3, and 0.473 3 mm respectively, the mean absolute percentage errors (MAPE) were 0.222 0, 0.510 1, and 0.051 0 respectively, and the coefficients of determination (R2) were 0.994 2, 0.993 8, and 0.999 3 respectively, which were close to 1. Most of the localization errors were distributed near 0, which basically conformed to the standard normal distribution. The localization accuracy of the localization model in the depth direction was higher than that in the horizontal direction. Compared to the traditional iterative localization method, the localization efficiency was stable, and it showed clear advantages in processing many localization tasks, while the localization error was reduced by about 5%. The damage location detected by the localization model basically matched the real crack location.ConclusionsThe proposed localization model shows good localization efficiency and accuracy. Compared to the traditional iterative localization method, it is not influenced by the initial iteration value or the iterative process, and it exhibits the advantages of stability and high efficiency. In addition, it maintains stable performance for new data points and shows strong applicability, making it a reliable reference for early warning of damage evolution based on nondestructive testing, with potential application to the damage detection of other materials in the future.关键词：acoustic emission;localization method;convolutional neural network;damage detection293|556|0更新时间：2026-01-23 - “在软土地基处理领域，专家建立了多元排水体复合地基固结解析模型，验证了理论解答的合理性，并分析了固结性状的参数敏感性，为实际工程提供依据。”摘要：ObjectiveThe combined composite foundation with long prefabricated vertical drains (PVDs) and short stone columns demonstrates significant engineering benefits in practical applications by integrating reinforcements with varying drainage capacities. However, its consolidation characteristics remain insufficiently understood, resulting in a lack of strong theoretical support for engineering practices. In this context, the consolidation of the entire foundation is considered in two parts: consolidation within the length of the drainage piles and consolidation beneath the bottom of the drainage piles. Previous research has shown that the primary and secondary consolidation settlements of the soil layer beneath the bottom of the drainage piles are often the main contributors to excessive post‒construction settlement. Therefore, particular attention is required for the consolidation theory of the combined composite foundation with multiple drains under this condition.MethodsTaking the four common layout forms of composite foundations with stone columns-PVDs in engineering into account, the consolidation analytical model for the composite foundation with long PVDs and short stone columns was established, which considered stone columns as the central element and PVDs as the outer boundary. The PVDs of the outer boundary were reasonably simplified into an outer drainage ring using the area equivalent method. The composite foundation with long PVDs and short stone columns was divided into two layers: the upper layer, which was the composite foundation within the length of the stone columns, and the lower layer, which extended from below the bottom of the stone columns to the depth of the composite foundation. A virtual pile was assumed to exist at the bottom of the stone columns, with consolidation parameters identical to the surrounding soil, to ensure continuity conditions of pore pressure and seepage between the upper and lower layers of the foundation. Based on the assumption of equal strain, the smear effect during the installation of multiple drains and the radial bidirectional seepage toward the stone columns and PVDs within the soil were fully considered. The consolidation control equations were derived separately for the upper and lower layers of the composite foundation by incorporating these factors. During the consolidation process, it was assumed that the surface of the foundation was drained, and the bottom was not drained. The classical assumption of equal flow around the drains was applied to the boundary between the stone columns and the soil, as well as the boundary between the PVDs. Utilizing the method of separated variables, linear equations, properties of singular matrices, and the trigonometric orthogonality of pore pressure in the composite foundation, the analytical solutions for pore pressure and average consolidation degree of the composite foundation, soil, stone columns, and PVDs under instantaneous loads were derived for the upper, lower, and overall layers of the composite foundation. In addition, employing the superposition method, the solution of consolidation under k‒level multi-stage instantaneous loading conditions was obtained. Finally, the analytical solution result of the composite foundation with long PVDs and short stone columns was compared to the settlement data of an embankment project in Shanghai. The analytical solution demonstrated strong agreement with the measured settlement data of the project, indicating that the analytical solution of consolidation for the composite foundation with long PVDs and short stone columns accurately predicted settlement with high rationality and reliability.Results and DiscussionsA parametric sensitivity analysis was conducted to investigate consolidation behavior and optimize the methods of this composite foundation. The results showed that quantifying the acceleration effect of the composite foundation with long PVDs and short stone columns compared to traditional composite foundations with long PVDs and short soil-cement piles, the former significantly accelerated the consolidation rate, with a maximum difference of 15.9%. Increasing the penetration ratio effectively enhanced the consolidation rate of the composite foundation, with a more significant impact on the later stages of consolidation. In addition, augmenting the permeability coefficient and compression modulus of stone columns accelerated the early stages of foundation consolidation, and this influence became increasingly pronounced as the penetration ratio increased. Unlike the impact of stone column parameters on consolidation, increasing the number and permeability coefficient of PVDs also accelerated the rate of foundation consolidation, with a stronger effect in the later stages. The smear effect of stone columns has a significant influence on the composite foundation, whereas the smear effect of PVDs exerts a relatively weaker impact on the consolidation rate of the composite foundation due to their smaller size.ConclusionsThe results indicate that the analytical solutions correspond closely with the measured settlement data, establishing a reliable basis for practical engineering applications of the composite foundation under study. In addition, in comparison to traditional composite foundations consisting of long PVDs and short soil-cement piles, the configuration with long PVDs and short stone columns demonstrates clear advantages in the consolidation rate. Therefore, this research provides both significant theoretical insights and practical contributions. The parametric sensitivity analysis reveals that the effects of long stone column parameters and short PVD parameters on the consolidation rate differ considerably. The consolidated solutions and characteristic analyses of the composite foundation with long PVDs and short stone columns can not only provide guidance for related engineering foundations but also act as a reference for optimizing the application of this technology to attain greater engineering benefits.关键词：consolidation;analytical solution;composite foundation;stone column;prefabricated vertical drain (PVD)212|387|0更新时间：2026-01-23
CIVIL ENGINEERING
- “复合材料薄壁带簧结构研究取得进展，专家建立了快速计算方法，为航空航天领域提供解决方案。”摘要：Composite materials possess the advantages of light weight, high specific strength, and high specific stiffness, which have attracted considerable attention in recent years. Thin-walled flexible deployable structures made of composite materials have become one of the research hotspots in aerospace applications. The composite thin-walled tape spring is one of the typical space deployable structures that can serve as supporting components for solar sails, wrapped-rib antennas, and similar systems. The composite thin-walled tape spring stores strain energy during the folding process and uses this energy to drive itself and other components to achieve deployment in orbit. The stiffness properties of the composite laminate determine the mechanical response of the tape spring and affect the stored strain energy. In addition, the folding process of the tape spring is a complex process involving large deformations.ObjectiveThe investigation of the stiffness properties of the composite laminate and the folding behavior of the composite thin-walled tape spring holds significant engineering value. First, the theoretical calculation method of the ABD matrix describing the stiffness properties of composite laminates is derived based on classical laminated plate theory. However, when the layering order and layering angle of the composite lamina are changed, the ABD matrix must be recalculated to obtain the values of each element, which is a time-consuming process. Therefore, to make the process faster and more convenient for engineering applications, a finite element method for obtaining the ABD matrix of composite laminates is proposed.MethodsThe commercial finite element software ABAQUS was utilized to create a representative volume element of the composite laminate, which was modeled as a square structure with two sets of opposing edges. During the model discretization, it was necessary to ensure that the mesh nodes on the opposite edges of the representative volume element corresponded so that periodic boundary conditions can be correctly applied in subsequent steps. Two reference points were then established, and the nodes on the two sets of opposite edges of the representative volume element were correlated with these reference points by imposing the degree of freedom equation. Thus, the displacement relationship between each edge node and the corresponding opposite edge node was constrained through the reference points. The successful implementation of the periodic boundary conditions was achieved by establishing a linear equation on the reference points, ensuring that the structure deformed with identical shapes on the opposite edges, meaning that the overall structure represented by the representative volume element was ideal with an infinite array in space and no boundaries. After modeling, based on the derived ABD matrix formula, six different loads were applied to the reference points of the representative volume elements, corresponding to the six kinds of unit strain loading conditions. The reaction force and reaction moment on the reference points under six different loading conditions were then solved through static analysis. For each loading condition, the obtained reaction forces and moments were utilized to calculate the value of each element in a column of the ABD matrix. Finally, the values of all elements in the ABD matrix were sequentially obtained using homogenization theory. The results of the finite element method were compared to the theoretical results, and the maximum deviation of each element value in the ABD matrix was approximately 0.73%, verifying the correctness of the numerical calculation method developed for obtaining the ABD matrix. This approach can be extended to calculate the ABD matrix of composite laminates with any material parameters, ply sequences, or ply angles. After obtaining the stiffness properties of composite laminates, a numerical simulation of the folding behavior of the composite thin-walled tape spring was performed.Results and DiscussionsThe tape spring structure has a length of 252 mm and a width of 40 mm, and was discretized using the shell element S8R5 in ABAQUS. The material properties of the tape spring were defined using the ABD matrix obtained previously through the developed method. Two reference points were created for kinematic coupling with the left and right edges of the tape spring to define the boundary conditions of the folding process. The rotational degrees of freedom of these reference points were linked to a virtual reference point, and the folding process was achieved by applying a rotational displacement to the virtual reference point to ensure uniform bending of both edges of the tape spring during the folding process. The influence of two implicit solvers (Static General and Dynamic Implicit) and one explicit solver (Dynamic Explicit) in the finite element software ABAQUS on the mechanical response of large deformation folding processes of the composite thin-walled tape spring was investigated. The implicit algorithm approached the equilibrium solution through continuous iteration, which has a very high computational cost for completing one iteration, and often encountered convergence difficulties for strongly nonlinear problems such as large deformations and contacts. The explicit algorithm employed the central difference method and used the result of one incremental step as the initial condition for calculating the next incremental step. This method did not require iteration or a convergence criterion and handled complex contacts and extreme discontinuities more effectively. However, the time increment step of this method needed to be sufficiently small to ensure the accuracy of the numerical simulation results. The change curve of the folding moment and curvature at the center of the composite thin-walled tape spring during the folding process was obtained through numerical simulation analysis under different solvers.ConclusionsThe numerical simulation results indicate that both implicit and explicit solvers can achieve accurate outcomes, and the mechanical responses of the composite thin-walled tape spring correspond well with the experimental results. Although the implicit solution method effectively addresses the numerical calculation challenges of the thin-walled tape spring under large deformation, it requires numerous iterations during the solution process. In contrast, the explicit solution method does not require iteration or convergence criteria, avoiding computational convergence issues caused by nonlinear factors such as large deformation. Therefore, it is more advantageous for numerically simulating the bending and folding processes of composite thin-walled tape springs.关键词：composites;thin-walled structures;ABD matrix;finite element analysis246|453|0更新时间：2026-01-23
- “在建筑材料领域，研究人员通过共沉淀法合成了不同铝硅摩尔比的纳米水化硅铝酸钙悬浮液，探究了其对水泥早期水化性能的影响，为低温下水泥水化提供了新思路。”摘要：ObjectiveNano-hydrated calcium silicate (C‒S‒H) is a new type of nano early strength agent that effectively improves the early mechanical properties of cement-based materials. Currently, more Al-rich supplementary cementitious materials, such as finely ground blast furnace slag (GGBS) and fly ash (FA), are increasingly used as admixtures to reduce cement consumption. Al can enter C‒S‒H to form C‒A‒S‒H gel during the hydration process of cement. This study prepares nano C‒A‒S‒H with different A/S molar ratios and explores its effects on the hydration of Ordinary Portland cement (P·O), Slag cement (P·S), Fly ash cement (P·F), and Portland composite cement (P·C) at low temperature.MethodsNano C‒A‒S‒H solution with different A/S ratios was prepared by the in-situ co-precipitation method using analytical-grade Ca(NO3)2·4H2O, Na2SiO3·9H2O, Al(NO3)3·9H2O, NaOH, and polycarboxylic acid polymer (PCE, solid content 40%). C‒A‒S‒H with a theoretical C/S molar ratio of 1.2 and A/S molar ratios of 0, 0.02, 0.04, 0.06, 0.08, 0.15, and 0.20 were prepared and labeled as S0, S002, S004, S006, S008, S015, and S020, respectively. The stability and particle size of nano C‒A‒S‒H were characterized by visual inspection and a zeta potential analyzer. The initial and final setting times of pastes containing 4% nano C‒A‒S‒H were tested based on the Chinese standard at 10 ℃ (Test methods for water requirement of normal consistency, setting time, and soundness of Portland cements, GB/T1346—2011). The compressive strength of mortars containing 0.04% nano C‒A‒S‒H was determined using a YAW-300 compressive and flexural integrated machine produced by Wuxi Jianyi Instrument. The pastes cured to specific ages were sampled for TG, SEM, and QXRD analyses to evaluate the effects of nano C‒A‒S‒H on the hydration progress of cement. In addition, the influence of optimized nano C‒A‒S‒H suspension on the early hydration of Portland cement, fly ash cement, slag cement, and composite cement was investigated.Results and DiscussionsThe results indicated that the average particle size of nano C‒A‒S‒H increased with the increasing Al/Si ratio. For instance, the average particle sizes of S0 and S002 were 97.4 nm and 181.2 nm, respectively. The C‒A‒S‒H suspension tended to become gelatinous, and the stability of the suspension diminished with the increasing Al/Si ratio. The initial and final setting times were effectively shortened compared to the reference paste. The reduction in both initial and final setting times became more significant with the increase in the A/S ratio. The initial setting times of the cement pastes containing S0, S008, and S020 were reduced by 16.3%, 53.6%, and 35.7%, respectively, compared to the reference paste, while the final setting times were reduced by 19.8%, 44.5%, and 27.5%, respectively. The 8, 16, and 24 h compressive strengths of mortars containing S008 increased by 120.0%, 151.5%, and 170.6%, respectively, compared to the reference mortar. Similarly, the 8, 16, and 24 h compressive strengths of mortars containing S015 increased by 130.0%, 164.5%, and 201.8%, respectively, compared to the reference mortar. However, no significant increase was observed for the 3 and 28 d compressive strengths of the mortars containing nano C‒A‒S‒H. The TG analysis indicated that the content of Ca(OH)2 in the cement paste was 4.19% after 16 h of hydration at 10 ℃, and the contents of Ca(OH)2 reached 5.29% and 5.80% after 16 h of hydration when S006 and S015 were used, increasing by 26.25% and 38.42%, respectively. In addition, the QXRD results indicated that the Ca(OH)2 content of the reference paste was 3.53% after 16 h of hydration, and the Ca(OH)2 content gradually increased to 6.79% with the increasing C‒A‒S‒H content, which was consistent with the TG results. The setting times of P·O, P·F, P·S, and P·C cement pastes mixed with C‒A‒S‒H seeds were effectively shortened. Nano C‒A‒S‒H exhibited the best adaptability to Ordinary Portland cement, and the initial and final setting times of the paste containing S006 were shortened by 26.01% and 25.05%, respectively.ConclusionsThe nano C‒S‒H suspension was a semi-translucent liquid that turned white when Al was incorporated. The particle size increases, and the stability of nano C‒A‒S‒H decreases with an increasing Al/Si ratio. Both nano C‒S‒H and nano C‒A‒S‒H enhanced the early strength of cement at 10 ℃, while the latter’s compressive strength was not significantly affected. The compressive strength initially increased with the rising Al/Si ratio and then decreased, consistent with the observations from SEM images. The nano C‒A‒S‒H seeds exhibited varying promotion effects on Ordinary Portland cement (P·O), Slag cement (P·S), Fly ash cement (P·F), and Portland composite cement (P·C). The degree of promotion was related to the clinker content in the cement.关键词：nano C‒A‒S‒H;Aluminum Silicon Ratio;early hydration;compressive strength of early hydration405|538|0更新时间：2026-01-23
MATERILAL SCIENCE & ENGINEERING
- “最新研究揭示了好氧颗粒污泥在水处理领域的潜力，发现其胞外聚合物对Cd2+的吸附效果显著，为污废水处理提供了新思路。”摘要：ObjectiveDue to the advancement of industry and the improvement of residents' living standards, the urban sewage treatment load continues to increase, making it urgent to develop cost-effective water treatment technologies. In recent years, aerobic granular sludge (AGS) has been increasingly applied in practical projects because of its high stability and good pollutant removal efficiency. This study examines the rapid cultivation of stable AGS, records the changes in appearance and morphology of inoculated activated sludge floc (AS) and AGS, and reveals, at the microscopic level, the reasons why the structure and stability of AGS are superior to those of AS. In addition, this study investigates the adsorption mechanism of Cd2+ by the extracellular polymeric substances (EPS) of AGS from a mechanical perspective and verifies whether AGS‒EPS exhibits a higher pollutant removal capacity.MethodsThe sludge was collected from the secondary sedimentation tank of a wastewater treatment plant in Hefei and was pretreated to obtain cleaner inoculating sludge. It was divided into two portions: one was used for research, and the other was concentrated and placed into the SBR reactor. The process of granular sludge cultivation was documented, including morphological transformations and variations in key indicators. The microbial community structure was analyzed through high-throughput sequencing of 16 s rRNA genes. Changes in EPS content before and after sludge granulation were determined, and the EPS was characterized using Three-Dimensional Excitation Emission Matrix Fluorescence Spectroscopy (3D-EEM) and Fourier Transform Infrared Spectroscopy (FTIR) to analyze the structural differences between AS and AGS. At the mechanical level, kinetic and thermodynamic models were employed to fit the adsorption experiments of AS‒EPS and AGS‒EPS for Cd2+.Results and DiscussionsDuring the domestication of AS into AGS, on the 85th day, the sludge morphology stabilized, and the sludge particle size increased from 31.59 to 442.72 μm. This experiment successfully produced phosphorus-rich granular sludge, with the phosphorus content of the sludge increasing from 25.93 to 86.27 mg/g. The results of 16 s rRNA gene high-throughput sequencing indicated that Chloroflexi served as the initial particle framework of the granular sludge, while Proteobacteria represented the core bacterial phylum responsible for sludge granulation. In addition, Candidatus_Competibacter and Candidatus_Accumulibacter were identified as the dominant bacterial genera involved in the sludge granulation process. Throughout the cultivation period, the EPS content increased from 79.18 mg/g VSS to 133.63 mg/g VSS. The 3D-EEM analysis confirmed that AGS‒EPS contained a higher level of proteins, and the reduction in protein decomposition and solubilization by cells during the domestication process facilitated microbial cell aggregation within the biofilm. FTIR and protein secondary structure analyses revealed that both types of EPS contained alcohol phenols, —OH, —CH, C—O, C—C from sugars, carboxylate groups from uronic acids, N—H from proteins, and alkane-like organic compounds. However, AGS‒EPS exhibited a more diverse composition of polysaccharides and a more stable cell structure. The granulation process altered the adsorption mechanism of AGS toward Cd2+. Adsorption kinetic studies demonstrated that the adsorption of Cd2+ by AS‒EPS involved both physical and chemical adsorption, whereas the adsorption of Cd2+ by AGS‒EPS was better described by the pseudo-second-order kinetic model, indicating stronger biochemical adsorption properties. Adsorption thermodynamic analyses indicated that the adsorption of Cd2+ by both types of EPS was best fitted by the Langmuir model, indicating a homogeneous monolayer adsorption process. The theoretical maximum adsorption capacity of AGS‒EPS was higher than that of AS‒EPS, at 617.09 and 542.90 mg/g, respectively.ConclusionsThe experimental process demonstrated that increasing the COD content in a high-phosphorus environment enhanced the phosphorus removal performance of sludge and positively influenced the cultivation of granular sludge. In the comparison between AGS and AS, 3D-EEM analysis revealed that AGS exhibited a higher protein level, while FTIR indicated that both had similar functional groups; however, AGS contained more abundant polysaccharides. The secondary protein structure confirmed that AGS possessed stronger cell aggregation, greater mechanical strength, and higher protein compactness. The by-product of granular sludge cultivation, AGS‒EPS, exhibited a stronger Cd2+ adsorption capacity, primarily through biochemical interactions, indicating that AGS holds greater potential in wastewater treatment applications.关键词：aerobic granular sludge;extracellular polymers;heavy metal;microbial community;biosorbent351|536|0更新时间：2026-01-23
- “在口腔诊室内，超声波洁牙操作产生的气溶胶迅速扩散，高风速可加速气溶胶排出，降低感染风险。”摘要：ObjectiveVarious respiratory viruses spread through aerosol transmission. Dentists and nurses are at a high risk of infection during ultrasonic scaling procedures because of exposure to patients’ respiratory secretions and the substantial aerosols generated by ultrasonic devices. At present, there is a lack of research examining the effect of different airflow velocities on aerosol generation during dental ultrasonic scaling procedures. This study develops a numerical model to simulate aerosol dynamics during scaling procedures in dental clinics under two distinct airflow velocity conditions.MethodsThe study precisely simulated ultrasonic scaling procedures performed by one dentist and one assistant nurse, based on an actual dental clinic at West China Hospital of Stomatology, Sichuan University. First, a physical model was constructed, and a tetrahedral unstructured mesh was generated using discrete computational domains. ANSYS Fluent 20.0 software facilitated the numerical simulations, employing the Realizable k-ε turbulence model and the Discrete Phase Model (DPM) for transient aerosol calculations. In addition, the Discrete Random Walk (DRW) model was applied to simulate particle dispersion induced by the airflow field, and the evaporation of aerosol particles was also simulated. The numerical model encompassed the initial 30-minute aerosol release procedure and the following 60-minute post-release clearance process. The study comprehensively analyzed the temporal and spatial distribution of aerosols within the dental clinic during ultrasonic scaling under two distinct airflow velocity scenarios (Scenario 1: V = 0.3 m/s and Scenario 2: V = 2.0 m/s). Quantitative analyses of aerosol deposition and distribution patterns across various surfaces within the clinic, particularly those frequently touched by healthcare personnel, were conducted. Finally, the spatial distribution of aerosols generated under the two wind speed scenarios was compared (Scenario 1: V = 0.3 m/s versus Scenario 2: V = 2.0 m/s). The regularity and distribution area of aerosol deposition on multiple surfaces in the consulting room, mainly those frequently touched by medical staff, were quantitatively analyzed, and the infection risks faced by medical staff under both scenarios were discussed.Results and DiscussionsIn the dental clinic, aerosols generated during ultrasonic scaling procedures rapidly spread throughout the clinic space under both airflow velocity conditions. As the procedures progressed (Scenario 1: by the 9th minute; Scenario 2: by the 3rd minute), aerosol concentrations remained consistently high until the treatment ended. However, with a higher airflow velocity at the inlet (Scenario 2: V = 2.0 m/s), aerosol levels inside the clinic significantly decreased, showing approximately 2.46 times lower particle concentration compared to Scenario 1 (V = 0.3 m/s). This reduction was attributed to enhanced aerosol diffusion and accelerated sedimentation within a shorter duration. After the treatment concluded, aerosol clearance required approximately 30.55 minutes in Scenario 1 and 5.55 minutes in Scenario 2. These findings indicated that maintaining relatively high ventilation speeds in dental clinics was advisable during respiratory disease seasons. Both scenarios presented a high risk of exposure of dental care personnel to infectious aerosols during ultrasonic scaling procedures; therefore, strict adherence to protective equipment requirements was necessary. In Scenario 1 (V = 0.3 m/s), aerosol concentrations at the dentist and nurse sampling sites decreased to zero after 30 minutes, whereas in Scenario 2 (V = 2.0 m/s), this occurred within only 6 minutes. Aerosol deposition rates were 78.93% in Scenario 1 and 87.90% in Scenario 2, with 37.55% and 52.67% of aerosols depositing on frequently touched surfaces within the clinic, respectively. The treatment console surfaces, which were most frequently touched by dental care personnel, exhibited higher aerosol deposition rates in both scenarios: 4.65% in Scenario 1 and 8.19% in Scenario 2. Surfaces such as computers and dental chairs showed higher deposition rates under Scenario 2 compared to Scenario 1. Considering that airflow-induced resuspended deposits served as important sources of airborne microbes, final disinfection of these corresponding surfaces after treatments was crucial to prevent cross-infection. These results emphasized the importance of meticulous surface disinfection after dental treatments to mitigate the risk of airborne pathogen transmission in dental clinics.ConclusionsCompared to Scenario 1, Scenario 2 exhibited significantly reduced aerosol concentrations and shorter clearance times, attributed to enhanced airflow dynamics that facilitated aerosol dispersion and sedimentation. Aerosol deposition rates on surfaces such as treatment consoles, computers, and dental chairs were higher in Scenario 2. Different airflow velocities influenced the peak aerosol concentration, the time to reach the peak concentration, and the aerosol clearance duration in the dental clinic. When the airflow velocity increased, the peak aerosol concentration decreased, while both the time to peak concentration and aerosol clearance duration were shortened. Regardless of the scenario adopted to control infection in the clinic, it remained essential for dental staff to maintain appropriate aerosol protection. Higher airflow velocity led to increased deposition of aerosol particles on accessible surfaces. The findings of this study provide valuable insights into how various ventilation schemes and airflow velocities influence aerosol transmission dynamics in dental care environments. These results highlight the critical importance of implementing effective ventilation strategies and rigorous surface disinfection protocols to mitigate the risk of airborne pathogen transmission during dental procedures, contributing to the optimization of infection control measures in similar clinical settings.关键词：dental clinic;aerosol;airborne transmission;numerical simulation;computational fluid dynamics202|465|0更新时间：2026-01-23
ENVIRONMENTAL ENGINEERING
- “报道：提出了一种复合脉波倍增平衡电抗器，有效抑制12脉波整流器输入电流谐波，降低总谐波失真至三分之一，具有成本低、易于实现等优点。”摘要：ObjectiveThe 12-pulse rectifier is widely used as the interface circuit between high-power electrical equipment and the power grid, particularly in applications such as subway traction, ship electric propulsion, and metal smelting. However, the strong nonlinearity of the rectifier diodes causes the 12-pulse rectifier to inject a significant amount of harmonics into the power grid, which seriously contaminates the grid. A hybrid pulse-multiplying inter-phase reactor (HPM-IPR) with pulse-tripling capability is proposed to effectively reduce the harmonics generated by the 12-pulse rectifier, resulting in the development of a 36-pulse autotransformer rectifier based on the HPM-IPR.MethodsFirst, the structure of the proposed HPM-IPR was introduced. It consists of an inter-phase reactor with double secondary windings and two sets of single-phase full-wave rectifier bridges. Then, the topology of the 36-pulse autotransformer rectifier with the HPM-IPR on the DC side was presented. Its main circuit employed a 12-pulse autotransformer rectifier, and the conventional inter-phase reactor in the rectifier was replaced by the HPM-IPR. The primary winding of the HPM-IPR was connected to the positive polarity output of the two three-phase rectifier bridges, while the two sets of single-phase full-wave rectifier bridges on the secondary side of the HPM-IPR were connected in series and parallel with the load, respectively. The winding voltage relationship of the P-type phase-shifting transformer was analyzed, and the winding turn ratio was designed based on the 30° phase shift requirement. Based on the polarity of the HPM-IPR input voltage and its relationship with the output voltage, the four auxiliary diodes in the two sets of single-phase full-wave rectifier bridges were operated alternately, resulting in four operating modes of the rectifier. In all working modes, one auxiliary diode was always connected in series with the load, while in two of the four modes, one auxiliary diode was connected in series and one in parallel with the load. The modulation of the four auxiliary diodes increased the output current level of the rectifier bridge to four, and based on the AC-DC side current relationship, the number of steps in the input current increased from 12 to 36. Thereafter, the functional relationship between the turn ratio of the HPM-IPR and the input current and output voltage of the rectifier was established. Based on the definitions of the total harmonic distortion (THD) of the input current and the ripple coefficient of the output voltage, the relationship surfaces between the input current THD, output voltage ripple coefficient, and the turn ratio of the HPM-IPR were obtained. The analysis results indicated that when the HPM-IPR was designed with the optimal turn ratio, the THD of the input current was reduced to one-third of its original value. The capacity of the HPM-IPR and the voltage and current stresses of the four auxiliary diodes under the optimal turn ratio condition were calculated, and the design methods of the HPM-IPR inductance and the selection criteria for the four auxiliary diodes were determined. The calculation results revealed that the capacity of the HPM-IPR was only 3.7% of the output power. The auxiliary diodes connected in series with the load were subjected to voltage and current stresses of 0.03Ud and Id, respectively, while the auxiliary diodes connected in parallel with the load experienced voltage and current stresses of 2Ud and 0.15Id, respectively. The auxiliary diodes in series with the load were selected as high-current, low-voltage diodes, whereas the auxiliary diodes in parallel with the load were selected as high-voltage, low-current diodes. The efficiency of the rectifier before and after adding the HPM-IPR on the DC side was analyzed, and the efficiency slightly decreased after adding the HPM-IPR.Results and DiscussionsA 1.5 kW experimental prototype was constructed, and the input current THD of the rectifier before and after employing the HPM-IPR was compared and analyzed. The experimental results showed that after implementing the HPM-IPR, the waveform of the input current improved from 12 steps to 36 steps, and the 11th, 13th, 23rd, and 25th harmonics in the input current were almost eliminated. The input current THD decreased from 10.9% to 2.1%, and the input current harmonics were effectively suppressed. The input current THD of the rectifier was tested when the output current was 3 A, 4 A, 5 A, 6 A, and 7 A to clarify the harmonic suppression effect under various load conditions. As the load current increases, the THD of the input current slightly decreases because the filtering effect of the leakage inductance of the phase-shifting transformer becomes stronger, making the input current smoother. After using the HPM-IPR, the input current harmonics of the rectifier were effectively suppressed over a wide load range, and the input current THD was reduced to less than 3%.ConclusionsThe results demonstrate that the proposed HPM-IPR increases the number of steps in the rectifier's input current from 12 to 36 and reduces the input current THD to less than 3% over a wide load range, achieving a nearly sinusoidal input current and effectively suppressing input current harmonics. The HPM-IPR scheme exhibited good harmonic suppression performance. In addition, the equivalent capacity of the HPM-IPR was only 3.7% of the output power and did not require active devices, indicating that the proposed HPM-IPR scheme possesses advantages of small equivalent capacity, ease of implementation, and low cost, along with substantial application value.关键词：multi-pulse rectifier;inter-phase reactor;pulse-multiplying;harmonic suppression;power quality402|754|0更新时间：2026-01-23
- “在电力设备领域，专家基于有限元法建立了换流变阀侧套管模型，探究了振动作用下阀侧套管表带触指的失效行为和规律，为增强其可靠性提供参考和指导。”摘要：ObjectiveThe converter transformer is the core power equipment of the converter and inverter in the UHV DC transmission system, and the valve-side bushing of the converter transformer, as the key accessory equipment for outgoing and current-carrying functions, plays an important role in ensuring the long-term safety and stability of the power system. During the actual operation of the converter transformer, the valve-side bushing is subjected to the combined effects of electricity, heat, mechanical force, and chemical reactions, as well as long-term vibration and other complex mechanical loads. The vibration load of the converter transformer is transmitted to the valve-side bushing through its internal structure, stimulating its vibration. Due to the lack of axial constraints between the strap contacts and the copper conductive rod contact surface, relative cyclic displacement is likely to occur between them, which can cause wear failure of the valve-side bushing strap contacts under long-term operation. Many incidents of wear deterioration in bushing strap contacts have occurred in actual engineering practice; however, there is currently a lack of research on the failure mechanism of valve-side bushing strap contacts under converter transformer vibration.MethodsThe finite element method was applied to investigate the failure behavior and mechanism of valve-side bushing strap contact under vibration. First, a model of the valve-side bushing of the converter transformer was established, and its modal analysis was conducted. Based on the finite element model, the vibration differential equation of the valve-side bushing strap contact was formulated. Then, the direction, frequency, and amplitude of the vibration load were varied, and a surface model was developed to describe the relationship between the cumulative relative displacement of the strap contact, the wear depth, and the vibration amplitude. The relative displacement of the strap contact was obtained for different contact areas of the strap contact and various sizes of the flange stiffener by altering the size of the strap contact and adding a stiffener to the bushing flange.Results and DiscussionsA simulation study was performed to analyze the relative displacement and wear characteristics of the strap contact by changing the vibration characteristics and bushing structure. The results showed that when the strap contact of the valve-side bushing was subjected to axial vibration, it produced a significant relative displacement, and its cumulative relative displacement stroke increased linearly with the amplitude of axial vibration acceleration. Under radial vibration, a considerable contact pressure was generated, and the contact pressure increased linearly with the amplitude of radial acceleration. When subjected to both axial and radial vibrations, the radial vibration increased the contact pressure between the strap contact and the copper guide rod, enhancing the contact surface resistance and reducing the relative displacement between them. The valve-side bushing exhibited a low inherent frequency, making it more sensitive to low-frequency vibration. At a vibration frequency of 100 Hz, the cumulative relative displacement stroke of the bushing strap contact was significantly larger than that under other single-frequency vibrations, while higher vibration frequencies corresponded to shorter cumulative displacement strokes. The contact pressure generated by radial vibration constrained the relative displacement of the strap contact caused by axial vibration, resulting in the wear depth of the strap contact not increasing monotonically with the rise in vibration acceleration amplitude. Changing the size of the strap contact has minimal influence on the relative displacement; however, it is recommended to use larger strap contacts to reduce contact resistance. The installation of a flange stiffener enhances the rigidity of the entire bushing structure, reducing the relative displacement of the strap contacts.ConclusionsThis study addresses the issue of wear and failure of the valve-side bushing strap contact under the vibration conditions of converter transformers. Based on actual engineering faults, a differential equation describing the vibration behavior of the valve-side bushing strap contact was formulated, and the finite element method was applied to analyze the relative displacement and wear depth between the strap contact and the conductive rod under various vibration loads and bushing structures. On this basis, a surface model was developed to represent the cumulative displacement stroke and wear depth of the bushing strap contact in relation to the vibration load amplitude, providing a reference for predicting the service life and wear characteristics of the bushing strap contact under long-term vibration conditions. The obtained research results and proposed improvement measures can enhance the reliability of the strap contact-type electrical connection structure of the valve-side bushing during vibration and are important for extending the service life of the valve-side bushing and ensuring the reliable operation of the converter transformer.关键词：converter transformer;valve-side bushing;long-term vibration;strap contacts;relative displacement wear187|460|0更新时间：2026-01-23
ELECTRICAL ENGINEERING
-
Optimization of Hyperspectral Image Classification Using Spectral-feature Based Approaches 增强出版 AI导读
“在遥感图像处理领域，专家提出了一种改进的基于深度像素对特征的高光谱图像分类方法，有效提高了分类精度和效率。”摘要：ObjectiveHyperspectral imaging technology captures image data across a wide range of wavelengths, providing rich spectral information for each pixel in the scene. This detailed spectral information enables precise identification and classification of various materials and land cover types, making HSIC a critical task in remote sensing. The vast amount of data contained in hyperspectral images presents significant challenges and opportunities for advanced image processing techniques, particularly those involving deep learning. Over the past decade, deep learning has revolutionized numerous fields, including image processing and classification. In the context of HSIC, deep learning techniques, especially those utilizing convolutional neural networks (CNNs) and recurrent neural networks (RNNs), have demonstrated remarkable improvements in performance. These methods effectively capture the complex and high-dimensional nature of hyperspectral data, extracting both spatial and spectral features to enhance classification accuracy. Methods based on spatial-spectral features have gained substantial attention due to their ability to integrate spatial context with spectral information. These methods can better discriminate between different classes by combining these two types of features. However, a limitation arises when applying sample partitioning strategies from natural image classification directly to hyperspectral images. This approach often leads to unintended overlap between training and test samples, particularly in spatially contiguous regions, which can artificially inflate classification accuracy and reduce the model's generalization capability.MethodsThis study proposed a unique approach that reduced the size of input image patches in spatial-spectral-based methods, enabling an effective separation of the training and test sets. When the image patches were reduced to a specific size, their features gradually transformed into pure spectral features. Considering this observation, this study refocused on the HSIC method based on spectral features and proposed an improved method that relied on deep pixel pair features for faster and more robust hyperspectral image classification. The proposed method improved the original pixel pair feature method in two main aspects. The first key improvement introduced was the distance-constrained pixel pair generation method. Traditional pixel pair methods often suffered from inefficiencies and the inclusion of redundant or irrelevant pixel pairs. The proposed method ensured that only meaningful and diverse pixel pairs were selected by incorporating distance constraints, enhancing the training process and the robustness of the model. This method not only efficiently constructed a large-scale pixel pair training set but also eliminated redundant and unreasonable pixel pairs, leading to more accurate and efficient model training. The second major enhancement was the result-reuse voting strategy combined with a classification method. This strategy optimized the classification process by reusing intermediate results, reducing the computational burden, and accelerating the overall classification time. This approach enhanced efficiency while maintaining high classification accuracy by avoiding redundant model calls. The result-reuse voting strategy ensured that the classification process remained both effective and efficient, making it suitable for large-scale hyperspectral datasets.Results and DiscussionsExtensive experiments were conducted on public hyperspectral image datasets to validate the effectiveness of the proposed method. The results demonstrated that the proposed method outperformed existing spatial-spectral feature-based methods in terms of classification accuracy. In addition, compared to standard pixel pair methods, the proposed approach significantly reduced computational time, making it more suitable for practical applications. The experimental results confirmed that the proposed method achieved higher classification accuracy in practical applications and significantly improved time efficiency compared to existing methods.ConclusionsAccordingly, this study addresses a critical challenge in hyperspectral image classification by proposing a novel approach that shifts the emphasis from spatial-spectral features to purely spectral features through the reduction of the input image block size. The introduction of a distance-constrained pixel-pair generation method and a result-reuse voting strategy significantly enhances classification efficiency and robustness. The experimental results validate the superiority of the proposed method, highlighting its potential for broader applications in remote sensing. This research contributes to the development of more accurate and generalizable HSIC models, paving the way for future advancements in hyperspectral image analysis by addressing the overlap issue between training and test samples. The study highlights the importance of considering the unique characteristics of hyperspectral data and provides a robust framework for leveraging spectral features to achieve high-precision classification. The proposed method represents a significant advancement in HSIC, providing a practical and efficient solution to challenges inherent in existing methods. Future research will explore the integration of additional constraints and optimizations to improve the performance and applicability of hyperspectral image classification techniques.关键词：hyperspectral image classification;deep learning;pixel-pair feature;sample division strategy375|436|0更新时间：2026-01-23 - “在状态估计领域，专家提出了一种边学习边估计的卡尔曼状态估计算法，有效提高了估计精度和鲁棒性，为实际工程应用提供新思路。”摘要：ObjectiveKalman filtering (KF), as a widely used state estimation algorithm, plays a crucial role in estimating system state variables. The high-precision Kalman state estimation algorithm requires accurate knowledge of model parameters and noise statistical characteristics. Otherwise, estimation performance significantly degrades, and filter divergence can occur. However, in practical applications, many system model parameters and noise statistical characteristics are often unknown or inaccurate. Therefore, a Q-learning-based Kalman filtering (QL‒KF) algorithm is proposed that learns and estimates simultaneously when model parameters and noise statistical characteristics are unknown.MethodsThe Q-learning policy iteration algorithm, which was divided into two parts, policy improvement and policy evaluation, was employed to address the issue of unknown model information. In the policy improvement stage, a state-action value function (Q function) that evaluated the estimated state value was defined. Then, a formula transformation was utilized to ensure that the estimated value depended only on observed values rather than model parameters, eliminating the need for model parameters. In addition, two adjustable weight matrices were introduced to calculate the Kalman gain, avoiding reliance on the system noise statistical characteristics. Then, an estimation policy for obtaining system state estimates was derived from the Q function. In the policy evaluation stage, the estimation of the Q function was transformed into the estimation of its information matrix, and the recursive least squares algorithm was applied to identify the information matrix. Afterward, based on the identified information matrix, the estimation policy was followed to execute the corresponding actions and update the estimated values of the state variables. Finally, the proposed algorithm was applied to estimate the state of a two-state polynomial system and the water level of a quadruple water tank system to verify the effectiveness and feasibility of the algorithm. In addition, the proposed algorithm was compared to a joint state and parameter estimation algorithm.Results and DiscussionsThe estimation performance of the QL‒KF algorithm was analyzed under conditions of unknown model parameters and noise statistical characteristics. A Monte Carlo experiment was conducted, and 50 Monte Carlo simulations were performed to enhance the credibility of the simulation. Uncertainty was introduced into each parameter to verify the robustness of the proposed algorithm. The root mean square error (RMSE) and the average RMSE (ARMSE) were used as performance evaluation metrics. For the two-state polynomial system, when both the system process noise and measurement noise were Gaussian noise, the simulation results showed that the RMSE of the QL‒KF algorithm exhibited a strong convergence trend, demonstrating the effectiveness of the algorithm. Because the initial estimates were randomly assigned and the Q-learning algorithm required some data accumulation during application, the initial RMSE was slightly larger and fluctuated, but showed a decreasing trend with an increasing number of iterations and gradually stabilized. Compared to the standard KF algorithm, when the model parameters were known in the KF algorithm, the RMSE value of the KF algorithm was low and very stable. However, when the model parameters of both algorithms were unknown, the proposed QL‒KF algorithm achieved significantly better estimation accuracy than the standard KF algorithm, demonstrating stronger robustness. Compared to the EVIU algorithm (joint state and parameter estimation algorithm), the RMSE of the QL‒KF algorithm was smaller and more stable after convergence, with an average ARMSE reduction of 34.66%, indicating higher estimation accuracy. It also demonstrated stronger robustness under parameter uncertainties. In addition, the algorithm required less computational time, reducing the average running time by 44.38%, and exhibited high real-time performance. When both the system process noise and measurement noise were non-Gaussian noise, the simulation results still showed that the RMSE of the QL‒KF algorithm exhibited a convergence trend, confirming the algorithm's effectiveness. When the system parameters were unknown, the estimation error of the proposed algorithm was lower than that of the KF algorithm and similar to that of the EVIU algorithm. The running time of the QL‒KF algorithm was reduced by 45.03% compared to the EVIU algorithm, indicating higher real-time performance. However, compared to the Gaussian noise system, the estimation error of the QL‒KF system increased, indicating that different types of noise affected the estimation accuracy of the proposed algorithm. For the quadruple water tank system, the RMSE of the QL‒KF algorithm for both state components showed favorable trends, demonstrating the effectiveness of the algorithm. Compared to the EVIU algorithm, the proposed algorithm exhibited stronger robustness under parameter uncertainties, with smaller estimation errors, an average ARMSE reduction of 79.93%, and a decrease in running time of 47.78%, indicating good real-time performance.ConclusionsThe findings indicate that the proposed QL‒KF algorithm can utilize only observations, without identifying system parameters, to estimate the internal state of systems when the model parameters and noise statistical characteristics are unknown. The estimation accuracy of the algorithm is influenced by the type of system noise. For Gaussian noise systems, the algorithm demonstrates high estimation accuracy, robust performance, and strong real-time capability. However, for non-Gaussian noise systems, the estimation accuracy decreases. Future work will focus on further improving estimation accuracy.关键词：Kalman state estimation;Q-learning;unknown model parameters;unknown noise statistical characteristics167|437|0更新时间：2026-01-23
COMPUTER SCIENCE AND TECHNOLOGY
- “在聚酰亚胺复合材料领域，研究者通过非接触应变测量技术评估了柔性太阳翼毯面缝纫连接方法的拉伸性能，发现4毫米缝合线迹接头具有最大拉伸载荷，为提高连接性能提供新思路。”摘要：ObjectivePolyimide fiber fabric composite laminate is a typical aerospace flexible film material. Achieving high-quality joining of these film composites remains challenging, and the joining quality has a significant influence on the service performance of spacecraft. This study investigates the sewing connection method of polyimide fiber fabric composite laminate used in the flexible solar wing blanket surface of spacecraft to improve the joining performance of these film composites.MethodsThe sewing test of polyimide fiber fabric composite laminate is designed to evaluate the sewing quality of the joint. Both the tensile failure progression and final failure modes are characterized through quasi-static tensile tests. The surface tensile deformation of the specimen is evaluated using digital image correlation (DIC) measurement technology. In addition, parameter tests of film sewing joints are conducted to examine the influence of sewing parameters on the connection performance, including sewing dislocation and sewing stitches. Finally, the tensile slip performance of the specimen is assessed through tensile slip tests, in which different sewing parameters are also considered.Results and DiscussionsDuring the tests, adhesive layer peeling, surface cracking, and fiber peeling in the sewing area are examined. The tensile load-displacement results show that the tensile failure process of the sewing joint of polyimide fiber fabric composite laminate exhibits a linear growth characteristic in the force‒displacement curve. Both the composite material laminate and the sewing joint remain in an elastic deformation state until the failure strength is reached, leading to multiple failure modes in the joint. The parameter study indicates that the highest average maximum tensile load of the specimen corresponds to a suture stitch length of 4 mm. In addition, the shorter the suture stitch length, the higher the average maximum tensile load value. This behavior likely results from the shorter suture stitches increasing the number of joining holes when the sewing length remains the same. When the specimen is loaded, the load distribution in the suture zone becomes more uniform, and the local maximum stress level at the joining hole decreases, which in turn delays tensile failure in the suture area. From the perspective of two-line dislocation specimens, 50% dislocation yields a larger average maximum tensile load value than 25% dislocation. This is likely because the dislocation alters the local stress distribution around the suture hole and improves the joining interaction of the specimen, leading to changes in the load transfer condition. The tensile loading progression of the specimen shows that, due to the stress concentration carried by the sutures, the specimen surface becomes severely wrinkled. Before tensile failure occurs, the surface of the specimen becomes wrinkled due to local stress concentration at the connecting hole. In addition, the suture zone demonstrates the weakest load-bearing capacity because the transverse tow is carried through the resin in the tensile direction. Therefore, a large number of tows are stretched and peeled off, and the adhesive layer on the surface of the suture area is destroyed. Fiber tow peeling occurs only in the suture area, where stress concentration in the suture joint is the main cause. The tensile failure resistance in other areas remains significantly higher. At the initial loading stage, the local strain at the end of the seam is the largest, and the local surface adhesive layer is destroyed early. As the external load increases, the local strain around the suture stitch further increases, and the glue layer in the suture zone becomes damaged. Then, fiber bundles in the suture zone peel and progressively deteriorate. Eventually, the fiber bundles in the suture area completely peel off, leading to total separation and failure of the suture joint. During the tensile failure process, the failure modes of the specimens under different parameters are similar. However, when the stitches are sutured at 8 mm, all specimens are pulled off at both ends of the suture area, in addition to wrinkles and transverse fiber bundle peeling. The main reason is likely that the stitches are too long, causing the suture holes under the same load conditions to experience greater local loads. In particular, the sutures at both ends of the suture-bearing area are relatively less constrained, resulting in suture pull-off. Further analysis of the displacement of the specimen perpendicular to the surface shows that the specimen exhibits obvious and varying wrinkles during the loading process. At the beginning of loading, no wrinkles appear on the specimen surface. However, as the load increases, distinct wrinkles form, protruding in the shooting direction. Then, the folds intensify, the middle area remains convex in the shooting direction, and the edges become concave backward. Finally, the degree of wrinkling reaches a maximum and is later reduced due to specimen destruction during further loading. The tensile slip test shows that the tensile slip of the specimen under all suture parameters remains less than 0.2 mm, indicating that the suture area exhibits good stiffness. As the number of stretches increases, the mean slip value shows a weak upward trend, and the suture zone demonstrates slight stiffness degradation. After tensile loading, no adhesive layer peeling, surface cracking, or fiber peeling occurs in the sewing area on either side of the specimens.ConclusionsThese results indicate that the sewing connection method demonstrates high reliability and stiffness in joining polyimide fiber fabric composite structures. This study provides valuable insights for the design of aircraft thin-wall composite structure joints.关键词：solar wing;flexible film;polyimide;composites;stitched connections;tensile properties239|341|0更新时间：2026-01-23
- “在粉体介质动轴密封领域，专家提出了新型犁刀混合机动轴密封装置设计思路及方案，验证了其密封性能和使用寿命，为提高动轴密封性能提供了可靠解决方案。”摘要：ObjectiveThere are challenges related to improving manufacturing and assembly accuracy to ensure adequate sealing performance. The sealing device of the plow blade hybrid motor shaft, which primarily relies on gas sealing, exhibits low sealing performance and a short service life.MethodsThis study proposed enhancing manufacturing and assembly accuracy and improving sealing performance. A new design concept and scheme for the plow blade mixing dynamic shaft sealing device were developed by applying the spiral conveying principle and integrating various forms of sealing in the powder medium dynamic shaft sealing device by combining the sealing structure and principles of the plow blade mixing dynamic shaft. Based on the theoretical calculation of the maximum leakage speed of the 2 000 L plow blade mixer material, the design objective aimed to ensure that the conveying speed of the spiral seal for the leaked material exceeded the material's maximum leakage speed, determining the design values of the main parameters of the spiral seal. Then, the theoretical maximum material accumulation in the gap between the spiral seal and the inner wall of the sealing seat was derived from the main parameter design values of the spiral seal. This maximum theoretical material accumulation served as the foundation for ensuring that the volume flow of the sealing gas after passing through the labyrinth seal surpassed the theoretical maximum material accumulation. Therefore, the main design values of the labyrinth seal were obtained. Thereafter, SolidWorks Flow Simulation software was employed to preliminarily verify the design results of the labyrinth seal and to assess whether the volume flow rate of the sealing gas after traversing the labyrinth seal exceeded the theoretical maximum material accumulation. The simulation outcomes confirmed that the labyrinth seal design satisfied the design criteria. Then, experimental research methods were utilized to determine whether the actual residual amount of material in the sealing seat of the dynamic shaft sealing device was less than or equal to the theoretical maximum accumulation amount when the sealing gas was not introduced, and whether the actual sealing gas volume flow rate after labyrinth sealing surpassed the actual residual material amount.Results and discussionsThe results revealed that 1) the continuous sealing time of the new dynamic shaft sealing device was
MECHANICAL ENGINEERING
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0