期刊：
Computers and Geotechnics,2025年177:106814 ISSN：0266-352X
通讯作者：
Jun Shen
作者机构：
State Key Laboratory of Intelligent Geotechnics and Tunnelling, College of Civil and Transportation Engineering, Shenzhen University, Shenzhen, Guangdong 518060, China;Key Laboratory of Coastal Urban Resilient Infrastructures (Shenzhen University), Ministry of Education, Shenzhen 518060, China;[Cong Zhang] School of Civil Engineering, Central South University of Forestry and Technology, Changsha 410004, China;[Xiaohua Bao; Junhong Li; Jun Shen; Xiangsheng Chen; Hongzhi Cui] State Key Laboratory of Intelligent Geotechnics and Tunnelling, College of Civil and Transportation Engineering, Shenzhen University, Shenzhen, Guangdong 518060, China<&wdkj&>Key Laboratory of Coastal Urban Resilient Infrastructures (Shenzhen University), Ministry of Education, Shenzhen 518060, China
通讯机构：
[Jun Shen] S;State Key Laboratory of Intelligent Geotechnics and Tunnelling, College of Civil and Transportation Engineering, Shenzhen University, Shenzhen, Guangdong 518060, China<&wdkj&>Key Laboratory of Coastal Urban Resilient Infrastructures (Shenzhen University), Ministry of Education, Shenzhen 518060, China
关键词：
Multivariate joint distribution model;Marine soft soil;Physical property;Correlation analysis;Vine copula
摘要：
To consider the highly nonlinear and complex buckling behaviour of various sections of cold-formed steel (CFS) built-up columns, experimental and finite element (FE) methods are commonly used for calculating their axial compressive capacity, although these methods are time-consuming and costly. This paper proposes machine learning (ML) methods to overcome the issues of traditional methods for predicting the maximum axial load capacity (MALC) of CFS built-up columns. A total of 3839 samples from more than 33 different types of sections were collected from 43 published papers, including 817 experimental data and 3,022 FE simulation data. A total of 15 characteristic parameters, such as the second moment of area, the radius of gyration, and the polar second moment of area, are considered when nine different ML models are trained. The robustness of these models was compared via the Monte Carlo simulation method, and the predicted results were compared with the calculated results from the current codes. The results show that the extreme gradient boosting (XGB) model has higher accuracy and smaller prediction errors in estimating the MALC. The proportion of data with a relative percentage error less than 10% in the prediction results of the AISI-DSM codes is 45.31%, whereas the XGB model achieves a proportion of 87.57%, and the results calculated from current codes tend to be more conservative, with a large deviation, which needs to be further considered.
作者机构：
[Jiang, Xueliang; Yang, Hui] Guangdong Provincial Key Laboratory of Green Construction and Intelligent Operation & Maintenance for Offshore Infrastructure, Guangzhou Maritime University, Guangzhou, 510725, China;[Wang, Haodong; Wei, Zhenzhen; Jiang, Xueliang; Shen, Bo; Yang, Hui] school of civil engineering, Central South University of Forestry and Technology, Changsha, 410004, China;[Yin, Jun] school of civil engineering, Central South University of Forestry and Technology, Changsha, 410004, China. yinjun0509@163.com
通讯机构：
[Yin, Jun] s;school of civil engineering, Central South University of Forestry and Technology, Changsha, 410004, China.
关键词：
HHT transform;Living stumps slope;Marginal spectrum;Seismic dynamic response;Shaking table test
摘要：
A large-scale shaking table test of a living stump slope with a geometric similarity ratio of 1:7 was designed and completed. The peak acceleration, acceleration amplification factor, and displacement response patterns of living stumps slopes under different types of seismic waves and excitation intensities were obtained. The time-frequency and energy variation characteristics were analyzed using the Hilbert-Huang Transform (HHT). The results showed that: (1) Regardless of the type of seismic wave, the peak acceleration and acceleration amplification factor of the living stumps slope surface are positively correlated with relative height and seismic excitation intensity. When the excitation intensity is ≤ 0.4g, the acceleration amplification effect is more pronounced; when the excitation intensity is > 0.4g, the acceleration amplification effect weakens. (2) Under the action of different seismic waves, the peak displacement of slope surface shows amplification effect along the elevation, and increases with the increase of excitation intensity. In addition, the incremental displacement gradually decreases from the toe to the top of the slope, which is expressed as D2 > D3 > D1 > D4 > D5. The peak displacement at the top of the slope is the greatest, but the incremental displacement is the smallest; the peak displacement at the toe of the slope is the smallest, but the incremental displacement is relatively large. (3) Regardless of the type of seismic wave, living stumps slope shows the characteristics of filtering the low-frequency components of the seismic waves and amplifying their high-frequency components. At the same time, the seismic Hilbert energy gradually accumulates along the elevation. PSHEA and PMSA significantly increase with elevation and excitation intensity, and they reach the maximum at the top of the slope. (4) The seismic Hilbert energy is positively correlated with the relative height and excitation intensity, and reaches the maximum at the top of the slope. With the accumulation of seismic Hilbert energy increases, the dynamic response parameters such as peak acceleration, acceleration amplification coefficient and displacement also increase synchronously, reaching the maximum at the top of the slope. The research conclusions can provide an experimental basis for the seismic design of living stumps slopes.
摘要：
In order to improve the utilization and development of bamboo resources and promote its application in the engineering field, the effects of different lattice cores and processing methods on the bending performance of laminated bamboo sandwich panels were discussed in this study. These structures featured distinct lattice cores, namely triangular lattice, square lattice, and Kagome lattice. The manufacturing process included using laminated bamboo as the raw material and adopting the interlocking method and partition method. Four-point bending tests were carried out on laminated bamboo sandwich panels with different lattice cores, and the bending performance of sandwich panels was discussed. The failure mechanism of sandwich panels under bending load, as well as the variation law of mid-span deflection, bending stiffness and ultimate bearing capacity, are analyzed, and the specific stiffness and specific strength of laminated bamboo sandwich panels with four different core layers were compared. At the same time, a fourpoint bending test model of laminated bamboo sandwich panels was established by using finite element software ABAQUS, and numerical simulation was carried out. Three kinds of sandwich panels with different lattice core processed by the interlocking method all show shear failure during the bending loading, among which the triangular lattice sandwich panel has the best bearing capacity, while the specific strength of the laminated bamboo sandwich panel with triangular lattice is also the largest. The failure mode of the laminated bamboo sandwich panel with square lattice processed by partition method is that the upper layer yields under compression and the lower layer is damaged in tension. Its bearing capacity and specific strength are better than those of the three sandwich panels processed by interlocking method, which are 21.6 % and 43.6 % higher than those of the triangular lattice interlocking sandwich panel respectively. The error between the simulation results and the experimental results is less than 5 %, which has good consistency and can effectively predict the bending performance of the laminated bamboo sandwich panels. The laminated bamboo square lattice sandwich panel processing by partition method has excellent bending performance among the four kinds of sandwich panels, which can better present the lightweight and high-strength advantages of laminated bamboo. The research results can provide an effective structural form and theoretical basis for the application of bamboo in the engineering field.
期刊：
IEEE Transactions on Geoscience and Remote Sensing,2024年62 ISSN：0196-2892
通讯作者：
Gao, Shijuan;Zhu, Jianjun;Fu, HQ
作者机构：
[Gao, Shijuan; Fu, Haiqiang; Zhu, Jianjun; Li, Yi] Cent South Univ, Sch Geosci & Infophys, Changsha 410017, Peoples R China.;[Zhao, Rong] Cent South Univ Forestry & Technol, Sch Civil Engn, Changsha 410018, Peoples R China.
通讯机构：
[Gao, SJ; Fu, HQ ; Zhu, JJ] C;Cent South Univ, Sch Geosci & Infophys, Changsha 410017, Peoples R China.
关键词：
Photonics;Forestry;Surfaces;Noise;Histograms;Background noise;Vegetation mapping;Densely forested areas;gradient-constrained morphological operation (GCMOO);Ice;Cloud;and Land Elevation Satellite-2 (ICESat-2);improved elevation frequency histogram statistics (imEFHS);subcanopy topography
摘要：
The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) has been widely used to obtain high-precision subcanopy topography. However, due to the vegetation cover over densely forested areas, the ground photons are sparse, which makes it difficult to accurately estimate the subcanopy topography over densely forested areas. In this article, we proposed a novel method for retrieving subcanopy topography over densely forested areas from ICESat-2/ATL03 data. First, the proposed method used an improved elevation frequency histogram statistics (imEFHS) method to obtain candidate ground seed photons (GSPs). In densely forested areas, the obtained candidate GSPs are easily misclassified as canopy photons. Therefore, we performed a gradient-constrained morphological operation (GCMOO) to identify erroneous GSPs. Finally, an erroneous GSPs refinement approach was derived to correct erroneous GSPs over densely forested areas. In addition, the subcanopy topography can be presented by the refined GSPs with cubic spline interpolation. ICESat-2/ATL03 data acquired over densely forested areas was selected for testing the proposed method. The results in the given test sites show that the proposed method can extract subcanopy topography accurately, with a root-mean-square error of 1.71 m over densely forested areas. We also compared the retrieved subcanopy topography results with National Aeronautics and Space Administration (NASA) ATL08 terrain samples. We found that the ratio of useful subcanopy topography results (residual <2.5 m) in the forest-intensive area was 87%. However, the ratio of useful ATL08 terrain samples was only 9%, which is lower than that of retrieved subcanopy topography results. In addition, in densely forested areas, the correlation (R-2) between the retrieved subcanopy topography results and the reference high-precision digital terrain model (DTMs) reached 0.93, which is much higher than that of the ATL08 terrain samples (R-2 =0.63 ).
期刊：
INDUSTRIAL CROPS AND PRODUCTS,2024年218 ISSN：0926-6690
通讯作者：
Wang, HQ
作者机构：
[Wu, Songlin; Li, Xianjun] Cent South Univ Forestry & Technol, Sch Mat Sci & Engn, Changsha 410004, Peoples R China.;[Wang, HQ; Zhao, Jinping; Wang, Hanqing; Li, Chengjun; Xu, Zhaoyang] Cent South Univ Forestry & Technol, Sch Civil Engn, Changsha 410004, Peoples R China.;[Zhao, Jinping; Wang, Hanqing; Li, Xianjun; Li, Chengjun] Hunan Engn Res Ctr Full Life Cycle Energy Efficien, Changsha 410004, Peoples R China.
通讯机构：
[Wang, HQ ] C;Cent South Univ Forestry & Technol, Sch Civil Engn, Changsha 410004, Peoples R China.
关键词：
Machine learning;Lignocellulosic biomass;Liquefaction;Bio-polyols yield;Hydroxyl value
摘要：
Amid the escalating demand for alternatives to petroleum resources and the imperative to decrease carbon emissions, there is an increasing interest in transforming lignocellulosic biomass into valuable chemicals. This study utilized machine learning models to analyze the acid-catalyzed liquefaction process of lignocellulosic biomass and to predict and characterize the yield and hydroxyl value (HV) of bio-polyols. A total of 612 yield data samples and 229 HV data samples were collected and analyzed. From these data, four machine learning models were constructed: Random Forest (RF), Gradient Boosting Regression (GBR), Gaussian Process Regression (GPR), and Support Vector Regression (SVR). The Bayesian Optimization Algorithm (BOA) was applied to refine the hyperparameters of the models, thereby enhancing their predictive accuracy. To clarify the effects of input features on predictive outcomes and to understand their interactive dynamics, one-dimensional and twodimensional Accumulated Local Effects (ALE) analyses were performed. The results show that the GBR model was particularly precise in forecasting the yield and HV of bio-polyols, with corresponding test set determination coefficients (R2) of 0.82 and 0.91. The most influential factors on yield were reaction time (15.1 %), lignin content (13.1 %), and glycerol mass (12 %), while for HV, they were glycerol mass (28.9 %), particle size (15.1 %), and liquid-to-solid ratio (11.1 %). The ALE analysis also exposed the intricate interaction mechanisms between feedstock composition and liquefaction conditions. The verification results confirmed that the optimal model displayed exceptional generalization capabilities, with a Mean Absolute Percentage Error (MAPE) of 9.18 %. Utilizing this model, a user-friendly software package has been developed to enable rapid and precise prediction of the lignocellulosic biomass conversion process. This research not only delivers strategies to cut down on experimental costs and time but also offers a novel perspective on the industrial utilization of biomassderived polyols.
期刊：
Indian Geotechnical Journal,2024年:1-10 ISSN：0971-9555
通讯作者：
Liu Sisi
作者机构：
[Zeng Feng; Liu Sisi; Huang Ying] School of Civil Engineering, Central South University of Forestry Science and Technology, Hunan, China;[Wang Shuisheng] China Construction Fourth Engineering Bureau Co., Guangzhou, China;[Lu Yu] Yunnan Tobacco Rebaking Limited Liability Company Shizong Rebaking Factory, Yunnan, China
通讯机构：
[Liu Sisi] S;School of Civil Engineering, Central South University of Forestry Science and Technology, Hunan, China
关键词：
Root mass density (RMD);Flexible reinforcement model;Root-soil composite;Shear strength;Vetiver
摘要：
Vetiver root-soil complexes were utilized, and direct shear tests as well as root weight density (RMD) measurement experiments were conducted on these complexes at the different depths h and growth periods (t), ranging from 1 to 16 months. RMD, different over growth periods and depths, was analyzed for its impact on the shear strength increments. The experimental results demonstrated that the variation rules of the shear strength increments and RMD with respect to the growth periods (t) were essentially identical. With RMD at its core, a meticulous examination of the flexible reinforcement model has been undertaken, combined with an empirical understanding of RMD’s evolution over time (t), we’ve crafted an enhanced flexible reinforcement model. This model includes growth periods (t) and depths (h) as critical parameters, thereby offering a comprehensive depiction of the dynamic impact of plant root systems on soil reinforcement across different stages of development and depths. The established model was employed to calculate and analyze the shear strength increments for the different growth periods (t) and depths (h), which were compared and analyzed against the experimental outcomes. The results indicate that both the measured and model-calculated shear strength increments exhibit a largely consistent of trend over different growth periods (t). The measured shear strength increment is all within the range predicted by the new calculation model, thereby indicating that the model calculation method is reasonable.
摘要：
In order to investigate the dynamic response of embankment slopes supported by wooden frame beams and bamboo anchor rods under train loading, this study conducted model tests on embankment slopes supported by wooden frame beams and bamboo anchor rods and carried out three-dimensional numerical simulations of the slopes. This study focused on analyzing the effects of train loading frequency, the peak value difference, and the peak value of the soil pressure on the embankment slopes. This study also analyzed the horizontal displacement of the slope surface, the internal forces in the support structure, and the slope safety factor. The results indicated the following: (1) The increase in loading frequency from 2 Hz to 3 Hz resulted in a significant increase in dynamic soil pressure, with a smaller increase observed upon further frequency increments. Moreover, increasing the load or peak value difference led to an overall increase in the maximum dynamic soil pressure. (2) Under various loading conditions, the axial force in the top anchor rod was significantly greater than that in the middle anchor rod. Additionally, the axial force in the rod body exhibited a pattern of larger forces near the anchorage end and smaller forces near the anchor head. The location of the maximum bending moment in the anchor rod transitioned from the anchor head to the anchorage end as the slope depth increased. The bending moment of the anchor rod increased with the loading frequency but decreased with an increase in the peak value, showing a minor influence from the upper and lower peak values. (3) With the presence of this support system, the slope safety factor increased by 20.13%. A noticeable reduction in the horizontal displacement of the slope surface was observed, with the greatest reduction in the top slope area, followed by the slope angle.
通讯机构：
[Yu, M ] C;Cent South Univ Forestry & Technol, Hunan Prov Key Lab Engn Rheol, Changsha 410004, Peoples R China.
关键词：
prefabricated semi-rigid concrete beam-column joint;finite element analysis;load-carrying capacity;node ductility
摘要：
This paper introduces a novel type of prefabricated semi-rigid concrete beam-column joint, aiming to examine its load-carrying capacity and seismic performance in comparison with a traditional cast-in-place joint. This study utilized the ABAQUS 2020 software to establish finite element models for both types of joints and conducted finite element analysis under low circumferential reciprocating displacement loads. When comparing the energy dissipation capacity, ductility, ultimate load-carrying capacity, stress mechanism, and damage mode, a comprehensive evaluation of the two types of joints was performed. Furthermore, this study investigated the impacts of various factors such as the axial compression ratio, concrete strength, reinforcement strength, and connector strength on the ultimate load-carrying capacity, ductility, and energy dissipation performance of the joints. Based on the findings, the newly combined joint exhibited a substantial 31.7% increase in its ultimate load-carrying capacity, along with a notable 7.23% enhancement in ductility and an improved energy dissipation capacity when compared with the cast-in-place joint. As a result, it can be concluded that the seismic performance of the new joint surpasses that of cast-in-place joints. Additionally, this study examined the impact of modifying relevant parameters on the seismic performance of the new prefabricated semi-rigid concrete beam-column joint.
关键词：
wood tar rejuvenator;interfacial adhesion mechanism;water damage resistance;molecular dynamics simulation
摘要：
This study utilized molecular dynamics simulation to investigate the adhesion process between wood tar-rejuvenated asphalt and acid/alkaline aggregate. Initially, various indicators including the contact area, cohesion coefficient, and interaction energy were employed to assess the adhesion effect under dry conditions. This revealed the action mechanism of the wood tar-rejuvenator in enhancing the adhesion performance between aged asphalt and aggregate. Subsequently, an asphalt–water–aggregate interface model was developed to simulate the water damage process of the asphalt mixture. This aimed to unveil the damage mechanism of water intrusion on the adhesion performance of the asphalt–aggregate interface and evaluate the water damage resistance of wood tar-rejuvenated asphalt through adhesion energy, stripping work, and the energy ratio. The findings indicate that wood tar-rejuvenated asphalt exhibits favorable adhesion properties with both acid and alkaline aggregates. The addition of wood tar-rejuvenated asphalt increased the interaction energy between aged asphalt and acid and alkali aggregates by 67.75 kJ/mol and 97.3 kJ/mol, respectively. The addition of a wood tar rejuvenator enhances the interaction energy between aged asphalt and aggregate, thereby increasing mutual attraction and enlarging the contact area. The adhesion between asphalt and aggregates hinges on the interaction between asphaltene and aggregates, and the wood tar rejuvenator reduces the diffusion ability of asphaltene in the attractive state of the aggregate, resulting in stable aggregation. Moisture intrusion increased the aggregation distance between asphaltene and aggregate by 14.1% and decreased the degree of aggregation by 24.0%, thereby reducing the interaction energy. The extent of damage caused by water intrusion is linked to the aggregation distance, with greater distances leading to deeper damage. Under wet conditions, the interaction energy of wood tar-rejuvenated asphalt increased by 78.2% in the acidic aggregate system and 98.1% in the basic aggregate system compared with aged asphalt. Meanwhile, wood tar-based rejuvenated asphalt improves the adhesion between aged asphalt and aggregate and reduces the stripping function of asphalt affected by water replacement, which results in the ER value of wood tar-rejuvenated asphalt being higher than that of the original asphalt by 0.12 and 0.22 in the acidic and alkaline environments, respectively, thus showing excellent resistance to water damage. This study provides new criteria for the selection of rejuvenators for waste asphalt, which will help in the future selection of superior rejuvenators for aged asphalt and reduce the possibility of choosing the wrong rejuvenator.
作者机构：
[Cai, Yong; Li, Liupeng; Chen, Haijun; Fan, Xiaoyue] School of Civil Engineering, Central South University, Changsha 410075, China;[Lv, Xiaoyong] School of Civil Engineering, Central South University of Forestry and Technology, Changsha 410004, China
通讯机构：
[Xiaoyong Lv] S;School of Civil Engineering, Central South University of Forestry and Technology, Changsha 410004, China
摘要：
This paper proposes an analytical method to investigate the out-of-plane vibration of thin-walled curved beams under a moving mass. This method comprehensively considers the warping stiffness, shear deformation, and rotatory inertia, as well as the inertia force caused by moving mass and damping. The Fourier sine integral transform and Laplace-Carson integral transform are used to decouple the system of equations for out-of-plane vibration of the thin-walled curved beam under a moving mass. Two dynamic response expressions in two directions involving vertical and torsional responses are easily obtained after using their inverse transforms. Numerical results demonstrate a strong agreement between the outcomes obtained from the proposed expressions and those obtained using numerical methods in existing literature. Furthermore, the effects of variables such as slenderness ratio, the center angle, velocity of the moving mass and eccentricity on the dynamic response of the thin-walled curved beam are discussed. For beams with small cross-sectional shear correction factor k and slenderness ratio, ignoring the shear deformation, rotatory inertia, and warping stiffness of the beam can cause obvious errors for the natural frequency of the beam. Moreover, when the speed and mass of the moving load are large, it is necessary to consider the inertial force caused by the mass for studying the out-of-plane vibration of the thin-walled curved beam. The influence of shear deformation is less affected by the velocity and mass velocity of moving loads, and primarily depends on the structural characteristics of the curved beam itself. For beams of the same length, shear deformation becomes more pronounced with increasing radius of curvature.
摘要：
Since many buildings are not equipped with fresh air systems, many people open windows to allow for natural ventilation while the air conditioner is running. However, prolonged window opening may cause unnecessary energy consumption. Additionally, the diverse ways of window opening lead to significant variations in indoor airflow distribution, consequently impacting ventilation efficiency. Therefore, this study used field experiments and computational fluid dynamics (CFD) to determine the air exchange efficiency and air change rate of singlesided natural ventilation in air-conditioning rooms by fitting the decay curves of tracer gas. CFD simulation was performed using the unsteady Reynolds -averaged Navier-Stokes (RANS) model to examine how window openings affect the room's air distribution, and the computed indoor CO 2 concentration was compared to the corresponding field experiment data. The results show that natural ventilation through the window opening induced constant heat and mass exchange between indoor and outdoor air, which caused extra cooling losses but was globally effective in eliminating indoor contaminants. The air exchange efficiency is influenced to varying degrees by factors such as the air supply register's position, window opening percentage, and window opening position. The air supply register's position exerted the most significant impact at 46.45%. Efficient indooroutdoor air exchange can be achieved by opening windows intermittently. When the window is opened from one side and the opening percentage is 100%, the indoor environment can be improved by opening the window for about 10 min every 39 min for X -direction air supply and about 7 min every 39 min for Y -direction.
