Hasil untuk "Engineering geology. Rock mechanics. Soil mechanics. Underground construction"

Jingyue Li, André Storhaug

With the advancement of Agentic AI, researchers are increasingly leveraging autonomous agents to address challenges in software engineering (SE). However, the large language models (LLMs) that underpin these agents often function as black boxes, making it difficult to justify the superiority of Agentic AI approaches over baselines. Furthermore, missing information in the evaluation design description frequently renders the reproduction of results infeasible. To synthesize current evaluation practices for Agentic AI in SE, this study analyzes 18 papers on the topic, published or accepted by ICSE 2026, ICSE 2025, FSE 2025, ASE 2025, and ISSTA 2025. The analysis identifies prevailing approaches and their limitations in evaluating Agentic AI for SE, both in current research and potential future studies. To address these shortcomings, this position paper proposes a set of guidelines and recommendations designed to empower reproducible, explainable, and effective evaluations of Agentic AI in software engineering. In particular, we recommend that Agentic AI researchers make their Thought-Action-Result (TAR) trajectories and LLM interaction data, or summarized versions of these artifacts, publicly accessible. Doing so will enable subsequent studies to more effectively analyze the strengths and weaknesses of different Agentic AI approaches. To demonstrate the feasibility of such comparisons, we present a proof-of-concept case study that illustrates how TAR trajectories can support systematic analysis across approaches.

Yu Lu, Kai Gu, Yuping Zhang et al.

SUN Zhoujing 1, YAN Liming 1, LI Wentao 1, 2 , WANG Jinghao 1, CHEN Yin 1

The traditional reactive magnesium oxide (MgO) carbonation technology is an emerging soil stabilization method. However, since this technique requires CO₂ to be introduced at specific concentrations and pressures, it is challenging to control in practical engineering applications, often leading to uneven solidification, especially for clayey soils. To address this issue, the sodium bicarbonate (NaHCO3) is used as a CO₂ carrier to treat the MgO-stabilized clay by an indirect carbonation method. The UCS, moisture content and XRD tests are conducted to investigate the effects of MgO and NaHCO3 ratios, curing age, temperature and acidic conditions on the mechanical properties and expansibility of clay reinforced by the indirect carbonation. The results indicate that the UCS initially increases and then decreases with the addition of NaHCO3, reaching its peak at an MgO: NaHCO3 ratio of 2:1. With extended curing time and elevated temperatures, the strength generally increases. In terms of strength and anti-expansion performance, the MgO: NaHCO3 ratio of 2:1 outperforms direct carbonation-stabilized soil. The XRD results show that the UCS of carbonated soil is mainly provided by hydromagnesite, or dypingite.

Kaistrenko, Victor M., Willis, Pascal, Razjigaeva, Nadezhda G. et al.

On October 4(5), 1994, one of the strongest tsunamigenic earthquakes in the history of the Southern Kuril Islands occurred, which was accompanied by large-scale tectonic manifestations. Intense shaking led to the collapse of the entire infrastructure on Shikotan and Kunashir Islands; 11 people died. The tsunami that followed the earthquake reached a height of about 10 m on the ocean side of Shikotan Island and was recorded along the entire coast of the Pacific Ocean. Generally, the study of the manifestations of this event had been carried out since October 1994 and ended in 1997 with the publication of extensive reviews. For the first time, the data on the specific effects on the ocean surface (seaquake) that accompanied this earthquake have been collected and systematized. In the following years, the interest in this extraordinary event initiated further research. At the same time, the study of the deposits of the tsunami of October 4(5), 1994, in the general context of study of ancient tsunami (paleotsunami) deposits in this region came to the fore. These studies allowed us to obtain estimates of the "preservation"/ "erasure" of the traces of this tsunami and more ancient events on the coast of the Southern Kuril Islands. Some important data presented in this review are being published for the first time.

ZHOU Pengfa 1 , SHEN Yusheng 1, 2, GAO Deng 1, HUANG Haifeng 1, ZHANG Xi 1, GAO Bo 1, TIE Mingliang 3

To meet the current demand for anti-seismic testing technology in underground engineering in China, a dynamic model container with hinged sidewalls for fault-crossing tunnels has been developed. The structures and operational principles of the model container are introduced, and a concentrated-mass model for soils, hinged sidewalls and springs is established. Using the 2-norm deviation of acceleration response as the criterion, the influences of spring stiffness on the boundary effects of the model container are analyzed. This analysis provides a process for determining the stiffness of the springs. Finally, large-scale shaking table model tests are conducted using the model container to investigate the impact range of fault interfaces on the tunnel structures under strong seismic actions. The results reveal that the boundary effects of the model container gradually diminish with an increase in the damping ratio of the model soils. With the increase in the spring stiffness, the 2-norm deviation between the model soil response and the free-field soil response, as well as the 2-norm deviation between the hinged sidewall response and the model soil response, initially decreases and then increases. The stiffness of boundary springs should be determined based on the minimum 2-norm deviation between the model soil response and the free-field soil response, following the principle of "flexible but not rigid". The results of the shaking table tests indicate that the model system can well reproduce the zonal impact characteristics of the tunnel near the fault interface. By analyzing the overall bending moment and axial force distribution characteristics along the longitudinal direction of the tunnel, the impact range of the fault interface on the tunnel is determined to be 4B (hanging wall) and 5B (fracture zone). The findings enrich the seismic testing technology for the tunnels in high-intensity seismic zones in China, providing valuable references for shaking table tests in underground engineering.

Hui Yu, Jiangping Hu, Shi-Xin Zhang

The search for non-ergodic mechanisms in quantum many-body systems has become a frontier area of research in non-equilibrium physics. In this Letter, we introduce Hilbert subspace imprint (HSI)-a novel mechanism that enables evasion of thermalization and bridges the gap between quantum many-body scars (QMBS) and Hilbert space fragmentation (HSF). HSI manifests when initial states overlap exclusively with a polynomial scaling (with system size) set of eigenstates. We demonstrate this phenomenon through two distinct approaches: weak symmetry breaking and initial state engineering. In the former case, we observe that ferromagnetic states including those with a single spin-flip display non-thermal behavior under weak U(1) breaking, while antiferromagnetic states thermalize. In contrast, the Z2-symmetric model shows thermalization for both ferromagnetic and antiferromagnetic states. In the latter case, we engineer the initial state prepared by shallow quantum circuits that enhance the overlap with the small target subspace. Our results establish HSI as a mechanism equally fundamental to non-thermalization as QMBS and HSF.

Leonardi Melo, Luís Gustavo, Dimmy Magalhães et al.

This study presents a comparative analysis of three U-Net-based architectures for semantic segmentation of rock art petroglyphs from Brazilian archaeological sites. The investigated architectures were: (1) BEGL-UNet with Border-Enhanced Gaussian Loss function; (2) Attention-Residual BEGL-UNet, incorporating residual blocks and gated attention mechanisms; and (3) Spatial Channel Attention BEGL-UNet, which employs spatial-channel attention modules based on Convolutional Block Attention Module. All implementations employed the BEGL loss function combining binary cross-entropy with Gaussian edge enhancement. Experiments were conducted on images from the Poço da Bebidinha Archaeological Complex, Piauí, Brazil, using 5-fold cross-validation. Among the architectures, Attention-Residual BEGL-UNet achieved the best overall performance with Dice Score of 0.710, validation loss of 0.067, and highest recall of 0.854. Spatial Channel Attention BEGL-UNet obtained comparable performance with DSC of 0.707 and recall of 0.857. The baseline BEGL-UNet registered DSC of 0.690. These results demonstrate the effectiveness of attention mechanisms for archaeological heritage digital preservation, with Dice Score improvements of 2.5-2.9% over the baseline.

Jin Cui, Yifei Zou, Siyuan Zhang

China's Chang'e 5 mission has been a remarkable success, with the chang'e 5 lander traveling on the Oceanus Procellarum to collect images of the lunar surface. Over the past half century, people have brought back some lunar rock samples, but its quantity does not meet the need for research. Under current circumstances, people still mainly rely on the analysis of rocks on the lunar surface through the detection of lunar rover. The Oceanus Procellarum, chosen by Chang'e 5 mission, contains various kind of rock species. Therefore, we first applied to the National Astronomical Observatories of the China under the Chinese Academy of Sciences for the Navigation and Terrain Camera (NaTeCam) of the lunar surface image, and established a lunar surface rock image data set CE5ROCK. The data set contains 100 images, which randomly divided into training, validation and test set. Experimental results show that the identification accuracy testing on convolutional neural network (CNN) models like AlexNet or MobileNet is about to 40.0%. In order to make full use of the global information in Moon images, this paper proposes the MRNet (MoonRockNet) network architecture. The encoding structure of the network uses VGG16 for feature extraction, and the decoding part adds dilated convolution and commonly used U-Net structure on the original VGG16 decoding structure, which is more conducive to identify more refined but more sparsely distributed types of lunar rocks. We have conducted extensive experiments on the established CE5ROCK data set, and the experimental results show that MRNet can achieve more accurate rock type identification, and outperform other existing mainstream algorithms in the identification performance.

Caner Erden, Alparslan Serhat Demir, Abdullah Hulusi Kokcam et al.

Soil compaction is critical in construction engineering to ensure the stability of structures like road embankments and earth dams. Traditional methods for determining optimum moisture content (OMC) and maximum dry density (MDD) involve labor-intensive laboratory experiments, and empirical regression models have limited applicability and accuracy across diverse soil types. In recent years, artificial intelligence (AI) and machine learning (ML) techniques have emerged as alternatives for predicting these compaction parameters. However, ML models often struggle with prediction accuracy and generalizability, particularly with heterogeneous datasets representing various soil types. This study proposes an automated machine learning (AutoML) approach to predict OMC and MDD. AutoML automates algorithm selection and hyperparameter optimization, potentially improving accuracy and scalability. Through extensive experimentation, the study found that the Extreme Gradient Boosting (XGBoost) algorithm provided the best performance, achieving R-squared values of 80.4% for MDD and 89.1% for OMC on a separate dataset. These results demonstrate the effectiveness of AutoML in predicting compaction parameters across different soil types. The study also highlights the importance of heterogeneous datasets in improving the generalization and performance of ML models. Ultimately, this research contributes to more efficient and reliable construction practices by enhancing the prediction of soil compaction parameters.

Luca Cocconi, Michalis Chatzittofi, Ramin Golestanian

The study of active solids offers a window into the mechanics and thermodynamics of dense living matter. A key aspect of the non-equilibrium dynamics of such active systems is a mechanistic description of how the underlying mechano-chemical couplings arise, which cannot be resolved in models that are phenomenologically constructed. Here, we follow a bottom-up theoretical approach to develop a thermodynamically consistent active solid (TCAS) model, and uncover a non-trivial cross-talk that naturally ensues between mechanical response and dissipation. In particular, we show that dissipation reaches a maximum at finite stresses, while it is inhibited under large stresses, effectively reverting the system to a passive state. Our findings establish a generic mechanism plausibly responsible for the non-monotonic behaviour observed in recent experimental measurements of entropy production rate in an actomyosin material and enzymatic activity in crowded condensates.

Cuimei XIA, Nan WANG, Jingyu LIU et al.

Objective Black carbon is a carbonaceous mixture formed by the incomplete combustion of biomass and fossil fuels. Since black carbon has a high carbon content and an aromatized molecular structure, it can persist in the natural environment, suggesting that black carbon is an important component of the global carbon cycle. Although pioneering studies have advanced the understanding of nature, circulation fluxes, and reserves of black carbon in the geosphere, it is still necessary to further clarify the source-sink processes of black carbon and explore the importance of black carbon in the global carbon cycle. Methods Here, this study reviews the available data on pathways, fluxes, and time scales of the global black carbon cycle and analyses and summarizes the migration mechanisms and constraints in the source-sink process of marine black carbon. Results This paper concludes that marine sediments serve as a large reservoir for black carbon, with a reserve of 569-1 380 Pg. Black carbon can persist over millennia in marine sediments, which is associated with a long turnover time as an "effective carbon sink" in the context of "carbon neutrality". Moreover, black carbon exerts a negative feedback effect on global warming due to its longer turnover cycle time than biomass carbon. The deposition of black carbon into the sea is an effective means to enhance carbon sequestration in the ocean and could be an effective way to implement the carbon capture, utilization, and storage (CCUS) strategy. Conclusion In this paper, we propose that human-induced production of black carbon from biogenic combustion may be actively regulated and that the deposition pathways and burial areas of black carbon into the sea should also be optimized given the premise of ecological and environmental friendliness based on a summary and assessment of the strategic value and climate significance of black carbon in marine sediments in the context of "carbon neutrality". Therefore, the positive role of black carbon in "negative ocean carbon emissions" can provide a basic theoretical framework for achieving the goals of "carbon neutrality".

Yuanhua XIE, Guowei ZHANG, Zongwei CAO et al.

Objective Surface displacement and crack development are the most intuitive manifestations of landslide deformation and evolution. Therefore, revealing the characteristics and evolution stages of landslide deformation and failure through displacement monitoring and surface crack survey is highly practical. Methods In this study, the Baishuihe landslide in the Three Gorges Reservoir Area is selected as the research case, and monitoring data from 2003-2016 for this landslide are systematically collected and analyzed, taking monitoring points ZG93 and ZG118 as example. First, the fractal parameter Hurst index R/S analysis is carried out for the displacement data. Second, the multiple fractal dimensions of the displacement of the two monitoring points are calculated every year. Finally, the theory of stage matching of landslide cracks is introduced, and the crack fractal dimension of the Baishuihe landslide is determinated. Results The results show that there is a certain positive correlation between the landslide displacement sequences, and the fractal dimension H index of the two monitoring points is greater than 0.5. The evolution process of the Baishuihe landslide can be divided into a steady deformation state (from June 2003 to June 2007), an accelerated deformation state (from June 2007 to December 2009), and a steady deformation state (from December 2009 to December 2015). According to the matching law of crack stages, the development of the landslide is the initial deformation stage. Conclusion The research results have a certain guiding significance for revealing the evolutionary state of reservoir landslides.

Hanane Mebarki, Noureddine Maref, Mohammed El-Amine Dris

Modelling the hydrological balance in semi-arid zones is essential for effective water resource management, encompassing both surface water and groundwater. This study aims to model the monthly hydrological water cycle in the Wadi Mina upstream watershed (northwest Algeria) by applying the Soil and Water Assessment Tool (SWAT) hydrological model. SWAT modelling integrates spatial data such as the Digital Elevation Model (DEM), land use, soil types and various meteorological parameters including precipitation, maximum and minimum temperatures, relative humidity, solar radiation and wind speed. The SWAT model was calibrated and validated using data from January 2012 to December 2014, with a calibration period from January 2012 to August 2013 and a validation period from September 2013 to December 2014. Sensitivity and parameter calibration were conducted using the SWAT-SA program, and model performance evaluation relied on comparing the observed discharge at the outlet of the basin with model-simulated discharge, assessed through statistical coefficients including Nash-Sutcliffe Efficiency (NSE), coefficient of determination (R2) and Percent Bias (PBAIS). Calibration results indicated favourable objective function values (NSE=0.79, R2=0.93, PBAIS= −8.53%), although a slight decrease was observed during validation (NSE=0.69, R2=0.86, and PBAIS= −11.41%). The application of the SWAT model to the Wadi Mina upstream watershed highlighted its utility in simulating the spatial distribution of different components of the hydrological balance in this basin. The SWAT model revealed that approximately 71% of the precipitation in the basin evaporates, while only 29% contributes to surface runoff or infiltration into the soil.

FAN Xu 1, 2, WANG Yongzhi 1, 2, LIANG Xiaocong 3, 4, CHEN Pingshan 3, 4, WANG Tiqiang 1, 2

Change of clogging and drainage performances of stone columns under many earthquakes is one of the focus issues in construction of artificial dredging island reef and port engineering. The dynamic centrifugal model tests on the composite foundation with stone columns in coral sand are carried out, and the change of clogging and drainage performances of stone columns is discussed through the image observation of the excavation profile of the stone columns, the grain-size distribution tests on the stone columns before and after earthquakes, and the change analysis of the dissipation rate of pore water pressure. The results show that liquefaction occurs at the buried point of 1.25 m, the pore pressure ratio reaches 1.0, and no liquefaction occurs at the buried point of 7.5 m under the seven times of intense dynamic loading. After the tests, no obvious fine particle infiltration is found in the sections of the 9 stone columns excavated from top to bottom, and the grain-size distribution curves of the stone columns tested before and after earthquakes are basically consistent, indicating that clogging pores do not appear in the stone columns. At the buried point of 1.25 m, the pore pressure dissipation rates are respectively 0.027 s-1 and 0.029 s-1 under the first and sixth times of intense dynamic loading, which proves that the drainage performance of the stone columns has not changed significantly.

Biao Shi, Feng Wu, Shuxin Li et al.

On the southeastern margin of the Ordos Basin, the mineral composition of marine-continent transitional facies deposits is complex. The shale, sandstone, coal, and related lithofacies frequently interact, and the lithology changes rapidly in the vertical direction. Due to the low resolution of conventional logging method and borehole enlargement which is a common while drilling, the commonly used methods for identification of lithology including high-quality shale which is prevailing in marine shale gas evaluation are less effective for the study area. First, deconvolution technology was used to improve the resolution of natural gamma rays, gamma rays without uranium and uranium logging curves. Then, a log cross-plot was used to identify the lithology including shale of marine-continent transitional facies, and the uranium-spontaneous potential curve overlap method was proposed to identify high-quality shale from marine-continent transitional facies. The results show that the deconvolution method can effectively improve the vertical resolution of natural gamma rays, gamma rays without uranium and uranium logging curves. The cross-plot of natural gamma-density logging data has a better effect on identifying the lithology of the marine-continent transitional facies, and the cross-plot of uranium logging data and gamma ray data without uranium can further identify three types of shale lithofacies (calcareous siliceous shale, siliceous clay shale and clay shale). In the marine-continent transitional facies, the newly proposed uranium-spontanous potential overlap method is better than the traditional ΔlogR method inidentifying high-quality shale. This research can provide theoretical support for reservoir evaluation of marine-continent transitional shale gas and improve the accuracy of high-quality shale identification.

Junsen Xiao, Kenta Tozato, Shuji Moriguchi et al.

Granular flow is affected by multiple parameters, which makes its study challenging. The discrete element method (DEM) is widely employed to simulate granular flow in consideration of particle motion, particularly when the effects of related parameters and particle shape on flow characteristics are being studied. In this study, different combinations of four input parameters (spring coefficient, friction angle between elements, coefficient of restitution, and bottom friction) were first obtained with the help of the Latin hypercube sampling method. Then, a series of simulations were performed using DEM with different sets of input parameters in consideration of different particle shapes and contact models. Radial basis function (RBF) interpolation was then employed to construct a response surface of run-out distance. Monte Carlo simulations were also conducted to obtain the contribution ratio of each input parameter. The result revealed that the bottom friction has a significant influence on the run-out distance, while friction angle between elements and spring coefficient account for a small proportion in the contribution ratio. Moreover, it was confirmed that the coefficient of restitution has a considerable contribution ratio in the front part of elements. The results also revealed that the influence of the particle shape and contact model on the contribution ratio was not as important in comparison.

Mahmoud Sesa, Hagen Holthusen, Lukas Lamm et al.

The development of tissue-engineered cardiovascular implants can improve the lives of large segments of our society who suffer from cardiovascular diseases. Regenerative tissues are fabricated using a process called tissue maturation. Furthermore, it is highly challenging to produce cardiovascular regenerative implants with sufficient mechanical strength to withstand the loading conditions within the human body. Therefore, biohybrid implants for which the regenerative tissue is reinforced by standard reinforcement material (e.g. textile or 3d printed scaffold) can be an interesting solution. In silico models can significantly contribute to characterizing, designing, and optimizing biohybrid implants. The first step towards this goal is to develop a computational model for the maturation process of tissue-engineered implants. This paper focuses on the mechanical modeling of textile-reinforced tissue-engineered cardiovascular implants. First, we propose an energy-based approach to compute the collagen evolution during the maturation process. Then, we apply the concept of structural tensors to model the anisotropic behavior of the extracellular matrix and the textile scaffold. Next, the newly developed material model is embedded into a special solid-shell finite element formulation with reduced integration. Finally, we use our framework to compute two structural problems: a pressurized shell construct and a tubular-shaped heart valve. The results show the ability of the model to predict collagen growth in response to the boundary conditions applied during the maturation process. Consequently, we can predict the implant's mechanical response, such as the deformation and stresses of the implant.

Ergin Isleyen, Sebnem Duzgun, R. McKell Carter

TONG Yan-mei, ZHANG Hu-yuan, ZHOU Guang-ping et al.

During the operation of underground repository of high-level radioactive waste, the highly alkaline solution generated by groundwater corroding lining concrete will diffuse into the bentonite buffer barrier, resulting in the degradation of barrier property. The KOH solution with different pH values was used to simulate the alkaline solution, and a one-year contact diffusion test at room temperature was conducted. Then, X-ray fluorescence spectroscopy (XRF), X-ray diffraction (XRD), scanning electron microscope (SEM) and energy dispersive spectrometry (EDS) were performed to investigate the effect of slow diffusion on the mineralogy of bentonite. The XRF test results show that when the pH of KOH solution was greater than 12.6, the content of silicon began to decrease, that is, montmorillonite, quartz, cristobalite and other Si-containing minerals in bentonite were dissolved. At the same time, the content of K increased, indicating that the bentonite had an ion exchange reaction with the alkaline solution, a large amount of K+ ions in the solution entered the montmorillonite crystal layer. The XRD test results show that the 001 peak of montmorillonite mineral started to shift to the right at pH=12.6, the peak widths widened, and the peak intensity reduced considerably. When the pH>13, the crystal interlayer space decreased from 1.385 3 nm (13.853 Å) to 1.221 0 nm (12.210 Å), indicating that the crystal layer of montmorillonite was compressed. With the increase of the pH value, the content of minerals such as montmorillonite and quartz decreased significantly, and the contents of illite, clinoptilolite and feldspar minerals increased slightly. The SEM test results show that part of the montmorillonite crystal layer overlapped with the pH of the solution increasing, and then some cracks and holes were generated. As a result, the cracks can accelerate the dissolution of montmorillonite. In the one-year contact diffusion test, the diffusion depth of the KOH solution with pH=13.8 exceeded 7.5 mm, and the newly formed illite crystallites were observed on the contact surface between the alkaline solution and the bentonite. It is confirmed that the strong alkaline solution will cause montmorillonite dissolution and illitization.

Lifeng Xu, Qi Li, Yongsheng Tan et al.

It is important to understand the process of multiphase carbon dioxide (CO2) leakage in faults for the risk assessment of carbon capture and storage (CCS). To quantitatively characterize the CO2 leakage process in the fault, pressure sensors, fiber Bragg grating (FBG) temperature and strain sensors were simultaneously used to monitor CO2 leakage in the fault. Ten experiments were carried out, including five groups of gaseous CO2 leakage tests with initial pressures of 1–5 MPa and five groups of liquid CO2 leakage tests with initial pressures of 6–10 MPa. The results indicate that when liquid CO2 leaked with an initial pressure of 7–10 MPa, the pressure and temperature of CO2 dropped rapidly in the first few seconds and then remained unchanged. The behavior that CO2 continues to leak while maintaining temperature and pressure unchanged is defined as “temporary pseudo-sealing (TPS)” behavior, which continues for the first 1/3 of the leakage period. However, this TPS behavior did not occur in gaseous CO2 leakage. If only the pressure and temperature data were used to evaluate whether CO2 leakage occurred, we would misjudge the risk of leakage in CCS projects during the TPS period. The causes and conditions of TPS behavior were further studied experimentally. The results show that: (1) TPS behavior is caused by the phase transition energy generated when liquid CO2 leaks. (2) The condition for TPS behavior is a small leak aperture (0.2 mm). Only a small leakage rate can make the phase transition energy and pressure change from a dynamic equilibrium, and (3) The compression zone caused by the Bernoulli effect and fault “barrier” could reduce the CO2 leakage rate and further promote the occurrence of TPS behavior. This study provides technical and theoretical support for the quantitative characterization of the CO2 leakage process in faults of CCS projects.