It is crucial to study the axial compression behavior of concrete-filled steel tubular (CFST) columns to ensure the safe operation of engineering structures. The restriction between steel tubular and core concrete in CFSTs is complex and the relationship between geometric and material properties and axial compression behavior is highly nonlinear. These challenges have prompted the use of soft computing methods to predict the ultimate bearing capacity (abbreviated as N_u) under axial compression. Taking the square CFST short column as an example, a mass of experimental data is obtained through axial compression tests. Combined with support vector machine (SVM) and particle swarm optimization (PSO), this paper presents a new method termed PSVM (SVM optimized by PSO) for N_u value prediction. The nonlinear relationship in N_u value prediction is efficiently represented by SVM, and PSO is used to select the model parameters of SVM. The experimental dataset is utilized to verify the reliability of the PSVM model, and the prediction performance of PSVM is compared with that of traditional design methods and other benchmark models. The proposed PSVM model provides a better prediction of the ultimate axial capacity of square CFST short columns. As such, PSVM is an efficient alternative method other than empirical and theoretical formulas.

The velocity pulse contained in near-fault ground motions have a tremendous impact on dam safety. Previous studies have mainly focused on the response of dams under near-fault seismic records without considering the obliquely incident seismic waves. In this study, the structure–soil interaction (SSI) is taken into consideration, and the nonlinear behavior of a conventional concrete roller-compacted concrete (CC-RCC) gravity dam under near-fault pulse records and non-pulse records is investigated with consideration of the obliquely incident P waves. On the basis of the dam site conditions, three groups of near-fault pulse records are chosen, and three corresponding non-pulse records are fitted by their acceleration response spectra. Combining with the viscous-spring artificial boundary, the wave input method is proposed to transform the near-fault seismic records into the equivalent nodal forces at the boundary of the foundation. The concrete damaged plasticity model is used for the nonlinear analysis. The results show that the pulse ground motions are more destructive than the non-pulse motions. The nonlinear behavior of the dam varies with the incidence angle of P waves and generally reaches a maximum at 60° and 75°, the worst damage occurs at the interface between different materials of the dam, and the spatial variation of its damage is very obvious under near-fault seismic records with various incidence angles. Therefore, the effect of the angle of obliquely incident seismic waves and near-fault pulse effect should be considered comprehensively in the seismic analysis of dams.

Time-dependent deformation of surrounding rocks during construction and service directly affects the structural safety of hydraulic tunnels. Specifically, the creep effect on the cross-fault diversion tunnels is particularly significant. In recent years, a variety of research has focused on the impact of creep on operating cross-fault diversion tunnels. However, there is a lack of consideration for the creep phenomenon during the construction-operation period. This study proposes a simulation method of segment linings deformation of cross-fault diversion tunnel combining deformation monitoring, creep constitutive identification, and creep during construction operation. According to the measured time-displacement data of the surrounding rock, the suitable creep constitutive model is selected by a regression-comparison method. Subsequently, the three-dimensional numerical model of the cross-fault tunnel is established, with the tunnel excavation and long-term creep calculation carried out under the creep effect of the surrounding rock. In addition, the intelligent inversion model was adopted to obtain the basic parameters of the surrounding rock. The results show that fault directly affects the rock mass characteristics near the fault. The radius of influence is about one time the tunnel diameter from the fault. The vertical deformation of segment linings of the vault and inverted arch tends to be subsidence when the fault is above the tunnel; however, the trend is inverse when the fault is under the tunnel. This work gives targeted engineering suggestions and provides a scientific basis for designing and constructing diversion tunnels under complicated geological structures.

In this paper, the early drying shrinkage coefficients of different hydraulic cement mortars are calibrated through laboratory experiments for moderate-heat Portland cement (MHPC) and low-heat Portland cement (LHPC). By developing an improved mesoscale modeling approach, a 3D highly detailed simulation of concrete was generated, which incorporates the phases of mortar, aggregates, and interfacial transition zone (ITZ). The simulation result is in good agreement with the concrete early drying shrinkage experiment, exhibiting an error of less than 4.99% after 28 days. Subsequently, the mesoscale model is employed to explain the influence of the ambient humidity, cement type, and aggregate volume ratio on the early drying shrinkage performance of concrete. The results show that the early drying shrinkage coefficient of the LHPC is approximately 82% of the MHPC. Additionally, the depth of ambient humidity influence is about 15 mm from the concrete surface after 28 days. The early drying shrinkage can be controlled by increasing ambient humidity via the LHPC or raising the aggregate volume ratio. The mass-loss rate of concrete decreases as the ambient humidity or aggregate volume ratio increases during the process of drying shrinkage. Furthermore, the results quantify the influence patterns of various factors on drying shrinkage, thereby facilitating their application in assessing the cracking time induced by early drying shrinkage in roller-compacted concrete (RCC) dams. This provides theoretical guidance for crack prevention in concrete structures and aids in developing strategies for the construction of crack-free dams.

Neural radiance fields and neural reflectance fields are novel deep learning methods for generating novel views of 3D scenes from 2D images. To extend the neural scene representation techniques to complex underwater environments, beyond neural reflectance fields underwater (BNU) was proposed, which considers the relighting conditions of on-aboard light sources by using neural reflectance fields, and approximates the attenuation and backscatter effects of water with an additional constant. Because the quality of the neural representation of underwater scenes is critical to downstream tasks such as marine surveying and mapping, the model reliability should be considered and evaluated. However, current neural reflectance models lack the ability of quantifying the uncertainty of underwater scenes that are not directly observed during training, which hinders their widespread use in the field of underwater unmanned autonomous navigation. To address this issue, we introduce an ensemble strategy to BNU that quantifies cognitive uncertainty in color space and unobserved regions with the expectation and variance of RGB values and termination probabilities along the ray. We also employ a regularization method to smooth the density of the underwater neural reflectance model. The effectiveness of the present method is demonstrated in numerical experiments.