The special large-slope V-shaped structure of underwater tunnels changes the ventilation characteristics during tunnel construction, making the traditional experience limited. Therefore, it is urgent to study the influence of the special structure on the safety of the air environment during construction. In this paper, a series of small-scale experiments were conducted to investigate the ventilation characteristics of V-shaped tunnels. The coupled effects of ventilation parameters (distance of duct outlet from working face $L_0$, air velocity at the duct outlet $u_0$) and structural characteristics (digging length $L_d$, slope of the uphill section $\theta$) were considered. The extreme slope of the V-shaped tunnel of 8% was considered. The flow field and pollutant transport law were determined by using CO as a tracer in the experiments. The results show that $u_0$ has a positive impact on the air return velocity, while $L_d$ has a negative impact, and neither of the other two factors has a significant effect. The transport characteristics of CO in V-shaped tunnels differ from those in flat tunnels, with the former tending to cause unconventional areas of high pollutant concentrations in the horizontal sections. Furthermore, the correlations between CO concentration and distance, ventilation time, and the influence factors discussed in this paper are derived from the experimental results. The conclusions provide guidance for the construction of V-shaped tunnels to prevent air pollution in the construction environment and to improve the working conditions of laborers. Additionally, it can also enrich the ventilation experience in tunnel construction.

Water leakage defects often occur in underground structures, leading to accelerated structural aging and threatening structural safety. Leakage identification can detect early diseases of underground structures and provide important guidance for reinforcement and maintenance. Deep learning-based computer vision methods have been rapidly developed and widely used in many fields. However, establishing a deep learning model for underground structure leakage identification usually requires a lot of training data on leakage defects, which is very expensive. To overcome the data shortage, a deep neural network method for leakage identification is developed based on transfer learning in this paper. For comparison, four famous classification models, including VGG16, AlexNet, SqueezeNet, and ResNet18, are constructed. To train the classification models, a transfer learning strategy is developed, and a dataset of underground structure leakage is created. Finally, the classification performance on the leakage dataset of different deep learning models is comparatively studied under different sizes of training data. The results showed that the VGG16, AlexNet, and SqueezeNet models with transfer learning can overall provide higher and more stable classification performance on the leakage dataset than those without transfer learning. The ResNet18 model with transfer learning can overall provide a similar value of classification performance on the leakage dataset than that without transfer learning, but its classification performance is more stable than that without transfer learning. In addition, the SqueezeNet model obtains an overall higher and more stable performance than the comparative models on the leakage dataset for all classification metrics.

Geopolymer-based high-performance concrete (GHPC) stairs have been proposed as an alternative solution that fulfills both lightweight and load-bearing requirements for prefabricated stairs. Nevertheless, the fire resistance properties of GHPC stairs remain uncertain. Therefore, this study first conducts high-temperature experimental research on GHPC test blocks. The experimental results demonstrated that the GHPC test blocks exhibited no signs of bursting at elevated temperatures and displayed less degradation in the compressive strength compared to that of ordinary concrete. Subsequently, GHPC-Z stair specimens with ribbed and unribbed configurations were fabricated to conduct a fire resistance test at elevated temperatures, followed by a comparative analysis of the resulting damage. The test results indicate that both types of GHPC stairs demonstrate exceptional performance without exhibiting bursting or noticeable cracks or structural spalling marks on their main bodies, despite experiencing vertical deformation. Furthermore, based on these findings, finite element models were established to simulate the fire-induced damage in GHPC and ordinary concrete stairs without ribbed folding plates. The simulation results illustrate that GHPC stairs possess commendable fire resistance capabilities along with an ability to effectively recover from high-temperature damage.

The widespread application of the drill-and-blast method in tunnel construction generates instantaneous high-concentration dust, posing severe threats to workers’ health and safety. However, existing contact-based monitoring techniques, such as filter membrane weighing and light-scattering sensors, are hindered by operational complexity, maintenance challenges, data latency, and an inability to capture whole-field distribution. To address these limitations, this study conducted in situ experiments to construct a dust image dataset containing rich “real-world noise”. Analysis revealed significant variations in visibility, contrast, and light-scattering intensity across different concentration levels, establishing a physical basis for visual feature extraction. Consequently, a dust concentration prediction method based on transfer learning and CNN was proposed. Comparative training of six network models, including ResNet50, demonstrated that VGG16 achieved the best performance with an R² of 0.9684, verifying the method’s feasibility. Furthermore, hyperparameter optimization (learning rate: 0.001; batch size: 32; dataset split: 8:1:1) was shown to further enhance prediction accuracy.