We proposed a new classification method for stress-bearing structures in very-deep roadways. We conducted tests for roadwayengineering mechanical properties, including rock mechanical tests and ground stress measurement of two caverns in very-deep wells. We suggested a classification method for stress-bearing structures based on shear stress. Tests revealed that rock strength in the shear direction was the lowest; the stress distributions of acoustic emission and hollow inclusion were highly similar. Based on the roadway-engineering mechanical properties of two caverns and numerical simulation and in-situ tests, the key bearing structures were classified based on the concentrated shear stress. In deep caverns, shear stress was more concentrated on the softer surrounding rock, the key bearing structure areas, and the more seriously fractured surrounding rocks. Using the loose circle in-situ test, we compared the classification method of the key bearing structures with other classification methods. The results revealed agreement with the classification methods used for the key bearing structures. The deformation in-situ test showed that the strata convergence of concentrated shear stress areas developed quickly. Therefore, the phenomena of concentrated shear stress and the expanded range of key bearing structures are the structural instability mechanisms of deep roadways.

A two-branch convolutional neural network (CNN) architecture for feature extraction in person re-identification (re-ID) based on video surveillance is proposed. Highly discriminative person features are obtained by extracting both global and local features. Moreover, an adaptive triplet loss function based on the original triplet loss function is proposed and is used in the network training process, resulting in a significantly improved learning efficiency. The experimental results on open datasets demonstrate the effectiveness of the proposed method.

We study the structural instability mechanism and effect of a multi-echelon support in very-deep roadways. We conduct a scale model test for analysing the structural failure mechanism and the effect of multi-echelon support of roadways under high horizontal stress. Mechanical bearing structures are classified according to their secondary stress distribution and the strength degradation of the surrounding rock after roadway excavation. A new method is proposed by partitioning the mechanical bearing structure of the surrounding rock into weak, key and main coupling bearing stratums. In the surrounding rock, the main bearing stratum is the plastic reshaping and flowing area. The weak bearing stratum is the peeling layer or the caving part. And the key bearing stratum is the shearing and yielding area. The structural fracture mechanism of roadways is considered in analysing the bearing structure instability of the surrounding rock, and multi-echelon support that considers the structural characteristics of roadway bearings is proposed. Results of the experimental study indicate that horizontal pressure seriously influences the stability of the surrounding rock, as indicated by extension of the weak bearing area and the transfer of the main and key bearing zones. The falling roof, rib spalling, and floor heave indicate the decline of the bearing capacity of surrounding rock, thereby causing roadway structural instability. Multi-echelon support is proposed according to the mechanical bearing structure of the surrounding rock without support. The redesigned support can reduce the scope of the weak bearing area and limit the transfer of the main and key bearing areas. Consequently, kilometre-deep roadway disasters, such as wedge roof caving, floor heave, and rib spalling, can be avoided to a certain degree, and plastic flow in the surrounding rock is relieved. The adverse effect of horizontal stress on the vault, spandrel and arch foot decreases. The stability of the soft rock surrounding the roadways is maintained.

Premature failure of rock bolts (cables) due to stress corrosion cracking (SCC) is a phenomenon that has been reported to occur in the underground environment. In the 1990s, many failure accidents of bolts which occurred in the United Kingdom were caused by SCC [1]. In this study, the corrosion behavior and failure mechanism of rock bolt (cable) samples obtained from the underground coal mine were examined and discussed. Macroscopic observation and weight loss tests were carried out for the bolts’ corrosion characteristics without failure. The results show that the bolts with short service time (1.5–2 yrs) underwent uniform corrosion. However, bolts with longer service time (3–8 yrs) experienced different pitting corrosion degrees. The corroding degree of different parts of bolt samples shows the following decreasing trend: bolt head > bolt end > free section. The absolute corrosion degree increased with the service time, while the corrosion rate was the highest in the early stage and dropped down in the later stage. At the same time, the macro‐ and micromethods were used to analyse the failure mechanism in the broken cable sample. Failure of one cable sample with a medium service life (6 yrs) was found to be controlled by the SCC. It was induced by long‐term action of O, Cl, and S in the surrounding rock environment and resulted in pitting corrosion. The pitting corrosion reduced the outer diameter of the rock cable and its bearing capacity, leading to the final fracture.

In order to explore the micro-cracking behavior and fractal characteristics of sandstone with different particle sizes under the condition of compression-shear test, this paper studies sandstone with different particle sizes by using high-strength concrete wrapped in cylindrical sandstone test specimen to carry out a compression-shear mechanical test, with the assistance of performing 3D optical scanning on the fracture surface of sandstone after shear failure by using MATLAB to re-establish the three-dimensional shape of the fracture surface. The roughness parameter of the joint surface is introduced to quantitatively calculate the fracture structural plane, which found that the roughness parameter manifests the damage of the structural plane well. The roughness of fracture surface is negatively correlated with the particle size of sandstone, that is, the roughness of fracture surface increases with the decrease of particle size. The fractal dimension of the structural trace line and the sampling area is calculated according to the digital model. Obvious anisotropic characteristics are observed in both dimensions, where the fractal dimension is positively correlated with the particle size of sandstone, in another word, the fractal dimension decreases along with the particle size. The above research results provide experimental support in scientific theory for clarifying the instability mechanism of sandy rock roadway under compression-shear conditions.