Search

In this paper, anthracite in Shanxi region is taken as a research object, and the pore structure of coal under different freeze-thaw times is characterized and analyzed by scanning electron microscope and nitrogen adsorption experiment. The SEM experimental results show that the micron-scale cracks on the surface of anthracite gradually connect after freeze-thaw, and the secondary crack channels are formed around the pores and air holes, so that the pores also begin to connect. Further, the quantitative information of the picture was extracted, and it was found that the surface porosity and fractal dimension of the coal sample changed significantly after three freeze-thaw cycles, which increased by 163.3% and 15.9% respectively, and increased with the increase of freeze-thaw cycles. The results of low-temperature nitrogen adsorption experiments show that the coal samples after freeze-thaw are mainly crack, and the pore structure changes from simple to complex. In addition, the maximum nitrogen adsorption capacity, specific surface area and pore volume increased by 125.7%, 69.3% and 72.6%, respectively. With the increase of freeze-thaw times, the growth rates of maximum nitrogen adsorption capacity, specific surface area and pore volume are significantly different. The results of joint characterization show that the number of pores and cracks increases to different degrees after freezing and thawing. Comparing the increase in the degree of fracturing (163.3%) with the increase in the total pore volume (30.68%), we find that the degree of change of the crack after thawing is higher than that of the pore. In this paper, the damage effect of coal samples after freezing and thawing with liquid nitrogen is investigated from an experimental point of view, which accelerates the great leap of cyclic cryogenic fracturing technology from basic theory to technical practice. [ABSTRACT FROM AUTHOR]

Loess landslide is one of the most harmful and serious geological hazards in the Loess Plateau of China. Early identification of potential loess landslide is an urgent need for its prevention. Traditional methods, e.g. support vector machines and decision trees, often suffer complicated data pre-processing, multitudinous causative factors, or low accuracy. This study aims to develop a high-performance loess landslide early identification model based on convolutional neural networks (CNNs). A case study was carried out in northwest Lvliang, China, where loess landslide is a major concern. Two hundred and six loess landslide cases were interpreted by comparing remote sensing images of two time phases, and were randomly divided into a training set (80%; 165) and a validation set (20%; 41). Four algorithms were developed, including a CNN structure with skip connection using data with (S–C) or without (S–N) slope crest and plain CNN structure using data with (P–C) or without (P–N) slope crest. The results show that the S–C structure is the most suitable for early identification of potential loess landslides because it achieved the highest overall accuracy (OA = 0.902) and largest area under the receiver operating characteristic curve (AUC = 0.932) on the validation set. [ABSTRACT FROM AUTHOR]