Xiaoyang Liu

Secondary compression studies can provide insights for evaluating the engineering potential and environmental impact of soil. The objective of this research was to investigate the effect of bound water content on the secondary compression of marine silty clay. To this end, a novel method was established based on thermogravimetric analysis (TGA) to determine the contents and limits of strongly bound water and weakly bound water for three typical marine silty clays including Tianjin mucky silty clay (TJ), Qingdao mucky silty clay (QD), and Weihai silty clay (WH). A total of 17 groups of uniaxial confined compression tests were performed for reconstituted samples at different absolute water contents under the condition of multistage loading to investigate their secondary compression characteristics. The results show that the initial dehydration temperature of strongly bound water (T_s) corresponds to the peak of the derivative thermogravimetry (DTG) curve. The values of T_s for TJ, QD and WH are 112.35 °C, 109.67 °C and 118.46 °C, respectively. The initial dehydration temperatures of weakly bound water (T_w) for TJ, QD and WH are 55.26 °C, 52.56 °C and 56.56 °C, respectively. The secondary compression coefficient C_α changes little before the strongly bound water limit and increases dramatically as weakly bound water content increases at the same vertical stress. A piecewise linear model and a quadratic polynomial model are established for calculating the average secondary compression coefficient from bound water content. Weakly bound water is the determining factor controlling secondary compression. Increasing the bound water content weakens the connection force and friction force among the particles and the viscosity of weakly bound water. The results will guide decisions on long-term settlement assessment and facilitate understanding of the secondary compression mechanism of marine silty clays affected by bound water.

Landslides cause huge damage to social economy and human beings every year. Landslide susceptibility mapping (LSM) occupies an important position in land use and risk management. This study is to investigate a hybrid model which makes full use of the advantage of supervised learning model (SLM) and unsupervised learning model (ULM). Firstly, ten continuous variables were used to develop a ULM which consisted of factor analysis (FA) and k-means cluster for a preliminary landslide susceptibility map. Secondly, 351 landslides with “1” label were collected and the same number of non-landslide samples with “0” label were selected from the very low susceptibility area in the preliminary map, constituting a new priori condition for a SLM, and thirteen factors were used for the modeling of gradient boosting decision tree (GBDT) which represented for SLM. Finally, the performance of different models was verified using related indexes. The results showed that the performance of the pretreated GBDT model was improved with sensitivity, specificity, accuracy and the area under the curve (AUC) values of 88.60%, 92.59%, 90.60% and 0.976, respectively. It can be concluded that a pretreated model with strong robustness can be constructed by increasing the purity of samples.