Some Investigations on Cracking of the Tunnel in Cold Region

研究结果表明：(1)温度可以用线性变换从均匀空间过程中定义，隧道隔热层的设计温度可以用随机算法用半变异函数插值。将设计温度的估算值与实测温度进行了比较，结果表明所提出的方法具有很好的效果。(2)由于距离隧道入口较远，大气温度的周期性变化可以转化为移动平均值和变窄幅度。(3)利用神经网络进行数据挖掘，有助于发现隧道衬砌裂缝产生的原因。(4)提出了半无限区域一维运动边界问题的数值分析方法。该方法可用于确定隧道衬砌后冻土深度。与其他方法的比较表明，该方法具有令人满意的精度。(5)为研究隧道衬砌热裂的机理，采用二维有限元法，采用界面元对隔热层进行传热和应力计算。采用了包含老化效应和随时间变化特性的具体模型。明确了混凝土早期弹性模量和混凝土导热系数对热裂的敏感性。

The project is summarized as follows :(1) Temperature can be defined from homogeneous spatial processes by means of a linear transformation, and the design temperature for tunnel heat insulator can be interpolated by Semi-Variogram using a stochastic algorithm. A comparison of the estimated design temperature and the observed temperature of actual data shows that the proposed method yields very good results.(2) The periodic changes of atmospheric temperature can be transformed to a moving mean value and narrowing amplitude due to distance from the entrance of the tunnel.(3) It is found that data mining procedure using neural network is useful for finding the cause of cracks in the tunnel lining.(4) A numerical method is developed for analyzing for the one-dimensional moving boundary problem in semi-infinite region. The method is useful for determining the depth of the frozen ground behind tunnel lining. A comparison with other solutions suggests that the proposed method provide satisfactory accuracy.(5) In order to investigate the mechanics of thermal cracking in the tunnel lining, heat transfer and stresses are determined using the two-dimensional finite element method and employing interface element for the heat insulator. A concrete model including ageing effects and time-dependent properties is adopted. It is made clear that the elastic modulus of concrete in early ages and thermal conductivity of concrete are sensitive to thermal cracking.