| As the temperature of warm frozen soil range with permafrost severe phase transition, its physical and mechanical properties are extremely easy to change on the influence of some factors such as confining pressure, they are uncertain and their distributions are random. Reasonable value of thermal and mechanical parameters is the premise and basis to solve the cold regions engineering problems. In order to optimize the engineering design and to ensure the project security, the theory of probability and statistics must be introduced to analyze this uncertainty and get reliable parameters.In this paper, through testing50Qinghai-Tibet Plateau frozen silt samples at-1.5℃can get the thermal parameters of warm frozen soil such as thermal conductivity and volumetric heat capacity, and the thermal diffusivity can be calculated. Based on two classic probability distribution mode-normal distribution and lognormal distribution, the distribution of the mentioned thermal parameters is analyzed, the variable coefficient and distribution parameters can be obtained.In order to study the creep characteristics of warm frozen soil, an artificial triaxial creep test of frozen silt under0.5MPa,1.0MPa,2.0MPa at-1.5℃temperature environment were conducted respectively. Considered the creep curve shape and their relationship with stress and time, a creep model of warm frozen soil under complex stress is given on the basis of creep power function formulas, which is verified to be practical by fitting the experimental curves with the equation curve.Next, a method of combination is proposed to expand the sample size of creep equation parameters. Based on the knowledge of probability and mathematical statistics, the creep equation parameters of warm frozen soil are analyzed and its coefficient of variation and probability distribution are determined. Then, according to the long-term strength equation can obtain the long-term strength (100years later) of warm frozen soil, three classic distribution mode (normal distribution, lognormal and Weibull distribution) and the K-S test method are used to confirm its distribution pattern. In the end, different reliability statistics of long-term strength of warm frozen soil are analyzed. |
