Study On The Reliability Of Thawing Settlement Of Road Subgrade In Frozen Soil Area Based On Random Field Theory

The evaluation of stability and reliability of permafrost roadbed is a complex engineering problem to be solved.The freezing and melting of ice in frozen soils makes the soil parameters more discrete and random.In this paper,the permafrost roadbed is taken as the research object.Considering the randomness of the thermal parameters(specific heat capacity and heat transfer coefficient)of the permafrost and the parameters of the external air temperature,the random temperature of the subgrade in the permafrost region is established based on the engineering background of a section of the Qinghai-Tibet Railway.The field analysis model uses Monte Carlo stochastic finite element method to calculate the reliability of frozen soil,and the results of the study have a certain reference for the stability of permafrost roadbed.The main content of the paper is as follows:(1)Combining CPT(static cone penetration)field measured data,using the random field theory and the method of solving the related distances,the average correlation distance of the frozen soil is calculated as the main parameter for the research reliability.(2)According to the random field basic theory and local average discrete method,the subgrade temperature field is discretized according to the average correlation distance of permafrost.The random variables affecting the temperature field of permafrost are analyzed.The covariance matrix of each random variable group and correlation coefficient matrix is calculated using Matlab.(3)Under the condition that the global average temperature rises by 2.6°C in the next 50 years,the temperature field in the frozen formation of the subgrade will be determined for 5,10,15 and 20 years regardless of the temperature field of the warmed natural formation.In the finite element calculation,the change of temperature field was analyzed from the natural ground center,the temperature change in the center of the roadbed top surface,the rise and fall of the frozen soil ceiling,and the size of the melting block directly below the roadbed,which was compared with the simulation of the stochastic finite element.(4)Using the ANSYS/PDS module,considering the correlation coefficient between random variables in random temperature field,Monte Carlo stochastic finite element method is used to calculate the randomness of subgrade temperature field,and the input parameters are tested randomly at the subgrade center.The nodal temperature was used for sensitivity analysis,and the upper limit sample of frozen soil was calculated by interpolation method.The random distribution probability of the upper limit of the frozen soil was calculated to lay the foundation for calculating the reliability of frozen soil subgrade settlement.(5)According to the principle of one-dimensional consolidation of Terzaghi,the structure of the permafrost roadbed is simplified.Considering the joint action of the roadbedstructure and the static load and dynamic load of the train,the ANSYS is used to establish the frozen surface below the natural upper limit and the natural upper limit.The subsidence and subsidence of permafrost roadbed at 10 cm,20cm,and 30 cm are used to construct the subgrade subsidence functional function,and the reliability and reliability indexes of 5 years,10 years,15 years and 20 years of subgrade construction of frozen soil are calculated.The innovation of this paper is to consider the randomness of the variation of the upper limit of the frozen soil and to study the reliability of the frozen soil subgrade as a bridge.