Calibration of PLAXIS frozen/unfrozen soil model according to results of laboratory tests and in-situ monitoring

Abstract

Frost heaving is a challenge for geotechnical engineers who deal with transportation infrastructure in a cold climate. To make a frost heave forecast and access reliability of soil masses engineers apply different numerical soil models calibrated by results of in-situ and numerical simulation tests. The recently developed frozen/unfrozen soil model by PLAXIS bv and NTNU was used for reproduction of laboratory and field tests data. The model describes frost heaving as a function of frost front moving rate and water migration rate from unfrozen zone and has 25 parameters: 7 general ones, 12 parameters that are responsible for solid strains, 3 parameters responsible for suction strains and 3 ones that account for coupling effects between the variation of the solid phase stresses and the cryogenic suction. The results of oedometer tests of silt in the unfrozen and frozen states and frost heave tests, presented in the article, allow us to obtain such parameters as initial segregation threshold value, elastic and elasto-plastic compressibility coefficients, rate of change in Young’s modulus with temperature. But we have to apply backward analysis in order to obtain a best-fit of simulated results with test results. Unfortunately, most parameters is to be obtained by only calibration method. Additionally, to validate the frozen and unfrozen soil model the results of geotechnical monitoring of the cold-storage building are used. The frost heave phenomenon was appeared for 29-years period of its operation. Results of in-situ monitoring and numerical simulation showed good correlation with depth of frost penetration but unsatisfied results of frost heave deformations.