Building and validation for liquefaction simulation using centrifuge model test

Abstract

The liquefaction phenomenon is caused by loss of soil strength when the rapid vibration such as an earthquake is applied under the instantly undrained condition. The liquefaction induced settlement, lateral spreading have a influence on the stability of the structure. In the U.S.A and Japan where strong earthquakes occur frequently, liquefaciton evaluation is an essential consideration in seismic design. Typical methods for evaluation liquefaction include field tests, laboratory tests, physical modeling, and numerical analysis. The field tests have major limitations in that they are costly and time consuming due to the actual ground investigation. In domestic, liquefaction evaluation is generally performed by using cyclic triaxial test and simple shear test. This is an appropriate method for analyzing the behavior of soil elements rather than the global behavior of the liquefied ground. As a physical modelling, centrifuge model tests can be performed to observe the global behavior of liquefied ground and the local behavior by installing sensors in the ground. In addition, It is possible to estimate economical and reliable data by comparing and verifying numerical analysis modeling with physical modeling experiment results. On the other hand, the disadvantage is that it is quite complex and difficult to implement a liquefaction model using the centrifuge model tests. There is no case of liquefaction study using the domestic centrifuge, and there is no specific procedure for implementing liquefaction model in domestic. In international study, an international collaborative research project called Liquefaction Experiments Analysis Project (LEAP) for liquefaction research was launched based on centrifuge model tests and numerical analysis. As participating in LEAP-2017 and LEAP-2018, centrifuge model tests were peformed to analyze liquefaction behavior with different relative density and to verify the generalized scaling law. For this study, it was possible to implement liquefaction model using the centrifuge based on the detailed experimental procedure provided by LEAP. In this study, a liquefaction simulation system was built by 6 steps of ground modelling, viscous fluid manufacture, saturation process, measurement, CPT, and input motion application to implement liquefaction model in the centrifuge model test. For the validation to the simulation system, a 5 degree inclination model with a different relative density was applied to the 1Hz sine wave of 0.15g for each model. Dilatancy spike, stress-strain curve, effective stress, and displacement were measured to evaluate the occurrence of liquefaction for each model. In order to overcome the limitations of centrifuge performance and size of the shaker, centrifuge model tests were conducted to verify the generalized scaling law. For the same prototype model, dilatancy spike, stress-strain curve, effective stress path, and displacement of liquefied ground were calculated and compared by conventional scaling law and generalized scaling law. The validation of the liquefaction simulation system constructed through the experiment results enabled us to secure the validity and reliability of the system. Based on the results of the system and the results of the study, efficient and reliable data estimation for the evaluation of structure behavior and liquefaction remediation in liquefied ground is suggested.