Guidelines from direct shear modeling in centrifuge for TDR landslide monitoring

Abstract

The suitability of using time-domain reflectometry (TDR) for landslide monitoring and assessment has been validated by field studies. This method can be used for detecting shear surfaces and monitoring shear displacements at depth, which can provide warning of potential slope failures or landslide reactivations. Although laboratory tests have suggested using the TDR reflection coefficient to estimate shear displacement, quantifying the relationship between these parameters is challenging because it is affected by various cable–grout–ground interactions. Investigating these interactions is challenging because of the limited artificial overburden pressure that can be applied in a laboratory. Therefore, this study improved the TDR applicability through high-gravity centrifuge modeling to simulate sliding at depth. Furthermore, a flexible coaxial cable was modified to improve its sensitivity, which was evaluated using the simple direct shear test. The modified coaxial cable with stiff grout had sufficient sensitivity to detect a small-scale shear displacement of 0.5 mm. Reversed-fault centrifuge modeling was conducted for sandy soil to simulate the shear plane under a gravity level of 55 g, and the shear plane was realistically generated and detected by a cable–grout–ground composite and cables located adjacent to an inclinometer casing. Notably, only the model-scale fault movements directly correlated with TDR reflection peaks and integration areas. Finally, guidelines for conducting TDR-based slope monitoring and directions for further research regarding TDR-based landslide centrifuge modeling are suggested.