Fiber Bragg grating-based performance monitoring of a slope model subjected to seepage

Abstract

In the past few years, fiber optic sensing technologies have played an increasingly important role in the health monitoring of civil infrastructures. These innovative sensing technologies have recently been successfully applied to the performance monitoring of a series of geotechnical structures. Fiber optic sensors have shown many unique advantages in comparison with conventional sensors, including immunity to electrical noise, higher precision and improved durability and embedding capabilities; fiber optic sensors are also smaller in size and lighter in weight. In order to explore the mechanism of seepage-induced slope instability, a small-scale 1 g model test of the soil slope has been performed in the laboratory. During the model’s construction, specially fabricated sensing fibers containing nine fiber Bragg grating (FBG) strain sensors connected in a series were horizontally and vertically embedded into the soil mass. The surcharge load was applied on the slope crest, and the groundwater level inside of the slope was subsequently varied using two water chambers installed besides the slope model. The fiber optic sensing data of the vertical and horizontal strains within the slope model were automatically recorded by an FBG interrogator and a computer during the test. The test results are presented and interpreted in detail. It is found that the gradually accumulated deformation of the slope model subjected to seepage can be accurately captured by the quasi-distributed FBG strain sensors. The test results also demonstrate that the slope stability is significantly affected by ground water seepage, which fits well with the results that were calculated using finite element and limit equilibrium methods. The relationship between the strain measurements and the safety factors is further analyzed, together with a discussion on the residual strains. The performance evaluation of a soil slope using fiber optic strain sensors is proved to be a potentially effective approach.