Abstract
This study presents experimental work aimed at investigating the feasibility of fibre Bragg sensors as an early warning system for sinkhole detection. To this end, small-scale plane strain, scaled, instrumented 1-g models were built to provide further insight into the formation of a sinkhole using Cullinan silica sand and wad material due to an embedded leaking pipe for a range of densities. In these models, the sinkhole was simulated by an inflated balloon. The tests included tests with a higher flow rate of leaking water within the sinkhole models, one with silica sand and the other with layers of wad material mixed with silica sand with a lower leaking flow rate, in silica sand wad material. FBGs were positioned horizontally and vertically in the models. The experimental results suggest that the upward propagation of the sinkhole process could be divided into four phases: Phase 1 - cavity formation, characterised by the time for the balloon to deflate, producing a stable arch to support the soil mass above the cavity. Phase 2 - the weathering process was characterised by the leaking time (infiltration of water) after the cavity was formed. Phase 3 - the collapsing process during which time the failure was induced. Phase 4 - the equilibrium period, where the soil mass falls over the sensors into the void (cavity) and reaches an equilibrium stage after the void (reservoir) is filled within the model. The failure was described as gradual in that the inflow of water into the land mass was slow and long. The collapse process in the sinkhole models occurred gradually as cover subsidence sinkholes or suddenly. It was also observed that the failure mechanism depended on the leaking pipe's flow rate and the soil mass's permeability. The internal strain distribution is not symmetrical from both sides of the cavity. According to the findings of this study, the fibre Bragg grating sensors are sensitive to ground movement and can identify collapse failure before it becomes visible. They have been observed to be highly useful in determining the strain distribution in different soil profiles for characterising and identifying the development of sinkholes and monitoring the collapse process, which might serve as an early warning signal for sinkhole disasters since the optical fibre sensing can provide a viable option for developing an early warning system. Efficient integration of an IoT-enabled FBG sensing system could detect precursor movement of soil mass due to leaking pipes to provide automated decisions.
Keywords: FBG, fibre Bragg grating, geotechnical monitoring, leaking pipe, optical fibre sensors, physical modelling, sinkhole, subsidence, wad material.