Generation and propagation of acoustic emissions in buried steel infrastructure for monitoring soil–structure interactions

Soil–structure systems (e.g. pipelines, pile foundations, retaining structures) deteriorate with time and experience relative deformations between the soil and structural elements. Whether a result of age, working conditions, or environmental conditions, deformations have the potential to cause catastrophic social, economic, and environmental issues, including limit state failure (fatigue, serviceability, ultimate). The UK spends £100s of millions a year spent on infrastructural maintenance; the early detection of deterioration processes could reduce this spend by an order of magnitude.

Techniques to monitor ground instability and deterioration are consequently increasing in use, with most conventional approaches providing localised information on deformation at discrete time intervals. Nascent technologies (e.g. ShapeAccelArray, fibre optics) are however beginning to provide continuous measurements, allowing for near real-time observations to be made, although none are without either technical limitation or prohibitive cost.

A novel monitoring system is proposed, whereby pre-existing and newly built steel infrastructure (e.g. utility pipes, pile foundations) are employed as waveguides to measure soil-steel interaction-generated AE using piezoelectric sensors. With this, a two-stage quantitative framework for understanding soil-steel interaction-generated AE and its propagation through steel structures is also developed where (stage 1) informs the creation of an adaptable sensor network for a variety of infrastructure systems, and stage (2) informs interpretations of the collected AE data to allow for decision makers to take appropriate action. Timely actions made possible by such a framework is of great significance to practitioners, having the potential to reduce the direct and indirect impacts of deterioration and deformation, whether long- and short-term.

Stage 1 used an extensive programme of computational models, alongside small- and large-scale physical models, to enable attenuation coefficients to be quantified for a range of soil types. It was shown that both the structure and bounding materials, i.e. the burial system, significantly influenced propagation and attenuation through steel structures. In free-systems, though, the frequency-thickness product was more influential; propagation distances of 100s of metres are obtained at products <0.5 MHz-mm but reduce to 10s of metres by 1 MHz-mm. Guidelines for three generic systems, free bound, soil bound, and soil bound with an internal water environment, were developed.

Stage 2 used a programme of large direct-shear box tests to allow for relationships between AE and normal effective stress, mobilised shearing resistance, and shearing velocity to be quantified. This enabled for quantitative interpretations of soil-steel interaction behaviours to be made using various AE parameters. Both the magnitude of values, and the rates of change of the parameters, could be used in the interpretation of behaviours. Shearing and stress conditions of sand could also be determined, increasing proportionally with AE activity, whilst the point at which full shear strength mobilisation occurs was also identifiable.