Assessment of resonance compaction effects on site liquefaction probability by considering spatial variability

Traditional methods for simplified soil liquefaction estimation, such as the cone penetration test, often rely on single-point data, failing to account for spatial variability, limiting their accuracy and comprehensiveness. In this study, we aimed to examine the effectiveness of the resonance compaction method for treating liquefiable soils by considering spatial variability. We used Bayesian compression sampling to generate 2D random field profile samples of the soil liquefaction safety coefficient to visualise the effects of resonance compaction. Based on the generated random field samples, the probability of liquefaction at each location on the 2D profile was calculated using Monte Carlo simulations. We observed that the resonance method was effective in reducing or eliminating liquefaction. However, owing to the spatial variability of the geotechnical body, the treatment effects varied at different locations. Thus, this approach provides a robust understanding of the effectiveness of the resonance compaction method for liquefiable foundations and highlights the significance of considering spatial variability through 2D liquefaction probability profiles. The results underscore the need for targeted additional treatments in areas with suboptimal outcomes to ensure comprehensive ground improvement.