RELIABILITY-BASED EVALUATION OF FINE GRAIN EFFECTS ON LIQUEFACTION RESISTANCE

Abstract

Liquefaction evaluation can use deterministic or probabilistic methods. The deterministic approach is limited by its inability to address uncertainty from soil complexity and heterogeneity, as well as the probabilistic nature of earthquakes. This method is often inadequate for accurate liquefaction analysis, as it does not reflect true field conditions. Conversely, probabilistic analysis allows for uncertainty and establishes a safety factor proportional to the associated risk. This research analyzes at how the influence of fine particles affects the likelihood of liquefaction, using the Lind-Hasofer reliability theory. This theory estimates the any probability of reliability (Ro) by converting the nonlinear limit state function into a linear form around the design point. The reliability of liquefaction (Ro) is determined using factors such as earthquake magnitude (Mw), maximum shaking strength (a_max/g), total pressure (σ_v), effective pressure (σ'_v), percentage of fine particles (FC), and SPT blow count (N_SPT). Results from 16 drilling locations with different amounts of fine particles and earthquake loads show that cyclic resistance increases with (N₁)_60CS​, but decreases when fines content exceeds 35% or when (N₁)_60CS < 13. The empirical relationship between SF and Ro (y = 8.898x^3.0536, R2 > 0.9267) highlights that some layers with SF ≥ 1 still correspond to low Ro < 0.8, indicating the limitations of deterministic analysis. Overall, the probabilistic approach provides a more realistic and risk-consistent assessment of liquefaction potential, making it more suitable for risk- and performance-based geotechnical design.