Abstract
This paper presents a quantitative method of organizing field observations concerning liquefaction and non-liquefaction during actual earthquakes. The ratio τc/σ^-V is evaluated for 13 sites affected by 8 earthquakes and plotted against relative density (τc is the average peak dynamic stress during the earthquake, and σ^-V is the vertical effective overburden stress, both evaluated at the critical depth). The effect of the duration of shaking is considered by means of a correction to τc/σ^-V. For recent fluvial deposits and uncompacted hydraulic fills, the critical level of τc/σ^-V for fine sands is about 0.15. This result is in reasonable agreement with results from laboratory investigation. More evidence is required concerning coarser and denser granular deposits.
