During- and post-flood damage-vulnerability window in unbound granular pavements: A degree-of-saturation threshold decision rule

• Single DoS threshold ( Do S c ) defines one damage-vulnerability window. • Do S c calibrated from RLT near stiffness peak and rising deformation risk. • Field DoS time series give a binary signal and exposure metrics for planning. • Case road: Do S c = 0.83; sustained vulnerability observed during flooding. • Guides during- and post-flood actions: axle limits, speed, drainage. Floods leave moisture-sensitive unbound granular pavements in a weakened state in which premature reopening accelerates rutting and permanent deformation. This paper advances a single decision rule that defines a damage-vulnerability window whenever field-measured Degree of Saturation (DoS) exceeds a site-calibrated critical threshold ( Do S c ) , derived from standard repeated-load triaxial (RLT) evidence. The threshold is anchored in resilient behaviour: Do S c is selected near the onset of stiffness decline together with deformation susceptibility under wetting and is then checked against field observations. Hourly DoS time series at depths are used to identify when the pavement is vulnerable. Two metrics summarise exposure for planning and communication: (i) the share of hours with DoS > Do S c and (ii) the vulnerable-to-non-vulnerable hour ratio. A flood-prone low-volume road case demonstrates the approach: with Do S c = 0.83, the pavement experienced a sustained vulnerability window during inundation, indicating proportionate controls such as temporary axle restrictions, speed management, and targeted drainage works. Normal operation resumes once DoS remains below Do S c with supporting site checks. The framework explicitly links flood phases to decisions: the during-flood phase corresponds to DoS > Do S c (elevated vulnerability), while post-flood recovery is tracked by the decline of DoS below Do S c and the duration of exceedance, informing the timing and intensity of interventions. Requiring only standard RLT data and field (or estimated) DoS time series, the method is minimal, transparent, and transferable to similar pavements, providing a clear moisture-aware cue for safer, faster reopening and integration with broader asset management.