Continuous deflection monitoring of flexible pavement under varying environmental conditions using embedded sensors

This study investigates the influence of pavement temperature on deflection measurements obtained from embedded geophones and accelerometers installed in an instrumented flexible pavement section on the E16 motorway in Borlänge, Sweden. The analysis is based on case-wise heavy-vehicle passages recorded during a continuous monitoring period spanning Autumn 2024 to Summer 2025, encompassing a wide range of axle configurations, gross vehicle weights, travel speeds, and near-surface pavement temperatures ranging from approximately −11 °C to over 24 °C. Vibration velocity and acceleration signals were processed to estimate peak pavement surface deflections, which were subsequently normalised by load and related to pavement temperature using correlation analysis, one-way ANOVA, and regression modelling. Two empirical temperature–deflection correction models, linear and exponential, were developed to normalise deflections to a reference temperature. Both models significantly reduced temperature-induced variability in the measured deflections; however, the exponential model provided more effective overall correction, consistent with the nonlinear temperature-dependent stiffness behaviour of asphalt materials. Temperature-corrected deflections showed substantially improved agreement with multilayer elastic responses computed using the ERAPave software, with coefficients of determination reaching approximately 0.86 and normalised RMSE values reduced to less than 10% after exponential correction. The results demonstrate that embedded geophones or accelerometers, when combined with appropriate temperature normalisation, can enable accurate, year-round pavement response monitoring at traffic speed in cold and seasonally variable climates.