The Roles of Deflection Basins and Slopes in Bayesian Updating using Traffic Speed Deflectometer Measurements

The traffic speed deflectometer (TSD) technology facilitates continuous, non-destructive pavement structural assessment under actual traffic conditions without disrupting traffic flow. The TSD measures deflection slopes using laser Doppler velocimeters, which can be processed to generate deflection basins. Both data types can serve as inputs for back-calculating pavement material parameters. However, when it comes to deploying TSD widely, challenges exist: (i) TSD data exhibit varying sensitivity to material parameters of different pavement layers, which affects the reliability of back-calculations, and (ii) different deflection basins can correspond to similar slopes, which may compromise the uniqueness of back-calculated parameters. While previous studies have utilized either basins or slopes for back-calculation, there is a lack of comparative studies clarifying their respective roles. This paper investigates the effectiveness of using deflection basins, slopes, and their combination for back-calculating material parameters. Starting with a simulated example, results showed that relying solely on deflection basins yielded less reliable estimates for surface and base layers but was effective for subgrade layers. In contrast, slopes were especially effective for base layers. Optimal parameter estimates were achieved by combining basins and slopes. These findings were corroborated using two case studies involving real-world TSD data. This study provides evidence supporting the potential for future advancements by equipping TSD systems with both deflection and slope sensors or measuring corresponding deflection data with other techniques. This also highlights the promise of the combined use of slopes and deflections in enhancing the reliability of structural assessments and facilitating implementation of TSD in practical applications.