講演情報

[Data_Collection-37]Development of a Frequency-based Metric for Characterizing Transverse Profiles

Aditya Kumar1, Aditya Medury1, *Animesh Das1 (1. Indian Institute of Technology Kanpur (India))

キーワード:

rutting、transverse profile、Fourier analysis

The structural distress along the transverse profile of pavement surfaces is characterized by rutting. Traditionally, rut depth measurements, collected using a combination of manual (straight edge method, wireline method) and automated (multi-point laser) measurements emphasize upon the maximum depth sampled along the transverse direction, ignoring the overall irregularities present in the transverse profile. Further, these measures are affected by the existing camber slope of the pavement, thus leading to measurement inconsistencies. In the present work, a frequency-based approach using Fast Fourier Transform is proposed which offers the advantages of rotational and translation invariance. The frequency decomposition approach provides a general framework to model and analyze the entire shape of the deformed transverse profile, which is otherwise sub-sampled (e.g., maximum depth) or simplistically aggregated (e.g., rut cross-section area) in existing traditional rutting metrics. The proposed methodology is illustrated using high-resolution transverse profiles collected from line lasers (at 1 mm sampling interval) along multiple kilometers of Indian highway. Using an empirical dataset, a comparative assessment of the proposed metric with the prevalent measures of depth and area-based summary statistics is undertaken, wherein salient differences, and similarities, across metrics is highlighted. Finally, the proposed approach is also leveraged to categorize different rut shapes based on the composition of the underlying dominant frequencies.