Federal Road Profile Model Generation Based on Road Scanner Data

Abstract

In Malaysia, the application of Mechanistic-Empirical approach in pavement design guide is still in early stage of introduction and many more researches need to improve it. One of the task needs is generating road profile model to evaluate the dynamic axle loads from vehicles. Therefore, this study aims to generate a few models of road profiles based on the real road profile data called reference road profiles model using mathematical method. The reference road profiles data are a measurement of new paved federal road in Malaysia. A profiles of two section roads (300 m length each) measured by road scanner were determined from Malaysia Public Work Institute (IKRAM). The Profile Viewing and Analysis software (ProVAL) was used to filter unwanted wavelength, visualized the measured road profile data and obtained the road profile International Roughness Index (IRI). Meanwhile, Matrix Laboratory software (MATLAB) was used to generate road profiles based on Dodds and Robson Power Spectral Density (PSD) approximation equation. The calibration between generated and reference road profiles were examined by correlation coefficient value. From the results, all generated road profiles with correlation coefficient value ranges between −1 < ρ < −0.7 or 0.7 < ρ < 1, gives different IRI values with the reference road profile. However, the differences of IRI values are acceptable (less than 5 %). Therefore, all 56 models of generated road profiles are applicable for long term pavement performance. Although, a road profile model with the same IRI value with a reference profile is the most applicable for long term pavement performance.