Repeatability and Agreement of Various High-Speed Macrotexture Measurement Devices

Abstract

Five high-speed macrotexture measurement devices were tested on a variety of asphalt and Portland cement concrete pavement surfaces to evaluate their repeatability and their pairwise agreement. Experiments were run under three speed conditions: highway speed, varying constant speeds, and various acceleration and deceleration profiles. Data were processed and reduced per current industry standards. A novel approach for outlier removal from line laser devices was developed. The pairwise device agreement was evaluated using a limits of agreement analysis. The results demonstrate good repeatability for each of the devices tested. The agreement analysis showed that not all high-speed distance triangulation devices can be used interchangeably for all pavement surfaces. Data acquisition speed was found to be a factor in macrotexture parameter calculation for two of the five devices. The effect of speed was found to be worse on randomly textured surfaces than on transversely textured surfaces. Finally, acceleration was shown to have an effect on the parameters produced by one of the devices, giving further credence to the fact that care should be taken to gather high-quality datasets for the critical pavement characteristic of macrotexture.