Comparisons of Five Computational Methods for Transverse Profiles

Abstract

Abstract Pavement funding allocation and rehabilitation/maintenance decisions are often based on the degree of rutting. Thus it is critical to have accurate rut measurement. In this study, five different sensor setups and computation algorithms were investigated which include: (1) 5-sensor stringline, (2) 3-sensor average, (3) 5-sensor maximum, (4) 7-sensor stringline, and (5) 7-sensor maximum. The 94 transverse profiles utilized in this study included six profiles from the standard 3.7-m wide pavement (with a well-defined rut basin) and 88 profiles from narrow pavements with a possible edge drop off. It was found that the 5-sensor maximum method yielded the best accuracy for both standard 3.7-m wide and narrow pavements. It is surprising to find that the 3-sensor average yielded better results than the 5-sensor stringline method. Also, regardless of the number of sensors (either 5 or 7) or pavement width (standard or narrow), the stringline method always underestimated the rutting. Based on the available data and analyses performed, it is concluded that, without altering the existing sensor configuration, the accuracy would improve significantly by changing the computation algorithm from 5-sensor stringline to 5-sensor maximum.