Relationship Between Flexible Pavement Cracking and Surface Deflections

The Kansas Department of Transportation (KDOT) has a comprehensive pavement management system (PMS), commonly known as network optimization system (NOS). The performance prediction of this PMS is based on the Markov process where the minimum time step is one year. Thus in support of NOS, annual condition surveys are conducted. Currently, KDOT uses manual distress surveys to collect cracking data which is slow, labor intensive and somewhat hazardous especially in the urban areas. KDOT uses a falling weight deflectometer (FWD) to measure pavement surface deflection. The objective of this study was to investigate whether cracking of flexible pavements can be predicted in terms of surface deflection and some other objectively measured variables. FWD deflection data, collected with a Dynatest 8000 FWD on the KDOT network were analyzed. Nonlinear regression procedure in the Statistical Analysis Software (SAS) and Solver in Microsoft Excel were also used. Sigmoidal Equivalent Fatigue Cracking (EFCR) and Equivalent Transverse Cracking (ETCR) models have shown low to high coefficients of determination. The higher the number of data points, the lower the coefficient of determination. The lower the mean absolute deviation difference (MAD), the better the agreement between observed and estimated cracking values. These models can be used to estimate EFCR and ETCR at network-level.