Long-term prediction of the performance and durability of pavement represents a critical and vital issue in the pavement surface type selection process by the Kansas Department of Transportation (KDOT) using the life-cycle-cost analysis. Accurate prediction of roughness progression on Portland Cement Concrete (PCC) pavements is very important since the current model used by KDOT is based on the pavement serviceability guidelines (1993 AASHTO Design Guide). In this study, dynamic Artificial Neural Network (ANN) and statistical analysis approaches were used to develop reliable and accurate time-dependent roughness (International Roughness Index, IRI) prediction models for the newly constructed Kansas Jointed Plain Concrete Pavements (JPCP). To achieve this objective, data used in the model development process include construction and materials data as well as other inventory items, such as, traffic and climatic related data, which reflect the section-specific local conditions in Kansas.

In the fall of 1984, the Kansas Department of Transportation constructed a project involving cracking, seating, and overlaying a portland cement concrete pavement. The project is located in Wyandotte County on K-7 highway, north of K-32. This project is on the west side of the Kansas City metropolitan area. Eight experimental sections and one control section were stipulated for this project. One section of regular construction was selected for the control section. Four of the eight test sections were designated to be cracked at 0.9 m (3 ft) centers, and the other four were cracked at 1.5 m (5 ft) centers. One half of each spacing pattern was saw cut at the quarter points of the 18.7 m (61.5 ft) panels [approximately 4.6 m (15 ft) spacing, minimum 127 mm (5 in.) deep, maximum 13 mm (0.5 in.) wide] prior to cracking. Four sections had minimal joint repair; four had normal joint patching. There have been no easy determinants as to why the transverse cracking is low in some sections, and high in others. In some years it appears to be one factor, but the next year the factor has appeared to change. After 10 years there is no difference attributable to the cracking spacing, probably because both spacings are too large to adequately distribute the thermal movements. Slightly better performance was noted in sections that have minimal joint patching. More recent projects have used rubblizing as a rehabilitation technique. These have not shown significantly better performance than the cracking method. More investigation needs to be done to quantify the effect of various components that are used in this technique.