Evaluation of cracking in asphalt pavement with stabilized base course based on statistical pattern recognition

Abstract

Selection of optimal treatment strategy for a cracked asphalt pavement requires efficient evaluation of cracking. Pavement cracking severity was usually evaluated just by crack width in the past. In this paper, a novel method for extracting both crack type and crack width information in asphalt pavement with chemically stabilised base course from falling weight deflectometer (FWD) test data based on k-nearest neighbour (k-NN) pattern recognition was put forward. FWD field tests and core drilling tests were conducted at cracks on four typical expressways and a database with 215 deflection basins was established. Three feature sets were selected or extracted from each deflection basin as inputs. And, cracks were categorised into six classes based on crack type and crack width as target outputs. Results indicated that most cracks with high width level were thermal cracks and it was of great importance to consider both crack type and width in cracking evaluation in asphalt pavement with chemically stabilised base. Considering the classification accuracy and the convenience of test, crack was advised to be located between sensor S300 and sensor S600 in FWD field test. Deflection ratio feature set was recommended as inputs of the k-NN classifier.