A fuzzy inference system for predicting pavement surface damage due to combined action of traffic loading and water

Abstract

This paper presents a fuzzy logic-based deterioration prediction models for gap and open-graded asphalt surfaces when both dynamic loading and shallow flooding coincide. The impact of aggregate size, load frequency, compaction levels, and environmental conditions was evaluated in a controlled laboratory testing to measure cracking and rutting performance of each mixture. A set of fuzzy logic was developed using the experimental data and then tested against randomly selected samples. The predicted cracking and rutting showed excellent agreements (95% correlation) with the experimentally measured values. The validation and sensitivity analysis showed that irrespective of aggregate gradation, mixture parameters (aggregate size, void contents), traffic parameters (loading frequency) and environmental factors (wet and dry condition) have a significant impact on model performance. Overall, the Fuzzy-based prediction model showed the potential to differentiate the performance of different asphalt surfaces and can be further developed to use in practical applications.