Multivariate regression analysis for rapid fatigue prediction in airport rigid pavements

The Airport Pavement Management System (APMS) of extensive air - side infrastructures demands a continuous measurement survey, up-to-date monitoring analysis, proactive rehabilitation and maintenance strategy. Reactive management protocols and current lack of integration in design, survey, and monitoring procedures undermine the effectiveness of any long-term decision making processes. In particular, rigid pavement are currently designed with numerical models and subsequently surveyed with NDT techniques as plate elementary units. Due to the high computational times and numerical resources required, FEA analyses are typically averaged on representative sample units, which inhibits the reliability of the assessment procedure. On the other hand, the use of theoretical closed-form solutions simplifies the physical model allowing for rapid evaluation, at the expense of overall accuracy. In this study, numerical solutions are taken as a reference for setting up a multivariate regressive analysis of rapid theoretical models, with varying pavement and traffic conditions. Hereby, a mechanistic-empirical fatigue condition assessment is directly established and validated to automatically infer Cumulative Damage Factor (CDF) calculation from previous tensile stress results. Relevant results demonstrate a good viability of the proposed method against current stress calculation and damage prediction models, thereby showing potential to contribute to the existing APMS frameworks.