Mechanistic model for transition probabilities

Abstract

One important input in a pavement-management system is the matrix of transition probabilities. A methodology is proposed that employs mechanistic distress submodels for evaluating pavement distresses in terms of fundamental material behavior and structural response to environment and traffic factors. This approach allows the distress modes to be modeled individually so that the appropriate measures for correcting deficiencies can be readily identified. The performance submodels are formulated within a stochastic framework for determining the transition probabilities. The formulation is based on state-of-the-art structural reliability methods, specifically, the first-order reliability method (FORM), and is applicable to individual and combined modes of distresses. It is able to handle numerous variables with various distribution types and correlations. The multimode distress case is an extension of the individual distress modes incorporating the concept of cut-sets and directional simulation. With the proposed methodology, the assumption of a time-invariance matrix of transition probabilities, as adopted in existing pavement-management systems, is no longer necessary.