Verification and Validation of Pavement Models

Abstract

Model development activities are increasing in pavement research and engineering applications. At the same time, there is ambiguity and a lack of consistency with regard to checking and quantifying credibility and suitability for the intended application. Specifically, usage of the terms Verification and Validation (V&V) is seen to vary across contributions. In this context, this paper was motivated by the desire to provide a best-practice reference that underlines the significance of the V&V terms, clarifies their definition, and promotes a more unified usage. Accordingly, the objective was to offer examples that introduce the V&V jargon, demonstrate basic V&V concepts and processes, and highlight certain fine points. A total of five distinct models were introduced and discussed: (1) load-related responses in asphalt pavements, (2) load-related responses in concrete pavements, (3) crack initiation and propagation in asphalt concrete, (4) linear viscoelasticity of asphalt concrete, and (5) water flow through asphalt concrete pores. A general conclusion from this work is that successful V&V efforts are closely linked to a clear definition of the intended usage, i.e., the specific reality of interest being targeted by the model. It is also concluded that documenting V&V efforts is integral to any computational model development. Doing so communicates to potential users the region of confidence for the model alongside the expected differences with the reality of interest.