In-situ tests and inspections for reliability assessment of RC buildings: how accurate?

Measuring the incomplete knowledge of the structural properties in as-built conditions is a formidable challenge in the performance-based seismic assessment of existing reinforced concrete (RC) buildings. Two basic sources of structural modeling uncertainties, which can directly affect the component demand and capacities and even influence the eventual structural collapse mechanism, are related to the mechanical properties of materials and the construction details. Recent European codes propose to consider the uncertainty in the knowledge of structural properties for an existing building by introducing an adjustment factor applied to the mean material strength, called Confidence Factor, whose value depends on the level of knowledge (KL) of structural properties. The latter (KL) is established as a function of number of in-situ tests and inspections available. The implementation of the code-based approach inevitably brings up several questions to be answered ranging from the implementation of the result of tests and inspections, and their relative measurement error. This work aims to introduce a Bayesian framework to quantify the relative error associated to the compressive concrete strength’s non-destructive ultrasonic tests. The proposed framework is applied to the test data available for an existing frame belonging to a pre-seismic code RC school building in Avellino (Campania), located in southern Italy. The framework also provides means of quantifying the relative weights of the concrete strength based on non-destructive test results for assessment of existing RC buildings.