Models for predicting corrosion-induced cracking, spalling/delamination of rehabilitated RC bridge decks

The corrosion of reinforcing steel in reinforced concrete (rc) bridge decks due to the application of de-icing salts in winter has been recognized as one of the major causes of highway bridge deterioration in North America. Corrosion-induced damage is usually manifested by longitudinal cracking, spalling, and/or delamination of the concrete cover due to the expansion of corrosion products accumulating around the reinforcement. This damage leads to loss or reduction of serviceability, safety, and service life of rc bridge decks. This research presents an analytical model that describes the mechanism of damage initiation and accumulation to predict corrosion-induced cracking, spalling and delamination of rc bridge decks. This model assumed the concrete deck cover acts as a thick-wall cylinder subjected to a uniform internal pressure due to the expansion of corrosion products. The prediction of the damage caused by corroding reinforcing bars was established by calculating the induced stresses in the surrounding concrete. Corrosion-induced damage of rc bridge decks was also investigated using the finite element method. For this purpose, the commercial finite element package ABAQUS/Standard was used (Hibbit Karlsson & Sorensen Inc. 2004). The finite element analysis was to simulate three main failure modes of corrosion-induced damage of rc bridge decks: longitudinal cracking, delamination and spalling of the concrete. In this study, a finite element model that was independent to parameters defining boundary locations was used. Based on this model, parametric investigations of the influence of various design variables on the concrete deck failure modes were conducted for both unprotected and overlaid rc bridge decks. Finally, propagation models were proposed by comparing the analytical models with the results obtained from the finite element modelling. These propagation models are formulated by correcting the analytical models with factors obtained from the comparison of both sets of results.