Service Life Modeling of Virginia Bridge Decks

TR Number
Date
2007-03-20
Journal Title
Journal ISSN
Volume Title
Publisher
Virginia Tech
Abstract

A model to determine the time to the End of Functional Service Life (EFSL) for concrete bridge decks in Virginia was developed. The service life of Virginia bridge decks is controlled by chloride-induced corrosion of the reinforcing steel. Monte Carlo resampling techniques were used to integrate the statistical nature of the input variables into the model. This is an improvement on previous deterministic models in that the effect of highly variable input parameters is reflected in the service life estimations. The model predicts the time required for corrosion to initiate on 2% of the reinforcing steel in a bridge deck and then a corrosion propagation time period, determined from empirical data, is added to estimate the EFSL for a given bridge deck or set of bridge decks.

Data from 36 Virginia bridge decks was collected in order to validate the service life model as well as to investigate the effect of bridge deck construction specification changes. The bridge decks were separated into three distinct groups: 10 bare steel reinforcement decks â 0.47 water/cement (w/c), 16 Epoxy-Coated Reinforcement (ECR) decks â 0.45 w/c, and 10 ECR decks â 0.45 w/(c+pozzolan). Using chloride titration data and cover depth measurements from the sampled bridge decks and chloride corrosion initiation values determined from the literature for bare steel, service life estimates were made for the three sets of bridge decks. The influence of the epoxy coating on corrosion initiation was disregarded in order to allow direct comparisons between the three sets as well as to provide conservative service life estimates.

The model was validated by comparing measured deterioration values for the bare steel decks to the estimated values from the model. A comparison was then made between the three bridge deck sets and it was determined that bridge decks constructed with a 0.45 w/(c+p) will provide the longest service life followed by the 0.47 w/c decks and the 0.45 w/c decks, respectively. From this it can be inferred that the addition of pozzolan to the concrete mix will improve the long-term durability of a bridge deck while a reduction in w/c appears to be of no benefit.

Description
Keywords
bridge deck, corrosion, concrete, service life, life-cycle cost analysis, chloride diffusion
Citation