Predicting Deterioration of Bridge Decks Using BogdanofPs Cumulative Damage Model
Different types of models, including empirical, mechanistic, statistical, and logistic haven been proposed in the literature for the prediction of bridge deck deterioration (Golabi and Shepard 1997; Frangopol et al. 1997; Weyers 1998; Stewart and Rosowsky 1998; Lounis and Madanat 2002; Morcous et al. 2003). The prediction of the deterioration and service life of bridge decks is a difficult task due to the complexity of the mechanisms involved, such as the penetration of chlorides into concrete, onset of corrosion, damage initiation, damage accumulation, and coupling effects of corrosion and other deterioration factors (e. g. traffic load, initial damage, etc.). This deterioration prediction is further complicated by the considerable uncertainty in the governing parameters of the damage initiation and accumulation models, as well as in the uncertainty in the models themselves.
The fluctuations from mean performance and mean life are sufficiently large and cannot be ignored without serious consequences owing to the large fluctuations in the in-service environment, deck design, initial damage, etc. Hence, a probabilistic modeling of the damage accumulation and service life is required to achieve reliable results. Furthermore, these models should be developed by using the available data collected during the regular or special inspections of bridge structures in the inventory of highway agencies. These models can be updated as more data become readily available and/or by using data from long term in-service exposure, accelerated laboratory testing, or from the use of fundamental mechanistic models (Golabi and Shepard 1997; Lounis and Madanat 2002; Morcous et al. 2003).