Guido Roelfstra
The management of existing road infrastructures is a multidisciplinary activity that involves structural engineering, material science, management, economics and ecology. The objective is to achieve maximum availability of road links at minimum societal costs. Recently, tools (Bridge Management Systems, BMSs) have been developed to help decision makers to determine the optimal management strategies within available resources. The condition development model is a key element of the BMSs. Currently this model is using empirical approach making use of data collected during inspections and is formulated with Markov chains. The assessment units are structural elements and these are classified into condition classes according to their visual appearance. The condition assessment is therefore rather subjective since no measurements are involved in it. Consequently, the condition development based on visual inspections will reflect this subjectivity. In order to improve the objectivity of condition forecasts physical and chemical phenomena governing deterioration has to be incorporated in the condition development model. Observations made on a bridge sample representative of the Swiss highway bridges showed that chloride-induced corrosion is the main deterioration mechanism. For this reason this research was focused upon modeling the chloride-induced corrosion. In this research, the condition development model is established, based physical and chemical phenomena involved in chloride induced corrosion. This analytical model is able to yield quantitative results on condition development. Accuracy and availability of input data govern the level of accuracy of such an analytical model. The existing knowledge on chloride induced corrosion and inspection techniques available on site are examined in order to develop a simple and reliable condition development model. Firstly, non-destructive testing methods are used in addition to visual observations, to determine quantitative values (permeability and thickness of the concrete cover) of parameters governing the segment deterioration. The structural elements are divided into areas i.e. segments where the parameters governing chloride-induced corrosion are assumed to be constant. It is proposed to divide structural elements into segments based on three concrete cover permeability classes, concrete cover thickness and three types of surface exposures to chloride contaminated water. Secondly, a chloride transport model is developed to simulate the real exposure of reinforcement to chloride contaminated water, taking into consideration the transport mechanisms of water and chloride ions. Simulations are performed for the various segment parameters that have been identified in the segmental approach. Thirdly, a relation is established between the free chloride ion content and the corrosion initiation probability. Using the results of the chloride transport model, the development of the corrosion initiation is determined for road bridges segments. An analysis is made to predict the corrosion initiation time of existing bridges as a function of their segment parameters (exposure and concrete cover permeability and thickness), but also to establish recommendations for the construction of new bridges. Fourthly, the corrosion rate i.e. loss of steel cross-section is then determined according to the concrete cover permeability and the exposure to chloride contaminated water. Fifthly, the results of the presented mode1 are mapped to conditions States as defined in the BMSs. Three degradation matrices (Markov chains) are determined, corresponding to three condition developments: favorable, normal and unfavorable. The three condition developments can be estimated for given quantitative values of the concrete cover permeability, thickness and surface exposure to chloride-contaminated water. In this way, the segment parameters may be used in BMSs to forecasts future condition of segments, elements and the whole structure. Finally the structural behaviour of the concrete road-bridges until failure is studied. The determinant failure scenarios are identified for each bridge type and a case study is performed on the sample representative for the Swiss highway bridges. For this purpose the determinant segments for the structural safety are identified. Additionally the required concrete cover depth for a service life of 50 and 100 years is expressed as a function of the permeability and the exposure to chloride contaminated water. This research outlines the need to measure the concrete cover permeability and thickness with non-destructive test methods as well as to determine the exposure to chloride contaminated water. These values are essential to predict the condition development of concrete road bridges.
EPFL2001
