The development of empirical chloride-induced corrosion rate prediction models for cracked and uncracked steel reinforced concrete structures in the marine tidal zone

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title

University of Cape Town

Empirical chloride-induced corrosion rate prediction models for cracked and uncracked reinforced concrete (RC) structures in the marine tidal exposure zone are proposed in this study. The data used to develop the models were obtained from parallel corrosion experiments carried out by exposing half of 210 beam specimens to accelerated laboratory corrosion (cyclic 3 days wetting with 5% NaCl solution followed by 4 days air-drying) while the other half were left to undergo natural corrosion in a marine tidal zone in Cape Town (Table Bay). The main experimental variables were pre-corrosion flexural cover cracking, cover depth and concrete quality (binder type and w/b ratio). Corrosion rate, half-cell potential and concrete resistivity were monitored bi-weekly throughout the experimental period. The experimental results show that even though each of the variables investigated affects corrosion rate in a certain manner, their combined influence is complex. In general, regardless of the exposure environment (laboratory or marine tidal zone), for a given concrete quality and cover depth, pre-corrosion cover cracking was found to result in higher corrosion rates than in uncracked concrete, but with the field corrosion rates being much lower than the corresponding laboratory ones. Even though corrosion rates in both the field and laboratory specimens increased with an increase in crack width, the influence of concrete quality and cover depth was still evident. However, the effect of cover cracking on corrosion rate diminished with increasing concrete quality. In the blended cement concretes, the effect of concrete quality is further diminished by the inherent high resistivities of these concretes. The increase in corrosion rate due to increase in crack width, regardless of w/b ratio and cover depth, was generally higher in the 100% CEM I 42.5N concrete specimens than in the blended ones. A framework is proposed that can be used to objectively compare predicted corrosion rates for specimens with similar concrete quality (influenced by binder type and w/b ratio) but different cover depths and crack widths. The framework, which incorporates the combined influence of cover depth, crack width and concrete quality (quantified using chloride diffusion coefficient) on corrosion rate, is the basis of the proposed corrosion rate prediction models for cracked concrete. Sensitivity analyses on the proposed models show that if any two of the three input parameters (cover depth, crack width and concrete quality) are simultaneously varied, their effect on corrosion rate is dependent on the value of the third (unchanged) parameter. Furthermore, (i) the initial cover depth was found to have no effect on the extent to which a change in cover depth affects corrosion rate; a similar trend was found in the case of sensitivity of corrosion rate to change in crack width , and (ii) the extent to which a change in either crack width or cover depth affects corrosion rate is dependent mainly on the concrete quality. In general, the sensitivity analyses showed that corrosion rate is more sensitive to change in concrete quality than crack width and cover depth. The proposed models can be used to (i) quantify the propagation phase with respect to a given performance limit using relevant corrosion-induced damage prediction models, and (ii) select suitable design combinations of cover depth, concrete quality and crack width to meet the desired durability performance of a given RC structure in the marine environment.

Includes bibliographical references.