Title:
Improved Electrochemical Determinations of
Chloride-Induced Steel Corrosion in Concrete
Author(s):
Daksh Baweja, Harold Roper, and Vute Sirivivatnanon
Publication:
Materials Journal
Volume:
100
Issue:
3
Appears on pages(s):
228-238
Keywords:
blast-furnace slag; corrosion; fly ash
DOI:
10.14359/12624
Date:
5/1/2003
Abstract:
It is clear that the accurate measurement of corrosion rates of steel in concrete is a developing science. Typically, estimations of service life of concrete structures and concrete performance are based on methods that measure concrete properties alone, an example being chloride diffusion. The problem with such techniques is that no direct account is taken for steel corrosion—a key factor influencing service life. Techniques like polarization resistance that directly measure corrosion rates offer much greater potential in reducing uncertainty in predicting service life and other fit-for-purpose criteria when compared with techniques that focus solely on concrete properties alone. In this study, corrosion rates of steel in concrete were investigated by examining long-term data on concrete slabs partially immersed in 3% NaCl (simulated seawater) solution. Using polarization resistance technique (corrosion current) Icorr values were determined from the polarization resistance Rp data using a range of Stern-Geary Constant B values. For high values of Rp corresponding to low corrosion activity, differences in B had little effect on the value of Icorr; but for low values of Rp , corresponding to active corrosion conditions, differences in B produced significant changes in calculated Icorr values. A model was developed that described the behavior of the reinforced concrete slabs partially immersed in simulated seawater solution in terms of steel corrosion rates. Data were divided into the three stages of corrosion previously described in the literature: the Quiescent Stage, the Active Stage, and the Breakaway Stage. Relationships between Ecorr and Icorr within a range of reinforced concrete were analyzed with a view to developing improved methods for deriving steel corrosion rates. Using polarization resistance techniques, relationships between Ecorr and Icorr were investigated, this being used as a criteria to model steel corrosion rates. Using a B value of 26 mV, widely reported in the literature, correlations between Ecorr and Icorr , although significant, showed some scatter. Better relationships were obtained when B values were calculated from Ecorr data using equations developed in this study. Improved correlations to steel weight loss through corrosion were also obtained using this procedure than those obtained using a B value of 26 mV. It is concluded that more accurate estimates of Icorr can be made when B values are computed from Ecorr results rather than adopting a single B value as described extensively in the literature.