Title:
Phenomenological Model of Corrosion Process in Reinforced Concrete Identified by Acoustic Emission
Author(s):
Masayasu Ohtsu and Yuichi Tomoda
Publication:
Materials Journal
Volume:
105
Issue:
2
Appears on pages(s):
194-199
Keywords:
acoustic emission; chloride concentration; corrosion; half-cell potential.
DOI:
10.14359/19764
Date:
3/1/2008
Abstract:
The corrosion process of reinforcing steel bar (reinforcing bar) in concrete is modeled phenomenologically as corrosion loss of three phases. At the first phase, corrosion initiates in reinforcing bar. Then the rate of the corrosion loss decreases at the second phase under aerobic conditions. At the third phase of anaerobic corrosion, expansion of the reinforcing bar due to corrosion products nucleates concrete cracking. To identify such a corrosion process as the corrosion loss in reinforced concrete, continuous acoustic emission (AE) monitoring was conducted in an accelerated corrosion test and a cyclic wet and dry test. Two periods of high AE activities were characteristically observed. To elucidate these two activities, AE waveform parameters and the slope gradient of AE amplitude distribution were investigated. At the first period of high activity, AE events were found to be of small amplitudes and cracking mechanisms were classified as the other-type cracks than a tensile crack. Although no corrosion products were visually observed on the surface of reinforcing bar, examination of the reinforcing bar surface by a scanning electron microscope (SEM) demonstrated that ferrous ions vanished at some regions of the surface. This suggests that the first high AE activity results from the onset of corrosion in the reinforcing bar, corresponding to the first phase in the phenomenological model. Corresponding to the second phase of the model where the rate of the corrosion loss decreases, AE activities become low. Then, following the second period of high AE activity, corrosion products were visually observed on the reinforcing bar surface. It suggested the nucleation of cracks in concrete around the reinforcing bar. Reasonably, AE sources were classified as tensile cracks and amplitudes of AE events were found to be fairly large. These AE events are associated with concrete cracking due to expansion of corrosion products that could be observed at the third phase of the model. Thus, the phenomenological model of the corrosion loss is in remarkable agreement with AE generating behaviors in reinforced concrete. By monitoring AE activities, the time of corrosion in reinforced concrete could be identified at the initiation of corrosion in the reinforcing bar and also at the nucleation of cracks in concrete. Even the latter AE activity is detected earlier than other nondestructive evaluation (NDE) techniques.