Transcripción del resumen del autor. Corrosion severity prediction is very useful information for maintenance planning. Acoustic emission (AE), a non-destructive testing method, can be applied to monitor corrosion severity. To improve its prediction reliability, a novel technique of applying acoustic source recognition is presented. In different acidic solutions and static corrosive potentials, stainless steel AISI 304 (UNS S30400) was used to study pitting corrosion, and carbon steel A36 (UNS K02600) was used to consider uniform corrosion. Consequently, each acoustic source was identified and used to predict the corrosion rate. In addition, the frequency spectrum and average frequency distribution of both corrosion types were studied. In conclusion, the frequency range can be used to classify types of corrosion. Either direct or indirect acoustic sources can be selected to monitor the corrosion severity. This method successfully increases the signal-to-noise ratio.

5