Chemical acoustic emissions from gas evolution processes recorded by a piezoelectric transducer

The information of frequency domain power spectra obtained from acoustic emission (AE) signals by fast fourier transform (FFT) program for different gases evolution processes are investigated. The different processes include different metals in dilute inorganic acids for evolving hydrogen gases, different carbonate/bicarbonate solutions dropping into dilute inorganic acids for evolving carbon dioxide gases, bubbling nitrogen into above-mentioned salt solutions of corresponding reaction products and the above-mentioned salt solutions background without gas evolution. Averaged power spectra from these systems and their corresponding characteristics are examined. The characteristic frequencies for a given system are found to be more or less reproducible, while the corresponding maximum power intensities are varied to a large extent but still kept up to the same order of magnitude. The AE signals can be used for chemical pattern recognition. The principal component analysis (PCA) with appropriate frequency selection procedure can give satisfactory classification of AE signals for different gas evolution systems. The back-propagation algorithm of multilayer feed-forward network can give better recognition of AE signals for different metal-acid buffer solution reaction processes. In addition, the AE energetic parameters are linearly related with solution pH, which opens some new possibilities for analytical application of AE signals. (C) 1998 Elsevier Science S.A. All rights reserved.