Concrete pillars, used as a structural support on the electrolysis process at zinc factories, are exposed to corrosive environmental conditions, due to sulfuric acid presence. In order to prevent irreversible damage to the involved structures, maintenance becomes of vital importance, while reducing costs is the main factor to be considered. Neural network as a model is a recently developed alternative to determine where and how the structural damage on concrete columns is taking place. This neural network model was fed by four-year maintenance registries data. Prediction showed that the most affected concrete components are those located near liquid zinc crucibles. Neural network model also helped to develop a more accurate preventive maintenance schedule and improving the annual repairs investment.
Corrosion prediction and annual maintenance improvement of concrete structural components using neural networks
2007-09-01
375093 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Corrosion prediction in aging aircraft materials using neural networks
Tema Archiv | 2000
|Corrosion Prediction in Aging Aircraft Materials using Neural Networks
British Library Conference Proceedings | 2000
|Aeroplane structural maintenance and corrosion prevention
Online Contents | 1996
|