Modeling the Carbonation Front of Concrete Structures in the Marine Environment through ANN

Control of the carbonation depth in reinforced concrete structures, as well as its prediction, has great relevance for the construction industry, since this process is directly related to the service life and durability of these structures. One of the challenges in carbonation modelling consists in understanding the complex relation between the governing parameters of the phenomenon, especially when dealing with structures located in marine environments. Thus, in the present work, the concrete carbonation was modelled through Artificial Neural Networks. Altogether, 91 Multi-Layer Perceptron networks were implemented. Results found in this work indicate the potential of RNAs in light of modelling concrete carbonation depth and that the obtained model properly describes the influence of CO
₂
action over the concrete structure relative to its distance to the ocean.