Disaggregation Analyses of the External/Internal Parameters for Chloride Diffusion in Concrete Structures: Artificial Intelligence by k-means Algorithm
Résumé
In this paper, the parameters involved in the chloride ions diffusion for concrete structures have been analysed. Usually, several factors for diffusivity and surface chloride concentration are introduced to consider all internal and external actions that contribute to the diffusion. However, the inter-correlations between them are neglected. This paper treats these inter-correlations. From experimental laboratory tests about 850 parameters and values have been filtered and collected. These values have been divided in 11 categories and combined with each other. More of 120 numerical analyses by using k-means algorithm have been carried out. This algorithm is considered an artificial intelligence (AI) to support the human limitations in managing and analysing an enormous quantity of parameter. Results provide disaggregated values where each parameter is correlated with other one. This allows to understand their weight and effects on chloride ions diffusion. This paper provides not only a new approach but also practical values for more accurate analyses.