Study of the Workability of Self-Compacting Concrete (SCC) Using Experimental Methods and Artificial Neural Networks (ANN)

Resumen

The self-compacting concrete (SCC) flows under its weight and does not require external vibration for compaction. However, its formulation requires careful calculation of its constituents. Three methods are considered: the first is an empirical method represented by an approach based on mortar optimization, a solution proposed by Japanese researchers who originally introduced the concept of self-compacting concrete; the second is a graphical method by Dreux-Gorisse used for ordinary concrete, which optimizes the composition of the aggregate skeleton by selecting fractions without additives and superplasticizers; and the third is a statistical method that we developed using an approach based on Artificial Neural Networks (ANN) built from a database from previous research projects. The objective is to characterize workability through an ANN model and compare it with experimental methods. Therefore, we focused on the slump flow, L-box, and sieve stability segregation tests.