Resumen

GONZALEZ SALCEDO, Luis Octavio; GUERRERO ZUNIGA, Aydée Patricia; DELVASTO ARJONA, Silvio  y  ERNESTO WILL, Adrián Luis. EXPLORING ARTIFICIAL NEURAL NETWORKS TO ESTIMATE COMPRESSIVE STRENGTH OF STEEL FIBER-REINFORCED CONCRETE. Cienc. Ing. Neogranad. [online]. 2012, vol.22, n.1, pp.19-41. ISSN 0124-8170.

ABSTRACT By designing and building concrete structures, the compressive strength achieved at 28-day curing typically represents the stability control specification of any work. Furthermore, reinforcing fibers into the cement based matrix has allowed a gain to their properties, as well as a high performance material. Technical literature states predictive formulations of compressive strength of concrete in function of a few composition parameters, such as water/cement ratio and the Portland cement. Also, there are formulations to find the proportion of the raw materials to get a defined compressive strength, specifically non-reinforced ordinary concrete. Besides artificial neural networks as a metaphor of biological neurons have been used as a tool to predict concrete compressive strength. The experience in this application shows an increasing interest to develop applications using fiber-reinforced concrete. In this paper, an artificial neural network has been developed to predict the compressive strength of steel-fiber-reinforced-concrete. The results prove that developed artificial neural networks may perform an adequate approximation to the actual value of the mechanical property.

Palabras clave : compressive strength; fiber-reinforced concrete; steel fiber; prediction; artificial intelligence; artificial neural networks.

        · resumen en Español     · texto en Español     · Español ( pdf )