The Mix design of normal weight concrete with artificial neural networks

IV SUMMARY Masters Thesis THE MIX DESIGN OF NORMAL WEIGHT CONCRETE WITH ARTIFICIAL NEURAL NETWORKS A. Tevfık BİLDİK Fırat University Graduate School of Natural and Applied Sciences Department of Civil Engineering 1998 Page 49 In recent years the artificial intelligence applications had been popular in computer aided engineering calculations. In this thesis the Artificial Neural Networks which is a artificial intelligence approach was used. The subject is the mix design of normal-weight concrete. Mix design methods subject to various standards were investigated and a data set was built using the Turkish Standard. A comparison had been made between the experimental results from the literature and the results of the trained data set. Good results had been found from this comparison. Also by training the data set about the fiber reinforced concrete which had been most popular in recent years, the slump values of fiber reinforced concrete were calculated. It was shown that the artificial neural networks had given suitable solutions to the problems which have ampirical expressions and which the existing solutions more than one. KEYWORDS : Artificial Neural Networks, Concrete, Fiber Reinforced Concrete, Mix Design