NEURAL NETWORK ANALYSIS OF ASR OF CONCRETE INCORPORATED WITH CHEMICAL AND MINERAL ADMIXTURE

摘要

Alkali Silica Reaction (ASR) is one major causes of deterioration of concrete structure. A large amount of research has been conducted to study the ASR of concrete, both theoretically and experimentally. Because ASR experiments are time consuming, it is desired to develop a model to predict or evaluate the ASR. This paper studies the effects of calcium nitrite (CN), fly ash, microsilica and their combination on ASR according to ASTM C227 (mortar bar test). The feasibility of using a neural network as an ASR predictor and an adaptive synthesizer has been studied also. The cascade correlation algorithm is adopted in the present investigation to predict ASR. It has been found that the method has the potential of becoming an effective tool in evaluation of ASR durability problems.