| 要旨 |
This paper presents a study on the parameters influencing the maximum width of early-age thermal cracks in massive RC abutments using neural networks. The parametric studies are based on a well-trained neural network which was trained using Yamaguchi prefecture dataset. The effect of unit cement content, width, thickness, lift height, reinforcement ratio, lift interval, ambient temperature and initial concrete temperature was studied. The results have shown the potential of neural networks to extract the essences of construction data and proposing recommendations to mitigate harmful thermal cracks. |