摘要
基于BP神经网络，建立了UHPC力学性能的预测模型，并利用相关试验数据验证模型的有效性。采用该模型系统研究了水胶比和聚羧酸减水剂掺量对于UHPC力学性能的影响，结果表明：在0．136～0．225的范围内，水胶比对UHPC力学性能的影响不显著，而且减水剂掺量越小，它的影响越不显著；聚羧酸减水剂掺量对UHPC力学性能的影响极为显著，掺量在0．7％左右的UHPC的力学性能最优。进一步通过试验验证了水胶比和聚羧酸减水剂掺量对UHPC力学性能的影响规律。对于今后UHPC的试配，在试验取得一定数据样本之后，能够通过该模型试验代替部分试配试验来减少试验工作量，同时为UHPC配合比设计和优化提供指导。
Based on BP Neural Network, a model for predicting the mechanical behavior of ultra high performance concrete was established.And the model was verified by the relative test data.Afterwards, the effects of the water-binder ratio and polycarboxylic acid type superplasticizer on the mechanical behavior of ultra high performance concrete were studied by the model test.The results showed that there was no significant influence of water-blinder ratio on mechanical properties of ultra high performance concrete in the range of 0.136 to 0.225.And the less superplasticizer was added, the less effect of water-blinder was taken.While the polycarboxylic acid type superplasticizer influenced the mechanical properties of ultra high performance concrete greatly.In term of mechanical behavior,the optimal dosage was about 0.7%.What was more,it verified the influence law of water-binder ratio and polycarboxylic acid type superplasticizer on the mechanical behavior of ultra high performance concrete by experimental tests.For the future designing of mix proportion, once enough simple data was obtained,model test could substitute for experimental test.In this way,the workload would be reduced and the guidance for design and optimization of ultra high performance concrete mix proportion will be provided.
出处
《混凝土》
CAS
CSCD
北大核心
2013年第3期18-22,25,共6页
Concrete
基金
国家自然科学基金(50908167)
关键词
BP神经网络
UHPC
力学性能
模型
BP neural network
ultra high performance concrete
mechanical behavior
model
