欢迎访问《硅酸盐通报》官方网站,今天是

硅酸盐通报 ›› 2017, Vol. 36 ›› Issue (7): 2447-2452.

• • 上一篇    下一篇

基于正交设计与BP神经网络优化制备钢渣代砂环保型泡沫混凝土

陈华;李辉;顾恒星;杨刚;陈伟   

  1. 西安建筑科技大学材料与矿资学院,西安 710055;中冶宝钢技术服务有限公司,上海 201999;西安建筑科技大学材料与矿资学院,西安,710055;中冶宝钢技术服务有限公司,上海,201999
  • 出版日期:2017-07-15 发布日期:2021-01-18
  • 基金资助:
    国家自然科学基金(50872105)

Optimizing Preparation of Environmental Protection Foam Concrete with Steel Slag as Fine Based on Orthogonal Design and Back-propagation Neural Network

CHEN Hua;LI Hui;GU Heng-xing;YANG Gang;CHEN Wei   

  • Online:2017-07-15 Published:2021-01-18

摘要: 以特殊钢尾渣作为掺和料,制备钢渣代砂环保型泡沫混凝土,即特殊钢尾渣泡沫混凝土.基于正交设计与BP神经网络考察各制备因素对特殊钢尾渣泡沫混凝土干密度与28 d强度的影响.结果表明,最优特殊钢尾渣泡沫混凝土的制备工艺参数:水泥用量74.9份、特殊钢尾渣用量30.2份、粉煤灰用量12.8份、水料比0.455、发泡剂用量309.7 g,其干密度628.49 g/cm3和28 d强度2.675 MPa.所建立的BP神经网络模型精确性高,即实验测试值与模型预测值的相对误差分别为3.854%与3.925%.特殊钢尾渣泡沫混凝土中C-S-H凝胶将特殊钢尾渣包裹,不仅能提高所制备泡沫混凝土的力学性能,而且能增强所制备混凝土中的应用性和安全性.

关键词: 泡沫混凝土;特殊钢尾渣;环保;正交设计;BP神经网络;优化制备

Abstract: Environmental protection foam concrete with steel slag as fine(special steel tailings foam concrete)was prepared with special steel tailings as admixture.The effects of every preparation factor on dry density and 28 d strength of special steel tailings foam concrete were studied by orthogonal design and back-propagation neural network.The results show preparation process parameters of special steel tailings foam concrete is amount of cement content 74.9 part,amount of special steel tailings content 30.2 part,amount of fly ash 12.8 part,ratio of water to material 0.455 and amount of foaming agent 309.7 g,its dry density 628.49 g/cm3 and 28 d strength 2.675 MPa.Established the BP neural network model accuracy is high,such as relative error of experimental results and model calculation is only 3.854%and 3.925%.Special steel tailings were wrapped in special steel tailings foam concrete,not only improved the mechanical properties of special steel tailings foam concrete,but also enhanced the applicability and safety of special steel tailings.

Key words: foam concrete;special steel tailings;environmental protection;orthogonal design;back-propagation neural network;optimizing preparation

中图分类号: