摘要
气泡轻质土为新型回填材料,可能受交通荷载和机器振动等动荷载作用。目前关于动荷载对气泡轻质土强度的影响缺乏统一认识,易造成设计不合理。开展循环荷载作用下气泡轻质土无侧限抗压强度试验,探讨密度、振动频率和循环次数对气泡轻质土动强度的影响。研究结果表明,气泡轻质土在循环荷载作用下动强度明显低于静强度,这可能因为其内部结构在动荷载作用下更易发生破坏。采用基于人工神经网络的机器学习方法建立气泡轻质土动强度预测模型,经模型训练后,预测值与试验数据达到良好的一致性。
Foamed concrete is a new type of backfill material,may be subjected to cyclic loading,such as traffic loading and machine vibration.The effect of dynamic loading on the strength of foamed concrete does not reach a consensus,which may cause an inappropriate design.This paper conducted unconfined compressive strength test of foamed concrete under cyclic loading and discussed the influences of density,loading frequency,and loading times on the strength of foamed concrete.The results show that the cyclic loading yielded lower strength of foamed concrete than static loading,as its internal structure is easier to fail under cyclic loading.Artificial neural network-based machine learning method was used to establish a model to predict the dynamic strength of foamed concrete.Good agreements between predictions and the real values were obtained after training the model using the test results.
作者
叶观宝
刘江婷
张振
饶烽瑞
陈佳祺
张甲峰
YE Guanbao;LIU Jiangting;ZHANG Zhen;RAO Fengrui;CHEN Jiaqi;ZHANG Jiafeng(College of Civil Engineering,Tongji University,Shanghai 200092,China;School of Mathematical Sciences,Tongji University,Shanghai 200092,China;Shanghai CAAC New Era Airport Design&Research Institute Co.,Ltd.,Shanghai 200335,China)
出处
《施工技术》
CAS
2020年第9期50-53,共4页
Construction Technology
基金
国家自然科学基金(41972272
41772281
51508408)
中央高校基本科研业务费专项资金(22120180106)。
关键词
气泡轻质土
抗压强度
循环荷载
BP神经网络
预测模型
foamed concrete
compressive strength
cyclic loading
BP neural network
prediction models
