首页|机器学习模型评估裹浆改性再生骨料形态特征及分布规律

机器学习模型评估裹浆改性再生骨料形态特征及分布规律

扫码查看
再生骨料(RA)的高吸水率、高压碎值、表面微裂纹及其不规则形态特征等缺陷限制了 RA的再生利用.裹浆改性再生骨料(SRA)表面形成的壳膜结构不仅能够提高RA的力学性能,且对RA的形态特征有着显著影响.为了评估裹浆改性对RA形态特征的影响,本文采用粉煤灰基地聚物浆体对再生混凝土骨料(RCA)和再生砖骨料(RBA)进行裹浆改性,并通过图像处理技术分别获取并建立裹浆改性前后RCA和RBA两种常见RA的形态特征数据集.此外,文中还结合机器学习技术提取数据集中的关键信息,从而量化RA的形态特征.经过定量分析发现,与改性前相比,改性后的RA轴向系数、棱角度和球形度的分布范围均有不同程度的改善,其中裹浆改性对棱角度改善效果最为明显,裹浆改性再生混凝土骨料(SRCA)最大改善幅度达132.2%,裹浆改性再生砖骨料(SRBA)最大改善幅度达69.2%.
Morphology Characteristics and Distribution Patterns of Slurry-Modified Recycled Aggregate Evaluated by Machine Learning
The defects such as high water absorption,high pressure crushing value,surface microcracks and irregular morphology of recycled aggregate(RA)limit its recycling and utilization.The shell film structure formed on the surface of slurry-modified recycled aggregate(SRA)not only improves the mechanical properties of RA,but also has a significant impact on morphological characteristics of RA.In order to evaluate the effect of slurry-modified on morphological characteristics of RA,this study used fly ash based polymer slurry to modify recycled concrete aggregates(RCA)and recycled brick aggregates(RBA),and obtained and established datasets of morphological characteristics of RCA and RBA before and after slurry-modified using image processing technology.In addition,the article also combined machine learning techniques to extract key information from the dataset,thereby quantifying the morphological features of RA.After quantitative analysis,it is found that compared with before modification,the distribution range of axial coefficient,angularity,and sphericity of modified RA has been improved to varying degrees.Among them,slurry-modification has the most significant effect on improving angularity.The maximum improvement amplitude of slurry-modified recycled concrete aggregate(SRCA)is 132.2%,and the maximum improvement amplitude of slurry-modified recycled brick aggregate(SRBA)is 69.2%.

recycled aggregateslurry-modificationquantification of morphological characteristicdistribution of morphological characteristicmachine learning

张嘉豪、陈正发、宋艳、陈昭言

展开 >

常州大学城市建设学院,常州 213164

常州大学材料科学与工程学院,常州 213164

武汉昭舍装饰设计有限公司,武汉 430070

再生骨料 裹浆改性 形态特征参数量化 形态特征参数分布 机器学习

常州大学人才引进项目

ZMF18020312

2024

硅酸盐通报
中国硅酸盐学会 中材人工晶体研究院

硅酸盐通报

CSTPCD北大核心
影响因子:0.698
ISSN:1001-1625
年,卷(期):2024.43(7)
  • 7