Strength investigation of tannic acid-modified cement composites using experimental and machine learning approaches

被引:4
作者
Li, Ning [1 ,2 ,3 ,4 ]
Kang, Ziye [1 ]
Zhang, Jinrui [1 ,4 ]
机构
[1] Tianjin Univ, Sch Civil Engn, Tianjin 300350, Peoples R China
[2] Tianjin Univ, Key Lab Coast Civil Struct Safety, Minist Educ, Tianjin 300350, Peoples R China
[3] Tianjin Univ, Key Lab Earthquake Engn Simulat & Seism Resilience, China Earthquake Adm, Tianjin 300350, Peoples R China
[4] Tianjin Univ, State Key Lab Hydraul Engn Intelligent Constructio, Tianjin 300350, Peoples R China
基金
中国国家自然科学基金;
关键词
Tannic acid; Cement composites; Compressive strength; Machine learning; SHapley Additive exPlanations; COMPRESSIVE STRENGTH; SILICA FUME;
D O I
10.1016/j.conbuildmat.2024.135684
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
The application of tannic acid (TA) as reinforcement material in cement composites can effectively improve its sustainable development. Nevertheless, the efficacy of TA can be influenced by multiple factors, thereby impeding its practical utilization. This study primarily examines the effects of curing regimes on mechanical properties. It is found that the optimal condition is achieved at 90 degree celsius water curing. Under this regime, the addition of 0.025% TA results in a notable enhancement of 34.6% in flexural strength and 11.4% in compressive strength at 3 days. Then, this study establishes four models - RF, XGB, SVR, and ANN - to predict the compressive strength of TA-modified cement composites. The outcomes highlight that the XGB model has superior accuracy and generalization capability. It achieves a high R2 of 0.92 on the test set and an average R2 of 0.83 in a 10-fold cross-validation. Furthermore, the SHapley Additive exPlanations analysis reveals that TA ranks as the third most significant influencing factor, following the age and water content of the composites. And variables like age, nano silica, silica fume, and TA demonstrate positive correlations. They contribute positively to the development of compressive strength of the materials.
引用
收藏
页数:13
相关论文
共 48 条
[1]   A scientometric review of waste material utilization in concrete for sustainable construction [J].
Ahmad, Waqas ;
Ahmad, Ayaz ;
Ostrowski, Krzysztof Adam ;
Aslam, Fahid ;
Joyklad, Panuwat .
CASE STUDIES IN CONSTRUCTION MATERIALS, 2021, 15
[2]   Random forest in remote sensing: A review of applications and future directions [J].
Belgiu, Mariana ;
Dragut, Lucian .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2016, 114 :24-31
[3]   Random forests [J].
Breiman, L .
MACHINE LEARNING, 2001, 45 (01) :5-32
[4]   The effect of carbon nanotubes and silica fume on compressive strength and flexural strength of cement mortars [J].
Chaipanich, Arnon ;
Rianyoi, Rattiyakorn ;
Nochaiya, Thanongsak .
MATERIALS TODAY-PROCEEDINGS, 2017, 4 (05) :6065-6071
[5]   Data-driven based estimation of waste-derived ceramic concrete from experimental results with its environmental assessment [J].
Chang, Qiuying ;
Liu, Lanlan ;
Farooqi, Muhammad Usman ;
Thomas, Blessen ;
Ozkilic, Yasin Onuralp .
JOURNAL OF MATERIALS RESEARCH AND TECHNOLOGY-JMR&T, 2023, 24 :6348-6368
[6]   Evolution of full-length HBV sequences in chronic hepatitis B patients with sequential lamivudine and adefovir dipivoxil resistance [J].
Chen, Chien-Hung ;
Lee, Chuan-Mo ;
Tung, Wei-Chih ;
Wang, Jing-Houng ;
Hung, Chao-Hung ;
Hu, Tsung-Hui ;
Wang, Jyh-Chwan ;
Lu, Sheng-Nan ;
Changchien, Chi-Sin .
JOURNAL OF HEPATOLOGY, 2010, 52 (04) :478-485
[7]   Machine learning for high-fidelity prediction of cement hydration kinetics in blended systems [J].
Cook, Rachel ;
Han, Taihao ;
Childers, Alaina ;
Ryckman, Cambria ;
Khayat, Kamal ;
Ma, Hongyan ;
Huang, Jie ;
Kumar, Aditya .
MATERIALS & DESIGN, 2021, 208
[8]   A novel approach to explain the black-box nature of machine learning in compressive strength predictions of concrete using Shapley additive explanations (SHAP) [J].
Ekanayake, I. U. ;
Meddage, D. P. P. ;
Rathnayake, Upaka .
CASE STUDIES IN CONSTRUCTION MATERIALS, 2022, 16
[9]   Densifying hydration products of alite by a bio-inspired admixture [J].
Fang, Yi ;
Wang, Jialai ;
Wang, Liang ;
Qian, Xin ;
Wang, Xiaodong ;
Liao, Wenyu ;
Chen, Peiyuan ;
Ma, Hongyan .
MATERIALS & DESIGN, 2023, 225
[10]   A renewable admixture to enhance the performance of cement mortars through a pre-hydration method [J].
Fang, Yi ;
Wang, Jialai ;
Qian, Xin ;
Wang, Liang ;
Chen, Peiyuan ;
Qiao, Pizhong .
JOURNAL OF CLEANER PRODUCTION, 2022, 332