Bonding behavior of concrete matrix and alkali-activated mortar incorporating nano-SiO2 and polyvinyl alcohol fiber: Theoretical analysis and prediction model. Issue 22 (15th November 2021)

• Record Type: Journal Article
• Authors: Gao, Zhen; Zhang, Peng; Guo, Jinjun; Wang, Kexun
• Is Part Of: Ceramics international. Volume 47:Issue 22(2021)
• Page Start: 31638
• Page End: 31649
• Publication Date: 2021-11-15
• Subjects: Bonding behavior -- Alkali-activated mortar -- Grey relation analysis -- Artificial neural network -- Weight contribution algorithm; Ceramics -- Periodicals; Céramique industrielle -- Périodiques; Ceramics; Periodicals; Electronic journals; 666
• DOI: 10.1016/j.ceramint.2021.08.044
• Languages: English
• ISSNs: 0272-8842

Abstract

Abstract: As an emerging construction material, alkali-activated mortar is considered as a sustainable alternative to cementitious composites for the repairing and reinforcement of existing defective buildings. Furthermore, the bonding performance of alkali-activated mortar and concrete matrix can be promoted by adding polyvinyl alcohol (PVA) fiber and nano-SiO2 (NS). In this study, the effects of PVA fiber and NS contents, alkali-activated mortar type, concrete strength grade, and interfacial roughness on the bonding behavior of two-interfaced shear samples were explored. Based on the experimental results, the grey relation analysis was applied to evaluate the significance of each factor on the bond properties of the alkali-activated mortar and concrete matrix. A prediction model of artificial neural network (ANN) was established considering the effects of alkali-activated mortar type, concrete strength grade, and interfacial type on the bond strength of the samples. The relevant factors affecting bond strength derived by grey relation analysis and weight contribution algorithm was compared and analyzed. Results of the two-interfaced shear test showed that the addition of PVA fiber and NS can significantly boost the bonding property of the samples, and the bond strength increases with the increase of concrete strength grade, alkali-activated mortar strength, and interfacial roughness. Grey relation analysis results indicate that the interfacial type has the most noticeable effect on the bond strength of the samples, followed by the concrete strength grades and the alkali-activated mortar types. The optimum bond strength is derived from PN-C40-III, which is alkali-activated mortar with 0.6% PVA fiber and 1.0% NS contents, concrete strength grade of C40, and interface of type III. The prediction results of the ANN indicate that the predicted values of the bond strengths of the samples are consistent with the experimental values (R = 0.982), and the importance of each factor towards the bond behavior derived by the grey relation analysis and weight contribution algorithm is ultimately consistent after normalization.