Effect of polymer binder on the mechanical and microstructural properties of pervious pavement materials. (28th March 2022)

• Record Type: Journal Article
• Authors: Shin, Yoonjae; Park, Hyeong Min; Park, Jinha; Cho, Hongdong; Oh, Seo-Eun; Chung, Sang-Yeop; Yang, Beomjoo
• Is Part Of: Construction & building materials. Volume 325(2022)
• Publication Date: 2022-03-28
• Subjects: Pervious pavement materials -- Polyurethane -- Epoxy -- Micro-CT analysis -- Machine learning; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2021.126209
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: Pervious pavements composed of polyurethane and epoxy binders were investigated. Mechanical properties and water permeability of specimens were measured. The internal structures were quantitatively analyzed through a micro-CT analysis. An ML was applied to develop a model that can predict the material performance. Abstract: Pervious pavements materials (PPMs) are prospective functional materials with the potential to make urban areas more environmentally friendly. In this study, the effects of polymer binders on the mechanical and microstructural properties of PPMs are analyzed experimentally and through machine learning. Polyurethane and epoxy, which are widely used due to their reasonable cost and high performance, are considered as polymer binders, and specimens with different aggregate size distributions are fabricated using these polymers. The mechanical properties of the specimens are analyzed using compressive and flexural strength tests, freeze–thaw durability tests, and water permeability tests. Then, the internal microstructure is characterized using micro-computed tomography, including the pore size distribution, pore sphericity, anisotropic ratios of the pores, pore tortuosity, and aggregate sphericity. The correlation between the mechanical properties and the pore structure is analyzed. The aggregate size distribution and the polymer viscosity affects the size of the internal pores, and thus the mechanical properties. Finally, machine learning is used to develop a model that can predict the microstructural properties and compressive strength of PPMs according to the aggregate size distribution and the properties of the polymer. The model was validated using experimental data, and the methods used in its creation could be used to derive a general model for PPMs.