Comprehensive study to accurately predict the water permeability of pervious concrete using constant head method. (15th November 2021)

• Record Type: Journal Article
• Authors: Abdelhady, Ahmed; Hui, Li; Zhang, Hengji
• Is Part Of: Construction & building materials. Volume 308(2021)
• Publication Date: 2021-11-15
• Subjects: Constant head method -- Porosity -- Prediction -- Pervious concrete -- Porous concrete; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2021.125046
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: Previous models didn't perform well in predicting the permeability ( Kc ) of PC. W/C, A/C had not a significant effect on Kc when using a large database. Effective porosity of PC has a key effect on Kc with nonlinear shape. Power form exhibited a good accuracy to predict Kc of PC. Abstract: The pervious cement concrete (PC) or Permeable concrete is a special type of concrete with no or little sand, which exhibited an excellent hydrological function in decreasing the water runoff. However, accurate prediction of its water permeability coefficient using the Constant Head Method (CHM) is still unreliable due to the previous models have been established using a small size of PC samples, also these models have not been evaluated so far to understand its behavior in a large database. Thus, the authors think that the convenient time to propose a reliable model to predict the water permeability of PC, especially in the existence of large experimental PC Samples in literature. In this study, the experimental data of PC from (2016–2021) that was conducted in the previous work has been collected to build a database of the PC samples, which calculated the permeability coefficient using CHM. This effort provides about 695 PC samples from 96 studies. Moreover, the mathematical models that were proposed in the literature (from 1995 to 2021) to predict the water permeability of PC, also have been collected and classified into five groups based on the mathematical forms. The collected models were 41 models from 30 studies during this period, which have been evaluated using the constructed database based on some statistical criteria to nominate a consistent model among them. From the analysis of results, all the examined models failed to significantly predict the Kc. Thus, a new reliable model has been proposed using Power forms, which exhibited a good relationship with the porosity of PC.