Abstract
The functional performance of pervious concrete pavement surfaces (such as hydraulic, acoustic, and frictional performances) is greatly influenced by the properties of its internal pore structure (such as effective porosity, intrinsic permeability, tortuosity, and pore size distribution). Nondestructive evaluation (NDE) using X-ray computed tomography (CT) and digital image processing (DIP) involves the crucial step of image segmentation of grayscale histograms, which can significantly affect subsequent pore structure analysis and fluid flow simulations. This paper presents a new discharge-based segmentation algorithm capable of predicting non-Darcy permeability of pervious concrete mixtures. The algorithm uses X-ray CT image-based finite-volume permeability simulations to determine the specific discharge at various hydraulic gradients. Experimental results of a falling-head permeability test were used to calibrate and validate the developed finite-volume models. The permeability simulation results from the developed thresholding algorithm were compared against simulation results obtained from 10 different global thresholding algorithms. It was found from the analyses that the developed discharge-based thresholding algorithm predicts non-Darcy permeability characteristics and the effective porosity of the pervious concrete mixtures more accurately than other global thresholding algorithms.
Original language | English |
---|---|
Article number | 04019179 |
Number of pages | 12 |
Journal | Journal of Materials in Civil Engineering |
Volume | 31 |
Issue number | 9 |
DOIs | |
Publication status | Published - 2019 |
Externally published | Yes |
Keywords
- Discharge-based segmentation
- Falling-head permeability
- Finite-volume modeling
- Global thresholding algorithms
- Non-Darcy flow
- Pervious concrete
- X-ray computed tomography