Drainage capacity of pervious pavement mixtures is commonly measured using a falling head permeameter at hydraulic heads much higher than expected in the field. Recent advancements in computational fluid dynamics (CFD)- and X-ray computed tomography (XRCT)-based modeling eliminates the laboratory challenges of maintaining lower hydraulic heads. However, improper characterization in digital image processing (DIP) and finite-volume simulations resulted in significant errors in permeability measurements and fluid flow behavior. In addition, past studies have identified non-Darcy fluid flow characteristics in pervious pavement mixtures following the Izbash and Forchheimer laws. This paper attempts to bridge this research gap by comparing the Darcy and non-Darcy permeability parameters at different laboratory and field hydraulic heads using advanced XRCT-based modeling. It was found from the analyses that the use of laboratory hydraulic head could result in significant underestimation of permeability parameters compared with the field hydraulic heads for Darcy and Izbash equations (by up to 73%), and overestimation for Forchheimer equations (by up to 216%). Fluid flow behavior in pervious mixtures was found to be in transition flow regime (neither laminar nor turbulent) at both laboratory and field hydraulic gradients. Overall, this study can help in a better fundamental understanding of the current limitations of laboratory measurements and the need for XRCT-based numerical modeling to bridge field and laboratory permeabilities of pervious pavement mixtures.

@en

The primary usage of pervious concrete mixtures in pavement construction is to improve the functional performance of pavements through increasing skid resistance and hydroplaning speeds (by dispersing the hydrodynamic pressure developed at the tire-pavement interface) and increasing the water infiltration and tire-pavement contact area at higher speeds, thereby reducing the wet weather accidents. Past skid resistance simulation models for pervious pavements considered the use of artificial pore grid models, but with today's advanced X-Ray computed tomography (XRCT) technologies, it is now possible to develop realistic pavement models for skid resistance simulation. This paper aims to investigate the effect of thresholding - a crucial step in image segmentation - on skid resistance of pervious concrete pavements, whose pore structures are derived from XRCT. It was found from the analysis presented in this paper that the various thresholding algorithms are found to be either under- or over-estimating the uplift and drag forces, as compared to the discharge-based thresholding algorithm. Errors in skid number due to thresholding are however found to be marginal (with up to 2 SN or 3% error).@en