CHAPTER 3: DEVELOPMENT OF NUMERICAL MODEL FOR SKID
3.1 Issues Considered in Modeling Skid Resistance on Porous Pavement
The wet skid resistance on porous pavement is a complex phenomenon involving various mechanisms. Its modeling requires both finite element method (FEM) and computational fluid dynamics (CFD) formulation. In the simulation of porous pavement skid resistance, following issues should be considered numerically in order to get more accurate results.
3.1.1 Tire-Pavement Contact
The interactions between tire and pavement are critical in skid resistance modeling. A numerical study conducted by Ong and Fwa (2007a) has indicated the importance of realistic contact modeling in the prediction of skid resistance on conventional smooth pavements. The adopted non-linear contact algorithm must
106
include friction computation so that tire footprints at different wheel loads, as well as contact and friction forces, could be reproduced closely.
3.1.2 Fluid-Structure Interaction
The interaction between water on road surface and the pneumatic tire is an important element in wet-pavement skid resistance and has to be properly addressed.
It includes the hydrodynamic pressure applied on tire tread by water and the effect of tire deformation on water flow. It should be a two-way coupling process, where fluid force is transferred to the solid model and solid deformation is transferred back to the fluid model. The iteration continues until errors in both pressure and displacement satisfy certain convergence requirements.
3.1.3 Tire Deformation Behavior
Tire deformations under loading have to be correctly described in the simulation model. The deformed tire geometry is essential to model water flow patterns and hydrodynamic pressure underneath and around the tire. The entire tire structure should be modeled with different material properties for different tire components (i.e. rim, sidewall and tread). The modeling of tire behavior can be calibrated through a comparison between the simulated tire footprints and experimental measurements under different wheel loads.
3.1.4 Turbulence in Fluid Flow
Previous studies (Chuai, 1998; Ranieri et al., 2012) have already shown that the water flow within porous layers is turbulent rather than laminar Moreover, water flow under tires is also found to be turbulent due to the high relative velocity and small flow channel dimensions (Schlichting, 1960). Various numerical formulations are available to date to simulate the turbulent flow and special care should be placed when selecting appropriate turbulence model and calibrating model parameters.
107 3.1.5 Multiphase Flow
In the moving wheel frame of reference proposed by Ong and Fwa (2007a), both water and air are moving towards the stationary tire. Although the uplift and drag forces applied on the tire tread by air are relatively small comparing to those by water, it is still necessary to properly capture the free surface at water-air interface because it provides a simple way to control the inlet water quantity during the simulation process. The modeling of water-air interaction could be attained using a multiphase flow algorithm such as the volume of fluid (VOF) method (Hirt and Nichols, 1981).
3.1.6 Drainage Capacity of Porous Media
Experimental studies have implied that the superior skid resistance on porous pavement may be a result of its inner drainage channels and surface macrotexture.
These features significantly influence the under-tire drainage capacity at high speeds, which is essential for deterring skid resistance loss with increasing sliding speed.
Therefore, it is important for porous pavement skid resistance models to accurately simulate the drainage capacity of porous surfaces. As drainage capacity depends on randomly distributed pores and channels within a porous layer, it is extremely challenging to reproduce such a pore structure in the porous pavement. As such, some reasonable simplifications are necessary to solve this complex problem numerically.
Four of the problems discussed above, i.e. the tire-pavement contact, flow- structure interaction, tire deforming behavior and turbulence flow, have already been addressed by Ong and Fwa (2007a) in their previous studies on skid resistance modeling of conventional smooth pavements. Although a different software package is used in their work, similar modeling methods and algorithms could be applied. The multiphase flow has been considered in another model proposed by Ong and Fwa (2006) in a numerical study on hydroplaning. That application, despite the lack of flow-structure interaction, provided useful insights on the modeling of multiphase
108
flow in skid resistance simulations. However, the drainage capacity of porous pavements has never been included in their skid resistance simulations. This research focuses on the modeling of porous surface layers using a geometrically simple pore structure network which is capable to provide the same drainage capacity as in-situ porous pavements. Parameters such as porosity, permeability, clogging percentage and outflow time may be used as indicators of drainage capacity in this study.