Dynamic Analysis and in Situ Validation of Perpetual Pavement Response to Vehicular Loading

Abstract

A three-dimensional (3-D) finite element (FE) model was developed to predict pavement responses to vehicular loading. The model incorporates measured tire-pavement contact stresses, continuous moving wheel loading, and hot-mix asphalt (HMA) viscoelastic characteristics. The model was fine-tuned using implicit-dynamic analysis and validated using pavement response from accelerated loading. Two tire configurations (dual-tire assembly and wide-base 455 tire) and three full-depth flexible pavement designs (HMA 152 mm, 254 mm, and 420 mm) were used in both FE modeling and accelerated loading tests. The predicted and calculated strain responses at the bottom of HMA were in agreement. Most important, the study shows that vertical shear strain in the upper 76 to 100 mm of the pavement surface is critical for thick pavement and is influenced by the 3-D tire-pavement contact stresses under each tire rib. However, the tensile strain at the bottom of HMA is affected mainly by the total wheel load. The vertical shear strain is responsible for near-surface fatigue cracking as well as HMA primary rutting. Top-down cracking could result from the local vertical shear strain in the upper 25 mm of the HMA where the effect of tire-pavement tangential stresses are the highest. In addition, the study concluded that wide-base tires cause higher longitudinal tensile strain at the bottom of HMA and compressive strain at the top of subgrade, where those responses are highly affected by the total wheel load. However, wide-base tires were found to cause less vertical shear strains near the surface than dual-tire assembly loading regardless of HMA thicknesses.

Numerical comparison of flexible pavement dynamic response under different axles

“…3D FEMs have recently been successfully used in pavement dynamic analysis (Zaghloul et al 1994, Shoukry and William 1999, Saad et al 2005,Al-Qadi et al 2008). There are various advantages in using a 3D FEM: first, 3D FEM allows consideration of complex behaviour of A. Sarkar 2 pavement material; second, simulation of different complex situations such as moving multiple loads; third, the analysis results may substitute for the laboratory or field test results and fourth, the analysis allows more realistic interface, footprint of the loaded wheel and boundary conditions than the analytical method.…”

Evaluation of the Mechanistic–Empirical Pavement Design Guide model for permanent deformations in asphalt concrete

“…It can be argued that the MnROAD test differs from the HVS test regarding ageing, healing and viscoelastic strain relaxation time. The comparably high axle load and tyre pressure used in the current HVS test could also have altered the stress state as shown by Al-Qadi et al (2008). Therefore, additional accelerated loading tests should be carried out to investigate the distribution of flow rutting as a function of depth.…”

“…This was achieved by accelerated pavement testing with a heavy vehicle simulator (HVS) subjecting two representative Swedish flexible pavement sections to repeated loadings under controlled climate conditions. Model results were compared to observation and the results were discussed in the light of experiences from other investigations (Agardh and Busch 2006, NCHRP 2006a, Al-Qadi et al 2008, Azari et al 2008 as well as the MEPDG documentation (NCHRP 2004(NCHRP , 2006b.…”

Simulation of pavement response to tire pressure and shape of contact area

“…Kim et al (2009) found that the domain size of 140-times the radius (R) of circular loading area in the vertical direction and 20-times R in the horizontal direction gives more accurate results using the 8-node isoparametric quadrilateral elements. Al-Qadi et al (2008) found that the horizontal location of the infinite boundary from the load centre needs to be greater than 900 mm to have the closest solution to the full-size reference finite element model. The location of the bottom infinite boundary element was recommended at a depth of 1100 mm, where the maximum compressive stress in the subgrade became insignificant at 1% or less of the maximum contact stress.…”