Improved consideration of multiple-axle loads in the structural design of flexible pavements

2017-03-01T01:33:37Z (GMT) by Moffatt, Michael Alexander
This research demonstrates that the observed performance of pavement materials subjected to multiple-axle loading can be used to improve the way flexible pavement design procedures consider the damage caused by different axle group and load combinations. Australian flexible pavement design procedures are used as the basis for considering this improvement. Fundamental to the Australian design procedures are established equal damage loads for each type of multiple-axle group. This study determined that the derivation of these equal damage loads was more heavily influenced by numerical calculations based upon assumed behaviour than on experimental data. It was also found that international design methods are similarly based upon assumed response-to-load behaviour of modelled pavement structures. In contrast, this thesis utilises performance data in order to determine the relative damage caused by different axle group-types and loads. In new work, the effects of multiple-axle loads on three pavement material types were examined. Accelerated pavement testing was used to assess the deformation of unbound granular pavements with a thin bituminous surfacing. It was determined that a 180 kN triaxle group caused 0.8 to 1.0 times the damage caused by a 80 kN single axle. Assuming a load-damage exponent of 4, the results confirmed that interaction between the axles of a multiple-axle group did not affect the deformation damage caused to the pavements. New consideration of an existing laboratory-based asphalt fatigue damage model - developed by others - demonstrated that there was no damaging or ameliorating effect of grouping strain pulses used to simulate multiple-axle groups. A similar laboratory-based study was undertaken to examine the effects of grouping load pulses on the flexural fatigue performance of a cemented material. Analysis of the resultant data demonstrated that damage was not affected by grouping pulses of equal load. Parametric three-dimensional finite element method and layer linear-elastic analyses indicted that the pavement structure plays a significant role in determining loads for multiple-axle groups that cause the same asphalt and cemented material fatigue damage as a Standard Axle load. Consequently, the summed-peaks pavement design model for determining the damage to a bound material resulting from an axle group with specific load was proposed. An improved design procedure based upon this model was proposed. The procedure specifically calculates the pavement damage resulting from each axle grouping type and each load within a traffic load distribution. An examination of the implications of design outcomes in using this improved procedure determined that reductions in both asphalt and cemented material thicknesses of up to 50 mm would result. These differences are significant, and represent improved material and construction cost savings - potentially of the order of A$50 million a year. A determination of the overall reliability of pavements that are constructed in accordance with the improved procedure is needed to fully implement the improved procedure. Future work should also consider the use of the summed-peaks model with the subgrade compressive strains used in current mechanistic design.