Development of New Fatigue Model for New Zealand Dense Graded Hot Mix Asphalts
The Mechanistic empirical pavement design method for flexible pavements is based on modelling certain modes of failures for different pavement materials. In the Australian and New Zealand guidelines, the mechanistic empirical pavement design is based on modelling fatigue and permanent deformation as the two major modes of failures. The Austroads guidelines use the Shell fatigue performance transfer function to model the fatigue behaviour of asphalt mixes. In this research, the fatigue behaviour of different mixes AC10, AC14 and AC20 with different types of binders 80/100 and 60/70 was thoroughly investigated. The Shell model significantly underestimated the measured fatigue life for all mixes. A wide range of properties of the examined mixes was considered; percent of air voids ranges from 1.2% to 11.4%, binder content (at optimum, ± 0.5 from optimum), and the flexural modulus ranges from 1600 to 4576 MPa. A new fatigue model was developed at the University of Canterbury. The Canterbury model was based on the bending beam fatigue results of 78 beams tested at constant strain mode at different strain levels range from 300 to 600 microstrains. The new model provides a much better matching to the measured data with no observed bias and it accounts for percentage of air voids in the total mix and the effective binder content instead of the total binder content that is currently included in the Shell Model.