The development of models to predict rutting in the asphaltic concrete and dense-graded aggregate layers of flexible pavements is reported. The models were developed from data obtained from repeated-load triaxial tests. Details of equipment and methodology are reported.

An asphaltic concrete base mixture was tested at temperatures of 45 F, 77 F, and 100 F. Longitudinal stress levels of 20 psi, 50 psi, and 80 psi were used. The tests were performed unconfined. Most of the specimens were tested to 10,000 cycles of loading. Three sequences of loading were used.

It is noted that temperature and stress level are extremely important in determining the magnitude of permanent deformation. However, the slopes of plots of permanent strain against number of stress repetitions appear to be largely unaffected by these two variables.

All the dense-graded aggregate specimens were compacted at optimum moisture content (4.7 percent) determined from AASHTO T-180. A number of the specimens were allowed to air dry until a moisture content of approximately two percent was reached. The specimens were then tested in that condition. One group of specimens was tested at optimum conditions; a !bird group was tested at conditions "wetter" than optimum. Longitudinal stress levels of 10 psi, 20 psi, and 30 psi were used. Confining pressures ranged from 5 to 15 psi. All specimens were loaded for 1,000 cycles. Each load cycle produced greater permanent deformation in the wetter specimens than in the drier ones. However, the magnitude of the initial load cycle decreased with increasing moisture content.

The two models described in this paper are to be used in tandem to predict rutting for flexible pavement at particular sites after developing the traffic and environmental histories for those sites.

Report Date


Report Number

No. 538

Digital Object Identifier



1980 Annual Meeting of The Association of Asphalt Paving Technologists (Louisville, Kentucky).