The objective of this thesis was to assess the repeatability of four-point bending method for estimating fatigue cracking of hot mix asphalt. This experimental research examined the influence of mix proportions, specifically asphalt and air void contents, on fatigue behavior of asphalt beam specimens in the laboratory tested with four-point bending method of loading. Following the relevant specifications, the strain-controlled mode of loading was selected. Fatigue failure is expected to occur at the cycle which the flexural stiffness of asphalt specimen reduced to 50% of initial amount. Up to a 1,000,000 cycles of loading were applied to each specimen.

Different tests at different temperatures and loading frequencies were conducted on a synthetic specimen and four different asphalt mixes. The applied strain varied between 200 and 600 μϵ, temperature varied between 4.4 and 37.5°C) and loading frequencies of 5, 10, and 15 Hz were used.

Based on repeated tests on a synthetic specimen, the device evaluated here was very repeatable. However, tests performed on hot mix asphalt specimens yielded a variability of up to 30% which can be mostly attributed to the variability of the specimens.

Temperature impacted the fatigue properties of a given mix the most. A decrease in temperature results in an increase in mix stiffness and consequently a decrease in laboratory fatigue life of specimens. The least fatigue resistance was obtained for specimens tested at 4.4°C. A number of specimens at that temperature failed before the full number of cycles could be applied. The differences in the fatigue properties of specimens tested at 21.5 and 37.5°C were more similar than those tested at 4.4°C.

The impact of the frequency on the measured parameters in the range of frequencies of 5 to 15 Hz is rather small and within the magnitude of uncertainties anticipated due to variations in the uniformity of the specimens. Typically tests performed at a load frequency of 15 Hz resulted in lower fatigue parameters than those at 5 and 10 Hz.

The nominal air void contents are directly related to the fatigue parameters. An increase in air-void content resulted in a decrease in the laboratory fatigue life and a decrease in mix stiffness.

Another important consideration in this test method is the quality of the specimens. With respect to fatigue performance, accurate control of air-void content is very important. A cyclic loading method was used to prepare the specimens in this study. It was found that the control of the density of the specimens after trimming was difficult. Several specimens with very close bulk densities after compaction yielded variable overall air voids after trimming. An ultrasonic device is recommended as a tool for quality control before fatigue testing of the specimens.