Development of a new fatigue performance test using the Hamburg Wheel Tracking Device
Restricted (Penn State Only)
- Author:
- Sahraei Joubani, Ali
- Graduate Program:
- Civil Engineering (PHD)
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- December 11, 2023
- Committee Members:
- Farshad Rajabipour, Program Head/Chair
Reuben Kraft, Outside Unit, Field & Minor Member
Mansour Solaimanian, Chair & Dissertation Advisor
Shihui Shen, Major Field Member
Aleksandra Radli¿ska, Major Field Member
Behnam Jahangiri Koohbanani, Special Member - Keywords:
- Asphalt Mixture
Fatigue Performance
Hamburg Wheel Tracking Test - Abstract:
- The main objective of this study was to develop a new test that would be simple to conduct and analyze for evaluating resistance of asphalt mixtures to fatigue cracking. For this purpose, a piece of equipment known as the Hamburg Wheel Tracking Test (HWTT) was used. HWTT has been widely accepted as a reasonable and reliable test to evaluate the rutting and moisture damage performance of asphalt mixtures but its use for evaluating fatigue resistance of asphalt mixtures is relatively a new concept. Its use as a fatigue test is advantages over current laboratory fatigue tests because of the possibility of considering various underlying support layers when the specimen is subjected to repeated loading. Ability to assess the asphalt mixture fatigue resistance in a layered system under repeated loading and in a laboratory environment provides the opportunity to integrate material design and pavement structural design for optimum performance. In addition, this test is practically feasible as a routine test method in terms of reliability, equipment availability, and data processing efficiency. The tests were conducted on a two-layer structure with the asphalt concrete slab as the top layer and the neoprene or an unbound aggregate base as the underlying layer. Based on the numerical analysis, a width of 6 inches was selected as the optimum width for this study. However, the experiment also included the 4-inch width for comparison. The thickness of the slab was selected at three levels: 1.0, 1.5, and 2.0 inches, with the 1.5-inch thickness being the main thickness applied to most of the slabs. In an asphalt pavement structure, the surface layer is often designed at a thickness of 1.5 to 2.0 inches. These tests were based on strain amplitude growth corresponding to three defined stages during the test (Early, Middle, Late). The impact of several factors on the test results was investigated. Those factors included the slab width and thickness, the underlying support type and condition, temperature of the test, the speed of tracking, and the type of asphalt mixture. The experiment showed that as the slab thickness become smaller, the strain amplitude and its corresponding growth rate become larger, exhibiting a relatively higher rate of fatigue damage. Using the unbound aggregate base proved to be challenging because of difficulty in achieving a smooth, even surface. A smooth surface was required to ensure proper bond of the gauges and collection of the data. Using synthetic rubber neoprene provided a smooth base and delivered a more reliable dataset compared to the aggregate base in this study. As expected, temperature played a significant role in the experiment, as a higher initial strain amplitude was observed at elevated temperatures. A higher strain and increased strain growth rate were observed at lower loading speeds. After determining the appropriate dimensions, temperature, and underlying support, tests were conducted on six different asphalt mixes. Strain development was monitored at the bottom of the slab for each mix. From the data collected, a new fatigue life model was developed to evaluate the fatigue performance of asphalt pavements when tested under the conditions explored in this research. The fatigue life in this model is defined as the number of HWTT load cycles at which the strain amplitude doubles, or equivalently, when the stiffness is reduced to half. This determination is made under the premise that the test operates in stress control mode. It was found that the rate of growth in strain amplitude during the test increased considerably when the binder content in the mix was decreased, validating the importance of ensuring adequate binder content in the asphalt mix. The SMA mixture demonstrated a higher rate of strain amplitude increase compared to the dense-graded mixtures. The mix with 35% RAP content and a PG 64-22 binder delivered lower growth in strain amplitude compared with other mixes. The IDEAL-CT test was also performed to draw comparisons. However, based on the outcomes, it showed no correlation with the results of the newly developed fatigue test. Based on the growth of strain amplitude throughout the test, a fatigue life model was developed. The fatigue life can be determined using this model, based on the number of cycles, with a criterion of either a 50% reduction in stiffness or a doubling of the strain amplitude. Finally, a numerical model was developed, and the observed strain response demonstrated that this model is capable of predicting both the type of response and the magnitude of strain amplitude with relative accuracy.