Correlating Laboratory Fatigue Endurance Limits to Field-Measured Strains

The most commonly used technique for predicting the fatigue performance of asphalt mixtures in the field is the bending beam fatigue test. However, there has been some difficulty in developing meaningful relationships between the laboratory and the field, sometimes extending to perpetual pavement design. To examine this issue, three analysis techniques compared measured strain data from embedded asphalt strain gauges under dynamic loading at the 2006 National Center for Asphalt Technology (NCAT) Test Track to laboratorymeasured endurance limits. These analysis techniques were originally used to evaluate fatigue endurance limits of two mixtures placed at the 2003 NCAT Test Track. This research was designed to further validate previous results. The first analysis compared the laboratory fatigue endurance limits of the mixtures to field-measured strains at a reference temperature (20C). The analysis was developed to determine if pavements that performed well in the field had field-measured strains less than the endurance limit. As with the previous study, a strong relationship between field-measured strains at a relative temperature and the laboratory fatigue endurance limit did not exist. Some test sections with field-measured strains lower than the laboratory fatigue endurance limit cracked. The converse was also true. The second analysis technique compared the fatigue endurance limit to a percentile on each section’s cumulative strain distribution. If a greater percentage of strains fell below the fatigue endurance limit, then one would hypothesize better performance in the field. As in 2003, the findings from the 2006 Test Track using this analysis technique were inconsistent. Test sections which showed no evidence of fatigue cracking had only 10% of their strain events below the fatigue endurance limit while other test sections which cracked had 50% of their strains repetitions below the laboratory threshold. The final analysis, using the concept of a laboratory to field strain ratio, showed the best relationship between laboratory fatigue endurance limits, field-measured strain data, and field performance. The strain ratio represented the nth percentile strain of a section’s cumulative strain distribution divided by the laboratory endurance limit. For example, the fatigue ratio for the 95th percentile strain is 2.76. The results of this final analysis verified prior findings that showed the fatigue ratio was the most appropriate analysis technique to use to correlate pavement performance with the laboratory fatigue endurance limit.