Development of a balanced cracking index for asphalt mixtures tested in semi-circular bending with load-LLD measurements

•A new index called Balanced Cracking Index (BCI) is developed based on Semi-Circular Bending (SCB).•Flexibility Index (FI) relies on slope at inflection point than the fracture energy.•The BCI was more repeatable and easier to compute as compared to the Flexibility Index (FI).•The BCI showed better correlation to field cracking performance as compared to FI and CRI.•All SCB-based indices failed to show the effectiveness of crumb rubber in the mixtures. In this study, a new Semi-Circular Bending (SCB) Balanced Cracking Index (BCI), is introduced as an alternative cracking index to the FI and CRI, which are based on the IL-SCB test (AASHTO TP 124). The high coefficient of variation (COV), high susceptibility to aging, and reliance on slope parameter in the previous index, called flexibility index (FI) was the motivation to develop a new index. The BCI is computed as fracture energy times the load-line displacement corresponding to 75% of the Pmax after the peak load (L75) is reached. The cracking behavior of 52 asphalt mixtures was investigated using three SCB indices: FI, CRI, and BCI, which was introduced herein. By analyzing identical data sets, it was found that the FI, BCI, and CRI indices were accompanied with significantly different coefficient of variability (COV) levels: 34.6%, 17.5%, and 11.4%, respectively. Although some general correlation between the indices was noted, the overall ranking of mixtures differed when comparing the three indices to field performance. The FI parameter is heavily driven by the slope of the load–displacement curve at the inflection point, and to that point, was not found to correlate with the fracture energy. Although fracture energy appears in the index, it is overwhelmed by the influence of the slope parameter. On the other hand, the BCI is driven both by the fracture energy and the L75 parameter. To illustrate the applicability of the BCI to predict cracking performance in the field, ten field sections were investigated. In contrast to other SCB indices, the BCI had the highest correlation with field data, suggesting its potential value in predicting the cracking performance of asphaltic pavement surfaces. The BCI was also shown to be superior in delineating groups of means, owing to its relatively low COV while having relatively high sensitivity to the effects of common mixture variables. All SCB indices investigated produced non-intuitive trends with respect to the effects of air voids and ground tire rubber on c