Performance evaluation of a newly introduced chemical additive for warm bituminous mastic based on fatigue response and microstructure homogeneity

The present study investigates the fatigue performance of warm bituminous mastic (WBM) containing a new chemical additive, PAWMA, compared with traditional chemical and waxy additives represented by Zycotherm and Sasobit, respectively. A 60/70 pen grade bitumen served as the base bitumen. Limestone filler was added at a proportion of 60% by the weight of the base bitumen. A new approach was developed to evaluate the microstructural homogeneity of WBMs. This was accomplished using maps obtained from elemental mapping analysis carried out through energy dispersive X-ray spectroscopy. For each map, an image processing method provided a homogeneity coefficient, HC. The fatigue behavior was investigated using the time sweep method enabled by a dynamic shear rheometer apparatus. Definitions of fatigue life based on the concept of dissipated energy were used in order to investigate the fatigue behavior over both LVE and NLVE strains. Comparing the HCs, it was found that all the different additives had a best value beyond which the coefficient dropped significantly. The optimum contents of PAWMA, Zycotherm, and Sasobit were determined as 0.4%, 0.1%, and 3%, respectively, leading to the highest HC values of 83, 75, and 75. The threshold for this optimum was found to be relatively similar for a number of the fatigue life definitions and most of them coincided with the maximum homogeneous point. The dissipated energy-based criteria were effective in assessing the fatigue behavior in the LVE regime, where the higher coefficient of variation (≈48%) and normalized range (≈1.10) indicated a clear distinction among the criteria. In contrast, in the NLVE regime, lower dispersion (CV ≈38%) and higher inter-criteria correlation (r ≈0.94) suggested that the distinction among the criteria was less pronounced. • Fatigue response and microstructure of WBM with a new chemical additive PAWMA were studied. • Image processing techniques effectively assessed the distribution of elements on the sample surface. • Even small amounts of additive had a notable impact on filler distribution uniformity. • Fatigue behavior was significantly different in the linear or nonlinear viscoelastic strain ranges. • The optimal PAWMA percentage resulted in the highest number of loading cycles.