Chemical Composition-Driven Performance Prediction Model for SBS-Modified Asphalt Using Multivariate Statistical

The performance of modified asphalt is closely related to its chemical composition. The further correlation study between chemical composition and performance gives a boost to the effective quality control of modified asphalt. The modified asphalt with a SBS modifier content of 4.5% was selected as the research object. The contents of asphaltene (As), resin (R), aromatic fraction (Ar), and saturated fraction (S) were tested using a rod-shaped thin-layer chromatography, and the succinic index (Ic=c), carbonyl index (Ic=o), and sulfoxide index (Is=o) were determined using a fourier transform infrared spectrometer. In addition, the phase angle (θ), complex shear modulus (G*), fatigue factor (G* sinθ), and rutting factor (G*/sinθ) of modified asphalt were obtained through dynamic shear rheological tests. At the same time, the basic performance indicators of SBS modified asphalt were tested. Based on the grey correlation theory, the correlation degree between various performance indicators of SBS modified asphalt and the four components was established, and the indicator factors for establishing the factor model were determined according to the magnitude of the correlation degree. Through factor analysis, a performance optimization model for modified asphalt was ultimately established. The results showed that, except for Δm, all other indicators had a good correlation with the four components. Among them, softening point (SP) had the highest correlation with As and R, while θ had the highest correlation with Ar and S. The modified asphalt indicators were reduced to two factors F ₁ and F ₂ through factor analysis, and the factor weights were calculated to obtain the principal component formula. Finally, the model was validated using modified asphalt with different SBS modifier dosages, and the results showed the feasibility of the model in assessing the performance of modified asphalt.