The prediction of bitumen properties based on FTIR and multivariate analysis methods

The aim of this research was to correlate the chemical and physical characteristics of bitumen samples with the Fourier Transform infrared (FTIR) spectroscopy. Based on the FTIR analyses of 32 bitumen samples of different refineries, viscosity and ageing states using the attenuated total reflection (ATR) technology with multiple reflections, the gained spectra were evaluated with the chemometrical approach including multivariate analysis methods. With the Linear Discriminant Analysis (LDA), the bitumen samples could be distinguished according to the refinery even if the grades and the ageing states of the samples varied. In addition, the Partial Least Square Regression (PLSR) enables the determination of linear combinations for the description of different chemical, conventional and rheological parameters including the asphaltene content casp, the softening point TR&B, the needle penetration PEN, the complex shear modulus |G∗| and the phase angle δ. Thereby, the separated consideration of the refineries allows the prediction of the parameters within the permitted limits according to the respective standards. Only for the softening point, the deviation of the calculated values exceeds the permitted deviations.