Improving reliability in the bitumen bond strength test on pavement materials

Abstract

The study of mechanical properties of pavement materials is an essential part for their initial characterization. Therefore, the reproducibility and constant improvement of the tests involved allow to obtain faithful results, reducing the potential errors and variability that might occur during sample preparation. For this reason, this study focuses in establishing criteria and nominal values to obtain an adequate finish in the aggregate sources. This intends to reduce the standard deviation of the uniaxial tensile test, carried out by means of the bitumen bond strength test on dry conditions. The aggregates were treated by a sanding process for 60 seconds, repeating such cycles 10 times per sample. Also, abrasive bands of varying grain numbers were used. A Pull-Off Tensile Strength (POTS) of 2.39 (0.29) MPa was achieved, where the value in parentheses corresponds to the standard deviation. This value was selected due to the best conditions found in the experiment. Furthermore, a linear model obtained by a factorial design demonstrated the main effects provide up to 84 % correlation with the pneumatic stress and up to 78 % with its standard deviation, which is considered satisfactorily since the p-value defined in this research as the approval probability for the recommended conditions is 6,56·10^-4, demonstrating statistical significance.