Fatigue life prediction of the modied asphalt mixtures with ANFIS modeling

Abstract. In this study, the eect of Styrene-Butadiene-Styrene (SBS) modication on physical and mechanical properties of hot mix asphalt, in terms of fatigue characteristics, are evaluated. As such, 87 samples were prepared using 15 dierent mixtures, and the repeated load-fatigue indirect tensile test was applied to these samples using a Suleyman Demirel University (SDU) asphalt tester apparatus. In addition, it is also used to analyze the eects of gradation, bitumen percentage, loading, and polymer type on fatigue behavior. 18 control samples were prepared to analyze the eect of SBS modication on fatigue performance. The other objective of the study is to present the reliability of the computer controlled repeated loading test apparatus. Linear regression and Adaptive Neuro Fuzzy Inference System (ANFIS) were used to analyze the results of the test, and models were developed and analyzed in terms of modeling. At the end of the study, the variables that are eective in improving the fatigue performance are addressed in detail, and the ANFIS environment is likely to be used in non-linear modeling of such studies.