Genetic programming to formulate viscoelastic behavior of modified asphalt binder

The objective of this research was to develop prediction models for complex shear modulus (G*) and phase angle (δ) bitumens modified with crumb rubber, styrene-butadiene styrene, polyphosphoric acid at low moderate temperatures. experiments consisted three different dosages each modifier added the original bitumen followed by measurement G* δ using dynamic rheometer (DSR) test in frequency sweep mode (21 loading frequencies from 0.1 100 Hz) seven temperatures: −22, −16, −10, 0, 10, 16 22 °C. Having experimental database, a robust genetic programming (GP) method used an individual model based on temperature, frequency, bitumen, dosage modifier. Results showed that GP successfully developed accurate meaningful expressions calculating as two main constitutive components viscoelastic behavior bituminous composites. Then, parametric study sensitivity analysis were performed better understand effect variables trend models. is most effective input variable amount accurately reflect temperature they behave linearly considered