Predicting the Recovery and Nonrecoverable Compliance Behaviour of Asphalt Binders Using Artificial Neural Networks

Additives are widely used to enhance the rheological and performance properties of asphalt binder satisfy demands extreme loading climatic conditions. Meanwhile, adding complexity behaviour that requires more time, effort, material resources during laboratory work. The purpose this research was use Artificial Neural Networks (ANNs) predict recovery (R) nonrecoverable compliance (Jnr) based on mechanical test parameters binder. A comprehensive experimental database consisting results frequency sweep Multiple Stress Creep Recovery (MSCR) using a dynamic shear rheometer (DSR) at five temperatures (46 ?C, 52 58 64 70 ?C). Prediction models for R Jnr modified with different contents fly ash, ash-based geopolymer, glass powder/fly styrene–butadiene styrene (SBS) were developed. ANNs model developed input (temperature, frequency, storage modulus, loss viscosity) one hidden layer neurons. pointed out hybrid 4%SBS binders achieved highest ability resist extremely heavy traffic recover deformation 60.1% 85.5% 46 respectively, compared other binders. Excellent R-values total data set 0.937, 0.997, 0.985, 0.987 Jnr3.2 unaged binder, aged R3.2 respectively. Therefore, is appropriate tool or temperatures.