Abstract: On the premise of maintaining the applicability of the Generalized Logistic Sigmoidal(GLS) model, fixing partial parameters can improve the convenience and application values of the model according to the characteristics of asphalt mixture. Based on test data of asphalt mixture containing storage, storage and loss, and dynamic moduli, three residual fitting functions were established respectively, and the differences of these functions using different shift factors to construct various master curves were compared. For the GLS model describing asphalt mixture, a new method was established to classify the sensitivity of the parameters, and then fixed values of the insensitive parameters were obtained based on 65 groups of test data. The results show that the GLS model is suitable to use the residual function Fit- $ E_{dyn} $ and the shift factor Kaelble for asphalt mixture; the parameters $ \lambda $ and $ \beta $ can be fixed; in addition, the parameter $ \gamma $ can also be fixed when the requirement for fitting accuracy is not high. This work can be used to further analyze the correlation between fewer model parameters of the simplified GLS model and material properties of asphalt mixture.
Keywords: asphalt mixture; Generalized Logistic Sigmoidal(GLS) model; master curve; shift factor; parameter sensitivity
合肥工业大学学报自科版