基于结构固有频率的沥青混合料动态模量及预估模型研究

To realize rapid and nondestructive testing of the dynamic modulus of an asphalt mixture, Free-free Resonant Test (FFRT) was studied. Firstly, based on the principles of natural frequency calculations and equivalent temperatures, the main dynamic modulus curves were constructed by FFRT and the repetitive loading method. Then, the feasibility of testing dynamic modulus using FFRT was expounded, and the influence of temperature, gradation type, and porosity on the dynamic modulus of the asphalt mixture was studied by the FFRT method. Hence, a dynamic modulus prediction-model database was constructed based on the experimental data of 584 groups. Based on this investigation, the back-propagation (BP) neural network method was used to establish the dynamic modulus prediction model and compared with the traditional Witczak prediction model. The results demonstrate that it is both reasonable and feasible to characterize the dynamic modulus of an asphalt mixture by measuring the natural frequency. The FFRT can extend the constructed maximum frequency of the dynamic modulus main curve from 10⁷ Hz to 10¹⁴ Hz using the repeated loading method (RLM). The main dynamic modulus in this wider frequency range was constructed, and the accuracy of high frequency prediction of the dynamic modulus was improved. The prediction results of the BP neural network model are more accurate than the Witczak model. In conclusion, the above study provides a technical basis for fast and nondestructive testing of the dynamic modulus.