Backcalculation Of Asphalt Pavement Stiffness Coefficient Based-FWD Deflection Basin Parameters

With the gradual improvement of China’s road infrastructure construction, and the construction tide is fading away, more and more pavement conditions of old roads need to be detected and evaluated. In order to detect pavement condition quickly, we usually use FWD to measure road surface deflection, which is use to back calculate the modulus of pavement and subgrade layers for evaluation of pavement structures. However, there are so many complex factors affecting pavement deflection. No matter what kind of analysis method, such as Iteration,database search, regression analysis, or artificial neural networks, is not enough to establish the mapping relation between road surface deflection basin and layer modulus accurately and directly, which results in inaccuration of layer moduli. So this paper proposes to back calculate layer stiffness coefficients(SC) to assess the condition of pavement structures based on parameters of FWD deflection basin.Firstly, this paper analyzes the influence of pavement materials, pavement thickness, the contact state between the layers of the pavement structures, environmental factors and load factors based on features of FWD deflection basin individually, fingding that the road surface deflection basin is most sensitive to the change of the modulus and the thickness. Secondly, it introduces the BP neural network’s calculation principles and algorithms applying to back calculation, and studies the back-calculated models of neural network structure design and the setting of key parameters of neural network. Thirdly, based on BP neural network and the database of deflection basin, building the layer moduli and SC respectively. Then it proves that the back-calculated SC is more accurate than back-calculated modulus by comparing theoretical and measured datas of deflection basin. Finally, applying the study of SC to the pavement condition assessment of the Foshan First Ring Road. Then establishing the evaluation standard of the pavement condition based on SC, which is used for comparing with PCI-based evaluation of pavement damage status, and the feasibility of this method is tested and verified. It provides guidance to evaluate the condition of asphalt pavement quickly and efficiently.