The theory of viscoelasticity has been a classic topic, but its application to the modeling of flexible pavement has not been so successful even though viscoelasticity has long been characterized in asphalt concrete (AC). This is mainly owing to that the time-dependent property of a viscoelastic material is always very challenging and also that flexible pavement is a layered structure which further complicates the modeling work. This work establishes an efficient and accurate semianalytical solution for primary response (stresses, strains, and displacements) of flexible pavement on the ground of layered viscoelastic theory (LVET).
Chapter I highlights the development of flexible pavement. On account of the viscoelastic AC, the theory of linear viscoelasticity is reformulated briefly from a mechanical standpoint. The material characterization and the constitutive equation for viscoelastic AC are reviewed.
In Chapter II, pavement response to a moving load is expressed by the convolution integral of the pavement impulse response and the moving load. This convolution integral is elaborated, and in so doing the history of viscoelastic modeling of flexible pavement is reviewed and summarized.
Pavement impulse response, the kernel of pavement response, is introduced in detail in Chapter III. Making use of the Prony series for a viscoelastic material, pavement impulse response is established numerically in the space domain and, for the first time, analytically in the time-domain. Such solution is called the semianalytical solution.
To verify the semianalytical solution established in Chapter III, Chapter IV establishes the semianalytical solution for pavement response under the stationary load. This semianalytical solution is verified with a finite-element-based method through numerical examples, after which, the well-known collocation method is examined.
Chapter V proposes the semianalytical solution for pavement response under moving dynamic load. The semianalytical solutions under various typical loading conditions are verified with the Abaqus solutions.
Main conclusions are summarized in Chapter VI. Future works needing further research are suggested.