**Autor:**Leena Korkiala-Tanttu (Korkiala-Tanttu, L.)

**Instituição:**Helsinki University of Technology

**Departamento:**Department of Civil and Environmental Engineering

**Cidade:**Finlândia

**Ano:**Janeiro/2009

**Abstract:**

An analytical-mechanistic method for the calculation of permanent deformations of pavements has been developed at the Technical Research Centre of Finland (VTT) over some years by the author. The calculation method is needed in the analytical design procedure of pavements. This research concentrated on the calculation method for permanent deformations in unbound pavement materials. The calculation method was generated based on the results of full-scale accelerated pavement tests along with the complementary laboratory tests together with finite element calculations. The objective was to develop a relatively simple material model for unbound materials, which is an analytical, nonlinear elasto-plastic model. The stress distribution studies of traffic load showed that it is very important to calculate stresses in pavements with an elasto-plastic material model to avoid false tensile stresses in unbound materials, especially when the asphalt layers are thin. The new material deformation model can take into account the amount of the loading, the number of vehicle or wheel passes, the deformation capacity of the material and its stress state. The strains in each layer and subgrade are calculated and converted to the vertical deformations and then summed to obtain the total rutting. The method was verified against two Finnish accelerated pavement tests. The results indicated that the material model gave tolerable results for the relatively high load levels used in these Heavy Vehicle Simulator (HVS) tests as the relative error was around ± 30%. For the structures with thicker bound layers and therefore lower stress state in the unbound layers, the method gave more reliable results. The material parameters have been defined only for the most common Finnish unbound materials in a few basic situations. The wider use of the method requires material parameter definitions for a larger range of materials. However, even in the current form the method can be applied in a relatively reliable way to compare the sensitivity of different structures against rutting. The most important factors affecting rutting were studied to find a method to include their effect on the calculation method. These factors were loading rate, stress history, temperature and the geometry of the road embankment. The modelled examples proved that the most important factor of rate effect is the change in stress state due to the change in the resilient properties of bound layers, while the rate effects on the unbound material itself has a smaller role. The accelerated pavement test proved that rut depth depends greatly on the temperature: the rut depth grows from 10% to 15% at +10 ºC and 20 to 25% at +25 ºC compared to rut depths at +5 ºC due to the changes in the stiffness of the bound layer. The unloading-reloading cycles have only a slight effect on the permanent deformation. The introduced geometric factor describes an average, structurally independent, increase in the rate of rutting, which depends on the steepness of the side slope and on the distance to the edge of the structure.