Currently, there is a need to replace empirical methods of pavement design with new methods that use mechanical concepts (empirical-mechanistic), as occurs in the transition from the current Brazilian method to the MeDiNa method. However, the MeDiNa software requires many input parameters that are obtained through inaccessible tests, and therefore end up being estimated through correlations. Therefore, the objective of this work is to perform a sensitivity analysis of the percentage of Estimated Cracked Area (TRI) and Wheel Track Sinking (ATR) in two flexible pavements, calculated by MeDiNa. For that, data from the BR-316 Duplication Project were used, from which, through empirical correlations, the input parameters of the reference scenario were estimated, such as: Modulus of Resilience (MR), Poisson's Coefficient (μ), Regression coefficients of permanent deformation (Ψi) and fatigue (K1). Subsequently, these input parameters were individually varied by ± 25% and the results of ATR and TRI were analyzed on two pavements of different traffic, both on the same subgrade. For the structures studied, the sensitivity analysis technique detected strong susceptibility of the ATR to variations in the regression coefficient Ψ2 and Poisson's coefficient in the subgrade and especially in the base where they generated increases of 41.0% and 18.1%, respectively, in addition to indicate that the TRI is greatly affected by the layer thickness and modulus of resilience of the coating, which recorded increases of 29.6% and 17.7%, respectively. Therefore, it is recommended to carry out these tests as a priority, to the detriment of the others, whose parameters could be obtained by correlations, in the present scenario of test restriction.
|
