ANÁLISE PROBABILÍSTICA DE UM PAVIMENTO DE CONCRETO SUBMETIDO A UM GRADIENTE TÉRMICO EM REGIME PERMANENTE

Abstract

Probabilistic analysis of a concrete pavement submitted to a steady-state temperature gradient is presented to study its stresses and displacements. The concrete floor consists of three layers: a concrete plate followed by a rolled concrete base under a polyurethane sub-base. Two 3D models are implemented through the finite element method with ANSYS software, one deterministic and other probabilistic. Uncertainties in geometry, thermomechanical material properties and thermal gradient are implemented by Monte Carlo simulation. This model describe how uncertainties affects output stresses and vertical displacements variables in the concrete slab. The hexagonal elements Solid170 and Solid185, with respectively one or three degree of freedom per node and eight nodes, are in the thermomechanical analysis of the pavement. Numerical results obtained in scientific literature are used to compare and validate those with are obtained by means deterministic model. Statistics and cumulative distribution functions associated with some responses of the plate are obtained using the probabilistic model. In addition, the probabilistic results show that the thermomechanical properties of the sub-base have less significance in the stresses and displacements suffered by the pavement. Young's modulus and thermal expansion coefficient of the plate, thermal expansion coefficient of the base and the thermal gradient are the most significant input variables between output parameters of the pavement.