Search for a command to run...
Abstract The application of uncertainty quantification (UQ) methods to realistic industrial-scale structures remains a challenging task, due to the typical geometric complexity, the existence of various sources of uncertainties and the high dimension of structural models. These challenges are even bigger when dealing with composite material structures, which are knowingly prone to uncertainties arising from manufacturing processes and environmental influences. In this context, the present paper intends to contribute to increase the maturity level of UQ techniques to composite aeronautic structures, under the combined effects of space-dependent material and environmental fluctuations. Variations in temperature, laminate thickness and fiber volume fraction are jointly considered, being represented as random fields discretized using the Karhunen-Loève Expansion (KLE), while fiber angles are treated as random variables. Aiming at expanding the range of situations possibly found in practice, both Gaussian and non-Gaussian random fields are considered within a methodology combining the Iterative Translation Approximation Method (ITAM) and KLE. A micromechanical model is used to represent temperature- and moisture-dependent material properties, capturing the coupled effects of environmental degradation of material properties and hygrothermally-induced stresses. Monte Carlo Simulation (MCS) is employed to perform uncertainty quantification for buckling loads and vibration natural frequencies of a regional aircraft composite wing structure modeled with a relatively high-dimension finite element model. Additionally, global sensitivity analysis based on Sobol’ indices is conducted to identify the most influential random parameters, where structural responses are approximated using artificial neural network (ANN)-based surrogate models. From the simulation scenarios analyzed, accounting for different values of standard deviations attributed to random variables and correlation lengths assigned to random fields, the statistics of structural responses are assessed. The significant spread of structural responses highlights the importance of incorporating the considered types of uncertainty in analysis and design procedures for achieving robust and reliable aerospace composite structures.