Quantifying the uncertainty associated with material properties in engineering analysis has become more important in the design of many components in the aerospace field. In this study, a new method is proposed to account for the uncertainty associated with the elastic modulus of materials used in aerospace components. A computerized tomography (CT) scanner is used to capture the porosity of the material and the corresponding uncertainty is represented with epistemic uncertainty. A stochastic upscaling method is used to find a homogenized modulus that correctly reflects the effect of defects inside of the material. This homogenized elastic modulus is then applied to a constructed finite-element model of an aerospace component so that the stochastic behavior can be correctly quantified. Simulations of the selected example, i.e., turbine blade, include both aleatory and epistemic uncertainty; thus, a P-Box is introduced to represent the response of the simulations. The stochastic upscaling method is applied again to match the P-Box response of the coarse scale model to that of the fine scale model. The obtained results show that the proposed framework not only significantly reduces complexity of the given engineering problem, but also produces accurate results which include both aleatory and epistemic uncertainty.