This research aims to develop a methodology allowing to extend the validation domain of aeronautical structural models to reduce reliance on costly physical testing while maintaining the current required safety standards . The core challenge is quantifying the risk of applying models to untested configurations. Using composite panel buckling as a case study, the approach employs a multi-fidelity strategy. This leverages low-fidelity analytical models to guide the extrapolation of sparse high-fidelity experimental data. Non-dimensional parameters are used to reduce the high-dimensional input space associated with composites multi-ply laminates. Finally, a probabilistic validation framework propagates uncertainties for a robust comparison between simulation and experiment.