Résumé : The report proposes a methodology for estimating the uncertainty in experiments and numerical simulations, to promote a constructive validation of computational approaches. Validation hierarchies are suggested for two combustion applications that attract increasing interest in the combustion field: a flameless and an oxy-fired furnace. Two guiding examples, a self-recuperative flameless unit and Fo.Sper furnace, are then identified to show the main steps involved in the V&V methodology. These include the determination of the total error in the experimental data, including the statistical reproducibility and experimental error, the evaluation of the numerical error and modelling uncertainty, and the quantification of the level of agreement between experimental data and numerical simulations. Finally, the consistency between numerical models and experimental data is discussed, showing the potential of V&V of identifying which values of the uncertain input parameters allow defining a consistent predictive model.