Abstract
The aim of this study is the assessment of the effect of input data uncertainties on the predictions achieved for a ventilated room by means of a computational fluid dynamics (CFD) technique. We first describe the characteristics of the model (experimental setup and numerical approach). This is followed by a review of sensitivity and uncertainty analysis methods in order to choose the most appropriate one. Taking into account the numerical model particularities, a differentiation analysis with finite differences is applied. The methodology is based on an automatic generation of the perturbation size imposed on the inputs. Experimental-numerical confrontations concerning the uncertainty domains are given for air mean velocity and temperature within the enclosure, as well as for room wall heat transfer calculations.