Abstract
The thermal properties and boundary conditions of real (not theoretical) thermal systems are imprecise or uncertain. As a consequence, the temperatures and heat fluxes exhibit variability which is usually expressed in terms of a standard deviation. Generally uncertain system parameters are assumed to be stochastic. However, in real life it is highly probable that they are better described by fuzzy variables. This paper describes how to compute the variability for parameters which are stochastic, fuzzy, or both. It is shown that, for the problem considered here, a first order perturbation analysis gives reasonably accurate estimates of the system variability for both types of parameters.