On the Effect of Material Properties Uncertainty on the Magnitude of Temperature Oscillations in Two-Component Periodic Laminate: Tolerance Averaging Modelling

Abstract

Over last decades composite materials gained even more interests in many industries due to theirs effective properties which may be apparently different (better) then constituents itself. By specific layout and distribution of composite components we can achieve desired properties in macro scale, e.g. high elasticity and low conductivity at the same time. On all interfaces, by perfect contact between phases, there will appear jump of gradient field (displacement or temperature) of unknown magnitude. This magnitude depends in fact, for the heat transfer problem, inter alia on the ratio of conductivities of composite constituents lying next to interface. Values of these oscillation magnitudes are case of our study here. The conductor under considerations is a two-phase, periodic laminate subjected to initial-boundary conditions assuring unidirectional heat flow, perpendicular to the laminas. Ratios of material properties are assumed as random variables of known probabilistic distribution. We will give an answer to the question: is the jump function of temperature gradient a random variable of Gaussian distribution In order to have a good description to considered problem we have decided on the use of tolerance averaging technique.