Analytical and Experimental Results of the Coefficient of Thermal Expansion of High-Modulus Graphite-Epoxy Materials

Abstract

The coefficient of thermal expansion (CrE) of an arbitrary laminate can be determined by the Classical Laminate Theory. However, the results obtained have shown that the CFE is very sensitive to some orthotropic elastic constants and to the laminate layup. In particular, the non-Hookean behaviour of an unidirectional lamina in the fibres direction should have to be taken into account to exactly predict the CTE. In order to verify the theoretical analysis, a new test facility has been designed to carefully measure the CTE in specimens made of advanced composite materials having a quasi zero value of CTE. The facility is basically composed of two thermostats and contained in a vacuum chamber (10−2 Pa). Four specimens are enclosed in the first thermostat and the operating temperature range is limited to −20°C up to 80°C. Four displacement transducers are enclosed in the second one at a temperature of 20°C and kept rigorously constant within 0.1°C. The specimens are mounted on Ultra Low Expansion (ULE) frames having a CTE of 1 × 10−8/°C. It has been possible to miniize the measurement error in the CTE, by a careful choice of displacement sensors and the high control of their thermal stability. The following CTEs have been obtained along the fibre direction, normal to the fibre direction and for a quasi-isotropic laminate, respectively, for two High-Modulus (HM) graphite/epoxy materials. GY70/M18: α1 = −1.04 × 10−6/°C; α2 = 33.9 × 10−6/°C; α x = -0.095 × 10−6/C M55J/M18: α1 = −1.02 × 10−6/°C; α2 = 35.2 × 10−6/°C; α x = −0.162 × 10−6/°C The analytical results have shown as a variation of ±1° only in the lamina orientation can change the CTE of the quasi-isotropic laminate up to ±50.5% of the theoretical value; as well as, a variation of ±5% in the physical and mechanical properties can change the CTE up to ±48%.