Effects of cyclic thermal loads on IM7/8552 composites

Derrick A Jensen, Purdue University

Abstract

Thermal cycles ranging from room temperature to 400°F and from room temperature to 550°F were performed on the IM7/8552 composite material. It was found through tension testing that the resin dependent strength properties, such as the σ12 and σ22 strength, decreased in strength as the number of thermal cycles increased. The σ12 strength was cycled at both 550°F and 400°F, and the σ22 strength was cycled at 400°F. It was also found while cycling at 400°F that the E1 and E2 modulus don’t change as the number of thermal cycles increase, and the σ11 strength degrades initially due to its dependence of the resin for transferring stress from fiber to fiber, but then it’s strength does not degrade further as the number of thermal cycles increase. The E1, E2 modulus, and σ11, σ22 strength were not tested for after cycling at 550°F. Curve fitting methods were applied to the data to establish a model capable of performing failure prediction of any IM7/8552 composite laminate thermally cycled within the limits of this research. The shear strength was tested for after thermal cycling from room temperature to 550°F and it was found that after five thermal cycles the strength reduced by more than 42%, and about 60% after 15 thermal cycles. While thermal cycling from room temperature to 400°F the E1 and E2 modulus were tested, as well as the σ11, σ22, and σ12 strengths. The σ22 and σ12 strength were certainly affected but not as critically as when thermal cycling to 550°F. The σ12 strength only reduced by 6% after 15 thermal cycles, and then after 100 thermal cycles it reduced by 16%. The σ22 strength reduced by 62% after 100 thermal cycles. It was expected it would drop more than the σ12 strength since the σ22 strength is more dependent on resin strength. It was determined that the cause of the degradation was due to the resin system of the composite material being burned off during thermal cycling. Keeping track of the mass property during thermal cycling helped obtain this conclusion. While thermal cycling to 550°F more than 3% of the mass was lost after five thermal cycles and more than 4% after 15 thermal cycles. X-rays of the [45/-45/45/-45]S specimen were performed to see if microcracking due to residual thermal stresses might be a contributing factor, but the x-rays showed no signs of microcracking within any of the specimen. Similarly, while thermal cycling at 400°F the mass dropped by 0.6% after 15 thermal cycles and 1.2% after 100 thermal cycles, with no signs of microcracking showing up on the [45/-45/45/-45]S specimen x-rays. The difference in the amount of mass loss between thermal cycling to 550°F and then to 400°F is attributed to the fact that 550°F is 150°F higher than 400°F. Two different curve fitting methods were used to best fit the data obtained from tension testing, a linear regression model using least squares method, and a first order exponential decay model. The linear model fit the modulus and strength data obtained from thermal cycling to 400°F the best. The exponential decay model fit the shear strength data from thermal cycling to 550°F and the mass data from both thermal cycling tests the best. If more thermal cycles were performed on the specimen showing a linear decreasing trend, it is assumed the data would eventually flatten out to some asymptotic value, much like the exponential decay model, since it isn’t possible for thermal cyclic loads to drive a material to failure all by itself.

Degree

M.S.E.

Advisors

Sun, Purdue University.

Subject Area

Engineering|Aerospace engineering|Mechanical engineering

COinS