Ultrasonic Studies of Composites Undergoing Thermal and Fatigue Loading
New composite materials possess attractive properties for use in advanced aircraft. A necessary requirement for their introduction into aeronautic use is an accurate understanding of their long term aging processes so that proper design criteria can be established. In order to understand those properties, these composites must be exposed to thermal and load cycles that are characteristic of flight conditions. Additionally, airline companies will require nondestructive evaluation (NDE) methods that can be used in the field to assess the condition of these new materials as they age. As part of an effort to obtain the required information about new composites for aviation use, we are performing ultrasonic measurements both in the NDE laboratory and in the materials testing laboratory at NASA. The materials testing laboratory is equipped with environmental chambers mounted on load frames so that composite samples can be exposed to thermal and loading cycles representative of flight protocols. Applying both temperature and load simultaneously will help to highlight temperature and load interactions during the aging of these composite materials. This study reports on our initial ultrasonic attenuation results from thermoset and thermoplastic composite samples. Ultrasonic attenuation measurements have been used reliably to assess the effects of material degradation. For example, recently, researchers have shown that by using frequencies of ultrasound on the order of 24 MHz, they could obtain adequate contrast in the evaluation of thermal degradation in these composites. This paper will present data that shows results at a lower frequency range. In addition, we report results on the frequency dependence of attenuation as the slope of attenuation with respect to frequency, beta = delta alpha (f) / delta f. The slope of attenuation is an attractive parameter since it is quantitative, yet does not require interface corrections like conventional quantitative attenuation measurements. This is a consequence of the assumption that interface correction terms are frequently independent. Uncertainty in those corrections terms compromises the value of conventional quantitative attenuation data.