Analysis of Uncertainties in Infrared Camera Measurements of a Turbofan Engine in an Altitude Test Cell

The infrared (IR)signature ofa jet aircraft engine in altitude operation is a key component for the design of effective IR countermeasures and low-emission engines. Predicting the signature with radiometric models is widely accomplished, but measurements in situ are crucial for model verification. The altitude test cell provides a venue for measuring the IR signature in a simulated altitude environment, but the facility is designed for testing engines, not IR imaging. As a result, the imaging in the test cell is laden with measurement uncertainty due to stray radiation from the facility structure, hot exhaust gases, and the measurement equipment itself. Postprocessing using correction factors is necessary to extract the engine signal from the stray radiation. The correction factors, however, inject an additional level of uncertainty in the measurements. The National Aeronautics and Space Administration (NASA) Glenn Research Center measured the IR signature of a jet aircraft engine in an altitude test cell in the summer of 2002.They reported measurement uncertainty as the foremost concern. With NASA's efforts as the prime motivation, this research investigated the uncertainties in measuring the IR signature of a General Electric F110-GE-129 turbofan engine inside an altitude test cell. The engine is measured by an IR camera immersed in the hotexhaust gases 35 feet downstream from the on-engine axis view. A protective enclosure and zinc selenide (ZnSe) window shield the camera from the heat and vibrations of the plume. The requirements for the IR measurement system include the apparent intensity and radiance of the visible engine surfaces in three bands of operation, two Medium Wave IR (MWIR) bands and one Long Wave IR (LWIR) band with a spatial resolution of one inch. To explore the extent of the measurement uncertainties, a radiometric model of the altitude test cell at NASA is formulated to quantify the engine and stray flux. To increase the fidelity of the model