Optical coherence tomography, recognized in biomedical circles as an excellent imaging method, is expanding its field of view into materials applications, most recently the nondestructive imaging of composites. Optical coherence tomography imaged a glass-reinforced woven composite. Dark blue features are very intense reflections from where the glass fibers cross over each other, visible from both the top and side views. Courtesy of NIST.With help from the technique's innovators at the Massachusetts Institute of Technology in Cambridge, researchers at the National Institute of Standards and Technology (NIST) are applying the technique to imaging composite parts so they can visualize and predict flow patterns during molding. Composites took wing with the development of stealth aircraft but have since reached beyond military applications to automotive and large-appliance markets. Like engineered plastics, they are composed of resin and reinforced with materials such as carbon fiber. But unlike plastics, the fibers are oriented and sometimes woven together to produce remarkably strong yet lightweight products. The research could provide a means to predict and improve how the resin permeates a fiber weave. Permeability is a big problem in composites, according to Joy Dunkers, NIST's lead researcher on the project. "If you could model a part to figure out how molds are filling, where there are dry spots, where you could put injection pourage or outlets, you could adjust cycle time. It would be a big advancement." NIST's researchers compared optical coherence tomography and optical microscopy in evaluating a glass epoxy and glass vinyl-ester composite. The data obtained from each were comparable. The difference was that optical coherence tomography obtained them more quickly and nondestructively."It provides 10- to 20-µm spatial resolution, which is fine unless you have 5-µm-diameter fibers. We're working on that," said Richard Parnas, group leader for polymer composites at NIST. Part of that work entails revamping the instrument for nonbiological material. "I can tell you for certain that you want to scan the reference mirror slower for composites -- say, at 80 kHz -- than for biological systems -- 400 kHz," said Dunkers. "In biological systems, you are often interested in time-dependent phenomena and would want a faster scanning system. Also, you are much less limited by scattering in biological systems, so you can afford to scan faster and sacrifice sensitivity." The same argument applies to confocal parameter, added Dunkers. "You really want to see as deep as possible, so you would want a large confocal parameter without sacrificing resolution. You can always make up for the increase in spot size by oversampling."