High-Resolution Moiré Interferometry for Quantitative, Low-Cost, Real-Time Surface Profilometry
- Project Type: Directed
- Directed Project Contributors: J. Josiah Steckenrider, J. Scott Steckenrider
Purpose / Abstract
This project involved the development of an inexpensive hand-held system to measure fastener flushness on jet aircraft.
Introduction / Background
The Air Force had an interest in developing an instrument for measuring the relative surface height of fasteners (rivets, screws, etc.) on its aircraft with a precision of 0.0005” (~1/8 a human hair width) without making contact with the surface being inspected. At the extremely high speeds at which many military aircraft are designed to fly, even the smallest protrusion or recess of a small rivet or screw head can cause significant turbulence in flight. Thus, the Air Force requested proposals from for an instrument design that would be hand-held and could ultimately be built in mass production for $5,000. Although there are already instruments available that meet this need, they cost significantly more (>$100,000) and are exclusively bench-top systems that cannot operate “in the field.” We proposed that we use the phenomenon of Moiré Interferometry to construct a system to meet the requirement using materials available in the lab here at Taylor. Over two summers we worked in a faculty-student collaborative research project to successfully develop a technology that meets the Air Force’s stated specifications, albeit in a bench-top format.
The instrument we developed works by the method of Moiré interferometry. Moiré interferometry is the phenomenon that occurs when two similar but slightly altered patterns are overlaid, such as occurs when looking through two layers of sheer drapes covering a window. Even though the individual threads in the sheer are too fine to resolve, when the two layers overlap a more coarse pattern of bright and dark bands appears due to the slight misalignment of those threads, and this coarse pattern is clearly visible to the eye. An example of this interaction is shown in Figure 1 in which the top two slightly different patterns are overlaid to form the bottom pattern whose repeat structure (Moiré fringes) is much more coarse than either of the constituent patterns themselves. These bright and dark bands represent a Moiré pattern, and provide detailed information about the threads’ alignments, even though they cannot be directly discerned. Our project involves optically projecting a pattern of extremely fine lines onto the contoured surface of an object, such as the face of a quarter, and overlaying that pattern (or more accurately an image of that pattern) onto the extremely fine lines in a digital camera’s focal array. What results is a coarse pattern of bright and dark bands in the camera image that represent the contour of the surface, much like topographical lines on a geographic map. By quantitatively measuring these Moiré patterns we can measure the height of the surface with a precision below 1/1000th of an inch. Having demonstrated this capability, we subsequently transitioned the technology during the 2015 academic year into a prototype device designed and built by engineering students in our senior-level Capstone course. Because the system involves computer hardware, software, electrical and mechanical components it is ideally suited for the multi-disciplinary teams of computer, electrical and mechanical engineering physics majors that work together in our interdisciplinary project courses.
To demonstrate the effectiveness of our approach, this process was carried out in real-time (that is, at normal video frame rates of 30 frames/second) on a US quarter with protrusions on the order of 200µm (0.008"). Figure 2 shows the profile of the quarter obtained using white light and telecentric lenses for projecting and imaging the sinusoidal fringes. Each individual letter on the surface of the quarter dollar is clearly distinguishable, and even the strands of hair are visible. The standard deviation of a flat region on the coin was 2.85µm, so that the minimum detectable step size (12.7µm) is 4.46 standard deviations away from the mean. This means that the minimum feature size required to meet the required sensitivity of deviation of 12.7 µm would be two pixels in a 640 x 480 pixel image. Since the fasteners in question would occupy a significantly larger percentage of the image, this method has demonstrated more than sufficient sensitivity to meet the required performance characteristics.
Resources / Links
J. Josiah Steckenrider and J. Scott Steckenrider, "High-Resolution Moiré Interferometry for Quantitative, Low-Cost, Real-Time Surface Profilometry", Applied Optics, 54 (28), pp. 8298-8305 (2015)