Field of View Determination for Shearography Nondestructive Evaluation John W. Newman, President Laser Technology Inc. Norristown, PA 19403 USA Introduction Shearography NDT is a highly developed, mature technology applicable to the structural inspection of a wide variety of applications, from buildings and civil engineered structures to aircraft and spacecraft panels to microcircuits. The field of view (FOV) used for these tests vary widely from 5-10 sq. meters to less than 1.0 sq. cm. In order to obtain the maximum throughput for in-process shearography inspection it is necessary to examine the important factors affecting the output performance of an NDT system as measured by the Probability of Detection and to calculate the maximum field of view for a specific application. Key factors affecting the field of view are camera pixel count in x and y direction, the maximum allowable defect size (mm. or inches) for a specific structure and application, defect indication definition (area), noise. This paper presents the step by step analysis and background to perform the field of view calculation to achieve maximum inspection throughput at minimum cost yet achieving the required defect sensitivity. The maximum field of view for a shearography camera is determined by the following variables: 1. Shear Camera CCD pixels in horizontal and vertical axis, Px and Py, in pixels 2. Dimensions, X and Y, of the largest allowable defect, Dx and Dy, in mm. or inches 3. Minimum pixel count required to define a defect indication, Ix and Iy in pixels 4. Optical Resolution of Shearography System 5. Signal to Noise ratio, S/N These variables are used to calculate the maximum field of view to achieve a specific probability of defect detection. The maximum allowable defect size, (Dx, Dy) is determined through engineering fracture mechanical techniques or failure analysis of the structure. Defects below this size are benign; defects that are large must be recorded for analysis and disposition. The critical question is to determine how large a field of view can be established to detect the critical flaw sizes 95% or more.
Definition of Defect Indication All optical inspection technologies requiring an operator to identify defects on a monitor, such as video imaging, shearography, thermography and real-time X-Ray require a definition of a minimum defect indication size, measured in pixels. A single pixel in an entire computer monitor screen can not practically be detected by an operator and used to define a defect. Camera pixel sensitivity is not uniform and all CCD detectors have missing pixels or pixels with variable electrical response to incoming light. Multiple pixels together having a measurable signal to noise ratio (S/N Ratio) must be used to define a defect. Extensive tests by government laboratories and industry have led to conventions requiring pixel counts of 5x5 pixels to 12x12 pixels, with a signal to noise ratio greater than 1.1 to be used to define a defect. Test at LTI have shown a 95% reliability/confidence factor is achieved with a defect definition of 10x10 pixels. Fig. 1 shows the PoD curves for three operators presented monitor indications having pixel counts 3x3, 5x5, 7x7 and 10x10. All operators detected the indication at the 10x10 size more than 95% of the time. For a typical monitor image, with 1350 pixels x 1024 pixels, these results indicate a defect measuring approximately 0.74% of the screen width x 1% of the screen height is detected by operators 95% of the time or more. Fig. 1 Probability of defect detection plotted as a function of defect pixel dimensions for three operators. While some operators will detect defects with smaller indication pixel counts, all operators will identify defects with a size of 10x10 pixels 95% of the time or more.
Calculating Shearography Image Scale, Is With a minimum defect indication defined as (Ix, Iy) =10x10 pixels, we can determine the required image scale or resolution required for any shearography camera. For the maximum field of view, the dimensions of the Largest Allowable Defect, (Dx, Dy) in mm. or inches, must allow detection better than 95% of the time, hence we can set Dx/10 and Dy/10 as equal to the Image Scale in mm/pixel. For example, with a maximum allowable defect of 9mm. the image scale, or Is, is 0.9mm/pixel. Calculating the Field of View With the Image Scale defined for a specific application, the field of view is then simply The Image Scale times the shear camera CCD pixel count in the x and y directions, Px and Py. Horizontal Field of View = Is (Px) Vertical Field of View = Is (Py) In the example above, where the maximum allowable defect is 9mm, the Is = 0.9 mm/pixel. For a shearography camera such as the LTI-5100HD, with a CCD resolution of 1350 x 1024 pixels, the field of view to image a 9 mm defect 95% of the time or more is: Horizontal FOV= 0.9mm/pixel x 1350 pixels = 1215 mm Vertical FOV = 0.9mm/pixel x 1024 pixels = 921.6 mm The total area of the field of view is 1215 x 921.6 mm = 1.12 sq.m Additional Factors The above calculation represents an idealized FOV calculation. In the real world, additional factors must be examined that may degrade image quality and reduce the probability of defect detection. Shearography NDT requires uniform stressing of the panel or structure to achieve identical strain change response of the homogeneous material. Variations in thermal load changes, acoustic signals can lead to variable defect detection response. It is important to apply load changes during shearography NDT in a highly repeatable manner. Second, the use of an NDT Reference Std. Panel, with a representative flaw sized to the Maximum Allowable Defect dimensions is critical. This panel should be used to evaluate defect sensitivity over the field of view. A weak indication, with a measured S/N ratio less than 1.1, will require a reduction in the Image Scale (Is) factor for the test.
Other factors, unique to shearography, such as interference from air currents or test panel movement due to acoustical noise can also cause a reduction in the necessary Is and hence overall field of view. Conclusion The determination of the maximum field of view for shearography NDT is a critical factor in the system output Probability of Defect Detection and resulting product quality. The FOV is application specific. Variables that must be defined for a specific application include the required Maximum Allowable Defect, shear camera pixel count and test data to support the definition of detectable defect area. Further, software tools for measuring the defect indication S/N ratio are critical to assuring correct establishment of the FOV for a given test panel to detect critical defect sizes. The author, John W. Newman, is President of Laser Technology Inc. Norristown PA 19073 USA, a company he founded in 1980. He has worked in the field of aerospace engineering, holography and shearography NDT since 1968. Mr. Newman is the Chairman of the ASNT Laser Methods Committee, Chairman of the ASTM Shearography Committee, E07.11.3, and has served on the NASA NDT Advisory Board for the last five years. He is the author of more than 80 papers and articles on Shearography NDT and developed the world s first commercial electronic shearography camera in 1987 working with Dr. Y.Y. Hung at Oakland University. Mr. Newman holds 38 patents in the field of shearography NDT technology.
Illustrative Examples Fig. 2 Deformation of brick wall of historic building. LTI-5100 with a FOV of 2x3 M. Fig. 3 Aircraft laminate panel, four defects with a FOV 1.2 x 0.9 M. Close up, 30x30 cm. Fig. 4 1 M. FOV Aluminum honeycomb core with graphite face sheets, 14mm defects