H. J. Jin

Synchronous thermal wave IR video imaging for nondestructive evaluation

We describe a system for real-time processing of infrared video images in synchronism with the time-dependence of the target objects temperature. The system can either be used either with periodic or pulsed heating of the target. With periodic heating, the system operates as if it were a collection of lock-in amplifiers, one for each of the quarter of the million pixels of the image. With pulsed heating, it operates as if it consisted of a similar number of box-car averagers. In both cases, the signal-to-noise ratio and temperature sensitivity of the infrared camera are improved. The technique lends itself to a wide spectrum of NDE applications.

IR Thermal Wave Tomographic Studies of Structural Composites

Vavilov et al [1] have recently described a technique for making tomographic thermal wave images. Their method involves recording a succession of thermal wave images after a flash-heating pulse, followed by a numerical pixel-by-pixel search of the images for the time at which the reflected thermal waves from subsurface features have their peak amplitudes. Since the peak time is related to the depth of the scatterer, this information enables one to separate the image into time (or depth) slices. The result is a thermal wave tomogram. Since their process involves post-processing and a search through a large number of stored images, it is memory-intensive, and is difficult to accomplish in real time. In the present paper, we report a thermal wave tomographic method which accomplishes the same result, but does so with real-time techniques which avoid the storage of a large number of images, and produces the tomogram without post-processing.

Real Time Thermal Wave Tomography

We report a thermal wave tomographic method that utilizes real-time techniques. A reference cooling curve, corresponding to a surface region beneath which there are no subsurface scatterers, is subtracted from the time dependence of every pixel of the image as it is generated. This is followed by a pixel-by-pixel determination of the peak contrast times of these subtracted curves and the peak times are stored as a 512 x 480 tomogram, all of which is done in real-time. We also store a peak-contrast image in which each individual pixel is displayed as it appeared when its contrast with the background was maximum. Currently, we are able to carry out this tomographic imaging at sampling rates up to 10 Hz. The process has been implemented with two different systems, both using an Inframetrics IR-600 Camera for data collection. In one of the systems the real-time processor is a series of DataCube boards on a VME bus controlled by a Sun 3/160 workstation. The other system is comprised of a Perceptics NuVision processor under the control of a Macintosh II microcomputer.

Infrared Thermal Wave Studies of Composites

Thermal wave techniques have received increasing attention for use in the inspection and characterization of composite materials. These techniques have the advantages of being contactless, rapid requiring access to only one surface of the object under inspection. In this paper we demonstrate the ability of the technique to image flaws in several types of composite materials and to measure the depths of the flaws.

Thermal Wave Detection and Analysis of Adhesion Disbonds and Corrosion in Aircraft Panels

Pulse-echo ultrasonic techniques are widely used for the purpose of nondestructive evaluation (NDE) of materials. Knowledge of the elastic wave velocities in a material permits time-of-flight measurements to be converted into measurements of the subsurface depths of various subsurface defects which are strong elastic wave scatterers. In recent years, the authors and others have utilized an analogous pulse-echo technique to study thermal wave scattering from subsurface defects [1–4]. In this technique, the waves are launched as the result of energy deposited at the sample surface from a suitably pulsed heat source, and the resulting time evolution of the surface temperature is monitored by means of an infrared (IR) camera. Descriptions of the detailed methods for processing the data stream from the camera in real time, so as to effect box-car or lock-in image processing, have been presented elsewhere [1–4].

Real-Time Thermal-Wave Imaging of Plasma-Sprayed Coatings and Adhesive Bonds Using a Box-Car Video Technique

We describe an IR thermal-wave system which does real-time, pixel-by-pixel, box-car averages of images with a pulsed heat source. The result is greatly improved sensitivity to subsurface defects.

Mapping thermal diffusivity

We describe applications of fast IR cameras to map the thermal diffusivity of a slab of material. The methodology is based on curve-fitting techniques like those described at the 1998 Review of Progress in QNDE. We will present a comparison between results obtained using a laboratory camera with those obtained using a portable camera suitable for field applications. The method has the potential for making sufficiently rapid and accurate measurements that it can be used to map the diffusivity over a slab for process-control applications, essentially in real time.—Research supported by the Institute for Manufacturing Research, Wayne State University.

Real-time asynchronous/synchronous lock-in thermal-wave imaging with an IR video camera

We describe an IR thermal-wave system which does real-time, lock-in imaging which is asynchronous with the camera’s scanning. This technique is compared with the previously reported synchronous version of lock-in imaging.

Colormap Based Image Display Enhancement

A number of well developed image enhancement techniques, basically categorized into global image processing and adaptive image processing, [1, 2] have been widely used in NDE applications, which help to present high quality images conveying information on voids, cracks and inclusions in samples. These enhancements apply certain algorithms on image data and change the values to generate particular effects. Since the visual effect of an image is determined by two factors, i.e. image data (pixel value) and colormap which assign color display for each intensity covering the whole range, the enhancement of the display of an image can be achieved by working on the colormap for a pseudo-color display without actually affecting the image data. A grey scale display can be treated as a special case of pseudo-color. With the advent of high quality color monitors and color output devices available for the scientific workstations whose colormaps can be easily redefined by users, it often is more efficient to manipulate the colormaps to achieve the desired visual effect. Since a typical image file with 8 bits/pixel data can easily contain 250,000 bytes while a colormap for the display of such image requires only 768 bytes for 24 bits/color (corresponding to 16.7 million different shades of colors), the gain in speed obtained by adjusting the colormap rather than the data can be very dramatic.

Imaging and quantitative measurement of corrosion in painted automotive and aircraft structures

Some of the authors have shown that it is possible to image and make rapid, quantitative measurements of metal thickness loss due to corrosion on the rear surface of a single layer structure, with an accuracy better than one percent. These measurements are complicated by the presence of thick and/or uneven layers of paint on either the front surface, the back surface, or both. We will discuss progress in overcoming these complications. Examples from both automotive and aircraft structures will be presented.—This material is based in part upon work performed at the FAA Center for Aviation Systems Reliability operated at Iowa State University and supported by the Federal Aviation Administration Technical Center, Atlantic City, New Jersey, under Grant number 95-G-025, and is also supported in part by the Institute for Manufacturing Research, Wayne State University, and by Ford Motor Company. Supported by a Grant from Ford Motor Company.