Detection and Characterization of Package Defects and Integrity Failure using Dynamic Scanning Infrared Thermography (DSIRT)

A dynamic scanning infrared thermography (DSIRT) system developed at the Univ. of Illinois Urbana-Champaign (UIUC) Packaging Lab relies on variation in transient thermal artifacts to indicate defects, and offers the possibility of characterization of many types of materials and structures. These include newer polymer and laminate-based structures for shelf-stable foods that lack a reliable, nondestructive method for inspection, which is a continuing safety issue. Preliminary trials were conducted on a polyester/aluminum foil/polypropylene retort pouch laminate containing artificially-induced failed seal and insulating inclusion defects ranging from 1 to 10 mm wide in the plane of the seal. The samples were placed in relative motion to a laterally positioned infrared laser, inducing heating through the plane of the seal. The emergent thermal artifact on the obverse side was sensed using either a bolometer camera or a thermopile sensor, with thermal anomalies indicating potential defects and the results of each sensors were compared. The bolometer camera detected defects to the limit of its measured optical resolution-approximately 1 mm at 20 cm-although the lower-resolution thermopile sensors were only capable of detecting 5 mm defects even at closer distances of approximately 5 mm. In addition, a supplementary magnification system was fitted to the bolometer camera which increased resolution but reduced field of view and would require a much higher frame rate to be useful. Automatic processing of the image data rapidly detected the model defects and can lead to development of an automated inspection system.  Much higher material throughput speeds are feasible using faster instruments, and the system is scalable.

Experimental spatial sampling study of the real-time ultrasonic pulse-echo BAI-mode imaging technique

The ultrasonic pulse-echo backscattered amplitude integral (BAI)-mode imaging technique has been developed to inspect the seal integrity of hermetically sealed, flexible food packages. With a focused 17.3-MHz transducer acquiring radio frequency (RF) echo data in a static rectilinear stop-and-go pattern, this technique was able to reliably detect channel defects as small as 38 microm in diameter and occasionally detect 6-microm-diameter channels. This contribution presents our experimental spatial sampling study of the BAI-mode imaging technique with a continuous zigzag scanning protocol that simulates a real-time production line inspection method in continuous motion. Two transducers (f/2 17.3 MHz and f/3 20.3 MHz) were used to acquire RF echo data in a zigzag raster pattern from plastic film samples bearing rectilinear point reflector arrays of varying grid spacings. The average BAI-value difference (deltaBAI) between defective and intact regions and the contrast-to-noise ratio (CNR) were used to assess image quality as a function of three spatial sampling variables: transducer spatial scanning step size, array sample grid spacing, and transducer -6-dB pulse-echo focal beam spot size. For a given grid size, the deltaBAI and CNR degraded as scanning step size in each spatial dimension increased. There is an engineering trade-off between the BAI-mode image quality and the transducer spatial sampling. The optimal spatial sampling step size has been identified to be between one and two times the -6-dB pulse-echo focal beam lateral diameter.