Flaw-detection system automates automotive inspection

Flaw-detection system automates automotive inspection

Capitalizing on its patented 3-D moiré technology, The Budd Company (Troy, MI), in conjunction with the Industrial Technology Institute (Ann Arbor, MI), has developed a method to automatically classify stamping defects in large, nonpainted, sheet-metal panels (see Vision Systems Design, March 1997, p. 32). In operation, the system uses moiré techniques to find scratches and dents in sheet-metal panels as large as 3 ft in diameter. Surface flaws as small as 0.0006 inches deep can be detected and classified.

In April, Budd awarded Medar (Farmington Hills, MI) a contract to commercialize the system. To produce the system, Medar will integrate VisionBlox software from Integral Vision (Farmington Hills, MI), a subsidiary of Medar, to provide a Windows interface for moiré inspection. "As a result," says Fred Sittel, senior staff engineer with Budd, "a complex technology will be turned into an easy-to-operate system that provides quantitative information about shape deviations and surface shape."

Moiré methods operate by superimposing two line grids on a part to produce a "beat" pattern. Using CCD cameras, a 3-D digital representation is extracted from the shape of this pattern. Through high-speed processing, the system can make a 3-D representation of the metal panel, consisting of 250,000 data points, in seconds. "The primary benefit of Budd`s design is the ability to use z-axis data to measure contours, profiles, straightness, flatness, roundness, cylindricity, and angularity characteristics in large sheet-metal panels,"says Sittel.

In operation, the inspection system acquires a reference image of a known good panel and compares new panels to the 3-D reference image. Deviations from the known panel are quantified and labeled as defects based on operator-selectable parameters. More information is available from Lisa Super at [email protected].