C. W. Swonger, Alex Klooster Jr., Dr. Carl C. Aleksoff And Dr. Bin Li 0000-00-00 00:00:00
The pace of progress in precision manufacturing in all industry sectors is creating an ever-increasing demand for new and innovative methods for noncontact inspection of manufactured part and assembly surfaces. Sometimes existing products can be found to perform the required function in terms of resolution, speed, robustness, field-of-measurement, depth-of-measurement, sensitivity and the many other factors that define an application’s requirements. But more often, the desired higher level of manufactured item quality and performance requires the development of new solutions that stretch the state of measurement and inspection technology in one or more critical directions. Often three-dimensional (3-D) inspection is a critical part of the solution. In most industrial and commercial sectors, the benefits of automated noncontact surface inspection are clear and substantial. They include most or all of the following factors: • Significantly reduced scrap or rejected product. • Reduced manufacturing costs and reduced manufacturing downtime. • Higher customer-perceived product quality. • Reduced warranty costs and product liability. • Higher product performance and customer satisfaction. The technologies of lasers, optics, dynamically controlled illumination, photonic sensing, image and signal processing (or vision) algorithms and software, and even mechanical part and surface positioning all come together as multidimensional ingredients of each advanced noncontact surface inspection solution. All of these technologies are advancing steadily albeit at different paces. What was not possible last year and last month is now often readily achievable with expert system engineering and integration. Technical solutions offered to satisfy proven large markets are often not as capable as can be provided to meet a specific focused application’s requirements. SOLUTION REQUIREMENTS DEFINITION The starting point of an advanced application solution is the careful, knowledgeable and quantitative definition of the complete set of application requirements. This involves much more than simply stating what must be detected or measured and how rapidly the solution must operate. As Albert Einstein famously said: “Things should be kept as simple as possible, but no simpler.” Surface inspection inherently entails detecting and often measuring surface phenomena—some considered to be defects and others considered to be just acceptable artifacts or desired features. These have some range of sizes, visibilities, optical spectral, scattering, polarization and other characteristics. The surfaces themselves also have important characteristics—perhaps consistent and perhaps highly variable in space or time. An application’s end-user and the solution development engineering team must work together to identify all parameters and trade-offs that will define the bounds of acceptable and optimum solutions for the application. A few of the usual questions include: 1. What are all of the types of phenomena (such as defects) to be detected or measured? 2. What range of 3-D or two-dimensional (2-D) sizes of those phenomena must be detected or measured? 3. What range of surface shapes, or what one nominal surface shape, is to be inspected? 4. What is the required probability of detection of those phenomena? 5. What will be the physical condition of the surface when inspected (such as dirt, dampness, oil, etc.)? 6. What is the allowable false alarm rate—the rate of detections when no defect actually exists? 7. What is the desired or required amount of surface to be inspected per unit time? 8. What are the spectral characteristics of the surface and the defects? 9. What will be the various system operator’s skill levels and form of output information that they desire? 10. What variety of phenomena on the surface can occur that should NOT be considered to be defects? Agreement on clear and quantitative answers to questions such as these is a key starting point for ensuring that all expectations are aligned and a responsive development can be proposed and executed. EXPLOITING THE PACE OF TECHNOLOGY PROGRESS A substantial range of technologies is available, each best suited to a subset of inspection applications defined by those answers. With today’s newest technologies, an advanced surface inspection solution can replace tedious, distracting and fatiguing manual tasks, detect defects not visible to a human operator, process surface areas faster than human visual perception can accomplish and perform detection, as well as the measurements that often are critical in judging defect severity. Recent underlying technology breakthroughs of particular value include: • High-speed computer-controlled dynamic illumination of surfaces with combinations of wavelengths and spatial patterns that bring out subtle but critical characteristics. • Real-time model-based algorithms that relate sensor data to the design reference data, which defines a correct surface for either a part or a complex assembly. • Knowledge-driven sensor perspective and positional control that allows curved surfaces to be rapidly captured and analyzed. • Interferometric sensing configurations that can accommodate large surfaces of more than one square meter. Along with advances such as these, the availability of ever-increasing lowcost computer power and the extended performance and reliability of light sources ranging from lasers to LEDs allow reliable systems to be devised for continuous operation on plant f loors or elsewhere. SUCCESSFUL DEVELOPMENT STRATEGIES The most prudent approach to accomplishing an advanced solution development is a step-by-step experimental and analytical series of planned project phases. These typically consist of: 1. An experimental and analytical feasibility investigation phase. This is designed to “home in” on the best overall approach to meeting the original requirements that were as defined as described above. The result will be a report of feasibility findings and a description of the proposed system concept. This phase may require weeks to a few months. 2. The engineering prototype phase. In this phase, a working implementation is created to demonstrate the fundamental capability that is required. At this stage, system packaging and peripheral features or functions may be deferred in the interest of achieving an economical “proof of principle.” Examples of deferred features might include speed or operator convenience functions. This phase may take from several months to as long as a year. 3. A fielded prototype (if quantities of systems will ultimately be desired) or a final system delivery phase. In this phase, a complete packaged system solution will typically be built, tested, validated and installed either in an evaluation setting or in the final end-use venue. This phase may require several months to as long as a year for completion depending on the operating and environmental requirements for the ultimate system. 4. If the desired solution will be desired in substantial quantity, then the fielded prototype phase may be followed by a manufacturing engineering phase aimed at cost minimization, producibility engineering, reliability engineering or other objectives. The result will normally be a production prototype. Several months will usually suffice for completing this phase. The manufacturer today who seeks to supply products with high perceived and actual quality at competitive costs must seriously examine how his product’s external and internal surfaces and appearance will be reliably and 100% inspected at high resolution, while not adversely impacting his production rates and other operating requirements. The specific types of variances and defects to be detected and avoided will differ across product classes. Development of noncontact optical inspection solutions, focused on a specific product’s characteristics and requirements, often can provide a level of product quality assurance that is not affordably achievable by either human visual inspection or by general-purpose inspection products. C. W. “Ron” Swonger is director of contract R&D, Dr. Carl C. Aleksoff is consulting optical scientist, Alex Klooster Jr. Is consulting electro-optical engineer, and Dr. Bin Li is chief engineer of lasers and optics at Coherix (Ann Arbor, MI). For more information, call (734) 922-4066, e-mail RonS@ coherix.com or visit www.coherix.com. QUALITY ONLINE For more information on noncontact surface inspection, visit www.qualitymag.com to read the following: “Noncontact Inspection: From Shadowgraph to Multisensor” “Noncontact Inspection Keeps Pace with Production” “Taking the Edge Out of Noncontact Measurement”
Published by QualityMagazine. View All Articles.