Examining for and eliminating particulate contamination is a vital step in the manufacturing process. Material contamination by micron-range-size particles has become a focus of modern quality assurance. Initially, it was the pharmaceutical and semiconductor industries that were occupied with particulate contamination in this range. However, as the functional density of parts and components has increased and expectations for production sequences and quality have grown, this topic also has caught the eye of the automotive and aerospace industries, along with their suppliers. Fierce global competition, along with increasingly rigorous national, international (such as the International Organization for Standardization, or ISO) and company standards, have helped make examining for and eliminating particulate contamination a vital step in the manufacturing process. In principle, nearly every production process and every manufactured product has specified cleanliness requirements. Solid particles in oils or fuels can damage motor components or hydraulic and lubrication systems. The production of precision parts and products in fine mechanical and micromechanical ranges is governed by stringent requirements regarding permissible levels of residue contamination. Such systems are typically used in sensitive areas of the medical device, aerospace and automotive industries. Damage caused by residual particulates can have catastrophic consequences, and in extreme cases, could even lead to total system breakdown. Filter residue analysis is an essential part of analyzing cleanliness for liquids and high-precision parts. Detecting, visualizing, identifying and reporting on particles found in automotive and aerospace inspection fi lters help engineers ascertain contamination present in lubricants, hydraulic fl uids, fuels and diesel emissions, and also is useful for monitoring cleanliness in engine blocks, transmissions, camshafts, crankshafts and other parts, which in turn affects the effi ciency and durability of fi nal products. MICROSCOPIC RESIDUE ANALYSIS Microscope-based residue analysis helps determine the size of particulate contamination down to the micron level. Particles are quantifi ed and then classifi ed in compliance with relevant standards. The analysis usually includes three steps: The object 1. Is cleansed with a rinsing fl uid in compliance with a predefi ned standard. 2. The rinsing fl uid is fi ltered and any retained particulates are visible atop the fi lter surface. Is placed underneath an optical microscope for automatic particle quantifi cation and analysis. Turnkey automatic fi lter inspection solutions consist of fi ve critical components: the optical microscope; a motorized XYZ stage and controller; a high-speed digital camera; a computer with dedicated image-analysis software for fi lter inspection and particle counting; and the stage insert plate and fi lter holder, to ensure fi lter planarity and minimize any potential focusing issues. Manufacturers typically require many or all of the following when working with automated fi lter inspection systems: • Adherence to international, national and/or company standards. • The ability to accurately count and reconstruct images of particles that exceed the size of individual image frames. Such particles must not be truncated or counted repeatedly. • Automatic focusing at each image frame to compensate for variations in fi lter height or tilt. • Data analysis fl exibility so technicians can revise, modify or reclassify results. • Accurate discrimination of refl ecting particles vs. non-refl ecting ones, plus fi bers vs. particles. • Automatic archiving of relevant data and generation of clear, useful Reports based on predefi ned company templates. Speed, e • ffi ciency and intuitive operation. HOW AUTOMATED SYSTEMS WORK In most modern microscope-based image analysis systems, system settings are defi ned once upon initial system confi guration by an administrator or advanced operator; these are typically password protected to ensure tamper resistance. Such presets may include the area defi nition of the fi lter to be scanned, software training for particle detection, and thresholds for image analysis, database fi elds and report template information. After such parameters have been defi ned, the operator secures the fi lter into the integrated holder and starts the scanning process. The motorized microscope scanning stage automatically positions the fi lter underneath the microscope so that images are acquired at every frame over the entire area of the fi lter. To maximize effi ciency, particles are typically detected on-the-fl y. Should large particles or fi bers exceed a single image frame, they are “stitched” together and reconstructed automatically so they are counted only once. Although the use of a fi lter holder can minimize variations in sample fl atness, some systems also offer an integrated tilt-compensation, or “predictive focus” algorithm, which extrapolates the predefi ned tilt, either linearly or non-linearly, during the scan. In addition, at high magnifi cations where the microscope depth of fi eld is minimal, image analysis solutions may offer additional options to ensure proper focus. Some of the more sophisticated solutions offer a highly accurate laserbased autofocus mechanism to achieve focus quickly and precisely at each frame. Using an automated microscope fi lter inspection system, a complete scan of a 47 millimeter fi lter can be accomplished in as little as three minutes. After the scan is complete, spreadsheets, histograms and reports can be generated automatically based on predefi ned standards and templates. The operator can address any discrepancies using a built-in feature that allows him or her to revise fi ndings interactively. The software also may archive results into a database automatically, eliminating the need for the operator to save results manually. Further, operators can perform simple or complex queries on all relevant data. CHOOSE THE CORRECT MICROSCOPE Among myriad microscope confi gurations available, one parameter is most critical: the minimum particle size that can be detected. In addition, there are specifi c guidelines that must be adhered to regarding industry, government, professional society and corporate standards. For instance, ISO 16232 (126.96.36.199.2) says: The magnifi cation, resolution and depth of fi eld are selected according to the optical lens. The key parameter for a precise measurement is the optical resolution (and not the magnifi cation) of the selected lens. This value is dependent on the light wavelength and the numerical aperture of the lens. The optical resolution of the selected lens should be equal to or less than 1/10 of the smallest particle. In the case of particles below 20 µm, the standard accepts that the resolution is less than or equal to 1/5 of the smallest particle size. To calculate the correct objective lens to use, here is a simple example. Let’s say particles as small as 5 micrometers (µm) must be detected in compliance with ISO 16232. The Rayleigh criterion formula can be used to calculate resolution (Resolution = .61 * / Naobj) Using a typical light wavelength of 550 nm, the resolution of one semi-apochromatic objective lens, for instance, is 0.75 µm. Since ISO 16232 states that the objective lens must resolve 1/5 of the smallest particle, or 1/5 of 5 µm (1 µm), this lens is acceptable. Some operators whose particles are 25 µm in size or greater can select a highresolution stereo microscope for fi lter inspection, since these instruments usually offer a larger fi eld of view and depth of fi eld, and typically cost less than upright microscopes. It is important to note that stereomicroscopes must have a repeatable magnifi cation detent to assure repeatable results. To distinguish between refl ecting and nonrefl ecting particles, polarized illum Nation may be required. Typically, one scan is performed with the polarizer in the open position, and a second scan is performed with the polarizers perpendicular to each other, or cross-pol. When the process requires automatic scanning of multiple fi lters simultaneously, a microscope with an extended throatdepth, such as a dedicated semiconductor microscope, may be considered. SELECTING THE CAMERA Which camera is best for automatic residue analysis? For highly accurate and fast residue analysis, key questions to consider include: What digital re • solution is required? • Is color or gray-scale imaging an appropriate choice? • What refresh rate (speed) can be achieved? For simple quantifi cation, a grayscale camera may be the appropriate choice. Because color information is usually of no value when counting, measuring and classifying particles, the absence of the Bayer mask—used in most color cameras—will decrease the exposure and overall scan times while minimizing distortion. However, if the process requires segregating particles of varying colors, a color camera is required. In addressing digital camera resolution, ISO 16232 states: The pixel resolution (µm / pixel) shall be = 1/10 of the smallest particle size to be measured, or = 1/5 for smaller particles (defi ned as below 20 µm). Using the above example of detecting 5 µm particles using a 20X objective lens, we can extend the example to calculate the calibrated camera digital pixel resolution. Take for instance, a high-speed monochrome or color camera with a pixel size of 6.45 µm. If the operator uses a 1.0X camera adapter, both cameras’ digital pixel resolution is 0.32 µm (6.45 µm ÷ 20). Since pixel resolution falls beneath the benchmark requirement of 1/5 of the smallest particle (1/5 of 5 µm, or 1 µm), both of these cameras meet the ISO 16232 requirement when using a 20X objective lens. As contamination has become a focus of modern quality assurance, many manufacturers have evolved from manually counting particles to the adaptation of a fully automated turnkey microscope-based system. With various solutions available, operators should determine if their system will produce results in compliance with appropriate international, national or company standards. In addition, systems should provide speed, fl exibility and repeatable data to meet individual process and operational requirements. Q Mario J. Gislao is industrial marketing and applications specialist for Olympus America Inc. (Center Valley, PA). For more information, call (914) 573-9739, e-mail Mario.Gislao@ olympus.com or visit www.olympusamerica.com/ seg_industrial/segi_home.asp.
Published by QualityMagazine. View All Articles.