© 2010 Quality Digest Magazine. All Rights Reserved.
Β
Responding to the needs of automakers and their suppliers, manufacturers of instruments that measure color are introducing new technologies to help control the quality of metallic flake, pearlescent, and other special-effect paints that have confounded optical instruments throughout the past 60 years.
New measurement techniques address the thorny problem that has troubled manufacturers since effect paints were introduced during the 1940s: how to get reliable and repeatable measurements of effect coatings that sparkle. The need to accurately quantify the hue shift of effect paints has become more critical as automakers assemble body panels, bumper fascia, and other parts made by several suppliers, each part coated with effect paints that must match under different illumination and observation angles.
Formerly the purview of a small band of car hobbyists who applied their own custom paint jobs, effect paints have become popular with the public and are now a central selling point for all major automotive manufacturers. The most popular automotive color worldwide for the past seven years has been silver with effect variations, according to DuPontβs 2007 βGlobal Automotive Color Popularity Report.β Even earth-toned, neutral colors offered by automakers contain some additives to yield special-effect appearances. Another major paint manufacturer with global operations, PPG Industries, headquartered in Pittsburgh, Pennsylvania, is focusing this year on how effect pigments can create innovations in the color and appearance of paints.
This burgeoning demand for effect paints has resulted in headaches for quality control personnel. Technicians on the production line may observe that a body panel and bumper fascia donβt match properly, but they canβt back up their human perception with data indicating why the mismatch is occurring. Consequently, manufacturers have spent an inordinate amount of time and resources trying to determine the root cause of painting problems through trial-and-error methods.
Trends indicate that, without proper instrumentation, problems with effect paints will only increase in magnitude. The quality control documentation required for a single change in automotive body color is considerable for all companies in a supply chain, and the task becomes enormous when one considers the number of color offerings in a model year. For example, PPG reports that it has developed 105 exterior-color and more than 25 interior-color concepts for automakersβ consideration for the 2010-2011 model years. Many of these colors contain effect pigments.
Seeing in three dimensions
Until recently, color-instrument manufacturers had only 2-D technology in their arsenals to measure the essentially 3-D problem of quantifying the sparkle of effect paints. But instrument manufacturers are now introducing new technology that captures data in three-axis form.
Comparing the versatility of the 3-D technology over 2-D is like comparing the data a technician can derive from a digital micrometer with a go/no-go gauge. The gauge can reliably tell if a part meets a single specification, but itβs severely limited in indicating when and why a process is drifting out of control. On the other hand, a digital micrometer can yield a great deal of information thatβs useful in determining the state of a process.
Similarly, a 3-D paint profile has applications far beyond simply quantifying if a part meets a color specification. By translating the spectrum of reflections from effect paints in three-axis form, new technologies can:
β’ Indicate whether a problem lies with process or formulation
β’ Help predict the practical limits to which a process or formulation can be adjusted to conform to a visual standard
β’ Provide a unique profile of an effect paint or surface that all companies in the supply chain can accept, which saves time and money when implementing a new program where several companies must produce parts perfectly matching in color and appearance
β’ Automate labor-intensive processes in producing test plates that serve as standards to manufacturing plants around the world
X-Rite Inc., headquartered in Grand Rapids, Michigan, a leading designer and manufacturer of color-measurement instruments and software, used the new 3-D technology to solve in two days a problem involving matching parts coated with effect paints that had troubled a major automaker for more than two months. Because the technology produces a unique profile for each effect paint--similar to how every person has a unique DNA structure--X-Rite coined the term βxDNAβ to describe its version of the 3-D technology. It combines the new MA98 instrument used to gather data and the software for analysis.
Theory behind technology
For many problems in science, finding the answer requires that the question be framed properly and broken down into components. The same proved true for unlocking the properties of color and appearance in effect paints. Engineers had good reason to believe that they could build a unique footprint of an effect paint and process, thanks to an inviolate law of science: conservation of energy. The law states that energy canβt be created or destroyed.
With the conservation-of-energy law in mind, engineers then applied a second law regarding light: When light falls on an object, itβs always reflected, refracted (bent), scattered, or absorbed.
By knowing the amount of light energy falling on the test surface, engineers had one side of the conservation-of-energy equation. They then used sensors arrayed at specific angles to measure the intensity and spectrum of the light emanating from the test surface and split that single energy value into the categories of reflected, refracted, and scattered light. From this, they could build a unique profile of the paint and how it was applied to the test surface.
The new technology probes beneath the surface of the paint to yield clues to the paintβs chemical composition and application process. Oddly enough, even materials as reflective as a mirror or absorptive as black carbon soot have some refractive and scattering properties. Also, materials do not refract all colors of the visible spectrum to the same degree. For instance, a certain material has a characteristic tendency to bend blue light at the lower end of the visible spectrum differently than red light at the higher end.
The fact that light penetrates below the surface of the material while itβs reflected, refracted, scattered, or absorbed is one key to unlocking effect-paint formulation and processing. However, modern coatings are a complex interaction of mixtures, layering, and application processes, and engineers also needed to make an assumption based on effective-medium theory to help them understand this interaction. Effective-medium theory states that a complex coating with many ingredients can be treated as a single homogeneous material when itβs measured from the test surface.
Challenges of effect paints
Although they can be treated as a simple material under the theory, effect paints are actually concoctions of tiny metal or mineral plates, additives, pigments, binders, and solvents custom-designed for a specific application. For example, a coating with three different layers and nine ingredients can be treated as if it were a single material by considering how the components are affected by their averages weighted by their distribution throughout the layers, the layersβ thickness, and the structure of the boundaries between adjacent layers.
With theory backing up the general approach, engineers took two steps to solve the problem of correlating human perception of effect paints with solid data: First, design an instrument that could βseeβ the differences in two samples better than the best-trained inspector, and second, develop a software program that could analyze and interpret these data in an easy-to-understand format.
The first step required adding sensors that would gather information from a third axis of light emanating from the test surface. Fortunately, the new technology could build on tried-and-true photodiodes, illumination sources, color-analyzer engines, and other optical components to handle the job. Unfortunately, adding just one additional axis meant that the instruments needed twice as many sensors and illuminators to gather the necessary data.
One version of the new instruments uses two illuminators and 11 sensors that measure 31 bands of the visible spectrum--from blue, representing the shortest waves, in the 400-nanometer range, to red, representing the longest waves, at the 700-nanometer range. The illumination sources are gas-filled tungsten lamps color-corrected to approximately 4,000 K, which produce results that can be translated into illumination C, D65, D50, A, F2, F7, F11, and F12 standards.
The two illuminators flash intense white light at 15Β° and 45Β° angles to the nominal of the test surface. Light emanating from the test surface is collected at 10 angles: -15Β°, 15Β°, 25Β°, 45Β°, 75Β°, 110Β°, 25Β°az90, 25Β°az-90, 60Β°az125.3, and 60Β°az-125.3. (The βazβ notation refers to the azimuthal rotation from spectral reference.)
The light collected at each of these angles is transmitted by fiber optics to the color analyzer, essentially a rugged stepper motor that precisely spins a wheel with 31 interference filters on its circumference. Interference filters are films that transmit only a certain wavelength of white light.
The 10 photodiodes on the opposite side of the filter wheel--one for each angle--measure the intensity of the sample light as it passes through the interference filters. For instance, the photodiode responsible for the 15Β° angle to nominal will measure 31 different data points when the instrument is activated. Each data point represents a wavelength of the visible spectrum. The elegance of this approach is that a single rotation of the wheel generates 31 points of a spectral curve from all 11 angles in less than 1 second.
One measurement of the test surface yields 310 data points in three-axis form--which amounts to a 3-D profile of the effect paint and how it was applied to the part. All of this happens in 1 second, and technicians can read measurements off the instrumentβs display in another second for a total measurement, calculation, and display time of approximately 2 seconds.
To make measurements both repeatable and reproducible, the instruments are equipped with three external pressure sensors around the viewing port that signal to the operator when he or she has applied sufficient force of the instrument against the test surface for measurements. Research has shown that the repeatability of an instrument is under 0.02 delta Eab on a white calibration plaque; 1 delta Eab corresponds to the point in color space where a trained human observer can detect a subtle difference in shade. Reproducibility, or inter-instrument agreement, is less than 0.18 delta Eab average on a reference Series II BCRA tile set.
Analyzing the data
All the data used to build the 3-D model of paint and process is useless without an interpretation method, so engineers and scientists had to develop sophisticated algorithms that could plot and analyze the data to provide useful information.
By itself, the 3-D portrait isnβt useful. The power of the new technology becomes apparent when one 3-D profile is compared with another. In that way, companies can do the following:
β’ Troubleshoot whether a problem on the shop floor is due to the manufacturing process or the paint formulation.
β’ Assess whether existing equipment can be adjusted enough to accommodate a new process.
β’ Develop more exact quality standards on the painting lines that indicate quickly when a process is going out of control.
β’ Predict whether a person will be able to perceive a difference in color and appearance when the formula of an effect paint, or the process used to apply the paint, are changed.
Similar to the way that a person can easily see differences between two photographs by placing and aligning one negative over the other, computer software mathematically superimposes one 3-D plot over another.
The differences in size, shape, and rotation of the two profiles provide clues about whether a process is out of adjustment from a standard, or if there are differences in reflective flake size and orientation between two parts from different companies.
Because the new generation of instruments can discern hue-shifting paints to a much finer degree than humans can, the question for quality control engineers and technicians isnβt whether differences exist but rather, will anyone be able to perceive the differences?
The new technology can act as a reliable predictor as to whether a person will be able to perceive a difference in two samples of effect paints--and why the samples look different.
This means that the new method can:
β’ Assist engineers in predicting whether subtle changes in a process or formulation will greatly affect the color and appearance of products
β’ Help technicians and quality personnel monitor processes that have subtle differences in process or formulation
Similar to 1 delta Eabβs definition as the distance in color space at which a human can typically begin to perceive a difference between a color and a slight variation of it, the new technology has a delta measurement of 1 unit that predicts when a person may be able to distinguish differences in two mating parts.
This type of comparison is critical for setting process and part tolerances. After an automaker selects a paint color for its sales catalogs, every company in the supply chain must spend time extensively documenting the production part approval process (PPAP) and the advanced product quality planning (APQP) approvals on the colorβs process and formulation. Using the new 3-D technology, companies can save a significant amount of time and money in this arena alone, and the applications go far beyond time and money saved in documentation.
A case study
A major European automaker experienced difficulties on its production line when body panels painted at the factory using a popular silver metallic color suddenly didnβt appear to match the bumper fascia supplied by an outside vendor. Although occasional small problems had occurred before that could be assigned to definite root causes, this time the differences in color and appearance where the body panels appeared darker than the bumper fascia were readily noticeable.
Process engineers at the automaker applied traditional root cause analysis to the problem, trying first to identify whether the matching problem was due to an internal process, changes in formulations by the paint supplier, or parts that were out of specification from the bumper supplier. Individuals working on the line told the engineers of a sudden difference in how the parts matched, and their floor inspectors, using an older generation instrument, indicated that there was a difference in its reflectance values. With these available data, the engineers reasoned that the root cause of the problem was due to paint formulation, and they asked for assistance from the paint supplier.
The paint manufacturer spent nearly two months unsuccessfully trying to match the color and appearance of the bumper fascias that appeared too light in comparison with the car body--even going as far as changing the formulation of the paint.
The new technology was used to develop 3-D plots of both the bumper fascia and the body panel, and an analysis indicated that although the plots showed similar shapes, their different positions in 3-D space highlighted a change in process rather than formulation.
This analysis provided another important clue: The difference in reflectance curves at the 25Β° angle out-of-plane measurements indicated a difference in the way that the flakes were oriented. The data all pointed to a mismatch or harmony problem due to processing differences between the bumper fascia and the car body.
The automaker then realized that it had altered a critical process parameter in prior months when it changed from a bell/air method of applying the base and final coats to body panels, to an entirely bell/bell method. The type of silver paint it was applying contained relatively low pigmentation and a relatively high reflective value due to the higher percentage of flakes in the paint medium. With the bell/air method, the paint flakes essentially had an opportunity to float in the paint medium and come to rest horizontally on the body panel in a random distribution. That orientation maximizes the reflective characteristic of the flakes, giving the paint a lighter appearance.
Prior to the process change, the color and appearance of the body panels and bumper fascia were in harmony because the bumper supplier used an air/air process that allowed the flakes to align themselves in a way similar to the automakerβs painting method. When the automaker implemented the bell/bell method, the flakes hit the body panel so that they were oriented on their edges uniformly, cutting their reflective quality and giving the paint a darker appearance.
Based on the analysis, the automaker adjusted the paint-supply speed in the bell/bell system and coordinated with the bumper supplier to achieve harmonious color matching in body panels and bumper fascia.
Thereβs no doubt that automakers and automotive suppliers will be the pioneers in employing this new technology, but 3-D analysis of surfaces with sparkle has potential applications in other industries, including appliance manufacturing and cosmetics formulation.