Testing…one…two…three…
By Jim Anderton
Repeatable, accurate testing and colour matching can be either a paint-by-numbers, or a sophisticated management tool. Either way, testing saves not only money but reputations as well....
Repeatable, accurate testing and colour matching can be either a paint-by-numbers, or a sophisticated management tool. Either way, testing saves not only money but reputations as well.
Ask any Quality Assurance (QA) manager or technician and they’ll agree: a micrometer is a beautiful thing.
It’s relatively cheap and when used correctly, yields results that are repeatable, accurate and, best of all, numeric. A measure of 12.25 centimeters (cm) is the same all over the world over — even in the U.S., which has resisted adopting the metric system.
Tensile strength, hoop stress, creep resistance and thermal stability — there are thousands of possible parameters to measure and tests to validate them. Each one has its place somewhere in plastics processing.
COLOUR IS SEXY
Breaking widgets in huge contraptions is the Hollywood idea of testing. But in the real world, sex sells. And nothing is as sexy in resin parts as colour.
Molders must ensure they’re molding parts with the colour their customer wants, and this is as crucial as any snap-fit tolerance or impact test.
Unlike physical testing, however, measuring colour with the uncalibrated human eyeball can create havoc in the plant and with the customer.
Instrumentation removes the subjectivity of human vision and makes colour a controllable property.
Colour, as a property of light, is worthy of an encyclopedia by itself. At its simplest, colour is about light and how the human eye makes sense of it.
Light from sources like the sun, fluorescent lamps or a flickering candle is reflected by a surface. The wavelengths of light enter the eye and are perceived as colour. That’s a huge simplification. The key concept, however, is that all three — ambient light, the surface being observed and the human eye — determine the outcome.
Everyone’s bought paint that looked great at the hardware store but looked terrible at home. That’s because the incident light has to be consistent for each measurement.
Because lighting is a major factor, the parts being shipped can appear a different colour on the customer’s loading dock.
The part’s surface is the second component, a variable that can be changed or controlled. The wild card is the human eye. Response to colour varies widely across individuals and it’s difficult to control for in a QA environment.
Enter technology. Modern, low-cost colourimeters and spectrophotometers allow molders and extruders to take formerly lab-only colour data readings right to the shop floor. These technologies allow for users to instantly verify colour and to ship parts that meet an objective, agreed upon standard with each shipment.
COLOURIMETER OR SPECTROPHOTOMETER?
However, each molder must determine the right device for their operation.
Colourimeters measure colour the way the human eye perceives it, said John McCasland, marketing services manager for the instrument services division of Konica Minolta Photo Imaging U.S.A. Inc. in Mahwah, N.J.
“(The colourimeter) uses a combination of three filters and detectors, red, blue and green,” McCasland added. “You take the colourimeter, put it on the piece of plastic and press the key. A light source illuminates the surface and it measures the light reflected from the surface.”
Colourimeters measure colour differences from a sample that’s good and a part that’s no good. These tools detect the difference with a resolution better than the trained human eye. It will tell you to what degree it’s off the standard, he noted.
“A spectrophotometer is a more sophisticated instrument. Instead of using filters, it measures colour wavelength by wavelength throughout the visible light range,” he explained. “Because it’s measuring by wavelength, you can measure colour in lighting other than standard daylight.”
A spectrophotometer, on the other hand, can be used for both colour matching and measuring colour difference, but costs about 1,000 than a colourimeter.
“If you need to measure colour under different light sources, you would use a spectrophotometer, whereas a colourimeter is good for colour difference under daylight conditions,” he explained. “The advantage of modern colour test instrumentation is the size and ease of use of modern units.”
“It operates as simply as a desk stapler,” declared Tim Mouw, lead application specialist for X-Rite Incorporated in Grandville, Mich. “I can lift it up and just lay the instrument on the surface and press it down like a stapler. Let go and you’re done.”
By that simple motion, users can take a measurement to use as a reference of a standard of the approved colour.
“The beauty of a portable is that you can take the instrument to the sample instead of the sample to the instrument. It can tell you right on the instrument screen whether the part passes or fails based on the tolerances you’ve set up in advance,” Mouw said.
With lab-quality instruments priced at less than $10,000 and some handheld models under $5,000, colour measuring equipment won’t bankrupt most processors. It isn’t, however, a trivial purchase. Yet many processors wonder why they should upgrade when their customers have been happy.
LOOKS ARE EVERYTHING
Molders should think of these tools as another instrument that sets QC between buyer and seller explained Jim Degroff, president of Color-Tec in Clinton, N.J. Even if an OEM has provided a colour sample, that doesn’t mean the colour of the molded part won’t change over time, even if it’s molded to specifications. That’s why agreeing on a sampling technique is crucial because the hard data taken by these tools won’t change,
“For example, you have to let the part cool to avoid ‘thermochromism’ (changes in colour with heat) and you document each lot, submitting the data as part of your certification,” he said. “Part of it is your sampling technique and part of it is the measuring technique. For the customer that bought too much and needs to send product back, he might say, ‘It’s the wrong colour.’ There, you have the numbers. Forget that you have two eyes and think about it as another gauge.”
Molders chasing OEM business, are likely communicating both up and down their supply chains as major customers demand tighter control and the paperwork that goes with it.
“The plastics industry continues to become more integrated with other supply chains from raw materials suppliers to large corporate accounts,” said Dr. Jan-Paul van Maaren, GretagMacbeth’s vice-president and general manager, Color & Appearance Business Unit in New Windsor, N.Y.
“(This is) driving demand for plastic components in automotive, consumer and business electronics, building/construction equipment and appliance industries,” he added. “Colour must be more tightly controlled not only across supply chains, but within the enterprise from product design and manufacturing to marketing and sales.”
Two variables are driving this integration. First is the consumer’s perception of quality. Consumers judge products based on how they look, and colour is a large part of that, he noted. Second, OEMs have greater need for speed, efficiency and more cost effective manufacturing.
PAPER TRAILS
Automating the paper trail that follows the test data is as important as the data-gathering technology itself for high-volume, low margin business that’s typical of OEM processing. Establishing a reliable digital workflow can drastically reduce or even eliminate colour errors. Plus, it helps suppliers meet aggressive market deadlines and cost goals, Dr. van Maaren added.
TOXIC SHOCK
Colour is the major attribute for many resin consumer goods, but it’s just one of hundreds of testable parameters ranging from aroma to radiation tolerance.
It may seem molding military aircraft parts that have to survive an atomic blast requires more testing than manufacturing teething rings.
But then again, how are plasticizer level
s in a baby’s mouth measured?
Testing properties like toxicity are an outsourced function for all but the biggest processors. However, many aspects of physical testing are available and affordable to small- and mid-sized processors.
For physical testing, the motivation for investing in equipment is similar to the case for colour matching: it’s either customer-driven, or a quality/productivity tool for internal use.
The equipment purchase equation is an easy one: cost to outsource versus cost of in-house testing. Inside the processor’s operation, however, the decision becomes a little more complicated.
Take nylon molders, for example, for whom moisture uptake is a real concern.
Moisture can kill resin properties and can be a serious issue to stable press operation.
A moisture meter capable of testing the resin pellets before they go into the hopper takes a crucial variable out of the loop before a crisis develops.
“No calibration is required,” said Kevin Butler, advertising manager for C.W. Brabender Instruments Inc. in South Hackensak, N.J. “[Moisture meters] will detect the water content down to 0.0005 per cent. [They are] quite affordable too. At US$10,000 to US$12,0000 [they] won’t break the bank.”
Butler noted test equipment capability has increased dramatically in the last decade, with performance-per-dollar rising noticeably.
“I think it’s due to the technology”, he stated. “Instruments aren’t as bulky and manufacturing costs have come down. So you get more capability and accuracy — it might be off only 0.25 per cent in the accuracy.”
Testing moisture content of the resin is an example of controlling a contaminant before it enters the barrel.
MELT CONTROL
For when the material is insider the barrel, there are other tests that can give a processor a good indication of material performance in the melt.
Capillary rheometers are traditionally used for measuring melt flow, but for many extrusion or injection operations, it’s performance against a known standard that counts, especially when ‘specing’ a resin from a compounder.
‘Melt flow index’ or ‘melt index’ is one specification. It can be defined as the number of grams of resin that can be squeezed through a 0.0825-in. hole by a force of 2.16 kg in 10 minutes at a temperature of 190C.
That’s a pretty esoteric parameter. By testing a combination that works in an application, it’s possible to hold compounders to a resin grade that works reliably with each shipment and to reject incoming material when it doesn’t.
Melt index testing compares to physical property instrumentation for finished parts as an initial purchase, said Keith Goble, product marketing manager, Polymer Testing Products Division at Dynisco Instruments in Franklin, Mass.
“I would say that a melt index tester would be more relevant to an injection molder than an impact tester,” Goble said. “Unless you’ve got an application on film-type products, I’d rather have a melt impact tester. From a price standpoint the melt index testers are actually some of the most affordable products. They give you a direct measurement of the flow characteristics. It’s crude by comparison to the capillary rheometer, but it is an industry standard and is used to measure the flow characteristics of polymers.”
Melt impact tests can compare how one polymer will behave relative to another, but the limitation is the range of shear rates available, he added. However, he said users can save costs by evaluating a material before processing it.
“I have found that a lot of users get a lot of useful information from them,” he noted.
However, depending on molder’s requirements, they do tend to outgrow what the melt index tester can provide, Goble added. Starting out, molders are likely more concerned with flow characteristics than physical properties.
“From a molder’s perspective, you should focus on spending the money on what goes into the machine as much as what comes out,” Goble said.
CUSTOMER IS NOT ALWAYS RIGHT
Testing the resin going into the barrel is useful, but giving the customer simply what they asked for, instead of what they wanted can waste time and money.
The scenario is painfully common: the customer designs a part, buys a mold and the part fails, despite the molder following the process and specifications to the letter. Can the molder substitute a different grade of resin to increase the part strength? Or how about quick mold modification that slows the cycle time?
When the processor ends up part of the customer’s product development and QA team, it may be necessary to break some parts. Classic tensile test instrumentation, for example, is now computerized to the point where an operator can generate good data without an engineering degree.
“If someone like a lab manager is setting up the methods and making sure everything is set up properly, the day-to-day operations and printing of reports don’t require a lot of training,” said Lorenzo Majno, business development manager for Instron Corp., in Norwood, Mass.
“With processing technology being what it is now, there are a lot of places where you can place traps to catch it before someone makes a mistake,” he added. “Where’s your time spent? In the real world, that’s something they should ask: how much time and effort is spent calibrating versus actually running the test?”
While the equipment has become very easy to use, the need to calibrate is still crucial for reliable, repeatable results. “It’s always struck me how people assume that the machine is calibrated,” Majno added. “It’s not necessarily the case.”
Calibration and verification is a very simple service provided by Instron as well as lot of other firms that will visit the customer and conduct the verification for them. Although, it’s not time-consuming, it’s necessary to check that the instrumentation has been calibrated and verified.
“Even simply picking it up and putting it down somewhere else in the lab means it needs to be re-verified,” Majno explained.
In the end, however, testing is about money. Testing equipment either makes a processor money by reducing the variability of incoming material and by validating a process, or it saves money by clarifying what the performance expectations are for incoming resin, the production process and in the finished parts.
Customers may demand testing capability, but properly executed, many processors will wonder how they managed before.