15 -1 15 Design of Experiments for Coatings 15.1 Introduction 15- 1 15.2 Standard Two-Level Factorial Designs 15- 2 Case study — Screening Factors thought to Affect a Spin Coater 15.3 Optimization via Response Surface Methods (RSMs) 15- 4 15.4 Mixture Designs for Optimal Formulation 15- 5 References 15- 6 15.1 Introduction The traditional approach to experimentation changes only one process factor at a time (OFAT) or one component in a formulation. However, the OFAT approach does not provide data on interactions of factors (or components), a likely occurrence with coating formulations and processes. Statistically-based design of experiments (DOE) provides validated models, including any significant interactions, that allow you to confidently predict response measures as a function of the inputs. The payoff is the identification of “sweet spots,” where you can achieve all product specifications and processing objectives. The strategy of DOE is simple and straightforward: 1. Use screening designs to separate the vital few factors (or components) from the trivial many. 2. Follow up by doing an in-depth investigation of the surviving factors. Generate a “response surface” map and move the process or product to the optimum location. However, the designs must be tailored for the nature of the variables studied: •Components in a product formulation •Factors affecting a process Tr aditionally, the experiments on formulations versus processing are done separately by chemists and engineers, respectively. Obviously, collaboration between these two technical professions is essential to the success of any study. Furthermore, mixture components can be combined with process factors into one design for final optimization. In other words, you can mix your cake and bake it too, but this should be done only at the final stages of development — after narrowing the field of variables to the vital few. We will devote most of this short discussion to process screening, because these designs are relatively simple, yet are incredibly powerful for making breakthrough improvement. Mastering this level of DOE puts you far ahead of most technical professionals and paves the way for more advanced tools geared to optimization of processes or formulated products. Mark J. Anderson Stat-Ease, Inc. Patrick J. Whitcomb Stat-Ease, Inc. DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC 16 -1 16 Top 10 Reasons Not to Base Service Life Predictions upon Accelerated Lab Light Stability Tests 16.1 Light Spectra 16- 1 16.2 Light Intensity 16- 4 16.3 Temperature Sensitivity of Materials 16- 4 16.4 Gas (Ozone) Fading 16- 5 16.5 Catalytic Fading 16- 5 16.6 Lux versus UV 16- 5 16.7 Light Stability Testing Standards 16- 5 16.8 Conclusion 16- 6 References 16- 6 The popularity of personal computers and digital cameras has ushered in an exploding new market of digital images printed from consumer printers. There is an endless combination of inkjet inks and commercial photo papers currently available in the marketplace. However, no one is really sure how long these printed images will remain lightfast. Image permanence is a big issue. Many OEM computer printer manufacturers, inkjet ink, and paper suppliers are rushing to develop a standardized light stability test protocol that will generate meaningful test data. But, this is inherently complex. There are a myriad of factors that can cause degradation of image quality besides ultraviolet (UV) light: ozone (or gas) fade, catalytic fading, humidity, dark stability, and temperature. Together or individually, each can wreak havoc on a treasured image. Following is a review of the major issues related to light stability testing of inks and substrates. 16.1 Light Spectra It must first be stated that there is no standard light spectrum to replicate indoor lighting conditions. However, a recent Kodak study concluded that indirect window-filtered daylight is the dominant indoor lighting condition in homes. 1 Let us review some widely used laboratory light sources for light stability testing of printed images. Eric T. Everett Q-Panel Lab Products DK4036_C016.fm Page 1 Thursday, May 12, 2005 9:40 AM © 2006 by Taylor & Francis Group, LLC Fluorescent Lamps • Xenon Arc Lamps Standard Temperature • Humidity • Dark Stability • Linearity of Degradation • Reciprocity Failure 17 -1 17 Under What Regulation? 17.1 Introduction 17- 1 17.2 Code of Federal Regulations 17- 1 17.3 Title 29 (Labor) 17- 1 17.4 Protection 17- 2 17.5 Biocides 17- 3 17.6 Testing 17- 3 17.7 Volatile Organic Substances (VOCs) 17- 3 17.8 Food and Drug Administration (FDA) 17- 3 17.9 Which Regulation? 17- 4 17.1 Introduction Civilization is based on laws and regulations for the common good. Way back, the law was as simple as, “don’t kill each other.” As time passed and technology grew, the laws and regulations became more complex to keep up with the technology. Regulations were and are issued by the federal government, the state government, the county government, and the local government. With everyone enacting regulations, confusion abounds. There are just too many regulations to compile in a single list. The following looks at some of the major regulations governing the coatings and inks industries. 17.2 Code of Federal Regulations The government of the United States of America discusses possible regulations and publishes the discus- sion in the Federal Register on a daily basis. When it is concluded that a regulation, or change in regulation, is needed, it will be published in the Code of Federal Regulations (CFR). This CFR is an enormous work comprising some 50 titles, each subdivided into a number of books. 17.3 Title 29 (Labor) In Title 29 (Labor) is section XVII (Occupational Safety and Health Administration, Department of Labor). Under this section is part 1910.1200, Material Safety Data Sheet (MSDS). The MSDS started by proposing safety considerations for asbestos during the process of cutting ships apart. It has progressed to cover almost all chemicals used in industry and commerce. The MSDS lists the manufacturer respon- sible for the product, the composition, and safety and health concerns. This form has gone, in practice, from seven sections to 16 or more. An example listing of the section titles is as follows: Arthur A. Tracton Consultant DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC II -1 II Coating and Processing Techniques DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC 18 -1 18 Wire-Wound Rod Coating 18.1 Introduction 18- 1 18.2 History 18- 2 18.3 Theory and Principle 18- 3 18.4 18.5 The Rod Coating Station 18- 4 Rod Station Variations 18.6 Advantages and Disadvantages 18- 7 18.1 Introduction Wire-wound metering rods have been used for more than 75 years to apply liquids evenly to flexible materials. They were the first tools used to control coating thickness across the full width of a moving web. The 1980s saw a new popularity in rod use because of improved quality and the industry trend toward shorter converting runs. Wire-wound rods are used in a wide range of applications but find their greatest appeal in the manufacture of tapes, labels, office products, and flexible packaging. The first rods were made of ordinary carbon steel, wrapped with music wire. Today’s metering rods use precision- ground core rods made of stainless steel, tightly wound with polished stainless steel wire at high speeds, on custom-designed winding machines. The resulting product is a laboratory-quality precision tool that can control coating thicknesses accurately within 0.0001 in. (0.1 mil). A typical wire-wound rod station Also called applicator rods, Mayer bars, equalizer bars, coating rods, and doctor rods, this equipment has found uses in a wide variety of production applications, from the manufacture of optical films to wallboard panels. Wire-wound metering rods look deceptively simple. A stainless steel rod is wound with a tight spiral of wire, also made of stainless steel. The wire can be so small that it is almost invisible to the naked eye, or so large that the windings look like the coils of a hefty spring. Today, the industry has standardized on stainless steel rods because they can be used with almost every coating liquid. Earlier problems with rust and corrosion have been virtually eliminated. Where abrasive wear is a problem, some converters use chrome plating to prolong the life of the rod because of the hard surface presented by chromium. Chrome has its drawbacks, however, as it builds up unevenly at the extreme tops of the wires, changing the shape of the wires and the resulting coating thickness. Also, if not applied properly, chrome can acquire pitting marks or can flake off, contaminating the coating bath or causing uneven coating. Several new products introduced since 1985 have further expanded the market for rods. Where streaking or rod cleaning is a problem, rods with a Teflon surface are available. Particles that might wedge between stainless steel wires tend to slide through, preventing buildups and subsequent streaks in the Donald M. MacLeod Industry Tech DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC Low Cost • Precise Coat Weights • Lower Setup Cost • Less Edge Film Thickness 18-4 Wear • Limitations is shown in Figure 18.1. Wire-Wound Rod Coating 18 -5 The web can be immersed directly into a tank (Figure 18.4); or an applicator can be rotated in the reservoir to transfer the liquid to the web at the top of its rotation (Figure 18.5). It is important to apply an excess of coating liquid at this station, to let the metering rod do its job. When an applicator roller is used in a rod coating system, the speed of the applicator is not a critical factor. In addition, the machine operator can adjust the applicator roller speed within a side range, even while the machine is running. The web passes over the metering rod, which may be stationary or may be rotated slowly. The rotation may be either in the same direction as the web or in the opposite direction. The choice of stationary rod depends on movement with different coaters and with different products. Establishing the ideal speed of rotation will also be different from job to job, and converters experiment to find the best procedure for each run. The most common procedure, however, is to rotate the rod slowly in the opposite direction to the movement of the web. The rotation flushes the coating material between the wires, keeping the wire surfaces wet, and preventing setting up and hardening of some liquids. The rotation also distributes any abrasive wear evenly on the wires and prevents flat spots from forming. The purpose of the metering rod is to remove excess coating liquid, allowing a measured amount to pass between the wire windings. The web should pass above the rod, to allow the excess liquid to fall back into the tank. The web, however, need not be perfectly horizontal, as long as the surplus coating can return to the tank through gravity. Metering rods for production coating can be made in a wide variety of sizes. The most common core rod diameters are quite small (3/16 and 1/4 in.), although sizes up to 1 in. diameter are also used. The main advantages of small-diameter rods are their low cost and ease of storing and handling. These thin rods must be supported in the coating machine, because they are not rigid and will deflect with pressure from the web. There are several types of rod holder in common use; the simplest is a square of a rectangular steel bar, with a “V” groove machined into it. This rod holder is mounted between the side frames of the coating machine, and the metering rods are placed in the grooves. The “V” groove FIGURE 18.4 Web immersion method. FIGURE 18.5 Applicator roller method. DK4036_book.fm Page 5 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC 18 -6 Coatings Technology Handbook, Third Edition should be ground and polished to minimize wear on the rod and should be mounted accurately, at a right angle to the direction of web travel and parallel to the idler rollers of the machine (Figure 18.6). The design of a rod coating station should ensure that the web makes intimate contact with the wires of the metering rod. The wrap angle, the angle between the web direction as it approaches the rod, and its direction as it leaves the rod, should be 15 ° for a heavy web tension or up to 25 ° for a light web tension (Figure 18.7). Web tension is a critical factor in the design of a rod coating station. With a wrap angle of 15 to 25 ° , the web must be tight enough to ensure intimate contact with the metering rod, yet not so tight that the web is deformed by the wires. Adhesives and some other liquids can solidify between the wire windings of the rod whenever the coater is stopped. Many coating machines have a “throw-off” feature, a mechan- ical means of separating the web from the rod automatically, whenever the machine is turned off. This allows for quick removal of the rod for flushing and cleaning before the coating material has a chance to congeal between the wires (Figure 18.8). This automatic releasing feature also simplifies rod changing between production runs. One method used by coating machine manufacturers is a rocker arm throw-off system. A series of idler rollers presses the web against the metering rod while the coating machine is running. Whenever the coater is stopped, the idlers automatically rise, lifting the web up, away from the metering rod. At the same time, a water flushing system can be triggered to remove coating material from the rods before it can set up. Other techniques involve lowering the metering rod and its holder when the coater is turned off. FIGURE 18.6 Rod holder. FIGURE 18.7 Wrap angle. FIGURE 18.8 Automatic throw-off. DK4036_book.fm Page 6 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC 19 -1 19 Slot Die Coating for Low Viscosity Fluids 19.1 Introduction 19- 1 19.2 Manifold Theory and Design 19- 1 19.3 Air Entrapment 19- 4 19.4 Lip Design 19- 4 19.5 Die Adjustment as It Relates to Manifold Design 19- 6 19.6 Coat Weight Adjustment 19- 6 19.7 Adhesive Selection 19- 6 19.8 Die Steel and Piping Selection 19- 6 19.9 Proximity versus Contact Coating 19- 6 19.10 Die Positioning 19- 8 19.11 Backup Roll Design 19- 12 19.12 Automatic Control 19- 12 19.13 Deckling 19- 14 Air Entrapment behind Deckling 19.14 Die Cleanup 19- 14 19.1 Introduction Slot head coating has spawned a wide range of designs, some quite radical in their concept. This chapter discusses conservative manufacturing experience along with the experience of a wide variety of processors currently utilizing the proximity or wipe-on method. 19.2 Manifold Theory and Design The primary purpose of a die is to define a width and provide an even coating in terms of cross-sectional thickness and smoothing. The manifold and coat-hanger section of the die is the main component in accomplishing uniform distribution. Smoothing will be addressed in a later section. There are two basic styles of manifold design in use today: coat-hanger shaped, with a volumetrically Harry G. Lippert Extrusion Dies, Inc. DK4036_book.fm Page 1 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC Lip Adjustment Design • Lip Wiping Face Design Die Control • Die-to-Roll Position Adjustment System Slot Head-to-Roll Position and Angle of Contact • Lip Opening Setup • Die-to-Roll Gap Setup Profiling • Die Support Design and Operation • Support and Positioning • Angle of Attack Position Adjustment • Lip Adjustment System Design Specifications • Die-to-Roll reducing cross section (Figure 19.1), and T-shaped, with a constant cross section (Figure 19.2). Slot Die Coating for Low Viscosity Fluids 19 -3 In either style, flow through the manifold is analogous to flow through a pipe in that there is an increasing resistance to material flow as the length increases. The wider the die (the longer the pipe), the greater the resistance to flow. It follows then that the primary criterion in a good die design is to ensure adequate flow to the ends of the die as the width requirements increase. The coat-hanger-style die utilizes a slot section (preland) with a varying length downstream of the preland section must decrease at the same rate it increases in the manifold section. If the sum of these two components is equal at any point in the overall flow stream, the result is an even flow. It can be seen that while the generic coat-hanger-style design is fixed, the overall dimensions may vary greatly, depending on a given die width, flow rate, or general coating material requirements. Generally, as the die gets wider, the length of the preland section ( B 1 ) must get longer and the manifold larger; as the flow rate increases, the manifold must get larger as well as the height of the gap at B 1 . The compensating preland section downstream of the manifold allows the die design to be varied greatly to suit a given application. These large internal designs are used for applications characterized by coating materials that vary greatly in viscosity levels or call for an extreme range of flow rates. Larger flow channels are less sensitive to rate and viscosity changes than are small channels. Small internal designs are used for materials that require a low residence time in the die because of thermal degradation, or high shear rates to prevent gelation (thixotropic materials). In analyzing the coat-hanger manifold, it must be emphasized that the manifold decreases in cross section as it approaches the ends of the die (dimensions A in Figure 19.1); this rate of reduction may also be changed to suit a specific application. Because material is flowing out of the front of the die along its entire width, less material is presented to the manifold as it approaches the ends of the die. The reduction in manifold volume is an attempt to keep the velocity of the material at the ends of the die to a maximum, to compensate for the lower flow rate, and to prevent carbonization of the resin or changes in viscosity in a thixotropic or dilatant adhesive. In summary, it can be seen that the coat-hanger manifold design can be modified to suit an application and still accomplish the primary criterion of even flow distribution. To adequately design a coat-hanger die, the following information is required: 1. its viscosity level at a given shear rate; this is required for all non-Newtonian or shear thinning fluids 2. Flow rate or range of rates 3. Material density at processing temperatures 4. General material characteristics, such as heat degradability or thixotropicity preland section; this is because of its inherent design. Rather, this style of die design relies on a larger manifold section to reduce the resistance to flow to the ends of the die; the larger the manifold, the less the resistance and the better the flow distribution. In theory, there can never be an even distribution, because no matter how large the manifold is, there will always be some pressure drop across it and, therefore, less flow to the ends of the die when compared to the center. The larger manifold has some drawbacks in that the residence time is greatly increased and flow at the ends of the die is nearly stagnant. The overall internal flow channel design cannot be increased or decreased to suit a given application, as it is restricted by the need for manifold size to achieve some flow distribution. If the flow presented to the die is not constant over time, if the fluid is not homogeneous in terms of temperature and mix, and because there are inherent errors in viscosity measurement and theoretical flow calculations, thickness variations will sometimes occur. To adequately adjust these flow variations, a flexible lip is required as a fine-tuning adjustment. Having multiple entrances or a pump within a die simply represents attempts at producing uniform distribution, minimizing the effect of transverse pressure drop, and simplifying the job of manifold design. DK4036_book.fm Page 3 Monday, April 25, 2005 12:18 PM © 2006 by Taylor & Francis Group, LLC manifold to compensate for this pressure increase (see area B, Figure 19.1). The pressure drop in the Rheology curve (see Figure 19.3) — a rheology curve, a fingerprint of a particular resin, predicts The T-shaped manifold in the constant cross-sectional style (see Figure 19.2) has no compensating . Positioning 19 - 8 19 .11 Backup Roll Design 19 - 12 19 .12 Automatic Control 19 - 12 19 .13 Deckling 19 - 14 Air Entrapment behind Deckling 19 .14 Die Cleanup 19 - 14 19 .1 Introduction . -1 17 Under What Regulation? 17 .1 Introduction 17 - 1 17.2 Code of Federal Regulations 17 - 1 17.3 Title 29 (Labor) 17 - 1 17.4 Protection 17 - 2 17 .5 Biocides 17 - . 15 -1 15 Design of Experiments for Coatings 15 .1 Introduction 15 - 1 15. 2 Standard Two-Level Factorial Designs 15 - 2 Case study — Screening