Once upon a time the beverage can was growing in worldwide usage. Bottles were near obliterated, at least in the US soft drinks market. Even beer in metal packaging was growing.
That was yesterday.
Today’s drinks markets are changing, as they have for the past several years. Today’s GenX’rs have grown up with the widest packaging variation ever available coupled with constant innovation, albeit in alternate packaging media. Plastic packaging is encroaching upon the metal beverage can’s once-dominant growth position. Glass bottles are making some inroads into the beverage can’s hold upon the beer segment. New drinks are being packaged in anything other than metal. Why?
Canmakers, financial analysts, equipment suppliers, and bottlers have been wrestling with this question for several years. There are different opinions; different answers. However, the most simple may be summed up within the following (true) story.
A friend visited the local pizzeria for lunch. She noticed the single-serve refrigerator no longer contained any cans, just row after row of plastic bottles. She asked the owner why this was. He answered in a quite matter-of-fact manner, “the kids don’t like them; they’re old-fashioned.” Keep in mind that this lunch spot sits between two schools.
Pretty disconcerting, isn’t it? But not surprising as one looks at the single-serve section in the local convenience store, or even along the aisles of the supermarket. Other packaging media are just screaming out to consumers, begging to be taken home, if even for just a try. Granted there are other issues, such as convenience, recycleability, transportability, “cold feel”, etc. But the fact is that more people are turning toward other packaging media besides metal for their beverage selections, at least in the United States. Why?
One of the reasons is obvious: the cans are having a harder time begging for attention amidst all the innovation occurring in other media. The can has become a commodity and is therefore currently trapped in a commodity pricing structure. Other media commands a premium price, making the cost of innovation that much more palatable. Sure, the can industry has attempted to invigorate the can via shaping, embossing, instant draft, etc. However, these marketing strategies are significant cost adders, something currently intolerable in our industry.
Speak to most anyone in the beverage can industry about what can be done to attract the consumer while maintaining the can’s price competitiveness and you will almost always receive the same response: “improve print quality.” We all know the billboard potential of the can, so why haven’t we made much progress in this area? At least not enough progress to ebb the tide of package competition?
The answers to these questions
require a little history. 
Circa 1970 continuous motion printing for beverage cans was invented. It comprised low speed (by today’s standards), "medium quality" printing ability in the US market, which was, and mostly still is, marketing the can as a commodity. This was what the industry wanted (and needed) at the time. As time progressed, equipment modifications were made to advance the technology to medium speed (again, by today’s standards), still with "medium quality" decoration. So far, so good. During the early to mid 1990s the technology was advanced to high speed, albeit with some "loss" in decoration quality. However, this was acceptable, as the beverage can was still the dominant force in its packaging segment relatively simple label designs were still acceptable. Concurrently it was clear that a quality market was emerging in several other countries where speed was not an important factor.
As the industry approached the millennium, however, one could see a marked "decline" in the print quality of at least the US market. Defects that once were frowned upon (and probably still are) became normal operating procedure in our business. Just pick up a can from a supermarket shelf and read the ingredients (if you can). Further complicating this is the inability for either the quality market or the commodity market manufacturers to enter the “impenetrable” area of both high quality and high speed.
Today’s halftone printing technology on two-piece cans is a huge diversion from what was done over thirty years ago. Today, as in other printing industries, we are in the business of optical illusion: fine dots creating images and colors that really “are not there.” Of course, we deal with different problems than in, say, the paper printing industry, as our substrate is essentially non-absorbent. We also face problems different than that in flat metal sheet printing, as we are attempting to print in the round without the additional advantage of individually curing each color for process printing.
Currently applied technology employs the simplest form of halftone printing, namely amplitude modulation. Simply put, AM achieves its “illusions” by varying the size of the halftone dots while maintaining the distance between them. When employed to create shifting tones of the same color, for example, the dots increase in size to correspond to a darker shade, and likewise decrease in size for lighter shades.
These dots may also be used to “create” colors. As an example, laying down red and black dots beside each other gives, at a normal reading distance, the illusion of a shade of maroon. The shade will depend upon the size of both color dots, as well as their spacing. Color illusion may also be achieved by laying one color upon another in wet-on-wet printing.
A derivative of amplitude modulation is frequency modulation, commonly referred to as stochastic (from the Greek for random) printing. Instead of varying the amplitude, or size of the dots, first order FM employs smaller, variably spaced dots of equal size, hence the modulation of frequency. This provides the opportunity for increased screen resolution, but is more difficult to print, as dot gain (both mechanical and optical) becomes more of an issue than with AM. This has been employed on a limited basis within our industry, due mostly to the lack of proper equipment.
In
turn, a derivative of first order FM is second order FM. In this technology
both the amplitude (or size) of the dots and their frequency (or spacing) is
varied. This allows for higher resolution and better-defined graphics, along
with the virtual elimination of moiré patterns (unintended dot color patterns
which can distract from the desired optical illusion). This, however, is yet a
step more difficult to produce, and to the best of our knowledge has yet to be
used in the two-piece industry.
Going back once again to the 1970s, it is important to note that at such low speed the variables of the process of printing were not as widely dispersed as they are today. Take ink as an example. Back then the ink had to deal with much less process variation, notably changes in machine temperature and dynamics due to varying line speed and, to a lesser degree, various ambient temperatures. However, today’s inks are formulated to work within a minimum range of 30 degrees F (17 degrees C) because of severe process inconsistencies, especially as line speeds vary and as ambient temperatures worldwide add to the confusion.
What effects does this all have on print quality? Taking this a step further, just what is print quality? Print quality on beverage cans may be broken down into several discreet, broad categories:
1). Static registration: the ability to set and hold register of colors one to another.
2). Color variation: the variation of color on a given decorator with given ink on a given label. Worse yet, the color variation between different lines in a plant. Worse yet again, the color variation from plant to plant and canmaker to canmaker.
3). Mechanical dot gain: the level (and predictability) of dot size increases from plate to blanket and blanket to can due to normal, natural physical phenomena.
4). Optical dot gain: the measure of “hard” versus soft dots. The (in)ability of a dot to conform perfectly to a set (usually circular) shape.
5). Dot shifting: the “moving” of dots on a blanket due to machine dynamics and/or the distortion of dot shape due to the same.
Let us take a look at these issues one at a time.
1). Static registration.
This is so elementary to printing, it will be given little press here. Today’s beverage can decorators are more than capable of holding initial registration within limits, allowing the potential of higher-quality printing. Plate making has become such a science, especially with today’s digital on and off-press technologies, that our industry has made great strides in the registration arena. Is there room for improvement? Sure, but some of the other issues surfacing for printing beverage cans deserve much more consideration.
2). Color Variation.
This is by far the canmaker’s largest decoration defect headache, in some extreme cases a nightmare. The problem has far reaching implications, as color variation is a problem between canmakers, plants, lines, and even the same printer during the same run. One of the major causes is temperature variation, in both machine and ink.
To understand this subject, one must once again examine some history. When continuous motion cylindrical printing was first invented circa 1970, speeds were low and so were print "quality" expectations. Cans were printed with line art and solid colors. The inks, although formulated for a wider temperature range than in other printing industries, were nonetheless more than adequate for the job at hand.
As
time marched on, speeds greatly increased. This produced wider variations in
machine temperatures and operating parameters, which in turn started to have
deleterious effects on the inks. Inker roller cooling was introduced to
partially help combat this problem, but even back then the effects of color
variation were beginning to surface. To further complicate the situation,
halftone printing was being introduced to the can, bringing with it an entirely
new set of additional process considerations.
The situation today has can lines running at extremely high speeds with canmakers attempting to simultaneously increase print "quality". In extreme cases canmakers are simply trying to control current quality. Additionally, because of these increased demands over time, inks manufacturers have been forced to formulate a wide range of inks to operate within a very wide process band, which as mentioned before is as high as 30 degrees F (17 C). Often no two plants within a canmaker’s organization use the exact same ink formulations for the same labels. Highly variable machine parameters, coupled with local ambient conditions, serve to create a lethal combination as canmakers attempt to quench the defects associated with color variation, of which there are many. This does not necessarily even begin to approach the subject of the intricacies of very high resolution halftone printing, which carries these problems as well as its own related directly to color.
The problem even becomes worse. Setting inkers today is more a magic art than anything else. Fountain key settings and roller pressures seem ambiguous, making repeatability almost impossible. How many canmakers have tried to repeat a label with little or no success? When one truly thinks about it, how could they? What data is there to go by? Precious little, to be certain.
3). Mechanical Dot Gain.
The transfer of ink in a dry offset
printing process is mostly mechanical. The plate transfers the image to the
blanket, which in turn transfers it to the can. The blanket (and the plate to a
point) is compressed in order to achieve this transfer, thereby the term
head
pressure (plate to blanket) and print pressure (blanket to can). If that were
not enough, add to this the complication that the mandrels holding the can are
cantilevered, causing some natural deflection during print. This is compensated
for by “toeing in” the mandrel, or presetting its axial displacement toward the
blanket to create a more even print under pressure.
Mechanical dot gain itself is not the culprit in dry offset printing. It is rather its unpredictability and variability that create, of course, unpredictable, variable gain. This is critical in halftone (screen) printing because the halftone dots are often used in discreet combinations to simulate other colors. However, if there is no predictability in the size of the dots that eventually appear on the can, color variation in these halftones will occur as the process runs. Worse yet, the variability in this dot gain within a single press run is often enough to create color variation off the same line during the same day.
Mechanical dot gain, therefore, also acts as a restriction on the screen definition possible. For a given tone percentage, if the line screen is increased, the dot sizes by definition decrease. Therefore, as the screens increase, the negative effects of unpredictable, variable mechanical dot gain increase as well, making it almost impossible to live with anything other than the meager line screens our industry is left with today. This is unfortunate, because increasing the line screen also increases definition of print, leading to the possibility of much better defined graphics.
Mechanical dot gain is, all things equal, mostly a pressure variation problem. As a press runs, the gap between the plate cylinder and blanket cylinder, and therefore the pressure of the plate to the blanket can change. This distorts the color separation when running a label, but even worse is hardly predictable. Maintaining consistent images becomes a very difficult task indeed. Additionally, there are also changes occurring at the gap between the mandrel and the blanket cylinder, yielding varying blanket to can pressure.
4). Optical Dot Gain.
This
is one of the least understood phenomena within the two-piece can printing
industry. Optical dot gain deals with the optical systems used by humans to
discern, in the case of this argument, color illusion. Also referred to as
“veiling,” it is loosely defined as the inability of the halftone dot to fill
the entire form (most often circular), for which it was defined. It is also
defined as the difference between a “hard” and “soft” dot, or even more simply,
a “sharp” dot and a “fuzzy” dot.
This characteristic obviously has a drastic effect on the clarity of the graphics being printed, as sharpness of the images may be directly correlated with optical dot gain. However, it is responsible for another, equally important aspect of high definition graphics, as it has a drastic effect upon the human optical system’s ability to be “fooled” into thinking a color is present that is actually being produced by the strategic placement of different color dots side by side.
When examining illusional color, especially in the transition from one illusional color to another, the human eye focuses on the circumference of the halftone dot. Of course, in the case of dot shapes other than circular, the eye is focusing on the perimeter. This is an amazing, but true phenomenon of the optic nerve, but a very important one when we consider our attempts to increase line screens. A simple example will explain.
Let us assume we are running a particular halftone at some particular screen. The value of both is almost irrelevant for the sake of this argument. If we were to magnify one dot and consider a square area of the can, such that this area is equally ½ line screen spacing all around the dot, we then calculate the circumference in that area upon which the eye is focusing, which of course is very straightforward.
Now let us assume that in an extreme case we would like to double the line screen. By definition that same square area of the can would now contain four dots, albeit half the diameter of the original dot. A simple calculation now tells us that we have twice as much circumference as before, with twice the room for error. Not only would our images look “fuzzy” with this increased line screen, but also the eye would have great trouble discerning the colors and color transitions, assuming we were dealing with excessive “veiling” in the first instance.
Optical dot gain is truly one of those unwanted effects produced by the entire printing system. Inks whose pigmentation and carrying agents have been compromised due to a decorating process that is out of control are one cause. Machine vibration, inconsistent print pressure between mandrels, and dot distortion also create significant optical dot gain problems.
5). Dot Shifting and/or Distortion
This is a phenomenon caused mostly by machine elements. On today’s modern two-piece container decorators, the blanket wheel system acts like a large flywheel. The bullgear, which is mounted to the same shaft as the blanket cylinder, drives the plate cylinder gear, which in turn is on the same shaft as the plate cylinder. This sets up torsional possibilities, and when coupled with the natural torsional properties of the blanket cylinder shaft, problems can occur, especially during line controlled speed changes.
When accelerating (or decelerating) several torsional moments are occurring in the print head, yielding several different torsional deflections. This in turn creates a time differential in when the various components are affected by these natural phenomena. If not properly engineered to be almost oblivious to these effects, the printing process will be affected to a great degree.
In
its mildest form, the dots become distorted, usually taking on elliptical form
in the case of circular dots. This affects the entire intended color separation
in much the way it is affected by mechanical dot gain. However, it is more
complicated in that the “gain” occurs only in one direction, that is the
direction of rotation of the blanket cylinder.
In its severest form, this distortion actually takes on the form of two discreet identifiable dots where there should only one. This is caused when one dot is laid on the blanket cylinder and printed on the can while a severe acceleration or deceleration occurs on the blanket cylinder shaft. The first dot will have left an inked “witness” on the blanket that is picked up by a can upon next contact. However, the second dot is also picked up, but in a shifted position due to torsional deflection. Although the first dot usually appears as more of a shadow, it is indeed recognizable.
So, where do we go from here?
The printing of two-piece cans is a process affected by many variables. In Six Sigma terminology we separate these broadly into “controlled” and “noise” variables. The noise variables are the ones that give us the headaches. However, when designing for Six Sigma quality, it is the norm to instill “robustness” in the design. This is simply assuring that the process be impervious to these unpredictable noise variables, or they be converted into controlled variables.
Is it possible for the two-piece can industry to get the decoration process under control, thereby improving existing quality? The answer is an undeniable YES. Furthermore, is it possible to then elevate the quality of printing on cans? ABSOLUTELY. The most important question, though, is will this transition require exotic technology, giant departures from current operational methodologies, or worse yet, disposal of existing printing equipment? ABSOLUTLEY NOT.
Equipment manufacturers have already built in several enhancements to allow canmakers to print with "high quality" at production speeds. Additionally, statistical methods are available to ensure that process outcomes are predictable by focusing on process input variables, rather than inspecting and "waiting to see" if the process is under control.
