Bron Wolff discusses how you can assess and improve your UV-dryer performance by learning from his company's experience.
UV inks have revolutionized graphics screen printing, but curing these finicky formulations remains a challenge to any shop interest in hig-quality, high-volume production. Besides matching ink characteristics with the correct curing parameters, printers must contend with power fluctuations that reduce dryer performance and high temperatures that can alter sensitive substrates and ruin registration on multicolor jobs. The printers at Joliet Pattern, Joilet, IL, know these problems well and decided to end them with a few modifications to their UV dryers. Joliet's production manager, Bron Wolff, describes how you can assess and improve your UV-dryer performance by learning from his company's experience.
I'd like to preface this article with a few simple statements that I believe to be fact. First, UV inks are not created equal. In fact, all colors within a given UV ink line will not cure the same under identical conditions. Second, light cures, not heat. Heat helps, but light--specifically UV light--is the key. Finally, more of anything is not necessarily better. This is especially true with UV curing, where more energy typically means more heat and more problems for the printer.
To be fair and honest, heat does play a part in curing. It improves the wetting properties of some UV inks, helps with crosslinking, and starts the post-curing process. But while a little heat can help the curing process, excessive heat can hinder it.
Initially, it wasn't heat that made us focus attention on our curing systems. It was unpredictable changes in curing-system performance. We found that jobs we cured successfully one week with particular lamp and belt-speed settings would fail to cure properly the next week, even though the settings remained unchanged and the curing unit was correctly maintained. As we eventually discovered, the problem didn't lie within our plant, but came from varying voltage levels supplied by our local utility company.
Powering UV-curing units
Power for manufacturing operations like ours comes from transformers located outside the plant. The power is rated in kilo volt-amps (KVA). A transformer rating of 25 KVA, for example, means that the transformer delivers 240 volts (triple phase) and 60 amperes per leg.
However, if a company is like Joliet Pattern, which is located in a large industrial complex with other manufacturing operations, it's likely to see periods of peak usage when the incoming power level drops. Such power variances will directly influence a UV dryer's ability to deliver a successful cure.
When you turn on a UV system and set the wattage level at which the lamps will operate (e.g., 125, 200, 300 watts), the system will draw a particular amperage until it is turned off. Think of voltage as a reservoir of electricity and amperage as a pipe through which the electricity flows from the reservoir. The higher the amperage, the wider the pipe.
You can monitor and record the power going into the curing unit, and as our experiences prove, you should do so regularly because it will change weekly, monthly, and seasonally. The power available from the transformer directly affects the voltage available at the UV ballast. Most ballasts for UV lamps are designed to operate at particular voltage levels, such as 208, 240, or 460 volts. The main thing to verify is that the correct amount of voltage is available to operate the curing system.
Using a voltmeter is the easiest way to find out what voltage is coming in. Begin by checking the voltage at the terminal that will power the curing unit. First, check the voltage with no load (all UV lamps off). The voltage should be within ±5 % of the expected rating. Next, power up the curing system and check the voltage again. The amps you are now drawing through the lamp should cause a big voltage drop. It is not uncommon to lose 20-40 volts on a 480-volt line (or 10-20 volts on a 240-volt line).
Note that this is just the loss experienced on one electrical line supplying one machine with one curing head. If you're running an inline printing system with multiple curing modules, you'll need to power them all up and check the voltage at each curing head. With multiple lamps drawing power simultaneously, you'll see an even greater voltage drop. The amount of the voltage drop can even vary from one lamp assembly to the next.
So what's the problem with low voltage? The problem is that when you vary the voltage and amperage provided to the lamp, you vary the radiant energy output (measured in millijoules) delivered by the lamp. If the lamp isn't delivering the proper energy to the print surface, it simply won't cure the ink.
To verify that you're getting sufficient curing energy from your system, you have to measure the energy output of the lamps using a radiometer. UV printing without a radiometer is like playing poker against a stacked deck. Radiometers make it easy to identify potential voltage drops. For example, by using a radiometer you'd quickly notice if a lamp-power setting of 200 watts and a belt speed of 70 ft/min gives you a different energy-output reading on different days, at different times of day, or at different times of the year.
Once we identified that voltage fluctuations were making it difficult for us to get a consistent cure, we invested in several step-down transformers. These units allow us to jack up incoming voltage levels when they fall below the levels we need to correctly operate our curing systems. However, due to component incompatibility and related concerns, several of our curing systems could not use step-down transformers, and we had to look for alternative solutions. So we turned our attention to the UV lamps.
Electricity is supplied to a UV lamp through a ballast, which is basically a transformer. The ballast takes the available voltage and converts it into the voltage supplied to the lamp. The incoming or primary voltage is directly proportional to the outgoing, or secondary, voltage to the lamp. The amount of voltage the lamp receives through the ballast directly affects the amount of energy (UV light) the lamp will generate.
Ballast manufacturers provide specification sheets for their products. As an example, consider a ballast for one of our old 72-in. lamps. It called for the lamp to use 2160 volts and 10 amps at a power density of 300 watts/in. However, in order to achieve the true 300 watts/in. over the entire 72-in. lamp, we needed to calculate the total power required for the lamp by applying Ohm's law, which states that volts x amps = watts. This gives us 2160 volts x 10 amps = 21,600 watts.
But wait. We also had to figure in a power factor, a number that is typically absent from spec sheets and represents the actual power delivered to the curing system. In our case, the factor was 0.92 (your local power company can tell you the power factor for your facility). When we multiplied this value against the actual wattage and divide by the lamp length, we found that the corrected theoretical output of the lamp was 21,600 watts x 0.92 ÷ 72 in. = 276 watts/in.
We wanted to check theory against reality, so we went back and measured the actual voltage being delivered to our system. We discovered that the 72-in. lamp was supplied with 2166 volts, but was only pulling 7.9 amps.
Going back to our formula, we calculated the actual power rating of our lamp as 2166 volts x 7.9 amps x 0.92 = 15,742 watts. Dividing this number by the arc length, we got 215 watts/in. as the lamp's real-world output--more than 22% lower than the projected output and more than 28% below the 300 watt/in. at which the lamp was rated!
The ballast dictates the current that reaches the lamp, and we couldn't change the 7.9 amps available unless we modified the whole power supply from the ground up. This option was too expensive, so instead we decided to increase voltage available to the lamp and the efficiency with which our lamps were converting current into curing energy.
We contacted our lamp manufacturer and explained our need for lamp that could handle increased voltage. We explained that the ballast spec sheet listed a maximum "open circuit voltage" of 2840, and that we wanted to bring our current voltage level from 2166 volts to about 2600 volts (leaving a margin of safety with the maximum open-circuit voltage). This would give us total lamp power of 2600 volts x 7.9 amps x 0.92 ÷ 72 in. = 262 watts/in., a bit shy of the theoretical 276 watts/in. but much better than the 215 watts/in. we originally started with.
Naturally, the lamp manufacturer thought we were smoking dope. But the real fun began when we explained our additional requests.
As I mentioned previously, one side effect of the curing process is heat. Only a small portion of the power used by a UV lamp is converted to UV-curing energy. Most of the power is converted directly to heat. The high levels of heat produced by conventional lamps causes a range of problems including substrate shrinkage, poor intercoat adhesion, brittle inks, plasticizer migration. The heat also makes the finished piece more difficult to diecut, overlaminate, and form.
We asked ourselves the following question: If heat is such a problem, and most of the power used by the lamps is converted to heat, why don't we change the curing system so that it converts more power into UV light and less into heat? This way, even if the voltage drops slightly, the more efficient lamps would still be able to provide enough energy to get a successful cure.
We started our exploration with a radiometer and a heat gun. The curing system was a 300 watt/in double-lamp unit with focused reflectors that were cleaned twice a week and lamps that were rotated regularly. We ran 4-mil vinyl, printed with red ink through the system at 50 ft/min, measuring and recording both the heat and the UV energy output. We then sped up the belt to 60 ft/min and repeated the measurement. We continued this exercise using the same type of vinyl and ink and speeding up the dryer 10 ft/min each time. At 100 ft/min, the print was still curing fine, but the amount of heat to which the prints were exposed only dropped marginally.
The thing to understand about UV curing is that the difference between dosage and irradiance. Dosage is the amount of light energy accumulated by the print/substrate during the time it spends under the lamp. Irradiance is the intensity of the energy, which influences the depth and speed at which the ink cures. The irradiance is influenced by the wattage of the lamp, the efficiency of the mercury gas within the lamp, the condition and focus of the reflectors, and age of the lamps.
More time inside a curing unit does not necessarily equate to a better cure--it only means more exposure to heat. That's why multiple passes through the dryer do no good with UV inks.
We wanted to see the effect of using a lower wattage setting on temperature and cure. So we cut the lamp power to 200 watts/in. and adjusted belt speed so that the printed substrate received the same energy dosage as it had at 300 watt/in. What we discovered was that the ink cured just fine and heat was also reduced. At that point, the metaphorical light bulb went off in my head! I though that if it worked at 200 watts/in., why not at 125 watts/in.? The fact that most printers use this power setting as "standby" mode when they break for lunch was far from my mind.
Initially, our 125 watt/in. setting was really delivering only around 84 watts/in. And when we tried to run the system at this power level, the lamps turned black at the ends--too little power was available to excite the mercury gas. So, we returned to our lamp manufacturers and asked them if the could increase the amount of mercury halide in the lamps. The amount they added was based on the electrical parameters of our equipment and the average power level our shop could supply to the equipment--these values differ from plant to plant.
Despite the increased mercury halide, the modified lamps still didn't produce enough energy to cure the inks, although they did keep heat down. So we began to consider other ways we could change them. In our next round with the lamp manufacturer, we asked the company to change the diameter of the lamp and the thickness of the quartz envelope. Because the heat is generated in the quartz around the lamp, the theory was that a smaller lamp diameter and thinner quartz wall would require less power and let more UV energy through. After several revisions, we received the modified lamp, which finally provided the irradiance we needed to get a successful cure without the heat.
Since those early experiments, we have gone through this process with every lamp and every UV dryer in the plant. As mentioned previously, it has taken several years for us to achieve successful results. The smaller diameter lamps we now use outperform the thicker lamps we used the past by delivering higher irradiance to our prints. The small diameter gives the lamp less surface area, which means there is less quartz to absorb the energy and convert it to heat. We can run these modified lamps at nearly double the power level (wattage) of the standard, large-diameter lamps before we even begin to approach the heat levels generated by the larger lamps.
On one of our inline printing systems, we have gone through six different generations of modified lamps. But the effort has paid off in effective and efficient curing with no heat-related damage on any of the materials we print, including vinyl, PVC, polycarbonate, paper, thermoforming films, polyethylene, polyester, and other plastics. In fact, I can't even recall the last time we ran a curing system at 300 watts/in.
As we were fine tuning our voltage adjustments and lamp designs, we also started to consider the effect of this low-temperature curing on our inks. Namely, we wanted to see what the results were with different ink/lamp/power combinations. Our goal was to make sure every color in multicolor jobs was curing properly, with no over or undercuring of any color by the time the job was complete,
We started by testing all the colors we typically print at different dryer speeds and power levels. Our first test with each color involved the lowest possible power settings and fastest possible belt speeds. We measured and continued adjusting belt speed down and/or lamp power up until the ink cured. We continued measuring and adjusting until we identified the belt and power settings at which each color became overcured and brittle. We now had an operating window that described the ideal curing parameters for each of our inks.
I'd like to draw your attention to Figure 1A, which charts our results in curing a yellow ink. Using a double lamp system with each lamp set at 125 watts/in. and a belt speed of 110 ft/min, we were able to cure the ink successfully. However, when different colors from the same ink line were cured under those conditions, these other colors failed. (An ink failed when it couldn't pass cross-hatch and tape testing or exhibited blocking after prints were stored overnight under a 50-lb weight.)
Our tests also showed how the curing window changed when the same ink was printed on different substrates. The nature of the pigment, photoiniators, and other ink components, along with the energy absorption/reflectance characteristics of the substrate all influence the curing window. Ink pigment, however, is the biggest obstacle in curing.
Through such testing and lots of mistakes, we learned to use the curing window for each ink to our advantage. Basically, the ink with the largest curing window goes first, while the color with the narrowest window goes last. We use this approach to determine color sequence on each job we print and set our dryers accordingly for each color. Our goal is to run each color at the lowest possible power setting and highest belt speed that will ensure a good cure.
We tend to run process colors faster than spot colors because UV light more easily penetrates transparent process-color inks, allowing us to cure them more quickly. On jobs that incorporate both spot and process colors, we run the process colors first. The spot colors, with their greater pigment load, require the most curing energy and have the greatest potential for exposure to heat in the curing system, so we save them for last when possible.
Whether jobs incorporate spot or process colors, the first colors typically get the lowest lamp settings and highest speeds. This reduces the potential that initial colors will become overcured as subsequent colors are also printed and cured. Like T-shirts underbases, our initial colors get a "flash" cure and rely on subsequent print/cure cycles to complete the cure.
Worth the effort
Our modified curing systems work. But recreating the same results in your operation will take time, effort, and help from lamp and curing-system manufacturers. Modified lamps will do you no good without the correct power. And increasing power without modifying the lamps will only give you more heat, not more energy to penetrate the inks.
What we developed is a system, one that's matched to our specific inks, substrates, and operating environment. It took us to years and countless hours to achieve, and involved a lot of sweat, time, and money. Before you commit to a project like this, ask your customers what they need. Our customers see the improvements and think our efforts were worth it. If your customers get better product quality and faster turnaround on their orders, they may, too.
Author's note: I read a lot and don't necessarily believe a lot of what I read, but every once in a while I get inspiration from someone else. Richard W. Stowe of Fusion UV Systems, Gaithersburg, MD, has written several articles on UV curing over the last five years that have appeared in various publications (including Screen Printing). His work helped us get through several issues as we were re-engineering our curing systems.
If you're thinking about a similar overhaul of your curing equipment, I recommend two things: patience and good maintenance/electrical personnel. I usually lack patience, but Joliet Pattern is lucky to have the talented Zmolek brothers, Rich and Steve. Without them, our UV project may never have seen the light.