The sensor is mounted in a connector with an inlet and outlet opening and installed in the supply line between the ink pump and the doctor chamber (seen in Figure 2). Influences such as machine vibrations or pressure pulsations from the diaphragm pump have no effect on sensor operation or measurement accuracy.
The sensor is maintenance-free—each cleaning cycle of the lines and doctor chamber ensures that the sensor is clean again, since it is automatically washed in solvent. As shown in Figure 3, only a very thin haze of color may remain on the sensor, which has no influence on its accuracy or repeatability. And because of the sensor’s robust construction, any necessary cleaning can be done with a solvent-soaked rag, with no danger of damaging the sensor or changing its calibration.
All sensors are separately connected via industrial grade cables (Figure 4) to their electronics units, and these communicate with an industrial grade computer. The computer controls a valve island, which in turn controls the pneumatically-actuated valves for dosing solvent. The system (Image 1) includes a touchscreen, next to the operator control panel, which operates the intuitive user interface of the viscosity control software. In the creation of the user interface it was obvious that it had to be clear, intuitive, effective and quick to operate.
The interface displays a dashboard, on which the operator can monitor the viscosity of all stations. Touch-sensitive controls enable the operator to switch individual stations on or off, enable automatic control and to set the viscosity limits. A separate station hub switches to a display that monitors the viscosity over time and allows adjustment of specific sensors and valves. Furthermore, the software notifies the operator when the viscosity changes are too large and helps by making the right correction to solve the problem.
Predictive Tracking Control
During printing, there is continuous evaporation of solvents. Evaporation increases with faster printing speed and rising ink temperature. Sensors measure the actual value of the viscosity and ink temperature once per second, enabling the software to calculate the temperature-compensated viscosity. This, in turn, enables the controller to determine whether the temperature-compensated viscosity falls within the desired tolerance. The controller will add a quantity of solvent that depends on the size of the deviation from the setpoint. During printing, it is possible to maintain a deviation of only 0.5 percent from the setpoint. Dosing valves used can add very small amounts of solvent that are necessary to achieve such fine control. Figure 5a and Figure 5b are of the same color with different scales, with vertical magenta lines indicating automatic solvent dosing.
The control system is extremely accurate because it can compensate rapidly for the evaporation that is occurring constantly during the printing process. To achieve such very small deviations from the setpoint, the system will sometimes dose as little as 10-g. of solvent every 30 seconds.
If an ink with too high a viscosity is added to the ink bucket, the control responds immediately by measuring the response to each solvent dose, with the subsequent dosing of solvent being adjusted accordingly, as in Figure 6. In the end, the setpoint is reached very gradually with very little overshoot. Besides the extremely accurate control, it’s possible to keep the viscosity stable when the level in the ink bucket is very low, just enough to pump the ink through the system.
QA & Standardization
An experienced operator knows what viscosity must be maintained for which types of ink in the particular process being used. This depends on the kind of ink—the Pantone color as well as special challenges, such as those presented by metallic and white inks, which have a somewhat different behavior with temperature than “normal” inks. Desired viscosity also depends on the type of substrate on which the printing is done.
To better understand the problem and its solution, we performed a series of experiments on the effects of ink dilution on print quality and measured ink viscosity. With these results, we know which viscosities have to be maintained for the type of substrate (paper, polyester, polyethylene, polypropylene).
In a first experiment, 10-kg. ink was 10 percent diluted, press running at 656.17 fpm, the polyester film was marked, and the press was stopped. The ink was diluted with a further 3 percent of solvent, the ink was circulated until the viscosity stabilized and the process was repeated a total of 15 times. The film was removed, all 15 segments were measured with a spectrophotometer, and photographs of the film segments were made for subjective visual evaluation.
Figure 7 shows the visual appearance of the printing quality at a series of dilutions. It depicts color density variation with ink dilution and viscosity.
At the lowest dilution (highest viscosity), too much ink is deposited and does not flow properly. Pinholes develop and overall quality is poor. Although the color between the pinoles is quite dense, the measured density is low, due to the high reflectivity of the pinholes. As dilution increases, viscosity decreases and the flow improves, but pigment loading decreases and the color becomes lighter. Each sample was measured with the spectrophotometer and compared with the digital PMS reference. Figure 8 and Table 1 show the Delta E 2000 and color density as functions of dilution and viscosity. Viscosity difference values are referred to sample 6, which is the target density.
This experiment shows that with this system, very accurate viscosity control can be achieved, with a viscosity bandwidth of 0.5 percent. By dosing very small quantities of solvent about every 30 seconds, the system enables very small variations in Delta E values to be achieved.
At the time these experiments were done, the customary viscosity bandwidth was ± 0.5 cup seconds (about ± 2.2-mPa.s) with the viscosity being checked manually every 15 minutes to 20 minutes. The amount of solvent that was then dosed was between 0.2-kg. and 0.5-kg. (depending on the ink coverage, type of solvent, anilox volume, machine speed and temperature).
We now have changed the process of printing a Pantone color, because we not only know which viscosities have to be maintained for the type of substrate, but can hold tight tolerances on this viscosity. Certain substrates require a higher viscosity due to the fact that the ink “sinks” too far and so the structure becomes visible, resulting in a decrease in color strength, while other substrates need a lower viscosity due to their smooth surface and good ink acceptance.
With the experience gained with this system, we know exactly which viscosity should be maintained for the type of substrate (polyethylene, polypropylene, polyamide, polyester, paper and biodegradable), and have actually determined a standardization for ourselves.
With the very first print, the color density of the Pantone color is measured and then the operator checks whether the ink has the correct viscosity for the relevant substrate. (The ink is usually not brought to the correct value in advance because the substrate may vary slightly in terms of surface quality, so we have some room to play with the viscosity for optimum results.)
In the older method, if a color had too high color density, we reduced it with varnish and/or with a different anilox roll. If in doubt, the viscosity was checked with a cup, which usually necessitated re-calibration of the relevant sensor.
Because we now have a more reliable measure for the initial temperature-compensated viscosity of the ink, its viscosity can immediately be adjusted automatically by diluting an ink to the correct value. Because the correct viscosity values are maintained, this leads to better ink transfer from anilox roll to printing plate and finally to the substrate. Contamination of the anilox roll can also be noticed earlier because we know which color strength should be reached with a certain viscosity.
Too high viscosity leads to poor transfer resulting in visual characteristics like opacity and “ghosting.” Due to a more accurate viscosity, the cell of the anilox roll is better emptied and the ink usually flows better, giving in a smoother ink layer and increased color strength. With increasing speed, the ink transfer decreases, but because the ink has the correct viscosity and performs optimally, these variations are smaller compared to our earlier method using cup-calibrated sensors.
In the last six months, we have improved color quality and are able to maintain much smaller deviations of Delta E 2000 values, especially during long runs. A result of tighter viscosity control is that the print inspection system sees far fewer errors in color strength deviations. Machine operators show high confidence in the accurate and repeatable values of the sensors and control system. This trust has led to our press achieving excellent print quality for jobs small and large.
In addition, we have performed neither maintenance nor calibration of the sensors and beyond our initial temperature-compensated viscosity parameter measurements for each ink, no further standardization of the viscosity values needs to be done. We now know what viscosities should be maintained for specific substrates. After printing each new order, we store the viscosity set values and use these for repeat orders of the same job.
mPa.s to Cup Seconds?
Although formulas exist to convert viscosity values from mPa.s to DIN cup seconds, we have found that abandoning cup seconds has several advantages.
Above all, it has changed the way we think about viscosity. As long as we thought in terms of cup seconds, keeping a tight control on viscosity seemed like an impossible task. Our expectations were limited by our previous experiences, so we set the bar lower than was necessary to achieve the kind of print quality we knew was attainable.
Furthermore, thinking in terms of cup seconds made us want to grab a cup and check the accuracy of these new sensors, with which we were unfamiliar at the time. But checking the accuracy of the sensors with a much-less-repeatable method can give the false impression that the sensors themselves are not repeatable!
It is only when we compared actual printing results using the new sensor system to what we were accustomed to that we saw the real value in thinking in the new, unfamiliar units. It enabled us to “think small,” to be able to see small variations in viscosity that were otherwise invisible. Furthermore, it let us get our viscosity under tighter control, which had direct positive impact on the quality of our final product, which is, after all, our main goal.
As printing speeds increase, and profit margins get tighter, “getting it right the first time” becomes much more important. An error in initial viscosity setting can result in producing several thousand meters of waste in no time at all. Tight control with an accurate sensor, combined with a responsive control system, has enabled us to streamline our printing process while improving color quality and reducing waste.
About the Author
Bert Verweel is the owner of Maasmond Papierindustrie bv Oostvoorne in the Netherlands. He did his bachelor’s degree in mechanical engineering from TU Delft, and has 25 years’ experience in flexographic printing, laminating, engineering, biological treatment of air emissions. Over the years, he has tested multiple types of inline sensors for ink viscosity monitoring and control.
Maasmond is a family owned company (founded in 1969) with 50 years of experience in converting of paper and plastics. For the converting of food and non-food packaging and labels, it houses a broad range of modern narrow and wide web flexographic printing presses (UV- and solvent-based inks), laminating (water- and solvent-based adhesives), and slitting, diecutting and perforating machines. Maasmond produces a wide range of high-quality labels and flexible packaging which adhere to the highest food packaging standards.