A constant and even ink transfer is of great importance for the final printing results. A variation leads to tonality changes for the individual printing stations which in turn causes color variation of the final product many times, and so large that the final product is unacceptable to the customer.

For the flexo- and rotogravure processes the ink is transferred via rolls where a pattern with defined area and depth has been engraved into the roll surface. The rotogravure process transfers the ink directly onto the substrate, whereas flexo transfers it by means of an engraved anilox roller to plates mounted on the print cylinder and these plates transfer it onto the.substrate. Ink is applied to the engraved roll, and the excess is doctored off with a doctor blade, or for some flexo machines by means of a rubber roll which squeezes off the excess amount of ink. For high quality flexo print a closed doctor blade assembly is used where the ink is pumped into this assembly which consists of two parallel doctor blades with seals in each end, which seal against the anilox roll. In this way a closed chamber is formed into which the ink is pumped, the excess doctored off and pumped back into the ink tank. Since each engraved cell of the anilox roll represents a constant volume, a constant ink volume per unit area is transferred to the plate on the print cylinder unaffected by line speed variation since the anilox peripheral speed follows the line speed.

A constant ink transfer presumes a few things; no dried ink may remain in the engraved cells, each cell must leave the same amount of wet ink each time it is transferring ink onto plate or substrate, no gas bubbles in the ink, which change ink density of the transferred ink should exist, or at least not vary, pigment concentration and pigment dispersion should be constant. Viscosity control is used to control pigment concentration- higher viscosity means higher pigment concentration. During the

printing process solvent or water evaporates from the ink, and additional solvent is added to compensate for this loss so that the ink viscosity all the time remains constant. Even small gas bubbles in the ink, which reduce ink density and pigment concentration, increases the viscosity which then causes the viscosity control system to add more solvent which further decreases pigment concentration. The viscosity control thus increases the error instead of decreasing it. Since some of the available volume in each cell is filled with gas instead of ink, the volume of transferred ink is further decreased. The empty cells will pump air bubbles into the ink when re-entering into the ink again, and it becomes obvious, that elimination of this problem is very important in high quality printing operations.

The foam problems when using waterbased inks are much more severe than those in solvent based. As mentioned above, the foam that is noticeable is less troublesome than the one in the form of small bubbles inside the ink circulation which suggests, that it is more difficult to obtain good quality prints when using waterbased than when using solvent based inks. Ink drying is different for solvent based and waterbased inks, which represents another probiJ|em. with waterbased inks. Solvent based inks dry in the conventional way, the solvent evaporates leaving the varnish and the pigments on the printed surface, whereas the water based solidify on the substrate surface through a combination of cross linking and evaporation. Cross linking is here the most important factor, and the evaporation of water is only necessary to have an acceptable substrate after it has been printed.

Printing on paper is easier since some of the water migrates into the substrate, whereas printing on plastics makes it necessary to evaporate all the excess water after the deposition of ink onto the substrate. This means, that the cross linking must happen very fast if one wants to obtain a high production rate. Air bubbles present in the ink circulation will contribute to a cross linking already before the ink is transferred to the substrate, which in turn aggravates the problem of filling the

cells in the anilox roll with ink that has already crosslinked. Crosslinked ink is very difficult to remove from the cells, and has until now only been possible to remove in special cleaning tanks, where a cleaning fluid has been used in combination with ultrasonic energy. Without this cleaning method the cells cannot be opened, and the roll will be permanently damaged.

Considering all of the above mentioned problems with air bubbles in the ink, it becomes obvious, that a reduction or elimination of foam is very important to arrive at constant ink transfer and not damage the anilox or rotogravure rolls. The most commonly used method is to add defoaming chemicals to the ink. Since they affect the surface tension they change the properties of the ink which can cause other problems. Since the printed substrate very often is used for food packaging, they will also have to be accepted by health authorities for this application.

A solution to above mentioned foam problems and to develop a method that prevents filling of gravured cells and has a potential to clean also those already filled with solidified ink is thus of great importance to arrive at high production rate, good quality and controllable printing process.

Ultrasonic energy contributes to a solution of all above mentioned problems, degassing, which reduces foam, cleaning, which prevents or cleans cells in rolls, ink dispersion for better color density and improved wetting between carrier and pigments. The ultrasonic energy can be added either externally or internally into the ink in the closed doctor blade assembly for flexo or ink tray for rotogravure processes. If cells are already filled with solidified ink one can add the ultrasonic energy to a cleaning fluid which will speed up the cleaning. This method is also used during color changes for complete cleaning as well of rolls as of doctor blade assemblies or ink trays.

One should therefore expect ink savings through improved color density and also a possibility to change over to a roll with

reduced cell depth with higher pigment concentration of the ink since the dispersion even for higher pigment concentrations results in a possibility for higher productivity, where drying of the ink is a bottle neck in the production. It will also make a meaningful viscosity control possible.

To obtain these results it will also be necessary to add ultrasonic energy to the ink already in the ink tank. As discussed later this will have to be done prior to the production start so that the ink is normalised by dispersion and degassing prior to start up. Unless one does this, the ink transfer will vary until such time as all ink is degassed and ink constituents completely dispersed.

The method described above has suggested ultrasonics via doctor blade assemblies or ink trays. Modifications are possible e.g. addition via anilox- or rotogravure rolls, but these are intended to be within the scope of this development.

EXPERIENCES

In order to verify above reasoning a couple of tests were carried out. In accordance with the first tests a standard ultrasonic double walled cleaning tank was made available together with a generator to drive the US transducers. An assembly for hanging the ink tin in this tank without metal to metal contact was manufactured, the ink tin was placed into this assembly, water with detergent added was filled into the cleaning tank up to 2/3 of the ink level in the tin, and the test could be started. The test was carried out using a process red ink for printing on paper. Foaming has always been a big problem associated with printing with this ink.

The ink tin was opened, and viscosity was measured and found to be 140 s using a 2 mm cup. When we switched on the US generators we could notice gas bubbles coming up to the ink surface. After one hour the viscosity was measured again and found it to be 45

s. Temperature and pH did change only marginally, temperature by less than 2 degrees C and pH by 0.1. The dramatic reduction of viscosity happened without any other addition than US energy. The viscosity remained unchanged during a two weeks storing of an ink sample.

Smear tests were carried out, and an addition of one litre of water at a time into the 18 litres of ink was carried out with smear and viscosity tests after each water addition. Color strength ws deemed to be the same up to 5 litres of water, and only slightly affected and satisfactory up to 7 litres of water. Viscosity was 12.5 s after 7 litres of water addition. This proves, that the pigment dispersion was working well. The dramatic viscosity change with US addition only, shows' tha even the acrylics has been well dispersed.

Printing tests were carried out during several weeks, and the first tests could be verified.

A water addition of 18% demonstrated better color strength than the normal at around 4% and resulted also in much improved printing quality. This water addition resulted in a viscosity of 16.5 s as compared to the normal 25 s. Normally, this should have caused much ink splashing, but this was not the case indicating that pigment wetting and surface tension had been influenced in a desirable way.

The same tests were also carried out using solvent based inks, and results were similar with one exception. Viscosity was only marginally affected. Degassing was observed, foam reduced and color strength improved. Tests were also made using a high concentrate solvent based ink with assistance from the ink supplier, Sun Chemicals. Color density for the different tests were carried out by means of a color densiometer. Ink type used was a high pigmented process blue ink viscosity controlled to 24 s for all tests. Anilox roll was a 140 lines/cm ceramic roll with 24 micron cell depth. Two test plates were used, one screened,

the other solid. The ink used for US dispersion test was predispersed during one hour prior to start-up. The printing station was equipped with a CARINT closed chamber doctor blade assembly. Test results are shown below:

Color density.

Screened plate normal ink: 1.4-1.6. US dispersed ink: 1.7-1.8

Solid plate normal ink: 1.5-1.7 US dispersed ink: 1.75-1.85

Visual inspection

Normal ink: Uneven inking and very undistinct dots for screen plate.

US dispersed ink: Even and good inking with very distinct dots.

Additional tests have been carried out on a Sund Emba machine for printing on card board. A 2.5 meter closed chamber doctor blade assembly was equipped with 2x8 US transducers driven by a 750W generator. The customer had severe filling problems of the anilox rolls. Hot water with detergent addition was circulated through the doctor blade assembly, and immediately after the US generator was started the water changed color. After a one week continuous US addition, sometimes to a cleaning fluid, sometimes to the ink, the roll was completely clean. It should be mentioned, that original cell depth was 45 micron, and cell depth at test start was 18 micron. It was also noticed, that foam was considerably reduced when US was added to the ink in the assembly. Other print technical advantages have not yet been evaluated, since color density changes will be difficult to evaluate with an ink transfer as high as the 45 micron cell depth will give.

Results from the above mentioned tests indicate, that theoretical advantages have been realized also in practice. So far all tests have been carried out in the flexo process, but most likely they will also materialize e.g. in the rotogravure process.

Either solvent emission is reduced by changing to high con¬ centrated solvent based inks with subsequent change to reduced cell depth, or by changing over to waterbased inks, the filling

of cells together with uneven ink transfer for other reasons, the problems encountered so far are possible to overcome completely. For flexo printing this assumes a change to closed chamber doctor blade assemblies instead of other inking systems, and that the ink must be normalised by US dispersion prior to viscosity control and start-up.

The methods described will also simplify order changes and cut stop time since cleaning of doctor blades and anilox rolls can be made without manual cleaning. Even changes between waterbased and solventbased inks will be very simple without ink transfer variations due to residual water or solvent remaining in the cells even after very thorough manual cleaning. The crew can therefore concentrate only on plate cylinder change and register problems with very much downtime reduction and mote production time as result of this.

Further tests with different types of inks and on different printing processes will have to be carried out, but unless no new problems will arise, which is unlikely to happen, one or several upgrade packages will be available latter half of this year. They will consist of ink tanks and doctor blade assemblies with US actuators mounted together and an anilox roll with suitable lines/cm and cell depth to fully take advantage of this new technology. Most existing doctor blade assemblies can be upgraded with US, but this will require that it is sent to our factory for mounting of actuators and tuning of system.