Professional Documents
Culture Documents
Measurement and Prediction Procedures For Printability in Flexography (MP3 Flexo)
Measurement and Prediction Procedures For Printability in Flexography (MP3 Flexo)
Stockholm 2002
Doctoral Dissertation
Royal Institute of Technology
Department of Numerical Analysis and Computer Science
---
Media Technology and Graphic Arts
Respondent: Laurent, Girard Leloup
TRITA-NA-0212
ISSN-0348-2952
ISBN 91-7283-306-8
ISRN-KTH/NA/R-02/12-SE
The time flexography was synonym for “potatoes printing” belongs to the
history. All around the world the flexographic process has proved its
ability to produce high quality on a very large amount of different
substrates. Flexibility and simplicity are namely major assets of the
process, which is the only conventional printing process showing a
positive growth during the last years.
To summarize, I would like to say that this project was in a lot of aspects
en enrichment for me and consequently I will keep a positive memory of
the time I spent making this research.
PAPER I
Girard Leloup, L.
“Statistical Analysis of the possibilities and limits of Flexographic
Process Modelling.” – Presented at the TAGA 2000 Conference -
Colorado Springs (USA) – Paper published in Conference Proceedings
TAGA 2000 – pp. 469-486
PAPER II
Girard Leloup, L.
“Analysis of the correlation between the print quality required in
Flexography and the tolerances in materials and paper
manufacturing in terms of printability.” – Presented at the FFTA Y2K
Forum - Orlando (USA) – Paper publis hed in Y2K FFTA Forum
Proceedings CD – “General Session”
PAPER III
Girard Leloup, L.
“Study of paper printability in flexography by controlled variation of
the surface roughness.” – Presented at the IARIGAI 2000 Conference -
Graz (A) - Paper published in IARIGAI 2000 Conference Programme –
pp.171-189
PAPER IV
Girard Leloup, L.
”A Flex-Odyssey…” – Presented at the FFTA Forum 2001 – Nashville
(USA) - Paper published in FFTA 2001 Forum Proceedings CD – “Paper
substrates Session”
PAPER V
Girard Leloup, L.
“Establishment and Comparison of Different Definitions and
Equations for a “Printability Coefficient” for the Flexographic
Process.” – Presented at the TAGA 2001 – San Diego (USA) - Paper
published in Conference Proceedings TAGA 2001 – pp. 80-97
PAPER VI
Girard Leloup, L.
“Prediction of the Substrate Printability in Flexography by using a
new established Printability Coefficient.” – APPITA 2002 - Rotorua
(NZ) - Paper published in Proceedings 56th APPITA Annual Conference
2002 – pp.459-464
1. Intr oduction 4
2. Background 5
3. Definitions 6
4. Related Research 12
5.1 Methodology 18
5.2 Research progression 20
• 5.2.9 Modelling 24
• 5.2.9.1 Modelling of the printing parameters 24
• 5.2.9.2 Modelling of the paper characteristics 25
7. Summary of results 35
8. Discussion 36
Acknowledgements 40
Included Papers
Paper I 48
Paper II 66
Paper III 82
Paper IV 102
Paper V 114
Paper VI 136
On one side, the paper industry worked on the structure and surface
properties of the different substrates. On the other side, the suppliers of
the printing industry concentrated their research on developing new
gravure technologies for anilox rollers, new types of printing plates and
mounting tapes, improving the sleeve technology, redesigning the printing
presses, defining new ink formulations ...
Why?
The flexographic process, born under this name in 1952, has moved in the
last years from a “marking process” to a “high qualitative printing
process”. The development of the different components of the
flexographic printing process has allowed an important improvement of
the achievable quality. Today the flexographic process is the printing
process that shows the largest rate of expansion. The previsions for the
next years are also very optimistic.
The discussion I had with a technical expert summarizes the situation; his
words: “Today we draw an arbitrary statement on the term “printability”
and I am afraid that this is not a very scientific way of determining it. If
you can directly give an absolute number to this very difficult to catch
parameter, it would help us very much to get a better understanding.”
3. Definitions
The word “printability” is the key concept of this thesis and it is also
interesting to spend some time looking at its origins and to explain its
significance. The noun “Printability” can be broken down into two parts:
“print” and “ability”. A first and logical approach would give as
definition “the ability to be printed”. The suffix “ability” comes from the
Latin “abilitas” (14 th century) and means capacity, fitness or tendency to
act or be acted on in a specified way.
“Physical and chemical parameters that influence the transfer, setting and
drying of inks on the substrate.”
This might look too simple but use it as a start. I might have overlooked
something.”
“Although we use the term on a daily basis I didn't really have a good
clear definition for it. When you asked this question I started thinking
how to define printability. I didn't come to a definition in terms of a
mathematical or physical formulation but I thought along the lines of a
more general definition: Printability is the ability to imitate life two-
dimensional on paper and/or board. The better a picture or print
represents its original life example in colour, sharpness, density, the better
the printability.”
“The definition of printability is like asking where does your lap go when
you stand up? I will look up some references in some text and would you
please clarify what aspects of printability you are looking for? For
instance, printability on a styrene soda bottle label is different from
printability on a cosmetic box.”
“Printability refers to the transfer of the ink onto the paper and
corresponding processes. The main part of printability is the ability to
achieve an even ink layer over the total image carrier in order to have a
true image reproduction without faults, which will reduce the image
quality. Printability deals also with the adhesion to the image carrier.
Many factors affect printability, such as ink transfer in the printing press,
ink transfer to the print carrier, ink setting and ink drying. Important
quality aspects are: ink distribution, ink drying, image resolution, colour
gamut, density range, dot gain, rub resistance. In short, one could say that
printability is factors that influence the print quality.”
The definition of printability I have chosen for this work is the following:
4. Related Research
4.2.1 Definition
The images that are analysed are not continuous but are made up of large
number of points. Each point, called a pixel, can take a value from a
limited range, which describes its appearance. For the printing industry,
this value is a grey level or a colour. The number of pixels in an image
determines the resolution of the image. The range of values determines
the limits for contrast and brightness.
The way the printing and paper industry is working with image analysis is
to remove all the unwanted information for the image, leaving the bits
needed. The final result is often just a single number.
The following list is not restrictive but should give an overview of the
image analysis instrumentation and software available and usually
implemented in the paper and printing industries:
Image analysis is a very helpful instrument for the paper and printing
industries. The paper industry profiting from medical research has long
used and contributed to the development of the technology.
In the research, reference will be made to visual perception, and for better
understanding of the work, a short description of what it returns is given
here.
Unconsciously, the perception of the visual scene can greatly affect mood
and behaviour. The ability to characterise light, colour and appearance
impinges on numerous aspects of the daily life and as a consequence,
there is strong industrial interest in ensuring that these measurements can
be made in such a way that they correlate to the visual perception
(Lindberg, 2001).
4.3.1 Definition:
From the optical point of view, the first stage in visual perception is the
formation of an image on the retina. The quality of this image sets the
limit on the quality of visual perception. The main factors that affect the
quality of the retinal image are: Diffraction, spherical aberration,
chromatic aberration and scatter (Thomson, 2002).
For the observation, the printed images are presented to the test panels
mounted in standard daylight viewing equipment placed in an
experimental room. The room is painted in a pale shade of grey,
illuminated by daylight-simulating light tubes. The intensity of the light at
the picture location is 2000 lux and over the table it varied from 1750 to
2000 lux.
Normally, different paper qualities are used for the printing trials. In order
to avoid the test panel judging similarity by comparing the unprinted,
tinted edges of the picture, the samples were mounted without an
unprinted edge surrounding the picture.
The pictures are attached to a grey paper, Y-value 60, by means of corner
slips, with two unprinted sheets of paper as backing.
a.2 Procedure
The samples were spread out on a large table covered by a dark grey
cloth. The purpose of the cloth was to neutralize the surroundings. D50
illumination, a spectral distribution of daylight that has a correlated
colour temperature of 5000 Kelvin, is used for the test.
b. Opinion ratings
All relevant print quality factors are, for practical reasons, interpreted by
using some sort of common scale. When particular print impairments
have been previously identified and defined, their magnitudes could be
assessed on impairment scales. Studies allow finding the important
dimensions. For an overview, see table 1. Each of these was put on an
impairment scale of perceptibility and annoyance, employing nine level-
dependent criteria ranging from ‘very annoying’ to ‘not perceptible’.
Often the criterion of greatest interest is the overall quality itself.
Therefore, jointly with the impairment scales, ‘overall quality’ was
assessed along with the quality of ‘colour rendition, sharpness, contrast
and detail rendering in shadows and highlights’. The quality scales had
nine levels ranging from ‘unusable’ to ‘excellent’.
The room was equipped with standardized viewing conditions, the same
as for the Proscale evaluation. Overhead lighting had a temperature of
5000 K and the illuminance was in accordance with the specification
followed by the graphic industry. The participants of the test panel sat by
the neutral grey table, described earlier, and had a pile of samples and the
reference image placed on small tilted table. Each sample was taken one
by one and compared to the reference images and ratings were entered
directly on a computerized form. Ratings were given to each of the
attributes compared to the reference image. The judges were encouraged
to move and handle the samples as under normal reading conditions in
order to get a good view of, for instance, gloss characteristics. A short
training session was conducted before the actual session started.
The analysis of the market and trends took place at different levels of the
paper and printing industries.
All these actions have allowed the identification of the major problems.
The next step was the formulation of the question, which is at the same
time the aim of the thesis: the relevancy of a “key number” to objectively
quantify the printability in Flexography. The methodology of the research
is the explanation of the way “to answer the question”. The way can be
divided in several stages, which however can be carried out
simultaneously. The different stages mostly interact together.
The link between the different actions will be described in detail in the
next part, called research progression.
Each parameter has been the object of a detailed analysis. For the printing
parameters, the analysis has been held in two steps: It has been
distinguished between the variations within the manufacture tolerances
and the absolute variations. The first type of variations could be seen as
“done” variations and the user cannot influence them but must
nevertheless consider them. The second type of variations represents the
product range of the market for a special item.
For the paper characteristics, the way has been different. First, a
qualitative analysis has been realised, the goal for which was to identify
the primary and secondary characteristics. The primary characteristics are
usual and easy to measure characteristics. They influence directly the
printability (Paper III). The secondary characteristics are subjacent and
influence the printability only indirectly, by modifying the primary ones.
The second step of the analysis was a quantitative study.
The objective of both procedures was to look at the parameter’s relevancy
and influence in the problem solving approach.
The evaluation method for the printed samples has been adapted to the
type of parameters to be analysed. However, all the methods were based
on the same fundamental principle. The evaluation has been carried out
with the help of the quality factors. The differentiation occurred by
varying the nature and number of quality factors, like the mathematical
instrument.
The statistical instrument has help by the construction of the equation, the
different modelling and the establishment of the printability prediction
procedure.
The quality factors (QF) are the heart of the flow diagram (figure 1),
called printability system. The QF are the inputs of the equation to
calculate the “Printability Coefficient” in the measurement way of the
system and at the same time the “blind” intermediate calculators in the
printability prediction procedure.
They have been selected in accordance to their relevancy for the
flexographic process. The relevancy itself has been the object of a double
check. It indeed goes through the subjective judgement of expert and non-
expert panels and the objectivity of measurements and statistical analysis
(Paper V).
The quality factors retained for the establishment of the equation are the
following: density, dot deformation, dot gain, edge sharpness and
mottling.
Printing Trials
Printed samples
Modeling Modeling
of the Measurement of the
print tool substrate
Quality
Factors
”P”
”Printability Coefficient”
The measured quality factors and row data from the printed samples
allowed in a second step (Paper VI) to model the print and the substrate.
These modellings are used in the prediction procedure to jump over the
For the density and the dot gain, the use of a densitometer is synonymous
with a standard method. The densitometer measures the density and
calculates the total (geometrical + optical) dot gain with the help of the
Murray-Davies formula.
The mottling, edge sharpness and dot deformation quality factors do not
lay out of spread portable device in the paper or printing industry. To
quantify this factor the industry resorts with the image analysis
technology. This has also been the case in this project. More information
about image analysis can be found in the chapter 4.1.
To test the relevancy of the equation, another data series (52 different
combinations) specific for the flexography has been confronted to the best
two equations. The ranking given by the equations has been attentively
observed and analysed by experts. The analysis comported two parts: a
visual judgment of the quality and a study of the partial quality factors
rankings. The validation of the equations for non-absorbent substrate has
been successfully led. The equations were also retained for the next stage,
which is the elaboration of the prediction procedure.
5.2.9 Modelling
The objective of the modelling was the achievement of the second part of
the research. The measurement of the printability was the result of the
first part. The gain of a “Printability Coefficient” and consequently of an
objective and reproducible evaluation is a big step. However, the method
requires printing trials and measurement operations. This makes the
method time and cost expensive and only adapted for a production quality
control. The method is not adapted for decision-makers or comparison
studies. For these kinds of problems, the modelling has been the
instrument allowing the establishment of a prediction procedure.
The modelling of the hardness of the printing plate, the pressure at the
plate/substrate nip and of the anilox roller volume took place in two
stages. The first one was to look at the influence in percent of the primary
parameters on the QF. Then, the corresponding influence of the final
“Printability Coefficient” was calculated by transitivity. The variation
range was divided in unit variation and the modelling was done for this
unit. A scale was created to judge the degree of importance in the
modelling.
To be able to predict the printability without any printing trials (see flow
diagram), it was necessary to define a reference. This reference has been
calculated by giving the quality factors the mean values for both the
printing and paper parameters. The reference is called P average, noted
Pav and will not directly be used in the numeric interpretation. It is a
graphic instrument (fig. 2).
The Min and Max values for P will allow the user to roughly locate the
quality of the printing and paper parameter combinations she/he has
simulated and calculated the “Printability Coefficient” for. It is also a
control to detect eventual errors by calculating: the calculated values
included have to be between Pmin and Pmax !
The next and last step of the prediction is to make the calculation of the
prediction values for “P” (Pp). For the calculation, the user needs to know
or choose values for the printing parameters and paper characteristics he
is going to use. These values (xi ) are the input for the evaluation of the
deviation to the standard values (a i). The standard values are the values
used in the research work to get the modelling and establish Pav, the
constant factor in the Pp calculation.
The deviation obtained will then be divided by the unit coefficient (c i ) and
multiply by (fi ), the, in percent, influence coefficient. The (a i), (c i) and
(fi ) values are values resulting from the research (Paper VI). The user has
these values at his disposition, which are constant.
The sum of the i indexes of the results for the different (xi ) will be called
“p” (small p).
p = ∑ {[ (xi - ai ) / ci ] * fi }
The limits for P are now known: Pmin (0) and Pmax (4,8) theoretical values.
The centre of the variation interval will be considered as “at least to
achieve quality” (P g) for the prediction stage. The figure 2 is a proposition
for a graphic representation of the different Pp calculated. The
representation allows a rapid and easy interpretation of the results. The
centre of the figure is the “low quality” area and the contour the “high
quality” area. The geometry of the figure varies with the number of
simulations (Pp) plotted (in the represented case an heptagon for the 7
predictions). The points located in the area between Pg and Pmax are the
points researched. The Pp points are the points for the calculated (with the
1
5,00
7 2
0,00
6 3
5 4
Pp Pm Pav
Pmin Pmax Pg
The last step of the procedure takes place after the printing stage. To
personalize the Pp the user has the possibility to calculate a ∆P, which will
give him an idea of the accuracy of his prediction. The ∆P can be used to
adjust the fi to his production more precisely with an iterative procedure.
Thus, the user can better fit the technical parameters, the tolerances as
well as indirectly the skill of the personnel for his company.
∆P = Pp – P m
This part provides the reader with an overview of the content of the
different papers. It is a link between the full version and the references in
the preceding chapters. The papers are presented in chronological order,
which at the same time is the logical evolution of the research work.
The first paper refers to the study, which had the objective of looking at
the possibility of rationalizing the research work. The rationalization,
which simultaneously is a simplification of the problem by eliminating
one variable, had the “printing press” as its central object. The idea was
to study the correlation between different printing presses and simulation
or proofing devices. Should the correlation be good enough, the printing
press parameters should be indirectly eliminated. This should allow
printing on different presses during the project and still compare the
obtained results.
The analysis started with a review of the printing devices. The review
served as basis to regroup the presses in different categories: modelling
devices type 1, modelling devices type 2, production presses narrow web,
production presses wide web and lab presses. In a second stage, one press
of each category was tested. The tests were carried out keeping as many
parameters as possible constant. The printing plates, the mounting of the
plates, the ink, the different substrates, the theoretical volume of the
anilo x roller, the pressure at the plate/substrate nip and, as far as it was
possible, the printing speed were the same.
The evaluation has been made in two steps, where the first step was
determinant to decide on the continuation of the investigation. The first
step was based on density measurements on a test chart. The test element
comported 18 measuring fields covering the complete range from 1 to
100%. The analysis of the results, both with the conventional
mathematical instruments and with MDA method, revealed a good and
“acceptable” qualified correlation.
The printing press was the central object of the first paper. The analysis in
the second paper took into account the materials and peripheries of the
flexographic process. In the same way it was possible to speak about
tolerances in press manufacturing, the variations and tolerances in
materials have been analysed.
For the printing parameters, the analysis covered the tolerance in the
material manufacturing. The manufacturing tolerances are the tolerances
given by the supplier at material delivery and are not to be confused with
the variation of the parameters. It is indeed not the same to get different
results due to the use of two different materials and to get different results
due to tolerances in the same material.
The evaluation was done in two stages. The first stage was a control of
the tolerances given by the different suppliers. The second stage was to
choose materials at the extremities of the intervals and to obtain a
numeric evaluation of the projected tolerance intervals for parameters like
density or dot gain.
For the paper characteristics, the work was more critical. The paper
industry is more “reserved” to fix production tolerances. The range is
very large and the explanation to justify this fact is: “We are
manufacturing the paper on-demand!” Moreover, it was possible to
proceed to a qualitative analysis for each separate characteristic and to
obtain a quantitative evaluation of the global influence.
The tolerances are given and can normally not be reduced by the user.
The study has shown the non-negligible influence of very small variations
within the tolerances. It is also necessary not to forget that these
tolerances and their consequences can, in the worst case, be added. The
analysis resulted in a very strong control of the material for the rest of the
research work and the advice to the industry to proceed in the same way.
The idea of the third paper was to study more precisely the influence of
the paper characteristics on the printability. At this stage of the work, the
equation was not yet available, but the different quality factors were
known. The printing substrate, paper or board, is a complicated subject.
The multiple interactions between the different characteristics do not
make the study easy: it has also been decided to work on one parameter.
For practical and technical reasons, the retained parameter was the surface
roughness.
The influence was quantified and analysed separately for each quality
factor. The results show significant variations for the successive
calendaring levels. A better surface quality is, for all the quality factors
except for solid area mottling, synonymous with an improvement of the
quality factor measured and, consequently, of the general printability.
The results were presented in a numeric and as well a graphic form
resulting from a multivariate data analysis. Both are correlated.
The surface roughness was the fixed parameter in this study, but the other
paper characteristics, like density, absorption or porosity, have also been
evaluated and will serve to determine the contribution of the surface
roughness in the noted improvement of the printability.
Title: ”A Flex-Odyssey…”
The first question was: How to read the flow diagram? The diagram has
two entries. Depending from the utilisation calculation / prediction or
measurement, the level of entry is different. Whereas in the prediction
case, the inputs are the printing parameters and the paper characteristics,
in the measurement “modus” the inputs are the measured quality factors.
The output is in both cases a value of “P”; called respectively Pp and Pm .
Due to their roles in both procedures, the quality factors are qualified as
“heart of the flow diagram”.
The second question was: What is the contribution of each box in the
diagram? By contribution, it is to be understood both the content and the
functionality of the box. The content of the box described the nature and
properties of the elements contained in the box. The functionality of the
box is a definition of the interaction potential of the content of the box
with the other boxes of the flow diagram.
The work done in the first part of the research and reported in the four
preceding papers gave all the data and information to start the modelling.
In this paper, the establishment of different equations for a “Printability
Coefficient” are presented. The construction has been realised in
successive steps.
The first step was the collection and handling of the visual perception
data. The visual perception data served as a skeleton for the construction
of “P”. The goal was not to fit the visual perception values with a
complicated equation but to take into account the data and use them to
adjust the equation. It was indeed decided from the beginning to look at
linear combinations of quality factors. This decision was based on two
factors.
After the handling of the data, several equations were tested for all the
printing processes. The equation was progressively improved to
simultaneously respect the profile of the visual perception values for all
the processes and reduce the distance between the objective and
subjective values. During this stage of the work, by eliminating the
certain negligible quality factors, a simplified and adapted equation for
the flexographic process was established.
The stage reached corresponds to the bottom of the flow diagram. The
results presented prior to this paper allow the calculation of a “Printability
Coeffic ient”. The result however, is not satisfying for an industry where
the time and cost aspects play a central role. The industry needs an
instrument to save time and only a prediction of the printability without
any trials can answer the demand.
The last stage of the research project was thus devoted to the development
of a prediction procedure for the printability. The previously established
and verified equations for the printability were the fundament of the
work. With the help of the equations, the influence of the quality factors
on the “Printability Coefficient” were analysed and quantified. The
quantification was characterized in percentage of a “P unit”.
The next step was to look at the influence of the primary parameters, both
printing parameters and paper characteristics variations onto the quality
factors variations. The calculation was done for the total variation range
of the different parameters and then reduced to a “united result” in
percentage. With the two precedent results and the definition of units (c i),
it was also possible to calculate first the partial and then the global
influence, in percentage of “P units”, of each primary parameters onto the
final “Printability Coefficient”. The result of the calculation is called unit
influence coefficient and indicated with fi .
The discussion part of the paper should answer the question “How to use
the Printability Coefficient?”
p = ∑ {[ (xi - ai ) / ci ] * fi }
The research carried out and the experience acquired during the
management of the project have led to results of different natures. The
results should profit to both the research field and the printing and paper
industries. Some results are solutions to a well-known problem and can be
applied as presented; other results are ideas or partial solutions, which
must be adapted before the implementation.
The technical results are principally the results obtained in the first part of
the project where the variables have been tested separately. They are from
different natures: results for the project itself or for further research
works, results for the flexographic industry and results for the final users.
These results can be seen as the contribution of the research work to the
standardisation of the flexographic process. They are at the same time
global and final results with a direct application in the paper and printing
industrial worlds.
8. Discussion
The second adjustment is the choice of the printing parameters and paper
characteristics. Of course the choice is the consequence of several
interviews and a scientific analysis of the parameters, but it could appear
that, for a specific production, one or more parameters have to be added
or removed. The structure of the results with the delivery of flexible
procedures allows the adaptation of the results. In this case, the research
work final results are a support for the final user to develop his own
procedure.
It is also probably the first major result and innovation of this work: a
new approach to quality control, more global, more communicative and
more final user and industry-oriented.
The relevancy of the work for the industry is justified by responding the
requirements of modern life:
The work done has shown the way to be followed and should be
completed by a systematic application of the procedures for different
productions. In this way, the different needed adjustments will “polish”
the procedures. In a next step, the problem can also be extended to the
flexible packaging industry, where the flexography is very present.
9. Conclusions
At this stage, it is time to give the future users the last recommendations.
Before printing, the users dispose of two equations to calculate the
predicted printability (Pp ). Depending of the precision and goal of the
calculation, the user will choose the influence coefficient (f i)
corresponding to P10 for a global approach or to P9 for a more detailed
one. After printing, the user will be able to measure the printability (Pm )
by measuring the quality factors. The next step will be, with the help of Pp
and Pm , the calculation of ∆P. The comparison between the ∆P value and
the predefined tolerance will help the user to personalize his printability
coefficient and achieve standards.
Finally, I would like to thank particularly my near family for all kinds of
support. They were close to me during the last year for Tanja, the last
years for my brother Quentin and all my long adventure until this 7th of
June 2002 for my parents Marc and Marie-Claude Leloup.
Bichard, W. (1992)
“The Inter-relationship among Air-Leak Roughness / Smoothness
Methods” - Pulp and Paper Canada, N. 93(6), pp. 147-152
Paper encyclopedia
http://graphics.tech.uh.edu/Costing/Paperencyclopedia.pdf
Plowman, N. (1991)
“Ink gloss mottle, Graphic Arts Monthly” - vol 63, N.5, pp.122-125
Paper II
Paper III
Paper IV
Paper V
Paper VI
Abstract:
The flexographic process is from the first approach based on a simple ink
transfer concept but if you look at it in more detail you discover a
multitude of variable parameters, which influence this ink transfer and
also the quality of the final printed result. Due to economic demands and
ever decreasing delivery times, it is impossible to run a job on a
production press for use as proof. Therefore the industry is looking to
develop systems with the ability to refine the flexographic process or at
least a part of the process. The goal of the project was to analyse the
correlation between production printing presses and these modelling
systems.
The MacMillan Bloedel Print Indicator & Sizing Tester was developed to
provide a rapid indication of print quality and sizing characteristics.
Using an analytical pumping system, the device dispenses microliter-
sized drops of ink onto a mylar foil. The drops are then drawn onto the
paper sample and down its length by a motorized blade at controlled
speed and pressure. The resulting printing length is directly related to ink
receptivity and surface topography.
The Flexo Proofer F.P. 100/300 is a reel to reel device which has been
developed by the Saueressig company. The substrate is stocked from the
top onto the presseur and stick on the roll for printed material. The plate
cylinder and anilox roller move hydraulically to the arrested presseur.
When the doctor blade comes into position, the ink will be provided with
a pipette and the print operation is started.
A testform with different types of elements was elaborated for the series
of trials. This testform should make it possible to establish the correlation
between the different presses by measuring the density and the dot gain at
different places and within the range of 1% to 100% by 18 steps. Circular
dots. a resolution of 2540 dpi and a 34 l/cm screen were selected so as to
reflect industrial reality. The testform is also quite particular in that it
allows study of the influence of the printing direction. Moreover, it
consists of other elements such as bar codes, a large solid area, a large
50% field and negative and positive text which will be used for further
investigations and to look respectively at the edge sharpness, the uniform
coverage and the cleanliness of the print.
The data acquisition was carried out using a Gretag D19C densitometer
and using the KeyWizard data collection software to import the values
into calculation tables.
A discussion with experts from the SCA and StoraEnso paper industry
companies led to the conclusion that to make the work credible it was
necessary to print on five different paper qualities (Table I).
Paper quality
Name: Type:
DPS Digital
"universal plate"
HOF Conventional
"flexible plate" (exposure latitude and image resolution)
HOS Conventional
clean image relief / high quality process printing
TDR Conventional
the "corrugated board plate"
Both the conventional and digital plate making workflow have been
considered. 2 different thicknesses have been tested and all the plates
presented different types of polymer, exposure time and hardness.
The values in the table correspond to our measurements and could differ
from the standard values given by the supplier. 2 plates have been used
for the HOF type therefore 2 values for the min. and max. thickness. The
plates were mounted using a 0.20 mm PVC tape for the 2.84 mm plates
and a 0.38 mm PE foam for the 1.70 mm plate.
The ink was the same for all the trials and was a cyan commercial water-
based ink. The ink was printed with a 28-30s viscosity (Frikmar cup
4mm).
For the IGT F1, it was not possible to keep the same parameters. The
speed was 0.3 m/s (18 m/min). This choice results from the study of
anterior research, which have shown that by printing with speeds above
0.6 m/s, the print density decreases and the coefficient of variation
increases. The pressures at the plate-substrate nip were 65 and 120 N
which, once established after several tests, should correspond to 100 and
200 µm respectively over Kiss Print. Moreover, after checking the
repeatability of the device, the decision was taken to print only 2 “good
samples”. The washing time was about 5 minutes between each sample,
corresponding to 190 hours for 25 samples! Lindström, Dölling and
Poustis looked at the repeatability in their work and came to the
conclusion that the repeatability expressed as the coefficient of variation
of print density was about 1%.
The Saueressig Flexoproofer was run at 50 m/min and the plates were
mounted diagonally to avoid the significant vibrations observed in the
normal configuration.
5. Results
A study of the diagrams shows that all the curves have the same profile
with the same ranking for the different presses regarding the ink transfer:
the DPS plate transfers the least ink, then the HOS and TDR have almost
the same comportment and finally the HOF is characterised by a higher
ink transfer.
Density IGT-SEU-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Density DFTA-SEU-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Density IMT-SEU-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Only the IGT F1 presents an inversion of the ranking: the HOS and HOF
plates show an ink transfer of about 35% less than expected. The reason
should be the following: the HOF and HOS plates both have a higher
degree of hardness and 65N represents in this case less than 100 µm due
to the higher resistance of the polymer against the action.
The fact that all the curves have the same profile enables us to look
closely at the influence of the presses for one type of plate and all the
different substrates. The DPS plate has been chosen to purchase this
study.
The next series of diagrams (figure 6-10) the represents the density
variations measured for each substrate with the printing press as a
variable parameter. The pressure at the plate-substrate nip was 100 µm
and the volume of the anilox roller 8 cm3/m2. The influence of the anilox
roller and of the pressure at the plate-substrate nip will be considered
separately shortly.
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Density SEU-DPS-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Density PK-DPS-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Density WT1-DPS-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
Density DPS-IMT-8-100
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
-0,40
% Film
Density WT1-DPS-8-200
2,80
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
This analysis proves that when taking care of certain limiting parameters
like width, speed or stability of the devices it is really possible to
determinate a correlation between the different printing presses and
modelling devices. The fluctuations measured stay in an acceptable area.
Only the coated quality was outside the tolerances. Of course the study of
the density variations have to be supplemented with further analysis to
quantify specific surface defects and edge sharpness but this research
would mean further investigation using multivariate statistical analysis
and image analysis.
References:
Chalmers, I.R.
“Flexographic printability of packaging grade papers” – Appita’97
I. The materials:
1. Printing plate:
Regarding the hardness you can measure two different values: the
hardness of the raw plate or the hardness after processing. The hardness
of the raw plate is not really relevant for the printer, therefore it is usually
not controlled. The plate hardness after processing influences the ink
Plate 1A Plate 1BPlate 1C Plate 2A Plate 2B Plate 2CPlate 3APlate 3B Plate 3CPlate 4APlate 4BPlate 4C
Position1 2,82 2,819 2,818 2,857 2,862 2,862 2,884 2,881 2,852 1,717 1,704 1,71
Position2 2,813 2,813 2,81 2,862 2,855 2,861 2,879 2,873 2,861 1,717 1,705 1,708
Position3 2,813 2,817 2,811 2,855 2,853 2,863 2,879 2,886 2,86 1,715 1,701 1,707
Position4 2,842 2,81 2,819 2,857 2,857 2,858 2,877 2,874 2,872 1,715 1,704 1,711
Position5 2,819 2,797 2,84 2,854 2,854 2,856 2,883 2,869 2,872 1,714 1,705 1,723
Position6 2,811 2,826 2,821 2,867 2,862 2,861 2,884 2,873 2,873 1,714 1,705 1,719
Position8 2,822 2,832 2,824 2,86 2,859 2,851 2,887 2,883 2,866 1,716 1,707 1,719
Position9 2,826 2,837 2,835 2,868 2,872 2,856 2,881 2,885 2,868 1,714 1,708 1,723
Average(plate) 2,821 2,819 2,822 2,860 2,859 2,859 2,882 2,878 2,866 1,715 1,705 1,715
V- -0,01 -0,022 -0,012 -0,006 -0,006 -0,008 -0,005 -0,009 -0,014 -0,001 -0,004 -0,005 -0,009
V+ 0,021 0,016 0,018 0,008 0,013 0,004 0,005 0,008 0,006 0,002 0,003 0,008 0,009
Average(box) 2,821 2,859 2,875 1,712
V-- -0,024 -0,008 -0,023 -0,011 -0,014
V++ 0,021 0,013 0,012 0,013 0,012
Average(charge) 2,852
V--- -0,045
V+++ 0,035
100
80
2,840 mm
60
% Print
2,825 mm
40
20
0
0 20 40 60 80 100
% Film
80
60
° Shore A
40
20
0
1 2 3 4
Type of plate
The study of the dot gain curves (figure 4) confirms the importance of the
Shore A hardness for the ink transfer. The difference measured from 10.8
% in average between the 20 and 60 % tones is very demonstrative. Of
course the large amplitude is due to the choice of the plates but in the case
of 4° Shore A tolerances, 4-5% difference in dot gain can be expected.
100
80
45° Shore A
60
% Print
40
37° Shore A
20
0
0 20 40 60 80 100
% Film
The tape or foam tape is one of the elements of the “tolerances sandwich”.
For tape thickness under 0.20 mm the small variations are not to be taken
into consideration. For tape over 0.20 mm and foams it is not possible to
neglect the tolerances. A study at the DFTA-TZ has shown that for 0.20 /
0.30 mm tapes the tolerances are about 0.020 mm and could reach 0.035
mm for 0.55 mm foam tapes. The deviations to the nominal value are in
most cases positive. For the foam tapes the dimensional stability to the
temperature, the memory of the foam and its compressibility are the most
important factors which condition its thickness during production. The
tape is sometimes considered as a secondary element but in reality its
choice and good usage are very important: choose the right tape with the
right plate and motive to print, control its thickness and mount all the
plates of a job with the same roll could avoid expensive problems.
The goal of the investigation was to look at the influence of the variation
in the viscosity in a range which corresponds to viscosity measurement
uncertainty. When measuring the viscosity manual with a cup, 2 factors
influence the measurement: the state of the cup (wear, clean/dirty) and the
person stopping the time. These factors lead to +/- 2s variations.
Moreover 4s is a realistic value for a maximum variation between two
manual controls of the viscosity during the production. The diagram
(figure 5) represents the density variations measured for 23, 25 and 27s
viscosity for a water-based ink. The curves are very similar and no
relevant density variation can be noted.
Tolerances in viscosity
2,2
2,0
1,8
1,6
1,4
Dednsity
1,2
1,0
0,8
0,6
0,4
0,2
0,0
0 20 40 60 80 100
% film
27 25 23
100
80
% Print
60
40
20
0
0 20 40 60 80 100
% Film
4. Sleeve:
The number of printers using sleeve technology is growing very fast. The
main advantage of this technology is its flexibility. With the same
cylinder you are able to print different repeat lengths, to use different
plates and tape thickness. Moreover the make ready time of the press is
considerably reduced.
What about the tolerances? Do we have the same precision we had with
cylinders? Delivery certificates guarantee a circularity of 0.025 mm and
tolerances for the external diameter of +/- 0.020 mm. These tolerances are
independent of the diameter and length of the sleeve. The 2 sleeves tested
(figure 6) presented a difference of 0.013 mm in diameter and the average
variation in dot gain measured is 2,2 % for the 20-60% part of the curve.
This demonstrates one more time the importance of control by material
delivery. It does not mean you have to refuse sleeves if you measure one
with +0.020 and an other one with –0.020mm but do not use it for the
same job in the press!
For this study 10 anilox rollers from different manufacturers have been
considered. 3 different theoretical volumes have been selected: 8, 10 and
12 cm3/m2. In the following table (figure 7) the results of the
measurements are presented. For each anilox roller, the volumes that were
ordered, delivered (values mentioned on the delivery protocol by the
supplier) and controlled have been compared. Different methods exist for
measuring anilox roller volume. The controlled volumes have been
measured with a confocal laser-scanning microscope, an expensive but
very accurate method. The difference between the ordered and delivered
volumes is on average 2,8 % which means +/- 0,28 cm3/m2 for the 10
cm3/m2 anilox roller. This value related to the other parameters of
influence during the ink transfer is insignificant. However the redaction
of the protocol sometimes seems not to be credible: 8.0 and 12.0 ordered,
8.0 and 12.0 delivered! Therefore the control method with the confocal
laser-scanning microscope. Then the variations increase to 9.4% on
average. This means +/- 0.94 cm3/m2 for the same 10 cm3/m2 anilox
roller. These tolerances are now to be taken into consideration and could
be amplified by inadequate cleaning of the anilox roller and wear.
12,0
10,0
cm3/m2
8,0
6,0
4,0
2,0
0,0
1 2 3 4 5 6 7 8 9 10
Anilox roller N.
Ordered Delivered Controlled
Figure 7: Anilox roller: ordered, delivered, controled theoretical
volumes
1. surface roughness:
2. surface tension:
The surface tension determines the wetting of the paper surface by the
ink. Wetting is one of the basic requirements for ink absorption and ink
transfer. The contact angle method and Hercules size test are standard
methods to quantify the wetting of papers and boards and the results given
in mN/m is often called surface energy. When printing on paper or board,
using water-based ink, the problem is that the range of the paper surface
energy is almost the same as the range of the equilibrium surface tension
of the inks, which is from 20 to 40 mN/m. From this fact results an
inhomogeneous wetting leading to a print mottle. Wilhelmy and Bristow
tests are other methods to measure the surface energy but they are not
suitable for predicting the print mottle (Zang). A study has shown that it
would probably be more relevant to adjust the dynamic surface tension of
3. water absorbency:
When printing with water-based ink on “non closed” substrates, the water
absorbency is a determinant parameter for the printability. But at the same
time the interaction with the ink is complex and needs a very prudent
approach. The water absorbency will give an indication of the ink
consumption but the relation between the ink consumption and the print
density is not linear. The reason is that the quantity of ink offered by the
anilox roller is limited: if this amount of ink is superior to the absorbency
of the substrate, then the density will increase with the absorbency, but if
the amount of ink is inferior, then a low absorbency substrate will show
the highest print density (the other characteristics of the 2 substrates are
supposed to be similar).
IGT, Cobb test, Unger (g/m2), K&N (%), Tainionkoski (%), STFI (g/m2)
are other procedures for measuring the water absorbency.
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
2,40
2,00
1,60
1,20
0,80
0,40
-
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% Film
This study attests the complexity of the high quality flexographic process.
It confirms the necessity for a standardisation of the process and the
necessity for the development of a tool to be able to measure and predict
the printability when varying the different parameters. The analysis of the
properties of various elements has shown that variations within the
material tolerances influence the printed results. If you now add these
tolerances or use materials with properties outside the tolerances you can
imagine the consequences of the print quality. Therefore the choice of
materials with the right properties and tolerances and careful checking of
the delivered materials should be the start of all high quality flexographic
printing production.
References:
Jensen, K.W.
”Flexo printability of Coated White-Top Liner with Consideration of
Convertability”
Graphic Arts in Finland 1989
Introduction:
Methods:
The next step of the work was to print on the paper. The printing trials
were carried out on the Flexocompact flexographic printing press at
framkom (2). The press is a 2-colour multicylinder press equipped with
chamber doctor blades and a warm air dryer. The press is seven years old
and has shown its repeatability in precedent research works. The speed
was maintained constant at 120m/min during all the trials. The ink was a
to the substrate adapted water based ink, cyan, printed with a viscosity of
27s (Frikmar 4mm). 2 different printing plates have been used both with a
1.70mm thickness: a digital universal plate with the DPS designation
(Hardness 56° Shore A) and a conventional plate with the designation
HOS (Hardness 71° Shore A). The anilox rollers were selected to cover a
large range of transferred ink volumes; from 3 to 12 cm3/m2. All of them
were laser engraved ceramic anilox rollers, engraved with the multi pulse
technology and had a 45° screen angle. Moreover the study was used to
test how a fine screening affect the printability: this has been done by
choosing for the 4,5 cm3/m2 volume 2 different screen rulings: 200 and
500 l/cm.
The following 2 measurements, bleeding and wic king had for goal to
quantify an important quality factor for the flexographic process: the edge
sharpness. The bleeding procedure has been used to measure the
sharpness of negative text and the wicking the sharpness of positive text.
Both procedures are based on the same measurement principle and return
2 values as results: the surface area and the calculated sharpness in mm2.
The surface area is the difference between the area of the scanned sample,
calculated with a method based on the comparison of pixe ls, and the area
of the reference sample, entered by the user. The sharpness is the surface
difference between the surface area and the ideal outline of the object
measured. The dot gain measurement was in this study used as test for
further development. The procedure used (based on the sharpness
measurement) differs from the conventional methods but the results let
expect a future for the method.
The last part of the methodology I would like to mention here is the use of
the SIMCA P8 multivariate data analysis software for the exploitation of
the results. The Principal Component Analysis (PCA) is a multivariate
projection method that is designed to extract and highlight the systematic
variation in a multivariate data matrix and the eventual outliers to the
considered model. This analysis has to be seen like a control and
complementary analysis to the conventional statistical analysis presented
in the results part of the paper and will be used for the discussion.
2,40
2,30
2,20
2,10
Density 2,00
1,90
1,80
1,70
1,60
1,50
1 2 3 4
Serie
2,30
2,20
2,10
2,00
Density
1,90
1,80
1,70
1,60
1,50
1 2 3 4
Serie
The figure 1 and 2 present the density variations observed for the plate
HOS with the different papers. Series 1 and 2 are the results with a low
ink volume and 3 and 4 with an high volume. For the white top qualities,
which have not been calendered differences are constated at low nip
pressure but disappear at high pressure. This result confirms the existence
of 2 roughness levels and in this case the “macro-roughness” of the both
papper is the same but it is probably a difference in the“micro-roughness”
which make the difference at low pressure. At high pressure the micro-
roughness is then also compensated.
For the calendered qualitites, the edge reels show a regular behaviour for
all the series but more pronounced at low nip pressure. The density is
increasing at the same time the Bendtsen values are decreasing. The same
2,00
1,50
Density
1,00
0,50
0,00
0 10 20 30 40 50 60 70 80 90 100
% Film
WT1E-HOS08200 WT1M-HOS08200 SEUE-HOS08200 SEUE1HOS08200
SEUE2HOS08200 SEUM-HOS08200 SEUM1HOS08200
Mottling 100%
13,00
12,80
12,60
12,40
12,20
Mottling index
12,00
11,80
11,60
11,40
11,20
11,00
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
The mottling index values for the 30% screen area are varying between
8.72 and 9.78 (figure 5), which corresponds exactly to the same range
(1.06 / 1.05) as for the solid area. It is however important to note 2
particularities.
- The first one is that the mottling index is in average 4.5 less than for the
solid area: the explanation is that the solid area mottling is at the same
time influenced by the uniform coverage and at low nip pressure the
mottling index measured is sometime accentuated by the non uniform
coverage and also the residual substrate grey value.
- The second particularity is that the variations are opposite: the
improvement of the surface roughness has for consequence a decreasing
of the mottling index. Interesting is the fact that at same Bendsten the
difference between the minimum and the maximum (8.72 – 9.37) is one
more time 0.65.
10,00
9,80
9,60
9,40
Mottling index
9,20
9,00
8,80
8,60
8,40
8,20
8,00
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
1,05
1,00
0,95
%
0,90
0,85
0,80
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
Bleeding
6,00
5,00
4,00
3,00
%
2,00
1,00
0,00
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
"Absolute" Sharpness
6,00
5,00
4,00
%
3,00
2,00
1,00
0,00
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
The dot gain has been measured following 2 methods. The first method is
the conventional one using a densitometer and the Murray-Davies formula
with the full tone and 50% densities as inputs: the result is an optical dot
gain.
The second method was tested during this study and used the wicking
function of the Papeye software. In this case the result is a geometrical
dot gain: a one 1cm2 50% screen area with a 34 l/cm screen ruling has
been used to calculate the total area of the points. The accuracy of the
method can be justified by the fact that each time 34x34=1156 points are
measured and the final value is the average (figure9).
35,00
32,50
30,00
% Print
27,50
25,00
22,50
20,00
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
35,00
32,50
30,00
27,50
25,00
22,50
20,00
%
17,50
15,00
12,50
10,00
7,50
5,00
2,50
0,00
WT1E-Moy WT1M-Moy SEUE-Moy SEUE1-Moy SEUE2-Moy SEUM-Moy SEUM1-Moy
Serie1 Serie2
Figure 10: Optical / Geometrical Dot gain for the different paper qualities
WT1E-
1
WT1M- SEUM
SEUE-
t[2]
0
SEUM1
-1 SEUE1
SEUE2
-2
-3
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
t[1]
Scores: t[1]/t[2]
Ellipse: Hotelling T2 (0.05)
Eng 50%
0.60
0.40
% W Sharp
0.20
p[2
]
0.00
-0.20
Mottling Apl
-0.50 -0.40 -0. 30 -0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50
p[1]
Loadings: p[1]/p[2]
WT1E-
SEUM
SEUE-
SEUM1
t[2]
0
WT1M- SEUE1
-1
SEUE2
-2
-3
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
t[1]
Scores: t[1]/t[2]
Ellipse: Hotelling T2 (0.05)
0.80
0.70
0.60
0.50
0.30
Trame 50%
0.20
0.10
0.00
-0.10
Mottling 30%
p[1]
Loadings: p[1]/p[2]
2
WT1E-
SEUM
t[2 SEUM1
] 0
SEUE- WT1M-
SEUE1
-1 SEUE2
-2
-3
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
t[1]
Scores: t[1]/t[2]
Ellipse: Hotelling T2 (0.05)
0.70
0.60
0.50
0.40
Eng 50%
0.30
p[2]
0.20
0.10
Mottling Apl
0.00
Trame 2%
-0.10
-0.20
-0.30
Mottling 30%
-0.40
-0.40 -0.30 -0.20 -0.10 0.00 0.1 0 0.20 0.30 0.40 0.50
p[1]
Loadings: p[1]/p[2]
Acknowledgements:
References:
Aspler, J., Byron, J.,Zang, H. and Nguyen, N., (1998), Print Quality of
Linerboard in Commercial Water-based Flexography, TAGA 98
proceedings, 749-774
Ginman, R. and Tiainen P.-R., (1988), Requirements of the flexographic
printing process on wrapping paper and packaging board, Adv. Print. Sci.
Techno, 19:210
Jensen, K.W., (1989), Flexo Printability of Coated White-Top Liner with
Consideration of Convertability, Graphic Arts in Finland 18:14-20
Plowman, N., (1991), Ink gloss mottle, Graphic Arts Monthly, vol 63,
N.5:122
Bichard, W. (1992), The Inter-relationship among Air-Leak Roughness /
Smoothness Methods, Pulp and Paper Canada, 93(6):147
Appendix:
1. Introduction
The Flexography started 50 years ago a long travel. The different stages
and the final destination of this Odyssey were not easy to predict. Today
it looks to be evident that this Odyssey has for objective the “total
quality” concept. The flexographic process is reaching its maturity and
competitiveness, quality, innovation, flexibility are the words qualifying
the flexography. This explains probably why the flexography is actually
the only one conventional printing process still growing. The objective of
this work is to predict and measure print quality in flexography, before
the actual job is printed. This quantification should take into
consideration the interaction between the substrate characteristics and the
printing parameters. Therefore it is preferable to speak of measurement of
the “PRINTABILITY”. The method should be objective by eliminating
the human judgement and the environmental factors, which currently lead
to an interpretation and not to a scientific measurement. To achieve these
requirements of simplicity and objectivity, the help of a “key number” is
needed. In the investigation the credibility of such a number called
“Printability Coefficient” will be studied.
_________________
*Research Corporation Media and Communication Technology
(Framkom), Royal Institute of Technology (KTH), Stockholm, Sweden
The “Printability Coefficient”, “P” is the final number resulting from the
input of measured quality quantificators like density, mottling, edge
sharpness cleanliness of solid area, dot gain and dot deformation. The
gain of the equation results from different steps in the project. The
achieved first steps of the research work allow today the introduction of a
flow diagram for the method. Moreover almost all the parts of the
diagram can now be described in details.
In this paper the results will be introduced in 2 steps: the first step is a
general explanation of the diagram “How to read it” and the second step
will be look at the contribution of the different “boxes”.
3. Results
Principle
Printing Paper
parameters characteristics
Printing Trials
Printed samples
Modeling Modeling
of the Measurement of the
Print Tool Substrate
Quality factors
”Printability coefficient”
2001 A FLEX-
FLEX-ODYSSEY
Figure 1: “Printability Coefficient”: The Flow Diagram
The flow diagram should help the technical people both from the paper
and the printing industries to take the right decision. The right decision is
a consequence of different requirements, which interact together. The
final product and its functionality have to satisfy the technical and
financial aspect of the project. It means criteria such as transport,
marketing have to be considered in the choice of the “Printability
Coefficient”: the quality has a price, the technique has limits, “P” should
help to find the best compromise!
The next step is to achieve the “make ready” press, to get the “first sheet”
and to compare the obtained result with the expected or required result. At
this stage appear different problems. The first one and probably the most
important is the objectivity of the judgement during the quality control.
Why is it a problem? There is actually no standard to measure the “total
quality” but only different partial qualities like density measurement, dot
gain calculation, measurement, of spectral values… This has for
consequence the difficulty to compare different “sheets” in view of “total
quality”. To solve this problem the method requires the measurement of
standard and process adapted quality factors. The quality factors are then
the input values for an equation, which will give as result a “Printability
Coefficient”.
. The design of the product: today the packaging industry does not
anymore work only with the transport function of the packaging but
as well with the practical aspect and the marketing strategy of the
final customer. The substrate must as the same time respond to the
mechanical, chemical criteria and the “buy me” criteria. The first one
b. the press:
c. the ink:
. The final product function and the substrate lead to the type
of ink, ink system to use. Food and cosmetic industry for
example will have very high exigencies that will not be the
same for wrapping paper or trivial packaging solutions. In a
second phase the runnability will play a major role: the
viscosity, the pH and the temperature of the ink have to be
This box in the flow diagram represents the output box for the first 2
boxes described above: the collected samples are a materialisation of the
combinations obtained by the materials and peripheries selection. At the
same time the printed samples are an input for the next box: the
measurement tool. The printing trials deliver printed samples with test
charts / elements, which will be used for the evaluation of the different
quality factors.
A large study has allowed to find out the relevant quality factors for the
printability in flexography. Two test forms with 13 test elements have
been printed and evaluated. The evaluation has been made at 2 levels: a
called “subjective evaluation” carried out by experts and non-experts
“P” is THE final result, the output and at the same time centre of the flow
diagram. The “Printability Coefficient” should be seen as an instrument
for evaluating or previewing (in this case “P” is simultaneously playing
an output and input function) the printability in flexography. It is a
number without unit to resume a lot of parameters, which interact
together. A high value of “P” corresponds to a high printability. The user
will define for is company the acceptable ∆P and be able to control his
production.
1,50
1,25 1,30
1,24
1,22 1,24
1,00 1,04 1,16
0,99 1,05
1,01
0,86 0,89 0,93
Printability
0,75
0,89
0,79 0,86
0,50
0,25
0,00
S4
S3
-0,25
1 2 S2
3 4 5 6 7
8 9 S1
10 11
12 13 14
15
1,40
1,20
1,00
P units
0,80
0,60
0,40
0,20
0,00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
5. Conclusion
Abstract: The Flexographic Process has the last years noted a great
improvement of the achievable quality. The quote part of the market for
the flexographic printing process is still growing; working groups meet to
achieve standards for this process at an international level. The industrial
partners: paper manufacturers, peripheries suppliers, printers and
converters are today looking for a common language to be able to judge,
measure the print quality. The following paper will propose solutions for
a universal language with the help of a “Printability Coefficient” for the
Flexographic Process. The results of the research present different
approaches for a mathematical modelling of the parameters influencing
the printability and their interactions.
The present work can be divided in different steps. Theses steps are the
following: preparation of the printing trials, printing of the samples,
measurement of the printability parameters, establishment and
comparison of different equations for a printability coefficient. The first
3 steps are very time expensive in both the elaboration and realisation but
are very important to achieve a valuable result in the final stage. The data
collection (steps 1-3) results partially from the work achieved in preview
research (1)(2) and (3).
The printing trials have been done on full-scale printing presses. These
presses are located for the flexographic process at research centres in
Germany (DFTA-TZ) and in Sweden (Framkom). The LEMO flexopress
is a six-year-old six-colour CI-Press, with a width of 1300 mm and
equipped with chambered doctor blades, CNC motors for the adjustment
of the nip pressures and an automatic regulation of ink viscosity. The
Aquaflex fle xopress is a one-year-old, five-colour production stack press,
with a 18” width. In the difference to the 2 others presses run with water
based ink, the Aquaflex press was run with UV-ink. The third press, a
Flexocompact eight-year-old, two-colour press with a 600 mm width is a
modified production press used for research work. The capability and
repeatability of the press have been tested in the past and show very good
results (1).
The printing parameters have been adapted to the substrate. For the
qualities 1-7, the printing plates are 1.70 mm DPS and HOS plates with
anilox rolle rs offering volumes of 8 and 12 cm3/m2. For the qualities 8-
11, the printing plate is a 1.14mm DPN plate with an anilox roller volume
measured by 3-5 cm3/m2. The different substrates allowed at the same
time to test different plates both conventional and digital and anilox
rollers from high volumes to very fine gravure and less volume.
Not only the substrate and printing parameters are variables in the
investigation. 2 different printing procedures have been chosen to validate
the printability coefficient. The first procedure (substrate 8-11) is an
optimisation of the result. This optimisation has for goal to allow a
correlation between the technical measurements and the visual perception.
The printing trials have been effectuated in 3 steps: the printing of a test
form for the realisation of a colour management profile, the gravure of
new plates with the profiles (UV and water based ink generated 2
different profiles) and the final printing by running the same densities.
- For the qualities 1-7: densities in 2%, 50% ton value and
in the solid area, edge sharpness divided in wicking and
bleeding components, dot gain by a densitometric (for the
50 %) and optical way (for the 2, 30 and 50%), mottling
in 30% screen and in the solid area, dot quality
(roundness) for 2 and 30% dots.
- For the qualities 8-11: densities CMYK in the solid area,
edge sharpness divided in blurriness and raggedness
components, dot gain by a densitometric way in 40 and
80% for CMYK, mottling for C and K as for Red
(R=100%M+100%Y), Green (G=100%C+100%Y) and
40% K, gloss, colour gamut, colour failure and grey
balance.
The first stage of the work is to test different equations for a printability
coefficient and to look at the correlation with the visual perception data
for all the printing processes. In a second step the equation will be
optimise for the flexographic process. This optimisation has for goals to
get a better correlation with the visual perception data and to simplify the
equation. The simplification will put in relief the components of first
priority, the printability parameters specific for the flexographic printing
3. Results
Dot Gain
160,00
140,00 134,67
131,41
MAX and MIN in % to mean value
121,98
119,78 121,33 120,20
120,00
110,26
102,02
98,95 100,00 100,00 100,00
94,87 96,75
100,00 93,17
90,58
85,42
82,25
80,00
61,02
60,00
48,16
36,88
37,34
40,00
20,00
0,00
DG40C DG40M DG40Y DG40K DG80C DG80M DG80Y DG80K DG2C DG30C DG50C
Figure 2: Amplitude of the Dot Gain variation for the different substrates
Density
120,00
100,00
MAX and MIN in % to mean value
100,00
90,34
80,00
60,00 54,76
0,00
C M Y K C2% C50% C100%
The results for the density are as expected: about the same variations in
CMYK, for the minimums and larger variation for the maximums due to
the inkjet process and its very high densities. C 2% with values around
100% can be explained with the small nominal value.
437,68
450,00
400,00
MAX and MIN in % to mean value
350,00
300,00
250,00
200,00
150,00
119,15
88,25
100,00
72,52
47,84
50,00 30,29 29,23
20,15
0,00
Raggedness Blurriness Bleeding Wicking
The edge sharpness is a typical quality characteristic for the inkjet and
flexographic processes. Raggedness, blurriness and wicking are
quantifications of the quality for positive lines and bleeding for negative
lines. One more time the extreme values are the maximum values. The
over 400% in bleeding correspond to a for low nip pressure. The
minimum values are between 20.15 and 72.52%.
Mottling
488,01
500
450 423,86
MAX and MIN in % to mean value
400
350
300
236,03
250 223,48
200
160,70
150
100 84,67
71,65 71,05 76,03 67,38
50
13,53 14,27 17,63 15,07
0
C K R G K40% C C30%
985,55
1000,00
900,00
808,86
MAX and MIN in % to mean value
773,13
800,00
700,00
600,00
500,00
400,00
300,00
200,00 153,70
98,47 100,00
75,12
100,00 45,43 50,98 52,17
18,34 16,69
0,00
Roundness 2% Roundness 30% Gloss Colour Gamut Colour Failure Grey Balance
To measure the roundness is a method to look at the dot quality (the ideal
circular point has a 1.00 roundness). This measurement gives together
with the dot gain a very good characterisation of the dots. The other
parameters presented in this diagram are more relevant to compare the
different printing processes than to judge each one separate. Grey balance
notices very surprising results: the use of ICC profile has been done for
the trials!
P1 - VP
1,40
1,20
1,00
Printability
0,80
0,60
0,40
0,20
0,00
0M
H
4H
4H
4H
0M
0M
2H
6U
1U
H
4H
4H
6U
4H
3H
H
6U
6U
I05
04
I08
14
K0
K1
V0
X0
X1
K1
V1
X1
I1
O0
O0
C1
C1
O0
K1
C1
P1 VP
with α i = +/ - 1
+1 for mottling, blurriness, raggedness, dot gain, colour failure,
grey balance
-1 for density, colour gamut, gloss
A i printability parameters
P2 - VP
1,20
1,00
0,80
Printability
0,60
0,40
0,20
0,00
0M
H
4H
4H
4H
0M
2H
6U
1U
H
0M
4H
4H
6U
4H
3H
H
6U
6U
I05
04
I08
14
K0
K1
V0
X0
X1
K1
V1
X1
I1
O0
O0
C1
O0
C1
K1
C1
P2 VP
0,25
0,20
0,15
0,10
0,05
0,00
1 2 3 4 5 6 7 8
-0,05
-0,10
-0,15
-0,20
-0,25
1,20
1,00
0,80
Printability
0,60
0,40
0,20
0,00
0M
H
4H
4H
4H
0M
0M
2H
6U
1U
H
4H
4H
6U
4H
3H
H
6U
6U
I05
04
I08
14
K0
K1
V0
X0
X1
K1
V1
X1
I1
O0
O0
C1
C1
O0
K1
C1
W
P3 VP
1,20
1,00
0,80
Printability
0,60
0,40
0,20
0,00
5H
4H
6U
4H
4H
6U
3H
6U
H
4H
1U
0M
0M
2H
4H
H
6U
4H
04H
I08
I10
14
K0
K1
V0
V1
X0
X1
K1
X1
I0
O0
O0
O0
C1
C1
K1
C1
W
P4 VP
P5 - VP
1,20
1,00
0,80
Printability
0,60
0,40
0,20
0,00
M
4H
6U
0M
4H
6U
4H
1U
3H
H
6U
4H
4H
0M
2H
4H
6U
04H
14H
I10
I05
I08
K0
X0
C1
C1
K1
X1
X1
K1
V0
V1
C1
O0
O0
O0
K1
P5 VP
with
αi = +1 for mottling, dot gain
αi = -1 for density
αi = 1/2 for blurriness, raggedness
A i printability parameters
P6 - VP
1,60
1,40
1,20
1,00
Printability
0,80
0,60
0,40
0,20
0,00
0M
H
M
1H
5H
H
2H
5H
5H
5H
H
7M
1U
5H
1U
5H
5H
7H
15H
I04
I09
I06
I11
03
05
O0
O0
X1
X1
C0
K0
V0
C1
C1
K1
K1
V1
S1
O0
W
O
P6 VP
with
αi = +3 for mottling
αi = +4 for dot gain
αi = -4 for density
αi = +1 for blurriness, raggedness
A i printability parameters
P5 and P6 are both using the same raw data as P4 but the difference is the
weighting of the different quality parameters. P5 is a trial to regroup
raggedness and blurriness under a sharpness factor by weighting both
parameters with 0.5. P6 is more ambitious and is recalculation of the
printability coefficient by distribution of 13 weighting. The weighting
corresponds to the number of measured fields for each parameter
(CMYK=4, mottling 40%, one colour, overprint = 3).
The second step was to confront the best model to the flexographic
process. This has been done by varying printing parameters like plate,
volume of the anilox roller, nip pressure for different substrates.
Parameters plots
1,00
0,80
Quality index
0,60
0,40
0,20
0,00
1
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
Mottling 30% Mottling Solid area Bleeding Wicking
Dot Gain 50% Density 2% Density 50% Density 100%
Roundness 2% Roundness 30%
Figure 14: Plot of the different quality parameters for the flexography
8,00
7,50
7,00
6,50
6,00
5,50
5,00
Printability
4,50
4,00
3,50
3,00
2,50
2,00
1,50
1,00
0,50
0,00
0
0
0
OS 0
00
DP 0
OS 0
1D 00
OS 0
M-D 200
OS 0
DP 0
1H 00
-DP 0
2H 00
10
UE 1220
20
UM 8100
1H 00
10
M-H 210
0
UE 8100
0
UE 200
UE 1210
UE 0820
UE 1220
UM 1210
UM 210
1D 0
0
82
81
82
81
10
2
22
UM 220
UE 081
10
22
12
1
12
T1 S08
UM 08
12
SE OS0
S0
S1
SE PS0
12
12
SE OS0
S0
08
S1
0
SE PS1
PS
OS
PS
0
1
SE OS
S
OS
OS
OS
S
PS
PS
PS
PS
DP
HO
O
HO
-D
-H
-H
1D
2D
-H
-D
2H
1H
-H
1D
1H
E-H
E-
E-
M-
E-
UE
M-
UM
UM
UE
UE
UM
T1
T1
T1
T1
T1
T1
T1
SE
SE
SE
SE
W
SE
SE
SE
SE
SE
SE
SE
SE
W
SE
W
P7 is an adaptation of the P2 model for the quality 1-7. The equation has
a similar construction. Moreover the quality parameters are this time
specific for the flexography:
with α i = +/ - 1
+1 for mottling 30% and solid area, wicking, bleeding, dot gain
2%, 30% and 50%, roundness 2% and 30%
-1 for density 2%, 50% and full tone
A i printability parameters
0,00
0,50
1,00
1,50
2,00
2,50
WT1E-HOS08100
WT1E-HOS08200
WT1E-HOS12100
WT1E-HOS12200
WT1E-DPS08100
WT1E-DPS08200
WT1E-DPS12100
with α i =
WT1E-DPS12200
WT1M-HOS08100
WT1M-HOS12100
WT1M-HOS12200
WT1M-DPS12100
WT1M-DPS12200
SEUE-HOS08100
SEUE-HOS08200
SEUE-HOS12200
A i printability parameters
SEUE-DPS08200
SEUE-DPS12100
SEUE-DPS12200
SEUE1HOS08100
SEUE1HOS08200
P8
SEUE1HOS12100
SEUE1HOS12200
combinations
SEUE1DPS08100
SEUE1DPS08200
SEUE2HOS08100
SEUE2HOS08200
SEUE2HOS12100
SEUE2HOS12200
MAX: 4.02 – MIN 1.14
SEUE2DPS08200
SEUE2DPS12100
SEUE2DPS12200
SEUM-HOS08100
SEUM-HOS08200
SEUM-HOS12100
SEUM-HOS12200
(10)
SEUM-DPS08100
SEUM-DPS12100
Figure 16: P8 for different plate/anilox/nip pressure/substrate
SEUM1HOS08100
SEUM1HOS08200
SEUM1HOS12100
SEUM1HOS12200
SEUM1DPS08100
P8 is a moderation of the printability coefficient P7 by weighting of the
135
SEUM1DPS08200
SEUM1DPS12100
SEUM1DPS12200
MAX: 1.52 – MIN: 0.60
P7 - P8
1,20
1,00
0,80
Printability
0,60
0,40
0,20
0,00
2D 200
0
M-H 2100
1D 00
0
DP 0
00
2H 00
0
1D 00
2H 00
0
E-H 100
00
0
M-H 100
M-D 200
UM 810
10
20
-HO 0
00
T1 0810
-HO 0
UE 1210
UM 0810
UM 210
UM 210
T1 1210
UM 220
10
0
82
T1 0820
21
2
1
82
2
12
SE S082
SE S122
12
SE S122
08
08
12
08
08
12
08
SE OS1
0
S1
SE PS0
SE PS1
1
1
SE OS0
SE OS1
PS
PS
PS
PS
OS
OS
OS
OS
PS
PS
PS
OS
OS
OS
OS
OS
PS
-D
-D
1D
E-D
-D
1D
1H
-H
-H
1H
1H
1H
M-D
E-H
E-
UE
UM
UE
UE
UE
UE
UE
UE
UE
UE
UM
UM
T1
T1
T1
T1
T1
SE
SE
W
SE
SE
SE
SE
SE
SE
SE
W
SE
P7 P8
4. Discussion
P1 is for example very well adapted to the offset process (O). The shape
of the curves are parallel and the distance between the plots are very
limited. P2 has been used to construct the flexographic printability
coefficient. P3 shows a very accurate compatibility with the Inkjet
process (X). P4 and P5 can be qualified as neutral equations which can be
used for example comparing to printing process: the curves do not
correlate as well as for the other equations but the point for point
5. Conclusion
The present study has delivered valuable results. The most interesting
result for the flexographic industry is that the possibility to have an
objective quantification of the printability has been proved. This
quantification called “Printability Coefficient” offers the printing and
paper industries a simple comparison instrument. The numbers obtained
can in the future be scaled and conduced to new development of the P7-
P8 equations but it has been shown that the choose of the parameters and
the form of the equation should be kept.
The next step of the work should be to use these equations for a prevision
of the printability and not only as a quality control instrument.
INTRODUCTION
The present paper is the last one of a series of papers regarding the results
of a research work, which had for final goal to obtain an objective
measurement and a prediction of the printability of different substrates
(paper/board) for the flexographic process. After an analysis of the
correlation between the different printing presses and modelling devices,
a study of the influence of the different printing parameters onto the ink
transfer, a quantification of the influence of the different paper
characteristics on the printability, a definition, establishment and test for
validation of different equations for a printability coefficient, the last part
of the work was to look at the possibility to use the chosen coefficient for
prediction of the printability of different substrates in flexography.
This paper is a presentation of the method used and results obtained for
different substrates, papers and boards printed in flexography. The
elaborated method should help the paper and printing industries to a better
communication and a common understanding of the word “printability”.
The use of a “Printability Coefficient” and the expected generalisation of
the method for all kind of peripheries and substrates should be a big step
in the standardisation of a flexographic printing process still in
development.
RESULTS
The equations P9 and P10 have been established specific for the objective
evaluation of the printability by flexographic printing. The correlation
with subjective evaluation has been successful tested.
with α i = +/ - 1
+1 for mottling 30% and solid area, wicking, bleeding, dot gain
50%, roundness 2% and 30%
-1 for density full tone
A i printability parameters (printing and paper)
The quality factors (mottling index, edge sharpness, dot gain, density, dot
deformation) are different contribution to the final “P”. The figure 1
shows the variation of “P” for the Min and Max values of the different
quality factors. Plotted are the P9 and P10 Min and Max values.
3,50
3,00
2,50
2,00
"P units"
1,50
1,00
0,50
0,00
Max
Max
Max
Max
Max
Max
Max
Max
Max
Max
Max
Min
Min
Min
Min
Min
Min
Min
Min
Min
Min
Min
Average
Average
Average
Average
Average
Average
Average
Average
Average
Average
Average
Mottling Mottling Mottling Bleeding Wicking Edge Dot Gain Density Roundness Roundness Roundness
Screen 100% Sharpness 2% 30%
P9min varies from 0.43 to 2.00, P9max from 1.84 to 3.27, P10m i n from 0.19
to 1.34 and P10max from 1.42 to 2.02.
The following figures (2) and (3) represent respectively for P9 and P10
the influence of the quality factors variations converted in “P units”.
0,36
0,93
1,03
0,23
0,91
0,18
0,47
0,52
0,22
0,91
For both “Printability coefficient” the largest ∆QF is to notice for the dot
gain quality factor. For P9 the smallest ∆QF?is measured for the density
and for P10 for the mottling.
QF influence in %
80,00
70,00
60,00
50,00
%
40,00
30,00
20,00
10,00
0,00
30
ing
00
ing
ity
M
0%
M1
s
s
ling
ed
ns
ick
es
es
ain
s2
Ble
s3
W
De
ott
dn
pn
tG
es
M
un
es
ar
Do
dn
lta
lta
dn
Ro
Sh
un
De
lta
De
un
Ro
De
lta
ge
Ro
De
Ed
lta
De
P9 P10
The figure 4 confirms the precedent results and gives indications about
the contribution of the primary factors. For mottling (17.07/15.24%) and
dot deformation [roundness] (32.32/26.22%) the contribution are similar.
For the edge sharpness the component bleeding with 57.93 - 39.67% to
compare with 22.56 - 15.70% for wicking is the major contributor.
The first step is to calculate the influence in percent for each seven
parameters. The results are obtained by dividing the ∆parameter
(∆parameter =ParaMax-ParaMin) by the correspondent ∆printability.
50,00
40,00
30,00
20,00
%
10,00
0,00
-10,00
% Anilox % Pressure % Hardness
-20,00
-30,00
Both positive and negative values are noticed for all the printing
parameters. This has for consequence an improvement or a deterioration
of one or other quality factors and also point one more time the
complexity of the interaction between the different parameters. The trends
are the same for three factors: mottling 30%, dot gain and roundness 30%.
The range of the influence is varying from 50.81% for the nip pressure to
63.04% for the plate hardness. The influence onto the global mottling for
the plate hardness is close to 0% (2.14%).
50,00
40,00
30,00
20,00
%
10,00
0,00
-10,00
-30,00
Against positive and negative values are read on the figure. Moreover for
the porosity the trend is clear: only the influence on the wicking shows a
significant positive value. The influence of the surface roughness and
absorption are positive for the mottling and edge sharpness quality
factors. It is also interesting to remark the similarity of the values. The
range of the influence is varying from 30.41% for porosity to 44.44% for
the surface roughness. The surface roughness, density and absorption
notice values around 0% for respectively the dot deformation, wicking
and mottling solid area/density quality factors.
The following figures represent the results for the printing parameters:
Anilox roller volume, plate/paper nip pressure and hardness of the
printing plate. The results are presented for P9 and P10.
Unit
Anilox Roller Volume 0.5 cm3/ m2
Plate/Paper Nip Pressure 10 µm
Hardness of the plate 1 Shore A
4,0
3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0
-0,5
-1,0
% Anilox / P9 % Pressure / P9 % Hardness / P9
-1,5
-2,0
-2,5
-3,0
-3,5
3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0
-0,5
-1,0
-1,5
% Anilox / P10 % Pressure / P10 % Hardness / P10
-2,0
-2,5
-3,0
-3,5
% Degree
< 0.25 minimal
0,25<βi<1 important
The dot gain with values over 1% seems to be the common factor to all
the printing parameters for both P9 and P10. The nip pressure shows as
well a great influence (-1.4%) on the bleeding. Comparable is the
influence of the anilox roller on the density (0.9%).
For P9 the variations are between –1.4 and +1.5 %, for P10 between –0.9
and +2.0 %, what means the same 2.9% range.
The following figures represent the results for the printing parameters:
Surface Roughness, Density, Absorption and Porosity. The results are
presented for P9 and P10.
Unit
4,0
3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0
-0,5
-1,0
% Surface Roughness / % Density P / P9 % Absorption / P9 % Porosity / P9
-1,5
P9
-2,0
-2,5
-3,0
-3,5
3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0
-0,5
-1,0
-3,0
-3,5
The first observation is that the proportional united influences are larger
for the paper parameters than for the printing parameters. For the density,
absorption and porosity, the influence onto the dot gain is over 2%. The
bleeding shows as well values around 2% for the surface roughness, the
density and the absorption (around 1.5% for P10). For the same
parameters give the mottling 30% values around 1%. A further value over
1% (1.2%) is the wicking for the porosity parameter.
For P9 the variations are between –2.4 and +2.7 %, for P10 between –3.2
and +3.7 %.
6,0 5,3
5,1
5,0
4,0
3,0
3,0 2,5 2,6
2,3 2,1
2,1
2,0
1,2
0,7
"P units" %
1,0
0,0
0
10
P9
10
P9
0
9
P9
P1
0
10
9
/ P9
-0,3
P1
/P
10
P9
/ P1
/P
/P
/P
-1,0
/
-0,5
/P
/
/
/P
x/
ilox
ity
P
/
on
s/
ess
re
ity
P
on
ess
ilo
re
ity
ros
ssu
ess
rpti
s
ity
An
ros
ssu
rpti
rdn
e
An
ns
rdn
hn
ns
Po
Pre
so
hn
%
Po
De
Ha
so
Pre
ug
%
De
Ha
-2,0
Ab
ug
%
%
Ab
%
%
Ro
Ro
%
%
ce
ce
rfa
rfa
-3,0 Su
Su
%
-4,0 %
-4,1
-5,0 -4,5
-6,0
Printing parameters:
Anilox roller volume: 10 cm3/m2
Nip (plate/substrate) pressure: 150 µm
Printing Plate Hardness: 64 Shore A
Paper parameters:
Paper Density: 735 kg/m3
Surface Roughness (Bendtsen): 550 ml/min
Absorption (Bristow): 13,55 cm3/m2
Porosity (Bendtsen): 152,5 ml/min
The mean values for P are: P=2.22 for P9 and P=1.20 for P10.
The Min and Max values will serve the user to judge the quality of the
combinations, she/he has simulated and calculated the “Printability
Coefficient” for.
Min and Max values are calculated by giving the (Ai /max Αi ) (9) the
values 1 (worst) or 0 (best).
The same approach allows to fix the worst achievable quality by 0 for
both P9 and P10. This is a logical consequence of the interaction of the
quality factors.
The limits for P are now known. The centre of the variation interval will
be considered as the “at least to achieve quality” (Pg) for the prediction
stage.
7 2
0,00
6 3
5 4
Pp Pm Pav
Pmin Pmax Pg
The next step is to make the calculation of a prediction for P (Pp). For the
calculation the user needs to know or choose values for the printing and
paper parameters. These values (xi ) are the input for the evaluation of the
deviation to the standard values (a i). The deviation will then be divided by
the unit coefficient (c i) and multiply by (fi), the, in percent, influence
coefficient. The sum on the i indexes of the results for the different (a i)
will be called “p”. Pp is then the result of the multiplication of (p/100 + 1)
by the average value of P (Pav ).
p = ∑ {[ (xi - ai ) / c i] * f i } (3)
The last step of the process takes place after the printing stage. To
personalize the Pp the user has the possibility to calculate a ∆P, which will
give him an idea of the accuracy of his prediction. The ∆P can be used to
adjust with an iterative procedure more precisely the f i to his production.
∆P = Pp – P m (5)
CONCLUSION
The demand of High Quality Flexography is growing very fast and the
exigencies are becoming higher and higher. All the sectors of the
flexographic fields work to improve their products. The luck of
standardisation and the diversity of the products on the market are
responsible for difficulties in the right choice of the materials. The
research work presented in this paper proposes a tool to solve this
problem. With the help of the “Printability Coefficient” and the
prediction procedure will the flexography manage the challenge of the
future!
REFERENCES
Armel, D., Cusdin, G., “ Plates & Printability ” – FQC FLEXO 04/2001