Measurement and Prediction Procedures For Printability in Flexography (MP3 Flexo)

Measurement and Prediction
Procedures for Printability in

Flexography
(MP3 Flexo)
Laurent, Girard Leloup
Stockholm 2002
Doctoral Dissertation
Royal Institute of Technology
Department of Numerical Analysis and Computer Science
---
Media Technology and Graphic Arts
Respondent: Laurent, Girard Leloup
Opponent: Dr. Patrice J. Mangin
Examination committee: Dr. Renke Wilken

Prof. Lars Järnström
Dr. Lars Palm
Supervisor: Prof. Nils Enlund
Chairman: Prof. Roger Wallis
Thesis for the degree of Doctor of Technology to be presented with due

permission by the Royal Institute of Technology, for public examination
and criticism, in room E3, at the Royal Institute of Technology, KTH,
Lindstedtsvägen 5, Stockholm, on June 7, 2002, at 10 AM.
TRITA-NA-0212
ISSN-0348-2952
ISBN 91-7283-306-8
ISRN-KTH/NA/R-02/12-SE
* Laurent, Girard Leloup, Mai 2002
Printed in France by ATO IMPRIMERIE - 8, Rue Jacquard - 54500 VANDOEUVRE
2 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

Abstract
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.
A literature study around the printability and quality control terms in

flexography, completed by interviews, help for the analysis of the
parameters, the planning of printing trials and the measurement of quality
factors. The next stage was the establishment of the procedures.
The present work should help the flexography to achieve needed

standardisation to assure the stability of the process. Flexography is
profiting and simultaneously suffering from its broadband of possibilities,
which make the decisions complicated in view of quality striving.
The thesis proposes a solution to objectively quantify the quality of the

achieved (measurement procedure) or achievable (prediction procedure)
result. The use of a “Printability Coefficient” obtained with the help of
measurement and prediction procedures will give the industry a further
weapon to improve and promote flexography.
“MP3 Flexo” is a quality control and communication instrument, which

accompanies the product from its conception (customer, publicity agency,
paper and supplier industries) to its realisation (printing industry).
The system as presented here can be implemented in the industry.
Key words: flexography, printability, objectivity, measurement,

prediction, quality control
Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002 3

Preface
The project was part of a national research program called T2F

(“Tryckteknisk Forskning”), financed at the same level by the Swedish
government and the paper/printing industries. A network of industrial
partners and research institutes throughout Sweden manages the projects.
This project was the first one and accompanied the building up of the
network.
In addition to the scientific results presented in this thesis, my work also

provided many results and experiences on another level. From the human
point of view the project brought some results about teamwork,
communication and industrial property in the paper and printing research
and industry. A few examples:
• There exists a real concurrence between the different research

institutes. This concurrence occurs at the personal, knowledge
and equipment levels. It could be a positive motivation factor but
also, in a case of a network, slow down some projects.
• Human affinity or the lack of it could be a relevant success
factor
• The industry is open to interesting projects and can offer help by
delivering materials or exp ert time
• Meetings, conferences and fairs visits are important inputs for a
research project. The exchange of information with other experts
guides the project by giving it the right orientation and avoiding
too theoretical work, work without interest for the industry or the
repetition of work already done.
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.

List of included papers
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

Contents
1. Intr oduction 4
2. Background 5
2.1 State of the art 5

2.2 Identification of the problem 5
2.3 Aim of the thesis 5
3. Definitions 6
3.1 What is flexography? 6

• 3.1.1 The official definition 6
• 3.1.2 Advantages of fle xography 7
• 3.1.3 Disadvantages of flexography 7
3.2 The definition of the “Printability” 8

• 3.2.1 Industrial experts 8
• 3.2.2 Pulp and Paper Association 11
• 3.2.3 Technical glossary / book 11
• 3.2.4 Printability definition in this thesis 11
4. Related Research 12
4.1 1988 … 2002 12

4.2 Image analysis and the Printing Industry 13
• 4.2.1 Definition 13
• 4.2.2 Hardware and Software for image analysis 13
• 4.2.3 Measurable parameters 14
4.3 Visual perception 15

• 4.3.1 Definition 15
• 4.3.2 Observation settings 16
• 4.3.3 Evaluation procedures 16
5. Methodology and progression of the research 18
5.1 Methodology 18
5.2 Research progression 20

• 5.2.1 Census of the parameters 20
• 5.2.2 Individual analysis of each component 20
• 5.2.3 Description of the evaluation method 21
• 5.2.4 Definition of the quality factors 21
• 5.2.5 Determination of the influence of each parameter 23
• 5.2.6 Acquisition of the evaluation methodology 23
• 5.2.7 Establishment of a “Printability Coefficient” 23
• 5.2.8 Test of the relevancy of the equation 24
• 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
• 5.2.10 Establishment of a procedure for the prediction

of the Printability 25
• 5.2.10.1 Definition of a reference 25
• 5.2.10.2 Determination of the Min and Max values 25
• 5.2.10.3 Theoretical and practical optimal quality 25
• 5.2.10.4 Prediction of the Printability 26
• 5.2.10.5 Judgement of the quality 26
• 5.2.10.6 After printing 27
6. A journey through six papers 28
6.1 The printing press parameter 28

6.2 The influence of the primary parameters 29
6.3 Deeper in the paper characteristics 30
6.4 A Flow diagram as directive line 31
6.5 Establishment of a “Printability coefficient” 32
6.6 The way back 33
7. Summary of results 35
7.1 Technical results 35

7.2 Results for the process 36
8. Discussion 36
8.1 The adjustments 37

8.2 Relevancy of the work 37
8.3 Proposition for further work 38

9. Conclusions 38
Acknowledgements 40
Literature cited and reviewed 41
Included Papers
Paper I 48
Paper II 66
Paper III 82
Paper IV 102
Paper V 114
Paper VI 136

1. Introduction
In spite of the appearance of a multitude of new information and

communication technologies, the Graphics Arts are still one of the leading
media support. This position is due to the variety of the products and the
constant reaction to new requirements. High quality is today, in addition
to the production speed and economical aspects, a must. Flexography,
which is one of the most-used printing processes in the packaging
industry, had also to raise the challenge to survive.
The complete environment of flexography has been continuously the

object of an evolution. The evolution has been so enormous and so rapid
that it is here possible to speak about a real revolution at the printing
industry scale.
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 ...
Today the professionals of the flexographic printing industry have at their

disposition all the technical elements to be able to produce a very high
and competitive quality. But they sometimes lose their way by navigating
through the multitude of parameters.
Why?
Because they do not have the help of a structured method, of an

instrument to objectively evaluate and compare the results achieved with
different parameter combinations.

2. Background
2.1 State of the art
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 years 2000-01 were strong for flexographic printing, as quality

achievements impressed package buyers and, near the traditional
flexographic products, more conversions from offset were noted (Armel,
2001). Computer-to-plate was most notable in this segment, providing a
competitive tonal range and detail rendition. Flexographic printing with
UV inks has matched offset printing in quality.
2.2 Identification of the problem
However, and unfortunately, flexography, which strives to compete with

the gravure and offset processes, is confronted with two main problems:
the absence of standards (as they exist for the offset process) and the
impossibility of previewing and measuring with objectivity the obtained
quality. The lack of standards has inhibited progress in reaching more
consistency from job to job and plant to plant, as well as the inability to
predict results without costly makeovers. Moreover a minimum amount of
systematic and disciplined training prevents companies from taking
advantage of its equipment capabilities.
2.3 Aim of the thesis
The present thesis will contribute to find a solution to the above-

mentioned problems, which could be obstacles for the competitiveness of
the flexographic process in the future. At the same time the work should
be a step, a contribution, to a necessary standardisation of the
flexographic process.
The main objective of this research work is to make it possible to measure

the printed quality in flexography and in a second step, to be able to
predict the print quality without any trials before the achievement of the
print. 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 interpretation and not to
scientific measurement. To achieve these requirements of simplicity and
objectivity, we need the help of a “key number”: In my investigations I
have studied the relevancy of such a number I call the “Printability
Coefficient”. The printability coefficient “P” is the final number resulting
from the input into an equation of measured quality quantificators called
quality factors, like density, mottling, edge sharpness, cleanliness of solid
area, dot gain and dot deformation. The gain of the equation will result
from different steps in the project. These steps can be seen as
intermediate goals: study of the influence of the paper characteristics,
study of the influence of the printing parameters and acquisition of the
quality input parameters. The modelling of the print and modelling of the
unprinted paper in view of printability will serve to establish the
printability prediction procedure, which could be seen as the end product
for the industry.
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
3.1 What is flexography?
3.1.1 The official definition
“Flexography is a method of direct rotary printing that uses resilient relief

image plates of rubber or photopolymer material. The plates are affixable
to plate cylinders of various repeat lengths, inked by a cell-structured ink-
metering roll, with or without a reverse-angle doctor blade, and carrying a
fast drying fluid ink to plates that print onto virtually any substrate,
absorbent or non-absorbent. Flexography is a rotary printing method,
which means for every revolution of the printing plate cylinder, an image
is produced.” (FFTA, 1999)

3.1.2 Advantages of flexography
• It can print on a wide variety of absorbent and non-absorbent

substrates
• It uses fast-drying inks, whether solvent, water-based or
ultraviolet (UV) curable ink
• It can print wet ink over dry ink to eliminate trapping problems,
back-trap contamination and setoff
• It uses resilient rubber or photopolymer image carriers (printing
plate) that can print millions of impressions
• Presses can accommodate a wide range of cylinder repeat lengths
to match customer print length requirements. Flexography is a
near total variable-repeat-length system
• Its inking system (anilox roller) can deliver a predetermined
amount of ink with minimum on-press adjustments
• It can print continuous patterns (gift wrap, wallpaper, floor tiles)
• It can print on extensible plastic films
• It can print on the reverse side of stretchable, transparent films
• It can perform coating and in-line laminating operations
• It is cost effective for many applications
• It enables fast turnaround time between jobs
• It can do short-run work profitably
• Presses can produce in-line, pressure sensitive labels in a
continuous operation
• Presses can produce many types of projects requiring down line
finishing such as: die cutting, kiss cutting, scoring, perforating,
embossing, hole punching, etc.
• The investment costs are lower than for the offset and gravure
processes
• The achievable quality is continually growing and now matches
the offset and gravure recognized qualities
3.1.3 Disadvantages of flexography
• It still shows some repeatability problems from order to order:

the stability of the process is not yet equal to that of gravure
• It has difficulties with the reproduction of motives with
continuous screen
• The plate making is time expensive
• It is characterised by a large dot gain, which has to be
compensated for

• It has a tendency to halo around the text: this can be seen as a
edge sharpness problem due to the elasticity of the printing form
• The large range of substrates and products confront the printer
with a lot of different situations and problems to solve
3.2 The definition of “Printability”
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.
To understand the meaning of “printability” it is necessary to confront

the word to its industrial context.
To get definitions that can be classified under the category “descriptive

definitions”, an interesting survey* has been carried out and gives the
following results to the question: “What is your definition of
printability?”
3.2.1 Industrial experts
Industrial experts are people working in the industry at a leading or

research level. They have been selected for their diversity. The diversity
must be understood as different sectors of the industry, countries,
educations and age classes.
“Printability is a term which describes the interaction between paper, ink

and press from the paper’s point of view.”
“A somewhat difficult question - define it along the lines of the digital

workflow:
A file to be printed is printable when

1. the file itself is processable in the environment under view and (not or)
2. the device for printing the file is able to fulfil the requirements as they
are encrypted in the file and (not or) sooner or later
3. the digital environment around a device is able to process the job ticket
as it is attached to the file.
* A list of the interviewed experts can be found in the chapter “References”.

In future, to some degree the information-feedback strategy (how does the
device communicate with controllers, servers, front end) will also be an
issue which will be decisive on whether or not a file to be printed is
printable.”
“For me printability is a word for calibrating media source material for

different quality printing output.”
“Physical and chemical parameters that influence the transfer, setting and
drying of inks on the substrate.”
“Excellent question! Would like to see the answers.”
I would like to explain printability starting with defining print quality in

measurable variables as:
1) Register (Colour to colour and colour to structure);
2) Edge sharpness (this would cover the edges of bar codes but also dot's
in half tone printing);
3) Colour (shade and variability in colour);
4) Cleanliness (dust spots in full tone area);
5) Mottling (random and pattern or stripiness in print related to the flutes
of corrugated board);
If we now look at printability of paper, we need to look at 2 to 5 and the
paper properties that influence them. The measurable paper properties that
can influence them are:
a) Porosity (influences: 2, 3, 5);
b) Colour (influences: 3);
c) Surface energy (influences 2, 3);
d) pH (influences: 2, 3);
e) Dust (influences: 4)
This might look too simple but use it as a start. I might have overlooked
something.”
“This is a tough one!

I would probably define printability as:
...the measure of likelihood of being able (or not) to reproduce (print) the
text or design in the colours selected or by the printing method chosen.
Printer’s point of view
“Printability is whether or not the printing ink will adhere to the
stock/material being used in a satisfactory manner”.

Agency point of view
“The suitability of the printing image and the paper/card/board to be
printed on.”
“Printability is a complex term but is influenced by several specific

factors. These include surface topography (smoothness), compressibility,
ink receptivity and holdout, ink vehicle absorption and dimensional
stability. Of course each characteristic has its own complex set of
variables. Sorry, but it is not so simple.”
“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.”
“…to me is the ability of a print substrate to present the information (to

be printed) so that it is perceived by the viewer to be “as” or “close to”
the original;
In the case where no original exists, you could perhaps say: perceived by
the viewer to be “as” or “close to” what is supposed to be the original.”
“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.”

3.2.2 Pulp and Paper Association
“The printability is how well the paper performs in a printing machine.”

(Canadian Pulp and Paper Association)
“Printability is a measure of the ability of a printing paper to provide a

high quality printing medium. It has been found that print mottle, or the
unevenness of print density, is the most important factor affecting
printability” (Pulp and Paper Centre, Toronto)
3.2.3 Technical glossary / book
“The combination of print quality characteristics that enhance the

reproduction of an original in any printing process.” (Glossary of Graphic
Communications)
“That characteristic of a material, such as paper, which permits high

quality printing, and which, though not capable of precise definition, is
generally judged visually in terms of uniformity of colour of the printed
areas, uniformity of ink transfer, quality of “black on white”, and rate of
ink setting and drying. In paper, hardness, smoothness, opacity, colour,
and pick resistance are some characteristics, which lend themselves to
good printability.” (Bookbinding and the Conservation of Books, A
dictionary of Descriptive Terminology)
“A broad paper characteristic, which refers to a paper’s suitability for

presswork.” (Glossary of paper terms, Southwestern Indian Polytechnic
Institute)
“Printability covers a range of paper properties affecting print results:

gloss, smoothness, whiteness, opacity, etc.” (Print Process)
“How well a paper performs with ink on press. Absorbency, smoothness,

ink holdout, and opacity all affect printability.” (Paper Encyclopedia)
3.2.4 Printability definition in this thesis
Reading the different definitions emphasises the disparity of the different

definitions. The paper industry of course does not have the same
definition as the IT industry, nor is it the same as the printing industry.
However, these 3 industries (at least 2 of them) are working in a much
confine connection. This observation justifies the necessity of an “own”
or “work definition” to clarify the objective of the research. This

definition, which is also a stipulative one, will have to be accepted to
understand and discuss the research work.
The definition of printability I have chosen for this work is the following:
“Achievable print quality resulting from the interaction between the

critical properties of the substrate and the printing parameters”
The main objective of the work is to give an objective measurement of the

printability for the flexographic printing process. This objective
measurement is nothing more than an operational definition of the
printability.
4. Related Research
4.1 1988 … 2002
Some researchers already tried in the past to objectively quantify the

printability. Some of them did that even specifically for the flexographic
process. The common factor for all the research work is to take the
substrate as the parameter for the study. The substrate, depending on the
work, is characterised by one to a multitude of properties, like surface
topography, absorption, porosity, optical properties, etc. Moreover, it is
possible to distinguish different schools in the research approach.
The classification can be done in two categories:
• By listing the variables considered in the study: the most

common studied single correlation is the influence of the
substrate properties by keeping other parameters constant (Zang,
1995), (Göttsching, 1997) and (Miller, 1997). A few papers
regarding the ink have been published (Ginman, 1988), (Jensen,
1989), (Steadman, 1993). The influence of the printing pressure
has also been studied as an independent factor (Pommice, 1989).
Then we have a few publications taking into account, near the
substrate, two or more other parameters. Lindström (1994) was
looking at the printing speed and anilox roller contributions;
Chalmers (1997) was taking into consideration both the ink and
printing pressure factors. The most ambitious attempts,
concerning the number of parameters have been realised by
Aspler (1998) and Armel (2001) by analysing the same
combination ink, printing plate and printing pressure.

• By looking at the chosen quality factors for the evaluation of the
printed samples: in the same way as for the variables we can here
note research using a single quality factor, the density
(Chalmers, 1997), the mottling (Armel, 2002), the missing dots
(Göttsching, 1997) or the unevenness (Miller, 1997). Lindström
(1994) and Steadman (1993) were analysing the density and
unevenness, Zang (1995) added the gloss whereas Pommice
(1989) concentrated his work on dot gain and density. In the late
80’s, the Finnish school with Ginnman (1988) and Jensen (1989)
developed a more global approach by looking simultaneously at
the density, contrast, gloss, dot gain and unevenness of the
printed samples. 10 years later, Aspler (1998) made a tentative
with the mottling, bar code reading and dot gain quality factors.
A common method for the measurement of the substrate properties or the

evaluation of the printing samples is to have recourse to the image
analysis. It was also interesting to give a brief overview of this
methodology.
4.2 Image analysis and the printing industry
4.2.1 Definition
“Image analysis is a discipline, which takes two-dimensional data from a

variety of sources, manipulates it and makes measurements on it”
(I’Anson, 2001). This data is usually what could be described as a digital
photograph obtained from a camera or a scanner.
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.
4.2.2 Hardware and software for image analysis
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:

Devices:
• Microscopy
• Ultraviolet viewing systems
• Infrared viewing systems
• Image enhancement systems
• Thin layer chromatography
• High-pressure liquid chromatography
• Gas chromatography
• Scanning electron microscopy
• X-Ray
• Fluorescence spectrometry
• Fourier transform infrared spectrophotometry
Software:
• Standard image processing software
• IDL
• ImagePro
• MBS
4.2.3 Measurable parameters
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.
Some of the current utilisations of the image analysis:
• Dirt and ink particle measurement and counting

• Filler distribution in paper
• Density profile through paper thickness
• Fibre length, coarseness
• Fibre contact ratio
• Fibre orientation
• Formation and flocculation
• Forming fabric, press felt, dryer fabric, press roll mark diagnosis
• Paper machine CD relative shrinkage profile
• Surface roughness
• Curl of paper
• Local gloss measurement
• Calender blackening
• Analysis of print mottle

• Measurement of dot deformation
• Evaluation of optical dot gain
• Edge sharpness determination (wicking, bleeding, etc.)
• Missing dots
4.3 Visual perception
The visual perception makes an attempt at a global quality statement,

which corresponds to the definition of the printability retained for this
research work. Lyne in 1979 already showed the way to follow by
introducing the multidimensional scaling for the print quality; Aspler
(1998) and his colleagues at PAPRICAN developed the method and
looked at the correlation with the objective measurements.
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).
In the quality control field, visual perception is a “human” instrument,

which allows a subjective evaluation of the product quality. For the paper
and printing industries, the products are printed samples and the criteria
for the judgement parameters, like colour rendition, sharpness, mottle,
contrast, etc.

4.3.2 Observation settings
The visual perception is based on the observation of one selected and

representative sample of each printed series.
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.
4.3.3 Evaluation procedures
a. The “Proscale” evaluation

a.1 Definition
Human psychological judgment has been computerized in Proscale.

Proscale is a measurement technique and accompanying software tool. It
offers a variety of techniques adapted to measuring paper and print
quality. Proscale allows the user to obtain simple ratings of overall
preference for a series of samples, as well as ratings on specific
describable aspects of the samples. It also provides a method for
obtaining multidimensional similarity and preference ratings from a
sample set, for associating these measurements with physical
measurement variables through regression and correlational techniques,
and for averaging the results from replicated samples within the set. It
provides a method of defining observer sub-groups whose judgments
about the sample set differ (Donderi, 1999).
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.

A test panel participated in the experiment. Everyone had normal or
corrected to normal vision and had been tested for colour vision with a
simple test The test panel was instructed to arrange the samples into
groups with respect to their similarity in print quality so that each group
had some characteristic in common that separated them from the other
groups. The participants of the test panel were free to arrange as many
groups as seemed appropriate. The groups were then assigned ratings
corresponding to each participant’s perception of the quality of the
samples in each group. As an attempt to understand the strategy behind
the grouping, the participants of the test panel were asked to characterize
each group of samples. Finally, the participants were requested to decide
whether the samples could be considered acceptable to a buyer or a
consumer.
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’.
Type of scale Scale levels

Quality scales
Overall quality 9 levels, ‘unusable’ to ‘excellent’
Colour rendition 9 levels, ‘unusable’ to ‘excellent’
Sharpness 9 levels, ‘unusable’ to ‘excellent’
Contrast 9 levels, ‘unusable’ to ‘excellent’
Rendition of details in highlight 9 levels, ‘unusable’ to ‘excellent’
Rendition of details in shadow 9 levels, ‘unusable’ to ‘excellent’

Tone quality (smoothness of
9 levels, ‘unusable’ to ‘excellent’
tone transitions)
Gloss level 9 levels, ‘unusable’ to ‘excellent’
Impairment Scales
9 levels ‘very annoying’ to ‘not
Colour shift
perceptible’
Patchyness
perceptible’
Mottle
perceptible’
Ordered noise
perceptible’
Gloss unevenness
perceptible’
Table 1: Opinion rating scales.
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.
5. Methodology and progression of the research
5.1 Methodology of the research
The analysis of the market and trends took place at different levels of the
paper and printing industries.
A literature survey around the terms quality control, measurement,

printability, print quality, flexography in different databases has covered a
large part of the research field.

The results of the literature survey have been completed between 1999-
2002 by:
• attending scientific conferences arranged by PIRA [Printing
Industry Research Association], FTA [Flexographic Technical
Association], STFI [Swedish Pulp and Paper Research Institute]
or TAGA [Technical Association for the Graphic Arts].
• visiting paper mill production in Sweden (SCA, StoraEnso),
Germany (StoraEnso) and Austria (Sappi), research sites in UK,
Germany and Sweden and printing plants from several large
companies (Å&R, TetraPak, Kobusch Folie, etc.). These visits
gave the occasion to interview employees and experts of both
paper and printing industries.
• Visiting fairs (DRUPA, PROFLEX, INFOFLEX, PrintPack,
etc.), which offered the opportunity to contact the supplier
industry. This opened the doors to companies like Bobst,
Dupont, BASF, Praxair, ManRoland…
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.
To answer the question, it would be tried at several printing trials at

different places both in Sweden and Germany. The variations of the
printing conditions and materials corresponded to the objectives of the
investigations (Papers I-III). To complete the printing trials, a calendaring
would be achieved at the STFI (Paper III).
Each printing trial was followed by an evaluation of the printed results to

get the required numeric results and to prepare the next trials. The
measurement part of the work has been carried out at different places,
depending of the nature of the measurements and the equipment needed.
Framkom [Research Corporation Media and Communication
Technology], DFTA-TZ [Technology Centre of the German Flexographic
Technical Association] and the paper labs of the paper industry are the
more often used resources (Papers III & V).
The link between the different actions will be described in detail in the
next part, called research progression.

5.2 Research progression
5.2.1 Identification of the influence parameters
The problem to be solved had a high grade of complexity. A detailed

analysis was needed to first set the parameters, which could have an
influence on the printability in flexography. Then it would be possible to
simplify the problem environment.
The question to answer corresponded in a certain way to building a bridge

between the paper and printing industries and their customers. This
bridge, the “Printability Coefficient”, should facilitate communication,
decision-making and avoid conflicts. The parameters could be classified
in three main categories, where only two of them have been analysed
here. The first category was called “printing parameters” and contained
all the materials, press and press peripheries involved in the flexographic
printing process. The second category, called “paper characteristics”
gathered the substrate properties, which primary influence the printability
(Paper II). The third category, which will not be taken into account, is
formed by the “human factors”.
5.2.2 Individual analysis of each parameter
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.

5.2.3 Description of the evaluation method
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.
Two mathematical tools have delivered the results: traditional statistical

evaluation and Multivariate Data Analysis (MDA). The MDA has been
employed to make comparisons, and with the help of the projection
method, has allowed a graphical summarization of the results for a
meaningful overview (Paper III). The main contribution of the MDA
method was the elimination of the parameter “printing press”.
The statistical instrument has help by the construction of the equation, the
different modelling and the establishment of the printability prediction
procedure.
5.2.4 Definition of the quality factors
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 Paper
parameters characteristics
Printing Trials
Printed samples
Modeling Modeling
of the Measurement of the
print tool substrate
Quality
Factors
”P”
”Printability Coefficient”
Figure 1: A Flow diagram: “The Printability system”
By varying the combinations of the printing parameters and paper

characteristics, printing trials has been carried out. The results of the
printing trials were a series of printing samples. The printing samples
were in the next step evaluated with the help of a measurement tool. The
quality factors were obtained at this stage. The quality factors were the
input of the Pi equations, which has been established in parallel (Paper
V). The output of these equations (P9 and P10) is the “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

costs and time expensive printing trials and quality factors measurement
stages.
5.2.5 Determination of the influence of each parameter
Each parameter has been analysed by its characterising properties. The

influence of the parameter has been determined by choosing materials
(substrates and periphery equipments) covering a large range of the
products available on the market. To guaranty the relevancy of the results,
the combinations have been chosen, such as the variations of the values
for the quality factors that show great amplitudes. Systematic printing
trials (Papers III and V) with variation of one parameter, keeping the
other constant, are allowed to test the influence.
5.2.6 Acquisition of the evaluation methodology
The acquisition of the evaluation methodology comported two parts. The

first part was a theoretical contribution. It corresponds to the development
of the flow diagram. The flow diagram represents the strategy for the
evaluation work (Paper IV). The second part was the selection of the
methodology or devices to get the values of the quality factors. The
quality factors are common and recognised factors in the quality control
sector of the printing industry. Moreover the lack of a standard gives the
user the freedom to choose his measuring methods.
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.
5.2.7 Establishment of a “Printability Coefficient”
The establishment of a “Printability Coefficient” for the flexography

process has been realised in successive stages. The first stage was to find
the best linear combination of the measured quality factors (QF). The
linear combination had to contain at least all the previously selected
quality factors and fit with the visual perception values (VP).

The VP -quality was evaluated for the flexographic samples by expert and
non-expert panels. More details about VP have been done in chapter 4.2.
The goal was to obtain an equation valid for the flexographic process by
minimising the distance between the values of both printability
(measured/calculated and visual) but at the same time keeping the shape
of the printability curves parallel for the other printing processes (Paper
V).
5.2.8 Test of the relevancy of the equation
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.
5.2.9.1 Modelling of the printing parameters
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.

5.2.9.2 Modelling of the paper characteristics
The modelling of the paper density, surface roughness, absorption and

porosity followed the same construction as the modelling of the printing
parameters. The choice of these characteristic resulted from two criteria:
the characteristics had to be standard parameters each paper supplier is
able to provide in the paper delivery protocol, and the influence of the
printability is great for the flexographic process (Papers III and VI).
5.2.10 Establishment of a procedure for the prediction of the Printability
The procedure should be seen as a conclusion of the research work and at

the same time as a manual for the user: “How can I get a value for “P”?”.
The establishment of the procedure followed a logical process. The first
steps are necessary for the construction and understanding of the
procedure but not to use it. The central part is the prediction itself. The
judgement and “after printing” parts are the analysis steps (Paper VI).
5.2.10.1 Definition of a reference
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).
5.2.10.2 Determination of the Min and Max values for P
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 !
5.2.10.3 Theoretical and practical optimal quality
The theoretical achievable quality is the quality calculated in the

preceding chapter (Pmin and Pmax ): 7 and 4 respectively for the printability
equations P9 and P10 (Paper VI). Moreover, the process has certain
limitations and tolerances, which have to be reflected in the judgement.
These limitations will reduce the interval of variation for “P” and centre
the mean value of Pa v. This new value for P will be fixed as goal for the

user: Pg. The correction is effectuated by considering the tolerances for
the quality factors (Paper VI).
5.2.10.4 Prediction of the Printability
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).
Pp is then the result of the multiplication of p by the average value of P

(Pav). Pp is also the “solution value”!
p = ∑ {[ (xi - ai ) / ci ] * fi }
P p = Pav * (p/100 +1)
5.2.10.5 Judgement of the quality
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

procedure described in 6.10.4) printability and the Pm points are the
measured (on printed samples with input of the quality factors in a Pi
equation) printability.
1
5,00
7 2
0,00
6 3
5 4
Pp Pm Pav
Pmin Pmax Pg
Figure 2: Visualisation of the Printability for different calculated Pp
5.2.10.6 After printing
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
Pm is with the P9 or P10 equations measured “Printability Coefficient”.

6. A journey through six papers
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.
6.1 The printing press parameter
Title: “Statistical Analysis of the possibilities and limits of

Flexographic Process Modelling.”
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 correlation and this validity have been confirmed by a

complementary study on other quality factors. The work was done on the
same samples for the factors mottling and edge sharpness.

6.2 The influence of the primary parameters
Title: “Analysis of the correlation between the print quality required

in Flexography and the tolerances in materials and paper
manufacturing in terms of printability.”
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.

6.3 Deeper in the paper characteristics
Title: “Study of paper printability in flexography by controlled

variation of the surface roughness.”
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 control of the surface roughness was obtained by calendaring the

substrate at different predetermined defined levels. The control
measurements were effectuated following the Bendtsen method. The
precision of the method could be discussed, but was appropriated in this
case due to the large range it offers and the intervals selected.
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.

6.4 A Flow diagram as directive line
Title: ”A Flex-Odyssey…”
To be able to understand the theory behind the numbers and particularly

behind the “Printability Coefficient”, it happened at this stage of the
research to be necessary to reveal and explain the flow diagram, which
the research is based on. The title of the paper reflects the course
flexography has followed the last 50 years. Flexography is today a high
quality printing process, which strives after a standardisation and needs
therefore tools like the “Printability Coefficient”.
The paper should present the flow diagram by answering two

fundamentals questions.
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 paper was a kind of platform to give the reader a simultaneous

summary of the work already done, an indication of the future work and
present the global strategy of the project. The flow diagram and the
answers to the two questions are the main results of this paper. The
missing answer is the answer to the question “How to use it?” This will be
the object of the final part of the work.

6.5 Establishment of a “Printability coefficient”
Title: “Establishment and Comparison of Different Definitions and

Process.”
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.
The quality factors are measurement and can be seen as secondary

parameters where the non-linear dependency is already eliminated. The
equation should be as simple as possible due to the finality of the final
product: an instrument for the industry and not only for the research.
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.
To be sure of the credibility of the equation, the equation was confronted

with a new data series.
The results were encouraging. The flexography disposes on a global and

objective quality control instrument. This is however not the final result
of the project and the research will be continued to the next phase. The
goal is a prediction of the printability.

6.6 The way back
Title: “Prediction of the Substrate Printability in Flexography by

using a new established Printability Coefficient.”
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?”
The user of the “Printability Coefficient” needs some references to be

able to compare his primary parameters. The references were fixed as the
mean values of the parameters used in the research work. The calculation
of the distance of the user’s values (xi) to the references values (a i) will
then be the first step of the calculation procedure. The obtained values
will be divided by the defined units (c i) and then multiplied by fi, the unit
influence coefficient. The sum on the index i is called “p”.
p = ∑ {[ (xi - ai ) / ci ] * fi }

The predicted “Printability Coefficient”, Pp is also the result of the
multiplication of Pav (average of “P”) by (p/100 +1).
P p = Pav * (p/100 +1)
The calculation of “P” is not sufficient to make a judgement of the

quality. The user needs a scale to be able to locate the values he has
calculated. The results can be confronted to the absolute quality or
theoretical optimal quality but it is more interesting to compare the Pp
values to the practical optimal quality. The centre of the variation interval
will then be defined as “at least to achieve quality” and called Pg.
As final consideration, the Pp points located in the area between Pg and

Pmax are the points researched.

7. Summary of results
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.
7.1 Technical results
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.
The principal technical results of the research can be resumed in the

following points:
• Correlation exists between the overall quality of the different

flexographic printing presses used in the production and the
modelling devices (Paper I).
• IGT F1 is, as a modelling device, a “standard” instrument in the
research world. The IGT is not the only construction on the
market, but, by respecting certain conditions, delivers satisfying
results, which can serve as start for further, more deeper
investigations: a good compromise (Papers I and II).
• The tolerances in material are not to be underestimated: a
receiving control can avoid trouble-shooting and save a lot of
time and money (Paper II)!
• An improvement of the paper surface roughness significantly
affects the printability. The influence is both direct, due to the
better surface, but also indirect, by modification of other
parameters like density or porosity of the substrate (Paper III).
• The Multivariate Data Analysis (MDA) is an instrument to get a
visual and fast overview of measurement data that can be used in
certain situations with acceptable results (Paper III).
• Mottling definition and measurement is a research project itself.
Different kinds and definition of mottling, different methods to
measure it coexist and the industry needs the establishment of a
standard.
• The flow diagram of figure 1 can help the user in the decision-
making action. The flow diagram is at the same time the map of

the project, the representation of the decision process and the
support of the explanation for the printability prediction
procedure (Paper IV).
7.2 Results for the process
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.
• It has been shown that an objective quantification of the

printability is credible. This quantification is an instrument for
the quality control. The objective quantification shows a strong
positive correlation with the ranking and scale of the subjective
quantification (visual perception) (Paper V).
• Equations are at disposal to numerically calculate the values for
the quantification: the “Printability Coefficient” (Paper V).
• A modelling of the printing parameters and of the paper
characteristics via the quality factors is the basis of the
printability prediction (Paper VI).
• A procedure for calculating “Pp” has been developed and the
final user has the possibility to use it as is or adapt it to his own
environment (Paper VI).
8. Discussion
The objective of the discussion part is to give the reader an instrument to

be able to judge of the quality and relevancy of the work. The reflection
should allow the reader to know how to use the results, how to profit from
them and eventually how to complete them.
The planed work was an ambitious project, according to the statements

from different experts. It represented a new approach to the quality
control in the flexographic process. The research works achieved and
published around the world before this research had the main objective of
studying one or several quality parameters or quality factors in detail and
the final results were helpful for one part of the industry. It was not a
question of a universal communication instrument. This work had the
ambition of establishing a technical and objective language between
different industries and level of users. As all “new products”, some

adjustments occurred during the research and further adjustments are
probably necessary for the generalisation of the utilisation.
8.1 The adjustments
The first adjustment, which could be seen as a restriction of the original

work ambition, is that the project “only” delivers procedures for the
measurement and the prediction of the printability and is not a complete
tool, as the end user could wish. The decision not to include the
measurement device in the final result was taken at an early stage of the
research. After an analysis of the market and contact with different
suppliers and developers, it has been clear that different measuring
devices for the quality factors are spread in the industry and that to fix the
utilisation of a specific one could have limited the universality of the
results. Moreover, standards for all the quality factors do not exist and the
work could not wait for the establishment of standards.
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.
8.2 Relevancy of the work
It is difficult to compare the results with results of preceding works due to

the lack of research in this specific area. The high quality flexography is a
new actor in the packaging industry and the reference for quality control
in the past had to be adapted. Moreover, a study of what has been done in
the printing industry shows very few similar works, which there should be
for other printing processes.
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:
• Credibility: The “Printability Coefficient” is a new quality

control instrument based on a scientific approach,

• Objectivity: A key number (P), and only one, can objectively
quantify the global quality of the printed result in flexography,
• Reality: The proposed number correlates with the visual
perception evaluation
• Economy: The prediction procedure eliminates the cost
expensive printing trials stage
• Simplicity: No need of complicated devices or high level of
education
The establishment of a key number is a necessary and important step to a

standardisation of the flexographic process.
8.3 Proposition for further work
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.
It can also relatively easily lead to the development of software, which

could be distributed by the different suppliers of the paper and printing
industries. A possibility should be to offer the customers the prediction of
a n e-service on the net: “Log in, enter your specific parameters, we
calculate the printability with our own product range”.
9. Conclusions
As conclusions to the research work it, seems important to remember the

main results and to give the future users the last advice for good usage of
the “Printability Coefficient”.
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.

As a basis of the research, the credibility of a key number has been
discussed. The characterisation of the significance behind the number has
been delimited by giving the printability a definition. Starting from this
definition, the architecture of the work, an equation has been built. The
equation was a partial result and represents the body of the printability
house. The testing and adjustment of the equation gave it the furniture.
The establishment of the prediction procedure completed the work by
delivering the key of the global quality control in flexography to the user.

Acknowledgements
I would like to express my sincere gratitude to a few persons who, during

the last past years, gave me opportunities, facilities, motivation, support
or love. Respecting this sequence, I would like to start by thanking
Anders Bovin and Prof. Nils Enlund for their engagement, making my
study in Sweden possible. I would like to continue by thanking the IMT /
Framkom management for the technical support. I formulate a special
thanks to the colleagues, who helped with my integration at the institute
and in the “cold” Sweden. I do not want to forget the “Hochschule der
Medien” in Stuttgart for its contribution in the final phase of my work.
An industrial work without industrial partners could not have been

possible. Therefore a particular “tack” to the companies that supported
my work by delivering materials, answering questions, making visits
possible… taking time in participating in my project.
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.

Literature cited and reviewed
Armel, D., Cusdin, G. (2001)

“ Plates & Printability ” – FQC FLEXO 04/2001
Aspler, J., Byron, J.,Zang, H. and Nguyen, N., (1998)

“Print Quality of Linerboard in Commercial Water-based Flexography” -
TAGA 98 proceedings, pp. 749-774
Bichard, W. (1992)
“The Inter-relationship among Air-Leak Roughness / Smoothness
Methods” - Pulp and Paper Canada, N. 93(6), pp. 147-152
Bookbinding and the Conservation of Books, A dictionary of Descriptive

Terminology
http://palimpsest.stanford.edu/don/dt/dt2682.html
Chabut, F., Girard Leloup, L. (2000)

“Influence of the printing press and paper surface roughness onto the
printability for the Flexographic Process” - Projet de Fin d’Etudes - EFPG
(F)
Chalmers, I.R. (1997)

“Flexographic printability of packaging grade papers” – Appita’97, pp.
335-341
Donderi, D.C., Aspler, J. (1999)

”Proscale: a fast new way to measure subjective print and paper quality” –
Preprint of PAPTAC Annual Meeting, Montreal, pg. B241
Foundation of Flexographic Technical Association, Inc. (1999)

“Flexography: Principle & Practices, 5t h Edition”
Ginman, R. and Tiainen P.-R., (1988)

“Requirements of the flexographic printing process on wrapping paper
and packaging board” - Adv. Print. Sci. Techno, N. 19, pp. 210-227
Girard Leloup, L. (2000)

Flexography and the tolerances in materials and paper manufacturing in
terms of printability.” - FFTA Y2K Forum - Orlando (USA)

Process Modelling.” - TAGA 2000 - Colorado Springs (USA)

“Study of paper printability in flexography by controlled variation of the
surface roughness.” - IARIGAI 2000 - Graz (A)

“Es tablishment and Comparison of Different Definitions and Equations
for a “Printability Coefficient” for the Flexographic Process.” – TAGA
2001 – San Diego (USA)

“A Flex-Odyssey…” - FFTA Forum 2001 – Nashville (USA)

“Prediction of the Substrate Printability in Flexography by using a new
established Printability Coefficient.” – APPITA 2002 - Rotorua (NZ)
Glossary of Graphic Communications - “Definition of Printability”
Glossary of paper terms, Southwestern Indian Polytechnic Institute

http://native.sipi.bia.edu/
Göttsching, Prof. Dr.-Ing. Dr.H.C.L., Ness, Dr.-Ing. C. (1997)

“Two-Dimensional Characterisation on the Surface of Paper with
reference to its Printability” – Institute für Papierfabrikation, Darmstadt
(D), EU Publication COST E11
I’Anson, S.J. (2001)

”The fundamental of digital image analysis and its application to paper
and board making problems” – http://pygarg.ps.umist.ac.uk/ianson
Jensen, K.W. (1989)

“Flexo Printability of Coated White-Top Liner with Consideration of
Convertability” - Graphic Arts in Finland N. 18, pp. 14-20
Karcz, J.P. (1998)

“Pinpoint Those Printing Problems” – Paperboarb Packaging (08/98), pp.
32-34

Lagerstedt, P., Kolseth, P.
“Influence of Surface Energetics on Ink Transfer in Flexo Printing” –
STFI, Stockholm, Sweden
Lindström, C., Dölling, R., Poustis, J.

“Evaluating the printability of liner board for flexography” - Paper
Technology (10/94), pp. 38-41
Lyne, M.B. (1979)

”Multidimensional scaling of print quality” – Tappi Journal (11/79), pp.
103-107
Meyer, Prof. K.H. (1996)

“Toleranzen in der Druckabwicklung und deren Auswirkung” – DFTA-TZ
Miller, B.W. (1997)

“Determining printability of liner-board manufactured for postprint
converters using water-based flexographic ink” – 1997 Process & Product
Quality Conference & Trade Fair, pp. 135-138
“Minutes” - Paptac Printing and Graphic Arts Committee Meeting –

10/1998 – Québec (Canada)
Murray-Davies equation: FD = [1-10(-D R)] / [1-10-(DV)] , D R is the

halftone density and DV the solid tone density
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
Pommice, J.C., Poustis, J., Lalanne, F.

“Testing the printability of board for flexography” – Paper Technology
(11/89), pp. 22-24
Print Process – www.printprocess.net
“Provtryckning 2000” DPC-FRAMKOM-STFI Research project

(Sweden)
Steadman, R., Woodall M. and Lesniak M. (1993)

“The flexographic printability of linerboard” - Appita’93, pp. 629-637

Thomson, D. (2002)
“Image formation in the eye” – http://www.city.ac.uk/optics/visual
perception/
Zang, Y.H., Aspler, J.S. (1995)

“Factors that affect the flexographic printability of linerboards” – Tappi
Journal (10/95), pp. 23-33
List of the experts, who participated at the survey “Definition of

printability”: Dijkstra, G.(WXS); Has, M. (OCE); Incontrao, R. (NAPIM);
Knudsen, C. (KTH); Lie, C. (KTH); Lindholm, G. (Tumbabruk),
Lindström, M. (IMT); Olsson, M. (Framkom); Parker, S. (Kalamazoo
Paper Chemicals); Streefland, W. (SCA Research); Wagner, L. (3M).

Paper I
Paper II
Paper III
Paper IV
Paper V
Paper VI

Statistical Analysis of the possibilities and limits
of Flexographic Process Modelling.
Girard Leloup, Laurent*
Flexography - Printability - Modelling - Correlation
Abstract:
Several research programmes are aiming at a standardisation of the

flexographic printing process. To obtain viable results, it is necessary to
distinguish between different steps in the scientific approach to the
problem. The first one is to formulate a ”useful” definition of printability
in flexography. This definition will then, at a final stage, allow a
statistical analysis based on parameters defined in advance. One of the
intermediate steps of the work consists of a comparison of the different
possibilities of modelling the flexographic process. The goal is less to
pass a judgement than to find a correlation between the different methods
for future investigations. In this study, we have compared 2 different
industrial production presses, a laboratory press and 2 simulation systems.
The results obtained are positive: the parameter printing press can, within
certain limits, be fixed as a constant for future investigations.
1. Objecti ves of the investigation
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.

2. Review of the modelling systems available in the market with short
description
An inquiry has shown that the number of available systems is not as

significant as expected. The reasons are once again the costs of the
development of such devices which can not be sold in large series and the
lack of knowledge of modelling the flexographic process.
References to 6 companies have been found which deal with development

and marketing of “modelling systems” for the flexographic process.
IGT Reprotest has developed the computerized F1 printability tester,

which consists of a combined inking unit with an engraved anilox roller,
doctor blade and a printing unit with printing form and impression
cylinder. The substrate is attached to a substrate carrier and placed on the
substrate guide, between the printing form and the impression cylinder.
With the aid of a pipette, a few drops of ink are applied to the nip
between the doctor blade and the engraved anilox roller. The ink is
transferred from the anilox roller to the printing form and from the
printing form to the substrate.
Two prints are automatically made since this the anilox roller is filled as
well as possible with ink. The second print will be used for the
evaluation.
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 Pamarco Flexo Proofer is a simple hand-held device, using an anilox

roller for dispensing a layer of ink on to the substrate. With the aid of a
micropipette, 250µl of ink are applied to the nip between the anilox roller
and the rubber roller. Then the proofer is drawn rapidly along the sample
to make the print.
The RK company supplies laboratory reel to reel presses. Their RK

Rotary Koater can be configured for flexography. The flexographic unit
consists of a printing head with micrometer adjustable pressure settings,
doctor blade assembly and an ink tray; also required are anilox rollers.
With sufficient drying, speeds up to 90m/min can be reached which allow
inks to be used at press viscosity.

The RNA-51 two shaft Printability Tester is a microprocessor controlled
unit. When fitted with an anilox roller and steel doctor blade assembly,
fast setting inks can be used. The motorized system keeps the anilox roller
in constant motion, thereby continuously bathing the anilox in ink and
subsequently doctoring it off.
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.
3. The selected systems for the trials in order to look at correlation
a. the printing devices:
In the following study we will compare 2 of the mentioned devices (IGT

and Saueressig), which represent the main families of modelling devices,
to production presses. The RNA system can be assimilated to the IGT F1
and the RK Koater to the Saueressig Flexo Proofer. The 2 other systems
reviewed, MacMillan Bloedel and Pamarco were not suitable for this
study due to the only partial modelling of the flexographic process they
offer. In the first case the system does not include any anilox roller nor
any photopolymer plate; in the second case no control of the pressure at
the nip and no constant speed are to be seen. Hence the choice not to
integrate these devices in the study.
As far as the production presses were concerned, the investigations were

carried out on 3 different presses.
The LEMO flexopress, located at the DFTA-TZ in Stuttgart (D) is a five-
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 second press is a Flexocompact seven-year-old, two-colour press

with a 600 mm width. This press located at the IMT in Stockholm (S) 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.
A new Soloflex eight--colours CI-Press with a 850 mm width was the
third flexopress in the programme. This press located at Windmöller &
Hölscher in Lengerich (D) represents a standard product on the market in

the middle-width range and is also equipped with chambered doctor
blades and an ink viscosity regulation system.
b. the data acquisition:
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.
4. Test methods and parameters
a. Description of the materials
a.1. the different paper qualities:
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
SEC Coated Liquid Packaging Board

SEU Uncoated Liquid Packaging Board
SCAWT1 White Top
SCAWT2 White Top
SCAPK Print Kraft
Table I: The different paper qualities
The chosen qualities represent 3 large product families. The

characteristics of the different substrates are presented in table II. It is

important to note the large amplitude of the paper characteristics: the
surface roughness (PPS) shows values between 1.9 and 7.2 µm, the
thickness between 157.0 and 269.0 µm and the weight between 138.7 and
193.0 g/m2. This dispersion will allow explanation of certain results and
eliminate the risk of error in the interpretation regarding correlation.
PPS Bendtsen Thickness Weight

(µm) (ml/min) (µm) (g/m2)
SEC 1.90 93 267.50 193.00

SEU 7.20 493 269.00 172.00
SCAWT1 6.45 277 157.00 138.70
SCAWT2 6.59 297 161.00 139.30
SCAPK 6.90 330 170.00 141.80
Table II: The characteristics of the different substrates
a.2. the different printing plates:
With the same objective of eliminating the risk of erroneous conclusions

due to coincidence, 4 types of photopolymer plates have been used for the
trials (Table III).
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"
Table III: the different types of plates
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.

Name: Hardness Thickness (mm)
(Shore A) min. max.
DPS 49 2.80 2.84

HOF 55 2.85 / 2.80 2.89 / 2.84
HOS 71 1.70 1.73
TDR 37 2.85 2.87
Table IV: the characteristics of the different plates
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.
a.3. the ink:
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).
a.4. the anilox rollers:
The anilox roller is a very important component in the ink transfer

process and is sometimes described as the “heart” of the process. The
problem is that different suppliers have different methods to engrave the
cylinders. This reality explains that the cells could have different forms,
depths and surface finishing and consequently the ink transfer will not be
the same! Moreover the anilox roller manufacturers use different ceramics
and deliver protocols with volume indications measured with more or less
accurate methods. In view of this problem and to moderate its influence, 2
different volumes have been selected for the trials: 8 and 12 cm3/m2. The
screen ruling, the exact volume and the supplier are shown in Table V.

Press: Screen ruling Volume Supplier
(l/cm) (cm3/m2)
LEMO 160 7.0/7.5 Zecher
100 13.0/11.6 Praxair
Saueressig 120 8.0 Saueressig
80 12.7 Saueressig
Flexocompact 140 7.9 Praxair
100 12.4 Praxair
IGT F1 180 8.0 IGT
120 12.0 IGT
W+H Soloflex 140 8.0 Zecher
120 11.0 Zecher
Table V: the anilox rollers characteristics
b. Description of the trials
The trials took place at different locations under external conditions

(Table VI) varying within a range not affecting the substrates and the
printing conditions. The procedure was the same for each trial, only the
planing was different for practical reasons. Depending of the press it was
easier to start by changing the plate, the paper or the anilox roller.
Press: Place Temp. Rel. humidity

(°C) (%)
LEMO DFTA-TZ 25-27 43
Stuttgart (D)
Saueressig DFTA-TZ 23 40-43
Stuttgart (D)
Flexocompact IMT 26 48
Stockholm (S)
IGT F1 StoraEnso 20 40-45
Falun (S)
W+H Soloflex W+H 23 50-55
Lengerich (D)
Table VI: the printing conditions

The first step, after checking the ink viscosity, was to obtain the Kiss
Print. The Kiss Print was important with regard to the rest of the trials
because fixing the reference for the following different pressures at the
plate-substrate nip. No standard method exists for obtaining the Kiss Print
and it is still a subjective procedure. To minimise the fluctuation,
elements have been tested but without success. Therefore it has been
decided that one person, the same for all the trials, would be responsible
for the “OK Kiss Print” with the help of a reference sample.
After realisation of the Kiss Print, it was printed at different plate-
substrate nip pressures: 100, 150, 200, 250 µm over the Kiss Print. The
printing speed was kept constant: 120 m/min. For each pressure 25
samples were cut after printing the complete series for the evaluation.
These samples represent approximately 1/10 of the total running time:
which, depending of the repeat length (Table VII), was about 45 to 60 s. It
was sufficient to allow the establishment of a stable situation.
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.
Press: Repeat length Width Speed

(mm) (mm) (m/min)
LEMO 480 1300 120
Saueressig 315 270 50
Flexocompact 600 600 120
IGT F1 530 50 18
W+H Soloflex 400 850 120
Table VII: Repeat length, width and speed

The particular features of the 3 production presses were that they were
equipped with chambered doctor blades and 2 of them had an ink
viscosity regulation system. In the case of the third one the viscosity was
manually controlled regularly. Drying was necessary for the 2 presses
with the short paper band before rewinding and the temperature was
maintained at between 70 and 75°C.
5. Results
The significant quantity of trials and measurements has produced a very

large database. To be able to make a relevant interpretation and a
comprehensible presentation of the main conclusions, it has been
necessary to extract only a part of the values. The goal of the project was
to analyse the correlation between production printing presses and
modelling systems. Different combinations with different types of plates,
anilox rollers, pressures at the nip and substrates have been tested and
evaluated.
Figures 1 to 5 represent the density variations measured for each press

with the 4 different plates. The pressure at the plate-substrate nip was 100
µm and the volume of the anilox roller 8 cm3/m2.
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
HOS HOF TDR DPS

Density Saueressig-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
HOS HOF TDR DPS
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
HOS HOF TDR DPS
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
HOS HOF TDR DPS

Density WH-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
HOS HOF TDR DPS
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.
The curves obtained show a large divergence between the comportment of

the uncoated substrates and the SEC coated one. For this last quality, only
the IGT F1 has a conventional density profile. This is probably due to the
fact that it was running at the lowest speed and also a more uniform ink
transfer. What is responsible for this abnormal result? The thickness of
the SEC is almost the same as the thickness of the SEU quality - it is also
possible to eliminate this parameter – and the study of the curves at the
200 µm pressure produced the same results. It seems also to be only the
coating which affects the result. This is confirmed by figure 11, which
shows the influence of the paper for the IMT press. In this diagram it is
easier to see the special comportment of the SEC substrate.

Density SEC-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
IGT Saueressig DFTA IMT W+H
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

Density WT2-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
2,80
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
SEU SEC WT1 WT2 PK
Figure 11: influence of the paper (plate DPS)
The diagrams demonstrate a good correlation between 4 of the 5 presses

and a less successful correlation for the last one. This does not signify
that it will not be possible to use the Saueressig Flexoproofer for trials in
the future but it is necessary to take into consideration the fact that the
density obtained is always about 20% higher than with the other presses.

The same conclusion is to be noted at the 200 µm nip pressure and with
the 12 cm3/m2 anilox roller. It is interesting to note one exception: the
curves of the IMT and Saueressig crossed at 85% for the 200 µm nip
pressure (figure 12). This is not due to a modification of the comportment
of the Saueressig flexoproofer but to high densities in the 0-85% range at
200 µm with the IMT press.
Regarding the correlation between the 4 other presses, it is necessary to
distinguish 3 different parts for the evaluation:
• 0 – 30% : the range of the density variation is on average 12%

between the IMT, DFTA and IGT presses and about 23% for the
W+H press. This means for a 0.35 average density, the IMT, DFTA
and IGT densities are between 0.33 and 0.37 and the W+H density
between 0.31 and 0.39
• 30 – 80% : in this part the density variation on average remains at

12% for the IMT, DFTA and IGT presses and is reduced to 16% for
the W+H press. (D=1.0 => 0.94<IMT,DFTA,IGT<1.06 and
0.92<IMT,DFTA,IGT,W+H<1.08)
• 80 – 100% : the last part of the curves shows a greater dispersion,

the percentage is common to the 4 presses and about 20%. (D=1.6
=> 1.44<IMT,DFTA,IGT,W+H<1.76)
A possible explanation for this divergence in the high tone area
could be the variation in the volume of the anilox rollers: the
printing was carried out with 8 cm3/m2 in all the presses but it
would be interesting to measure the volume of all the anilox rollers
with the same method!
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
Figure 12: the “crossing” at 200 µm

6. Conclusion
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:
Steadman, R., Woodall M. and Lesniak M.

“The flexographic printability of linerboard” – Appita’93
Chalmers, I.R.
“Flexographic printability of packaging grade papers” – Appita’97

“Evaluating the printability of liner board for flexography” – Paper
Technology (10/94)

Analysis of the correlation between the print
quality required in Flexography and the
tolerances in materials and paper
manufacturing in terms of printability.
Due to the fast development of new technologies, it is today necessary to

have another approach to the flexographic process: a better
comprehension and control of the process.
It could be a fatal error to think that the process control has to start when
you are printing: in some cases it is already too late! This control begins
at an early stage of the materials delivery. To be able to preview the
future print quality it is important to know the tolerances you can expect
from your suppliers for the delivery of all the different elements such as
the inks, anilox rollers, photopolymere plates, mounting tapes, sleeves
and substrates you are using and the influence of these tolerances on the
final printed result.
This paper presents the results of an investigation carried out with
suppliers and users in a project, a step towards standardisation, with the
goal of measuring the printability of different substrates in flexography.
I. The materials:
1. Printing plate:
The printing plate as a raw material is characterised by two quantifiable

parameters: its Shore A hardness and its thickness. Plate suppliers
distinguish tolerances between two productions, two charges and
tolerances within a plate or a box. The official statement is a thickness
variation of +/ - 0.010 to 0.015 mm within a plate or within a box. At the
same time they certify variation in the range +/- 0.025 mm between two
productions.
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

transfer and must be considered carefully. Suppliers indicate tolerances
within a range of +/- 2 °Shore A.
In this study we have considered 4 different types of plates and two
different thicknesses. Each plate has been made 3 times with the same
film. The thickness has been measured in 8 positions and the Shore A
measurement has been repeated 3 times.
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
Figure 1: Tolerance in plate thickness
The table above (figure 1) presents the results of the thickness

measurements: the average plate is the variation within the same plate and
the eight different positions measured, the average box is the variation
between plates from the same box and the average charge is the variation
between two deliveries of plates. If we compare the results with the
values given by the suppliers, we can observe that the variations within a
plate are as expected, the variations within a box are at the limits of the
tolerance area but still under the tolerances but the variations between two
productions are larger than mentioned by the supplier. The fact that the
tolerances between two charges are about +/- 0.040 mm instead of +/-
0.025 mm is not a dramatic problem but justifies a certain work
methodology.
It is of course interesting to know that variation in thic kness exists in the

plate and how large it is but the most interesting thing is to know the
influence of these variations on the final printed result. To study the

problem 2 plates from the same type and also with the same Shore A have
been selected and printed at the same time on the circumference of the
plate cylinder to be sure to have the same kiss-print. The specificity is
that these two plates have a thickness difference located in the middle of
tolerance scale: if we eliminate the case to print a job with two different
charges, the tolerances are +/- 0.015 mm and the middle tolerance 0,015
mm. Plate 1 and 2 had a thickness of 2.825 mm and 2.840 mm
respectively. Figure 2 shows the dot gain curve calculated with the
Murray-Davies formula for the two plates. The interesting part of the
curves is the middle tones part between 20 and 60%. In this part we can
observe a translation of the curve from plate 2 to a higher dot gain. The
study of the measured values quantifies this difference with an average of
4.7 %.
Tolerance in plate thickness
100
80
2,840 mm
60
% Print
2,825 mm
40
20
0
0 20 40 60 80 100
% Film
Figure 2: Dot gain and plate thickness

Tolerances in plate hardness
80
60
° Shore A
40
20
0
1 2 3 4
Type of plate
Figure 3: Hardness expected / measured
The diagram (figure 3) presents the result of the Shore A hardness

measurements. The values measured tie in very well with the values and
tolerances given by the supplier. The tolerances for the measured plates
are included between +0/-0 and -1/+2° Shore A, which conforms to the
+/-2 ° Shore previously mentioned.
In the worst case the difference between 2 plates could be 4° Shore A. In
order to see the influence or not of this factor, we printed, with the same
procedure and test form we used in the thickness investigation, two types
of plates with 8 ° Shore A difference and the same thickness.
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.

Tolerances in plate Shore A
100
80
45° Shore A
60
% Print
40
37° Shore A
20
0
0 20 40 60 80 100
% Film
Figure 4: Dot gain and plate hardness
2. Tape or foam tape:
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.

3. Ink viscosity:
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
Figure 5: Viscosity and density

Tolerance in sleeves wall
thickness
100
80
% Print
60
40
20
0
0 20 40 60 80 100
% Film
Figure 6: Dote gain and sleeve diameter
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!

5. Anilox roller:
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.
Tolerance in Anilox roller volume
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

6. Cylinders:
The impression cylinders, plate cylinder and anilox roller are

manufactured with tolerances in circularity and diameter. The
measurement of these tolerances was the topic a thesis at the HDM in
Stuttgart (D). The analysis of the result leads us to accept 5 to 10 µm as a
standard for both multi- or central impression cylinder presses.
II. The substrate:
The objective is to select the measurable paper characteristics which

influence the printability in flexography and to study the effect of their
variations on the printed results to be able to model the paper in terms of
printability.
1. surface roughness:
Surface roughness is one of the most important characteristics of the

paper regarding its printability. Surface roughness can be measured with
different procedures but the most common are the Bendtsen and PPS
methods both based on air leak measurement.The PPS method delivers
values in micrometers and the Bendtsen method in ml/min. Both methods
present advantages and disadvantages: PPS gives a good physical
representation of the surface topography but the Bendtsen method is more
accurate due to the wide range of the variations (400 ml/min instead of
5,3 µm). The variations in surface roughness influence the ink transfer
and are responsible for a form of mottling.
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

the ink, which is higher and suits the ink transfer mechanism better, to the
surface energy of the substrate.
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.
4. porosity / air permeability:
The porosity of a substrate is theoretically defined as the ratio of the void

volume within a porous material to its total volume. Due to the
difficulties involved in measuring this volume, the air permeability, which
is the rate at which air passes through the paper under specific conditions
is currently used in the paper and paperboard industry to quantify the
porosity of a substrate. Dalphond and Koller have shown a good
correlation obtained between the measurement of Gurley, Parker Print
Surf and Sheffield. However it is important to note that the measured air
permeability of a substrate is not suitable when comparing porosity of
substrates with different thicknesses, the thickness of the substrate
affecting the air permeability. These two conclusions, the correlation
between surface roughness and the non-suitability of the method for
thickness variations leads us to consider that the porosity of the substrate
is not a specific parameter of prime importance for the determination of
the printability. A good exploitation of the surface roughness
measurements and a consideration of the water absorbency will take care
of this parameter and also simplify the modelling of the substrate.

5. pH:
Water-based flexographic inks are generally formulated with resins

soluble in water under alkaline conditions (ammonia and alkanolamines
are used as solubilising agents) and they also show a pH range of 8.0-9.0.
The acidity of the substrate neutralises the solubilising amines and at the
same time a drying of the ink film occurs. This drying effect is suitable in
multicolour printing to eliminate the problem of the pearling of the ink
and assure a good ink trapping. However the relation pH of the substrate/
pH of the ink has to be kept under control to avoid the development of
tack in the ink films.
6. density of the paper:
Density is defined as the quotient weight/thickness and is given in g/cm3.

Paper density is a characteristic of its compressibility. Compressibility
plays an important role in combination with the plate/mounting tape
hardness for the cleanliness of the solid area and the edge sharpness.
The following diagrams (figures 9 to 14) represent density variations

measured for 5 different paper qualities (figure 8) with 2 different
printing plates, 2 anilox rollers and 2 pressures. For each diagram the
different elements have been kept constant and only the paper quality
varies. It is easy to observe that: first, the variations in paper properties
affect directly the printing quality (in this case the density), second the
variations are depending of elements such as the printing plate (hard/soft),
the theoretical volume of the anilox roller (8/12 cm3/m2) or the pressure
at the plate-substrate nip (100/200 µm).
Paper quality PPS Bendtsen Thickness Weight

(µm) (ml/min) (µm) (g/m2)
SEC Coated Liquid Packaging Board 1,90 93 267,50 193,00

SEU Uncoated Liquid Packaging Board 7,20 493 269,00 172,00
WT1 White Top 6,45 277 157,00 138,70
WT2 White Top 6,59 297 161,00 139,30
PK Print Kraft 6,90 330 170,00 141,80
Figure 8: papers characteristics

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
SEU SEC WT1 WT2 PK
Figure 9: hard plate – 8 cm3/m2 – 100 µm

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
SEU SEC WT1 WT2 PK

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
SEU SEC WT1 WT2 PK

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
SEU SEC WT1 WT2 PK
Figure 12: soft plate – 8 cm3/m2 – 100 µm

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
SEU SEC WT1 WT2 PK

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
SEU SEC WT1 WT2 PK

Conclusion:
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
Meyer, Prof. K.H

”Toleranzen in der Druckabwicklung und deren Auswirkung” – DFTA-TZ
Zang, Y.H., Aspler, J.S.

”Factors that affect the flexographic printability of linerboards” – Tappi
Journal 10/95

Study of paper printability in flexography by
controlled variation of the surface roughness.
Laurent, Girard Leloup

framkom – KTH, Stockholm, Sweden
Introduction:
The flexographic printing is in a period of rapid development. The

process achieves a high quality and allows to print of a large range of
different substrates. These substrates show very different properties. The
quality of the printed result related to the substrate properties and to the
printing parameters is called printability. In a first step of the research
work, the influence of the variation of the printing parameters has been
studied. This paper presents the results of the second phase of the work,
which intend to correlate the paper properties to the printability. The
problem is that the different paper characteristics are dependant variables
and interact. It is also very difficult to study all of them at the same time.
Therefore the chosen way was to control the surface roughness of the
substrate by calendering the paper in several steps.
Methods:
To be able to integrate this study in the research concept about the

establishment of a “Printability Coefficient“ it was necessary to keep the
same type of paper we used in the first step. The paper was an uncoated
Liquid Packaging Board (SEU) with a basis weight varying between 232
and 236g/m2 and a thickness varying between 0.382 and 0.398mm. Both
reels from the middle and reels from the side of the jumbo reel have been
used. The paper was coming from a production and was a conventional
product without special modification for the investigation. After the
measurement of the initial Bendtsen surface roughness values, it has been
renounced to the idea to calender at the same time a white top testliner.
This paper showed almost the same values for the middle and the side
reels for both 140g/m2 and 200g/m2 qualities. The variations measured
were in the range of the tolerances of the Bendtsen process. However the
white top testliner (WT1) has been used to compare the behaviour of 2
different paper qualities with the same surface roughness.

Calendering was performed at the STFI (1) on the pilot calender. The
calendering was a steel/steel nip calendering as used in the production of
this type of paper. The centre roll was heated to 200°C and the speed was
about 300m/min. The paper was calendered using different line loads. The
first rolls were used to find appropriate line loads giving the requested
Bendtsen values. Five line loads were tested and Bendtsen values for
those levels were measured. The following reels, which were destined to
be printed, were then calendered at the line loads given by the first trials.
The aims of the calendering were first to obtain paper with constant
values of Bendtsen surface roughness and second to create a possible
comparison between edge and middle reels, liquid packaging board and
testliner qualities.
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 testform was a new one, developed to allow the measurement of

several printability factors. For this study we only considered the 34l/cm
screen ruling. Measured were the density for 18 different tonal values (1-
100%), the mottling for a solid area and for a 30% screen area, the
bleeding, the wicking and the dot gain for the 50% tonal value. For each
printing plate/anilox roller/substrate combination, 3 different pressures at
the plate/substrate nip were printed. The first one was the Kiss Print to
define the “zero point“, then the pressure has been increased to 200µm by
2 100µm steps. After stabilisation of the printing conditions, samples has
been collected at each stage for the future measurements.

The measurements were effectuated in 2 steps. The first step was to
measure the optical density with a Gretag Macbeth D19C densitometer.
For each series the measurements were carried out on 3 samples and
exported to an Excel sheet using the Key Wizard interface. The average
values have been used for the evaluation. The second step was the
measurement of the mottling, wicking, bleeding and dot gain values using
the Papeye (3) software. The method is based on an image analysis
system with a scan of the sample as input for the different calculations.
The mottling analysis module gives a mottling index as result. This index
is calculated with the following procedure: The sample is scanned with a
resolution of 300 dpi, the application of a wavelet mathematical
transformation allows to calculate values for different resolutions, then
using a double weighting of the different points (a first weighting taking
care of the resolution of the human eye and second one to enhance the
local maximum), an average is calculated. This average is the called
“mottling index“. This index varies between 0, corresponding to the
mottling of a plastic surface printed with a neutral colour and 100 to a
chequered black/white surface with edge length of 2.7mm.
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.

Results:
First it is necessary to mention the paper qualities obtained by the

calendering. The following 7 paper qualities has been used in the rest of
the study:
SEUE LPB Edge reel not calendered

800ml/min
SEUE1 LPB Edge reel calendered at 75 kN/m
550ml/min
SEUE2 LPB Edge reel calendered at 130 kN/m
300ml/min
SEUM LPB Middle reel not calendered

550ml/min
SEUM1 LPB Middle reel calendered at 75 kN/m
300ml/min
WT1E WT Edge reel not calendered

300ml/min
WT1M WT Middle reel not calandered
300ml/min
(LPB: Liquid Packaging Board - WT: White Top 140g/m2)
Influence of the paper surface roughness on the printed density:
The achievable print density is a function of several parameters from both

the printing press and peripheries and the substrate properties. In this
study we analysed 8 printing plate/anilox roller/pressure combinations
confronted to 7 different paper qualities. The combinations called series
are the numbers of observations and the paper qualities the variables in
the following diagrams.

Influence of the PSR on the density - DPS
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
WT1E WT1M SEUE SEUE1 SEUE2 SEUM SEUM1
Figure 1: Full tone density – DPS plate
Influence of the PSR on the density - HOS
2,30
2,20
2,10
2,00
Density
1,90
1,80
1,70
1,60
1,50
1 2 3 4
Serie
WT1E WT1M SEUE SEUE1 SEUE2 SEUM SEUM1
Figure 2: Full tone density – HOS plate
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

trend is to see for the middle reels except for the serie 3 with the DPS
plate!
However the figure 3 illustrates that the density variations are not so
significant: the average is 0.07 for an average full tone density of 2.01,
which represents a variation of 3.5%.
Influence of the PSR on the print density
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
Figure 3: Density variations for tonal values between 0% and 100%
Influence of the paper surface roughness on the motlling:
The mottling is an important parameter to quantify the printability. The

mottling measured in this study corresponds to the “global” mottling. The
global mottling is the residual mottling from the substrate modified by the
print mottling. The used method to calculate the mottling index allows
only in a few special cases to distinguish the both types of mottling. It is
only if the unprinted and printed papers have the same mottling curve
profiles, that the possibility exists to separate the contributions of the
paper and print mottling. In these cases the mottling index is a value
usable for comparison or global statements.
- Solid area (100%):
The figure 4 is a representation of the average mottling index measured in

a solid area for each paper qualities. The values are varying between
11.87 and 12.92 which are quite high values but a narrow range of
variation. Regarding the ranking of the different qualities and the
correlation with the surface roughness the measurements prove that the

calendering is influencing the mottling index. The variation of the
mottling is opposite to this of the surface roughness: the improvement of
the surface quality creates an increasing of the mottling for both edge and
middle reels. The comparison of the 4 qualities with the same
(300ml/min) Bendtsen value conducts to the conclusion that it exists a
correlation between the surface roughness and the mottling index. But the
Bendtsen characteristic is not the only one factor: it has been measured a
0.65 difference (12.27-12.92) between the minimum and the maximum
values.
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
Figure 4: Mottling index for a 100% area
- screen area 30%
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.

Mottling screen 30%
10,00
9,80
9,60
9,40
Mottling index
9,20
9,00
8,80
8,60
8,40
8,20
8,00
Figure 5: Mottling index for a 30% area
Influence of the paper surface roughness on the edge sharpness of the

print:
- wicking:
To measure the wicking sharpness of a motive (text, line) is a method to
quantify the edge sharpness for positive elements. A low wicking is
synonym to a high sharpness. The figure presents the wicking sharpness
in %. The values fluctuate between 0.87 and 1.01% (figure 6). For the
edge reels the wicking reduces with the improvement of the surface
roughness what is the opposite for the middle reels. In this case it would
be interesting to control the absorption of the paper because it seems that
the effect of the absorption and the surface roughness are combined. A
too closed paper structure contributes to a spreading on the ink, which is
in this case is intensified by the calendering (see SEUM/SEUM1).
Wicking
1,05
1,00
0,95
%
0,90
0,85
0,80
Figure 6: Wicking measured for the different paper qualities

- bleeding
The bleeding can be considered as the negative effect of the wicking. The
bleeding is used to characterise the edge sharpness of negative motives. In
combination with the wicking the bleeding creates an “absolute”
sharpness. The trends for the bleeding are the same as for the wic king,
only the amplitude of the variations is greater (3.6% / 0.14% - Figure 7).
This is due in major part to the deformation of the fotopolymer printing
plate.
Bleeding
6,00
5,00
4,00
3,00
%
2,00
1,00
0,00
Figure 7: Bleeding measured for the different paper qualities
The “absolute” sharpness is built by calculating the average of the

bleeding and wicking sharpness. The value confirms the results of the
partial sharpness with the same ranking and values from 1.42 to 3.26%
(figures 8).
"Absolute" Sharpness
6,00
5,00
4,00
%
3,00
2,00
1,00
0,00
Figure 8: “Absolute” sharpness for the different paper qualities

Influence of the paper surface roughness on the dot gain in middle ton
(50%)
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.
DG= [(1-10E( -D50%)) / (1-10E( -D100%))] – 50 [%] (1)
with: DG: Dot Gain

D50%, D100%: Density for respective the 50% and 100
%areas
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).
DG= [(WA/1156) x 50/ TPA] – 50 [%] (2)
with: DG: Dot Gain

WA: Wicking Area
TPA: Theoretical Point Area
The method using the Murray-Davies formula gives variation from

10.05% to 16.75%, the values with the geometrical method are varying
between 23,51% and 31.18% (figure 10). The correlation between both
methods can be seen on the figure 10: the trends for the variations are
exactly the same and the rapport geometrical/optical dot gain is about 2
for all the paper qualities. Regarding the fluctuation of the dot gain, a
better surface topography leads to a reduction of the dot gain but this
result has to be nuanced by the fact that the amplitude of the variations is
limited (about 2.5%).

Dot Gain 50%
35,00
32,50
30,00
% Print
27,50
25,00
22,50
20,00
Figure 9: Geometrical Dot gain for the different paper qualities
Dot Gain comparison
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
Serie1 Serie2
Figure 10: Optical / Geometrical Dot gain for the different paper qualities

Discussion:
The figures 11 to 16 have been created using the SIMCA P8 multivariate

data analysis software. These have for aims to obtain a graphic
representation of the printability of the different paper qualities. The
diagrams have to be considered as scores/loadings pairs. In the loadings it
is possible to localise the position of the different parameters which has
been measured (density, mottling index, wicking,…). The lecture of the
scores informs about the quantitative influence of these parameters onto
the paper quality plotted. The centre of the ellipse would represent the
ideal point or the best compromise without to take care of the density.
However the achievable density can not be excluded from a printability
evaluation. Therefore the 3 different pairs of plots (scores/loadings) with
the density for the respective 2 (“trame 2%), 50 (“Eng 50%”) and 100%
(“Aplat”) tonal values. To take into consideration the fact that all the
parameters except the density are expected minimum and the density
maximum, it is necessary to consider the line joining the centre of the
ellipse to the density point in the loadings plot. The point P situated on
the line at 1/5 to the centre of the ellipse is defined as the ideal point. The
distance d between the point P and the paper points plotted in the scores
plots is characterisation of the Printability.
Using this definition of the printability and a graphical evaluation, the
results are very encouraging. For each density the WT1M, SEUM1
qualities are the nearest points and for SEUE2 quality the result is similar
for 2 densities. This would lead to the conclusion that the 3 precedent
qualities are the best printability.

Roughness – Solid area density
Distribution of papers (average)
3
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)
Figure 11: Scores for the 100% Density Printability evaluation
Roughness – Solid Area density
Eng 50%
0.60
0.40
% W Sharp
0.20
p[2
]
0.00
-0.20
Mottling Apl
-0.40 Mottling 30%

Aplat
-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]
Figure 12: Loadings for the 100% Density Printability evaluation

Roughness – 50% screen density
Distribution of papers (Average)
3
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]

0.90
% W Sharp
0.80
0.70
0.60
0.50
0.40 Eng 50%

p[2]
0.30
Trame 50%
0.20
0.10
0.00
-0.10
Mottling 30%
-0.20 Mottling Apl
-0 .5 0 -0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 0.30 0 .4 0 0.50
p[1]
Loadings: p[1]/p[2]

Distribution of papers (average)
3
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]

0.90
% W Sharp
0.80
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]

Conclusion:
This study has shown that by influencing different components of the

printability matrix, the surface roughness (Bendtsen) of the substrate has
a great contribution to the quality of the printed result. The contribution is
positive for all the measured parameters excepted for the solid area
mottling index. This leads to the conclusion (confirmed by the discussion)
that an improvement of the surface roughness improve at the same time
the global and objective printability of the substrate. Further research
works will be carried out to complete the results and quantify the part of
the surface roughness influence in the total influence of the paper
parameters.
Acknowledgements:
I would like here to thank Francoise Chabut, Student at the EFPG in

Grenoble, F for her help in this project by doing measurements and
discussing the results.
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): STFI: Swedish Pulp and Paper Research Institute

(2): framkom: Research Corporation for Media and Communication
technology, Stockholm, Sweden
(3): Papeye: a software from Only Solutions (Kassel – D)

Bendtsen Roughness: The Bendtsen tester, first introduced in 1939,
measures the rate of air-flow, in ml/min, between the paper surface and a
narrow annular surface, which rests on the paper against a smooth glass
plate. The air supply is maintained at a constant pressure and, with the aid
of an air-flow meter, the instantaneous rate of the airflow across the paper
surface is measured. The resistance to the airflow between the annulus
and the paper surface decreases with the roughness of the paper.

A Flex-Odyssey…
Keywords: Flexography - Printability Coefficient – Flow diagram -

Quality Control – Objectivity - Measurement
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

2. Objectives of the investigation
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
3.1 How to read it?
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 is a representation on paper of the way a job is planed
and executed in the industry. A printing product has been concept and has
to be realised. It could be a simple or a very complicated item but the
approach of the work and the way to follow to the final product are
similar.
Before running the printing press, have to be decided:
- the substrate and its properties

- the press configuration and its peripheries
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”.
3.2 Contribution of the different “boxes”:
3.2.1 Paper characteristics:
a. factors to take into consideration:
. 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

does normally not interest the creative people but the technical
people have to find a solution to be able to realise the product!
. The economical parameters: the costs have to be minimized. The
packaging represents in Europe in average 6% of the final product
selling price and the substrate represents a large part of the packaging
costs. Some customers do only look at the quality of the final
product: in this case the decision is very simple and the choice is
oriented to the substrate with the best printability. However the
majority of the customers are looking for the price and need an
instrument to select the best compromise: a “Printability
Coefficient”.
. The printing process: in the choice of the paper is the printing
process a relevant player. It is necessary to know in advance which
technical constraints will appear in the printing process. The type of
press (CI, stack, multi cylinder), the ink system (water based, solvent
based, UV) and the drying technology are some of the most important
parameters to take into consideration.
. The converting exigencies: not forget that the printed product is not
always the final product! The paper, board has by its density,
thickness, … to guaranty the transformation (die cutting, gluing, …)
of the printed product into a final product without to damage the
technical and esthetical qualities of the before achieved result.
. The motive to be print: depending of the required quality and the
difficulty of the motive to be print certain paper properties become
more or less importance: typical example are the whiteness or the
surface roughness of the paper .
. Skill of the company / employees: the strategy of the company can
here be the critical point: What is the best investment or the
investment of high priority: invest for the education and motivation
of the employees and save money by buying a cheaper paper or pay
the price for a high quality paper to obtain the same result? Is it
really the same result? Is a compromise the right alternative?
b. the quantificators for the substrate:
This part of the diagram has been described in a precedent paper

under the title: “Analysis of the correlation between the print quality
required in Flexography and the tolerances in materials and paper
manufacturing in terms of printability.”
The substrates properties, which influence directly the printability are

the following:
. surface roughness
. water absorbancy

. surface energy
. porosity / air permeability
. pH
. homogeneity of the surface (mottling)
. density
Some other characteristics of the substrate should influence the

printability but indirectly: it means these characteristics like paper
formation, relative humidity, … are secondary variables, which
change the values of the above primary variables.
3.2.2 Printing parameters:
a. the anilox roller:
. The choice of the anilox roller depends of the substrate and

its characteristics, the nature of the motive to be print and the
specifications of the prepress work (screen ruling, colours
separation, colour management…). To define are the angle of
the screening, the theoretical volume of ink transferred, the
cell geometry (it is a directly consequence of the engraving
technology used by the supplier) and the screen ruling.
b. the press:
. The press must be adapted to the type of product to be print.

The size of the product, the substrate, the inking / drying
systems, the design and the quality required are the relevant
factors. A further parameter is the control of the pressure at
the nip printing plate / substrate: this adjustment is not really a
characteristic of the press but a problem at the interface
printer/printing press. A well-equipped press (with CNC
motors) will reduce the risk of error from the user. Not to
forget is the next step of the work and its requirements: the
converting.
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

adapted and a stability of these parameters during the
production has to be assured.
d. the printing plate:
. The printing plate summarizes the work down in the prepress

stage: it is the support of the information to be reproduced and
the rightness of this information has to be proofed. The
content and the integrality of the information such as the
rightness of the technical data (screen ruling / dot size, dot
shape, relief depth, …) are primordial factors for a good
result. Moreover the printing plate is by its own characteristics
an important element in the ink transfer process. The
composition of the photopolymer, its hardness and the
thickness of the printing plate as well as its relief depth allow
with different mounting techniques several combinations for
an optimisation of the ink transfer.
3.2.3 Printing trials / Printed samples:
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.
3.2.4 Measurement Tool:
The measurement tool is based on optical measurements of the different

test charts and test elements present on the printed sheet. The
measurements are effectuated with the help of densitometers,
spectrophotometers and image analysis hardware and software. A detailed
description of the different methodologies can be found in 2 precedent
publications: “Study of paper printability in flexography by controlled
variation of the surface roughness” and “Provtryckning 2000”.
3.2.5 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

panels and an objective evaluation done by several measurements and
measurement technologies. The selection of the quality factors has been
realised by first looking at all the measured parameters and then
eliminating the non relevant or with less influence for the flexographic
process. The result is a confirmation of the by the flexographic experts
expected result. The quality factors, which passed the tests are the
following: density, dot gain, dot deformation, mottling and edge
sharpness.
3.2.6 Modelling of the Print / Modelling of the Substrate:
Both modelling are complementary and must be considered as a level

itself in the flow diagram. They will allow to jump over the printing trials
and measurement of the quality factors steps and consequently to reach
the prediction stage of the project. By choosing a “Printability
Coefficient” and using the modelling boxes, it will be possible to define
the substrate characteristics and printing parameters corresponding to the
wished quality.
3.2.7 Printability Coefficient:
“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.

4. Discussion
“A Flex-Odyssey…”: the title of this paper tell us that the Flexography is

an interesting and attractive challenge but at the same time still an
adventure. The strategic development of an industry is often based on
risks but the control of the risks is necessary to achieve the goals. Goals
for the flexographic printing are diversity, competitiveness and quality.
The tools to achieve these different goals are not yet all present on the
market. The “Printability Coefficient” with its flow diagram as handbook
should be one more and not only one more but also a new type of
instrument. The efficiency and simplicity of the instrument will give it a
good acceptance in an industry where the education level shows very
large discrepancy.
"P" - "Printability Coefficient"
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
Figure 2: Representation of the ”Printability Coefficient”
The precedent figure shows a representation of the “Printability

Coefficient” by variation of the quality factors. The upper and over lines
represent the limits (min/max) of the achievable printability, the middle
line the average for the 15 values plotted. The user has the possibility to
draw a further line representing the “target or goal printability” (figure 3)
and can also very easily read on the figure for which combinations the
printability will be acceptable or not.

1,60
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
Measured Average Max Min Target
Figure 3: Representation of the printability with the “Target Printability”
5. Conclusion
The existence of instruments for a measurement and prevision of the

printability by a “Printability Coefficient” has been presented. The flow
diagram explains the way “P” will be calculated and proves the logical
and at the same time closed to the real work way it is based on. The
establishment of the equation has been shown previously and the figures
in the discussion part illustrate it. The only part missing is the procedure
“How to use it?” for “P”, the “Printability Coefficient”. This and further
improvement of the equation will be presented in the final results of the
work.

References:
(1) Girard Leloup, L.

terms of printability.”
FFTA Y2K Forum - Orlando (USA)

Process Modelling.”
TAGA 2000 - Colorado Springs (USA)

surface roughness.” - IARIGAI 2000 - Graz (A)Steadman, R., Woodall
M. and Lesniak M.
(4) “Provtryckning 2000” Research project (Sweden)
“The flexographic printability of linerboard” - Appita’93


Technology (10/94)

Establishment and Comparison of Different
Definitions and Equations for a “Printability
Coefficient” for the Flexographic Process.
Keywords: Flexography - Printability Coefficient - Equation - Quality

Control
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.
1. Objectives of the investigation
The work presented in this paper should be seen as a contribution to the

actual effort for a standardisation of the flexographic process. The main
objective of the investigation is to deliver the different actors in the
packaging industry a key number to be able to objectively and with a
guaranty of repeatability evaluate the quality of a printed product. This
number called “Printability coefficient” is going in a first time to be used
as an instrument in the quality control stage of the process and in a second
time to be a tool for the conception phase of a new printed product. It will
facilitate the dialogue between the different partners and also reduce the
lost of time and money due to a luck of a universal quality language.
_________________
*Research Corporation Media and Communication Technology

2. Methodology of the investigations
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).
2.1 Preparation of the printing trials:
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).
In this study will be mentioned other printing processes like lithography,

electrophotography and inkjet. These processes are not the main focus of
the present report but have been used as references for the establishment
of the different printability coefficients. Some of the printing trials
evaluated for the work are described in details in a project called
“Provtryckning 2000” (4).
The preparation of the trials consists to a selection of different substrates,
a definition of the printing parameters, an elaboration of the printing
procedure and the coordination of the different research resources
involved. The substrates range covers with 11 qualities (figure 1) from
matt and gloss paper, uncoated liquid packaging board, liner to high
coated boards a broad palette of the flexographic products.

N. Designation Specification
1. SEUE LPB-Edge reel not calendered
2. SEUE1 LPB-Edge reel - calendered at 75 kN/m

3. SEUE2 LPB-Edge reel - calendered at 130 kN/m
4. SEUM LPB-Middle reel - not calendered
5. SEUM1 LPB-Middle reel calendered at 75 kN/m
6. WT1E WT-Edge reel not calendered
7. WT1M WT-Middle reel not calandered
8. IG Highly coated matt paperboard

9. IA Highly coated gloss paperboard
10. SB Gloss Highly coated gloss paper
11. SB Matt Highly coated matt paper
(LPB: Liquid Packaging Board - WT: White Top 140g/m2 – I: High

quality board 220-250g/m2 – SB: Graphic Art paper 130g/m2)
Figure 1: Table with the different substrates
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.
2.2 Printing of the samples:
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.

The second procedure (qualities 1-7) is the direct printing of a known test
form by varying the ink transfer parameters. For different ink quantity, a
range of nip pressure has been printed with a cyan water based ink. The
results are “good”, “less good” and “bad” samples, which allow a scaling
of the printability coefficient.
2.3 Measurement of the printability parameters:
To be able to establish a printability coefficient a large volume of

measurements are necessary. The measurements methods are already
described in precedent papers (3) and (4). The results of the
measurements are going to be presented with diagrams and tables will
show the principle values. All the values are available but not presented
here!
The measured parameters are:
- 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.
Not all the parameters will be retained in the “final” printability

coefficients but all of them have been available and tested to find the best
compromise.
2.4 Establishment and comparison of different equations:
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

process. The obtained result will then confronted to the qualities 1-7 by
varying the printing parameters and comparing the values with the
expected results.
3. Results
3.1 The results of the measurements:
In this part will be presented the results of the different measurements.

The goal of this study is not to find the combination substrate / printing
parameters to obtain the best quality but to establish a relevant printability
coefficient. Therefore it does not appear necessary to present the results
of the measurement as raw data but more interesting to show the variation
in the results for the different measurements. The amplitude of the
variation should be seen as a quality factor for the printability coefficient.
All the following diagrams (figures 2 to 6) has been build on the same
principle: the minimum, the maximum and the mean value of each
measured parameters have been picked up, then the difference between
the max and min to the mean value have been calculated in % of the mean
value and plotted by property for the different substrate quality groups.
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
Qualities 8-11 Qualities 1-7
Figure 2: Amplitude of the Dot Gain variation for the different substrates

The dot gain diagram shows a regular repartition of the amplitude on
both side of the mean value for all the measurement except for DG80K
and DG50C. The summed amplitude is about 200%.
Density
120,00
100,00
100,00
90,34
80,00
60,00 54,76
40,14 39,40 40,43

40,00 32,39
30,41
28,12
24,93
21,39 21,30
20,00 12,69 12,87
0,00
C M Y K C2% C50% C100%
Figure 3: Amplitude of the Density variation for the different substrates
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.

Edge Sharpness
437,68
450,00
400,00
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
Figure 4: Amplitude of the Edge Sharpness variation
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
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%
Figure 5: Amplitude of the Mottling variation for the different substrates

The mottling diagram can be interpreted at 3 levels. The first conclusion
is only a confirmation of the large influence of the printing process on
the mottling values: the max values are very far away from the mean
values. A second observation is that the minimisation of the mottling is
not quantitatively dependent from the colour, trapping or screening of the
measured surface. Moreover the mottling is a factor, which is for a
specific process (flexography in this case), very stable in the amplitude of
the variation: 13.53 - 17.63% for min/max C and C30%.
Other Parameters
985,55
1000,00
900,00
808,86
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
Figure 6: Amplitude of “other parameters” variation
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!
3.2 The establishment of the printability coefficient:
3.2.1 Definition of the printability
The printability of a substrate has different definitions depending who is

speaking and in which context. The paper industry has of course not the
same definition as the IT industry but not either the same as the printing

industry. However are these 3 industries (at least 2 of them) working in
very eng connection. The admitted definition of printability for this work
is the following: Measurement of the printed result quality related to the
substrate properties and the printing parameters.
3.2.2 Different equations
a. correlation to the virtual perception data:
The establishment of a printability coefficient for the flexographic

process has been done in successive steps. The first step was to find the
best linear combination of the measured printing quality parameters,
which fits with the visual perception (VP) quality evaluated for the
flexographic samples. The goal was to obtain an equation valid for the
flexographic process by minimising the distance between both printability
(measured/calculated and visual) but at the same time keeping the shape
of the printability curves parallel for the other processes. The results of
the different printability values (P) are presented in the figures 7 to 17.
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
Figure 7: Plot P1 – Visual Perception

P1 is a linear combination of all the quality parameters measured for the
substrates 8-11. For the parameters with several measurements, like
densities, mottling, … the arithmetic mean has been calculated and used
as input for the equation.
P1=1- [(1/(max ∑αi A i )) * ∑αi A i ] (5)
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

Distance to VP printability.
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
Figure 9: Distance P2 to the Virtual Perception printability
P2 is a simplification of P1. The parameters qualified as “other

parameters”: gloss, colour gamut, colour failure and grey balance has
been eliminated from the equation. The αi coefficients are still +/- 1. The
figure 9 shows the distance of P2 to the visual perception printability. All
the values are in a range between -0.16 and + 0.15. These values show
that the equation can be used as basis for a credible printability
coefficient.

P3 - VP
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
P3 is a trial to take in consideration by an other way the fact which high

values for density, colour gamut and gloss are suitable for a good quality.
P3=1- [(1/S * (∑(1−A i )+ ∑A j )] (6)
with S= max (∑(1−A i)+ ∑Αj )
A j : mottling, blurriness, raggedness, dot gain, colour failure,

grey balance
A i: density, colour gamut, gloss

P4 - 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
P4 is built on the P3 model by the same principle as P2 on P1.
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

P5=1- [(1/(max ∑αi A i )) * ∑αi A i ] (7)
with
αi = +1 for mottling, dot gain
αi = -1 for density
αi = 1/2 for blurriness, raggedness
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
P6=1- [(1/(max( ∑αι Αι )) ∗ ∑αι Αι ] (8)
with
αi = +3 for mottling
αi = +4 for dot gain
αi = -4 for density
αi = +1 for blurriness, raggedness
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).

b. Definition of a flexographic specific printability coefficient:
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
The figure 14 is only here to show the necessity of a printability

coefficient. The interpretation of such a diagram is impossible and a
method is needed to visualise the results.

P7
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
Figure 15: P7 for different plate/anilox/nip pressure/substrate

combinations
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:
- uniform coverage: density, mottling solid area, mottling 30%

- line quality/edge sharpness: bleeding, wicking
- dot quality: dot gain, roundness
P7 = ∑αi (1−(Αi /max Α i )) (9)
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

Printability
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
+1 for dot gain

WT1M-DPS08100
the same weighting (=1).

WT1M-DPS08200
WT1M-DPS12100
WT1M-DPS12200
SEUE-HOS08100
SEUE-HOS08200
roundness 2% and 30%

SEUE-HOS12100
SEUE-HOS12200
SEUE-DPS08200
SEUE-DPS12100
SEUE-DPS12200
SEUE1HOS08100
SEUE1HOS08200
P8
SEUE1HOS12100
SEUE1HOS12200
combinations
SEUE1DPS08100
SEUE1DPS08200
P8 = ∑αi (1−(Αi /max Α i ))
-1/3 for density 2%, 50% and full tone

SEUE1DPS12200
SEUE2HOS08100
SEUE2HOS08200
SEUE2HOS12100
SEUE2HOS12200
MAX: 4.02 – MIN 1.14
SEUE2DPS08200
SEUE2DPS12100
SEUE2DPS12200
SEUM-HOS08100
SEUM-HOS08200
SEUM-HOS12100
SEUM-HOS12200
(10)
SEUM-DPS08100
Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

SEUM-DPS08200
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
same contribution for a quality quantification and to give them together

quality parameters. The principle of the weighting is to regroup factors of
+1/2 for mottling 30% and solid area, wicking, bleeding,
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
Figure 17: Comparison P7-P8 printability for different plate/anilox/nip

pressure/substrates combinations
4. Discussion
The investigation has delivered 8 equations for a printability coefficient.

The way the work has been conduced gives a logical orientation to the
discussion for the establishment of a printability coefficient for the
flexographic process. But before to study in details the final propositions
P7 and P8, it is interesting to have a look at the P1-P6 coefficients. P1-P6
has been obtained by trying different linear combination of selected print
quality parameters. The amplitude of the variation (see results part I) and
the large number of substrates tested allow to make a credible analysis.
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

deviation is almost constant. P6 finally seems to be the right alternative
for the digital printing technology (I and K).
The flexographic specific part of the diagrams (U and V) is described by

the equation P2. The form of the equation has also been kept for the
second phase of the work and the establishment of the coefficient P7 and
P8. The modification brought to P2 to obtain P7 and P8 had for objective
to more precisely take care of the different flexographic specific
parameters, both printing and quality parameters. The results confirm
both the expected results and the visual judgement.
The choice to use P7 or P8 depends of the panel of substrates to be

tested: P8 could be employed for a first approach and P7 for a finer
study.
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.

References:

terms of printability.”
FFTA Y2K Forum - Orlando (USA)

Process Modelling.”
TAGA 2000 - Colorado Springs (USA)

surface roughness.” - IARIGAI 2000 - Graz (A)
(4) “Provtryckning 2000” Research project (Sweden)

“The flexographic printability of linerboard” - Appita’93

Technology (10/94)

Prediction of the Substrate Printability in
Flexography by using a new established
“Printability Coefficient”
LAURENT, GIRARD LELOUP
Research Corporation Media and Communication Technology

Keywords: Flexography - Printability Coefficient - Prediction - Quality

Control - Objectivity – Measurement – Printing parameters – Paper
Characteristics
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.

OBJECTIVE OF THE INVESTIGATION
The objective of this research is to achieve the last step to an objective

prediction of the printability for the flexographic process. The work is
based on the equations presented at the TAGA conference in Mai 2001
(4). The equations has been improved and P9 / P10 will be used as final
version. The research will look at the influence of the different parameters
onto the printability, define the limits between the theoretical and
practical results and at the end propose a procedure to reach the final goal.
The procedure will contain an equation and “How to use it” method.
RESULTS
Equations for a “Printability Coefficient”
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.
P9 = ∑αi (1−(Αi /max Αi )) (1)
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)
P10 = ∑αi (1−(Αi /max Αi )) (2)

with α i =
+1/2 for mottling 30% and solid area, wicking, bleeding,
roundness 2% and 30%
+1 for dot gain
-1 for density full tone
A i printability parameters (printing and paper)

Influence of the quality factors variations
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.
Influence of the QF variations
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%
P9 Min P9 Max P10 Min P10 Max
Figure 1: Influence of the Quality Factors variations
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.
Quality factors range in “P units”
The following figures (2) and (3) represent respectively for P9 and P10
the influence of the quality factors variations converted in “P units”.

QF influence in P units - P9
0,36
0,93
1,03
0,23
0,91
Delta Mottling Delta Edge Sharpness Delta Dot Gain

Delta Density Delta Roundness
Figure 2: Influence of the Quality Factors in “P9 unit”

QF influence in P units - P10
0,18
0,47
0,52
0,22
0,91
Delta Mottling Delta Edge Sharpness Delta Dot Gain

Delta Density Delta Roundness
Figure 3: Influence of the Quality Factors in “P10 unit”
For P9: 0.23< ∆QF<1.03

P10: 0.22<∆QF<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.

Percentage of each QF range from the total P variation
The figure 4 is a more detailed representation of the results of the

influence of the quality factors. The mottling, edge sharpness and dot
deformation quality factors are decomposed in their primary components.
All the factors are related to the total “P” variation and the results are
plotted in percentage of ∆P.
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
Figure 4: Influence of the Quality Factors in %
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.
Influence of the primary parameters variations onto the quality

factors (QF)
To reach the final goal, a prediction of the printability, it is necessary to

go further down in the hierarchy and to look at the influence of the
primary printing and paper parameters. By transitivity it will then be
possible to calculate the global contribution of the primary parameters
(printing and paper) to the “Printability Coefficient” variations.
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.

The printing parameters (PP) variations
50,00
40,00
30,00
20,00
%
10,00
0,00
-10,00
% Anilox % Pressure % Hardness
-20,00
-30,00
Mottling 30% Mottling 100% Mottling % Bl Sharp

% W Sharp Sharpness Dot Gain Density
Round 2% Round 30% Roundness
Figure 5: Contribution of the printing parameters
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%).

The paper characteristics (PC) variations
50,00
40,00
30,00
20,00
%
10,00
0,00
-10,00
-20,00 % Surface Roughness % Density P %Absorption % Porosity
-30,00

Figure 6: Contribution of the paper characteristics
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.
Partial contribution of each parameter to the final “Printability

Coefficient”
The printing parameters
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.

The results are given is percent per unit. The intermediate results from the
precedent chapter are reduced to a “united result”. The units for the
printing parameters can be read in the table below.
Table 1: Units for the printing parameters
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

Figure 7: Partial contribution of the PP on the QF (P9)

4,0
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

Figure 8: Partial contribution of the PP on the QF (P10)
To be able to interpret the different figures a scale has been established.

The scale allows a correspondence between the proportional unit
influence and the greatness of the influence. The table should be
compared to the, for ∆E, known table. The scale is using the nominal
values.
Table 2: Degree of importance
% Degree
< 0.25 minimal
0,25<βi<1 important
1<βi<2 very important

>2 principal
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 paper parameters
The following figures represent the results for the printing parameters:
Surface Roughness, Density, Absorption and Porosity. The results are
presented for P9 and P10.
Table 3: Units for the paper characteristics
Unit
Surface Roughness 100 ml/min

Density 100 kg/m3
Absorption 0.1 cm3/ m2
Porosity 5 ml/min
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

Figure 9: Partial contribution of the PC on the QF (P9)

4,0
3,5
3,0
2,5
2,0
1,5
1,0
0,5
0,0
-0,5
-1,0
-1,5 % Surface % Density P / P10 % Absorption / % Porosity / P10

-2,0 Roughness / P10 P10
-2,5
-3,0
-3,5

Figure 10: Partial contribution of the PC on the QF (P10)
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 %.
Global contribution of each parameter to the final “Printability

Coefficient”
Each parameter influences the different quality factors positively or

negatively. By acting of one or more primary parameters it is possible to
control one or more quality factors. If the goal is to control the whole
printability it is interesting to look at the global contribution of each
primary parameter. The user can then decide at which level to work and
which are the parameters with a principal importance.

7,0
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
Figure 11: Global contribution of the PP and PC on “P”
The study of the figure 11 suggests to divide the primary parameters in 3

categories. The first one will regroup the porosity and the absorption and
will be called “first importance parameters”, the second one the anilox
roller volume, the plate hardness and surface roughness, called “priority
parameters” and finally the third one the nip pressure and the density,
called “less important parameters”.
The parameters in the last category are qualified as “less important”
parameters and not as non-important parameters. The influence below 1%
is not to underestimate and means only that the influences on the different
quality factors are neutralizing themselves.
The plotted values are the results for a decrease of one unit from the
standard value. This means for example: the surface roughness will be
improved with 100 ml/min and the absorption with 0.1 cm3/ m2 :
P10p = Pav * (1+ 0.26 + 0.53)

= 1.29

DISCUSSION
Establishment of a prediction procedure for the printability
Definition of a reference (average values)
To be able to predict the printability without any trials it is necessary to

have a reference. This reference will be constructed by taking the average
values for both the printing and paper parameters. As printability for the
reference configuration will be associated the mean value of P calculated
in the experimental phase of the project.
The reference shows also the following characteristics:
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.
Determination of the Min and Max values
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.
P9min = 7*0 + (-1)*1= -1 P10min= 4*0 + (-1)*1= -1

P9max = 7*1 + (-1)*0= 7 P10max = 4*1 + (-1)*0= 4
Min and Max values are calculated by giving the (Ai /max Αi ) (9) the
values 1 (worst) or 0 (best).
Theoretical and practical optimal quality
The theoretical achievable quality is the quality calculated in the

precedent chapter: 7 and 4 respectively for P9 and P10. Moreover the

process has certain limitation and tolerances, which has to be reflected in
the judgement. The correction is effectuated by considering the tolerances
for the quality factors.
Mottling: a mottling of 2 will be considered as good as a 0 mottling, 12
as the maximum => 2/12=0.17
Edge Sharpness: an edge sharpness of 1 is to assimilate to an ideal result
and 10 to the worst case => 1/10=0.10
Density: 1.6 is the target fixed for the density, it is to compare to the 2.2
max value => 1-(1.6/2.2)=0.28
Dot Gain: 15% is fixed as the top quality limits, 40% are expected as
maximum => 15/40=0.38
Dot Deformation (Roundness): 1.15 will be seen as a non-significant
deformation and 2.30 as a maximum deformation => 1.15/2.30=0.50
The restriction defined give for P9 / P10 the following results:

P9best = 7 – (0.17*2+0.10*2+0.38+0.28+0.50*2)
= 4.80
P10best = 4 – (0.17+0.10+0.38+0.28+0.50)
= 2.57
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.
=> 0.0 < P9 < 4.80

0.0 < P10 < 2.57
Judgement of the quality
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.
P9g = 2.40 P10g = 1.29

1
5,00
7 2
0,00
6 3
5 4
Pp Pm Pav
Pmin Pmax Pg
Figure 12: Visualisation of the Printability for different calculated Pp
The figure 12 is a proposition for a representation of the different Pp

calculated. The points located in the area between P g and Pmax are the
points researched.
Prediction of the Printability
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)
Pp = Pav * (p/100 +1) (4)

After printing
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)
Pm is with the P9 or P10 equations measured “Printability Coefficient”.
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

Flexography and the tolerances in materials and paper
manufacturing in terms of printability.” - FFTA Y2K Forum - Orlando
(USA)

Process Modelling.” - TAGA 2000 - Colorado Springs (USA)

“Study of paper printability in flexography by controlled variation of
the surface roughness.” - IARIGAI 2000 - Graz (A)

“Establishment and Comparison of Different Definitions and
Process.” – TAGA 2001 – San Diego (USA)

”A Flex-Odyssey…” - FFTA Forum 2001 – Nashville (USA)
“Provtryckning 2000” DPC-FRAMKOM-STFI Research project

(Sweden)
“Minutes” - Paptac Printing and Graphic Arts Committee Meeting –

10/1998 – Québec (Canada)
Prof. Dr.-Ing. Dr.H.C.L. Göttsching, Dr.-Ing. C. Ness, “Two-

Dimensional Characterisation on the Surface of Paper with reference
to its Printability” – Institute für Papierfabrikation, Darmstadt (D)
Armel, D., Cusdin, G., “ Plates & Printability ” – FQC FLEXO 04/2001

Measurement and Prediction Procedures For Printability in Flexography (MP3 Flexo)

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Measurement and Prediction Procedures For Printability in Flexography (MP3 Flexo)

Uploaded by

Copyright:

Available Formats

Measurement and Prediction

Procedures for Printability in

Laurent, Girard Leloup

Opponent: Dr. Patrice J. Mangin

Examination committee: Dr. Renke Wilken

Supervisor: Prof. Nils Enlund

Chairman: Prof. Roger Wallis

Thesis for the degree of Doctor of Technology to be presented with due

* Laurent, Girard Leloup, Mai 2002

Printed in France by ATO IMPRIMERIE - 8, Rue Jacquard - 54500 VANDOEUVRE

2 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

A literature study around the printability and quality control terms in

The present work should help the flexography to achieve needed

The thesis proposes a solution to objectively quantify the quality of the

“MP3 Flexo” is a quality control and communication instrument, which

The system as presented here can be implemented in the industry.

Key words: flexography, printability, objectivity, measurement,

Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002 3

The project was part of a national research program called T2F

In addition to the scientific results presented in this thesis, my work also

• There exists a real concurrence between the different research

4 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002 5

2.1 State of the art 5

3.1 What is flexography? 6

3.2 The definition of the “Printability” 8

4.1 1988 … 2002 12

4.3 Visual perception 15

5. Methodology and progression of the research 18

6 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

• 5.2.10 Establishment of a procedure for the prediction

6. A journey through six papers 28

6.1 The printing press parameter 28

7.1 Technical results 35

8.1 The adjustments 37

Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002 7

Literature cited and reviewed 41

8 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

In spite of the appearance of a multitude of new information and

The complete environment of flexography has been continuously the

Today the professionals of the flexographic printing industry have at their

Because they do not have the help of a structured method, of an

Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002 9

2.1 State of the art

The years 2000-01 were strong for flexographic printing, as quality

2.2 Identification of the problem

However, and unfortunately, flexography, which strives to compete with

2.3 Aim of the thesis

The present thesis will contribute to find a solution to the above-

The main objective of this research work is to make it possible to measure

10 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

3.1 What is flexography?

3.1.1 The official definition

“Flexography is a method of direct rotary printing that uses resilient relief

Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002 11

• It can print on a wide variety of absorbent and non-absorbent

3.1.3 Disadvantages of flexography

• It still shows some repeatability problems from order to order:

12 Laurent, Girard Leloup ∗ Royal Institute of Technology ∗ 2002

3.2 The definition of “Printability”

To understand the meaning of “printability” it is necessary to confront

To get definitions that can be classified under the category “descriptive