Helsinki University of Technology, Laboratory of Paper Technology

Reports, Series A19

Espoo 2004

MULTILAYERING OF FINE PAPER WITH 3-LAYER HEADBOX

AND ROLL AND BLADE GAP FORMER

Ari Puurtinen

Dissertation for the degree of Doctor of Science in Technology to be presented with due
permission of the Department of Forest Products Technology, for public examination and
debate in Council Room at Helsinki University of Technology (Espoo, Finland) on the 14th of
May 2004 at 12 noon.

Helsinki University of Technology

Department of Forest Products Technology
Laboratory of Paper Technology

Teknillinen korkeakoulu
Puunjalostustekniikan osasto
Paperitekniikan laboratorio
 2

PREFACE

This dissertation is based on research carried out at Metso Paper between 1998 and
2003. Experimental studies were conducted at the Metso Paper Technology Center
and VTT Processes in Jyväskylä. All work was completed under the guidance of
Metso Academy and funded by Metso Paper. The main goal of the research was to
elucidate the behavior of fibers and fillers in the paper thickness direction when
layering with a 3-layer headbox and a roll and blade gap former.

I would like to express my thanks to Professor Hannu Paulapuro of the Laboratory of

Paper Technology at HUT for his guidance of my work.

For my supervisors at Metso Paper, Pekka Pakarinen and Dr. Johan Grön (also one
of my cowriters), I owe a debt of gratitude for their inspiration, comments, and
critique.

Thanks are also due my former bosses, Lauri Verkasalo (M-Real) and Hannu
Korhonen (Metso Paper), who made it possible for me to carry out my research work
and encouraged me to forge on.

I would like to express special thanks to my cowriters, Terhi Saari (VTT Processes)
and Antti Oksanen (VTT Processes), for their attention to my work and for many
innovative discussions.

I would also like to thank the pilot paper machine and laboratory personnel at
Rautpohja and the laboratory personnel at VTT Processes for their numerous special
arrangements to accommodate my experimental and analysis work.

I also want to convey my appreciation to various Metso Academy students and other
key people within the Metso Academy for many fruitful discussions related to the
academic questions of my dissertation. Special thanks go to Dr. Kari Räisänen for
his guidance and encouragement.

Finally, I would like to express my deepest gratitude to my wife Eija and my son
Tuomas for their patience and support of my work.

Jyväskylä, May 2004

Ari Puurtinen
 3

Helsinki University of Technology, Doctoral Thesis

Laboratory of Paper Technology, Reports, Series A19
P.O. Box 1000, FIN-02015 HUT, http://www.hut.fi 14.5.2004
Title: Multilayering of Fine Paper with 3-layer Headbox and Roll and Blade Gap Former
Author: Ari Puurtinen Supervisor: Professor Hannu Paulapuro
Opponents: D.Sc. (Tech) Ari Kiviranta and Professor Risto Ritala

ABSTRACT

The aim of this study was on the layering of fine paper with a 3-layer headbox and roll blade
gap former. It consisted of methodological development and laboratory scale studies, as well
as layering studies on the pilot paper machine scale. The potential of layered structures in
papermaking and the phenomena affecting layer mixing were also studied.

Filler distributions were characterized by two parameters - filler distribution shape and
symmetry factors. The method developed made it possible to apply a statistical approach to
experiments when studying the impact of typical wet end control parameters on the control
of filler distributions and the contribution of filler distributions to paper quality parameters. A
method by which the center layer fiber distribution may be characterized from the final paper
is also presented. The method is based on the use of dyed fibers and a spectrofotometer.

In addition, a laboratory device, a multilayer handsheet former for making layered structures,
is presented. The process closely resembles a real continuous multilayer process by
providing water phase interactions between layers during forming and two-sided dewatering
characteristics. In a case study, retention chemical and filler layering with WFC base paper
was performed by the multilayer handsheet former. Surface roughness decreased and
brightness increased when the filler content in paper surfaces was increased. At constant
filler content, paper strength increased as the filler concentration in the paper surfaces
increased.

The distinction between roll dewatering and blade dewatering in terms of layer mixing as well
as differences in the mixing behavior of fibers and fillers in different dewatering phases was
studied. Major fiber mixing takes place during the free jet and roll dewatering phase. In blade
dewatering, fiber movements are related to formation improvements. The greatest mobility
differences between fibers and fillers were found in the blade dewatering phase in roll and
blade dewatering. Filler distributions were significantly affected during blade dewatering, with
the mobility of the fibers remaining insignificant.

The relationship between typical fine paper quality parameters and filler distributions was
studied. Formation improvement and densification of the paper were achievable at the same
running conditions. Low porosity was achieved by evenly distributed constituents (good
formation) and a local rich filler concentration somewhere in the paper structure in the paper
thickness direction. Oil absorption two-sidedness was minimized by symmetrical filler
distributions. The combination of all these quality factors required the filler distribution to be
symmetrical and as u- or turned u-shaped as possible.

The use of additives layering as a part of paper quality control was also studied.
Multilayering was found to be an effective tool for the control of filler distributions. On the
other hand it was shown that, in order to get optimized impact of the multilayering on quality
properties, the other process variables also need to be controlled.
Keywords: layering, fine paper, 3-layer headbox, multilayer handsheet former, filler distribution, fiber mobility, additives
layering, layer mixing
ISBN 951-22-7034-X , ISSN 1237-6248 [email protected]
 4

TABLE OF CONTENTS

LIST OF SYMBOLS AND ABBREVIATIONS 7

INTRODUCTION 8

OBJECTIVES AND STRUCTURE OF THE STUDY 10

MATERIALS AND METHODS 11

Multilayer handsheet former 11

Pilot paper machine 13
Analysing of filler distributions 14
Characterization of filler distributions 16
Calculation of filler distribution estimates 17
Characterization of fiber distributions 19

POTENTIAL OF LAYERED STRUCTURES 20

Fiber layering 20
Additives layering 21
A laboratory study concerning chemical layering of WFC base paper 21

LAYER MIXING 24

Headbox flow and vanes 24

Free jet and jet impingement 26
Roll dewatering 26
Blade dewatering 28

THE USE OF ADDITIVES LAYERING AS A PART OF PAPER

QUALITY CONTROL 35

Estimates 36

CONCLUSIONS 38

REFERENCES 39

APPENDICES

Appendix 1. The multilayer handsheet former. 45

Appendix 2. Test arrangements at pilot paper machine environment. 46

Appendix 3. Standard laboratory measuring methods. 50

 5

Appendix 4. Regression coefficients for filler distribution shape and 51

symmetry factors, Fue and Fsyme . Relates to the
publications III and V.

Appendix 5. The headbox and former parameters related to the 53

publications III, IV and V.

Appendix 6. Estimated vs. measured, porosity, formation index and oil 55

absorption two-sidedness. Relates to the publication IV:
Controlling filler distributions for improved fine paper
properties.

PUBLICATIONS:

I Puurtinen, A., Saari, T., Oksanen, A., Multilayer Handsheet Former. Paperi ja Puu-
Paper and Timber, 85(2): 92-95 (2003).

II Puurtinen, A., Saari, T., Grön, J., A laboratory study on the chemical layering of
WFC base paper. 15th PTS CHT Symposium: Chemische Technologie der
Papierherstellung, München, Germany, 37/1…37/11 (2002).

Puurtinen, A., Saari, T., Grön, J., A laboratory study on the chemical layering of WFC
base paper. Professional Papermaking, published by Wochenblatt für
Papierfabrikation, 1(1): 22-25 (2003).

III Puurtinen, A., Wet end control of filler distributions using additive layering
principles. Paper Technology, 44(4): 33-40 (2003).

IV Puurtinen, A., Controlling filler distribution for improved fine paper properties.
Accepted (13.6.2003) for publication in Appita Journal.

V Puurtinen, A., Oksanen, A., Control of fiber mobility and filler distribution in wet end
layering. NPPRJ, 18(2): 217-225 (2003).
 6

LIST OF SYMBOLS AND ABBREVIATIONS

The following symbols and abbreviations are used in this summary:

Fsym Symmetry factor

Fu Shape factor
Fts Top side filler content
Fbs Bottom side filler content
Fcl Center filler content
Fsyme Estimate for the symmetry factor
Fue Estimate for the shape factor
Ftse Estimate for the top side filler content
Fbse Estimate for the bottom side filler content
Fcle Estimate for the center filler content
Psym Retention chemical dosing symmetry factor
Pu Retention chemical dosing shape factor
Vane length Layering vane length
Q Headbox flow rate
FRvac Forming roll vacuum
LBvac12 Combined vacuum of the suction boxes 1 and 2 in
the LB-loading unit
BW Basis weight
CIE CIE-color coordinate system
L* Lightness
b* Yellow / blue axis
CTMP Chemi thermomechanical pulp
SC SC paper
WFC Woodfree coated paper
MFP Metered film press
MSP Metered size press
HSWO Heat set web offset
CS Coherent structure
MD Machine direction
CD Cross direction
IMH Mixing intensity at a specific location in the free jet
N Crowding factor
FFT Fast fourier transform
RMS Root mean square
TS Top side
BS Bottom side
Poroe Estimate for the porosity
Formae Estimate for the formation index
Oilabe Estimate for the oil absorption two-sidedness
 7

INTRODUCTION

The multilayering of paper has been commonly practiced in the manufacture of

paperboard and tissue grades and it has also been attempted with printing and
writing paper grades. In conventional paper manufacturing, the raw materials are
mixed. However, the objective of multilayering is to produce better paper from the
same raw materials, or the same grade of paper from lower quality raw materials. It
is thus motivated by the need to control factors affecting the cost and quality of the
paper.

Layered paper structures can be created in many ways. Separately formed sheets
can be bonded at suitable stages of the process and additional layers can be formed
on top of an existing sheet of paper. These methods are common in the manufacture
of paperboard where the goal is to produce layers with distinct fiber structures. The
coating of paper can also be loosely interpreted as a multilayering process.
Multilayering can also be accomplished using a so-called multilayer headbox where
layers are joined after passing through layering vanes either inside or outside the
headbox. The main motivation for using a multilayer headbox is the simplicity and
cost-effectiveness of the process compared to the equipment and systems needed
for separate sheet formation.

A number of experiments have been conducted to study and observe the potential of
multilayering in creating layered paper samples. Less attention has been paid to the
extent to which the layers tend to mix together during the continuous paper
manufacturing process.

One of the main goals of this work was to clarify the process parameters affecting
layer mixing during roll and blade gap forming with a 3-layer headbox. Attention was
paid to the impact of different dewatering phases on layer mixing and the mixing of
various raw materials.

Gap forming produces layered paper structures by itself. The ZD filler distribution
can be adjusted by changing the forming conditions. The impact of changes in the
forming conditions on the filler distribution also depends on the paper stock used. In
the manufacture of SC paper, for example, TMP-based stock produces a deep U-
shaped filler distribution, whereas GW-based stock yields a smoother distribution.

This study examined the filler distribution of fine paper and factors affecting fiber
migration in the paper sheet’s center layer. The results were examined from the
perspective of control factors (machine parameters) and dewatering during
formation. The goal was to find the most significant factors affecting center layer fiber
migration and the formation of the filler distribution. A secondary goal was also to
estimate what factors will produce selective fiber layering (layer purity) and the
desired filler distribution simultaneously.

The study examined the dependence of some key fine paper quality factors on the
filler distribution. The examination was based on empirical models created to depict
the dependence of filler distributions and the quality variables examined on the
control parameters.
 8

Laboratory scale sheet molds are fairly widely used tools in the study of
papermaking. This type of research includes raw material studies to determine the
potential of stock components, or the partial or complete simulation of manufacturing
processes. One of this study’s goals was to introduce a laboratory sheet mold that
could be used to partially simulate multilayer headbox layering and to investigate the
paper technology potential of multilayering using purely layered paper samples. The
multilayer sheet mold approach was also used to develop the methods used to
determine the purity of the layers.
 9

OBJECTIVES AND STRUCTURE OF THE STUDY

The main objective of this work was to study the process parameters affecting layer
mixing during roll and blade gap forming with a 3-layer headbox and to review the
significance of layered structures on fine paper properties. One objective was also to
distinguish various wet end mechanisms affecting the filler distributions and fiber
mobility in the paper thickness direction during the paper making process. The work
was based on laboratory and pilot paper machine studies. The main emphasis was
on the study of the role of the filler distributions.

The main objective was divided into the following sub-objectives:

1. To review the current knowledge of:

• the importance of layered structures
• various layering techniques
• process variables controlling filler distributions and fiber mobility
• the methods used to evaluate filler distributions and fiber mobility

2. To introduce and test a laboratory device for making layered paper structures.

3. To develop a method to characterize filler distribution, and to evaluate the role of

different wet end processes in controlling it.

4. To evaluate the significance of filler distribution on some fine paper quality

properties.

5. To develop methods to determine fiber mobility in the paper thickness direction

and to evaluate the role of different process variables in controlling it.

6. To distinguish different process variables controlling filler distribution and fiber

mobility in the paper thickness direction.

Publication
I Multilayer Handsheet Former 1,2
II A laboratory study on the chemical layering of WFC base 1,4
paper
III Wet end control of filler distributions using additives 1,3,6
layering principles
IV Controlling filler distribution for improved fine paper 1,3,4
properties
V Control of fiber mobility and filler distribution in wet end 1,5,6
layering
 10

MATERIALS AND METHODS

Multilayer handsheet former

Laboratory handsheet forming is a widely used method to produce paper sheets

under controlled circumstances. The basic research into the potential of fibers and
furnishes with respect to paper quality issues is often based upon laboratory
handsheet studies. Some of these methods may also be helpful in producing layered
sheet structures [1-9].

A common way to create a layered sheet is to filter new layers through already
formed layers [10-14]. Another possibility is to produce layered sheet structures by
bonding wet, single layered sheets together under pressurized conditions [15,16].
One task in the laboratory sheet forming field is to some extent to imitate the
continuous forming processes [1,2,4]. Two common features of the above-mentioned
laboratory formers are that basically they can be used to produce layered structures,
but that the dewatering is single-sided. Only a few laboratory scale methods are
capable of two-sided water removal [17-19].

A characteristic of the roll-blade gap forming process is two-sided dewatering and

the consequent influence that has on paper structure and quality properties. The
process enables a symmetrical paper structure which at least for most printing paper
grades is a desired quality property. When simulating the paper structure produced
by a modern gap forming process, it is necessary to create a symmetrical paper
structure in order to study the role and interaction of different additives (chemicals,
fillers) on paper quality properties.

In this study, a novel method of producing layered handsheets (Fig. 1) is presented

[I]. The basic idea of the multilayer sheet former is to create a layered sheet structure
between two fabrics. An advantage of this handsheet forming method is its
symmetrical dewatering characteristics, which resemble the dewatering
characteristics of the gap forming process.

Fig. 1. Multilayer handsheet former.

 11

Fig. 2 shows the design principle of the equipment. The sheet former's chamber can
be divided into two to three parts using removable divider plates. The fabric frames
are mounted in a vertical position at both ends of the chamber. The movements of
the fabrics are controlled by lateral pneumatic cylinders. The dividing plates in the
chamber are removed during the sheet forming process by vertical cylinders to allow
a two to three layered sheet to be formed between the fabrics.

Typical process variables:

• air mixing time 15 s and intensity 100 - 150 l/min,
• air exhausting time 12 s,
• forming speed from 15 * 10-3 m/s to 30 * 10-3 m/s (movements of fabrics)
• constriction force 250 N.

In order to make good quality stratified sheets, all these process variables have been
optimized in line with fundamental research.

Moving fabric Moving fabric

Lateral cylinder Lateral cylinder

Chambers that contain

components to be layered

Fig. 2. Multilayer handsheet former.

Fig. 3 shows the operating principle of the multilayer sheet former. The pulp and
chemicals (fillers, retention chemicals, wet end additives) are added manually into
each of the chambers. After a stage of mixing with compressed air, the actual sheet
forming begins by pushing the fabrics towards each other, which removes water from
the surface layers (Fig. 3, step 1 -> step 2). When the fabrics are moved close to the
dividing plates (step 2), the vertical cylinders remove the dividing plates (step 3). The
middle layer will then be dewatered through the surface layers and thus create a
multilayer sheet between the fabrics (step 4).
 12

1. 2.

3. 4.

Fig. 3. Operating principle of the multilayer handsheet former.

A more detailed depiction of the multilayer handsheet former has been presented in
Appendix 1.

Pilot paper machine

Test papers were made on the pilot paper machine located at the Metso Paper
Technology Center at Rautpohja, Finland. The furnish entering the headbox was
separated into three layers. The main stock feed line upstream of the headbox was
divided into three equal-sized lines. Each line had its own feed pump and machine
pressure screen. The headbox was a three-layer hydraulic headbox. The turbulence
generator had six rows of tubes, which were split into three sets of equal-numbered
(two/two/two) rows. The layers after the turbulence generator were separated by 2
mm thick flexible layering vanes. The last 50 mm length of the free end (downstream
end) of the 2 mm layering vanes were tapered to a final thickness of 0.7 mm. (Fig. 4)

headbox

retention chemical dosage

polymer bentonite

surface layer fibers

center layer fibers
surface layer fibers

Fig. 4. The additives and fiber layering wet end arrangements.

 13

suction loading
elements } elements (LB)

forming pre-loading
roll element

headbox

Fig. 5. The former.

The former was a roll-blade gap former equipped with loadable blades. The wrap
angle, i.e. the area where the outermost fabric is producing gap pressure against the
forming roll, was 15 degrees. The radial width of the forming roll suction zone was
also 15 degrees. The unit following the forming roll is called an LB-loading unit. The
loading side of the LB-loading unit consisted of one pre-loading element and four
loading elements. The fixed side of the LB-loading unit (opposite the loading side)
consisted of two suction box zones. The lower zone was located opposite the
loading blades (Fig. 5).

The forming fabrics were two-layer fabrics designed for fine paper. The retention
chemical dosage application took place after the division of the main stock line. The
retention chemical was introduced into the stock before the feed pump. The
bentonite was introduced after the feed pump and before the machine pressure
screen. Retention chemicals used were bentonite Hydrocol O and Allied Colloid’s
cationic PAM Percol 47. The line introducing furnish to the center layer could be
temporarily connected to a separate furnish source, which made it possible to use
different furnishes in the center layer than in the surface layers (Fig. 4).

More detailed depiction of the test equipments has been presented in Appendix 2.

Analysing of filler distributions

The filler distribution in paper in the paper z-direction is determined by splitting the
sheet into 10-15 layers by means of adhesive tape and then by analysing the filler
content of each layer by an incineration method in accordance with the SCAN P5:63
standard. Paper splitting technique itself is laboratory handiwork and is based on
defined laboratory work instructions.

The analysis method of filler distribution is assosiated various types of error sources.
Errors can be basically devided into two main categories: errors related to the
 14

analysis technique including the splitting technique itself and further sample
treatment such as sample weighting and ashing, and errors related to the paper's
structural issues such as homogeneity of the paper structure and material properties.

In planar directions (MD and CD) the paper homogeneity is mainly related to the
paper making process stability. In the paper z-direction the overall paper process
stability but also paper z-directional internal structure is affecting the paper
splittability in the paper z-direction. Studies related to the mechanisms and
measurements of the layered orientation structure of paper sheet, Erkkilä [20] found
that the splittability of the paper sheet was not dependent on the direction the
splitting has been conducted.

Repeatability of the splitting technique and the accuracy of the whole filler content
measurement chain has been evaluated by analysing the filler distributions from 5
individual points in a single A4 size sample and comparing the results (Fig. 6). The
sample was a production machine made fine paper sample.

30

25
Filler content, %

20

15

10

0
0 20 40 60 80 100
% of basis weight

Fig. 6. An example of typical filler distribution.

The average standard deviation of the filler content measurements is 0.91 and it's
standard deviation 0.23. The average standard deviation and it's standard deviation
for the percentage basis weights are 1.56 and 0.62 . The result shows that the
splitting as a technique in this particular fine paper case is relatively well repeatable
and the accuracy of the filler content can be considered to be within +/- 1.2 % units.

Tape splitting method is widely used in the existing literature [20-31].

 15

Characterization of filler distributions

To be able to apply a statistical approach to the analysis work, the filler distributions
needed to be converted into a more appropriate form. In this study the
characterization of the filler distribution was defined based upon two characteristics,
the symmetry factor, Fsym and the shape factor, Fu. The mathematical definition of
the filler symmetry factor is as follows:

Fsym = Fts/Fbs-1, when Fts/Fbs>=1, (1)

Fsym = -(Fbs/Fts-1), when Fts/Fbs<1, (2)

where Fts is the top side filler content and Fbs is the bottom side filler content. The
shape factor of the filler content, Fu is defined,

Fu = (Fts+Fbs)/(2*Fcl)-1, when (Fts+Fbs)/(2*Fcl)>=1, (3)

Fu = -(2*Fcl/(Fts+Fbs)-1), when (Fts+Fbs)/(2*Fcl)<1, (4)

where Fcl is the center layer filler content.

For the calculation of the Fsym and Fu, the representative top, center and bottom
side filler contents from the original measured filler distributions are needed. For the
determination of representative values, filler distribution estimates for each of the
original distributions were made. The estimated distributions were then used to
define the needed values for the calculations of the Fsym and Fu. In these
experiments the 4th degree polynomian fit was used. Fig. 7 shows an example of a
series of measured filler distributions (solid lines) and estimated distributions (broken
lines). Values marked with squared dots (three in each picture) represent the values,
and these were used to calculate the Fsym and Fu values related to each filler
distribution.
c)
F ille r c o n te n t, %

a) b)
30 30 30
25 25 25
20 20 20
15 15 15
10 10 10
P o ly m e r s p lit P o ly m e r s p lit P o ly m e r s p lit
5 5 5
3 3 /3 3 /3 3 % 5 0 /0 /5 0 % 6 0 /0 /4 0 %
BS 0 TS 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
% o f b a s is w e ig h t
d) e)
30 30
25 25
20 20
15 15
10 10
P o ly m e r s p lit P o ly m e r s p lit
5 5
7 0 /0 /3 0 % 1 0 0 /0 /0 %
0 0
0 20 40 60 80 100 0 20 40 60 80 100

Fig. 7. Typical filler distributions at different retention chemical splits. Measured filler
distributions are represented by solid lines and estimates by dashed lines. Estimates
are 4th degree polynomes, 80 gsm paper.
 16

The pictures in Fig. 8 show a few examples for geometrical interpretation of the
symmetry (Fsym) and the shape (Fu) factors.

a) b) c)

Filler content
Fsym > 0
Filler content

Filler content
Fsym = 0 Fsym < 0
Fu =0 Fu =0 Fu =0

BS
Basis weightTS Basis weight Basis weight
d) e) f)

Filler content
Filler content

Fsym > 0
Filler content

Fsym = 0 Fsym = 0
Fu >0 Fu <0 Fu >0

Basis weight Basis weight Basis weight

Fig. 8. Examples of geometrical interpretation of filler distribution characters, Fsym

and Fu.

Fsym measures the balance between filler surface contents. When both surfaces
have equal filler content, Fsym is 0 (a). If the bottom surface filler content is higher,
Fsym becomes a negative value (b). If the bottom surface filler content is lower,
Fsym is a positive value (c). The filler distribution shape factor measures the
relationship between the average filler content in the surfaces and the center layer
filler content. When the average filler content in the surfaces equals the center layer
filler content, Fu is defined as 0 (a). If the center layer filler content is lower than the
average filler content in the top and bottom surfaces, Fu is a positive value (d).
Where the center layer filler content is higher than the surface layer content, Fu is a
negative value (e). The characterization of the filler distribution is based upon a
combination of these two characteristics, Fsym and Fu (f) [III].

Calculation of filler distribution estimates

In this study models for filler distributions were created and used to study the
relations between the filler distributions and some quality parameters. Models were
also used to calculate the relative effect of different wet end components to control
the filler distributions.

The filler distribution models are based on the estimates for the filler distribution
shape and symmetry factors, Fue and Fsyme. In the following the estimates for the
factors are calculated using some of the wet end control parameters. The
mathematical definition of the factors can be stated in the following way:

Fue = f(Psym, Pu, Vane length, Q, FRvac, LBvac12) (5)

and
Fsyme = f(Psym, Pu, Vane length, Q, FRvac, LBvac12), (6)
 17

where Psym = retention chemical dosing symmetry factor,

Pu = retention chemical dosing shape factor,
Vane length = layering vane length,
Q = headbox flow rate,
FRvac = forming roll vacuum,
LBvac12 = combined vacuum of the suction boxes 1 and 2 in the LB-
loading unit.

A regression analysis method was used to calculate the coefficients for the factors
presented in Equations 5 and 6. The regression analysis design chosen was based
upon the principles of one of the response surface methods, which in this case was
Central Composite Surface Design. The main reason for using the chosen design
was that it allowed the use of individual factors at five levels, i.e., center points, cube
points and star points. The commercially available computer program that was
selected allowed various experimental designs to be used. In the analysis, the
statistical significance level p was set to be less than or equal to 0.05 [32].

The calculation of filler distribution estimates was based on filler distribution shape
(Fue) and symmetry factor (Fsyme) estimates, which can be calculated from the
models (Equations 5 and 6). The estimates for filler content in the paper surfaces
(Fbse and Ftse) and the center area (Fcle) were calculated using Equations 1-4.
Finally, in order to visualize the filler distribution estimate, the second order
polynome was fitted through the calculated filler center and the surface contents. An
example of such a visualization can be seen in Fig. 9, where the original measured
filler distribution is presented by solid lines and the estimate, which was based on
the model, is presented by broken lines.

a) b) c)
30 30 30
F ille r c o n te n t, %

25 25 25
20 20 20
15 15 15
10 10 10
P o lym e r sp lit 5 P o lym e r sp lit P o lym e r sp lit
5 5
3 3 /3 3 /33 % 5 0 /0 /5 0 % 6 0 /0 /4 0 %
BS 0 TS 0 0
0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100
% o f b a s is w e ig h t
d) e)
30 30
25 25
20 20
15 15
10 10
5 P o lym e r sp lit P o lym e r sp lit
5
7 0 /0 /3 0 % 1 0 0 /0 /0 %
0 0
0 20 40 60 80 100 0 20 40 60 80 100

Fig. 9. Typical filler distributions (solid lines) at different retention chemical splits.
Estimates (broken lines) are based on the model for 80 gsm paper.
 18

Characterization of fiber distributions

Lloyd et al. [33] used dyed (blue) fibers in the center layer for the identification of
mixing during three-layer stratified forming. A method defined as surface ply variation
(SPV) was also introduced. SPV was measured by scanning samples with a photo
scanner and then by analyzing these images. SPV was used to calculate the
variation of the blue content of the surface within the 0.3-30 mm wave length range.

In studies performed by the author [V], parts of the center layer’s long fiber fractions
were dyed with a blue color in order to identify their position on the final sheet . The
fibers were colored using a method developed by VTT Processes. In this method,
the final dyed pulp mixture contains dyed fibers but the surrounding water remains
dye-free. The portion of dyed fibers in the center layer pulp mix was approximately
25 %. The undyed pulp fraction in the center and surface layers consisted of 70 %
hardwood and 30 % softwood.

The center layer fiber (dyed) deviation in the thickness direction was analyzed by
laminating the final paper to separate layers in the thickness direction and then
measuring and calculating the portion of colored fibers in each layer. An example of
a laminated laboratory sheet made with a multilayer handsheet former [I], so that the
center layer fibers are dyed blue, is presented in Fig. 10.

BS TS
Fig. 10. Laminated laboratory multilayer former sheet, with the center layer dyed
blue.

Using white and blue backgrounds CIE L* and b* color space values are determined
from the layers. The distribution of dyed fibers in the thickness direction is calculated
by using the L*, b* values and calibration curves. The calibration curves are used to
correct layer grammage variation. In the figures, the tracer fiber distribution is
presented as a function of grammage from the bottom side to the top side (Fig. 11).
The calculation method was developed by VTT Processes.
 19

B
b
C

A c

Fig. 11. The distribution of tracer fibers in laboratory-made handsheet

samples.Tracer fibers are added to the center layer prior to forming. A) 70 gsm
sheet where tracer fibers are applied equally to the furnish. B) 70 gsm test sheet
where tracer fibers are layered in the middle layer (30 gsm). C) 110 gsm test sheet
where tracer fibers are layered in the middle layer (30 gsm). The dash lines
represent theoretical tracer fiber distributions in the thickness direction. b) 70 gsm
sheet and c) 110 gsm sheet.

The standard laboratory measuring methods are presented in Appendix 3.

POTENTIAL OF LAYERED STRUCTURES

A number of experiments have been conducted to study and observe the potential of
the method of creating layered paper samples. For the most common applications
the potential is related to raw material economy and/or improved quality parameters
[22,23,34-40]. Depending on the philosophy behind the layering, various advantages
can be seen compared with the conventional way of making paper.

Fiber layering

Häggblom-Ahnger et al. [24] discussed the potential of multilayering office paper

grades. They reported improvements in critical properties such as bending stiffness,
formation and opacity. A further possible benefit would be basis weight reduction
while maintaining stiffness and opacity at the same level. The reduction of furnish
costs by using CTMP furnish instead of chemical pulp in the middle layer has also
 20

been discussed. One possibility is to put recycled furnish in the middle layer while
retaining high quality virgin pulp on the surfaces. In studies concerning the optimum
location of softwood sulfate pulp in three-ply office paper Häggblom-Ahnger [41]
found that bending stiffness, formation and smoothness were improved by locating
the softwood in the middle ply instead of in the outer layers. The optimum stiffness
and surface properties could be achieved by using softwood and eucalyptus sulfate
pulps in the middle layer and applying birch sulfate to the outer layers.

Häggblom-Ahnger [42] also discussed the fractionation of softwood sulfate pulp in

three-ply copy paper. The layering of the fine fractions in the outer plies improved
bending stiffness. On the other hand, by locating the fine fractions of softwood
sulfate in the middle layer, Scott Bond was clearly improved.

Additives layering

Kinnunen et al. [25] discussed the role of different forming concepts and additives
layering principles to control filler and fines distribution in SC paper grades. They
reported that it was possible to improve optical properties by increasing surface filler
content. With multilayering it seemed to be possible to achieve a small brightness
increase without opacity loss. Odell [26] also discussed the advantages of
multilayering. He reported significant improvements in the gloss and dot gain from
the pilot printing of the multilayered SC papers when the surface filler content was
increased by layering retention chemicals into the paper surfaces.

Harwood [22] mentioned the possible advantages to be gained by layering additives

such as starch, filler, size and retention aids selectively, rather than dispersing them
across all the paper. In three ply layering, the addition of starch to the center layer
(rather than to all plies) increased tensile and bond strength.

Häggblom-Ahnger et al. [24] reported experiments with consistency layering. They

noticed that consistency layering had a significant impact on formation.

A laboratory study concerning chemical layering of WFC base paper

The experiments performed by the author et al. [II] studied different possibilities to
create desired surface properties by layering certain components in a WFC base
paper. The study was executed on a multilayer laboratory handsheet former [I]. One
of the main tasks was to reduce the openness of the paper surface through
concentrating fillers on the paper surfaces. The grammage of the sheets was 80 gsm
with rosette-shape PCC filler (median particle size of 1.5 µm) contents of 22%, 26%
and 36%. The chemical pulp used was a 65 / 35 % mixture of hardwood kraft pulp
beaten to SR° 24 and softwood kraft pulp beaten to SR° 20.

The study was based on previous findings that showed that increased filler content in
paper surfaces has several beneficial effects on either metered film press (MFP) or
metered size press processes (MSP). Forrström et al. [43] showed that the coat
 21

weight formed on woodfree base paper in filmpress precoating depended mainly on

the coating color film on the application roll and the base paper surface porosity. Low
base paper surface porosity and/or high filler content on the surface of the sheet
decreased coating color penetration into the base paper structure and the coating
color stayed better on the surface. Grön et al. [44] studied LWC base on a metered
film press (MFP) process. The main findings were that, when coating color is
transferred to a base paper, a paper with high filler content demanded less
premetered film to achieve the desired coated paper properties. In particular, base
paper with the highest filler content (20 %) also exhibited a denser paper surface and
kept the color more effectively on the paper surface. It was shown by Ahlroos et al.
[45] that increasing the base paper filler content levels had a positive effect on
surface roughness, print gloss and print mottle in HSWO Printing. It was also found
that coat weight uniformity was improved by pre-calendering as well as by higher
filler amounts in the base paper. Studies with WFC base paper and a metered size
press (MSP) by Dickson et al [27] showed that improved coating coverage can be
obtained through the use of more filler in the surface of the base paper, higher levels
of kaolin clay in the coating formula, lower nip load, and softer size press rolls.

In this study it was found that concentrating fillers in base paper surfaces promoted
quality characteristics such as brightness and surface roughness. Surface roughness
decreased and brightness increased when the filler content in paper surfaces was
increased. The brightness and surface roughness change followed the surface filler
content change (Figs. 12 and 14) rather than changes in the filler total content (Fig.
13). At the same time, with constant filler content, paper strength increased as the
filler concentration in paper surfaces increased (Fig. 15).
90

89 B rig h tn e s s
m e a s u re m e n t
88
Brightness, %

to p s id e ,
87 w ith o u t s ta rc h

86 to p s id e , w ith
s ta rc h
85
b o tto m s id e ,
84 w ith o u t s ta rc h

83 b o tto m s id e ,
0 10 20 30 40 50 60 w ith s ta rc h

F ille r s u r fa c e c o n te n t, %

Fig. 12. Brightness as a function of paper surface filler content.

 22

90

89 B rig h tn e s s
m e a s u re m e n t
88
Brightness, %

to p s id e ,
w ith o u t s ta rc h
87

86 to p s id e , w ith
s ta rc h

85
b o tto m s id e ,
84 w ith o u t s ta rc h

83 b o tto m s id e ,
0 .0 1 0 .0 2 0 .0 3 0 .0 4 0 .0 5 0 .0 6 0 .0 w ith s ta rc h

F ille r c o n te n t, %

Fig. 13. Brightness as a function of filler total content.

1400
Bendtsen roughness, ml/min

1200 R oughness
m e a s u re m e n t
1000 to p s id e ,
(uncalendered)

w ith o u t s ta rc h
800
to p s id e , w ith
600 s ta rc h

400
b o tto m s id e ,
w ith o u t s ta rc h
200
b o tto m s id e ,
0 w ith s ta rc h
0 10 20 30 40 50 60
F ille r s u r fa c e c o n te n t, %

Fig. 14. Surface roughness (Bendtsen) of uncalendered paper as a function of paper

surface filler content.

4 0 .0
F ille r c o n te n t in p a p e r 2 0 %
F ille r to ta l
F ille r s u rfa c e
c o n te n t
Tensile strength, Nm/g

c o n te n t
3 5 .0
L a y e re d p a p e r
s tru c tu re F ille r s u rfa c e
c o n te n t, w ith
s ta rc h

3 0 .0
F ille r to ta l
c o n te n t, w ith
E v e n (n o n la y e re d ) s ta rc h
p a p e r s tru c tu re

2 5 .0
0 10 20 30 40
F ille r, %

Fig. 15. Tensile strength as a function of paper surface filler content and filler total
content.
 23

Smoother base paper also showed smoother paper surface properties after coating.
It was also shown that the dusting tendency of fillers in laboratory calendering could
be significantly decreased by using starch. Layering offers an opportunity to intensify
the use of starch by using layering techniques and targeting the starch at the paper
surfaces.

LAYER MIXING

Layering with a multilayer headbox is based on the idea of separating the layers from
each other as long as possible. The consolidation of the different layers takes place
after the separating vanes, inside or outside the headbox. After consolidation during
sheet formation, the different layers tend to mix together. Mixing of layers does occur
in the free jet and continues in the various dewatering phases in the forming section.

Mixing can be different in different dewatering phases. In roll-blade gap forming

during the roll dewatering phase, the water volumes are significantly bigger than in
the blade dewatering phase. Thus it is concluded that the mixing is also different.
Furthermore, the raw materials to be layered may behave differently during different
dewatering phases. The behavior of fibers may be different than the behavior of
fillers. The behavior of the fillers is also related to the retention level [28] and the
retention mechanisms. Lindström [46,47] categorizes the retention mechanisms in
the following way: charge neutralization, patch model of flocculation, hetero
flocculation, bridging flocculation (adsorption flocculation, sensitization flocculation,
complex flocculation), network flocculation and volume restriction flocculation.
Depending on the dewatering phase, the retention mechanisms of fillers may vary
from that of fibers and thus give space for independent movements of fillers. The
goal may still be to control the layerability of both fibers and fillers.

Headbox flow and vanes

In the first stage of layer consolidation just after the vanes, mixing is governed by the
hydrodynamic state of the flows to be consolidated and phenomena related to the
vanes themselves.

Baker et al. [48] studied the impact of three dimensional headbox flows on tissue
layer purity. They observed the existence of both spanwise and streamwise
structures within the wake of headbox dividers. They stated that three dimensional
flows can have a significant effect on layer purity and that the potential mechanisms
responsible for such flows include the formation of swirling flows at the diffuser tube
bank, three dimensional wakes formed at divider tips and instability of the free jet
surface. They also found that the use of extended flexible slice lips can improve layer
purity, though the mechanisms for these improvements were not identified.

Jetsu et al. [49] discussed the inherited nature and the formation of coherent
structures (CS) of suspension flow when using vanes. The CS is geometrically three
dimensional and can be separated into MD and CD components (which are
dominant) with distinctive characteristics of their own. The amplitude and wave
 24

length of the CS was found to be related to the dimensions of the vane. By

increasing the effective thickness of the vane, it was possible to increase the
amplitude and wave length of the CS .

Lloyd et al. [33] studied layer mixing during three-layer stratified forming. They
concluded that a combination of vane tip vortices and channel turbulence is thought
to cause significant mixing of the different layers during stratified forming. Shorter
vanes led to mixing at floc level, while longer vanes produced mixing at the fiber
level. Longer vanes in a contracting nozzle meant higher flow velocities, both at the
vane tips and in the flow channels. This was thought to produce more intense and
smaller scale turbulence, which tended to mix the layers more at the fiber level than
at the floc level. Very long vanes caused complete turbulent mixing of the different
layers and very poor sheet formation. The other study done by Lloyd et al. [50]
concerned the effect of vane shape on layer mixing. They used stepped vanes with
various configurations. The aim of the steps was to break down the boundary layers
formed upstream along the vanes, and thus reduce downstream vortex generation.
They noticed that the turbulence introduced by the steps increased the layer mixing.

Lepomäki et al. [51] suggested that it is important to keep the amplitude of coherent
structures in a headbox low enough for stochastic turbulence to break and obscure
them. The amplitude and frequency of coherent structures are always a function of a
specific dimension of the element generating them. For this reason, all disturbances,
such as steps, have to be minimized after fluidization.

Li et al. [29] recommended a vane sheet with a tip thickness of 0.5 mm for 2-layer
and 3-layer stratification applications. For a very thin divider sheet, it is anticipated
that that there will be some concerns about cross machine uniformity problems and
the sheet life.

Söderberg et al. [52] studied the influence of the contraction ratio on headbox flow,
using dyed fibers in different layers. As a side result [53] they found that the optimum
contraction giving the lowest layer mixing was somewhere between the low and high
contraction cases.

Andersson [30,31] presented layering results achieved with a layering headbox using
“air-wedge” technology. Rigid layer separator vanes terminating outside the slice
opening made it possible to run different individual jet velocities in different layers. It
was found that the optimum layer purity was achieved with a jet velocity profile such
that the highest velocity of the jet was on the forming roll side.

Parsheh [54] studied mixing in a liquid jet out of a stratified headbox by using a
model. He found that the relative difference of the vane length and the contraction
length was one of the most important factors. Moreover, a vane shorter than the
contraction results in the lowest mixing.
 25

Free jet and jet impingement

The second critical stage in layer mixing is the free jet. Li et al. [29,55] studied the
mixing characteristics of a water jet as a function of the distance from a stratified 2
and 3-layer headbox nozzle. They put salt water into the middle layer and measured
the fluid conductivity profile in the z-direction. A mixing intensity at a specific location
in the free jet, IMH, was introduced. IMH increased with distance from the slice, and
was inversely affected by slice opening and headbox flow rate. It was found that a
shorter divider sheet stickout tended to provide a lower degree of mixing. However, a
longer divider sheet stickout may be preferable for the situation with a shorter free jet
length.

The role of the impingement of the jet between fabrics in roll-blade forming,
particularly from the layer mixing point of view, has not been discussed widely.
However, it is known to be a critical part of the paper forming process.

Roll dewatering

Studies done by the author [V] with 3-layer headbox using dyed fibers showed that
layer mixing increased when the headbox flow was increased (Fig. 16). In the graph
the vertical axis shows the headbox flow increase, i.e. in all points labeled A,
headbox flow was low. B denotes medium headbox flow and C high headbox flow.
The horizontal axis shows the forming section dewatering intensity. In the first
column from the left (denoted by 0) the dewatering of the LB-loading unit was
minimized. Only minimum vacuums in the LB-loading unit vacuum boxes were used
and only a pre-loading blade was used. Moving from the first to the second column
(denoted by 1) and further to the third column (denoted by 2), the relative share of
LB-loading unit dewatering (blade dewatering) increased at a given headbox flow.

A change in blade dewatering intensity did not have an effect on fiber mixing. Blade
dewatering intensity was changed from zero blade force and minimum vacuums in
the LB-loading area to maximum blade force and maximum vacuums. On the other
hand Li showed that the mixing intensity at a specific location in the free jet, IMH, was
inversely affected by the slice opening and headbox flow rate [29]. So it was
concluded that fiber mixing due to increased headbox flow took place at the roll
dewatering phase.
 26

100 % 100 % 100 %

Center layer fiber content

Center layer fiber content

Center layer fiber content

C0 C1 C2
0 0 0
0 Basis weight 100 % 0 Basis weight 100 % 0 Basis weight 100 %
BS TS
100 % 100 % 100 %

Center layer fiber content

Center layer fiber content

Increasing flowrate
Center layer fiber content

B0 B1 B2
0 0 0
0 Basis weight 100 % 0 Basis weight 100 % 0 Basis weight 100 %
100 % 100 % 100 %
Center layer fiber content

Center layer fiber content

Center layer fiber content

A0 A1 A2
0 0 0
0 Basis weight 100 % 0 Basis weight 100 % 0 Basis weight 100 %

Increasing blade dewatering

Fig. 16. Distribution of the center layer fibers.

The distributions were relatively symmetrical, and with low and medium flow rates
the layer purity of the surface layers was significantly better than with a high flow
rate. Approximately 20 % of the total basis weight in both surfaces remained
unmixed (low and medium headbox flows) or mixing was constant (high headbox
flow).

The impact of roll dewatering on the behavior of filler distributions has been
discussed in the studies [III,V] performed by the author. The role of the headbox flow
in the control of the filler distributions depends significantly on the basis weight.
Using the definition of the filler distribution shape and symmetry factors, it was found
that with 80 gsm paper, the correlation between shape factor and headbox flow was
significant, whereas with 50 gsm paper it was negligible. The explanation for that is
assumed to be due to the differences in the web structure after initial build up of the
surface layers in the roll dewatering phase.

With 80 gsm paper, the filler distribution u-shapedness increased as a function of

increased flow, i.e. the filler movement was emphasized when the headbox flow was
increased. This gives an indication (first assumption) that after the creation of the
surface layers, a consistency profile builds up between the surface layers. This firstly
allows filler movement during dewatering after initial mat build up and secondly filler
particle entrapment in the formed surface mats. The second assumption is that
before a surface layer can act as a filter and the creation of a consistency profile is
possible, a certain number of fiber layers (layer thickness) are needed. For
papermaking fibers the basis weight of one single fiber layer is 5-10 gsm [56,57]. 50
gsm paper has approximately 20 gsm fibers per side (fillers excluded), so the
theoretical number of fiber layers per side is 2-4. In their studies Wildfong et al. [58]
found that the viscous resistance coefficient increased as the mat formed. They also
 27

found that at lower basis weights the increased fines level did not have a significant
impact on the resistance values, presumably due to the low initial retention present
during early mat deposition. An increase in the viscous resistance coefficient with
fine paper furnish as well as with newsprint furnish could be seen after 20 gsm of
material had been deposited.

Based on the second assumption and Wildfong's findings, it is concluded that after
initial mat formation with 50 gsm paper, no excess material for consistency
difference creation between layers remained. Thus a not measurable amount of free
filler movement took place after initial mat formation and the impact of headbox flow
on the filler distribution shape factor remained insignificant. This is probably also the
reason why the effect of the former parameters on the 50 gsm paper shape factor
remained insignificant.

Blade dewatering

Fiber mobility is related to layer mixing, the main part of which takes place in the
headbox, jet and forming roll. Fiber mobility during the blade dewatering phase is
generally related to formation improvement. Filler distribution shapes are also
connected to layer mixing, particularly when filler loading into various layes has been
boosted with retention chemical layering. On the other hand fillers may still have
mobility when fibers have already settled into their final positions in the paper
structure. Mobility differences between fibers and fillers may be at their greatest
particularly in the blade dewatering phase in roll and blade dewatering.

Zhao et al. [59] studied the effect of consistency on pressure pulses in blade gap
formers. They stated that to improve formation, relative movement must take place
among the fibers so that the fibers move from the flocs to the zones between the
flocs. Thus, the fiber displacement distance should be of the order of a floc diameter
if formation is to be affected in a significant way.

Nordström et al. [60] showed that the effect on formation of a change from roll to roll-
blade dewatering was found to be determined by a balance between improving and
impairing mechanisms. This balance was influenced by jet quality and blade force
and was different in different wave length ranges. Higher blade forces led to a
deterioration in both the small-scale and large-scale formation. Norman [61] stated
that the effects of dewatering pulses on paper properties are positive as well as
negative. The main positive effect is the improvement of large scale formation. The
effect of the blade pulses is larger the poorer the formation level after initial roll
dewatering.

Kerekes et al. [62] presented two aspects of a fiber suspension which determine the
degree to which it may lead to good formation - uniformity and mobility. A uniform
suspension tends to produce good formation. Mobility, meaning the ease with which
fibers can move relative to one another, also affects formation. A crowding factor, N,
defined as the number of fibers in a spherical volume of diameter equal to the length
of the fiber, has been used to characterize flocculation of fibers in a water
suspension. The mobility of fibers and their uniformity of distribution were shown to
 28

change dramatically over the range 1<=N<=130. At N~=1, fiber mobility was high. As
N>1, the suspension tended to become non-uniform for fibers with a high-aspect
ratio (length/diameter) and uniform for low-aspect ratios. In the range 60<=N<=130,
fiber mobility decreased significantly. Kerekes et al. [63] stated that both fiber length
and coarseness also affected fiber mobility. It was postulated that, like uniformity,
mobility is a key factor in formation, since it determines the extent to which
suspension uniformity can be altered during drainage on the paper machine. In layer
mixing studies performed by the author [V] it was found out that layer mixing
increased when the headbox flow was increased (Fig. 17). On the other hand it was
also found that surface uniformity two-sidedness started to gradually worsen when
the headbox flow was increased and that the top surface became more uniform. On
the bottom side, surface unevenness increased and, at the highest flow rate, clearly
observable faults (crushing) could be seen. The flow level, which caused visible
unevenness in the paper, was chosen consciously and was here considered to be
"high" headbox flow. Figs. 17 and 18 show the scanned pictures of the paper bottom
and top side surfaces. Dyed fibers in the paper center layer were used. The crushing
originated most probably in the roll dewatering (gap) area. In the C0 case, the
following dewatering treatment (after the roll dewatering phase) was very gentle,
particularly on the bottom side (Fig. 17). The reason for crushing is assumed to be
the too fast densification of the fiber mat after a certain threshold headbox flow
during the initial dewatering phase. This led to uncontrollable suspension flows in the
paper surface plane directions, resulting in visible fiber clumps.

Bottom side

C0 C1 C2
Increasing flowrate

B0 B1 B2

A0 A1 A2
Increasing blade dewatering

Fig. 17. Scanned bottom side surfaces at various trial points. Dyed fibers used
in the center layer.
 29

Top side

C0 C1 C2
Increasing flowrate

B0 B1 B2

A0 A1 A2
Increasing blade dewatering

Fig. 18. Scanned top side surfaces at various trial points. Dyed fibers used in
the center layer.

The area of disturbance in the paper thickness direction can be visualized by means
of the following figure (Fig. 19), where the scanned pictures of all the separate paper
layers in cross direction at various test points are presented. The formed sheets
were typically divided into 8-10 individual plies in the paper thickness direction. The
paper splitting technique used was the same as for the filler distributions
assessments. Each picture in Fig. 19 is composed of scanned pictures of each
individual plies (width 50 mm). In final picture the plies are arranged side by side
according to their position in paper thickness direction resulting a picture which
represents the paper’s z-directional structure. The pictures were prepared from the
undyed samples. The uneven pattern in the bottom surface of the paper (in the
paper thickness direction) can be seen in figures B0, B1, C0, C1 and C2. It is
relatively easy to make a distinction between the crushed areas and the floccy areas
by visually inspecting Figs. 17, 18 and 19 at the same time.
 30

TS

“flocky” C0 C1 C2
“sm oothening”
“cru she d”
BS
In c reas in g flo w rate

B0 B1 B2
“sm oothening”

A0 A1 A2

In c reas in g b la d e d e w aterin g
Fig. 19. Scanned layers at various trial points. In each picture (A0, A1, ..),
representing a single trial point, pictures of each individual ply have been ordered
according to their positions in the paper’s thickness direction. Each picture typically
consists of 8-10 individual plies.

The bottom side unevenness (crushing) can also be seen in the floc size distribution
charts in Fig. 20. Floc size distribution in the thickness direction of the paper was
determined from the images of each layer using FFT-transformation. RMS (Root
Mean Square) values were calculated from the wave length bands 0.2 – 1.0 mm, 1.0
– 10 mm and 10 mm – 100 mm from Fourier Power Spectra. Floc size is estimated to
be half of the wave length. A small RMS value indicates a low amount of flocs while,
correspondingly, a high RMS value means a high number of flocs.
 31

35
RMS 35 35

RMS

RMS
C0 C1 C2
0 0 0
0 Basis weight 100 % 0 Basis weight 100 % 0 Basis weight 100 %
BS TS
flowrate

35 35 35
RMS

RMS

RMS
0.2 - 1.0 mm
1.0 - 10 mm
10 - 100 mm
0
B0 0
B1 0
B2
Increasing

0 Basis weight 100 % 0 Basis weight 100 % 0 Basis weight 100 %

35 35 35
RMS

RMS

RMS
0
A0 A1 A2
0 0
0 Basis weight 100 % 0 Basis weight 100 % 0 Basis weight 100 %
Increasing blade dewatering

Fig. 20. Floc size distributions. Floc categories 0.2-1 mm, 1.0-10 mm and 10-100 mm.

When estimating the role of the LB-loading unit in improving formation

characteristics, changes in floccy and crushed areas of the paper structure can be
used as a "tracer" of changes made by the LB-loading unit. If the structure is
"smooth", the deflocculation power of the LB-loading unit cannot be identified using
paper samples even if the pulsation created by the LB-loading unit is strong enough
to break the flocs [61].

Fig. 19 shows that when the blade dewatering intensity is increased, floc breaking
(and smoothing of the crushed structure) starts first inside the paper web (see
changes in floccy and "crushed" areas in Fig. 19). The region where the floc
breaking starts can be estimated to be 1/3 of the web thickness measured from the
paper bottom side in case of the highest headbox flow. The floc breaking capability
is also related to the headbox flow. The forming roll area dewatering remained
relatively constant with different headbox flows, but the LB-unit dewatering increased
as a function of increased headbox flow and blade loading intensity. Thus it is
concluded that the incoming consistency of the LB-loading unit is related to the
headbox flow and therefore also to the floc breaking capability.

Zhao et al. [59] showed that increased consistency in the blade section leads to
increased peak pressure (picture a in Fig. 21). Further, according to Zhao, to create
relative fiber displacement between the mat and the remaining undrained
suspension, the shear stress must be higher than the yield strength of the undrained
stock, but not of the formed mat. Bennington [64] showed that the yield stress of the
undrained suspension depends strongly on the consistency. Using Zhao's findings, it
is possible to explain the behavior of floccy and crushed areas due to blade forces
by introducing the assumption of a consistency profile of the web in some phase of
 32

the LB-loading unit. This hypothesis of the floc breaking power of the LB-unit could
be stated according to Fig. 21.
Paper
center surface

Consistency or yield stress

Blade pulse gradient

consistency
high
low Consistency or
Pressure

yield stress
formation
improvement Pulse gradient

Blade

a b

Fig. 21. a) The effect of web consistency on blade pressure (Zhao et al.). b)
Hypothetical relationship between web consistency, blade pulse intensity (blade
force) and steepness (gradient) in terms of the floc breaking capability of the LB-
loading unit.

The capability of the LB-loading unit to break flocs (improve formation) is related to
the blade pulse intensity (blade force), steepness (gradient) and web consistency at
given positions in the paper thickness direction. Picture b in Fig. 21 shows a
hypothetical situation (consistency/yield stress of the undrained suspension and
pulse gradient) at the moment of peak pressure (picture a in Fig. 21). On paper
surfaces, due to the higher consistency, blade peak pressure is higher than in the
paper center area where consistency is lower. This means that the pressure pulse
gradient (rate of change) decreases when approaching the paper center area. The
higher the consistency difference between the surfaces and the center area is, the
bigger the change in pulse gradient. On the other hand it is known that in order to be
able to improve formation in the blade section, the consistency has to be low
enough. With a combination of proper consistency profile and blade pulse intensity
and gradient, the area where formation improvement is possible can be found. With
low basis weights it is assumed that the role of the consistency profile is diminished
and the floc breaking capability is related only to the blade pulse intensity and pulse
shape. According to the hypothesis, in the case of a straight consistency profile the
pulse gradient is also the same straight across the paper web.

It was also shown that filler distributions can be significantly affected during the blade
dewatering phase, with the mobility of fibers simultaneously remaining insignificant.
Fig. 22 shows that the headbox flow has a strong influence on filler distribution. On
the other hand it is also shown that significant changes in filler distribution result from
changing the blade dewatering intensity.

The process of modeling the filler distributions by characterizing the filler distribution
symmetry and shape factors by means of the general wet end controlling parameters
has already been introduced in this paper. Filler distributions can also been modeled
by using only relative dewatering numbers: forming section total dewatering related
to the case A0 (lowest headbox flow and minimized blade dewatering) and forming
 33

area dewatering two-sidedness value (TS/BS). The mathematical definition for the
estimates of the characters, Fue and Fsyme, can be stated in the following way:

Fue = f(total former dewatering, dewatering two-sidedness (TS/BS)) (7)

and
Fsyme = f(total former dewatering, dewatering two-sidedness (TS/BS)). (8)

Using the estimated filler symmetry and shape factors and Equations 1-4, the
estimates for top, bottom and center filler contents can be calculated, and further
estimates can be made for the filler distributions. The original and estimated filler
distributions are presented in Fig. 22. Solid lines represent the original filler
distributions and dashed line the estimates.
20 % 20 % 20 %

F ille r c o n te n t
F ille r c o n te n t

F ille r c o n te n t

C0 C1 C2

0 0 B a sis w e ig h t 1 0 0 % 0 0 B a sis w e ig h t 1 0 0 0% 0 B a sis w e ig h t 1 0 0 %

flo w ra te

BS TS
20 % 20 % 20 %
F ille r c o n te n t

F ille r c o n te n t

F ille r c o n te n t

B0 B1 B2
In c re a s in g

0 0 B a sis w e ig h t 1 0 0 % 0
0 B a sis w e ig h t 1 0 0 0% 0 B a sis w e ig h t 1 0 0 %
20 % 20 % 20 %
F ille r c o n te n t

F ille r c o n te n t

F ille r c o n te n t

A0 A1 A2

0 0 B a sis w e ig h t 1 0 0 % 0
0 B a sis w e ig h t 1 0 0 0% 0 B a sis w e ig h t 1 0 0 %
In c re a s in g b la d e d e w a te rin g

Fig. 22. Original filler distributions (solid lines) and estimates (broken lines). The
estimates are based on model. 80 gsm paper.

Results show that the forming section dewatering numbers determine the filler
distribution characters relatively well. The tests were performed with 80 gsm fine
paper.

The shape factor responded similarly to the headbox flow rate changes as was the
case in the other study [III] performed by the author. The correlation of forming
section parameters with filler distribution shape characters was less significant with
80 gsm paper, because the vacuum adjustments in the LB-loading unit were limited
and trials were performed with constant blade pressure.
 34

THE USE OF ADDITIVES LAYERING AS A PART OF PAPER QUALITY

CONTROL

Paper quality requirements depend greatly upon the end users' needs.
Requirements for coated base papers can differ significantly from those for copy
papers. Base paper quality parameters such as porosity and surface roughness are
important factors affecting the quality of coated paper. The correlation between the
smoothness of base paper and the smoothness of coated paper is clear. Base paper
properties should also be symmetrical in order to avoid the need for asymmetrical
treatment of the base paper [65-67]. High base paper surface filler content
decreases coating color penetration into the base paper structure and the coating
color stays better on the surface [43-45]. For copy paper grades the ideal filler
distribution needs to be symmetrical and the surface filler content lower than the
average filler content to reduce dusting [24,26].

It has been shown that dewatering conditions during drainage are responsible for the
movement of fines and fillers in the thickness direction of a sheet [III, 24-26,65], while
later process phases, such as pressing, no longer have a significant effect on filler
distributions [68]. The role of different forming concepts on the control of fines and
filler distributions can vary. Fourdriniers show low fines and filler content on the wire
side. Hybrid formers reduce top side fines and filler content which results in a more
symmetrical distribution in which the filler is concentrated in the center of the sheet.
Gap former results show that the fines and filler distributions are symmetrical and
more or less U-shaped [25].

Even if the importance of the z-directional distribution of fillers and fines on paper
quality has been noted, less attention has been paid to how to control the
distributions. In many cases multilayering has been considered for its potential to
improve some paper quality properties. Less attention has been paid to the role and
use of the multilayering process as a tool in the control of critical quality parameters.
This is important, particularly in those cases when other process variables are also
present.

In the study performed by the author on fine paper [IV], one of the main goals was to
find out whether there is any relationship between filler distributions and other paper
quality parameters. The study was done by using a 3-layer headbox and roll and
blade gap former. Wet end additives layering principles were used to control the filler
distribution together with the normal headbox and former running parameters. The
models were created and used to study the relations between the filler distributions
and the chosen quality parameters. The parameters were extracted from the pilot
paper machine trial results using a statistical approach. The quality parameters
chosen in this case were formation, porosity and oil absorption two-sidedness. The
main focus of the study was on 50 and 80 gsm woodfree papers.
 35

Estimates

The calculation of filler distribution estimates was based on Equations 5 and 6. The
next step was to calculate the estimates for the porosity (Poroe), formation index
(Formae), and oil absorption two-sidedness (Oilabe). The mathematical definition for
the factors can be expressed in the following way:

Poroe = f(Psym, Pu, Q, FRvac, LBvac1, LBvac2), (9)

Formae = f(Psym, Pu, Q, FRvac, LBvac1, LBvac2), (10)

and
Oilabe = f(Psym, Pu, Q, FRvac, LBvac1, LBvac2), (11)

where Psym = retention chemical dosing symmetry factor,

Pu = retention chemical dosing shape factor,
Q = headbox flow rate,
FRvac = forming roll vacuum,
LBvac1 = vacuum level of the lower suction box in the LB-loading unit,
LBvac2 = vacuum level of the upper suction box in the LB-loading unit.

A regression analysis method (Central Composite Surface Design [32]) was used for
the calculation of the coefficients for the factors presented in equations (9), (10) and
(11). In Appendix 6 has been shown the estimated vs. measured, porosity, formation
index and oil absorption two-sidedness for both 50 and 80 gsm papers.

The results showed a clear dependence of the quality factors in question on the filler
distribution characteristics. Formation improvement and densification of the paper
were achievable at the same running conditions. The strong correlation between
formation and porosity can be seen in Fig. 23, where all the 80 gsm paper beta
formation values have been plotted against porosity values. In the picture, beta
formation index is prepared by dividing the measured beta formation value by the
square root of the sheet's basis weight in that particular point. The figure is made
using the original measured data.
0.8
Beta formation index, √(g/m2)

0.7

0.6

0.5

0.4

0.3
200 300 400 500 600 700 800 900
Porosity, ml/min
Fig. 23. Beta formation index as a function of porosity (Bendtsen). 80 gsm paper.
 36

Low porosity favored evenly distributed constituents (good formation) and a local rich
filler concentration somewhere in the paper structure in the paper thickness direction
(Fig. 24a). Figure shows all combinations of the filler distribution characteristics that
are possible when the flow rates and former vacuums are stepped within their
operating range.

The oil absorption two-sidedness minimum favored symmetrical filler distributions

(Fig. 24b). The optimum regions could be found in the area in which the filler
distribution symmetry factor had its minimum value, i.e. when the difference between
filler surface contents in the filler distribution was minimized. The optimum value
was, in practice, independent of the shape (factor) of the filler distribution.

Filler shape factor Filler shape factor

a b

Fig. 24. a) Porosity (Bendtsen) and b) oil absorption two-sidedness as a function of

filler distribution shape and symmetry factors. 80 gsm paper.

Multilayering was found to be an effective tool for the control of filler distributions. On
the other hand it was shown that to get optimized impact on quality properties,
control of the other process variables is also needed. However the correct use of
multilayering can be used to expand the field of solutions. Fig. 25 shows an example
of all possible formation and porosity combinations achievable using headbox and
forming section parameters within a certain range a) without and b) with the use of
multilayering. It can be seen that, for example at a given formation level, the use of
the additives layering principle for the control of filler distributions makes lower
porosity numbers possible.
 37

Beta formation index ,√(g/m2)

Beta formation index ,√(g/m2)

a b
Fig. 25. Beta formation index as a function of porosity (Bendtsen). a) Retention
chemical equally distributed. b) Retention chemical dosing shape factor
corresponded with a split change from 30/40/30 % to 43/14/43 % (bottom/middle/top
side) and split symmetry factor corresponded with a split change from 26.7/33.3/40%
to 46.7/33.3/20%. Squares represent the original data points. 50 gsm paper.

CONCLUSIONS

In this study the main focus was on layering fine paper with a 3-layer headbox and
roll blade gap former. The study consisted of methodological development and
layering studies on a laboratory scale as well as on a pilot paper machine scale.
Studies concerning the potential of layered structures in papermaking and the
phenomena affecting layer mixing were performed.

The studies showed that filler distribution shape and symmetry factors correlate
clearly with certain quality parameters of fine papers, such as porosity, formation and
oil absorption two-sidedness. Low porosity was achieved in cases in which the paper
had good formation (evenly distributed constituents) and a filler distribution
characterized by local rich filler content (increased filler concentration) somewhere in
the paper’s cross-section. Oil absorption two-sidedness could be minimized by a
filler distribution that was as symmetrical as possible. It was shown in multilayer
handsheet former research that the layering of filler has a clear influence on paper
brightness and in-plane strength properties. When the filler was layered in the paper
surface layers, it was possible to affect the base paper’s surface roughness to the
extent that the smoothness achieved by the layering was also visible in pre-
calendering and in the coated final product.

With respect to layer mixing during the different stages of former dewatering, it was
observed that different types of paper constituents behave in different ways. The
mixing of the layers was different in roll dewatering compared with blade dewatering.
In blade dewatering the roughest fractions, such as fibers, were mobile to the extent
that formation improvement was possible – but no more. On the other hand, major
changes could be obtained in filler distribution shape and symmetry factors by
adjusting the blade dewatering. Based on the results, a model was created by which
 38

it was possible to explain the observations made during the study concerning the
factors affecting formation improvement in the LB-loading unit of the gap former.

In summary it can be said that, given the layering technology in use it will be
essential to acquire more accurate control over the behavior of all the factors
affecting layering before it is possible to take a number of quality factors into account
in order to achieve an overall quality optimum. This is true regardless of whether the
question concerns a filler-layered product or a product in which the goal is to layer
the fibers. However, there is much evidence to show that, with careful optimization, it
will be possible to achieve the layering objectives without neglecting other important
quality factors.

REFERENCES

1. Helmer, R.J.N., Covey, G.H., Raverty, W.D., Vanderhoek, N., Preliminary

development of a laboratory former for oriented sheets. Appita J., 53(6): 465-470
(2000).

2. Moller, K. , Jonsson, L. , Ek, R., A new handsheet former simulating fourdrinier

forming. Svensk Papperstidning, 82(9): 267-271 (1979).

3. Trepanier, R.J., Dynamic Forming: Parameters for making oriented sheets in the
laboratory. Pulp Pap. Can. 90(4): 45-48 (1989).

4. Wildfong, V. J., Genco, J. M., Shands, J. A., Bousfield, D. W., Filtration

mechanics of sheet forming. Part 1: Apparatus for determination of constant-
pressure filtration resistance. Beloit , University of Maine. J. Pulp Pap. Sci., 26(7):
250-254 (2000).

5. Bergström, J.I., Peel, J.D., Fundamentals of Structured Webforming. 18th

EUCEPA Conference London, 120-130 (1979).

6. Htun, M., Fellers, C., The invariant mechanical properties of oriented handsheets.
Tappi, 65(4): 113-117 (1982).

7. Stöckmann, V. E., Performance of stratified sheets. TAPPI, 57(10): 108-110

(1974).

8. Erickson, L. D., Physical properties of stratified sheets. TAPPI, 60(10): 113-115

(1977).

9. von Herzen, L., Täyte- ja muut lisäaineet hienopaperin monikerrosrainauksessa.

Diplomityö, Teknillinen korkeakoulu, Puunjalostustekniikan osasto, 118p. (1996).

10. Stöckman, V.,E., Effect of bonding intensityprofile variations on performance of

paper. Annual Meeting Preprint, TAPPI, Atlanta, Ga., 39-45 (1974).
 39

11. Nanko, H., Mukoyoshi, S. , Ohsawa, J., Multi-layered handsheet making with
laboratory sheet machine. Jpn Tappi, 41(3): 49-54 (1987).

12. Moore, G.K., Multiply web structure and performance. Multi-ply forming forum,
Atlanta, GA, USA,13-21 (1998).

13. Sauret, G., Trinh, H.J., Lefebvre, G., Versuche zur Herstellung mehrlagiger
Industriekartons im Laboratorium. Das Papier, 23(1): 8-12 (1969).

14. Bristow, J.A., Pauler, N., Multilayer structures in printing papers. Paper I-3.
Eucepa, Manufacturing Processes and Control in the Paper and Printing
Industries, 1-26 (1982).

15. Hamelink, J., The Effect of Fiber Distribution upon Stiffness of Paperboard.
Tappi, 46(10): (1963).

16. Ranger, A. E., Flexural stiffness of multi-ply boxboard. Paper Technology, 8(1):
51-57 (1967).

17. Sivén, S., Manner, H., Twin wire retention tester. 2000 Tappi Papermakers
Conference, Vancouver, BC, Canada, 767-773 (2000).

18. Sivén, S., Manner, H., High speed retention tester. PAPTAC 86th Annual Meeting,
A37-A41 (2000).

19. Hammock, C., Garnier, G., Laboratory twin former, part 1: Design and
performance. Tappi J., 83(10): 13 p. (2000).

20. Erkkilä, A-L., Mechanisms and measurements of the layered orientation structure
of paper sheets. Thesis for the Degree of Licentiate of Philosophy. Department of
physics, University of Jyväskylä, Finland, Laboratory report 2/1995, 65 p. (1995).

21. Erkkilä, A-L., Pakarinen, P., Odell, M., The Effect of Forming Mechanisms on
Layered Fiber Structure in Roll and Blade Gap Forming. TAPPI 99 - preparing for
the next millennium, Atlanta, GA, USA, 2, p. 389-400 (1999).

22. Harwood, J., W., Stratification of Paper Grades. Tappi Journal, 73(5): 115-122
(1990).

23. Häggblom-Ahnger, U., Layering of office paper. Paperi ja Puu- Paper and
Timber, 80(7): 508-513 (1998).

24. Häggblom-Ahnger, U., Pakarinen, P., Odell, M., Eklund, D., Conventional and
stratified forming of office paper grades. Tappi Journal, 81(5): 149-158 (1998).

25. Kinnunen, J., Lloyd, M.D., Nygren-Konttinen, A., A new world of papermaking -
layered SC-paper. Paperi ja Puu- Paper and Timber, 80(3): 1-5 (1998).

26. Odell, M., Paper Structure Engineering. Appita Journal, 3(5): 371-377 (2000).
 40

27. Dickson, R., Forsström, U., Grön, J., Coating coverage of meter size press pre-
coated paper. TAPPI Coating & Graphic Arts Conference, May 1-3, Washington,
DC, USA, 167-184 (2000).

28. Tanaka, H., Luner, P., Côté, W., How retention aids change the distribution of
filler in paper. Tappi 65(4): 95-99 (1982).

29. Li, A.C., Rogers, T.D., Shands, J.A., Mixing intensity analysis of a 2-layer
stratified headbox jet. Tappi Journal, 83(12): 16 (2000).

30. Andersson, I., Simultaneous forming of multi-layer products with the KMW air
wedge technology. Papex ´82 International - Papermaking in the Eighties, PITA
Harrogate, Yorkshire, 86-97 (1982).

31. Andersson, I., Multilayering on a roll former. SPCI World Pulp and Paper Week -
New Available Techniques, SPCI Stockholm, 157-163 (1984).

32. Montgomery, D., C., Design and Analysis of Experiments. Fourth Edition, John
Wiley and Sons, 704p. (1996).

33. Lloyd, M.D., Norman, B., Layer mixing during three-layer stratified forming: the
role of vane length and mix-wire speed difference. Tappi Journal 81(11): 194-202
(1998).

34. Attwood, B. W., Multi-ply web-forming - past, present, and future. TAPPI, 63(8):
78-82 (1980).

35. Koran, Z., Multi-ply forming – improves properties and cuts costs. Paper
Technology, 32(9): 38-42 (1991).

36. Bergström, J.I., Stratified web formation, current and potential applications. Norsk
skogindustri, 31:(5), 133-137, 1977.

37. Cumming, R.M., Reducing market pulp consumption or its costs 1 – By improved
machinery and control. Paper, 189(6): 346-376 (1978).

38. Ewald, J.L., Forming layered products allows best use of different furnishes, Pulp
& Paper, 137-140 (1981).

39. Collins, N.J., Colley, J., Layered Paper Structures- the Influence of Furnish
Geometry on Physical Properties. Appita, 33(2): 119- 123 (1979).

40. Odell, M., Multilayering, method or madness. XI Valmet Paper Technology Days,
53-61 (1998).

41. Häggblom-Ahnger, U. Optimum location of softwood sulfate pulp in three-ply

office paper. Tappi Journal, 82(6): 181-187 (1998).
 41

42. Häggblom-Ahnger, U. Fractionation of softwood sulfate pulp in three-ply copy

paper. 2nd ECO PAPER TECH Conference, Helsinki, 195-206 (1998).

43. Forsström, U., Dickson, R.J., Grön, J., Strichdeckung und Struktur von holzfreien
Papieren, die in der Filmpresse vorgestrichen wurden. Wochenblatt für
Papierfabrikation, 128(14/15): 985-992 (2000).

44. Grön, J., Ahlroos, J., Influence of base paper filler content and pre-calendering on
metered film press coating- Part I: A coating study. TAPPI Coating/Papermakers
Conference, May 4-6, New Orleans, LA, USA, 899-914 (1998).

45. Ahlroos, J., Alexandersson, M., Grön, J., Influence of base paper filler content
and pre-calendering on metered film press coating- Part II: Paper and print
quality. TAPPI Coating/Papermakers Conference, May 4-6, New Orleans, LA,
USA, 915-925 (1998).

46. Lindström, T., Some fundamental chemical aspects on paper forming. 9th
Fundamental Research Symposium 'Fundamentals of Papermaking, September
17-22, Cambridge, UK, 309-412 (1989).

47. Lindström, T., Retention aids - their mechanism of action and interactions. Wet
end chemistry conference and COST workshop, May 28-29, Gatwick, UK,
40(1997).

48. Baker, J.K., Beuther, P.D., Farrington (Jnr), T.E., Impact of three dimensional
headbox flows on tissue layer purity. AIChE Symposium Series, Advances in
Pulp and Papermaking, 132-142 (1995).

49. Jetsu, P., Kellomäki, M., Karema, H., Salmela, J., Lappalainen, T., Piirto, M.,
Coherent structures of suspension flow and their inheritance in paper. 12th
Fundamental Recearch Symposium, Oxford, September, 18 (2001).

50. Lloyd, M.D., Norman, B., Use of stepped vanes to alter headbox turbulence
levels during stratified forming of woodfree paper. Appita Journal 52(2): 98-104
(1999).

51. Lepomäki, H., Huovila, J., Karema, H., Control of paper properties through
headbox fluid dynamics, ZELLCHEMIN Annual Meeting 2001, T44-T47 (2001).

52. Söderberg, D., A comparison between the flow from a paper machine headbox
and a low Reynolds number water jet. Tappi Engineering Conference, Anaheim,
1155-1172 (1999).

53. Norman, B., Söderberg, D., Overview of forming literature 1990-2000. STFI
Report, PUB 3, 116 p (2001).
 42

54. Parsheh, M., Aspects of boundary layers, turbulence and mixing in a plane
contraction with elastic vanes, envisaging potential application to paper
manufacturing. Licentiate Thesis. Technical reports from Royal Institute of
Technology Department of Mechanics FaxenLaboratoriet, Sweden (1999).

55. Li, A.C., Neill, E.B., Rogers, T.D., Fundamental studies of mixing in a stratified
headbox jet. Paptac, 86th annual meeting, Montreal, A53-A56 (2000).

56. Niskanen, K., Kajanto, I., Pakarinen, P., Paper structure. Papermaking Science
and Technology 16. Paper Physics, edited by Gullichsen, J., Paulapuro, H.,
Niskanen, K., chapter 1, 14-53, [Helsinki, Finland: Fapet Oy, 324p. (1998)]

57. Retulainen, E., Niskanen, K., Nilsen, N., Fibers and bonds. Papermaking Science
and Technology 16. Paper Physics, edited by Gullichsen, J., Paulapuro, H.,
Niskanen, K., chapter 2, 54-87, [Helsinki, Finland: Fapet Oy, 324p. (1998)]

58. Wildfong, V.J., Genco, J.M., Shands, J.A., Bousfield, D.W., Filtration Mechanics
of Sheet Forming. Part II: Influence of Fine Material and Compression. Journal of
Pulp and Paper Science, 26(8): 280-283 (2000).

59. Zhao, R. H., Kerekes, R. J., The effect of consistency on pressure pulses in blade
gap formers, Pap. Puu, 78(1-2): 36-38 (1996).

60. Nordström, B., Norman, B., Effects of Headbox Jet Quality and Blade Pulse
Force on Formation, Retention, and Mechanical Properties During Roll-Blade
Forming with Curved Blades. Nordic Pulp & Paper Research Journal, 1, 33-40
(1995).

61. Norman, B., Hydrodynamic developments in twin-wire forming – an overview.

Paperi ja Puu – Paper and Timber 78(6-7): 376-381 (1996).

62. Kerekes, R.J., Schell, C.J., Characterization of Fibre Flocculation Regimes by a

Crowding Factor. J. Pulp Paper Science, 18(1): 32-38 (1992).

63. Kerekes, R.J., Schell, C.J., Effects of fiber length and coarseness on pulp
flocculation. Tappi Journal 78(2): 133-139 (1995).

64. Bennington, C. P. J., Kerekes, R. J., Grace, J. R., The yield stress of fibre
suspensions. Can. J. Chem. Eng. 68, 748-757 (1990).

65. Lehtinen, E. , Introduction to pigment coating of paper. Papermaking Science and

Technology 11. Pigment Coating and Surface Sizing of Paper, edited by
Gullichsen, J., Paulapuro, H., Lehtinen, E., chapter 1, p. 14-27, [Helsinki, Finland:
Fapet Oy, 2000, 810p.]

66. Hietanen, S. , Woodfree base paper. Papermaking Science and Technology 11.
Pigment Coating and Surface Sizing of Paper, edited by Gullichsen, J.,
Paulapuro, H., Lehtinen, E., chapter 3, p. 46-59, [Helsinki, Finland: Fapet Oy,
2000, 810p.]
 43

67. Grön, J., Rautiainen, P., Coating solutions for woodcontaining and woodfree
paper grades. Coating conference, Toronto, Canada, 457-467 (1999).

68. Szikla, Z. and Paulapuro, H., Z-directional distribution of fines and filler material in
the paper web under wet pressing conditions. Paperi ja Puu- Paper and Timber,
68(9): 654-664 (1986).
 44

APPENDICES

APPENDIX 1

The multilayer handsheet former.

Moving fabric Moving fabric

height

Lateral cylinder Lateral cylinder

Chambers that contain

components to be layered

width
The multilayer handsheet former.

Main dimensions:
• width 750 mm , depth 250 mm and hight 345 mm

Fabric frame dimensions:

• depth 250 mm and hight 345 mm

Fabrics used:
• typical production 2-layer fabrics for newsprint grades
• air permeability 6500 m3/h*m2
• no of yarns/ cm2: 59.5 yarns/ cm * 46.4 yarns /cm

Typical process variables:

• air mixing time 15 s and intensity 100 - 150 l/min,
• air exhausting time 12 s,
• forming speed from 15 * 10-3 m/s to 30 * 10-3 m/s (movements of fabrics)
• constriction force 250 N.
 45

APPENDIX 2 1/4

Test arrangements at pilot paper machine environment.

headbox

retention chemical dosage

polymer bentonite

surface layer fibers

center layer fibers
surface layer fibers

The additives and fiber layering wet end arrangements.

2
5

4 6
1 3

The 3-layer headbox.

 46

APPENDIX 2 2/4

Coding:

1. header
2. dilution feed
3. manifold tube bank
4. turbulence generator
5. slice chamber
6. layering vanes

Additional statistics:
• headbox slice chamber contraction 1/8..1/12 depending the slice opening
• layering vanes: 2 mm thick flexible layering vanes; the last 50 mm length of the
free end (downstream end) tapered to a final thickness of 0.7 mm; sharp tip
 47

APPENDIX 2 3/4

10
11

12

The roll and blade gap former.

18
15

14
17

16

13

The forming area of the roll and blade gap former.

 48

APPENDIX 2 4/4

Coding:

7. headbox
8. forming roll
9. breast roll
10. 1st suction roll
11. high vacuum suction box
12. pick-up roll
13. forming roll vacuum zone
14. LB shoe 1 suction element
15. LB shoe 2 suction element
16. LB pre-loading element
17. LB loading elements
18. suction box

Additional statistics:

• wrap angle (the area where the outermost fabric is producing gap pressure
against the forming roll) 15 º
• the radial width of the forming roll suction zone 15 º
• forming roll diameter 1600 mm
• fabrics: typical 2-layer fine paper fabrics; air permeability 5500..6500 m3/h*m2
• fabric tensions 8 kN/m
 49

APPENDIX 3

Table 1. Standard laboratory measuring methods.

Filler content Incineration method according to the SCAN P5:63

Furnish drainability (SR° value) Shopper-Riegler device calibrated according to the
SCAN-C 19:65
Beta formation Ambertec beta formation tester
Oil absorption Cobb-Unger oil absorption measuring method
according to the SCAN-P 37:77
Bendtsen Porosity Bendtsen measuring device calibrated according to
SCAN-P 21:67
 50

APPENDIX 4 1/2

Regression coefficients for filler distribution shape and symmetry factors, Fue and
Fsyme . Relates to the publication III: Wet end control of filler distributions using
additives layering principles.

50 gsm paper

Fsyme = k11*Psym+k12*Q+k13*Q*Q+k14*LBvac12+k10, where

k10=26.88
k11=0.30
k12=-0.26
k13=0.00064
k14=0.020

Fue = k21*Pu+k20, where

k20=0.19
k21=0.030

80 gsm paper

Fsyme = k11*Pu+k12*Psym+k13*FRvac+k10, where

k10=0.030
k11=-0.060
k12=0.28
k13=-0.033

Fue = k21*Vane length(rel)+k22*Pu+k23*Q+k24*FRvac+k25*Pu*Q+k20, where

k20=-6.46
k21=0.0043
k22=0.63
k23=0.031
k24=-0.022
k25=-0.0029
 51

APPENDIX 4 2/2

Regression coefficients for filler distribution shape and symmetry factors, Fue and
Fsyme . Relates to the publication V: Control of fiber mobility and filler distribution in
wet end layering.

Fsyme = k11* Total former dewatering+k12* Dewatering two-sidedness (TS/BS)+

k13* Total former dewatering * Dewatering two-sidedness (TS/BS) + k10, where

k10=-2.81
k11=2.28
k12=1.86
k13=-1.41

Fue = k21 * Total former dewatering + k20, where

k20=-1.41
k21=1.44
 52

APPENDIX 5 1/2

The headbox and former parameters related to the publication III: Wet end
control of filler distributions using additives layering principles.

Table 2. Headbox and former parameters. Publication III.

50 gsm paper 80 gsm paper

min max min max
Polymer dose symmetry, Psym -9 0 -9 0
Polymer dose shape, Pu 0 4,5 0 4,5

Layer vane length relative -49 +26 -49 +46

to HBX nozzle, mm
Head box flow rate, l/s*m 190 230 190 210

Forming roll vacuum, kPa -2 0 -20 -5

LB-shoe vacuums 1&2 -25 -10 -10 0

(combined), kPa

The headbox and former parameters related to the publication IV: Controlling
filler distributions for improved fine paper properties.

Table 3. Headbox and former parameters. Publication IV.

50 gsm paper 80 gsm paper

min max min max
Polymer dose symmetry, Psym -9 0 -9 0
Polymer dose shape, Pu 0 4,5 0 4,5

Head box flow rate, l/s*m 190 230 190 210

Forming roll vacuum, kPa -2 0 -20 -5

LB-shoe 1 vacuum, kPa -10 -5 -5 0

LB-shoe 2 vacuum, kPa -15 -5 -5 0

 53

APPENDIX 5 2/2

The headbox and former parameters related to the publication V: Control of

fiber mobility and filler distribution in wet end layering (V).

Table 4. Headbox and former parameters. Publication V.

Headbox flow rate, l/s*m 150,170,200

Forming roll vacuum, kPa -20, -10
Loading blade load pressure, kPa 0, 10, 20
LB-shoe 1 vacuum (Lb-blades), kPa -12, -5
LB-shoe 2 vacuum (next to Lb-shoe 1), kPa -20, -5
Dyed fibers in center layer not in use, in use
 54

APPENDIX 6

Estimated vs. measured, porosity, formation index and oil absorption two-sidedness
for both 50 and 80 gsm papers. Relates to the publication IV: Controlling filler
distributions for improved fine paper properties.
Bendtsen porosity ml/min, 50 gsm paper Bendtsen porosity ml/min, 80 gsm paper
Po rosity, 50 g sm pap er Po rosity, 80 gsm p aper
1200 900
y = 0.7429x + 203.15 y = 1.0875x - 18.9

Estim ated p oro sity, m l/m in

1100 800
Estim ated valu es, m l/m in

1000 700

900 600

800 500

700 400

600 300

500 200
500 600 700 800 900 1000 1100 1200 200 300 400 500 600 700 800 900

Measured values, m l/m in Measu red p orosity, m l/m in

A) B)
Estimated porosity as a function of measured porosity with A) 50 gsm paper and B)
80 gsm paper.

F orm atio n ind

Formation ex,√50
index g2sm
(g/m ), 50pap
gsmerpaper Formation n ind√ex,
F orm atioindex (g/m80
2 g sm
), 80 pap
gsm er
paper
1 0.9
y = 1.0646x - 0.0431
y = 1.0142x - 0.008
0.9 0.8
Estim ated valu es
Estim ated valu es

0.8 0.7

0.7 0.6

0.6 0.5

0.5 0.4

0.5 0.6 0.7 0.8 0.9 1 0.4 0.5 0.6 0.7 0.8 0.9

Measured values Measured values

A) B)
Estimated formation index as a function of measured formation index with A) 50 gsm
paper and B) 80 gsm paper.

Oil ab sorption tw o-sidedn ess, 50 gsm p ap er Oil ab sorption tw o-sidedn ess, 80 gsm p ap er
0.85
1
0.98 y = 0.9062x + 0.071
y = 0.6246x + 0.3405 0.8
0.96
Estim ated valu es
Estim ated valu es

0.94
0.92 0.75

0.9
0.88 0.7

0.86
0.84 0.65

0.82
0.8 0.6
0.8 0.85 0.9 0.95 1 0.6 0.65 0.7 0.75 0.8 0.85

Measured values Measured values

A) B)
Estimated oil absorption two-sidedness as a function of measured oil absorption two-
sidedness with A) 50 gsm paper and B) 80 gsm paper.