Fluid Mechanics Papermaking

FL43CH09-Alfredsson ARI 15 November 2010 13:28
ANNUAL
REVIEWS Further
Fluid Mechanics
Click here for quick links to
Annual Reviews content online, of Papermaking
including:
• Other articles in this volume
• Top cited articles Fredrik Lundell,1,2 L. Daniel Söderberg,2,3
• Top downloaded articles
Annu. Rev. Fluid Mech. 2011.43:195-217. Downloaded from www.annualreviews.org
• Our comprehensive search and P. Henrik Alfredsson1

by California Institute of Technology on 05/18/13. For personal use only.
1
Linné FLOW Center, KTH Mechanics, Royal Institute of Technology, S-100 44 Stockholm,
Sweden; email: [email protected], [email protected]
2
Wallenberg Wood Science Center, Royal Institute of Technology, S-100 44 Stockholm,
Sweden
3
INNVENTIA, S-114 86 Stockholm, Sweden; email: [email protected]
Annu. Rev. Fluid Mech. 2011. 43:195–217 Keywords

First published online as a Review in Advance on fiber suspension, fiber orientation, stability, jet flow
August 19, 2010
The Annual Review of Fluid Mechanics is online at Abstract

fluid.annualreviews.org
Papermaking is to a large extent a multiphase flow process in which the
This article’s doi: structure of the material and many of the relevant properties of the final
10.1146/annurev-fluid-122109-160700
product are determined by the interaction between water and the wood
Copyright c 2011 by Annual Reviews. fibers. The dominant feature of a suspension composed of wood fibers and
All rights reserved
water is its inherent propensity to form bundles of mechanically entangled
0066-4189/11/0115-0195$20.00 fibers, known as fiber flocs. However, the phenomena apparent throughout
the papermaking process are not unique but in fact have a generic fluid
dynamical nature.
195
1. INTRODUCTION TO PAPER MANUFACTURING
1.1. What Is Paper?

Stock: the suspension
of pulp in water that is The word paper has its origin from papyrus, a plant that grows along the Nile River and was
fed to the paper used in ancient Egypt to make written records. However, paper as we know it is made from
machine cellulose fibers, mainly from wood; hence it is based on a renewable resource. After use, it can
Pulp: the fibrous be recirculated or burnt to produce electricity and/or heat. In almost all parts of our lives we
material prepared by encounter paper products: books, newspapers, magazines, packaging material for food and other
chemically or
products, cleaning towels, toilet paper, coffee filters, and cups, just to name a few. Regardless
mechanically
separating fibers from of the application, chemistry, physics, and fluid mechanics play roles in the production process.
wood, fiber crops, or Although the end product can be quite different, the paper-manufacturing process is basically the
waste paper same.
It is believed that the first paper sheets were manufactured in China more than 2,000 years
ago. In that case, fibers from plants were released by wetting and beating them into a fiber-water
mixture. The mixture, which today we call stock or pulp suspension, was then drained through a
sieve; the fibers remained on the sieve, were left to dry, and then were formed into a paper sheet.
In principle, paper manufacturing mostly follows the same principle even today, although it has
developed into a continuous process.
Most paper is made through a wet process in which the cellulose fibers are released from
the wood (so-called virgin fibers) through either a chemical and/or mechanical process and then
are mixed with water to a fairly low concentration, 0.1%–1%. However, in many processes,
recirculated fibers are also mixed with virgin fibers or used by themselves. In principle, the idea of
the paper-manufacturing process is to get rid of the water and keep the fibers, distributing them
in such a way that a homogeneous paper sheet results. Depending on the end paper product, the
characteristics vary, with important ones including the so-called basis weight and the density. For
newsprint (which is fed at high speed through a printing press), one wants the fibers to be oriented
mainly in the direction of the machine, making it strong in that direction, whereas for high-quality
magazines or laser printer paper, one would instead like to have an isotropic distribution of the
fiber direction in the paper plane.
1.2. The Paper Industry

The amount of paper products produced across the world was estimated to be nearly 400 million
tons in 2008. In the United States, the annual per capita consumption of paper is approximately
290 kg per year, whereas it is 130 kg in Europe and approximately 30 kg in Asia.
Half of the raw material for paper (i.e., the pulp) is based on recycled paper, and the other
half is so-called virgin pulp. The basis for the virgin pulp is mainly wood raw material, but annual
plants such as bamboo are also used.
Wood pulp comes from softwood trees such as spruce, pine, fir, larch, and hemlock and hard-
woods such as eucalyptus, aspen, and birch. The species defines the morphological properties of
the pulp such as fiber length and diameter. During the past few years the use of fast-growing
trees (e.g., eucalyptus) from plantations in the southern hemisphere has increased. Generally pulp
based on species that have longer fibers is used for products for which strength is required, such
as paperboard used for packaging.
The pulp and paper industry is a large consumer of energy. The annual global energy con-
sumption is more than 6 EJ, with 75% representing thermal energy (fossil fuels and biomass) and
25% representing electricity. This is more than 1% of the world’s total energy consumption, and
today there is a strong industry focus on decreasing this consumption.
196 Lundell · Söderberg · Alfredsson

1.3. This Paper

This review serves two purposes. The first is to introduce papermaking to a person with a back-
ground in fluid mechanics. As a tutorial, this approach is somewhat different from the typical
introductions to papermaking, as the engineering area traditionally is presented as chemical engi-
neering from the process perspective and mechanical engineering from the machine perspective.
The second purpose is to show in detail some interesting fluid mechanics problems that arise in
papermaking and areas in which fluid mechanics can help to understand the physics and possibly
aid in solving problems.
Section 2 describes the main process steps once the delivered wood or recirculated fibers reach
the paper mill. Section 3 gives a description of the basic material (the cellulose fiber) and briefly
introduces fiber suspension fluid mechanics. Section 4 presents in detail the fluid dynamics of the
forming section during which a fiber suspension at 0.1%–1% concentration is transformed into a
thin fiber suspension sheet on a permeable forming fabric. In Section 5 future developments and
fluid mechanical opportunities are discussed.

2. THE PAPER MILL AND ITS MAIN FLUID DYNAMICS

PROCESS STEPS
There are many different process steps from raw material to the final product in a paper mill.
Some mills start with wood, whereas others buy pulp. Figure 1 shows the full process chain,
from wood logs entering the mill to paper on rolls coming out. A modern paper machine can
be more than 200 m long, the width of the paper sheet can be more than 10 m, and the paper
can be produced at speeds up to 2,200 m min−1 (37 m s−1 ). Fluid mechanics play important roles
during many phases of the process, but in this short description we focus on the process steps
after the wood has turned into pulp, i.e., the mechanical treatment of the fibers (refining), through
the process of making the pulp suitable for entering the paper machine (cleaning, fractionation,
and separation), the spreading of the fiber suspension into a thin layer on the machine fabric
(formation) in which most of the water is drained and the fiber network formed (consolidation),
and finally the removal of most of the water from the fibers (pressing). Furthermore, drying, and
sometimes also the coating of the sheet, involves fluid dynamics. Another interesting aspect is the
fluid-structure interaction between the fast-moving paper sheet and the surrounding air.
Raw material Pulping Additives Refining Fractionation
Sheet Pressing Drying Surface Product

forming treatment
Figure 1
The main process steps of paper manufacturing.
www.annualreviews.org • Fluid Mechanics of Papermaking 197

Fiber flocs between the bars

Stator plate Rotation
Rotor plate
Pulp Rotor
suspension Energy
Plate gap
Stator
Refined fibers
Figure 2
The refiner (left) and a close-up (right) of the refining gap, showing how the fibers and fiber flocs are mechanically treated between the
rotor and stator.
2.1. Refining
The papermaking process starts with the blending of pulps of different types and the addition
of various chemicals and inorganic fillers. In addition, the fibers are subjected to the mechanical
treatment necessary to make them suitable for the production of the desired sheet of paper. The
mechanical treatment is performed as a continuous process known as refining.
The principle of the refining process is that two working surfaces in relative motion squeeze
the fiber suspension passing between them (see Figure 2). These surfaces are not smooth but
are covered with bars (millimeter scale) of various patterns. As the surfaces move relative to each
other, the fiber suspension passes through the narrow gaps formed between opposing bars. This
treatment aims to change the fiber structure in a controlled way, improving both fiber flexibility and
the surface properties of the fiber. The refining process has been extensively studied throughout
papermaking history and is one of the high–energy consumption processes of the paper mill.
2.2. Cleaning, Fractionation, and Separation

Prior to the forming of the actual paper sheet, the pulp suspension prepared in the refining process
is treated in various ways depending on the final product and on the type of paper machine, but the
suspension is normally diluted to a concentration of less than 1% before it enters the paper machine.
This is, with present-day technology, an essential step to avoid excessive fiber flocculation, i.e., the
appearance of fiber flocs, which would otherwise give poor paper quality. In addition, unwanted
particles in the suspension that may have a detrimental effect on both product properties and
machine runnability have to be removed. Screens and hydrocyclones are used to achieve this.
In addition to cleaning the suspension, these processes are also able to sort fibers into different
fractions on the basis of morphological properties. By using fibers from the different fractions, it
is possible to some extent to tailor the properties of the final product, thereby making it suitable
for different purposes.
The possibility of fractionating the pulp has a high potential for selective fiber treatment, but
it is used in only a few industrial installations. This breakthrough has so far been hindered by the
low efficiency of the fractionation process. Of particular interest is the possibility of fractionating
pulp on the basis of bending stiffness, which has been made possible by the use of hydrocyclones
(see Figure 3). However, the fluid mechanics behind the fractionation in the hydrocyclone are
not understood, and the present process requires too much water, too much pumping energy, and
too large investments in equipment.

Thin-walled fibers
Thick-walled fibers
Hydroclone
Figure 3
Schematic showing how a hydrocyclone can separate thin-walled and thick-walled fibers in the fractionation
process.
2.3. Sheet Forming

In the early days of papermaking, so-called handsheets were made by draining the suspension
through a stationary wire. The step from handsheet forming of paper to a continuous process
was made possible by the Frenchman Nicholas Louis Robert in 1798. He invented the sheet-
forming process that still is used today in modern paper manufacturing. In this process, a planar
fiber suspension jet, produced by a nozzle or headbox, impinges on a moving endless permeable
belt referred to as the wire or forming fabric. The forming fabric traps the fibers as the water is
drained through the fabric. At first, drainage was through gravity; however, this process is too slow
to be effective. By introducing rotating rolls below and in touch with the fabric, it was possible
to create a large negative pressure pulse when the fabric leaves the roll. In this way, drainage can
dramatically increase as Taylor (1957) demonstrated via analytic and experimental analysis. This
type of paper machine is today referred to as a Fourdrinier machine, and it is widely used in paper
manufacturing.
Early Fourdrinier machines used an open headbox mounted on top of the forming table. At
the bottom of the headbox, a slit allowed the fiber suspension to be continuously delivered on top
of the wire. The speed of the jet was controlled by the height of the water column in the headbox.
To increase speed, it was necessary to increase the height of the water column. If the speed is
doubled, the water level must be four times higher. This initiated the introduction of modern Wire: the traditional
permeable weave used
closed headboxes, in which the necessary pressure is supplied by pumps.
to trap fibers from the
In the 1950s experiments were performed in which the fiber suspension was introduced between stock while water is
two forming fabrics, wrapped around a rotating roll directly after jet impingement. This forming allowed to drain
process is called twin-wire or gap forming. The main aim of these experiments was to improve the Headbox: used to
dewatering capacity. In fact, the introduction of two forming fabrics did not double the dewatering distribute the pulp
capacity but rather quadrupled it. This is a result of water being removed at two sides and reduced onto the wire in the
dewatering resistance because the fiber mats are only half of that in single-sided dewatering form of a plane liquid
jet
(Figure 4). In addition, the dewatering pressure, δp, can be made very high because it is simply
given as Forming fabric: the
modern version of the
δp = T/R, wire weaved as
multilayered structures
where T is the outer wire tension and R is the radius of the roll. During the drainage through the optimized to improve
drainage without
forming fabric(s), the fiber concentration in the suspension increases from the headbox concentra-
causing marks in the
tion to 3%–5%. By applying local pressure pulses to the developing network after the forming roll, paper
especially when using deflector blades, one can improve the homogeneity of the fiber structure,

Water flowing in a structured fiber network
Fiber suspension Fiber suspension

Forming fabrics
Water flowing in a structured fiber network Water flowing in a structured fiber network
Figure 4
Consolidation of the fiber network on one wire (left panel ) or between two wires (right panel ).
although no physical explanation has been given. Even in the fastest paper machines, 90% of the
water is removed within a few meters from the jet impingement; after that point the fibers will
not move relative to each other, and the network of fibers is said to be consolidated.
2.4. Pressing
After the forming section, the sheet still contains more than 80% water by mass, with the exact
amount depending on the machine speed and paper grade. As it is extremely costly in terms
of energy to remove water by evaporation, mechanical dewatering through pressing is used to
remove as much water as possible before the final drying of the paper. The traditional pressing
process involves squeezing the fiber network together with one or two supporting felts in a nip
between two rolls. The applied load compresses the web, and the water is pressed into the felt.
An increased efficiency in the press section not only reduces the cost for drying, but also improves
the runnability of the paper machine.
A modern paper machine usually has two or more nips in the press section. In addition, in a
modern press, a curved press shoe often replaces one of the rolls in the nip, and this can extend
the press nip length from a few centimeters up to 25 cm. When the fiber network is compressed,
water is pressed out from the web. This is a highly dynamic process. Consequently, the press load
is taken up both by a deformation of the network as such and by the flow of water through (and
ultimately out of) the network. The process is further complicated by the fact that the water is
trapped at three levels: between the fibers, in the hollow core of the fiber, and within the fiber
wall.
Forming section: 3. CELLULOSE FIBERS AND FIBER SUSPENSIONS

part of the paper
machine comprised of 3.1. The Cellulose Fiber
the headbox and
wire(s), where the The raw material in papermaking is composed of wood fibers that are mixed with water to form
sheet structure is a suspension referred to as pulp. Pulp is primarily made from wood chips subjected to chemicals,
determined heat, and/or mechanical grinding, although recirculated cellulose fibers are becoming increasingly
Nip: place of important. These processes disintegrate the wood into cellulose fibers with a typical length of
intersection where one 1–2 mm and a diameter of approximately 20 μm (see Figure 5). The cellulose fiber is shaped as a
roll touches another
such as between the
cylinder in which the hollow core is referred to as lumen. The fiber is seldom straight, however;
press rolls it is flexible, and its shape can be affected by both the flow and the interaction with neighboring
fibers.

Latewoo
Latewood
ood
arlywood
Earl
Earlywood
Figure 5
Cellulose fibers in a tree (left panel ) and fibers after the pulp-making process (right panel ). Depending on the
tree’s growth season, the fiber walls can be thicker or thinner, which influences their flexibility. Earlywood
grows during the spring, whereas latewood grows during the summer.
In general the behavior of a fiber suspension depends on many different parameters, such as
concentration (given as either volume or mass), the density ratio between the fiber material and
the fluid, the fiber aspect ratio, fiber surface properties, and flexibility, as well as the viscosity
of the fluid (or, rather, the Reynolds number). However, the fiber suspension used in paper
manufacturing does not always behave as well as a dilute idealized fiber suspension consisting of
a Newtonian liquid with nonbuoyant rod-like stiff particles (cylinders). A dominant feature of a
suspension composed of wood fibers and water (above a certain low concentration) is its inherent
propensity to form bundles of mechanically entangled fibers, usually called fiber flocs (Mason &
Manley 1957). The main mechanism of fiber flocculation is the mechanical bonding of fibers to
each other, but both the complex morphological properties of the cellulose fibers and the chemical
environment play roles as well (see, e.g., Kerekes 2006).
3.2. The Pulp Fiber Suspension

The behavior of the pulp fiber suspension depends on the fiber length-to-diameter ratio and fiber
concentration, among other things, but also on the flow itself. There is an intricate interaction
between the flow and the fibers, which may influence transition between laminar and turbulent
flow as well as the turbulence as such.
3.2.1. Flocculation and the crowding factor. A paper sheet can be characterized by its mass
per unit area, which is referred to as its basis weight. Generally the main quality measure of a
sheet is the mass variation in the plane, i.e., the basis-weight variation, which is referred to as
the formation. Variation of the basis weight can easily be seen by holding a paper sheet in front
of a light source, by which the presence of fiber flocs can then easily be detected. As mentioned
above, the fiber concentration has a strong influence on the behavior of the suspension, but it
cannot in itself determine whether flocculation will occur. Soszynski & Kerekes (1988) defined
the propensity for cellulose fibers to flocculate in terms of a dimensionless number, referred to as
the crowding number, N. It is defined as
2
2 L
N = Cv , (1)
3 d
where Cv is the volume concentration of the fiber material in the suspension, L is the fiber length,
and d is the diameter of the fibers. N can be viewed as the number of fibers inside the volume of
a sphere with diameter L, and Figure 6 shows N as function of the concentration for different
L/d. The density of cellulose is approximately 50% higher than water, but due to the lumen and

1,000 L/d = 1,000 Network
100
N 10
L/d = 100 Semidilute
0.1
Dilute
L/d = 10
0.01
0.01 0.1 1.0 10

Cv (%)
Figure 6
Crowding factor as a function of fiber volume concentration (Cv ) for different length to diameter (L/d ) ratios.
swelling of the fiber wall, the volume concentration can be twice as high as the mass concentration.
It should be noted that a value representing volume concentration is difficult to obtain. Hence the
crowding number is usually estimated using the mass concentration.
Flocculation does not generally occur when N 1, as each fiber then can rotate freely
and seldom hits another fiber. Such a suspension is usually called dilute. Handsheets used for
laboratory testing are usually made at N ≈ 1, which gives an even sheet structure and thereby a
good formation. The range 1 < N < 60 is called the semidilute range; here fibers flocculate, and
it is in this range that most papermaking operations are performed. Kiviranta & Dodson (1995)
showed the usefulness of the crowding number as a tool to characterize the conditions during the
sheet-forming process. In the concentrated region, i.e., N > 60, fiber mobility is greatly hindered,
as the suspension forms networks.
Fiber flocs are detrimental when high-quality paper is the goal, and various strategies to disrupt
fiber flocs by generating suitable flow conditions during the sheet-forming process that minimize
flocculation have been tried. High-level turbulence, normal shear, and highly extensional flows
are candidate strategies. In a numerical study of flocculation in turbulent flows, Steen (1991)
introduced a measure for the rate of rupture and aggregation in pipe flow and backward-facing
step flow. Shear flow is another flow condition in which the dispersion of fiber flocs occurs. It
has been shown, however, that extensional flow is more efficient in dispersing fiber flocs. This
was first demonstrated by Kao & Mason (1975) and experimentally verified by Norman et al.
(1977). Studies of floc rupture in extensional flow have also been made by Lee & Brodkey (1987)
and James et al. (2003). Owing to increased computational possibilities, direct simulations of
fiber flocculation have been performed, including the morphological properties of wood fibers.
Initial attempts simulating fiber flocculation were made by Schmid et al. (2000). Building on their
model, Lindström & Uesaka (2009) investigated the effects of fiber aspect ratio, concentration,
and interparticle friction on the stress tensor of the suspension in the steady state and on the
tendency of fiber flocculation to occur.
3.2.2. Effects of fibers on turbulence. Not only does turbulence affect fiber flocculation,
but turbulence is also strongly affected by the presence of the fibers. Lundell et al. (2005) per-
formed visualizations of decaying turbulence with and without fibers. Without pulp, small-scale

Fluidized Plug
Pressure drop
Turbulent flow
(Fluidized flow)
Transition flow
Water Plug flow
Bulk velocity
Figure 7
(Left panel ) Typical pressure-drop variation as a function of flow velocity in a pipe flow of a fiber suspension in the semidilute range.
(Right panel ) Illustrations of the different flow regimes of a pulp suspension in a pipe flow.
structures seemed to decay first, whereas the large-scale structures remained, as expected. How-
ever, with pulp, the small-scale structures remained active throughout the decay process. The
authors suggested that the small-scale structures visible in the flow with fibers do not represent
turbulent fluctuations remaining from the beginning but are a result of fibers or fiber flocs that
create a shear directly surrounding the fiber or floc as it is rotated by the large-scale structures.
Such a mechanism could take energy from larger scales and deposit it at scales that are on the
order of the fiber or fiber floc size.
Bennington & Mmbaga (2001) showed that fibers change the nature of dissipation of the
turbulent energy. In a suspension, the turbulent energy is not only dissipated to heat at small
scales, as it is in pure water, but dissipation can occur also through particle interactions. In a
fiber suspension, they actually found that the main part of the dissipation occurs because of fiber
interaction; however, for a suspension of beads, only a small part of the dissipation resulted from
particle interaction.
Another well-known effect is that the turbulent friction drag (e.g., in pipe flow) decreases
when fibers at low concentrations are added to the flow. The fibers are presumably affecting the
turbulence generation cycle at the wall [see, e.g., Paschkewitz et al. (2004), who performed a direct
numerical simulation of turbulent flow with fibers in a channel].
At higher concentrations, the flow characteristics depend on the flow velocity, and for a fiber
suspension in a pipe at low flow velocities, it behaves like a plug flow. The plug flow is characterized
by an almost uniform velocity profile over the cross section of the pipe, except close to the
wall, where there exists a thin lubricating water layer. As the velocity increases, the shear in the
lubricating layer will be high enough to cause rupturing in the suspension at the individual fiber
level and/or at the floc level. If the velocity increases further, the center plug flow gradually breaks
up and becomes fluidized. These different situations are illustrated in Figure 7. Because of these
effects, the pipe-flow pressure drop will not increase monotonously with the flow rate, but it
actually decreases when the lubricating layer becomes thicker for an intermediate flow rate.
4. FLUID MECHANICS OF SHEET FORMING

In this section we discuss in some detail the sheet-forming process, which is the process in which
fluid mechanics plays a key role in the final quality of the paper. This is the part of the papermaking
process in which the fiber concentration is the lowest (<1%). Although the fiber suspension at

these low concentrations also can show non-Newtonian behavior, considerable understanding can
be achieved without taking this complication into account.
The sheet-forming process takes place in the headbox and starts with a flow distributor designed
to deliver the fiber suspension uniformly across the full width of the headbox (which can be more
than 10 m wide). The diffuser downstream of the flow distributor gives a large pressure drop
and is used to help disperse the fiber flocs. In the contraction part of the headbox, the flow is
accelerated and becomes more homogeneous. The acceleration helps in dispersing the fiber flocs
but also orients the fibers. From the headbox, a two-dimensional (2D) jet exits at a high speed. The
stability and spanwise uniformity of the water jet are important for the resulting homogeneity of
the paper product. The jet lands at one or between two moving wires, and the speed of the wire(s)
is usually close to that of the jet. The fiber orientation can be somewhat controlled by adjusting
the velocity ratio between the suspension jet and the wires. For the double wire (which is the
most common presently), the initial dewatering takes place by passing the wire system through
rolls or by so-called blade forming. After this initial part (e.g., one or two meters) the fibers are
interlocked, and the fiber distribution in the paper is more or less finalized.
4.1. Headbox
The task of the headbox is to transform the pipe flow of the pulp suspension to the paper machine
into a thin spanwise homogeneous suspension jet, which can be delivered onto the wires. The
headbox is fed with the fiber suspension from the flow distributor. Figure 8 shows an example of
a modern headbox consisting of a flow distributor, a diffuser, and a contraction.
4.1.1. Flow distributor. The flow distributor is a manifold usually with a bypass flow, and its task
is to generate a constant pressure prior to the entrance of the fiber suspension to the tube bank. The
manifold is usually designed with a gradually decreasing circular or quadratic cross section. Where
the pipe flow enters the manifold, the cross-stream area is large, and in the small area end there is
Pulp suspension
to headbox
Headbox
old
nif
Ma ank
eb
Tub
n
ctio
ntra
Co
10 mm
Overflow Headbox jet
30
m s –1
Fibe =1
r flo 0m ,80
c pa 5–1 0m
min –
ssin
g th 1
rou
gh
con
trac
tion
Figure 8
A modern headbox design.

an overflow to make it possible to control the pressure. The uniformity of the headbox jet (i.e., the
variation of the streamwise velocity across the machine width) is of vital importance if a spanwise
uniform product should be obtained. Historically, the manifold shape has been designed to give a
constant static pressure across the machine width. Today it is possible to combine computational
fluid dynamics analysis with optimization methods to find an optimal design (Hämäläinen et al.
2000). Although such methods are used in many engineering applications, they require valid
constitutive models for the fiber suspension.
4.1.2. Headbox diffuser. The fiber suspension is usually turned from the flow distributor into the
contraction through a tube bank consisting of a sequence of round tubes with increasing diameter
and abrupt increases of the area from one tube to the other. The tube bank serves two main
purposes. At the upstream side, the open area is low, around 10%, giving a high-pressure drop
necessary to produce a homogeneous cross-machine velocity distribution. This is usually followed
by a diffuser with pipes of discrete steps in diameter to increase the open area. The typical length
of each tube is approximately 5–15 times the tube diameter. The steps in diameter produce high
levels of turbulence and increase the dispersion of fiber flocs. At the downstream side, the open
area is maximized to reduce the wakes that form behind the separating walls.
To optimize the design of the tube bank, Hauptmann et al. (1990a,b) analyzed the effect of
turbulence generation by backward-facing steps and wakes based on experiments using hot-film
anemometry in water. They studied the generation and decay of turbulence and the turbulent
length scales. Comparisons of the results were made with visualizations of the flow behavior of the
headbox jet. Shariati et al. (2000) numerically solved the flow through a manifold, tube bank, and
contraction. The results clearly showed the effect of recirculation, generated by flow separation
in the tube bank. As the complete flow was solved, the effect of nonuniform flow conditions in the
manifold channel showed a strong influence on the behavior of the flow inside the tubes depending
on their individual position.
4.1.3. Flow through the contraction: calculations and measurements of fiber orientation.
The contraction part of the headbox typically has straight walls and a contraction ratio in the
range 5–15. The acceleration of the flow in the contraction affects both the fiber flocs and the
fiber orientation of the final paper sheet. Recently, this flow case has been studied extensively,
both experimentally and numerically. Among others, Nordström & Norman (1994) showed that
the contraction ratio affects several important properties of the final paper. Mastering the fluid
mechanical sources of these effects would enable not only the optimization of the process, but also
a more exact tailoring of the product.
The present modeling and calculation abilities do not allow full dynamical simulations of this
process, although ambitious attempts demonstrating the potential of such approaches have been
made (e.g., Lindström & Uesaka 2008). Furthermore, experimental studies are, if not impossible,
quite difficult due to the fibers. The main interest therefore is fiber orientation behavior at low
concentrations.
The fiber orientation is typically modeled by a Fokker-Planck equation following Advani &
Tucker (1987), which for the case of slender fibers with circular cross section following the fluid
and 2D flow (i.e., the flow is homogeneous in the spanwise x3 direction) reads

∂ ∂ 1 ∂ 1 ∂
U1 +U2 = Dr − sin θ φ̇
∂ x1 ∂ x2 sin θ ∂φ sin θ ∂φ

1 ∂ ∂
+ Dr sin θ − sin θ θ̇ , (2)
sin θ ∂θ ∂θ

CR = 16.7 CR = 25
a b Ullmar 1998 c
Hyensjö &
2.0 Dahlkild 2008 2.0
x2 θ
1.5 1.5
x1
φ Ψ Ψ
1.0 1.0
x3
0.5 0.5
0 0
−1 0 1 −1 0 1
φ φ
Figure 9
(a) Definition of coordinate system and orientation angles for fibers. (b,c) Calculated (line; Hyensjö & Dahlkild 2008) and measured
(dots; Ullmar 1998) orientation distribution at the end of the contraction for contraction ratio (CR) 16.7 (b) and 25 (c).
where (θ , φ) is the probability density as a function of the orientation angles defined in
Figure 9a. Equation 2 can be viewed as a conservation equation for , which is modified by
diffusion through rotational diffusivity Dr and by fiber rotation through the terms including θ̇
and φ̇. To solve Equation 2, one needs (a) an initial (upstream) boundary condition, (b) boundary
conditions at the headbox walls, (c) a mean flow U1 , U2 , (d ) estimates of the rotational diffusivity
Dr , and (e) relations for the orientational velocities θ̇ and φ̇, which typically are functions of the
local mean velocity gradients.
The rotational velocities are usually taken as those of an inertialess ellipsoid in linearly varying
Stokes flow obtained analytically by Jeffery (1922). It is worth mentioning, however, that the
motion of nonspherical particles, taking fluid and particle inertia into account, is an active field
of research in which both series expansions (Subramanian & Koch 2005, 2006) and various direct
simulation methods (Altenbach et al. 2007, Ding & Aidun 2000, Lundell & Carlsson 2010, Qi
& Luo 2003) considering fluid and/or particle inertia have been used. It is clear that the inertial
effects in a paper machine can be substantial, and a complete description of the rotational velocities
is not available.
Nevertheless, a mean flow calculated by computational fluid dynamics with some turbulence
model can be used together with the rotational velocities of Jeffery (1922) and some closure for the
rotational diffusivity. Estimates of Dr have been based on measured fiber orientation distributions
by, e.g., Nordström & Norman (1994), Asplund & Norman (2004), and Parsheh et al. (2005).
Some recent examples range from 1D approximations to full 3D rotation in 2D velocity fields
(Hyensjö & Dahlkild 2008, Hyensjö et al. 2007, Olson et al. 2004, Parsheh et al. 2006). An example
result from Hyensjö & Dahlkild (2008) is shown in Figure 9b,c for flow through contractions
with contraction ratios of 16.7 and 25, respectively. The strain in the accelerating flow gives the
fibers a tendency to align in the flow direction. For a suspension with an initially homogeneous
angular distribution, the flow through the contraction makes it become focused around φ = 0,
increasingly so for increasing contraction ratios. In these calculations, a constant Dr (different

values for the two contraction ratios) was used throughout the contraction and was adjusted so
that the final orientation distributions match the experimental values.
This approach gives a good match with experimental data even though the physical modeling
is far from complete. It is thus clear that measurements of rotational diffusivities, together with the
turbulence of the flow, are necessary for further development of the method. Such measurements
can be made either in Lagrangian numerical simulations or experiments.
Another approach is to derive dispersion coefficients from studies of fiber motion in homoge-
nous turbulent flow. This was done by Olson & Kerekes (1998), who showed that the dispersion
coefficients decrease as the ratio of fiber length to the Lagrangian integral length scale of the
turbulence increases. Actually, this ratio must also be taken into account when calculating the
rotational velocities of the fibers, as shown by Shin & Koch (2005), based on data from fiber
tracking in turbulent flow fields.
The effect of fiber concentration on dispersion coefficients has been studied in a combined
experimental/numerical study by Krochak et al. (2008). Dispersion relation closures in the litera-
ture that account for hydrodynamic interactions (Folgar & Tucker 1984, Koch 1995) gave decent
results up to the concentration at which mechanical interaction between the fibers becomes sig-
nificant (crowding numbers N below 15).
An additional aspect is the wall boundary condition of the fiber orientation. Measurements
by Carlsson (2009) show that, depending on the fiber aspect ratio and level of shear at the wall,
fibers tend to orient themselves either in the flow direction or perpendicular to this direction at a
wall. Furthermore, these tendencies can have an impact on the fiber concentration in the direct
vicinity of the wall. So far, the results are incomplete and inconclusive regarding the details of
these phenomena.
4.2. The Headbox Jet

As stated above, the uniformity of the headbox jet is vital for uniform paper properties. Variations in
the grammage, fiber flocculation level, and local fiber orientation can be coupled to the conditions
of the flow and the fluid in the headbox jet. Those conditions are affected by upstream flow
conditions as well as the dynamics of the jet itself.
A well-known problem that can arise in the final product is related to streaky structures through
an inhomogeneous fiber distribution. This is believed to originate from streamwise-oriented vor-
tices in the headbox jet. Söderberg & Alfredsson (2000) identified four possible mechanisms caus-
ing streaks in the headbox jet (Figure 10). Figure 10a illustrates how vorticity inside a converging
channel will be stretched as it moves downstream. This is usually referred to as vortex stretching,
and the occurrence of streaks inside the jet will depend on the turbulence level in the nozzle.
Concave curvature of the nozzle walls (Figure 10b) may be another possible source of streak
structures as the curvature may result in a centrifugal instability that gives rise to so-called Görtler
vortices in the boundary layer, a mechanism already proposed by Parker (1972). A third possible
way to obtain streaks is through so-called transient disturbance growth inside the boundary layer
close to the channel walls (Figure 10c). Such disturbances may be triggered by the presence of
turbulence outside the boundary layer and may cause the growth of streaky structures well below
the critical Reynolds number (see, e.g., Matsubara & Alfredsson 2001).
The three mechanisms mentioned above all originate from inside the nozzle. However, there is
a possibility that the plane liquid jet emanating from a contracting channel (Figure 10d ) develops a
2D wave instability as the jet leaves the nozzle. Söderberg & Alfredsson (1997) showed that strong
streamwise aligned streaky structures could be created inside the jet coupled to the breakdown of
the 2D surface waves. Although that study was performed under idealized conditions, Söderberg

a b
c d
Figure 10
Different types of streak creation: (a) vortex stretching, (b) centrifugal instability, (c) transient growth in the boundary layer, and
(d ) instability and breakdown of the waves.
(1999) showed that this type of wave and the related breakup can be found in the flow of a pulp
suspension in a real headbox jet.
The stability of the jet can be studied through linear stability analysis, but the analysis becomes
complex as it includes the liquid jet and the surrounding air, as well as the free-surface conditions
including surface tension. In addition, the jet flow develops in the downstream direction, and it
relaxes from a boundary layer–type profile toward a flat profile, giving rise to inflectional points
in the velocity distribution (Figure 11).
To study the stability, one may choose to make a local analysis, i.e., assuming that the flow
is locally parallel and homogenous in the downstream x direction ( y is the cross-flow direction).
The variables are made nondimensional according to
x = x ∗ /a, y = y ∗ /a, u = u ∗ /U m , v = v ∗ /U m , and p = p ∗ /(ρU m2 ),
where the superscript ∗ indicates an unscaled variable. The velocity fields in the relaxational liquid
jet (subscript l ) and the surrounding gas (subscript g) are divided into basic and disturbance flow
z x
Figure 11
Relaxation of a jet flow exiting a channel nozzle.

fields:

ul,g = {U l,g (y; X ) + ul,g (x, y, t), vl,g (x, y, t)}, (3a)

p l,g = Pl,g (X ) + p l,g (x, y, t). (3b)
A slow coordinate X = x, where = 1/Rel 1, has been introduced, which governs the variation
of the basic flow. Both the basic flow and the disturbance (denoted by a prime) are assumed to be
2D, i.e., independent of z.
A normal-mode ansatz is made, in which the disturbances are assumed to have the form
{u , v , p } = {û(y; X ), v̂(y; X ), p̂(y; X )}e i(αx−ωt) . (4)
Here α and ω are the wave number and frequency, respectively, and one of them can be considered
complex, depending on whether one chooses the spatial (α = αr + iαi ) or temporal (ω = ωr + iωi )
problem. The decomposition given in Equation 3 and the ansatz given in Equation 4 are inserted
into the Navier-Stokes equations, and linearization is performed that results in the well-known
Orr-Sommerfeld equation,
−1
(iαU − iω)(D2 − α 2 )v̂l,g − iα D2 U v̂l,g = Rel,g (D4 − 2α 2 D2 + α 4 )v̂l,g , (5)
where D = d /d y. This means that the flow locally, i.e., at the position X, is considered to be
parallel. Note that there is one equation for the liquid jet and one for the gas flow outside, which
are coupled through the boundary conditions at the interface. The boundary conditions are given
by Söderberg & Alfredsson (1998) and involve Reg , We, ρ̃, and μ̃ (i.e., the gas Reynolds number,
Weber number, density ratio, and viscosity ratio, respectively), defined as
ρg U m a ρU m2 a ρg μg
Reg = , We = , ρ̃ = , and μ̃ = .
μg γ ρl μl
To have a basic flow field in the surrounding gas without making assumptions about the geometry
and without having to solve the flow of the gas phase, the solution to Stokes’ first problem, the
infinite starting plate (see, e.g., Schlichting 1999), can be used as an approximation. The time
dependence in the solution to the Stokes problem can be transformed into a downstream distance
with Um . As far-field boundary conditions in the gas, the disturbance amplitudes, û and v̂, are set
to zero. The boundary conditions, the mean flow distributions together with the Orr-Sommerfeld
equation, constitute an eigenvalue problem that gives eigenfunctions v̂(y) only if the parameters
satisfy the dispersion relation
D(α, ω; Re, We, X ) = 0. (6)
Söderberg & Alfredsson (1998) observed surface waves in their experiment as shown in
Figure 12a. At Re = 700, no waves can be seen. As the velocity (or equivalently Re) increases,
waves start to appear in the image. At Re = 920, the waves break down starting at x ≈ 40
and spread in the streamwise direction; i.e., both waves and the breakdown can be seen down to
x ≈ 100. As the velocity (Re) is increased even more, the breakdown moves upstream.
The stability problem defined by the dispersion relation given in Equation 6 actually gives
rise to five different modes that can become unstable, three of sinuous character and two of
varicose character. The modes are to various extents associated with the relaxational inflectional
velocity profile, effects of surface tension, and aerodynamic drag on the jet. Figure 12b shows
the eigenvalue spectra for the spatial growth problem. In the eigenvalue spectrum, four of the
modes are unstable (αi < 0), and the fifth mode is a sinuous mode (eI ) that has a positive αi . Also,
Figure 12c shows the amplitude of the eigenfunctions for these modes in both liquid and gas. As
can be seen, all eigenfunctions decay far away from the free surface.

a Re = 700 Re = 830 Re = 920 Re = 1,050 b

Odd
20 Even
0.4
60
0.2
x αi
100
0
140 I III II
180 –0.2
0 1
αr
c
eI eII eIII oI oII

6
y
4
0
0 1 0 1 0 1 0 1 0 1
v v v v v
Figure 12
(a) Shadowgraph visualization of the breakup of a plane liquid jet at ( from left to right) Re = 700, 830, 920,
and 1,050. At the top of the image the nozzle is visible, and the downstream position x is shown on the
vertical axis. The four images are taken from the central part of a plane jet and represent an increase in Re
from left to right. (b) Eigenvalue spectra with spatial formulation at ω = 1, Re = 1,000, We = 25.5, and
x/Re = 0.07 of sinuous or even modes (red circles) and varicose or odd modes (blue circles) with the
classification of the type shown (I, II, and III). (c) Absolute value of the associated eigenfunctions, |v̂|. The
sinuous (even) modes are denoted eI , eII , and eIII and the varicose (odd) modes are denoted oI and oII , with I,
II, and III representing their types. The gas-liquid interface is at y = 1.
The mechanism behind the breakdown of the waves was later identified by Söderberg (2003)
as an absolute instability in the jet coupled to the velocity profile relaxation. Tammisola (2009)
has extended the local analysis to a global analysis of the flow instabilities.
4.3. Initial Dewatering

As the jet hits the forming fabric, the dewatering begins. A modern roll-and-blade forming section
is shown in Figure 13a. Typical dimensions are R ≈ 800 mm, and the thickness varies with the
type of paper produced. The initial jet thickness is typically in the range of 10 mm (newsprint) to
30 mm (board). The thickness after the roll depends on the process conditions and can be as small
as 1 mm or less. After the initial dewatering, the web is transferred to the press section where
stronger mechanical pressure is used to press out as much water as possible.

12
a T T b
10
Pressure (kPa)
8 T/R = 7.3 kPa
0
−400 −300 −200 −100 0
(mm) from wire/roll separation
0.5
h3 c
T R 0.4
0.3
h2
α
0.2
h1 u2
Holm & Söderberg
0.1 (2005)
u1
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
T
We0.5
Figure 13
(a) Sketch of a roll-blade former. The suspension jet (blue) is sprayed in between two wires ( green dotted lines), which are first wrapped
around the roll. After the jet suspension has passed the roll, blades are pressed toward the wires (only the first three are shown here).
The suspension thickness is highly exaggerated in the figure. (b) Pressure measurement made between the forming wires in a full-scale
paper machine. The distance √ is along the wires from the nominal point of wire/roll separation. (c) Measured nondimensional wave
number α as a function of We compared with the theoretical prediction of Holm & Söderberg (2005).
4.3.1. Dewatering on the roll. As the jet is sprayed in between the forming fabrics and they are
wrapped around the roll, the pressure of the suspension is increased from the ambient pressure pamb
to p amb + T/R, where T is the wire tension (N m−1 ) and R is the roll radius. This pressure increase
is used to dewater the suspension, as the pressurized water is pressed through the wires, both into
the perforated roll and out to the ambient room. Furthermore, the pressure increase is coupled to
a velocity decrease (and actually an initial thickness increase, although the rapid dewatering makes
this increase difficult to detect) via Bernoulli and mass conservation (for a definition of the various
variables, see Figure 13a), so that

2T ρR
u2 = u 21 − h 2 = u1 h 1 . (7)
ρR ρ Ru 21 − 2T
It is necessary that u2 is real, i.e., that u 21 > 2T/ρ R, for the suspension to be able to penetrate
in between the wires. If the jet velocity is too low, or the wire tension too high, the system will
collapse, and the jet will bounce back instead of continuing forward between the wires.
The jet impingement has been studied and modeled, taking free surfaces and wire permeability
into account (Audenis & Dahlkild 2001, Dalpke et al. 2004). Once the suspension has penetrated
in between the two wires, the overpressure between the wires is initially balanced mainly by the

flow resistance as water is drained through the growing fiber networks and wires. Further on,
when the network fills the gap between the wires, the fiber network might carry load as well. The
key process parameter is the drainage rate, which is a function of the pressure in between the wires
and the permeability and thickness of the fiber networks together with the drainage resistance of
the wires. Such measurements were made by Martinez (1998), and Zahrai et al. (1998) analyzed
how the thickness of the fiber network develops during drainage.
At first sight, it could be expected that the pressure would be constant (equal to T/R) as long
as the wire is wrapped around the roll. However, measurements, e.g., by Holm et al. (2005) in
a model of a roll-former system and Holmqvist & Söderberg (2010) in a pilot paper machine,
show that the pressure is oscillating (a stationary variation in the laboratory frame of reference).
An example of the pressure profile from the pilot machine is shown in Figure 13b. The pressure
profile shows clear oscillations from −150 mm to 0 mm. Holm et al. (2005) showed that these
oscillations are related to variations in the distance between the roll and the outer wire. Holmqvist
& Söderberg (2010) showed that Holm & Söderberg’s (2005) analysis gives
2π h 3 √
α= = We, (8)
λ
where α is a nondimensional wave number, h3 is the suspension thickness close to the wire/roll
separation (see Figure 13a), λ is the wavelength of the last oscillation in pressure profiles like
the one in Figure 13b, and the Weber number is defined as We = ρU 2 h3 /T, where ρ is the
suspension density and U is the suspension velocity. Note that for the formation of this wave, the
fabric tension acts as surface tension. Figure 13c shows that this analysis explains the variations
in pulse wavelength measured by Holmqvist & Söderberg (2010).
4.3.2. Blade dewatering. Once the fabrics have left the roll, the blade-forming unit follows. In
Figure 13a, the first three blades of such a blade-forming unit are shown. As explained by Norman
(1979), the blades force a deflection of the fabrics. Just as on the roll, such a deflection together
with the fabric tension gives a pressure pulse of high amplitude. This pressure pulse has two effects:
It drives dewatering and affects the formation of the resulting paper sheet. In particular, pressure
pulses over the blades can tear flocs apart and improve formation. This is achieved by the rapid
acceleration/deceleration associated with the pressure pulse. The longitudinal velocity variations
result in shear against the fabrics, which move with a constant speed. Note that these velocity
variations might also destroy the paper by tearing the network apart.
Pressure pulses of high amplitude were measured by, e.g., Zhao & Kerekes (1995) and Zahrai
et al. (1997), and they have been modeled with increasing complexity by Green & Kerekes (1996),
Zahrai et al. (1998), Roshanzamir et al. (1998), Green (2000), and Holmqvist & Dahlkild (2006),
among others. The pressure pulse profile and amplitude depend on the blade force and shape, wire
tension, and drainage resistance. The coupling between settings of the blade section, the hydro-
dynamics of the blade-fabric-suspension system, and the resulting process and paper properties is
not fully understood. However, pressure pulses can be predicted fairly well, at least the qualitative
effects of parameter changes. The effect of the pressure pulses on web consolidation is still not
fully understood, however.
One of the key factors is the drainage resistance that is connected to the permeability of the
fiber web building up on the web. Several attempts to model this complex, inhomogeneous porous
media have been made. A key feature of this process is the shear, and this feature was modeled for
flocculated suspensions by Holmqvist & Dahlkild (2008) with consolidation models originating
from soil mechanics. It was shown that shear increases the drainage by decreasing the load-bearing
capacity of the network (and thereby increasing the pressure of the liquid phase). The complexity

of the drainage-resistance problem in the papermaking process is noted by Wildfong et al.’s (2000)
experimental results indicating that plugging of the fabric openings by fine material can contribute
more to drainage resistance than the permeability of the compressed fiber web during drainage.
5. FUTURE DEVELOPMENTS AND POSSIBILITIES

There are several interesting trends in the paper industry for which fluid mechanics both from
engineering and scientific points of view may play key roles. The main future challenges for the
papermaking industry are to reduce energy and raw material consumption. This is to some extent
a coupled problem as the need for reduced-fiber raw material for producing a product with a given
functionality normally also gives a reduction in energy consumption.
Apart from direct efficiency improvements of individual unit processes, there is a definite
possibility for improving the utilization of the fiber raw material. This can be achieved by modifying
and selecting fiber properties together with improving the material (paper) structure. Future
developments of techniques for fiber modification and selection will be facilitated by an increased
understanding of the fluid physics governing the refining process as well as mechanisms that can
be used for fiber fractionation.
Regarding improvements of the material structure of paper, there are two main routes. One
is to reduce the variability of the structure, which to a large extent is a result of flow variations in
the sheet-forming process. Specifically, reducing the streaks and large-scale fiber flocculation will
result in a paper sheet in which the weakest spots are relatively stronger.
Another possibility is to design the sheet as a layered structure in which the material (fiber)
composition can be varied throughout the sheet thickness. This is not a new concept in the paper-
making industry, and several publications have appeared in the scientific and technical literature
since the 1970s. This can be achieved using a concept called stratified forming. In contrast to
multi-ply forming, which is standard for high-basis-weight products such as paperboard, stratified
forming produces a layered structure with one single headbox. In the headbox, so-called vanes are
used to separate layers with different fibers. Recent developments of concepts for stratified form-
ing have all been based on flow control as a tool for reducing layer mixing in the sheet-forming
process. By utilizing the base-bleed effect, the mixing can be controlled, which results in improved
layer purity (Söderberg & Lucisano 2005).
Fluid mechanical challenges that arise from the papermaking process can be divided into three
categories: (a) the modeling of fiber suspensions (with correct orientation and concentration dis-
tributions) as such, (b) experimental methods providing the results necessary for validation of these
models, and (c) deeper understanding of the coupling between rheology (of a given pulp) and its sus-
pension characterizations (e.g., fiber lengths, morphology, fiber-fiber friction, and concentration).
Of course, these three areas provide input to each other. On a more detailed level, basic subjects
that need further study are the motion of fibers in flow at elevated Reynolds numbers, models
for the hydrodynamic interaction of fibers, and resulting redistributions of fibers in space. Increas-
ing the complexity somewhat, fiber motion in turbulent flow is an important link, closely related
to the coupling between fibers and turbulence.
Even though these issues might be challenging enough, understanding the flocculation and
consolidation of fibers, together with the water flow and mechanical deformation of the structure,
is a truly intriguing physical problem, for which full knowledge of fiber-fiber interaction (hydro-
dynamical, mechanical, and possibly chemical) is necessary. In fact, the behavior of a flocculated
suspension is a field of study in its own right.
All these issues provide numerical as well as experimental challenges. Presently, comparison
between experiments and numerics is mostly made based on a final result, such as the resulting fiber

orientation. However, with increased understanding of the physics and improved measurement
and computer capacity, it is now possible to compare the actual physical processes involved. It can
be assumed that such comparisons will lead to a relatively fast development of the understanding of
idealized fiber suspension flows, but also, within a longer perspective, flows with complex cellulose
fibers.
SUMMARY POINTS
1. Paper is one of the most important materials worldwide, and its most common applica-
tions include printing, hygiene, and packaging applications.
2. In the paper machine, a suspension of cellulose fibers (stock) with typically 0.5%–1%
solid content is transformed into an (almost) dry paper sheet.
3. Mechanical and other properties of the final paper depend on the mass and orientation
distributions of the fibers and other material in the sheet.

4. The mass and orientation distributions are determined by the interaction between the
fibers and the flow during the forming process.
5. During the forming process, a number of fundamental fluid mechanical phenomena are
at hand, such as fiber suspension flows, turbulent mixing, shear flows, interfacial flows,
and flow in porous media.
6. The knowledge of the fluid physics in the papermaking process is considerable. Never-
theless, it is far from complete.
7. There is a distinct potential for process as well as product innovations based on improved
fluid mechanical understanding of fiber suspension flows.
DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.
ACKNOWLEDGMENTS
We would like to acknowledge the inspiration of two of our colleagues, Profs. Bo Norman and Fritz
Bark. Both were instrumental in forging the cooperation between the applied area of papermaking
and fundamental fluid mechanics through the FaxénLaboratory that was operational from 1995 to
2005. The photographs in Figure 5 were kindly provided by Joanna Hornatowska at Innventia.
We would also like to acknowledge our former and present graduate students working in this area;
some of them can be recognized from their theses in the Literature Cited. Our research within
this area is or has been sponsored by the Swedish Energy Agency, the Swedish Research Council,
the Linné FLOW Center, the Knut and Alice Wallenberg Foundation, the Wallenberg Wood
Science Center, and the EU.
LITERATURE CITED
Advani S, Tucker C. 1987. The use of tensors to describe and predict fiber orientation in short fiber composites.
J. Rheol. 31:751–84

Altenbach H, Naumenko K, Pylypenko S, Renner B. 2007. Influence of rotary inertia on the fiber dynamics
in homogenous creeping flows. Z. Angew. Math. Mech. 87:81–93
Asplund G, Norman B. 2004. Fibre orientation anisotropy profile over the thickness of a headbox jet. J. Pulp
Paper Sci. 30:217–21
Audenis G, Dahlkild AA. 2001. Impingement of an inviscid plane liquid jet on a porous wall: a fixed domain
method. Nord. Pulp Paper Res. J. 16:274–83
Bennington CPJ, Mmbaga JP. 2001. Liquid phase turbulence in pulp fibre suspensions. In Science of Paper-
making, Trans. 12th Fund. Res. Symp., ed. CF Baker, vol. 1, pp. 255–86. London: Mech. Eng. Publ.
Carlsson A. 2009. Near wall fibre orientation in flowing suspensions. PhD thesis. Dep. Mech., KTH, Stockholm.
165 pp.
Dalpke B, Kerekes RJ, Green SI. 2004. Modelling jet impingement and the initial drainage zone in roll
forming. J. Pulp Paper Sci. 30:65–70
Ding EJ, Aidun CK. 2000. The dynamics and scaling law for particles suspended in shear flow with inertia.
J. Fluid Mech. 423:317–44
Folgar F, Tucker CL. 1984. Orientation behavior of fibres in concentrated suspensions. J. Reinf. Plast. Compos.
3:98–119
Green S. 2000. Modeling suction shoes in twin-wire blade forming: results. J. Pulp Paper Sci. 26:53–58
Green SI, Kerekes RJ. 1996. Numerical analysis of pressure pulses induced by blades in gap formers. In TAPPI
Eng. Conf., vol. 1, pp. 185–92. Atlanta: TAPPI
Hämäläinen J, Mäkinen RAE, Tarvainen P. 2000. Optimal design of paper machine headboxes. Int. J. Numer.
Methods Fluids 34:685–700
Hauptmann E, Vyse R, Mardon J. 1990a. The wake effect as applied to modern hydraulic headboxes. I:
Introduction and practical illustrations of the wake effect; elementary hydrodynamics. Pulp Paper Canada
91(9):88–97
Hauptmann E, Vyse R, Mardon J. 1990b. The wake effect as applied to modern hydraulic headboxes. II:
Design and operation of perforated roll headboxes, hydraulic or filled headboxes. Pulp Paper Canada
91(10):47–57
Holm R, Söderberg D. 2005. A theoretical analysis of the flow stability in roll forming of paper. Nord. Pulp
Paper Res. J. 20:212–16
Holm R, Söderberg D, Norman B. 2005. Experimental studies on dewatering during roll forming of paper.
Nord. Pulp Paper Res. J. 20:205–11
Holmqvist C, Dahlkild A. 2006. A flexible approach for modelling flow in multi-component blade formers.
Nord. Pulp Paper Res. J. 21:73–81
Holmqvist C, Dahlkild A. 2008. Consolidation of sheared, strongly flocculated suspensions. AIChE J. 54:924–
39
Holmqvist C, Söderberg D. 2010. Drainage pressure oscillations during roll forming. Proc. PaperCon 2010,
Atlanta, PAPERCD-10. Atlanta: TAPPI
Hyensjö M, Dahlkild A. 2008. Study of the rotational diffusivity coefficient of fibres in planar contracting
flows with varying turbulence levels. Int. J. Multiphase Flow 34:894–903
Hyensjö M, Dahlkild A, Krochak P, Olson J, Hämäläinen J. 2007. Modelling the effect of shear flow on fibre
orientation anisotropy in a planar contraction. Nord. Pulp Paper Res. J. 22:376–82
James DF, Yogachandran N, Loewen MR, Liu H, Davis AMJ. 2003. Floc rupture in extensional flow. J. Pulp
Jeffery GB. 1922. The motion of ellipsoidal particles immersed in a viscous fluid. Proc. R. Soc. Lond. 102:161–79
Kao S, Mason S. 1975. Dispersion of particles by shear. Nature 253:619–21
Kerekes RJ. 2006. Rheology of fibre suspensions in papermaking: an overview of recent research. Nord. Pulp
Paper Res. J. 21:598–612
Kiviranta A, Dodson C. 1995. Evaluating Fourdrinier formation performance. J. Pulp Paper Sci. 21:379–84
Koch DL. 1995. A model for orientational diffusion in fibre suspensions. Phys. Fluids 7:2086–88
Krochak PJ, Olson JA, Martinez DM. 2008. The orientation of semidilute rigid fiber suspensions in a linearly
contracting channel. Phys. Fluids 20:073303
Lee CW, Brodkey RS. 1987. A visual study of pulp floc dispersion mechanisms. AIChE J. 33:297–302

Lindström SB, Uesaka T. 2008. Particle-level simulation of forming of the fiber network in papermaking.
J. Eng. Sci. 46:858–76
Lindström SB, Uesaka T. 2009. A numerical investigation of the rheology of sheared fiber suspensions. Phys.
Fluids 21:083301
Lundell F, Carlsson A. 2010. Heavy ellipsoids in creeping shear flow: transitions of the particle rotation rate
and orbit shape. Phys. Rev. E 81:016323
Lundell F, Söderberg D, Storey S, Holm R. 2005. The effect of fibres on laminar-turbulent transition and
scales in turbulent decay. In Advances in Paper Science and Technology, Trans. 13th Fund. Res. Symp., ed. SJ
I’Anson, vol. 3, pp. 19–34. Bury Lancashire, UK: Pulp Paper Fundam. Res. Soc.
Martinez DM. 1998. Characterizing the dewatering rate in roll gap formers. J. Pulp Paper Sci. 24:7–13
Mason SG, Manley RSJ. 1957. Particle motion in sheared suspensions: orientations and interaction of rigid
rods. Proc. R. Soc. Lond. A 238:117–31
Matsubara M, Alfredsson PH. 2001. Disturbance growth in boundary layers subjected to free stream turbu-
lence. J. Fluid Mech. 430:149–68
Nordström B, Norman B. 1994. Influence of sheet anisotropy, formation, z-toughness and tensile stiffness of
reduced feed area to a headbox nozzle. Nordic Pulp Paper Res. J. 1:53–59
Norman B. 1979. Principles of twin-wire forming. Sven. Papp. 82:330–36

Norman B, Moller K, Ek R, Duffy G. 1977. Hydrodynamics of papermaking fibres in water suspensions. In
BPBIF 6th Fund. Res. Symp.: Fibre-Water Interaction in Papermaking, ed. H Corte, pp. 195–250. London:
Br. Paper Board Ind. Fed., Tech. Div.
Olson JA, Frigaard I, Chan C, Hämäläinen JP. 2004. Modeling a turbulent fibre suspension in a planar
contraction: the one-dimensional headbox. Int. J. Multiphase Flow 30:51–66
Olson JA, Kerekes RJ. 1998. The motion of fibres in turbulent flow. J. Fluid Mech. 377:47–64
Parker JD. 1972. The Sheet Forming Process. Atlanta: TAPPI
Parsheh M, Brown M, Aidun C. 2005. On the orientation of stiff fibres suspended in turbulent flow in a planar
contraction. J. Fluid Mech. 545:245–69
Parsheh M, Brown M, Aidun C. 2006. Investigation of closure approximations for fiber orientation distribution
in contracting turbulent flow. J. Non-Newton. Fluid Mech. 136:38–49
Paschkewitz JS, Dubief Y, Dimitropoulos CD, Shaqfeh ESG, Moin P. 2004. Numerical simulation of drag
reduction using rigid fibres. J. Fluid Mech. 518:281–317
Qi D, Luo LS. 2003. Rotational and orientational behaviour of three-dimensional spheroidal particles in
Couette flows. J. Fluid Mech. 477:201–13
Roshanzamir A, Green S, Kerekes R. 1998. Two-dimensional simulation of pressure pulses in blade gap
formers. J. Pulp Paper Sci. 24:364–68
Schlichting H. 1999. Boundary Layer Theory. New York: McGraw-Hill
Schmid CF, Switzer LH, Klingenberg D. 2000. Simulation of fibre flocculation: effect of fiber properties and
interfiber friction. J. Rheol. 44:781–809
Shariati M, Bibeau E, Salcudean M, Gartshore I. 2000. Numerical and experimental models of flow in the
converging section of a headbox. In TAPPI Papermak. Conf. Trade Fair, vol. 2, pp. 685–93. Atlanta: TAPPI
Shin M, Koch D. 2005. Rotational and translational dispersion of fibres in isotropic turbulent flows. J. Fluid
Mech. 540:143–73
Söderberg LD. 1999. A comparison between the flow from a paper machine headbox and a low Reynolds
number water jet. In TAPPI Eng. Conf, vol. 3, pp. 1155–72. Atlanta: TAPPI
Söderberg LD. 2003. Absolute and convective instability of a relaxational plane liquid jet. J. Fluid Mech.
493:89–119
Söderberg LD, Alfredsson PH. 1997. Experiments concerning the creation of streaky structures inside a plane
water jet. In Forming Bonds for Better Papermaking, Proc. 1997 Eng. Papermak. Conf., TAPPI Eng. Conf,
pp. 1205–22. Atlanta: TAPPI
Söderberg LD, Alfredsson PH. 1998. Experimental and theoretical stability investigations of plane liquid jets.
Eur. J. Mech. B Fluids 17:689–737
Söderberg LD, Alfredsson PH. 2000. Experiments concerning the origin of streaky structures inside a plane
water jet. J. Pulp Paper Sci. 26:395–400

Söderberg LD, Lucisano M. 2005. Reduction of layer mixing through hydrodynamic control. In Advances
in Papermaking, 13th Fund. Res. Symp., ed. SJ I’Anson, pp. 83–105. Bury Lancashire, UK: Pulp Paper
Fundam. Res. Soc.
Soszynski RM, Kerekes RJ. 1988. Elastic interlocking of nylon fibres suspended in liquid, part 2: process of
interlocking. Nord. Pulp Paper Res. J. 3:180–84
Steen M. 1991. Modeling fibre flocculation in turbulent flows: a numerical study. TAPPI J. 74:175–81
Subramanian G, Koch DL. 2005. Inertial effects on fibre motion in simple shear flow. J. Fluid Mech. 535:383–
414
Subramanian G, Koch DL. 2006. Inertial effects on the orientation of nearly spherical particles in simple shear
flow. J. Fluid Mech. 557:257–96
Tammisola O. 2009. Linear stability of plane wakes and liquid jets: global and local approach. Licentiate thesis,
TRITA-MEK 2009:04. R. Inst. Technol., Stockholm. 157 pp.
Taylor G. 1957. Fluid dynamics in a papermaking machine. Proc. R. Soc. Lond. A 242:1–15
Ullmar M. 1998. On fibre alignment mechanism in a headbox nozzle. Licentiate thesis, R. Inst. Technol.,
Stockholm. 73 pp.
Wildfong V, Shands JA, Bousfield DW, Genco JM. 2000. Drainage during roll forming: model validation
using pilot papermachine data. In TAPPI Eng. Conf. CD-ROM. Atlanta: TAPPI
Zahrai S, Bark F, Norman B. 1997. An analysis of blade dewatering in a twin-wire paper machine. J. Pulp
Zahrai S, Martinez MD, Dahlkild AA. 1998. Estimating the thickness of the web during twin-wire forming.
J. Pulp Paper Sci. 24:67–72
Zhao RH, Kerekes RJ. 1995. Pressure distribution between forming fabrics in blade gap formers: thin blades.
J. Pulp Paper Sci. 21:97–103

FL43-FrontMater ARI 15 November 2010 11:55
Annual Review of
Contents Fluid Mechanics
Volume 43, 2011
Experimental Studies of Transition to Turbulence in a Pipe

T. Mullin p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1
Fish Swimming and Bird/Insect Flight
Theodore Yaotsu Wu p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p25

Wave Turbulence
Alan C. Newell and Benno Rumpf p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p59
Transition and Stability of High-Speed Boundary Layers
Alexander Fedorov p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p79
Fluctuations and Instability in Sedimentation
Élisabeth Guazzelli and John Hinch p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p97
Shock-Bubble Interactions
Devesh Ranjan, Jason Oakley, and Riccardo Bonazza p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 117
Fluid-Structure Interaction in Internal Physiological Flows
Matthias Heil and Andrew L. Hazel p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 141
Numerical Methods for High-Speed Flows
Sergio Pirozzoli p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 163
Fluid Mechanics of Papermaking
Fredrik Lundell, L. Daniel Söderberg, and P. Henrik Alfredsson p p p p p p p p p p p p p p p p p p p p p p p 195
Lagrangian Dynamics and Models of the Velocity Gradient Tensor
in Turbulent Flows
Charles Meneveau p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 219
Actuators for Active Flow Control
Louis N. Cattafesta III and Mark Sheplak p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 247
Fluid Dynamics of Dissolved Polymer Molecules
in Confined Geometries
Michael D. Graham p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 273
Discrete Conservation Properties of Unstructured Mesh Schemes
J. Blair Perot p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 299
Global Linear Instability
Vassilios Theofilis p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 319
v
FL43-FrontMater ARI 15 November 2010 11:55
High–Reynolds Number Wall Turbulence

Alexander J. Smits, Beverley J. McKeon, and Ivan Marusic p p p p p p p p p p p p p p p p p p p p p p p p p p p p 353
Scale Interactions in Magnetohydrodynamic Turbulence
Pablo D. Mininni p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 377
Optical Particle Characterization in Flows
Cameron Tropea p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 399
Aerodynamic Aspects of Wind Energy Conversion
Jens Nørkær Sørensen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 427
Flapping and Bending Bodies Interacting with Fluid Flows
Michael J. Shelley and Jun Zhang p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 449
Pulse Wave Propagation in the Arterial Tree

Frans N. van de Vosse and Nikos Stergiopulos p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 467

Mammalian Sperm Motility: Observation and Theory
E.A. Gaffney, H. Gadêlha, D.J. Smith, J.R. Blake, and J.C. Kirkman-Brown p p p p p p p 501
Shear-Layer Instabilities: Particle Image Velocimetry Measurements
and Implications for Acoustics
Scott C. Morris p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 529
Rip Currents
Robert A. Dalrymple, Jamie H. MacMahan, Ad J.H.M. Reniers,
and Varjola Nelko p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 551
Planetary Magnetic Fields and Fluid Dynamos
Chris A. Jones p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 583
Surfactant Effects on Bubble Motion and Bubbly Flows
Shu Takagi and Yoichiro Matsumoto p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 615
Collective Hydrodynamics of Swimming Microorganisms: Living Fluids
Donald L. Koch and Ganesh Subramanian p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 637
Aerobreakup of Newtonian and Viscoelastic Liquids
T.G. Theofanous p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 661
Indexes
Cumulative Index of Contributing Authors, Volumes 1–43 p p p p p p p p p p p p p p p p p p p p p p p p p p p p 691

Cumulative Index of Chapter Titles, Volumes 1–43 p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 699
Errata
An online log of corrections to Annual Review of Fluid Mechanics articles may be found
at http://fluid.annualreviews.org/errata.shtml
vi Contents

Fluid Mechanics Papermaking

Uploaded by

Copyright:

Available Formats

Fluid Mechanics Papermaking

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fluid Mechanics Papermaking

Uploaded by

Copyright:

Available Formats

FL43CH09-Alfredsson ARI 15 November 2010 13:28

• Our comprehensive search and P. Henrik Alfredsson1

Annu. Rev. Fluid Mech. 2011. 43:195–217 Keywords

The Annual Review of Fluid Mechanics is online at Abstract

1. INTRODUCTION TO PAPER MANUFACTURING

1.1. What Is Paper?

1.2. The Paper Industry

196 Lundell · Söderberg · Alfredsson

1.3. This Paper

ﬂuid mechanical opportunities are discussed.

2. THE PAPER MILL AND ITS MAIN FLUID DYNAMICS

Raw material Pulping Additives Refining Fractionation

Sheet Pressing Drying Surface Product

www.annualreviews.org • Fluid Mechanics of Papermaking 197

Fiber flocs between the bars

2.2. Cleaning, Fractionation, and Separation

198 Lundell · Söderberg · Alfredsson

2.3. Sheet Forming

www.annualreviews.org • Fluid Mechanics of Papermaking 199

Water flowing in a structured fiber network

Fiber suspension Fiber suspension

Forming section: 3. CELLULOSE FIBERS AND FIBER SUSPENSIONS

200 Lundell · Söderberg · Alfredsson

3.2. The Pulp Fiber Suspension

www.annualreviews.org • Fluid Mechanics of Papermaking 201

1,000 L/d = 1,000 Network

0.01 0.1 1.0 10

202 Lundell · Söderberg · Alfredsson

Water Plug flow

4. FLUID MECHANICS OF SHEET FORMING

www.annualreviews.org • Fluid Mechanics of Papermaking 203

Overflow Headbox jet

204 Lundell · Söderberg · Alfredsson

www.annualreviews.org • Fluid Mechanics of Papermaking 205

206 Lundell · Söderberg · Alfredsson

4.2. The Headbox Jet

www.annualreviews.org • Fluid Mechanics of Papermaking 207

x = x ∗ /a, y = y ∗ /a, u = u ∗ /U m , v = v ∗ /U m , and p = p ∗ /(ρU m2 ),

208 Lundell · Söderberg · Alfredsson

{u  , v  , p  } = {û(y; X ), v̂(y; X ), p̂(y; X )}e i(αx−ωt) . (4)

www.annualreviews.org • Fluid Mechanics of Papermaking 209

a Re = 700 Re = 830 Re = 920 Re = 1,050 b

eI eII eIII oI oII

4.3. Initial Dewatering

210 Lundell · Söderberg · Alfredsson

www.annualreviews.org • Fluid Mechanics of Papermaking 211

212 Lundell · Söderberg · Alfredsson

5. FUTURE DEVELOPMENTS AND POSSIBILITIES

www.annualreviews.org • Fluid Mechanics of Papermaking 213

distributions of the ﬁbers and other material in the sheet.

214 Lundell · Söderberg · Alfredsson

www.annualreviews.org • Fluid Mechanics of Papermaking 215

Norman B. 1979. Principles of twin-wire forming. Sven. Papp. 82:330–36

216 Lundell · Söderberg · Alfredsson

www.annualreviews.org • Fluid Mechanics of Papermaking 217

Contents Fluid Mechanics

Volume 43, 2011

Experimental Studies of Transition to Turbulence in a Pipe

Theodore Yaotsu Wu p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p25

{u , v , p } = {û(y; X ), v̂(y; X ), p̂(y; X )}e i(αx−ωt) . (4)