Physical Testing of Paper

Physi
calTes
ting
ofPaper
RomanE.Popi
l
,Ph.
D.
Physical Testing of Paper
Roman E. Popil, Ph.D.
A Smithers Group Company
Shawbury, Shrewsbury, Shropshire, SY4 4NR, United Kingdom

Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118
http://www.polymer-books.com
First Published in 2017 by
Smithers Pira
Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK
©Smithers Information Ltd., 2017
All rights reserved. Except as permitted under current legislation no part

of this publication may be photocopied, reproduced or distributed in any
form or by any means or stored in a database or retrieval system, without
the prior permission from the copyright holder.
A catalogue record for this book is available from the British Library.
Every effort has been made to contact copyright holders of any material
reproduced within the text and the authors and publishers apologise if
any have been overlooked.
ISBN: 978-191024-292-6 (Softback)

978-191024-293-3 (ebook)
Typeset by Argil Services

P
reface
The ‘Physical Testing of Paper’ is a broad term which covers the

measurement of physical properties of various papers, paperboards,
corrugated boards and corrugated board boxes. Perhaps ‘wood-
fibre based laminate composite structures’ could be substituted as
being more encompassing than the generic ‘paper’ but that would
make a rather long title. Quite often, this vast subject receives only
cursory treatment in institutions where papermaking is taught as an
ancillary course to a chemical engineering curriculum concerning the
manufacturing processes of paper. An unfortunate and misleading
impression encountered here is that paper testing is simply a matter
of pressing a button on some ready-made instrument and recording
the resulting number and is therefore undeserving of much attention
or focus – not so! I have countered this somewhat condescending
oversimplification by stating in retaliation that chemistry by contrast,
appears to be, to the casual hallway passer-by, simply mixing various
chemicals in a beaker while wearing a white lab coat! This amusing
discourse continues to this day with my colleagues with the results
being as can be imagined.
This book arises from my experience gained from entering the paper
industry with an experimental physics background some decades
ago. In this approach, test methods that have been documented as
standards are subject to scrutiny, interpretation, improvement and
improvisation. There is the requirement and desire to understand
what is going on from a fundamental mechanics view when a paper
sample test piece undergoing a test is being pulled or torn, bent or
poked. Repeatedly, I find myself required to explain in layman’s
terms to my testing lab clients what it is that is going on in their
iii
submitted paper samples that disallows them to be nicely printed,

written on without ink blobbing onto the next page, glued together so
the assembled carton does not fall apart or folded without cracking.
There is meaning and satisfaction in linking the results of various
different tests to diagnose some end-use issue such as gluability,
printability or corrugated box storage longevity. The significance
of the results obtained from various testing methods is what is
emphasised in this volume rather than the details of the test methods,
which are documented elsewhere.
In this book I have focused on the basic physical tests that have
demonstrated interrelationships and solved problems over the years.
Many more different tests are occasionally requested but have not
resulted in providing significant insight and are not included here for
brevity. These include ZD out-of-plane tensile, Scott Bond, folding
resistance, surface abrasion resistance, surface friction, surface
electrical resistivity, water absorption and water resistance in its
many forms. I refer the interested reader wishing to delve further to
the references cited below for more details of the testing methods,
as well as the other listed texts for the paper physics associated with
paper testing.
The book is written from a personal view based on my initial

experience as an instrument developer for a newsprint production
company in Canada, followed by similar activity for a sensor scanner
manufacturer in California. Here, new measurement methods
and techniques were developed to provide unique paper property
characterisation that was not otherwise available by conventional
testing. Since 2003, I have managed the Physical Analysis Lab at the
Institute of Paper Technology and Science in Atlanta, GA, USA, now
renamed the Renewable Bioproducts Institute as part of the Georgia
Institute of Technology in Atlanta, GA, USA. Here, I have encountered
various testing requests from a variety of paper manufacturers and
end-users to address a wide variety of quality issues. Although much
of the activity still involves the apparent tedium of repeatedly pressing
a button on an instrument and recording the resulting numbers, there
is excitement, which I convey to my clients and any students within
iv
Preface
earshot, that lies in realising what all the numbers mean and how
they can be used together to solve a particular problem.
Thus, I invite the reader to share the joy and excitement of paper
testing by taking the time to explore some of the fundamentals of
paper physics related to property measurements, to take a curious
investigative and skeptical approach to the subject and finally, realise
how data obtained from testing can be used to tell a story that is not
only informative, but also gratifying.
Bibliography
1. TAPPI Testing Methods, Tappi Press, Atlanta, GA, USA.
2. Handbook of Physical Testing of Paper, 2nd Edition, Eds.,

R.E. Mark, C.C. Habeger, Jr., J. Borch and M.B. Lyne,
Marcel Decker, Inc., New York, NY, USA, 2002.
3. K.J. Niskanen in Paper Physics Book 16 of Papermaking

Science and Technology, Fapet OY, Helsinki, Finland, 1998.
4. J.E. Levlin and L. Söderhjelm in Pulp and Paper Testing,

Fapet Oy, Helsinki, Finland, 1999.
5. J.D. Peel in Paper Science and Manufacture, Angus Wild,

Vancouver, Canada, 1999.
6. G.A. Smook in Handbook for Pulp and Paper Technologists,

Angus Wild, Vancouver, Canada, 2002.
v
vi
C
ontents
1 Introduction –What is Paper?................................................ 1

1.1 Paper Structure, Paper Models: Fettuccine, Straw ....... 5
1.2 Formation of Paper: In-plane Lumpiness .................... 7
1.3 Hydrophilicity – Paper Really Sucks – Water
That Is! . ..................................................................... 9
1.4 Paper Property Variability – Why Paper Testing is
Necessary ................................................................. 14
1.5 Will it be Paper or Plastic? – Why Ask?..................... 16
1.6 There are Two Sides to Every Sheet of Paper ............ 16
1.7 Twin Formers Nomenclature .................................... 17
1.8 The Many Types of Paper.......................................... 18
1.9 Summary................................................................... 19
2 Tensile Properties................................................................ 23
2.1 Basic Mechanics of the Tensile Test........................... 25
2.2 Effect of Test Specimen Size on Tensile Strength........ 28
2.3 A Study of the Effect of Sample Size and
Deformation Rate . ................................................... 29
2.4 Units, Breaking Length ............................................. 33
2.5 Summary................................................................... 34
3 Ultrasonic Testing of Paper.................................................. 37

3.1 Introduction.............................................................. 37
3.2 ZD Ultrasonic Measurement..................................... 42
3.3 Correlation of Ultrasonic Results with Mechanical

Properties ................................................................. 44
3.4 Application of ZD Ultrasonics . ................................ 49
3.5 Summary................................................................... 53
4 Bending Stiffness of Paper and Corrugated Board: The

Connection to Caliper and Tensile Stiffness......................... 57
4.1 The Bending Elasticity Theory .................................. 57
4.2 A Comparison of Different Methods of Measuring
Bending Stiffness........................................................ 61
4.3 Bending Stiffness of Corrugated Boards .................... 67
4.4 A Comparison of Three- and Four-Point Bending
Results for a Series of Boards . .................................. 70
4.5 Summary................................................................... 77
5 Compression Testing of Paper, Board and Boxes:

Relationship to Tensile Testing, Elastic Modulus and
the Influence of Artefacts..................................................... 79
5.1 Introduction.............................................................. 79
5.2 Compression Testing of Corrugated Board................ 96
5.3 Three Different Edge Compression Strength Test
Methods.................................................................... 97
5.4 Bending during Board Compression Testing............ 103
5.5 Experimental Observations – Effect of Test
Specimen Height . ................................................... 108
5.5.1 C-Flute Board............................................. 108
5.5.2 E-, F- and N-Flute Boards .......................... 111
5.6 Facing Buckling during Board Compression
Testing..................................................................... 115
5.7 Compression Strength of Boxes – McKee
Formula................................................................... 128
5.8 Summary................................................................. 131
viii
Contents
6 Testing Methods for Measurement of the Writability and

Printability of Papers......................................................... 135
6.1 Background............................................................. 135
6.2 Controlled Application of Ink: Using the Bristow
Wheel...................................................................... 137
6.3 Bleed-Through/Show-Through Measurement ......... 141
6.4 Water Drop Contact Angle and Angle Change
Rate . ...................................................................... 146
6.5 Cobb Water Absorption Test .................................. 149
6.6 Caliper and Basis Weight ........................................ 150
6.7 Air Permeability or Porosity.................................... 151
6.8 Surface Roughness by Air Leak and Contacting
Stylus Profilometer................................................... 152
6.9 Sizing Test Ink Penetration – Hercules Size Test ...... 157
6.10 Results and Analysis ............................................... 158
6.11 Physical Testing for Bank Cheque Ink-Jet
Printability . ............................................................ 166
6.12 Summary................................................................. 169
7 ‘Beater Curves’ to Evaluate Pulp Potential: Burst, Tensile,

Tear, Opacity..................................................................... 173
7.1 Background............................................................. 173
7.2 Tear Testing of Paper............................................... 174
7.3 Burst Testing of Paper.............................................. 177
7.4 Pulp Beating ........................................................... 179
7.5 Results of Physical Properties from Pulp Beating .... 182
7.6 Summary................................................................. 185
Abbreviations............................................................................ 189
Index......................................................................................... 193
ix
x
1
Introduction –What is Paper?
One of my former physics professors upon hearing that I had started

working in the paper industry as a research scientist, dismissed all
the alleged complexities of paper structure with a wave of the hand
and saying that paper is ‘just reconstituted wood’. Those three
words, although being a typically glib oversimplification, are actually
true and are helpful to understand the nature of those things we
usually call paper. However, it was indeed an 18th century physicist,
Antoine de Reamur [1], who suggested in 1719 that his observation
of the maceration of wood by wasps to form their nests could be
replicated by humans to make something similar and useful. At the
time, paper was made from disintegrated cotton rags which were
increasingly in short supply as demand increased. Wood is wood
and paper is paper and, other than some wallpaper products that are
deliberately printed to look like wood, they do not look the same.
The similarity, however, is that all paper consists of conglomerates
of bonded fibres, all of which were once the living cells of a tree or
plant [2]. These fibres are able to form paper because they consist
of cellulose, which is hydrophilic, and so are able to absorb and
retain water. The individual fibres are also typically small, widths are
approximately 30 μm, about one-third of the width of a human hair,
and 1 or 2 mm long, sometimes longer or shorter depending on the
wood or plant species [3]. If a piece of paper is torn, these fibres can
be seen by the eye sticking out along the edge of the tear. Once the
fibres have been separated from the plant, by mechanical or chemical
means, dispersed in water and then strained, the hydrophilicity of
the cellulosic surface draws the fibres together and as the strained
mat dries, enough water becomes entrained to form hydrogen bonds
between the individual fibres, forming a matrix called paper [4]. If the
1
process is reversed, i.e., a copious amount of water is added to paper,

the water/paper mix is sufficiently agitated to disperse the paper into
separated fibres and the resulting slurry is strained and then dried,
the mat will reform [5] albeit with some irreversible changes to the
fibres and loss of sheet strength. Hence, the rather dismissive moniker
of ‘reconstituted wood’ is justified using this wet-laid strained slurry
process by which most paper is made. Thinking of paper as consisting
of agglomerations of fibres can be visualised by the analogy of fallen
pine tree needles on a road, as shown in Figure 1.1.
Figure 1.1 A naturally occurring paper analogy can be mats and

clumps of intertwined pine ‘straw’ laying on a road. Repeated
traffic has aligned the ‘straws’ somewhat in a left to right direction
in this photograph
Here, the aspect ratio of fibre width to length is similar to paper

magnified to a scale of 100×. Thus, paper consists of a mat of
cellulose-containing fibres that are bonded together via the hydrogen-
bonding mechanism occurring through fibre contact. On the scale
of approximately 1 mm, which can be easily observed through a
low-power microscope, the surface of paper in transmitted light can
2
be said to resemble a plate of cooked fettuccine pasta. The wood

fibres, which originated as hollow tubes in the papermaking stock
dispersion, have become transversely compressed and pressed dry in
the paper. They have thus collapsed to form flat ribbons with a dog
bone-shaped cross-section, as may be gleaned from the 2-photon
fluorescence scanning laser confocal shot [6] of Figure 1.2 and the
optical cross-section of newsprint shown in Figure 1.3. Chemically
prepared paper, referred to as kraft, (German for ‘strong’) forms
flattened ribbon fibres, obvious in Figure 1.3, where the newsprint
shown in cross-sections consists of flattened kraft fibres as well as
thick-walled uncollapsed mechanical pulp fibres. Thick-walled fibres
are also attributable to latewood, a result of fibre walls thickening
as tree growth slows down during colder temperatures, creating the
rings that we see when looking at the cross-section of a cut tree.
Figure 1.2 A confocal 2-photon fluorescence scanning laser

microscopy shot showing a layer of paper approximately 1 mm in
square area (image used for the front cover)
3
63.9 µm 52.0 µm 60.6 µm
Figure 1.3 Optical microscopy cross-section of newsprint
So, here is a model to bear in mind, paper is reconstituted wood

that looks like a plate of fettuccine pasta! Thinking of paper as
‘reconstituted wood’ is not entirely original.
Some undergraduate structural-engineering classes are tasked with

a team challenge exercise to design and build a scale model bridge
structure using uncooked pasta as the building material. The idea is
to compete in terms of how much load the model bridge will bear
before collapsing. Obviously and intuitively, the ultimate load such
a model bridge will bear is a function of the structural design, but
is also dependent on the properties of the integral building block,
in this case, the choice of pasta. Using rigatoni (a short tube with a
fluted external surface) may provide advantages over long slender
fettuccine or spaghetti strands. Similarly when building churches,
using cut stone blocks provides an advantage over adobe mud bricks,
although the wall thickness in both cases may be over 3 m.
The properties of paper are similarly determined, to a large extent, by

the fibre properties which are dependent on the wood or plant species,
pulp type and any subsequent mechanical or chemical treatments.
Softwood unbleached chemical pulp (kraft) fibres [7] are preferred
for strength properties compared with hardwood bleached fibres
[8], which are instead preferred for office copy paper. Mechanical
pulp fibres, the real ‘reconstituted wood’, are comparably coarse in
4
comparison to chemical pulp fibres and are found in newsprint and

coated publishing grades. The advantage of kraft pulp is that much of
the non-bonding lignin has been removed, allowing the cellulose fibres
to bond effectively and also become irreversibly highly bleached.
Mechanical pulp, traditionally groundwood or thermomechanical
refiner pulp, contains the lignin originally present in the wood which
interferes with fibre bonding and cannot be permanently bleached,
thus causing the familiar yellowing of newsprint or paperback books.
1.1 Paper Structure, Paper Models: Fettuccine, Straw

Thinking of paper as a matrix of bonded fibres, flat wooden
toothpicks, fettuccine pasta and so on, is helpful in appreciating that
the properties of paper are directional. A partial alignment of pine
needle straws in Figure 1.1 can be imagined as a mattress made of
straw and accordingly by analogy, paper, consisting of fibres bonded
at their contact points, will be highly compressible in the out-of-plane
z direction of paper (ZD). This property allows the reproduction of
images onto paper using contact methods such as letterpress, offset,
rotogravure and flexography, and imparts tissue tactile softness.
Most paper is made using the wet-laid process at speeds of 1,000 to

10,000 ft/min. The fibres are separated from the wood matrix and
thinly suspended in water to a concentration of 0.5% or less [9].
This stock slurry is sprayed from a long slot, called the slice of the
headbox, onto a wide moving straining mesh belt, historically called
the ‘wire’, through which much of the water drains leaving a mat
of wet fibres on the moving wire to be later lifted off the wire and
compacted and dried to form the finished paper product. Spraying
the fibre slurry at one velocity out of the paper machine headbox
slice onto a draining wire moving at a different velocity, along with
the finite time required for the water to drain through the wire, has
two significant effects on the resulting fibre mat: orientation [10]
and formation [11]. The hydrodynamics that occurs during a few
milliseconds on the moving wire orientates fibres along the machine
direction of machine-made paper (MD). The paper fibre mat, in
5
various states of dewatering, drying and compacting, is pulled

under tension in the MD causing stresses and further preferential
orientation in the MD. The consequence is that most machine-made
papers are strongest in the MD orientation, whereas the cross,
orthogonal direction to the machine direction of paper (CD) is often
weaker by approximately 1.5× or more. Thus, we have the principal
directions of paper defined as MD, CD cross-direction and ZD shown
schematically in Figure 1.4.
Fibres: former wood cells, typically

1–3 mm long 50 µm wide
ZD
CD
MD
Paper consists of 30–50% air and a network of bonded fibres

aligned predominantly along the MD
Figure 1.4 Schematic representation of paper structure showing

the three principal directions of oriented machine-made paper.
Fibres are depicted as dashes on the surface
This directionality in paper may be appreciated by attempting to

rip rectangular pieces out of newsprint without using scissors. Most
6
newspapers are pulled left to right in the reading direction. A paper

roll will be made with the MD perpendicular to the axis roll, as
shown in Figure 1.5.
Sample cut
along the MD
MD
Paper from
mills is usually
supplied in
roll form
Figure 1.5 A machine-made paper roll with the MD indicated
When tearing across the page along the print direction, the tear
usually follows a nice straight line along the direction of tear as this is
along the fibre direction. However, when attempting to tear newsprint
down or up the page which is the CD, the tear will propagate across
the fibres and will not be in a straight line, leading one to ultimately
reach for the scissors in order to proceed without frustration.
1.2 Formation of Paper: In-plane Lumpiness

During the paper machine former stage, the fibres in the slurry on
the moving drainage wire become aligned but also tumble as the
water is drained, which allows them to entrain neighbouring fibres
and intertwine [12]. This leads to paper becoming non-uniform in
7
mass on a small scale, i.e., of the order of fibre dimensions. This non-
uniformity can be seen by eye by holding paper up against a light
and noting the light and dark areas, the most discernible would be
approximately 5 mm in size. Light areas represent lower mass than
the surrounding darker areas. Figure 1.6 is a transmitted light image
of a 64 mm square of corrugated board using a night vision camera
sensitive in the near infrared wavelength region, which scatters less
through paper. Beside that image is a standard reflected light image
of the same linerboard separated from the fluting which now shows
dark iodine-stained glue lines and the agglomeration of starch from a
faulty adhesive application. The transmitted image shows a mottled
appearance corresponding to the non-uniform distribution of fibres
in clumps that are roughly the same scale between the glue lines,
approximately 8 mm in this case. These light areas, corresponding
to lower local mass, will accept ink differently leading to undesirable
mottle in printed images [13], and will be weaker in strength than
adjacent darker areas leading to potential failure at those points when
the paper is placed under tension. A considerable amount of effort in
paper manufacture is focused on minimising the severity of this fibre
clumping. Fast drainage speeds and low slurry consistencies are two
of the simpler strategies already mentioned. Paper non-uniformity is
an inevitable result of being comprised of contacting fibres, much like
the clumping of straw strewn on a floor or a road as in Figure 1.1.
Paper consists of clumps of intertwined fibres sticking to each other

via hydrogen bonding at their contact points. Thus, the surface is
actually an ill-defined boundary which depends on the degree of fibre
compaction and collapse. This out-of-plane structure complicates the
measurement of paper roughness [15], caliper [16] and porosity [17].
In-plane, the non-uniformity originating from the clumping of fibres
influences the test results making them affected by test specimen size,
and introduces variability or a coefficient of variation (cv) of several
per cent in nearly all measurements of physical properties, many of
which are proportional to the basis weight. The inherent variability
of paper causes considerable consternation for circumstances where
paper properties are required to meet stringent quality criteria.
The average value of a highly variable paper property must be set
8
high to avoid samples that are below a set product specification.

The proportionality of strength properties on basis weight means
manufacturers must often meet the marketing specifications by
increasing the basis weight of their products at a loss of profitability.
Figure 1.6 Left is a transmitted light image through corrugated

board showing fluting glue lines and right is the corresponding
reflected light image. Reproduced with permission from S. Johnson
and R. Popil, International Journal of Adhesives and Adhesion,
2015, 59, 105. ©2015, Elsevier [14]
1.3 Hydrophilicity – Paper Really Sucks – Water That Is!

The hydrophilic nature of cellulose enables paper to hold together
and allows it to be recycled [18]. However, the propensity to retain
moisture has consequences for paper testing, i.e., the moisture
exposure history of a sample has to be considered [19]. An extreme
example of the effects of moisture, which many people experience,
is the accidental wetting of the pages of a book. Here, the pages of
the book expand, warp out-of-plane and once dried will never be
9
the flat smooth thin pages that they once were. In paper testing,
samples of course, must be kept dry, but the ambient humidity must
also be consistent and controlled. All paper contains some degree
of moisture due to equilibration with ambient humidity. In high-
humidity environments such as refrigerated rooms or uncontrolled
warehouses of the southern US, the moisture content percentage in
paper (%M) on a wet basis (i.e., weight of water/total weight as per
Equation 1.1) can be 12% or more:
wet – dry (weight)

%M = × 100 (1.1)
wet weight
In an arid environment, such as Mexico City in winter, the paper

moisture will be around 4%. Some press rooms in Mexico City find
that sheets of boards for printing curl up into tubes and will not run
through the press. In southeastern US, stacks of boxes containing
milk jugs topple over crushing the plastic milk jugs they contain.
These phenomena are consequences of the hydrophilicity of paper.
Accordingly, paper is manufactured to specifications for an

environment at 50% relative humidity (RH) and room temperature
(RT). The moisture of paper in these conditions is usually
approximately 7.5%. Rolls and reams of paper products are wrapped
in moisture-proof covers during shipping so that the paper will
have the expected specifications at 50% RH. Thus, to be consistent,
paper testing is performed at 50% RH with an RT of 23 °C [20]. If
testing were to be performed at a lower humidity, the paper moisture
would become lower and many strength properties would become
proportionally higher. Conversely, if paper were to be tested at a
higher humidity than 50%, the paper moisture would become higher
and the strength properties would be proportionally lower from those
specified at 50% RH. Figure 1.7 shows a moisture history curve of
a paper’s moisture content adjusting to relative ambient humidity.
10
26
24 The same sample from
22 a humid environment
Moisture content (%)
20
18
Sample from high
16 humidity attains
14 8% M at 50% RH
12
10 Bring a sample
8 from 20 to 50%
RH, then % M is
6 7%
Condition sample 4
at 20% RH, its %M
will be 4% 2
0
0 20 40 60 80 100
RH (%)
Figure 1.7 A hypothetical paper moisture curve showing the

history cycle of increasing and decreasing ambient humidity
If paper is left to equilibrate to a high-humidity environment (>50%),

as happens with unwrapped rolls of paper left in an uncontrolled
press room or warehouse for long periods of time, and tested at 50%
RH, the strength properties will be measurably lower than when
originally measured. This is another consequence of the hydrophilic
nature of paper and can be described as the moisture hysteresis of
paper. When paper is exposed to high humidity and equilibrated to
a lower humidity, it will have a slightly higher humidity than paper
brought from a lower humidity to the testing standard humidity of
50%. The aim for a paper testing laboratory is to obtain consistent
reproducible property values. Incoming samples often have an
unknown moisture history so the standard practice is to equilibrate
paper samples to a low-moisture state followed by equilibration to
the 50% RH and perform the testing at 50% RH.
11
The question often arises as to what the duration of preconditioning/

conditioning periods should be prior to testing. In production mills,
paper quality is required to be verified quickly when machine or stock
changes are made. The Technical Association of the Pulp and Paper
Industry (TAPPI) method protocol of 24 h sample storage at 20%
RH followed by 24 h at 50% RH is intended to encompass reams
of paper sheets and boards. However, a single sheet suspended in
moderately moving air will equilibrate to 90% of its final moisture
value in 15 min and fully in 2 h. Some mill testing labs have adopted
the practice of placing samples into a 1 kW microwave oven for 10 s
to ensure the paper is dried and then suspending the sheets in 50%
RH moving air for 2 h prior to testing strength properties.
Figure 1.8 shows the exponential change in time of a moisture-

dependent property of dry paper acclimating from a dry state to
its equilibrium point in a humid ambient environment. Of course a
longer time is better, but a minimum can be 1 h for a single sheet in
circulating air flow or 5 h for a corrugated board [21]. Nonetheless,
urgency requires rapid measurements in a mill production situation
and measurements are taken of samples at various moistures. In such
cases, a corrective factor may be applied if the moisture content of the
sample is known at the time of measurement, as product specifications
are based on paper equilibrated to 50% RH. One study, requested by
a mill for softwood kraft linerboard, required obtaining a moisture
correction for measurements of compression strength, which is
important for corrugated box manufacture. The specific compression
strength of interest here is called the short-span compression test or
strength (SCT). Measurements for a moisture corrective factor were
made for the mill samples equilibrated in various humidity values
from 20 to 80% and the results are shown in Figure 1.9.
12
%M or other
moisture-dependent
property 1 h for single sheets
~ 5 h for corrugated boards
time
Figure 1.8 Hypothetical %M, strength, basis weight versus time

for dry paper equilibrating in a humid environment
1.5
1.4
1.3 SCT at 50% RH

1.2
SCT corrective multiplier
80% RH
1.1
0.9
0.8 SCT%M/SCT50%RH = 0.531e0.089%M
0.7
20% RH
0.6
4 5 6 7 8 9 10 11
% Moisture content
Figure 1.9 SCT of kraft linerboard measured at various humidities

relative to the standard value of 50% RH
13
Basic theory predicts that the dependence of strength properties

on moisture will follow exponential behaviour, and the results
are expressed as a corrective factor to the values at 50% RH in
Equation 1.2. An increased paper moisture content has the effect of
decreasing the fibre elastic modulus and accordingly, strength values
follow the same trend.
strength%M = strength50% RH × 0.531e0.089(%M) (1.2)
To a very good approximation, followed by many instruments that

correct for moisture, a linear interpolation of the strength moisture
dependence is used and shown in Equation 1.3:
strength50% RH = strength(%M) × [(%M ×)0.07 + 0.47] (1.3)
1.4 Paper Property Variability – Why Paper Testing is

Necessary
All paper has variability on a millimetre scale due to it being
composed of bonded fibres, as discussed above. The manufacturing
process introduces yet another variability on the scale of the test
specimen size of several centimetres. Paper is often manufactured in
high volumes on a high-speed paper machine which produces at its
end, a jumbo roll of paper on the reel being some 20 tonnes perhaps
20 to 60 ft wide, a schematic representation is shown in Figure 1.10.
There are inevitable profiles in the CD [22] that arise from stock
velocity differences, which fall to zero at the edges of the forming
wire, drying temperature differences across the machine and lack of
tensile restraint in the CD.
14
Front or Back or
tending side drive side
MD
Figure 1.10 Schematic representation of the jumbo roll at the dry

end of a paper machine
Paper machine rolls used for pressing and calendering introduce a

pressure profile that may not be compensated by the eventual wear
of the roll or the supporting bearings. Along the MD, pulsation
in the pressures at the stack headbox will manifest as variation in
basis weight, any press or calendering rolls that are not perfectly
round will also introduce a basis weight variation. Press felts with
any defects, forming wire anomalies and uneven seams, introduce
periodic artefacts in the sheet [23]. On a longer time scale, there are
also variations in the raw material of the fibre stock furnish arising
from both variations in the stock preparation process and the supply
of the raw material, wood or recycled materials. The combination of
all these variations requires any testing of paper to consist of repeated
tests on the replicates of the paper sample to account for the inherent
variability. Typically, 10 repeats are made for each sample when
testing for tensile or compression strength. Repeat measurements are
used to examine the result such that if the cv exceeds a typical value
of 7% or so, then either the sampling method, sample preparation
or the instrument becomes suspect and requires investigation.
15
1.5 Will it be Paper or Plastic? – Why Ask?

Typical variability in any paper can be expressed as the cv,
(standard deviation/average value) of whatever property is of
interest. The clumping of fibres from the wet-laid straining process
leads to a mass variation of around 6 to 8% [24]. Strength is
usually proportional to paper mass so it follows that paper strength
properties also display a cv of this order. This variability is what
governed the retailers’ answer to whether paper or plastic would
be used to bag groceries, i.e., usually plastic. A shopping bag is
made to meet strength specifications but to reduce basis weight it
will not have a strength value much above the specified lower limit.
The variation in strength will mean there will be weak spots in
the paper bag where a tear may initiate. Paper under tension will
not stretch very much, the fibres become pulled from one another
and a tear opens up at a weak low basis weight point which then
propagates along similar low basis weight points, resulting in
groceries falling onto the pavement. Plastic bags do not have the
cv% variability of paper and the stretch to failure exceeds that of
paper by approximately 400×, allowing one to realise long before
failure that a bag is about to fail.
1.6 There are Two Sides to Every Sheet of Paper

Paper sheets have two sides that usually appear to be the same but
are always different to some extent. The wet-laid stock straining
process of forming paper imposes asymmetry in the cross ZD profile.
Traditionally, the two sides of paper refer to the fourdrinier paper
machine, where sheets are formed by draining water via gravity on
the moving straining wire mesh and the resulting mat is pressed free
from water on the top by a moving belt of an absorbing felt fabric.
Accordingly, fourdrinier-made papers are said to have a ‘top’ or ‘felt’
side and a ‘bottom’ or ‘wire’ side. The direction of the free water
being drained from the fibre mat is towards the wire side. The stock
consists of fibres which inevitably have a length distribution such that
there is usually a larger proportion of shorter fibres and submillimetre
16
fibre fragments called ‘fines’. The directional drainage of the stock

combined with the interstices of the forming wire results in the loss
of the fines stock fraction at the wire side. Moreover, contact of the
mat with the wire will superimpose the topography of the wire fabric
onto the bottom side of the sheet.
The consequence is that the top or felt side of the sheet will often
be smoother, the surface having fewer voids, and the bottom or
wire side will be rougher with more voids between the long fibres.
The structure results in the two sides of the sheet absorbing water
differently. Printing and writing will appear differently on the
two sides of the sheet [25], especially when using aqueous inks.
Applying a coating will also have different results on a fourdrinier
sheet.
Air flow is somewhat affected by the two sidedness, much like

forced air-furnace filters, papers are more permeable when air flow
is towards the more open porous wire side. The in-plane ductility of
the sheet is also affected by two sidedness. Fibres in the wire side will
be surrounded with less bonded fines, which act as a filling cement to
transfer stresses so that when in tension, the wire side will have lower
strength. This is consequential whenever paper is stressed in a curved
geometry, such as in the burst test where paper is punctured while
being secured over an expanding rubber diaphragm or, when paper
is folded or scored when making carton packaging. When folding
paper, it is preferred to have the felt side on the outside of the fold
so it can endure the higher stresses from the imposed curvature and
thereby minimise the likelihood of cracking.
1.7 Twin Formers Nomenclature

Obviously, two sidedness of paper is not a desirable property, so
considerable effort has focused on forming technology to minimise
or eliminate the asymmetry. There are twin-wire formers, roll formers
or top former retrofits. The idea here is to dewater from both sides
of the sheet as simultaneously as possible; nonetheless, some two
17
sidedness remains. The paper is carried to the press section by the

conveying wire side of the paper (CW) while the opposite side of the
sheet is in contact with the backing wire (BW) side of paper.
Whenever water is drained from a consolidating fibre mat, the fibre

fines and filler-particulate matter are carried with the water so that
in the case of twin-wire forming, the middle of the cross-section
of the sheet is proportionally devoid of matter compared with the
CW or BW surfaces. The lower density in the middle of the sheet
on twin-wire formed papers makes the paper weaker in the middle
so that if the sheet is split by applying tension to both sides of the
sheet, as done in some types of paper tests, the twin-wire formed
sheet will split evenly through its thickness. In contrast, fourdrinier-
made papers will split out-of-plane proportionally weighted to the
felt side since the felt side has retained more of the fines which serve
to bond themselves and the fibres to each other.
1.8 The Many Types of Paper

The term ‘paper’ here will apply to all wood fibre-based materials
formed by the wet-laid fibre stock draining process. Most people
are familiar with the common office copier paper, which is made
of bleached kraft chemical pulp. This type of paper is usually
fourdrinier-made, of medium basis weight or weight per unit area, and
is comparatively porous and rough, which is of no consequence for
non-contact printing or writing with viscous inks. Papers encountered
in journals will often be comparatively lightweight, coated and
compressed to a higher density to allow a high-gloss surface and
good reproduction of images. To reduce costs, the base of such papers
will be mechanical pulp, traditionally groundwood but more likely
thermomechanical pulp mixed in with some kraft. Bleaching of pulp
is chemically expensive and weakens the strength of cellulose fibres,
so corrugated boxes are made from unbleached kraft of high basis
weight, as strength is proportional to basis weight. Paper roll cores
are tubes made of many bonded layers of heavyweight unbleached
kraft. Similar heavyweight papers are cartonboards for packaging
18
consumer goods that are often coated on one side to allow colour
printing.
On the opposite scale to the heavy basis weight of packaging

and printing papers are the tissue and towel products. Here, the
combination of absorbency and softness with strength require low-
density bleached kraft. Towel papers contain wet strength polymeric
chemicals, whereas sanitary tissue may contain lubricants to produce
more tactile softness. The demand for these products to have a bright
white appearance requires bleached chemical kraft pulp, and the
tactile softness marketing specification requires the fibre species to
be predominantly hardwood, often eucalyptus from South America.
1.9 Summary
Describing paper as ‘reconstituted wood’ accounts for wood fibre
wet-laid paper being comprised essentially of self-bonding matrices
of fibres. The fibres are usually much longer, with lengths extending
to millimetres, than wider and are flat in cross-section. Papers have
a wide range of densities, with higher densities required for writing
printing and low densities are required for absorbency. Mechanical
pulps produce weaker less bright papers such as newsprint, whereas
chemical pulps produce weaker but often bleached and white papers.
Containerboard corrugated packaging is predominantly made with
unbleached chemical pulp for optimal strength per basis weight.
Machine-made paper has an orientation where fibres aligned in
the MD provide increased strength compared with cross-direction
CD. Fibres lying flat in-plane provide a high compressibility in the
ZD. Paper fibres, being comprised of the polysaccharide polymers
cellulose and hemicellulose, are hydrophilic and retain moisture
absorbed from the ambient atmosphere. The amount of moisture
in the paper proportionally affects its strength properties, thus care
must be taken to ensure paper samples have the desired moisture
level prior to testing.
19
References
1. D. Hunter in Papermaking: The History and Technique of an
Ancient Craft, Dover Publications, Inc., Mineola, NY, USA,
1978.
2. G. Smook in Handbook for Pulp and Paper Technology,

Angus Wilde Publications, Vancouver, Canada, 1992.
3. C. Ververis, K. Georghiou, N. Christodoulakis, P. Santas and

R. Santas, Industrial Crops and Products, 2004, 19, 3, 245.
4. U. Hirn and R. Schennach, Scientific Reports, 2015, 5, 10503.
5. R.C. Howard and W. Bichard in MRS Proceedings,

Cambridge University Press, Cambridge, UK, 1992, 266, 195.
6. S.M. Potter, Current Biology, 1996, 6, 12, 1595.
7. R.A. Horn in Morphology of Wood Pulp Fiber from

Softwoods and Influence on Paper Strength, FSRP-FPL-242,
Forest Products Lab, Madison, WI, USA, 1974.
8. R.A. Horn in Morphology of Pulp Fiber from Hardwoods

and Influence on Paper Strength, FSRP-FPL-312, Forest
Products Lab, Madison, WI, USA, 1978.
9. C.J. Biermann in Handbook of Pulping and Papermaking,

Academic Press, San Diego, CA, USA, 1996.
10. T.R. Hess and P.H. Brodeur, Journal of Pulp and Paper
Science, 1996, 22, 5, J160.
11. D.S. Keller and P. Luner, Review of Scientific Instruments,

1998, 69, 6, 2495.
12. R.J. Kerekes, Nordic Pulp and Paper Research Journal, 2006,
21, 5, 98.
20
13. J-P. Bernie, H. Pande and R. Gratton, TAPPI Journal, 2006,

5, 10, 28.
14. S. Johnson and R. Popil, International Journal of Adhesives

and Adhesion, 2015, 59, 105.
15. R. Xu, P.D. Fleming, A. Pekarovicova and V. Bliznyuk,

Journal of Imaging Science and Technology, 2005, 49, 6,
660.
16. W.A. Wink and G.A. Baum, TAPPI Journal, 1983, 66, 9,

131.
17. C. Hii, Ø.W. Gregersen, G. Chinga-Carrasco and Ø. Eriksen,

Nordic Pulp and Paper, Research Journal, 2012, 27, 2, 388.
18. M.A. Hubbe, R.A. Venditti and O.J. Rojas, BioResources,

2007, 2, 4, 739.
19. W.A. Wink, TAPPI Journal, 1961, 44, 6, 171.
20. TAPPI T 402-om-88: Standard Conditioning and Testing

Atmospheres for Paper, Board, Pulp Hand Sheets and Related
Products, 1992.
21. S. Allaoui, Z. Aboura and M.L. Benzeggagh, Composites

Science and Technology, 2009, 69, 104.
22. S.M. Hoole, S.J. L’anson, M. Ora, T.N. Ashworth, D. Briggs,

B. Phillips and R.W. Hoyland, Paper Technology, 1999, 40,
10, 63.
23. S. McLeod, Z. Nesic, M.S. Davies, G.A. Dumont, F. Lee,

E. Lofkrantz and I. Shaw in the Proceedings of the Dynamic
Modeling Control Applications for Industry Workshop, IEEE
Industry Applications Society, 30th April–1st May, Vancouver,
Canada, 1998, p.59.
21
24. J.M. Hellawell, Paper Technology and Industry, 1973, 14, 1,

24.
25. I.I. Pikulik and J.D. McDonald, TAPPI Journal, 1987, 70, 4,
75.
22
2
Tensile Properties
The most fundamental and informative physical evaluation for paper

is probably that of the tensile test. It provides the elastic properties
of paper which are directly related to fibre quality and the level of
fibre bonding is most aptly described in the 1982 paper by Page and
Seth [1]. To a large extent, most of what is required to be known
about a paper sample can be gleaned from a tensile test provided it
is interpreted correctly. Therefore, tensile testing is routinely used in
manufacturing operations for quality control.
Simply stated, the tensile test is a strip of paper clamped at both ends
and pulled at a constant rate until failure [2]. The load at failure is
the tensile strength (St), which is satisfactory for many purposes but
much more can be gleaned from consideration of what happens while
the strip is pulled to fracture.
The tensile test pulls paper either in the machine direction of

machine-made paper (MD) or the machine direction of paper (CD),
so strips are cut such that the length is along either the MD or CD.
As described in Chapter 1, the MD is the length direction of paper
pulled from a roll. The fibres comprising the paper are aligned in the
MD and so strength properties are generally higher in the MD than
the transverse CD. The strip is firmly clamped at either end leaving
a free gap of several inches in length. A load cell is attached to one
of the two clamps and is affixed to a frame. A typical universal test
frame fitted with clamps for tensile testing is shown in Figure 2.1.
23
Figure 2.1 Photograph of a test frame for tensile testing, left-hand

side shows a fixed load cell clamp, test strip and another clamp
attached to a moving cross-head, and right-hand side shows
close-up detail
In this case, the test frame is connected to software which controls

the test and collects the load cell and displacement data for analysis.
The cross-head is connected to motor-driven screws and the rate of
downward motion is typically set to 25 mm/min. The motion of the
cross-head increases the tension load ‘F’ along with increasing the
displacement along the strip length of ‘Δl’. A graph of tension load
‘F’ versus time for typical copy paper would look like Figure 2.2.
24
Tensile Properties
E, Sb are calculated from the

linear slope
Fmax
St is this
point
∆F
∆l
Displacement (from the encoder on the belt

Tensile ‘slack’
drive, or better, an extensometer on the sample)
Figure 2.2 A hypothetical load-displacement curve for a paper

tensile test
2.1 Basic Mechanics of the Tensile Test.

Once slack is removed from the sample, if it is mounted loosely, the
load increases steadily in what is called the elastic region. The rise
in the load, ‘ΔF’, in typical units of newtons ‘N’, and the change in
displacement Δl of a strip with free length ‘l’ divided by the width ‘w’
of the test sample provide a useful quantity called the tensile stiffness
‘Sb’ defined by Equation 2.1:
S b = DF # l
Dl # w (2.1)
25
N/mm or equivalently kN/m are the units for Sb. Strength for paper
is also cited per unit width so also carries force per unit width
N/mm. The Sb is defined as the elastic modulus (E) multiplied by
the effective caliper ‘t’ in Equation 2.2. The modulus is the material
resistance to mechanical deformation given by Hooke’s law and for
paper and many other materials is directionally dependent [3], so a
strip cut along the MD is ascribed a MD suffix or conversely a CD
suffix for the case of a CD test.
E MD,CD # t = S b MD,CD (2.2)
Strain is the relative displacement ‘Δl/l’ denoted in mechanics as ‘ε’

and stress, the force per unit area, is denoted ‘σ’. In the case of the
uniaxial tensile test strain defined by Equation 2.3:
Dl MD,CD
fMD,CD =
l MD,CD (2.3)
Here for simplicity and as an approximation, we are ignoring the

small Poisson effects, i.e., when a tensile load is applied on a paper
strip cut along one direction, there will also be some dimensional
changes in the orthogonal and out-of-plane directions, which can
be described by a directional Poisson ratio ‘ν’. Similarly, for stress
applied to a paper strip we have Equation 2.4:
DF MD,CD
vMD,CD =
t#w (2.4)
So, the Sb can be cast into the familiar form of Hooke’s law in
Equation 2.5 as:
26
Tensile Properties
Sb MD,CD vMD,CD
E MD,CD = =
t fMD,CD (2.5)
E appears in formulas for other mechanical properties such as bending

stiffness and the speed of sound in paper, which will be explored later
in this book. E is dominated by fibre quality in most papers and can
be used to predict what the paper strength will be. The ‘t’ in the above
equation is the ‘apparent’ thickness of paper, but as stated previously,
the surface of paper is rather ill defined due to the combination of
surface unevenness and paper compressibility, however this issue will
be resolved in a later chapter.
A yield point is reached where departure from linearity occurs in the

tensile load displacement curve as the load increases further with
increasing displacement. Paper stretched beyond this point is said to
be in non-recoverable plastic deformation. Metals stretched beyond
this point are said to be strain hardened. If tensile load is relaxed
beyond this yield point, a permanent deformation results. If tension
is increased beyond the yield point, the displacement increases until
the sample strip fractures and the tensile load drops to zero. The load
where the strip fractures is called the ultimate strength or strength
of the paper. There is usually a relationship between the strength
and the linear reversible Sb values; however, strength can be altered
somewhat through the inclusion of additives in the pulp stock leaving
the Sb unchanged.
The work ‘W’ done on the sample during the tensile test is the area
covered by the load displacement curve defined by Equation 2.6:
lf
W= # F dl
0 (2.6)
where ‘lf’ is the displacement or elongation at failure, and ‘dl’ is the

incremental displacement.
27
This is useful to describe the elastomericity of the sample and is a

combination of the sample failure strength and stretch or strain to
failure. Samples consisting of long, highly curled fibres would have
a high strain to failure, and so a higher W, than samples comprised
of short straight fibres. Although the strain to failure (lf – l0)/l0 is
usually larger in the CD (~4%) than the MD (~1.5%) for most
machine-made papers, W is lower for the CD due to the lower
strength. The tensile work of a sample is sometimes interpreted as
a measure of sample resistance to in-plane tear and may be useful
for a comparative ranking of samples for potential tear resistance.
However, measurements of fracture toughness, which are tensile tests
of specifically nicked samples [4], are more accurate for this purpose.
The TAPPI method for reporting the tensile work or, equivalently,
the energy absorbed (tensile energy absorbed) is normalised to the
transverse area under tensile load, i.e., area of the test strip between
the clamps and so the units are J/m2.
2.2 Effect of Test Specimen Size on Tensile Strength

Since paper consists of non-uniform clumps of fibres, the size of the
sample test piece has an effect on the results [5]. The length direction
of the sample strip aligned along the MD or CD, as accurately as
possible, should be several times the width. When the length of a
tensile test span between clamps is shortened to less than 1 mm,
the fibre network effects become diminished and the results are
considered to be dominated by the fibre St [6]. Indeed, zero span
tests are used as a quality check of fibre quality [7] and interpreted
using the ‘Page equation’ model for St [8].
Dimensions of the strip for paper tensile testing should be several

times the length of the longest fibres. Intuitively, strength can be
understood to become lower with longer strip specimens, as paper
is non-uniform in mass distribution due to the clumping of fibres.
Therefore, the larger the test piece, the greater becomes the likelihood
of stressing a weak part in the test piece.
28
Tensile Properties
Two sizes are commonly used: 10 × 1” and 160 × 15 mm. The

latter is more popular as it is also a convenient size when testing
laboratory-made handsheets and the 15 mm width strips are also
used for short-span compression, hygroexpansivity testing.
Paper fibres are composed of polymeric materials that exhibit

viscoelastic properties, which means that the rate of mechanical
deformation also has an effect on the results of a test. Generally,
strength properties increase approximately 7.5% for every decade
change in timescale [9, 10]. Paper will appear to be stronger if it
becomes stressed at a faster rate. In most paper tests, two common
deformation rates, probably originating from a historical available
equipment convenience, are either 0.5 or 1 inch/min.
The TAPPI method T 494 contains a footnote that if a 1” wide

sample is used with a 7” span length, the displacement rate should
be 1 inch/min, but if the 160 × 15 mm sample is used instead, the
displacement rate should be 0.5 inch/min [2].
2.3 A Study of the Effect of Sample Size and

Deformation Rate
What happens when we use two different rates and two different
sample sizes in the tensile test? Are the differences significant?
Changing the rate from 0.5 to 1 inch/min significantly speeds up the
testing time, especially since the usual protocol stipulates that 10
replicate tests be made for a sample. Are 10 replicate tests necessary,
would 20 be better or would 5 suffice?
Let’s examine the possibilities. As representative paper samples for

this testing exercise, we will take a typical unbleached kraft linerboard
of 205 g/m2, bleached kraft copier paper of 75 g/m2, lined kraft
writing paper of 56 g/m2 and newsprint of 49 g/m2. A comparison will
be made between sample test piece specimen size and displacement
rates. The matrix of experiments is listed in Table 2.1.
29
Table 2.1 Matrix of test piece size and deformation rates for
the tensile test
Test Piece Size Displacement rate (inch/min)
7 × 1” 1
7 × 1” 0.5
160 × 15 mm 1
160 × 15 mm 0.5
Results for a typical lightweight lined bleached white notebook paper

and a typical medium weight unbleached kraft linerboard are shown
in Figure 2.3 for the MD and Figure 2.4 for the CD. To compare
results for the different basis weight samples, the strength value in
N/mm results are divided by the sample basis weight so that the
results are presented as indices, hence the odd looking mixed units
of N-m2/g-mm. Tensile properties scale linearly with basis weight, so
dividing value by the basis weight is a standard practice to compare
samples of differing basis weights. For the case of the notebook
paper, the largest difference occurs when using the larger 1” wide
strip compared with the 15 mm strip.
0.085 9
56 gsm notebook paper
0.08 8.5
205 gsm kraft linerboard 56 gsm notebook
0.075
MD strength index (Nm2/mm-g)
MD stiffness index (Nm2/mm-g)
8
205 gsm kraft lin
0.07 7.5
0.065 7
0.06 6.5
0.055 6
0.05 5.5
0.045 5
0.04 4.5
15 mm & 15 mm & 25.4 mm & 25.4 mm & 15 mm & 15 mm
1 inch/min 0.5 inch/min 1 inch/min 0.5 inch/min 1 inch/min 0.5 inc
a)
30
Tensile Properties
9
per
8.5
oard MD stiffness index (Nm2/mm-g) 56 gsm notebook paper
8
205 gsm kraft linerboard
7.5
6.5
5.5
4.5
25.4 mm & 25.4 mm & 15 mm & 15 mm & 25.4 mm & 25.4 mm &
in 1 inch/min 0.5 inch/min 1 inch/min 0.5 inch/min 1 inch/min 0.5 inch/min
) b)
Figure 2.3 Tensile test results for MD strength (a) and stiffness (b)
for notebook and kraft linerboard papers using two different strip
widths and elongation speeds
0.04 3.5
3.3
205 gsm kraft linerboard 56 gsm
0.035 3.1
CD strength index (Nm2/mm-g)
CD stiffness index (Nm2/mm-g)
205 gsm
2.9
0.03 2.7
2.5
0.025 2.3
2.1
0.02 1.9
1.7
0.015 1.5
15 mm & 15 mm & 25.4 mm & 25.4 mm & 15 mm &
1 inch/min 0.5 inch/min 1 inch/min 0.5 inch/min 1 inch/min
a)
31
3.5
3.3
3.1
CD stiffness index (Nm2/mm-g)
205 gsm kraft linerboard

2.9
2.7
2.5
2.3
2.1
1.9
1.7
1.5
4 mm & 25.4 mm & 15 mm & 15 mm & 25.4 mm & 25.4 mm &
ch/min 0.5 inch/min 1 inch/min 0.5 inch/min 1 inch/min 0.5 inch/min
b)
Figure 2.4 Tensile test results for CD strength (a) and stiffness (b)
for notebook and kraft linerboard paper using two different strip
widths and elongation speeds
Sb are also lower when testing a 15 mm strip compared with the

25 mm width, approximately 23% lower in the MD and 37%
in the CD. Thus, testing a smaller width results in lower strength
and stiffness values. This may be expected on the basis of paper
non-uniformity commonly called formation, which is most easily
visualised when looking through a sheet of paper backlit by a
bright light, although radiographs are less prone to scattering effects
from fillers [11] and are more accurate representations of the mass
distribution in paper. However, the mottled appearance of typical
paper illuminated from the back is due to the clumping of fibres
during the wet-laid process and is described as a crowding factor [12],
which relates to the effect of longer fibres entangling others forming
clumps in suspension then later forming the flocs we see when the
suspension is drained. So floc size is dependent upon the longer fibres
32
Tensile Properties
in suspension and are typically approximately 4 to 5 mm in diameter.

The 15 mm strip width therefore may not be wide enough to cover
several flocs across its width and so produces a lower St value. The
effect of change in elongation rate from 0.5 to 1 inch/min appears
not to be as consistently significant for both 15 and 25 mm wide
strips. The expectation is that a lower elongation speed would result
in lower strength and stiffness values.
In summary, when interpreting tensile results, care should be taken

to note the test specimen dimensions and the elongation speed. This
becomes important when the subtle effects of different treatments
or chemical additives are evaluated using separate sets of data that
may have been obtained using different tensile testing parameters.
For laboratory convenience, testing 15 mm samples at the elongation
speed of 25 mm/min is convenient and the results are not significantly
different when using a 15 mm wide strip tested at the standard
12 mm/min speed.
2.4 Units, Breaking Length

Paper strength is always specified per unit width, e.g., N/mm. As
mentioned before, it is common practice to normalise St and Sb values
by the basis weight ‘β’ to account for the linear increase of St with
basis weight. This way, the effects of fibre furnish or species changes,
strength additives or filler content can be assessed independently
of basis weight. Another common equivalent method is to also use
breaking length ‘BL’, which is the calculation of the length of paper
that would cause it to break under its own weight. Since the strength
of force to failure ‘F’ in this case is given by Newton’s law, F = mg,
we have breaking length defined as:
tensile strengh (N)

BL = F/w = (2.7)
gb kg
sample width (m) # basis weight c m # 9.8 m2
m2 s
33
which using the usual units of St , ‘F’ in N/mm, basis weight in

g/m2 becomes:
F ^ mm
N
h
BL ^km h = 102 # g (2.8)
b^ m h
2
2.5 Summary
St is perhaps the simplest yet most fundamental of the physical tests
for paper. A tensile test produces the strength stiffness and work
values for a sample. Strength and stiffness relate to the E of the sheet,
which is principally governed by the fibre quality for a given basis
weight and sheet density. Although the test method is comparatively
devoid of testing artefact and inherent high variability associated with
any other tests, tensile results are dependent on the test parameters
such as sample size, sample orientation and elongation rate. The
capability of obtaining the Sb and in turn the E for the two principal
directions, MD and CD, allow the inference of other mechanical
properties of paper.
References
1. R.S. Seth and D.H. Page in The Role of Fundamental
Research in Paper Making, Mechanical Engineering
Publication, London, UK, 1983, p.421.
2. TAPPI T 494: Tensile properties of paper and paperboard

(using constant rate of elongation apparatus), 2016.
3. R.W. Perkins in Handbook of Physical Testing of Paper, Eds.,

R.E. Mark and J. Borch, CRC Press, Boca Raton, FL, USA.
2001, 1, 2.
34
Tensile Properties
4. P. Makela and C. Fellers, Nordic Pulp and Paper Research

Journal, 2012, 28, 2, 352.
5. A. Hagman and M. Nygards, Nordic Pulp and Paper

Research Journal, 2012, 28, 2, 295.
6. TAPPI T 231: Zero span breaking strength of pulp (dry zero-

span tensile), 2016.
7. W.J. Batchelor, B.S. Westerlind, R. Hagglund and P. Gradin,

TAPPI Journal, 2006, 5, 10, 3.
8. D.H. Page, TAPPI Journal, 1969, 52, 4, 674.
9. R.C. Moody and J.W. Koning, USDA Research Note FPL-1212,

April, 1966.
10. R.E. Popil, BioResources Journal, 2012, 7, 2, 2553.
11. D.S. Keller and J.J. Pawlak, Journal of Pulp and Paper
Science, 2001, 27, 4, 117.
12. R.J. Kerekes, and C.J. Schell, Journal of Pulp and Paper
Science, 1992, 18, 1, J32.
35
36
3
Ultrasonic Testing of Paper
3.1 Introduction
One of the most useful tests for detecting the changes in relative sheet
quality is sonic propagation through and along sheets. Rediscovered
for paper applications in the 1960s [1], the technique is often
underutilised and overlooked for paper characterisation in favour of
mechanical measurements. A good review of ultrasonic measurements
and applications for paper can be found in Waterhouse [2]. A major
selling point for ultrasonic measurements is that it is comparatively
quick, requires no specifically prepared sample test piece size and is
non-destructive. These features are very useful in a testing laboratory,
allowing first pass screenings of sample sets which can be followed by
standard mechanical testing to ascertain any differences that may be
of interest to investigate such as the effects of various pulp treatments
or stock additives to enhance properties.
Many mechanical properties of a paper sheet are affected by the

paper elastic modulus ‘E’. The discussion regarding the mechanics
of paper as an orthotropic solid can be extensive, however [3], for
the purposes of quality testing related to mechanical properties of
interest, the immediate concern here will be sonic propagation in
the three principal directions of machine-made paper: the machine
direction of machine-made paper (MD), the machine direction of
paper (CD) and the out-of-plane z direction of paper (ZD). The main
principle that is useful for the sonic testing of paper is the relationship
between paper modulus and the speed of sound, ‘V’, as longitudinal
waves can be simply given by the approximation in Equation 3.1:
E , tV 2 (3.1)
37
where the ‘apparent density’ of the paper test sheet is the basis weight
divided by its caliper, preferable and more accurately, it is the ‘soft-
platen’ caliper, for reasons that will be explained shortly.
The modulus relationship in Equation 3.1 becomes approximately

10% more accurate with the inclusion of Poisson constant ratio terms,
which are calculated [4] using the measurements of in-plane shear
sonic wave propagation and results from orthotropic solid mechanics.
However, in the author’s experience, the added complexity in doing
so does not provide a useful advantage.
As sonic propagation in paper is directionally dependent, as is the

modulus, both ‘E’ and ‘V’ are written with suffixes 11, 22 or 33
corresponding to MD, CD and ZD, respectively. Typically, for many
papers commonly encountered, ‘E’ and ‘V’ are smaller in the CD
compared with MD by approximately 1.5 times or more, and ZD
values are approximately 30 times or more lower than either MD
or CD values. The ratio of MD to CD values is attributable to the
combination of fibre orientation and drying stresses acquired during
paper manufacture [5–7]. The much lower ZD values [8] reflect the
compressibility and bonding level of the fibre layers in the paper
structure.
The speed of sound in paper is most commonly measured using

pairs of bimorph transducers placed lightly onto the surface of the
test sheet. These are comprised of a metal paddle, a few millimetres
wide, which is made to vibrate when an alternating voltage, of typical
frequency 80 kHz, is applied to the piezoelectric crystals adhered
on either side of the paddle. The paddles vibrate perpendicularly to
their width producing a longitudinal sound wave that propagates
along the surface of the sheet. Another transducer is used to detect
the transmitted sound wave. Sound speed is calculated from the
known distance of the spacing between the transducers divided by
the time of travel of the sound wave pulse determined electronically.
Commercial instruments, the Lorentzen & Wettre (L&W) tensile
stiffness orientation (TSO) [9], Nomura Shoji SST 250 [10] or
the SoniSys usually report the in-plane velocity squared, which is
38
effectively ‘E/ρ’, known as the specific modulus or tensile stiffness

index (TSI).
A popular application of sonic in-plane testing exploits the sensitivity

of the method to the MD/CD ratio [11]. Paper machines are often
several metres wide so that the web, in varying states of consolidation,
is under tension. This, along with many open draws in the machine,
all lead to a cross-machine reel profile in the MD/CD ratio in the
resulting dry paper at the end of the reel. A highly varying MD/CD
ratio of edge rolls, compared with rolls selected from nearer the
centre of the machine, cause runnability issues once the paper is run
through a printing or converting process [12], therefore it is desirable
to minimise the MD/CD profile across a reel through a programme of
iterative adjustments of headbox stock flows, stock jet to wire speed
ratios, open draw tensions, drying strategies and so on.
An example is provided below where cross-reel strips from a

fourdrinier paper machine producing linerboard were submitted for
profile analysis. The cross-reel strips were supplied as rolls that were
3 ft or longer in the MD, crudely cut with a knife by walking along
the length of the jumbo reel at the end of the paper machine. The
strip is laid out flat on the floor, the edge of the roll is assumed to
be the actual MD, perpendicular lines to the edges are drawn across
the roll and test strips are cut across the reel at regular intervals. It is
important to realise in profile studies that there is a variation in MD
measurements as well as a CD. The objective is to observe whether a
significant profile exists in the CD. In Figure 3.1 below, a L&W TSO
tester [9] was used to make four measurements in the MD at various
successive positions across the reel. This instrument consists of a
circular array of transducers that instantly provide measurements of
in-plane V2 upon electronic activation of alternating opposing pairs.
39
2.90
42# Reel 11 4/13/15
2.80
42# Reel 10 10/1/14
2.70
TSI MD/CD ratio
2.60
2.50
2.40
2.30
2.20
0 2 4 6 8 10 12 14 16
Position from front edge (ft)
Figure 3.1 MD/CD stiffness profile for the 42# samples. The
‘typical smile’ MD/CD profile seen here is largely due to edge
shrinkage and some edge flow. Error bars are standard deviations
of 4+ measurements
The quantity of interest in reel strip profile measurements is the ratio

of VMD2/VCD2, the MD/CD specific stiffness ratio. Figure 3.1 shows
the typical dip in the middle of the cross-reel strip that is seen in many
fourdrinier-style paper machines that are not optimised to have a flat
profile. The lower ratio in the middle, around position 8, is caused
by the combination of stock edge flows on the forming fabric and
drying shrinkage stresses [11]. The paper web in most paper machines
endures many open draws while under MD tension and so inevitably
shrinks in the CD during the manufacturing process. The contribution
of edge flows may be assessed by examining the stiffness orientation
40
angle [13], which is the clockwise measured angle of the maximum

V2 with respect to the assumed MD from the sample strip edge.
Orientation angle profiles in Figure 3.2 show a slight rise or dip

in the middle of the machine (positions 6 through 8) indicating an
opportunity to flatten the profile by adjusting the stock to wire speed
ratio on the paper machine, the so-called rush/drag ratio. Orientation
is measured clockwise with respect to the assumed direction of the
true MD, which is 0°. Negative values indicate or orientate towards
the paper machine operator or control side, whereas positive values
indicate or orientate towards the drive side. An overall negative
profile suggests that there is too much recirculation flow causing a
skewed overall flow towards the operator side of the paper machine.
2
TSO angle (º)
0
0 2 4 6 8 10 12 14 16
-2
-4
-6
42# Reel 11 4/13/15
-8
42# Reel 10 10/1/14
-10
Figure 3.2 Corresponding TSO angles for the reel strips shown in
Figure 3.1
41
3.2 ZD Ultrasonic Measurement

Speed of sound through the test sheet is used to determine the
ZD modulus [14]. In this case, the transducers are 1 MHz driven
piezocrystals coupled to plastic contacting delay blocks ending in
neoprene sheets that contact either side of the sheet under a pressure
of 50 kPa, the standard measurement, or 20 kPa for towel or tissue
sheet measurements. A close-up of the transducers of one commercial
development, the SoniSys instrument, is shown in Figure 3.3. This
shows the central top and bottom ZD transducers which have black
neoprene rubber contacting tips with concentric pairs of in-plane
transducers behind. These transducers are made to rotate around the
axis of the ZD transducer axis and drop down, via actuators, to take
in-plane measurements at various angles on the test sheet. The action
of the transducers is shown in Figure 3.4. When the orientation of
the paddles are edge to edge, the propagation of the waves that are
detected are shear waves. This has the rather esoteric application
of calculating the in-plane Poisson ratios when the measurement is
made at 45° with respect to the MD of the sheet [4].
Figure 3.3 Close-up photograph of the transducer contacting ZD

and in-plane probes of the SoniSys instrument in their retracted
vertical position. The test sheet is placed onto the back rubber mat
42
a) b) c)
Figure 3.4 Transducers in action in the SoniSys instrument. a) The

ZD transducer is in contact with the sheet, b) the ZD transducer
is retracted and the in-plane transducers are placed close to one
another making a measurement and c) the in-plane transducers are
taking a corresponding measurement spaced further part
The distance that is used to calculate the speed of sound through

the sheet (ZD) is the sheet thickness or caliper. Paper caliper is
affected by the combination of the sheet compressibility and surface
roughness. For embossed towel and tissue, experimental evidence
shows that caliper does not decrease appreciably once the measuring
platen caliper is at 20 kPa or lower [15]. For dense printing, writing
papers, the higher pressure of 50 kPa provides a measurement that
correlates closely with the thickness of a stack of sheets [16]. Since
paper is non-uniform on the scale of the longest fibres, a circular
platen diameter is chosen to cover several flocs or visible clumps of
paper, 20 to 30 mm is typical. Surface roughness is compensated by
using soft rubber covers on the platens [17]. The rubber conforms to
the surface undulations and provides a more accurate measurement
of the caliper. This soft-platen measurement of caliper is known to
provide what is termed the mechanical equivalent, or effective caliper,
as shown by Setterholm [18].
The significance is that the soft-platen caliper corresponds to the

same caliper ‘t’ that can also be calculated from first principles of
mechanics, namely that the elastic modulus ‘E’ is related to the tensile
stiffness ‘Sb’ as:
43
S b = Et (3.2)
Bending stiffness ‘D’ is the resistance of a sheet to bending. In the

case of two-point beam bending, a specimen of a prescribed width
and length is secured at one end and the force required to deflect
the sheet test piece to a prescribed angle is measured using bespoke
commercial testing instruments. The most common form of the
instrument is the Taber version which measures bending moment
[19]. Linear elasticity theory derives D in this circumstance to be:
D = Et
3
12 (3.3)
Combining the equation for D and Sb, the caliper that works is then:
t= 12D
Sb (3.4)
Since the objective of ultrasonic measurements is to have

measurements that correlate with physical properties, the ‘t’ used in
the calculations to convert sound speed to modulus is the soft-platen
caliper [20], which is usually less than the caliper measured by hard
platens by a few per cent.
3.3 Correlation of Ultrasonic Results with Mechanical

Properties
Ultrasonic measurements conveniently correlate to many measured
mechanical properties and in principle can be used as a quality check.
The fundamental relationship that is used in this regard is between
44
the Sb and the specific stiffness, or sound speed squared, and sheet
basis weight ‘β’:
(3.5)
Many strength properties are related to the Sb. For example, as shown
by Seth and Page [21] for well consolidated papers, the stress–strain
curve shape is largely dependent on the modulus, as determined
by the fibre quality, and only the strength values are affected by
approximately 25%, as determined by the degree of fibre bonding.
Therefore, sonic propagation can be used to assess sheet quality for
a given sample set most easily by multiplying the basis weight of a
sheet with the ultrasonically measured specific stiffness. In Figure 3.5
below, a variety of paper and plastic film samples were tested for Sb
using a universal testing machine and also for in-plane ultrasonic
specific stiffness using a L&W TSO unit. There is a convincing
correlation between the sonic calculated stiffness βV2 and mechanical
Sb. Typically, ultrasonic measured equivalents of physical constants
are 30–50% higher than mechanically measured counterparts because
of the viscoelastic nature of paper.
Many strength properties are dependent on ‘E’ and correspondingly

the Sb, so that ultrasonics may be used instead of mechanical testing
for a given sample set. In one example, unbleached softwood kraft
pulp handsheets were prepared from slurries having undergone
various dry strength additive treatments, the objective being to
evaluate their effects in terms of standard short-span compression
test or strength (SCT) [23] and out-of-plane crush [corrugated
medium test (CMT)] [24]. The latter is a common qualifying test
for corrugated board medium grades measured on test strips which
are first sent through a laboratory corrugator to become fluted, the
fluted strip is affixed to firm backing tape and then crushed out-of-
plane on a compression tester. CMT is considered to be relevant to
45
the flat crush resistance of corrugated board in shipping containers

and SCT is the major contributor to the compression strength of
corrugated board and boxes. The results are shown in Figure 3.6,
where a good correlation of the physical results with mechanical
values for compression strength is apparent.
1400
1200
TSl_CD x β (N/mm)
y = 1.75 x -73.30
R2 = 0.98
1000
800
600
400
200
100 200 300 400 500 600 700 800
Mechanical Sb (N/mm)
Figure 3.5 Results from a lab study using a L&W TSO instrument
on a variety of paper samples and comparing the results to
standard mechanical Sb. Reproduced with permission from R.E.
Popil in the Proceedings of TAPPI PaperCon 2010, 2–5th May,
Atlanta, GA, USA, 2010. ©2010, TAPPI [22]
46
65
CMT = 53.21 × TSO Sb -8.82
R2 = 0.95
55
CMT (lb/6 in.) or SCT (lb/in/)
45 CMT
SCT
35
SCT = 17.72 × TSO Sb + 0.73
R2 = 0.97
25
15
0.9 1 1.1 1.2 1.3 1.4
TSO Sb (N/mm)
Figure 3.6 A comparison of out-of-plane corrugated strip crush

results with Sb calculated as basis weight × VCD2
Units for CMT and SCT values in Figure 3.6 have been kept in
standard popular English formats of lb/6 inches for CMT and lb/in.
for SCT. The relationships shown in Figure 3.6 indicate that ultrasonic
measurement can predict mechanical performance in a sample set
of similar grades.
A range of linerboard and furnish mediums from southeast region

USA mills were analysed for SCT and ring crush test (RCT) [25]
compression tests. RCT is similar in concept to SCT, but also contains
a component of bending failure [22], as it involves the compression of
a ring structure as the test piece. Again, as shown in Figure 3.7, the
mechanical strength results plotted against the CD TSO Sb obtained
from the ultrasonics measurements, demonstrate good correlation.
Note that the linear relationship between SCT and ultrasonic stiffness
for the machine-made liner and medium set represented in Figure 3.7
47
is different to the handsheet sample set shown in Figure 3.6. Note

also that the RCT is lower in value than SCT although both are
compressive strength. RCT being lower than SCT is attributable to
RCT having bending of the test piece structure contributing to its
failure. The advantage in using ultrasonics to test sheet quality is that
no sample cutting or preparation is required as long as the sample
can be placed into the testing instrument. The variation in ultrasonics
measurements is often smaller, approximately 3%, compared
with 7 to 10% variation for many mechanical measurements.
The disadvantage is that the correlation between elastic sonic
measurements and inelastic mechanical strength measurements varies
if the composition of the sheets changes markedly via substantial
changes in furnish or additives. Therefore, if used for routine quality
screening, the correlation between ultrasonic stiffness and mechanical
measurements should be checked whenever changes are made to the
product under evaluatation.
60
SCT = 27.49 (TSI_CD × β) - 1.93
R2 = 0.93
50
SCT
40 RCT/6
Linear (SCT)
SCT or RCT (lb/in)
Linear (RCT/6)
30
20
10
RCT = 17.73 (TSI_CD × β) - 2.58

R2 = 0.98
0
0.1 0.6 1.1 1.6 2.1
TSO Sb (kN/m)
Figure 3.7 SCT and RCT for a selection of linerboard and

corrugating medium of varying basis weights versus ultrasonic Sb
48
3.4 Application of ZD Ultrasonics

In the out-of-plane ZD, sonic propagation is affected by the degree
of bonding between fibres [26]. Debonder additives are applied to
towel or tissue products in an attempt to improve bulk, increase
absorbency and apparent softness [27]. ZD sonic measurements are
useful in assessing the potential efficacy of debonder treatments. In
Figure 3.8 below, a series of standard laboratory handsheets were
prepared using various levels of a conventional debonder ‘A’ and
two experimental debonders ‘B’ and ‘C’ at conventional application
levels of 1.5, 3 and 5 lb/tonnes dry weight of fibre mixed into the
stock slurry of a British Standard Sheet Machine handsheet former.
220
200
High-pressure platens
Low-pressure platens
180
ZD modulus (MPa)
160
140
120
100
Base A 1.5 B 1.5 A 3.0 B 3.0 A 5.0 C 5.0
lb/ton lb/ton lb/ton lb/ton lb/ton lb/ton
Sample ID
Figure 3.8 Ultrasonic ZD measurements at two different platen

pressures for a series of handsheets prepared with differing levels
of various debonders
The vertical axis in Figure 3.8 shows the ZD modulus obtained using
SoniSys equipment, where the time taken for a 1 MHz longitudinal
49
wave to traverse the soft-platen caliper along with the measured

basis weight produces the ZD modulus. There are two sets of ZD
modulus data shown in Figure 3.8. One set is taken using a caliper
pressure of 50 kPa and the other at a pressure of 20 kPa. The lower
pressure produces less compression of the sheet so that overall, lower
ZD moduli are obtained compared with the corresponding values
obtained at the higher pressure. Both sets of data indicate that relative
to the base sheet, which has no debonding additive, agent ‘B’ applied
at the 3 lb/tonnes level can be expected to be the most effective
treatment. Indeed, the decrease in ZD modulus was found to be
accompanied by a significant measured increase in surface roughness,
shown in Figure 3.9, and also a corresponding significant increase in
compressibility, as measured by the difference in caliper at 50 kPa
to that at 20 kPa. The ZD attenuation (loss of signal strength in dB
from transmission) in the case of 20 kPa caliper pressure showed a
good correlation with the measured Sheffield roughness.
15
14.5
14
13.5
ZD attenuation (dB)
13
12.5
12
11.5
50 kPa platen pressure
11
20 kPa platen pressure
10.5
170 120 270 320
Sheffield roughness (mL/min)
Figure 3.9 ZD attenuation versus Sheffield roughness for TAPPI

handsheets with varying debonder dosage
50
Note that the 50 kPa ZD attenuation results show no correspondence

with the increased surface roughness, suggesting that the 50 kPa
caliper pressure suppresses the surface roughness compared with
20 kPa.
For this sample set, i.e., exploring the potential effects of debonding
agents, no other significant differences were measured in strength
properties, indicating that the ultrasonic measurements were actually
more sensitive in detecting the expected effects of the chemical
additives.
Since ZD sonic propagation is dependent on the coupling between

fibres it can, in principle, detect tissue softness which is defined
as the perception of softness by tactile feel using the thumb and
fingertips of the hand. This is a contentious topic in itself, addressed
by several researchers in the past and summarised by Hollmark [28].
The consensus is that softness comprises a combination of physical
attributes such as low D, low surface friction, high compressibility,
and a microscale roughness attributable to loose fibre ends on the
surface. In sonic propagation, loose fibres on the surface, along with
loose interfibre contacts in the absence of bonding, compromise an
impedance mismatch leading to attenuation of the sonic signal. The
attenuation A (dB) of the signal can be measured as a method to
predict the softness using the multiple linear regression of the form:
Softness = aZ + bA/b + c b (3.6)
where ‘a’, ‘b’ and ‘c’ are empirical coefficients 247.6, 19.1 and 0.82,
respectively, or others as may be obtained from regression analysis,
‘Z’ is the impedance which is defined as the density multiplied by
the sound velocity or equivalently, Z is also the basis weight ‘β’
divided by the time of flight of the signal, ‘A’ is the attenuation of
the transmitted signal which is determined by Fourier analysis of the
transmitted and received signals. An example of this application is
51
shown in Figure 3.10 where a selection of commercial paper towels

were evaluated.
30
Softness
30
Handle
30
Attenuation
30
30
30
30
30
30
30
Good Best Typical Bad
Towel paper by apparent feel
Figure 3.10 Comparison of the softness formula values, handle

measurement and the attenuation for a towel sample set
Softness is calculated from the measurements using the regression

formula and comparing the results with a tissue handfeel
measurement, the so-called ‘Handle-o-meter’, and ZD attenuation.
‘Handle’ is measured as the force in grams required to insert the
tissue with a sheet metal edge into a 20 mm or smaller slot and is
also attributed as ‘Drape’ [29]. A lower value of handle is expected
to reflect a softer feeling tissue product. For this sample set of paper
towels, the distinction of softness, as calculated by the regression
formula, is not significant, the handle better reflects the differences
52
in basis weights of the sample set, however, the attenuation of the

papers correlate well with the apparent feel of the sheets.
3.5 Summary
The speed of sound in paper is readily measured electronically
using contacting transducers in commercially available equipment.
The relationship between the velocity of sound squared and elastic
modulus provides a convenient quality check of paper test samples
that is often related to end-use physical properties of interest, such
as tensile or compression strength. Measuring the directionality of
the sound speed squared across paper machine-wide strips permits
profile optimisation through iterative adjustments of the paper
machine headbox stock flows and stock jet to forming wire speed
ratios. Speed of sound through the sheet is affected by the level of
intrafibre bonding and quality of the contact with the contacting
transducers. This allows the potential of the measurement of sound
speed to discern the effects of pulp stock chemical additives and can
also be applied to the relative measurement of paper tissue or towel
softness.
References
1. J.K. Craver and D.L. Taylor, TAPPI Journal, 1965, 48, 3,
142.
2. J.F. Waterhouse, TAPPI Journal, 1994, 77, 1, 120.
3. R.W. Perkins in Handbook of Physical Testing of Paper, Eds.,

R.E. Mark and J. Borch, CRC Press, Boca Raton, Fl, USA,
2001, 1, 2.
4. R.W. Mann, G.A. Baum and C.C. Habeger, TAPPI Journal,

1980, 63, 2, 163.
53
5. T.R. Hess and P.H. Brodeur, Journal of Pulp and Paper

Science, 1996, 22, 5, 160.
6. D.W. Vahey, J.M. Considine, A. Kahra and M. Scotch in

Proceedings of Progress in Paper Physics Seminar, 2–5th June,
Otaniemi, Finland, 2008, p.2.
7. M. Titus, TAPPI Journal, 1994, 77, 1, 127.
8. C.C. Habeger and W.A. Wink, Journal of Applied Polymer

Science, 1986, 32, 4, 4503.
9. G. Lindblad and T. Furst in The Ultrasonic Measuring

Technology of Paper and Paperboard, AB Lorentzen and
Wettre, Kista, Sweden, 2001.
10. T. Nomura, Japan TAPPI Journal, 1994, 48, 1, 215.
11. S. Loewen and W. Foulger, Pulp and Paper Canada, 2002,

103, 5, 42.
12. J. Paukku, M. Parola and S. Vuorinen in Proceedings of the

International Printing and Graphic Arts Conference,
4–6th October, Vancover, Canada, 2004, p.245.
13. M. Santos, J. Perdigo and J. Velho, Insight – Non Destructive

Testing and Condition Monitoring, 2007, 49, 3, 146
14. C.C. Habeger, W.A. Wink and M.L. Van Zummeren, Institute
of Paper Chemistry Technical Paper Series, 1988, August,
No.301.
15. D.S. Keller, C. Feng and Y. Huang in Proceedings of the

International Paper Physics Conference, 10–14th June,
Stockholm, Sweden, 2012, p.32.
16. TAPPI T 411: Thickness (caliper) of paper, paperboard, and

combined board (Revision), 1997.
54
17. W.A. Wink and G.A. Baum, Institute of Paper Chemistry

Technical Series, 1983, March, No.133.
18. V.A. Setterholm, TAPPI Journal, 1974, 57, 3, 164.
19. TAPPI T 489: Bending resistance (stiffness) of paper and

paperboard (Taber-type tester in basic configuration)
(Revision), 1992.
20. TAPPI T 511: Thickness of paper and paperboard (soft-platen

method) (Revision), 2002.
21. R.S. Seth and D.H. Page in The Role of Fundamental

Research in Paper Making, Mechanical Engineering
Publication, London, UK, 1983, p.421.
22. R.E. Popil in Proceedings of the Tappi Paper Conference,

2–5th May, Atlanta, GA, USA, 2010.
23. TAPPI T 826: Short span compressive strength of

containerboard, 2013.
24. TAPPI T 809: Flat crush of corrugating medium (CMT test),

2011.
25. TAPPI T 822: Ring crush of paperboard (rigid support

method), 2011.
26. Y.C. Pan, C. Habeger and J. Biasca, TAPPI Journal, 1989,

72, 11, 95.
27. J. Liu and J. Hsieh in Proceedings of the TAPPI Papermakers

Conference and Trade, 16–19th April, Vancouver, Canada,
2000.
28. B.H. Hollmark in Handbook of Physical and Mechanical

testing of Paper and Paperboard, Marcel Dekker, New York,
NY, USA, 1984, 1, 497.
55
29. O.C. Hansen, L. Marker, K.W. Ninnemannn and O.J. Sweeting,

Journal of Applied Polymer Science, 1963, 7, 3, 817.
56
4
Bending Stiffness of Paper
and Corrugated Board:
The Connection to Caliper and
Tensile Stiffness
Generally, the resistance of objects to a mechanical action such as

applied stress is called stiffness. We have already encountered tensile
stiffness (Sb) as the reaction or resistance of paper to tension. Similarly,
the resistance of sheets or boards to bending action is termed bending
stiffness(es) (D). D ties in the Sb and the caliper of the board or sheet
and can, in principle, be calculated from tensile test measurements
and caliper. D of paper is important to specify the paper rigidity
required for printed tickets, business cards, folding cartons and in
converting processes such as printing or folding. Conversely, low D
is desired for towels, tissues and non-wovens where conformability
or drape of the sheet products are desired features. D in corrugated
boards limits the outward bulging of the side panels of boxes under
load and contributes to the compression strength of boxes.
4.1 The Bending Elasticity Theory

The simple elasticity theory provides the formula for the D of a one-
dimensional beam [1], which we can apply to the sheet of paper per
unit width:
D = Et # t
2
12 (4.1)
where Et, the elastic modulus ‘E’ multiplied by the effective caliper
‘t’ [2], is the Sb obtained from the tensile test. The effective caliper
57
removes the surface roughness contribution to roughness by using

caliper-measuring platens in the standard gauge which are covered
with soft neoprene. Measurement of paper D is achieved by applying
a moment, the simplest example is the two-point application where a
length of sheet is fixed at one end and a force is applied to the other
end. The deflection that the test specimen undergoes in reaction to
the force is related through the elasticity theory. For the two-point
bending of a beam of length ‘L’ and width ‘w’, standard Euler
mechanics elasticity analysis [1] provides the relationship:
60cF ^Nh L2 ^mm 2h

D ^mN - m h = (4.2)
rw ^mm hac

between the force ‘F’ subjecting a beam to a circular arc deflection of

‘α’ degrees. The commonly used Taber style of instruments measure
the force required to achieve 15° of deflection of a 38 × 50 mm test
specimen. This feature of Taber instruments is sometimes a source
for confusion [3] as it reports the bending resistance in terms of the
bending moment ‘M’, and not the D. The relationship between the
moment and D is:
D = ML (4.3)
3aw
where ‘L’ is the span (50 mm), ‘w’ is the width (38 mm) and ‘α’ is
the bending angle in radians (15°). It is useful to apply this formula
to convert from Taber moment units ‘M’ in grams force-cm to D in
mN-m, which is given below:
58
Bending Stiffness of Paper and Corrugated Board: The Connection to
Caliper and Tensile Stiffness
grams force-cm (4.4)
The formulas here all assume that beam strains are in the linear elastic
region of the stress–strain curve for the material. This requires the
assumption that the deflection is small, the moment point curves the
beam around a circular arc and that the beam is long relative to its
thickness, such that out-of-plane shear strain is negligible. Therefore,
when testing samples, consideration must be given to limits in sample
size and deflection to ensure linear elasticity.
The Taber [4] and Lorentzen and Wettre (L&W) bending resistance
measuring instruments are shown in Figure 4.1. Both instruments
typically use a 38 × 50 mm test specimen prepared using the punch
cutter shown in Figure 4.2. The length of the test specimen is cut
in the direction of interest to be measured, either in the machine
direction of machine-made paper (MD) or the machine direction of
paper (CD) for machine-made papers. The Taber turns the mounted
sample, which is subjected to a counterweight, until a 15° deflection is
attained. Different counterweight arrangements are applied according
to a table that ascribes the counterweights required for the range of
bending resistance being measured. Similarly, but simpler in design,
the L&W instrument bends a clamped end of the test specimen
through a rotation of 5° and measures the resulting force at the
other end by a contacting load cell. Different specimen lengths and
deflection angles can be selected for both instruments depending on
the stiffness of the sample.
59
a) b)
Figure 4.1 The Taber a) and L&W b) two-point D testers

commonly used in the paper industry to measure the bending of
board and paper samples
Figure 4.2 Both the Taber and the L&W bending measuring
instruments use a punch cutter that prepares 50.2 × 38.1 mm test
specimens, the length being along the MD or CD of the sheet
60
4.2 A Comparison of Different Methods of Measuring

Bending Stiffness
An example of measuring the D of several samples and their
relationship to Sb and soft and hard calipers follows. Seven different
samples were tested and Table 4.1 displays the designation, grade,
basis weight, soft caliper and density for each sample used in this
study.
Table 4.1 List of the various samples used in this Sb study,

basis weights, soft caliper and density
Sample Description Basis Soft Density
ID weight caliper (kg/m3)
(g/m2) (μm)
A 42# Brown kraft linerboard 212 277 766
B 26# Neutral sulfite semichemical 130 191 679
medium
C Lightweight bleached kraft 75 92 823
D Newsprint 45 57 808
E Lightweight coated 47 44 1057
F Mylar transparency film 146 98 1500
G Synthetic paper 156 97 1610
Reproduced with permission from C.G. Carson and R.E. Popil, TAPPI
Journal, 2008, 7, 12, 17. ©2008, Technical Association of the Pulp and
Paper Industry [4]
These samples were used to acquire hard caliper, soft caliper, a ‘stack’
caliper, and the MD and CD calculated effective thicknesses. The hard
and soft calipers were based on the Technical Association of the Pulp
and Paper Industry (TAPPI) methods T 411 and T 551, respectively.
The hard caliper was measured on an Emveco 200A and the soft
61
caliper on an Emveco 210-DH caliper measuring instrument. The

‘stack’ caliper was measured by stacking 12 specimens of a sample
and dividing the result by 12 (results of the measurements are given
in Figure 4.3). The expectation of TAPPI T 411 is that the hard
caliper result is a prediction of the thickness of a stack of specimens.
0.35
Hard caliper
0.30
Soft caliper
0.25
Caliper (mm)
Stack average
0.20
0.15
0.10
0.05
0.00
A B C D E F G
Sample ID
Figure 4.3 A comparison of hard and soft calipers for a range of

samples listed in Table 4.1. Reproduced with permission from
C.G. Carson and R.E. Popil, TAPPI Journal, 2008, 7, 12, 17.
©2008, Technical Association of the Pulp and Paper Industry [3]
The comparison of hard and soft calipers show the largest differences
occur for samples of a low density, such as samples A and B
(Table 4.1), which are also rougher surfaces compared with synthetic
(sample G) or coated magazine journal paper (sample E).
62
The effective mechanical equivalent thickness ‘teq’ was also calculated.

This parameter, proposed by Setterholm, calculates the thickness from
the relationship between St and Sb, which is derived as:
S b = Et
D = Et t eq 2
12
t eq = 12 D
Sb (4.5)
where ‘E’ is the elastic modulus.
The tensile measurements for the samples in Table 4.1 were made
on an Instron 1122 universal test machine according to the TAPPI
T 494 method. The tensile measurements provide values for Sb by:
S b = dF # L
dx max w (4.6)
which is the Instron Series IX™ software St slope algorithm. ‘L’ is the
gauge length (178 mm), ‘w’ is the width of the sample (25.4 mm),
and ‘te’ is the soft caliper measured and entered separately. D were
measured on the L&W instrument using the default 5○ deflection.
Figure 4.4 shows that the effective thickness, as calculated by the

stiffness to D ratio, matches the measured caliper using the soft-
platen method. Therefore, the soft-platen caliper can also be used
to calculate the D once the Sb is known, as will be shown in the
following. If we measure D using a different parameter or different
instruments how can we know which is right? Table 4.2 shows the
D of the variety of samples described measured using three different
methods.
63
0.3
Effective thickness (mm)
teq = 0.99 te + 0.0054

R2 = 0.99
0.2
0.1
0
0 0.05 0.1 0.15 0.2 0.25 0.3
Soft caliper (mm)
Figure 4.4 Comparison of the calculated effective Setterholm

thickness with soft caliper measurements. Reproduced with
permission from C.G. Carson and R.E. Popil, TAPPI Journal,
2008, 7, 12, 17. ©2008, Technical Association of the Pulp and
Paper Industry [3]
Table 4.2 Sb results using different methods

Sample MD CD
ID 5° L&W 30° L&W Taber 5° L&W 30° L&W Taber
A 9.235 8.106 8.932 5.051 4.466 4.800
B 2.278 1.966 2.012 0.891 0.812 0.799
C 0.358 0.352 0.333 0.144 0.141 0.144
D 0.102 0.099 0.101 0.016 0.017 0.017
E 0.065 0.057 0.058 0.031 0.030 0.027
F 0.431 0.469 0.465 0.417 0.437 0.448
G 0.393 0.409 0.398 0.360 0.390 0.384
Paper Industry [4]
64
Note, for example, that for the paper samples A–E, the stiffness
at 30° is smaller than at 5°. The angles and spans for the samples
were selected to obtain sufficient instrument sensitivity. The Taber
instrument tests were all performed with its default 15° deflection. We
will use calculations from the elasticity theory to check the accuracy
of D measurements, which requires the assumption that the deflection
is small, the moment point curves the beam around a circular arc
and that the beam is long relative to its thickness, such that out-of-
plane shear strain is negligible. These assumptions are not necessarily
true during standard D measurements, so their conditions should be
checked when invalidating a measurement. Stress/strain tensile test
curves show that strains below 0.2% were within the linear elastic
region of the curve. For the two-point method, the two-point bending
strain ‘ε’ can be estimated as:
f e = r # t e ac
120 # l (4.7)
Table 4.3 shows a comparison of the angles, spans and equivalent

strains used in the L&W measurements for the seven different
samples. These particular parameters were selected to optimise
accuracy. By comparison with stress/strain curves, we should expect
the stiffnesses at 5° to be higher than those at 30°, since the elastic
modulus at non-linear strains is lower than those at linear strains.
Lightweight grades cannot be measured with the L&W instrument
using default settings (50 mm span, 5° deflection), so large angle
deflections are required at the risk of underestimating the Sb from
non-linear strain.
65
Table 4.3 Listing of various test specimen Sb measurement

parameters: free spans, resulting corresponding strains at two
deflections and the Taber ranges used for the comparison of
results
Sample L&W spans (mm) Bending strains ε (%) Taber range
ID MD CD 5° deflection 30° deflection
A 25 25 0.14 0.87 10–100
B 20 20 0.13 0.75 1–10
C 15 15 0.080 0.48 1–10
a a
D 10 5 0.074/0.15 0.44/0.89 1–10
E 10 10 0.058 0.70 1–10
F 15 15 0.085 0.51 1–10
G 15 15 0.084 0.51 1–10
a
The two numbers here correspond to the MD 10 and CD 5 mm spans,
respectively.
Paper Industry [4]
Measurements of D listed in Table 4.2 were compared with calculated

D using the elastic modulus from tensile measurements by:
E = dF # L
dx max w # te
E t e3
c m c mE
vE 2 + 3 vt 2
D= ;1 !
12 E te (4.8)
using the Instron Series IX™ software modulus algorithm with ‘te’
the soft caliper measured and entered separately. Averages denoted by
brackets were calculated based on five or more repeat measurements
and the term in square brackets reflects the propagation of relative
errors based on the standard deviations denoted by ‘σ’.
Figure 4.5 compares the D of seven different papers measured using

three different methods, i.e., L&W 5° bend angle, L&W 30° bend
66
angle and Taber stiffness (15° bend angle), to the theoretical D

calculated using the measured elastic modulus.
10 10
L&W 5°
EMDt3/12 (mN-m)
L&W 5° L&W 30°
ECDt3/12 (mN-m)
1 L&W 30° 1 Taber
Taber
0.1 0.1
0.01 0.01
0.01 0.1 1 10 0.01 0.1 1 10
DMD (mN-m) DCD (mN-m)
Figure 4.5 The theoretical D plotted against the experimental

stiffnesses by three different methods: L&W 5°, L&W 30° and the
Taber instrument. The results from all three methods correlate well
to theory, including results of the lightweight grades. Reproduced
with permission from C.G. Carson and R.E. Popil, TAPPI Journal,
2008, 7, 12, 17. ©2008, Technical Association of the Pulp and
Paper Industry [3]
The results shown in Figure 4.5 indicate that the various measurements
are valid within experimental error when compared with the expected
results from calculations. Thus, D measurements can be checked using
the combination of tensile and soft caliper measurements utilising
the equations from the linear elasticity theory.
4.3 Bending Stiffness of Corrugated Boards

D is important for corrugated boxes [5], as the amount of panel
outward bulging that occurs during vertical load application
needs to be limited. The term flexural rigidity [6] is often used
67
for corrugated boards, although the relationship between flexural

rigidity and D involves a Poisson ratio term which is usually not
known nor measured for corrugated board. In the case of corrugated
boards, the so-called sandwich beam theory for D applies to a good
approximation:
D = E#t#h
2
2 (4.9)
where ‘E’ and ‘t’ are the modulus and caliper of the outside
linerboard and ‘h’ is the caliper of the board. In this case, hard caliper
measurements suffice. The equation indicates that the medium does
not contribute to the D [7] other than providing spacing between the
linerboard facings producing the board caliper ‘h’. More accurate
calculations [8] show that the contribution of the fluted medium to
D of the board structure is approximately 5%. The sandwich beam
approximation is also useful for consideration of increasing the D
of thicker paperboards, where greater gains are obtained when the
outside facing layers are made to have high Sb.
The arrangement for testing the D of corrugated boards uses the four-
point method shown schematically in the diagram of Figure 4.6. The
load is applied at the ends of the test specimen by weights designated
as ‘ P ’. The board is subjected to moments at either end from the
2
weights ‘ P ’ at a fixed distance, in this case 0.13 m, from the fulcrum
2
points designated as triangles. The deflection of the board ‘y’ is related
to the board’s D through the relationship:
^ 2Ph0.13 L 2
D=
8wy (4.10)
68
‘L’ is the length of the board between the pivot points, typically
0.2 m cut along either the MD or the CD, ‘w’ is the width which is
typically 0.1 m.
L
P P
2 2
y
0.13 m
Figure 4.6 Schematic of the principle of the four-point bending

arrangement for measuring D
The equipment is shown in Figure 4.7. The weights for ‘ P ’ are

2
supplied in incremental gram units.
Figure 4.7 Four-point bending instrument made by L&W
69
So, for typical situations with boards cut into 12 × 4 inch pieces,
the deflection ‘y’ panel read-out in mm, the convenient form of the
formula becomes simply:
(4.11)
For C-flute corrugated board, ‘ P ’ is 120 for CD and 220 for MD,
2
and deflections are approximately 1 mm or less providing values of
15 N-m for the MD and 8 N-m for the CD.
The four-point bending method eliminates the effect of shear in

the measurement [9] by the combination of the symmetry of load
application and clamping restraint of the board test piece. In the
alternative common three-point method [9], a test piece board
is instead supported at either end and is simply deflected in the
middle, and the load at the middlepoint is measured. For corrugated
board, MD shear develops mostly from a relative displacement of
the linerboard facings when the test pieces become bent. CD shear
for corrugated board is considerbaly restricted by the facings that
are adhered to the flute tips at the glue lines. Shear increases with
shorter test pieces, and experiments show that test pieces have to be
longer than 26” in the MD or longer than 10” in the CD to match
the higher D values obtained with four-point bending.
4.4 A Comparison of Three- and Four-Point Bending

Results for a Series of Boards
An example of the difference between four- and three-point bending
results is shown in Figure 4.8. In this case, a heavyweight single-
wall C-flute board was cut into lengths that were supported at two
points 14” apart and deflected about 2 mm at the centre using the
cross-head of a tensile testing machine. The cross-head was fitted
with a load cell that measured the resulting force ‘P’ caused by
70
deflecting the board. This provided the three-point D according

to the equation:
D = PL
2
48wy (4.12)
The same boards were tested using the four-point equipment. One
series of boards was taken directly off the corrugator and cut to size
on a computer-aided design (CAD) table, another series of boards
was sent through the converting slitter-scorer, which makes the flaps
of a box, and a third series consisted of samples cut from box blanks
made by the folder-gluer.
25
23 MD Four-point
CD Four-point
21 MD Three-point
19 CD Three-point
17
D (N-m)
15
13
11
9
7
5
CAD cut Scored board Box blank
Figure 4.8 Comparison of four- and three-point D for a

heavyweight C-flute corrugated board sample at three different
points in the converting process as indicated. Error bars represent
the 95% confidence interval of repeat measurements for each
average value
71
All converting stages can introduce some degree of crushing of

the board. In Figure 4.8, it is obvious that the D of the MD is
greater than the CD by a factor of approximately two, reflecting
the Sb orientation of the linerboard facings. The three-point D are
lower than the corresponding four-point values by 30% or more.
Boards that have been subjected to converting processes also show
significantly lower MD values than the CAD cut board. This can be
attributed to the effects of loss of shear stiffness caused by crushing
of the board, which can also be detected by a correponding change
in caliper measurement, as shown in Figure 4.9.
4.52
4.52
Caliper (mm)
4.47
4.42
4.37
4.32
CAD cut Scored board Box blank
Figure 4.9 Comparison of the caliper of singe-wall board samples

at different converting stages
Many manufacturers are interested in the possibility of replacing

heavyweight single-wall board with lightweight double-wall board.
72
The idea here is that the loss in compression strength from reduction
in basis weight can be compensated by an increase in D, as occurs
with the increase in caliper of double-wall board. A description of
the composition of the sample set are detailed in Table 4.4 – boards
A, H and I are single-wall, the rest are double-wall.
Table 4.4 Decription and composition of a test corrugated

board sample set
Sample ID and Basis weight of board components (g/m2)
description
Liner Medium Liner Medium Liner Total
weight
A – Heavyweight 337 112 337 – – 786
single-wall
B – Super lightweight 88 112 88 112 88 488
double-wall
C – Lightweight 98 112 98 112 98 518
double-wall
D – Lightweight B-B 98 112 98 112 98 518
but with kraft
E – Medium weight 127 112 127 112 127 605
double-wall
F – Heavyweight 151 112 127 112 151 653
double-wall
G – Heavyweight B-B 151 112 49 112 151 575
‘X’ flute
H – Heavyweight 274 112 274 – – 660
single-wall kraft
I – Medium weight 254 112 254 – – 620
C-flute with kraft
J – Medium weight 161 112 161 112 161 707
kraft B-B flute
B-B: Multi-walled board consisting of two B-flute medium layers
The four- and three-point D were compared for this sample set. The
four-point always measures higher than the three-point, particulary
73
in the MD, as Figure 4.10 shows. Some of the lighter linerboards

used in preparing samples B and C, were made using novel headbox
technology producing a sheet with no orientation, this results in the
D of MD and CD not being significantly different from each other
as is usually the case.
25
MD Three-point
20 CD Three-point
MD Four-point
15 CD Four-point
Sb (N-m)
10
0
l
te
ft
B
al
al
al
al
al
af
af
B-
ra
lu
w
w
kr
kr
lk
’-f
e-
e-
e-
e-
ft
e
ith
ith
ra
bl
‘x
gl
bl
bl
bl
al
in
tk
ou
ou
ou
ou
w
tw
w
-B
ts
e-
td
gh
td
td
td
tB
e
bu
gl
ut
gh
ei
gh
gh
gh
gh
in
gh
-fl
ei
w
t-
ts
ei
ei
ei
ei
ei
tC
w
gh
m
w
w
gh
vy
gh
iu
ht
ht
vy
ei
vy
m
ei
ea
ed
w
lig
ig
ei
ea
iu
ea
w
-H
ht
-L
M
ed
-H
-H
h
er
ig
ig
id
-M
J-
up
A
-L
-H
M
F
G
-S
I-
H
B
Figure 4.10 Comparison of three-and four-point D measurements

for the sample set of corrugated boards
D of corrugated board and loss of shear stiffness from crushing are

consequences of predicting the stacking strength [box compression
test strength (BCT)] of vertically loaded boxes. The McKee equation
for vertical BCT [10] is based on the analysis of the bending and
failure of panels, and results in the semiemprical formula in metric
form:
74
(4.13)
with the board edge compression strength in kN/m, the MD and

D of the board DMD and DCD in N-m, and the box perimeter (2 ×
{length+width}) Z in m. The geometric mean of the D term D MD D CD
is raised to the ¼ power so that a 10% change in D will affect BCT
only by approximately 2.5%. The change in the geometric mean D
for the set of boxes averages approximately 18% smaller between
three- and four-point measurements, thus BCT predicted using the
formula can be expected to be approximately 5% smaller when using
three-point D values.
Although all BCT measurements and predictions are within

experimental agreement, Figure 4.11 shows that when using three-
point D, the McKee formula becomes a more accurate predictor of
the actual BCT values. The mean difference between predicted and
actual BCT values is 367 N when using four-point values compared
with a difference of 181 N when three-point D is used instead. Since
boxes subjected to vertical load are not constrained to restrict MD
shear, three-point bending appears to be more relevant for a more
acurate prediction of BCT using the McKee formula.
The question remains what is the optimal three-point D test span

which is relevant as a predictor for a particular box. For long boxes,
shear stiffness should not matter since three-point D is about the same
as true four-point stiffness once the length exceeds 60 cm in the MD.
In fact, a formula to convert from three-point ‘D3’ measured at two
different lengths ‘L1’ and ‘L2’ to four-point ‘D4’, L1 > L2 is:
D 3,1 D 3,2 ^L 21 - L 22h

D4 =
D 3,2 L 21 - D 3.1 L 22 (4.14)
75
7.0 Actual BCT

Full McKee formula Three-point
6.0 Full McKee formula Four-point
BCT (kN)
5.0
4.0
3.0
2.0
l
ft
B
al
al
al
al
al
af
af
ut
B-
ra
w
-w
w
kr
kr
-fl
lk
e-
e-
ft
e
e
le
’
th
ith
ra
bl
‘x
gl
bl
l
b
b
a
i
in
ou
tk
u
w
tw
w
-B
do
o
do
ts
e-
td
gh
d
tB
e
bu
gl
ut
gh
ht
ht
ht
ei
gh
sin
gh
-fl
ig
ei
t-
w
ei
ei
ei
ei
C
w
e
gh
ht
m
w
w
w
w
vy
t
g
gh
iu
vy
ht
ei
vy
m
h
ei
ea
ed
lig
ig
ea
ei
iu
ea
w
-H
ht
-L
M
ed
-H
-H
h
er
ig
ig
id
-M
J-
up
A
-L
-H
M
F
G
-S
I-
H
B
Figure 4.11 Comparison of actual BCT with the BCT predicted by

the McKee formula using three- or four-point D values
For the case visited in Figures 4.7 and 4.9–4.11, the box dimensions
were 61 × 41 × 66 cm, so that MD shear should not be significant,
however, the selected test length of 36 cm still provides a better BCT
prediction.
Deflection in D measurements should be limited to allow the

assumption of linear elasticity theory to apply. In particular, the
strain ε of the outerliner must be kept below 0.05%, which can be
calculated using the formula:
(4.15)

76
where the strain is given in percentage, the deflection ‘y’ in mm, ‘t’
is the board caliper and ‘L’ is the test span between the two fulcrum
points.
4.5 Summary
Two-point D is used for paper and some board materials, whereas
four-point D is used for corrugated board. Both measurements depend
on the Sb, ‘E × t’. For paper, it is more accurate to use the soft-platen
caliper if D as ‘Et3/12’ is to be calculated as a check of measurement
accuracy. In this case, the modulus ‘E’ can be determined from the
analysis of a tensile test and the calculation of ‘D = Et3/12’ can be
compared with measurements of ‘D’. For corrugated board, the D is
approximated by the sandwich beam model ‘Eth2/2’, where ‘Et’ here
is now the Sb of the outerliners and ‘h’ is the board caliper. Three-
point bending provides smaller values than four-point, depending on
the test piece length, because of the effect of shear. Three-point D
values may provide a more accurate prediction of BCT when using
the McKee equation.
References
1. S.P. Timoshenko and J.M. Gere in Mechanics of Materials,
D. Van Nostrand, New York, NY, USA, 1972.
2. V.C. Setterholm in Handbook of Physical and Mechanical

Testing of Paper, Volume 2, Eds., R.E. Mark and K. Murakmi,
Macel Dekker, New York, NY, USA, 1984.
3. C.G. Carson and R.E. Popil, TAPPI Journal, 2008, 7, 12, 17.
4. TAPPI T 489: Bending resistance (stiffness) of paper and

paperboard (Taber-type tester in basic configuration)
(Revision), 1992.
5. T.M. Nordstrand, Composite Structures, 1995, 30, 4, 441.
77
6. TAPPI T 836: Bending stiffness, 4 point method, 2016.
7. R.E. Popil, Bioresources, 2012, 7, 2, 2553.
8. L. Carlsson, C. Fellers and P. Jonsson, Das Papier, 1985, 39,

4, 149.
9. R.C. McKee, J.W. Gardner and J.R. Wachuta, Paperboard

Packaging, 1962, 47, 12, 111.
10. R.C. McKee, J.W. Gardner and J.R. Wachuta, Paperboard

Packaging, 1963, 48, 8, 149.
78
5
Compression Testing of Paper,
Board and Boxes: Relationship to
Tensile Testing, Elastic Modulus
and the Influence of Artefacts
5.1 Introduction
Compression strength of paperboard is the dominating factor that
governs the stacking strength of corrugated boxes. Considerable effort
in the industry is focused on measuring the compression strength of
boxes, corrugated board, and the linerboard and medium that comprise
the corrugated boards. Compression strength is also considered in
the design of folding carton packaging. Increasing the compression
strength relative to the basis weight ‘b’ of the components remains a
major development area. Success has been achieved through the wet
pressing of sheets during the forming process to obtain high density
and high strength. The use of strength additives in the stock also
increases strength with the added expense of the chemicals and changes
in dewatering properties of the stock. In addition, starch application
at the dry end of the paper machine (PM) can provide some benefit,
as can adjustment of the PM variables such as jet to wire speed ratio,
open draw tensions, jet impingement angle and others.
Compression and tensile strengths (St) are related, although the

mechanisms of failure are different. St failure is due to a combination
of fibre fracture and fibre pull-out separation from the matrix. Close
visual examination of a paper fracture edge will reveal long fibres
extending out from the edge indicating fibre pull-out. Whereas
compression failure consists of a combination of fibre buckling and
separation of fibre layers at a visible crease, which is out-of-plane
from the sheet. Both compression and St are related to the fibre
modulus and the degree of fibre bonding; theory models exist [1]
which can provide useful insight towards optimisation.
79
Shallhorn and Gurnagul [2] adapted the tensile Page equation [3]
derived for St for compression strength modelling as a combination
of fibre axial failure and fibre Euler buckling. They produced their
analysis in the form of a harmonic average similar to the Page
equation, analogous to resistors in parallel:
1 = 1 + Cw c t f - 1 m2
vc v o ; 2aE f t 3 E t
(5.1)
where the sheet compressive strength per unit basis weight at limiting
high-density ‘σo’ (Nm/g) is related to the intrinsic fibre axial strength
‘σf’ (N) by:
v f = 8 Cv o
3 (5.2)
with ‘C’ being the fibre coarseness (g/m), ‘w’ the fibre width, ‘t’ the
fibre thickness, with t being much less than w, ‘Ef’ the fibre elastic
modulus, ‘α’ is an efficiency factor ~1, ‘ρf’ and ‘ρ’ are the fibre
(cellulose) and sheet densities, ‘σc’ is the sheet compression strength
index (Nm/g) equated to the short-span compression test or strength
(SCT), as measured by the Technical Association of the Pulp and
Paper Industry (TAPPI) T 826 method [4]. Inspection of the equation
shows that compression strength decreases with fibre wall thickness
‘t’, increases with wet pressed sheet density ‘ρ’, and refining is more
effective at increasing compression strength for a given increased
density compared with wet pressing, as refining also decreases fibre
coarseness. Significant differences in σo were noted, for example,
between spruce fibres (Ef = 56 GPa) and coarser thicker Douglas
fir fibres (Ef = 48 GPa). At low-density fibre buckling dominates,
therefore a larger fibre modulus and larger fibre thickness increase
80
Compression Testing of Paper, Board and Boxes: Relationship to
Tensile Testing, Elastic Modulus and the Influence of Artefacts
compressive strength. At high density, fibre compressive strength

dominates, hence a low fibril angle (high Ef) and low fibre thickness
produce a larger compression strength. Fibre modulus is related to
the sheet elastic modulus ‘Ep’ through the relationship derived by
Page and Seth [1] for randomly oriented sheets:
E P = 1 E f 61 - k/RBA@
3 (5.3)
where ‘k’ is a constant that is specific to a pulp and RBA is the fibre-
relative bonded area, which can be determined by optical scattering
coefficient measurements. Experimentally, Shallhorn and Gurnagul
inferred Ef as being just Ep/3 with sheet modulus Ep obtained at the
highest level of wet pressing, i.e., 7 MPa, the assumption is that at
this pressure the sheet elastic modulus is close to its maximum.
Figure 5.1 shows a hypothesised superposition of compression and

tensile stress–strain data for a typical fourdrinier-made linerboard
sample. At low strains, i.e., 0.5% and lower, the response of the
sample to compression or tensile strain is linear and to a first
approximation is described by Hooke’s law:
v = Ef (5.4)
σ is the stress, i.e., the applied force divided by the width multiplied
by the effective caliper of the test strip, ε strain is the displacement
of the test strip divided by its initial free length and ‘E’ is Young’s
modulus in this approximation. For an orthotropic solid such as
paper, the strains, stresses and modulus are all directional, here we are
principally concerned with the cross machine direction of paper (CD)
direction and the directional notation, in Hooke’s law, is dropped.
Handsheets are randomly oriented and so the MD CD notation does
81
not apply in this case as well. The modulus at low strains is the same
whether the sample is under compression or tension, as Figure 5.1
indicates.
50
45
MD tension
40
35
Stress (MPa)
30
CD tension
25
20 MD compression
15
10
CD compression
5
0
0 0.5 1 1.5 2 2.5 3 3.5 4
Strain (%)
Figure 5.1 Superimposed stress–strain data for tensile and

compression tests for a typical machine-made linerboard sample
Machine-made paper (MD)/CD modulus and strength ratios are

approximately the same, about 2 for fourdrinier papers. St are also
about twice as large as the corresponding compression strengths.
At high levels of sheet consolidation and bonding, the S t and
compression strength properties are related and determined by the
fibre quality and somewhat affected by the degree of bonding. The
relationship between compression and tensile measurements allows
the opportunity, in some circumstances, to estimate the compression
strength of papers from tensile tests when compression tests are
unavailable or impractical to perform. The compression testing of
paper becomes subject to artefacts when dealing with lightweight
papers, as buckling of the test specimen can produce an anomalously
82
low compression strength value. In such cases, if tensile data are

available for higher basis weights, the correspondence between St and
compression strengths can be extrapolated to infer the compression
strength at low basis weight.
An example of this follows, using the ring crush test (RCT) [5], SCT
and tensile test results in combination. The marketing of linerboard
and corrugating medium predominantly use the RCT as a quality
criterion. This consists of cutting a strip ½ × 6 inches, the length
being in the MD of a sheet, placing the test strip in a circular grooved
fixture and crushing the assembly in a compression tester with the
platen advancing at ½ inch per minute. Figure 5.2 shows a sample
in the RCT fixture where the compression tester platen has been
allowed to progress beyond the peak load.
Figure 5.2 Photograph of an RCT sample of high basis weight

after the test. Reproduced with permission from R.E. Popil in the
Proceedings of the TAPPI PaperCon 2010, 2–3th May, Atlanta,
GA, USA, 2010. ©2010, Technical Association of the Pulp and
Paper Industry [6]
83
A distinct repetitive buckling pattern going along the length of the

sample is evident in Figure 5.2, as is the top edge having rolled due
to contacting the compressing platen. Although the diameter of the
ring is some 50 mm and the height of the sample is 12.5 mm, it can
be shown from simple mechanics analysis that such a short paper
cylinder is still subject to buckling when placed under vertical load.
The buckling load ‘σcr’ of a thin shell ring, based on the analyses of
Timoshenko and Gere [7], is:
v cr = E CD t
1 - v 12 v 21 R (5.5)
where ECD is the CD modulus, ‘t’ is the thickness of the test specimen,
‘R’ is the radius of the ring fixed at 24.2 mm, the Poisson ratio term
under the square-root sign adds a few per cent correction to the
buckling load and does not vary significantly with different papers.
The criterion for this equation to be valid is that the ratio of the
buckling column height ‘l’ over R must be sufficient to fit at least
one buckling wave and is stated as:
l = 1.72 Rt (5.6)
This criterion is fulfilled for the standard sample of t = 0.3 mm and

RCT strip width of 12.4 mm, of which 6 mm protrudes outside of
the fixture base, so buckling in the RCT test can be expected from
a theoretical perspective. Buckling of a sample is undesirable as it
will lower the peak load when a sample is tested under compression.
Accordingly, a different geometry for compression testing has been
devised in the SCT.
The underlying principle in the SCT test method is that the sample
is constrained by clamps eliminating any buckling during the test.
84
If the test sample strip under compression is considered as a simple

elastic beam, then peak stress by buckling from load applied to the
two ends of the beam is, from first principles, given by:
v max = kr E
2
CD
m2 (5.7)
where ECD is the modulus of the test sheet in the CD, which is the
usual test direction of interest, k is a constant depending on the end-
point conditions and ‘λ’ is the important slenderness ratio defined as:
m=2 3 l
t (5.8)
where ‘l’ is the length of the sample strip and ‘t’ is its thickness. Plots
of σmax versus λ show that the compression stress stabilises once λ <30.
For linerboards, this translates to strip lengths of approximately
1 mm and to accommodate the range of linerboards, the standard has
been set with the free unclamped strip length for a compression test
being 0.7 mm. Thus, the SCT test has 15 mm wide strips clamped at
two ends with a free span of 0.7 mm, a fixed-load cell attached to one
clamp measures the peak force once the opposite clamp is advanced.
The point of this arrangement is to eliminate any buckling of the
test specimen and thus obtain the compression strength without any
sample bending. Unfortunately, testing such a small sample area of
15 × 0.7 mm subjects the values to a comparatively high variability
related to the samples’ mass non-uniformity on the mm scale, also
commonly called paper ‘formation’. In fact SCT, like many strength
characteristics, is proportional to basis weight. Accordingly, a high
variation in basis weight can be expected to also result in a high
variation in strength values.
85
A comparison of SCT values to RCT is instructive to illustrate the

influence of bending in the latter. A sample set consisting of 18
softwood linerboards and medium manufactured in the US southeast
were tested for RCT, SCT, basis weight, caliper and ultrasonic stiffness
using the Lorentzen & Wettre (L&W) tensile stiffness orientation
(TSO) instrument [8]. The results are presented in Table 5.1. The
buckling to compression stress ratio is calculated as:
v cr = E CD # t .
1 05
SCT R
t (5.9)
since SCT is given per unit width, dividing SCT by the caliper
produces the compression stress, ‘R’ is the ring radius of 24.2 mm, the
Poisson term of Equation 5.5 ‘√(1 – ν12ν21)’ has been approximated to
be constant at ~0.95 and ‘ECD × t’ is the tensile stiffness (Sb) provided
by the L&W TSO in Equation 5.10 as:
(5.10)
where ‘β’ is the basis weight and the factor 0.571 with offset
41.8 accounts for the difference between ultrasonic stiffness and
mechanical St obtained by a separate experimental correlation [6].
A plot of the critical buckling to compression stress ratio is shown

in Figure 5.3, where it is evident that buckling can be expected for
basis weights less than 81 g/m2, as the ratio is less than 1 at basis
weights below 81. Shallhorn and Gurnagul [10] noted that the
elasticity theory formula for the critical buckling stress for thin-
walled tubes is an overestimation by a factor of approximately 2
compared with experimental data. This then brings the basis weight
86
where the critical buckling stress becomes less than compression

stress to approximately 180 g/m2, this latter estimate is plotted as
a dashed line in Figure 5.3.
Table 5.1 Data for a set of linerboards and medium used for
commercial corrugated board in the southeastern US
Sample Caliper Basis weight TSI_CD CD SCT RCTCD
(mm) (g/m2) (km/s)2 (N/mm) (N/mm)
Liner A1 0.232 150.1 4.59 3.26 1.79
Liner A2 0.268 171.1 4.88 3.93 2.22
Liner A3 0.305 195.3 5.46 3.86 2.95
Liner A4 0.392 249.5 4.7 5.39 3.45
Liner A5 0.420 275.5 4.9 5.47 3.57
Liner A6 0.476 293.9 4.69 5.86 3.78
Liner B1 0.600 370.4 4.43 8.06 4.62
Liner B2 0.427 277.1 5.13 6.35 3.69
Liner B3 0.725 437.1 3.78 8.96 4.71
Liner B4 0.327 206.3 5.51 4.59 2.87
Liner B5 0.663 393.1 3.96 7.14 4.67
Liner B6 0.260 167.0 5.24 3.80 2.17
Liner B7 0.524 323.8 3.97 6.03 3.66
Liner B8 0.229 149.6 4.55 2.81 1.81
Liner B9 0.322 195.8 5.43 5.01 2.88
Liner B10 0.319 205.4 4.68 3.88 2.56
Liner B11 0.203 128.9 5.13 2.50 1.14
Liner B12 0.259 177.6 4.97 3.00 1.92
Medium 1 0.146 78.5 3.58 1.32 0.44
Medium 2 0.192 88.3 3.76 1.50 0.64
Medium 3 0.197 95.7 4.12 1.75 0.74
Medium 4 0.208 126.9 4.21 2.62 1.31
Medium 5 0.251 163.7 4.57 3.88 2.16
TSI: Tensile stiffness indices, CD and MD
Reproduced with permission from R.E. Popil, BioResources Journal,
2012, 7, 2, 2553. ©2012, North Carolina State University [13]
87
4
y = 0.0072x + 0.42
3.5
R2 = 0.93
Buckling/compression stress
2.5
(Buckling stress/2)
2
1.5
0.5
50 150 250 350 450 550
Basis weight (g/m2)
Figure 5.3 Ratio of the calculated buckling to compression stress

for the sample set of commercial linerboards and medium listed
in Table 5.1. The squares represent the ratio with the calculated
buckling stress divided by 2. Reproduced with permission from
R.E. Popil in the Proceedings of the TAPPI PaperCon 2010,
2–3th May, Atlanta, GA, USA, 2010. ©2010, Technical
Association of the Pulp and Paper Industry [6]
Therefore, buckling of samples under compression can be expected

for low basis weights, possibly as much as 180 g/m2. Buckling of
samples at low basis weight should produce a lower compression
strength than that provided by SCT. This is shown in Figure 5.4 for
the sample set presented in Table 5.1, where the units for RCT have
been converted to be the same as SCT, i.e., units for both are kN/m
or equivalently, N/mm.
88
10
Compression strength SCT or RCT (kN/m)
9
SCT = 0.0202 (β) + 0.069
8
R2 = 0.95
7 SCT
6 RCT
5
4
3
RCT = 2.71 ln(β) + 11.68
2 R2 = 0.97
1
0
50 150 250 350 450 550
Basis weight β (g/m2)
Figure 5.4 Comparison of SCT and RCT results for the sample
set in Table 5.1. Reproduced with permission from R.E. Popil in
Proceedings of the TAPPI PaperCon 2010, 2–3th May, Atlanta,
GA, USA, 2010. ©2010, Technical Association of the Pulp and
Paper Industry [6]
Figure 5.4 shows that at basis weights below 200 g/m2, SCT is larger
than RCT by approximately 1.7 times. The higher SCT values can be
attributable to the role of bending, which lowers the RCT. At higher
basis weights, the difference between SCT and RCT increases and
the upper free edge of the test strip becomes rolled due to contact
with the advancing platen of the compression tester. The best fit
regression analysis in Figure 5.4 shows that SCT is linear with basis
weight, whereas the RCT relationship best fit is logarithmic with
basis weight. Nonetheless, RCT is still a preferred specification for the
marketing of corrugated board component linerboards or medium,
89
as the variability of RCT is approximately two-fold lower than for

SCT. The results in Figure 5.4 show error bars representing 95%
confidence intervals for 10 measurements per sample. The coefficient
of variation, expressed as a percentage ratio of the standard deviation
to the average value, averages 7.8% for SCT compared with 4.7%
for RCT. The lower variability for RCT allows the production of
linerboard at a lower basis weight that will meet a specified marketing
specification. However, mill producers of linerboards prefer SCT
as a quality criterion, as it appears to be more easily controlled
through papermaking operations such as pulp stock refining, wet
pressing density, paper basis weight, softwood/hardwood ratio, stock
additives, and so on.
Besides being subject to buckling, rolling edges and not tracking basis
weight, RCT is also two-sided, meaning results can be significantly
different depending on the side facing the inside of the cylindrical
ring undergoing vertical compression. The majority of corrugating
medium papers are made on fourdrinier PM such that the drainage of
the stock on the forming wire is one-sided, leaving the top side more
filled with filler and fines than the wire side. The top side is therefore
somewhat weaker than the wire side. If the cylindrical buckling of
the test piece is primarily outward, then the outside of the ring is
subjected to tension and the inside to compression. Figure 5.5 shows
the results of 18 samples of medium from two paper machines at a
mill where three measurements were made for each sample with the
top side of the sample being inside the test ring and three outside
of the ring.
In most of the cases shown in Figure 5.5, RCT is a few per cent
higher and sometimes significantly so when the stronger wire side is
on the inside of the buckling test cylinder. RCT is also sensitive to
the cutting method. Parallel sides must be maintained so it is best to
use a punch cutter rather than a guillotine-style cutter. Furthermore,
test strips must be punched one at a time rather than several sheets
at a time, otherwise edges are obtained that curl when tested, leading
to lower than expected RCT values.
90
55
50
Top side inside

45
Top side outside
RCT (lb/6 in.)
40
35
30
25
20
1 3 5 7 9 11 13 15 17 2 4 6 8 10 12 14 16 18
PM 1 PM 2
Figure 5.5 Comparison of RCT (in standard English units of lb/6 in.)
for 18 samples with different side orientations as indicated (1 lb/6
in. = 0.0292 KN/m)
To satisfy long-established marketing requirements that specify RCT,

with its combined bending and compression, and to connect it to the
less artefact-prone SCT, a model can be devised based on the empirical
result for structures that fail through combined compression and
bending with the form for the maximum load ‘Pmax’ as:
Pmax = a Pcom b
c m
Pcr Pcr (5.11)
where ‘Pcr’ is the critical buckling load, ‘α’ and ‘b’ are constants,
and ‘Pcom’ is the intrinsic compression strength of the material. Thus,
91
taking RCT for Pmax and SCT for Pcom, the proposed model for RCT
becomes:
RCT = a ^SCT hb ^v cr h1 - b
(5.12)
The availability of ultrasonic TSO instruments [8] on automated

testing equipment installed in many mills makes use of the above
equation in a convenient way to obtain an RCT value. Since the
(E × t) term for the formula for the buckling stress ‘σcr’ is equivalent
to TSICD × β, the ‘R’ is constant in the RCT, we obtain by substitution
into the previous formula:
RCT = C ^SCT hb ^ TSI CD b h1 - b

(5.13)
with ‘C’ being a new constant. Keeping the units shown in Table 5.1,
RCT and SCT in N/mm, TSICD in (km/s)2 and β in g/m2, a three
parameter fit of the model equation to the data determined the model
constants to be:
RCT = 0.0326 ^ SCT h0.46 ^TSI CD b h0.54

(5.14)
which produces an average residual error of 0.177 N/mm compared

with a simpler linear model fit of RCT = 0.596 × (SCT), which
produces an error of 0.246 N/mm, as shown in Figure 5.6.
92
10
9
8 SCT = 1.54 (RCT) + 0.38
SCT or RCT (N/mm)
R2 = 0.95
7
6
5
4
3
2 Model = 0.89 (RCT) + 0.34
1 R2 = 0.98
0
0 1 2 3 4 5
Actual RCT (N/mm)
Figure 5.6 Plot of the RCT model fit (round points) and the SCT
values (square points) versus actual RCT for the samples listed in
Table 5.1
The same method was applied to another set of challenging

lightweight linerboards consisting of 100% recycle furnish. The
same analysis was applied but yielded different results, as shown in
Figure 5.7.
The basis weight and SCT data in Figure 5.7 were obtained from an
automated testing machine. The RCT results in the plot came from
manual laboratory testing but is preferably replaced by automated
testing or otherwise calculated from available data using the existing
automated instruments. Note that, as before, despite the fact that
both measurements describe compression strength, RCT is always
less than SCT, ranging from 3.5 times lower at a low basis weight
to 1.3 times lower at a higher basis weight.
93
4.0
SCT = 0.0174β + 0.231
R2 = 0.94
3.5
3.0 RCT
RCT or SCT (N/mm)
SCT
2.5
2.0
1.5
RCT = 0.0219β - 1.57
R2 = 0.97
1.0
0.5
0.0
70.0 90.0 110.0 130.0 150.0 170.0 190.0
Basis weight (g/m2)
Figure 5.7 A plot of SCT and RCT versus basis weight for samples
from a 100% recycle mill
Therefore, a similar analysis was undertaken as before to find a

predictive relationship for SCT using a two parameter fit model. The
least amount of scatter and best correlation for this particular data
set was found to use a combination of SCT and St data. St in the CD,
dubbed (St,CD), like Sb, can also be expected to be related to the CD
elastic modulus for a given sample set.
94
2.5
R2 = 0.94
2
Actual RCT (N/m)
1.5
0.5
0
0.6 0.8 1 1.2 1.4 1.6
Model (SCT and St,CD) RCT (kN/m)
Figure 5.8 Actual RCT versus a fitted model of RCT using SCT
and St for lightweight 100% recycled pulp linerboard
The fitted model for RCT represented in Figure 5.8 was determined
to be:
RCT = 0.42 SCT 0.86 ^S t,CDh0.14

(5.15)
with RCT, SCT and St,CD all in units of N/mm. The average error
between actual and model values is 0.27 N/mm. The advantage of
such a model is that it uses the data available from automated testing
machines avoiding the necessity of error-prone manual testing of RCT.
95
5.2 Compression Testing of Corrugated Board

The interest in testing SCT and RCT is primarily to optimise the
compression strength of corrugated board. Shipping containers made
from corrugated board are required to meet compression strength
specifications in order to be certified. The compression strength of
boards determines the stacking strength of boxes. The compression
testing of boards is faster and easier than the compression testing of
boxes hence, considerable attention is paid to the edge compression
strength test (ECT) of board in box plants for production quality
control. The effects of the quality of selected components and
converting operations, such as the inadvertent crushing of board
and bonding quality of the fluted medium to the linerboard, can be
measured by ECT.
Similarly, for the compression testing of linerboard, consideration of

the contribution of bending or other artefacts must be considered in
order to obtain an optimal compression strength result. Details, such
as specimen preparation, fixtures or constraints, have an effect. The
board compression test of interest for most box-making operations
is the edgewise compression in the CD ECT. In general terms, and
to a good approximation, it is related to either SCT or RCT through
a linear length-weighted summation of the individual strengths that
make up the corrugated board [11]:
(5.16)
(5.17)
where the medium to linerboard length ratio is the take-up factor

‘α’ and is typically 1.43 for C-flute boards. When using the RCT or
96
SCT summation it has been found that the form of the ECT equation
assumes that the components all fail at equal deformation. The
medium in most common boards is usually at a lower basis weight
than the linerboards, so it will fail at a lower strain and strength
when compression tested by itself alone; however, when it is fluted,
the curvature imparts a structural strength such that the failure at
equal deformation holds, as observations and numerical simulations
show. The constant ‘C’ is less than 1, typically 0.7, reflecting the loss
of strength of the components due to the converting process. The
formula can be extended to multiwall boards by summing up the
length-weighted SCT compression strengths of the ‘i’ linerboards and
‘j’ fluted media with their respective take-up factors:
(5.18)
5.3 Three Different Edge Compression Strength Test

Methods
When ECT values are substantially less than those predicted by
the formula, it may be attributable to the method chosen not being
suitable for the type of board being tested. Test specimen size,
specimen preparation and compression strain rate all have an effect
on the values. Three different methods are in common use. The
traditional TAPPI T 811 method [12] hardens the vertically loaded
edges of the test specimen by immersion of the edges in molten wax
to a prescribed depth. A different height of the test specimens is cut,
depending on the flute size, to limit bending, as may be assessed
by calculation of the slenderness ratio and ensuring that it is less
than 30. The vertical edges must be cut squarely and parallel to one
another, and Table 5.2 cites the various test specimen dimensions for
various types of boards. Test pieces are supported vertically between
the compression tester platens by machined steel blocks, which are
97
removed from the test piece once contact is made by the advancing
platens. The TAPPI T 811 method is approved for shipping box-
stamp certification.
Table 5.2 Common flute sizes, TAPPI T 811 method test

height, caliper and the slenderness ratio. 50 mm was selected
arbitrarily as the test specimen height for E- and F-flutes
Flute size TAPPI T 811 method test Board caliper Slenderness
(height in mm) (mm) ratio (λ)
A 50.8 7.9 22.2
B 31.8 3.2 34.4
C 38.1 4.8 27.5
E 50 1.6 108
F 50 1 173
The inconvenience of cutting different specimen heights for different

board types and dipping edges in molten wax is removed in the
TAPPI T 839 method [13], which uses a fixture consisting of two
pairs of clamps that hold the test specimen vertically aligned with the
compression platens. This is the method often used for quality control
checks in box plants. The specimen dimensions remain constant for
all board types and a specialised cutter, using parallel blades, ensures
squareness and parallelism of the vertically loaded edges. A problem
with the TAPPI T 839 method [9] can develop with high basis weight
boards where rolling edges occur, or at lightweights, where the spring-
loaded clamps introduce outward bowing of the facing of the test
piece. When using the TAPPI T 839 method, the test pieces should be
98
examined after compression to ensure that a visible crease, extending

through the width of the specimen, is clearly visible on both sides
of the board. Figure 5.9 shows two commercially available TAPPI
T 839 method ECT clamp fixtures. The springs provide a pressure of
approximately 10 psi onto C-flute boards when the clamp is released
to hold the board test piece firmly in place for the compression test.
a) b)
Figure 5.9 Sumitomo (a) and Emerson (b) TAPPI T 839 method
ECT clamp fixtures
Figure 5.10 illustrates several types of failure that can occur with the
clamp method. Heavy basis weight boards often display a folding
compression of the edge at the top or bottom. The accepted test failure
is observed as a crease going across the test piece at the unclamped
middle portion. Failure is guaranteed to occur in the middle portion
of the test piece once a tapered profile is cut into the test piece using
the rotating knife fixture shown in Figure 5.11.
99
Bad
Good Crushed
Crushed
Figure 5.10 Examples of postECT tested board specimens. From

the left, a sample with an MD edge roll, crease not at the midde;
next, a good sample with a visible crease at the free unclamped
middle of the test piece; a crushed board showing outward
buckling failure; rightmost sample, a neckdown crushed sample
still showing a crease at the middle of the test piece
Figure 5.11 Fixture with rotating blades for cutting neckdown

board pieces for the TAPPI T 838 method ECT
Figure 5.12 shows differences in ECT values using the different

methods on a series of laboratory-made A-flute boards all comprising
100
of 205 g/m2 linerboards but with varying basis weight of fluted

medium ranging from 68 to 205 g/m2. The largest disagreement
between predicted values and measurements occurs for the TAPPI
T 839 method for low basis weight samples. The least disagreement
overall for the sample set occurs for the TAPPI T 838 method. At
high medium basis weights, the TAPPI T 839 method provides
higher values than the TAPPI T 811 method, as may be expected
since the board is under greater restraint when in the clamping test
fixture. The large disagreement for the TAPPI T 839 method can be
understood by the effects of the clamps crushing the sample. What
appears to happen in the case of boards of low basis weight medium
or boards, that have undergone crush damage, is that the unclamped
facings of the test specimen have a small convex curvature at the
onset of increasing load and the curvature increases to produce a
noticeable bending failure rather than compression failure, as shown
in Figure 5.10 and Figure 5.13.
11 T 839 clamp method

10 T 811
9 T 838 routered
ECT (kN/m)
8 Predicted values
7
6
5
4
3
68 78 88 88 98 127 161 205
A-flute medium basis weight(g/m2)
Figure 5.12 ECT values for A-flute boards made with 205 g/m2
linerboards but different weights of medium as indicated.
Reproduced with permission from R.E. Popil, BioResources
Journal, 2012, 7, 2, 2553. ©2012, North Carolina State University
[9]
101
Figure 5.13 The TAPPI T 839 method clamp test fixture with
an insert showing a close up of the clamps crushing a low basis
weight board. Reproduced with permission from R.E. Popil,
BioResources Journal, 2012, 7, 2, 2553. ©2012, North Carolina
State University [9]
Much corrugated board is crushed to some degree during manufacture

either from the transfer press rolls or through die-cutting scoring
and slitting operations that are not optimally maintained. A-fluted
medium weakened from crushing can produce an anomalously low
ECT value, similarly to the lightweight medium results shown in
Figure 5.9.
In one study, a series of laboratory-made boards were run through a

motorised laboratory roller press nip running at a speed of 50 ft/min
with the roll gap separation set at various percentages of the original
uncrushed caliper, ranging from 90 to 60%. A survey of box plants
showed that a typical transfer roll gap is approximately 80% of the
initial board caliper. In these investigations, boards had a fluting
weight of 127 g/m2 (26 lb/1,000 ft2) and were of type A-, B- and
C-flute, the final calipers after nip crushing were measured and the
ECT results acquired using the TAPPI T 839 method. In addition,
one series of C-flute boards with a lighter fluting weight of 112 g/m2
(23 lb/1,000 ft2) was also tested using both TAPPI T 839 and T 811
102
methods, and the results are summarised in Figure 5.14 as the ECT
loss as a function of caliper reduction. Generally, the crushing data
indicates that ECT reduction is proportional to the corresponding
decrease in caliper. The lightweight board 23C data (open circles and
triangles in Figure 5.14) indicate that the waxed-end method results
(triangular points) have lower ECT loss values compared with the
clamp method results for low caliper reductions, as expected, but
the situation reverses at high crush values.
41
36 C-flute
B-flute
Reduction in ECT (%)
31
A-flute
26 T 839 23C board
T 811 23C board
21
16
11
6
1
1 6 11 16 21 26
Reduction in caliper by crushing (%)
Figure 5.14 Reduction percentage in ECT value of boards

progressively crushed by a rolling nip
5.4 Bending during Board Compression Testing

Bending during compression testing always lowers the value of the
peak load. Bending of the whole test piece or, on a smaller scale, of
103
the linerboard facings can occur. Bending of the entire test piece is
controlled by limiting the test piece height, whereas the bending of
the facings between the flute lines (interflute buckling) is governed
by the facing stiffness and flute spacing. Corrugated board bending
stiffness ‘D’ is approximated as a sandwich structure consisting of
facings Sb ‘E × t’ separated by caliper ‘h’ such that the medium has
a negligible contribution and thus:
2
D , Eth
2 (5.19)
The vertically loaded buckling load ‘P’ of a corrugated board

approximated here as a sandwich beam is considered to consist of
a combination of beam buckling ‘PE’ and shear ‘PS’ [14], expressed
as the harmonic mean in the form:
1 = 1 + 1
P PE PS (5.20)
The Euler beam buckling load ‘PE’ per unit length is π2EI/H2 or
π2Db/H2 with ‘I’ being the second moment of area, ‘H’ is the height
of the vertically loaded beam and ‘Db’ is the beam bending stiffness.
The properties of relevance here are the elastic modulus (ECD) and
machine direction of paper bending stiffness (DCD), as the vertical
loading in the ECT is along the direction of the fluting or cross-
direction, CD. If the ends of the beam are rigidly held or constrained
from pivoting, then Euler beam buckling theory predicts the bending
load to be 4π2DCD/H2.
Shear in corrugated board is the propensity of one facing to become

displaced relative to the other when the board is subjected to stress.
104
Shear stiffness or shear rigidity is regarded as significant due to the

following considerations. The shear rigidity, PS, for corrugated board
has been measured using a torsion pendulum method [15]. Essentially,
this method involves using a long strip sample of board clamped at
either end, with the upper end fixed to a supporting frame and the
lower attached to a dumbbell that is free to rotate. The board is set
to undergo a twisting motion as shown in Figure 5.15, the frequency
of the twisting motion is converted to a shear rigidity using the
D measurement and results from mechanical analysis.
Figure 5.15 Torsion pendulum measurement of the twisting

frequency of a corrugated board strip in the MD
105
For typical C-flute board with medium basis weight of 127 g/m2,
shear rigidity in the CD, designated R44, is 54 kN/m, which is much
larger than the corresponding MD value of 9.8 kN/m. Buckling loads,
determined by π2Db/H2 for typical C-flute samples having a DCD of
5 N/m and height in the range of 50 to 32 mm, are calculated to be
20 to 48 kN/m, which are comparable to the measured shear rigidity.
Valid ECT measurements for any board must not have bending, which
means that the critical buckling load ‘P’ of a board must be greater
that its compression strength ECT. Otherwise, as the compression
platens advance and the load increases, the board will first fail by
beam buckling rather than by compression. Upon substitution and
rearranging the Plantema equation (Equation 5.20) this criterion,
that P > ECT, takes the form:
r2 D CD PS $ ECT
r D CD + PS H 2
2
(5.21)
This equation is used for comparison with experimental data to

assess the effect of test specimen height on ECT. This allows a check
of whether the selected height is optimal for a particular board.
A selection of various corrugated board was chosen to evaluate the

effects of height on ECT. The relevant properties of these boards is
given in Table 5.3. D was measured using the four-point method and
shear rigidity by the torsion pendulum method.
What is interesting to note in Table 5.3 is how close the simple

sandwich beam calculation for ‘D’ is compared with the actual four-
point method measured values and the more sophisticated formulas
or calculations, such as those of Carlsson, Ranger or Nordstrand,
which take into account the small contributon of the fluted medium.
106
Table 5.3 Relevant corrugated board properties used in the evaluation of test piece height on ECT
Board properties DCD MD
shear
Board Take- Liner Medium Liner Medium Board Four-point Sandwich Ranger Nordstrand Carlsson
rigidity
flute up CD CD caliper caliper tm caliper measured beam formula formula calc.
R44
factor modulus modulus tl (mm) h (mm) (N-m) formula (N-m) (N-m) (N-m)
(kN/m)
α El (GPa) Em (GPa) (mm) (N-m)
C 1.42 2.09 1.61 0.29 0.25 4.21 5.2 5.37 5.28 5.31 5.45 54.3
E 1.27 1.72 1.13 0.24 0.19 1.64 0.7 0.56 0.44 0.44 0.6 20
F 1.25 1.74 1.14 0.29 0.22 1.37 0.41 0.47 0.31 0.32 − −
N 1.2 1.74 1.14 0.29 0.22 1.07 0.23 0.29 0.16 0.16 − −
Reproduced with permission from R.E. Popil, BioResources Journal, 2012, 7, 2, 2553. ©2012, North Carolina State
University [13]
107
5.5 Experimental Observations – Effect of Test

Specimen Height
5.5.1 C-Flute Board
Two series of parallel experiments were conducted using commercially

produced board WC 4226C 42 (205 g/m 2 liner and 126 g/m2
medium), obtained off the corrugator, and similar laboratory-
produced board using the same components. One series of test
specimens had their ends embedded perpendicularly in a quick-setting
epoxy resin, Alumilite™, producing a self-supporting test piece that
is vertically compressed on its edge in a universal testing machine
(UTM) compression tester, shown in Figure 5.16.
a) b) c)
Figure 5.16 a) and b) an ECT sample embedded in supporting

epoxy resin strips, front and side views, and c) mounted between
the platens of a compression tester. Reproduced with permission
from R.E. Popil, BioResources Journal, 2012, 7, 2, 2553.
©2012, North Carolina State University [9]
108
This is a variation of the TAPPI T 811 method with the horizontal test
piece ends encapsulated in hardened epoxy resin platforms instead
of being dipped in molten wax. The test pieces were prepared using
fixtures that supported the cut boards vertically while the setting
epoxy was poured. The epoxy platforms restrict pivoting at the ends
and ensure failure will occur away from the edges in the same manner
as the impregnated hardened wax. Test pieces were compression
tested in a Model 1122 Instron UTM using Series IX software to
record load and deformation at a cross-head velocity of 12.5 mm/min.
Figure 5.17 shows the experimental data of the resin-embedded ends

of the commercial WC 4226C C-flute board blank (205 g/m2 liner,
127 g/m2 medium single-wall) sample set. The data set suggests that
the free span between the glued ends, i.e., a specimen height of up to
80 mm, will provide a constant value of ECT. The fall in ECT value
attributable to the onset of beam bending is in general agreement with
the estimate provided by the beam buckling criterion equation, which
is in Figure 5.17, and is used to plot height versus buckling load using
the board properties detailed in Table 5.3. Boards that have an ECT
at a height above the curve value are expected to buckle, whereas
ECT values under the curve value represent boards that are failing via
compression with no bending. Error bars denote the 95% confidence
intervals of 10 repeat measurements at each selected height.
An alternative specimen testing arrangement used a Sumitomo

TAPPI T 839 method ECT clamp fixture, which was disassembled
to accommodate differing sets of machined aluminium block spacers
to separate the upper and lower clamps from the default standard
setting of 11.2 to 110 mm. Similar results for various heights are
shown in Figure 5.18 using the same commercially produced C-flute
board used for Figure 5.17.
The data of Figures 5.17 and 5.18 indicate that a WC 4226C

test piece height of 60 mm or less ensures an ECT will not be
influenced by beam buckling provided the ends are embedded in
resin platforms or are held in the clamp fixture. Throughout this
discussion, ‘free span’ is the height of the board that is free of
109
epoxy resin or is not in contact with the jaw faces of the TAPPI
T 839 method fixture.
11
10
Buckling load
9
ECT (kN/m)
5
0 20 40 60 80 100 120 140
Height (mm)
Figure 5.17 ECT versus the span length for C-flute specimens with
the horizontal edges embedded in resin. The dashed curve is the
calculated buckling load which indicates the height at which beam
bending is expected to dominate the failure mode. Reproduced
with permission from R.E. Popil, BioResources Journal, 2012, 7,
2, 2553. ©2012, North Carolina State University [9]
The error bars in the clamped method data shown in Figure 5.17
are somewhat higher than the resin-embedded end method results
shown in Figure 5.18. This is probably partly attributable to some
degree of compression of the edges during the clamp method versus
resin-embedded ends, and possibly some lack of true parallelism
of the clamps with the progressive insertion of machined block
assemblies to increase the spacing between the clamps. The expected
ECT for this board from calculation is 8.9 kN/m, which is consistent
with measurements for spans less than 60 mm using both methods.
110
Observations of lower than expected values of ECT at very short

heights, reported by McKee in 1961, were probably an artefact.
10
9.5
Buckling load
9
ECT (kN/m)
8.5
7.5
6.5
0 20 40 60 80 100 120
Height (mm)
Figure 5.18 ECT as a function of the free span height for C-flute
specimens using a modified Sumitomo clamp to extend the test
specimen unclamped height to 110 mm. The dashed curve is
the buckling load using the measured D and shear stiffness.
Journal, 2012, 7, 2, 2553. ©2012, North Carolina State
University [9]
5.5.2 E-, F- and N-Flute Boards
Compression strength measurements for smaller flute boards allow

calculation of the potential stacking strength for storage or transport
applications. Miniflute boards are replacing the use of folding cartons
111
in many instances. There are no accepted standards for compression

measurement in this case, so it is useful to qualify a current ECT
method for miniflutes. Samples of commercially made E-, F- and
N-flute board blanks were cut to various lengths and tested in a
similar fashion to that described above for the C-flute board. These
boards all had a liner of 207 g/m2 and fluted medium of basis weight
112 g/m2.
The results for resin-embedded edges of E-flute specimens are shown

in Figure 5.19. A fall in ECT occurs with increasing free span height,
as expected from increased bending occurring at larger heights.
Calculations for the buckling-load curve in Figure 5.19 are the same
as for C-flute but using DCD and PS values for E-flute from Table 5.3.
9.0
Buckling load
8.0
ECT (kN/m)
7.0
6.0
5.0
4.0
5.0 15.0 25.0 35.0 45.0 55.0 65.0
Height (mm)
Figure 5.19 ECT as a function of the free span length for resin-
embedded edges of E-flute test specimens. Reproduced with
permission from R.E. Popil, BioResources Journal, 2012, 7, 2,
2553. ©2012, North Carolina State University [9]
112
Figure 5.20 shows the ECT versus free span height for Sumitomo
clamped E-flute board. Figures 5.19 and 5.20 suggest free spans of
25 mm or less will suffice to provide a representative compression
strength value for E-flute. The default free span for a 50.8 mm
Billerud cut square test piece in the TAPPI T 839 method is 11.2 mm;
therefore, the TAPPI T 839 method can be expected to provide a
reliably accurate value for this single-wall E-flute board.
9
8.5
8
7.5 Buckling load
ECT (kN/m)
7
6.5
6
5.5
5
4.5
4
0 10 15 20 25 30 35 40 45
Height (mm)
Figure 5.20 E-flute specimens cut to various lengths and tested in

a Sumitomo clamp modified with extensions (dashed curve is the
calculated buckling load). Reproduced with permission from
R.E. Popil, BioResources Journal, 2012, 7, 2, 2553. ©2012,
North Carolina State University [9]
A similar study was conducted employing 42-23F and 42-23N

microflute boards using the TAPPI T 839 method clamp only and set
113
to provide various free spans. The data shown in Figure 5.21 display
a plateau in ECT at a height of 9 mm or less.
8 7
7.5
7 6
Buckling load Buckling load
6.5
5
ECT (kN/m)
ECT (kN/m)
6
5.5 4
5
3
4.5
4 2
3.5
3 1
5 10 15 20 25 30 35 5 10 15 20 25 30 35
Height (mm) Height (mm)
a) b)
Figure 5.21 ECT versus free span height using the TAPPI T 839
method clamping fixture for 42-23F (a) and 42-23N (b) board.
Dashed curves represent calculated vertical buckling loads.
Journal, 2012, 7, 2, 2553. ©2012, North Carolina State University
[9]
The plateau regions of constant ECT values in Figure 5.21 appear

to be overestimated by the buckling condition, i.e., P > ECT,
Equation 5.21, with the shear rigidity for E-flute as an approximation
for F- and N-flutes. Figure 5.21 indicates that the TAPPI T 839 clamp
method can be used for F- and N-flutes if the free span does not
exceed 9 mm. This can be easily achieved by cutting the microflute
samples to a height of 47 mm or smaller prior to their insertion into
the TAPPI T 839 method clamping fixture.
The data presented here augment and substantiate previous

observations of the effect of specimen height on ECT [16–20].
ECT values will decrease with increasing specimen height due to
114
specimen buckling rather than specimen compression. Specimen

heights for optimal ECT measurement fall approximately within the
limits estimated by sandwich structure buckling. The implication is
that the beam mechanics shown here can be applied to other board
configurations to estimate an appropriate specimen test height once
a measurement or a calculation of the D and measurement of shear
rigidity are available.
5.6 Facing Buckling during Board Compression Testing

If corrugated boards undergoing a vertical compression test are
observed closely, quite often the facings exhibit a pattern of surface
dimples [21]. These are actually sinusoidal out-of-plane deformation
waves extending along the glue lines resulting from panel buckling
of the facings [22] due to the applied vertical load. Whenever the
compression strength is greater than the load required for buckling,
i.e., the critical buckling load, the linerboard facing will buckle in a
wave pattern. To a good approximation, the critical buckling load
for the facings are given by the formula for a simply supported panel
of width ‘w’ and MD and CD bending stiffnesses ‘D1’ and ‘D2’:
Pcr = 4r D21 D 2
2
(5.22)
Kw
‘K’ is a factor to account for the constraint of the plates, which = 1 for
the case of a simple support, i.e., the edges are free to rotate. So, the
criterion for linerboard buckling to occur stated more specifically is
that the CD SCT must be greater than ‘Pcr’, much the same criterion
previously discussed for boards not to have bending expressed in
Equation 5.21. The critical buckling load for the linerboard facing
is calculated using the flute glue line spacing, e.g., it is 8 mm for
C-flute and the geometric mean of the bending stiffness √(D1D2).
The constant K in Pcr accounts for whether the panels are simply
(= 1) or rigidly supported (= 2).
115
D is measured using the Taber instrument of L&W with two-point

bending measurement. When using the Taber instrument, recall that
the instrument read-out values must be multiplied by 0.164 to convert
the read-out of bending moment of grams-force to D in mN-m.
A patterned dimpling of linerboard facings is commonly observed

in corrugated fibreboard containers that have been placed under a
vertical-stacking load and exposed to a high-humidity environment.
The same dimpling or localised buckling is also observed to occur
during the ECT of large-flute (A or C) lightweight corrugated boards.
A time sequence of video recordings of the ECT of various boards
were made with the frames synchronised with load-displacement
data. An example of a few selected frames from one such sequence
is shown in Figure 5.22, which tested a WC 3526C board mounted
in supporting resin platforms.
a) b) c)
Figure 5.22 Excerpted time sequence video snapshots of a glancing

angle illuminated ECT test of a WC 3526C-flute board:
a) 5.7 kN/m load, 1.3% strain, some buckling is evident,
b) peak load 6.7 kN/m, 1.6% and c) postfailure 4.5 kN/m, 2.1%,
crease forms joining microplate buckling crests. Reproduced with
permission from R.E. Popil, BioResources Journal, 2012, 7, 2,
2553. ©2012, North Carolina State University [9]
Glancing angle collimated illumination from the right-hand side in

Figure 5.22 was used to highlight features on the linerboard. Visible
116
dimpling of the facing is observed to progressively form as the load

increases near to the peak value. Once the peak load is reached, a
crease immediately forms which progressively develops joining the
crests of the buckled areas across the width of the sample. For boards
with a small flute size (B or E) or heavyweight linerboard facings,
no buckling patterns are observed, only the formation of a crease
at peak load.
The fluted medium is assumed not to buckle due to the additional

stiffness imparted by its curvature. Numerical simulations of ECT,
using the material constants for the laboratory-made boards used
in a study at the author’s laboratory, were made using a non-linear
constitutive model geometric non-linear effects [23]. Patterned out-
of-plane dimpling occurs in the simulations, which corresponds to
experimental video observations. The magnitude of the out-of-plane
displacement of the buckling pattern of the 205 g/m2 linerboard
facings is as much as 0.3 mm in the simulations. The computations
indicate no corresponding buckling for the 127 g/m2 medium, as
indicated in Figure 5.23.
a) b) c)
0.30
0.18
0.06
-0.06
-0.18
-0.30
Figure 5.23 Progressive out-of-plane displacement maps, a) to c),

for the C-flute medium from a non-linear numeric finite element
analysis (FEA) ECT simulation by Haj-Ali and co-workers.
Journal, 2012, 7, 2, 2553. ©2012, North Carolina State
University [9]
117
Figure 5.23 shows the progressive calculated out-of-plane

displacements for the medium for C-flute simulation data presented
in [23]. The numeric results indicate that the fluted medium does not
buckle like the facings during a compression test. In Figure 5.23a,
the failure initiates with the outer facing (not shown in Figure 5.23)
starting to show a patterned buckling and a Tsai-Wu failure criterion
(a summation of directional stresses exceeding yield stresses) is met
locally corresponding to the start of a formation of a crease, however,
the medium (Figure 5.23a) shows no patterned buckling. Figure 5.23b
is the calculated out-of-plane displacement at the ultimate failure
stress level, which is determined by the Tsai-Wu criterion linking
several failed regions across the width of the board. At this point in
the simulation, the FEA model of Haj-Ali and co-workers [23] shows
a prevalent predicted patterned buckling for the facings at failure
resembling Figure 5.22b and Figure 5.22c, however, the medium in
Figure 5.23 shows negligible buckling. A third calculation result is
shown in Figure 5.23c postfailure, where the vertical displacement
is 2X the failure initiation displacement.
Video recordings were also made of laboratory-prepared A-flute

127 g/m2 test pieces adhered to clear plastic film and subjected to
vertical load to see if any fluted medium buckling occurred. In this
case, calculations indicate that the buckling load for the flat portion
of the fluting is 12.6 kN/m and for the curved portion, of radius
1.5 mm, the calculated buckling load is 209 kN/m. These calculated
buckling load values are much larger than typical single-wall A- and
C-flute ECT values. Indeed, the video recordings of A-fluting adhered
to plastic film test pieces showed no buckling prior to a compression
crease failure, which always coincided with plastic film creasing.
Hence, video observations, and analytical and numerical simulations

all indicate that the 127 g/m2 medium will not buckle in ECT, but
the linerboard facings of 205 g/m2 will buckle, provided that their
compression strength exceeds their panel buckling load for a given
flute spacing. If the linerboard facings buckle, it is expected that their
ultimate strength will be governed by the empirical expression for
plate failure load, ‘Pz’:
118
Pz = cP mb P cr1 - b (5.23)
where, per unit length, ‘Pm’ is the intrinsic compression strength of

the material, ‘Pcr’ is the critical buckling load given by Equation 5.22,
and ‘c’ and ‘b’ are empirical constants. Equation 5.23 is the basis for
the derivation of the McKee equation for the compression strength of
boxes consisting of linked buckling panels with Pm being the ECT of
the corrugated board. By similar reasoning, ECT consists of buckling
linerboard plates for which the appropriate Pm is now taken as the
SCT. Thus, for a single-wall board, the model for ECT, which is
considered as the equivalent of Pz, is written as:
(5.24)

Application of Equation 5.24 to multiwall boards is straightforward

by the addition of similar terms to account for additional board
components and different medium take-up factors α.
A large series of commercial and laboratory-made corrugated boards

were considered for their buckling potential, calculated simply as the
ratio of the buckling load of the interflute linerboard panel to the
compression strength. In Table 5.4, the relevant physical properties
of the components were measured: caliper, basis weight, CD SCT,
MD and CD Taber stiffnesses.
119
Table 5.4 Summary of various corrugated board properties used in the study of buckling effects on ECT. The last column
120
is Pcr/SCT which when ≤1 indicates possible linerboard buckling
Board sample ID Board description Board Basis SCTCD Liner SCTCD Taber RMS TAPPI T Buckling
caliper weight liner density medium moment 839 ECT load/
(mm) (g/m2) (kN/m) (g/cm3) (kN/m) (gf-cm) (kN/m) SCT
WC 3526C 35 Commercial C-flute 4,140 542 3.9 0.62 2.1 25.4 7.5 0.72
WC 3526C 35 Same as above but lab 4,100 583 3.9 0.62 2.1 25.4 8.4 0.72
IPST made
WC 3526E 35 Commercial E-flute 1,620 521 3.9 0.69 2.1 25.4 8.1 3.68
WC 4226C 42 Commercial C-flute 4,210 605 3.4 0.68 2.1 31.4 7.9 1.02
IPST made
WC 5626C 56 Commercial board 4,350 748 5.3 0.84 2.1 65.7 11.2 1.34
IPST made
GP 262C 26 Commercial C-flute 3,838 420 2.4 0.68 2.1 8.5 5.3 0.38
TI 4223E 42 Commercial E-flute 1,691 581 3.8 0.72 1.7 27.0 8.0 3.99
TI 3523C 35 Commercial C-flute 4,057 547 3.7 0.70 2.4 20.2 7.8 0.60
TI 5623C 56 Commercial C-flute 4,207 751 5.4 0.70 1.8 63.8 10.8 1.29
IPST 4226C 35 Asymmetric lab board 4,214 580 4.6 0.70 2.4 42.2 8.6 0.95
IPST 4226C 55 Asymmetric lab board 4,443 687 4.6 0.70 2.4 42.2 10.6 0.95
IPST 4226E Lab-made multiwall board 5,679 1,019 4.6 0.70 2.4 42.2 16.3 0.95
4226C 42
IPST 4226A 42 Lab-made A-flute 5,171 624 4.6 0.70 2.4 42.2 8.5 0.81
IPST 4226B 42 Lab-made B-flute 3,044 599 4.6 0.70 2.4 42.2 9.9 1.64
IPST 4226A Lab A/B-flute multiwall 7,933 1014 4.6 0.70 2.4 42.2 14.9 0.81
4226B 42
IPST 4226A Lab A/C-flute multiwall 9,015 1011 4.6 0.70 2.4 42.2 14.9 0.81
4226C 42
IPST 4226B Lab B/C-flute multiwall 6,959 1026 4.6 0.70 2.4 42.2 15.3 1.64
4226C 42
IPST 4226C 42 Lab-made B-flute 4,252 613 4.6 0.69 2.4 42.2 9.5 0.96
IPST 4226C 42 Lab-made B-flute 2nd set 4,215 588 4.6 0.69 2.4 42.2 9.6 0.96
IPST 4226E 42 Lab-made E-flute 1,660 582 4.1 0.70 2.3 33.4 7.9 4.58
trial
IPST 4226E 42 Lab-made E-flute 2nd set 1,642 591 4.6 0.70 2.4 42.2 10.5 5.16
N&W 100 low Lab-spliced handsheet 4,063 421 1.8 0.40 2.4 20.4 4.7 1.18
press
N&W 100 med Lab-spliced handsheet 3,958 420 2.1 0.52 2.4 11.4 4.5 0.56
press
N&W 100 high Lab-spliced handsheet 3,956 427 2.6 0.62 2.4 2.1 4.7 0.09
press
press
press
press
press
121
122
press
press
N&W 100 100 Lab-spliced handsheet 4,357 390 1.5 0.44 2.4 8.9 4.5 0.61
psi
psi
N&W 100 1,900 Lab-spliced handsheet 4,507 390 2.2 0.80 2.4 2.3 4.5 0.11
psi
psi
psi
N&W 160 1,900 Lab-spliced handsheet 4,025 535 4.0 0.85 2.4 13.3 7.0 0.34
psi
Formette 100 Lab-spliced handsheet 4,105 410 1.1 0.29 2.4 17.4 3.9 1.69
0 psi
50 psi
0 psi
50 psi
N&W: Noble and Woods
Reproduced with permission from R.E. Popil, BioResources Journal, 2012, 7, 2, 2553. ©2012, North Carolina State
University [13]
Table 5.4 is a summary of various corrugated board properties used

in the study of buckling effects on ECT. The last column is Pcr/SCT,
which when ≤1 indicates possible linerboard buckling.
One series of boards produced on a pilot single-facer corrugator

contained embedded-spliced handsheet linerboards of various basis
weights pressed to various densities. These are designated by the
N&W prefix and the Formette prefix in the sample ID in Table 5.4.
This is similar to the investigation previously reported by Whitsitt
in 1988, where linerboard handsheets of various basis weights
were pressed to different densities to obtain a range of increasing
compression strengths. Generally, increasing sheet density through
wet pressing will increase the compression strength while decreasing
the caliper. The furnish composition of the linerboard and medium
were constant throughout the series of prepared handsheets. The
prepared linerboard handsheets were individually manually spliced
into conventional linerboard rolls that were run through the pilot
single-facer corrugator at IPST using conventional corrugating
operating parameters. Double-backing of the prepared single-face
combined board samples was performed manually using a metered
rolling nip, applying Stein–Hall starch adhesive to the flute tips
of single-facing samples, followed by a hot plate press adhering
linerboard sheets to the glued single-wall.
A second series of boards, designated by the IPST prefix or suffix in

the sample ID in Table 5.4, was prepared using commercial linerboard
and medium of various flute sizes and combinations utilising the
IPST pilot-plant facilities, as described in detail by Schaepe and
Popil [24]. In this second series, the linerboard and medium were
the same throughout and exhibited the properties of the C-flute
board data shown in Table 5.4. The ratio of buckling load Pcr to
SCT shown in the last column in Table 5.4 shows that many of the
boards investigated could be expected to display interflute buckling
whenever the ratio is less than 1. For the case of multiwall board
samples, a single-ratio value is reported in the table column for the
facing expected to buckle. Video recordings of the ECT testing of
these boards confirmed the prediction of Pcr Equation 5.22 with K
123
set to 1 for buckling onset. Specifically, Equation 5.22 predicts that

A- and C-flute boards will buckle when placed under vertical load,
but B- and E-flute boards will not. The smaller flute size, ‘w’, for B-
and E-flute board makes the ratio Pcr/SCT > 1 and so no buckling is
expected. Video recordings of the ECT of multiwall boards of both
single-face and double-back sides simultaneously [21] showed the
separate development of interflute buckling or compression failure
depending on the flute size adhered to the specific facing.
A third series of boards selected for ECT buckling investigations

consisted of commercial linerboards and medium supplied directly
from the corrugator as unconverted cut sheets. The linerboard and
medium for these boards were provided separately for their physical
property measurements.
The analysis is divided into the types or classes of boards. In each

case, a multiparameter fit for the constants C, C', K and b was applied
to the sets of data using the model represented by Equation 5.24.
Fitting was performed by iteratively reducing the error between the
model and actual data using numerical routines such as the Solver
algorithm in MS Excel.
Figure 5.23 shows the comparison of predicted ECT values from

the model to actual ECT values for the series of laboratory-made
boards using spliced handsheets. In cases where buckling of the
facings occurs, the model places the predicted adjusted values closer
to the identity line, reducing the error. Constants C', C, b and K were
determined to be 0.72, 0.7, 0.65 and 0.96, respectively. K being close
to 1 supports the model of linerboard plates between glue lines being
simply supported at the unloaded vertical edges of the glue lines. The
regression r2 value increases from 0.91 to 0.94 when the buckling
model, Equation 5.24, is used instead of Equation 5.18, and the mean
error correspondingly reduces from 0.19 to 0.16 kN/m.
Buckling of the liners limits the compression strength of the

corrugated board. Papermakers strive to produce liners with the
highest possible compression strength at the lowest possible basis
124
weight. Large amounts of refining energy and high-pressure wet

pressing produces sheets of high strength but at a lower caliper,
which will result in a lower D, although the elastic modulus will
have increased due to the increased sheet density. Boards prepared
and tested to produce Figure 5.24 using spliced handsheets were
examined for the relationship of caliper, compression strength SCT
and D as a result of increasing wet pressing.
13
SCT sum model, r2 = 0.91
12 Buckling model, r2 = 0.95
11 Identity line
Predicted ECT (kN/m)
10 N&W 200 high press
8
7 N&W 300 low press
N&W 100 high press
6
5
4
3
3 3 3 3 3
Actual ECT (kN/m)
Figure 5.24 Comparison of the buckling model versus actual ECT

values for a set of boards made with spliced handsheets pressed
to high densities. Reproduced with permission from R.E. Popil,
BioResources Journal, 2012, 7, 2, 2553. ©2012, North Carolina
State University [9]
The controlled wet pressing of handsheets allowed a large range

of densities to be prepared such that relationships between the
SCT, √D1D2 and basis weight could be determined, as shown in
Figure 5.25.
125
Geometric mean bending stiffness (gf-cm)

1 400
1 100 gsm 350
160 gsm
1 200 gsm 300 100 gsm
300 gsm 200 gsm
1 250
SCT (kN/m)
300 gsm
1 200 160 gsm
1 150
1 100
1 50
1 1
0.35 0.45 0.55 0.65 0.75 0.85 0.95 0.35 0.45 0.55 0.65 0.75 0.85
Density (g/cm3) Density (g/cm3)
Figure 5.25 Relationships between SCT (left) and geometric mean,

D (right), versus density for the handsheets used in the corrugated
board test results shown in Figure 5.24
Previously shown by Whitsitt [25], and evident in Figure 5.25,

increasing the compression strength SCT through wet pressing
decreases the D and ultimately places a limit on the ECT that may
be gained through wet pressing. Increased wet pressing results in
higher density and hence SCT increases, in addition, from first
principles, more fibre bonding may be expected [2]; however, the Sb
simultaneously decreases as a result of reduced caliper. Therefore, the
increase in SCT due to a wet pressing density increase is compromised
by a corresponding decrease in D and this phenomenon is accounted
for in the buckling model (Equation 5.24) for ECT. The respective
values for SCT and √D1D2 for each basis weight sample set (100,
160, 200 and 300 g/m2) was substituted into the buckling model with
fitted constants C and b to produce a series of predicted ECT values
for each basis weight class as a function of the wet pressed density.
Figure 5.26 summarises these calculations which show optimal
linerboard densities are reached for maximum ECT in lightweight
linerboards (<200 g/m2) adhered to 126 g/m2 C-flute medium.
126
9
200 gsm
160 gsm
Predicted ECT (kN/m) 8 100 gsm
3
0.35 0.45 0.55 0.65
Liner density (g/cm3)
Figure 5.26 Summary of calculations for predicted ECT using the

buckling model (Equation 5.24) and fitted constants derived from
the handsheet data set of Figure 5.24
Figure 5.26 shows how once a certain density is exceeded, gains

in SCT become offset by losses in D with the result that ECT does
not increase. Thus, there is a limit to the benefits of wet pressing
to increase SCT as, when incorporated into the corrugated board
structure, the increase in compression strength becomes offset by
the decrease in buckling load.
Application of the buckling model to predict ECT was also applied

separately to the laboratory-made board sets (IPST prefix in
Table 5.2), as well as the commercially made board set of Table 5.2.
Whenever SCT > Pcr, the predicted ECT values were calculated using
the buckling model, otherwise the linear summing SCT calculation
was made. The summary of the analyses listed in Table 5.5 shows
that in all three sets of boards investigated, the overall error in ECT
prediction [mean square-root error (MSE), last column] is reduced
and the correlation with actual values improves.
127
Table 5.5 Summary of SCT sum and buckling model fits for
several corrugated board data sets. MSE is the mean square-
root error for each fitted model
Data set ECT model C or C' b K r2 MSE
(kN/m)
All IPST lab- SCT sum 0.70 – – 0.91 0.19
made board Buckling 0.72 0.65 0.96 0.94 0.16
Commercial SCT sum 0.77 – – 0.90 0.18
boards Buckling 0.80 0.71 1.16 0.97 0.09
IPST-spliced SCT sum 0.70 – – 0.91 0.16
handsheet boards Buckling 0.65 0.85 1.18 0.95 0.14
5.7 Compression Strength of Boxes – McKee Formula

The compression strength of boxes [26] is considered to be the
combined compression and buckling of four joined panels so that
the general formula applies again:
Pmax = a Pcom b
c m
Pcr Pcr (5.25)
Such that Pmax here for a panel of length ‘L’ is the box compression
test strength (BCT)/L, Pcom is the ECT (kN/m) of the corrugated board
and Pcr is the buckling load (N-m) of the panels of the box given by:
4r ^ D 1 D 2h
Pcr =
L2 (5.26)
128
So that the BCT becomes:
1-b
BCT = C ECT b c D 12D 2 m L

L (5.27)
The box is assumed to be square such that its perimeter ‘Z’ is simply
4L. The exponent ‘b’ turns out to be ¾ so that the compression
strength formula takes the form:
(5.28)
A further simplification occurs with the assumption that ECT is

proportional to the Sb of the liners E1,2 × t, ‘h’ being the board caliper,
using the sandwich approximation for the board D:
E 1,2 2
D 1, 2 , th
2 (5.29)
The simplified form of the McKee equation as used in trademarked

CAPE and TOPS computer-aided design models (commercially
available box and pallet design software packages) is:
BCT = C l ECT # h # Z (5.30)
The assumption and substitution of D1 and D2 by ECT, thus removing

the power dependence of BCT on ECT, ignores the contribution of
the medium to ECT. Thus, predictions that are more accurate can
be made using the formula using both ECT and D.
An example of improved accuracy in predicting BCT is presented

for 15 different commercially prepared boards provided by the same
129
box plant. The samples represent a variety of heavyweight single-wall

boards and lightweight double-wall boards. All boxes had the same
dimensions L × W × D of 24 × 12 × 18 inches. ECT was measured
using the TAPPI T 839 clamp method, D in the MD and CD using
the four-point method [27], and caliper with a digital micrometre
accurate to 1 μm.
Figure 5.27 shows that the simple McKee formula overestimates the
actual McKee, using the full formula McKee brings predicted values
closer to the actual. The mean error using the full formula is 0.36 kN,
whereas using the simplified formula the mean error is 1.12 kN.
Actual BCT
7 McKee BCT
Simple McKee
6
BCT (kN)
2
te
l
t
al
l
ft
ft
B
lu
af
al
al
al
al
B-
ra
a
w
’-f
w
w
kr
kr
lk
e
e-
e-
e-
ft
‘x
e
bl
ith
ith
ra
bl
gl
bl
bl
al
ou
-B
in
tk
ou
ou
ou
w
tw
w
tB
td
ts
e-
td
gh
td
td
e
bu
gl
ut
gh
gh
gh
ei
gh
gh
gh
in
-fl
-B
ei
ei
w
ei
ts
ei
ei
ei
tC
w
w
w
tB
m
w
w
gh
vy
vy
vy
gh
iu
ht
ht
gh
ei
ea
ea
ea
ed
lig
ig
ei
iu
w
ei
-H
-H
-H
-L
M
ed
w
h
er
ht
ig
id
-M
J-
up
G
A
-H
ig
M
-S
-L
I-
H
B
Figure 5.27 Comparison of actual BCT and predicted BCT values

using the full and simple McKee formulas for boxes made from
the corrugated boards listed in Table 4.4 (B-B: multiwalled board
consisting of 2 B-flute medium layers)
130
5.8 Summary
The compression strength of paper, board or boxes will combine
buckling with compression failure. The SCT test method eliminates
buckling from its short-test span, however, this introduces a
comparatively high variability in the results which is attributable
to paper formation. RCT can be related to SCT using an empirical
compression buckling model. ECT will involve the buckling of the
entire sample if the height of the test specimen is too high, and will
display an out-of-plane buckling pattern on the facings in larger
flute (A or C) boards and lighter-weight liners. More accurate ECT
values can be predicted and attained if a compressive buckling model
is used. Box compression consists of combined compression failure
and buckling, and is a better prediction for BCT that agrees with
actual values obtained when both ECT and D measurements are used
in the full McKee model.
References
1. D.H. Page and R.S. Seth, TAPPI Journal, 1980, 63, 6,113.
2. P. Shallhorn, N. Gurnagul and S.Y. Ju, Nordic Pulp and

Paper Research Journal, 2004, 19, 2, 130.
3. D.H. Page, TAPPI Journal, 1969, 52, 4, 674.
4. TAPPI T 826: Short-span compressive strength of paperboard,

2013.
5. TAPPI T 822: Ring crush of paperboard (rigid support

method), 2011.
6. R.E. Popil in Proceedings of the TAPPI PaperCon 2010,

2–3th May, Atlanta, GA, USA, 2010.
7. J.M. Gere and S.P. Timoshenko in Mechanics of Materials,

Wadsworth, Belmont, CA, USA, 1984.
131
8. G. Lindblad and T. Furst in The Ultrasonic Measurement

Technology on Paper and Board, Lorentzen & Wettre, Kista,
Sweden, 2001.
9. R.E. Popil, BioResources Journal, 2012, 7, 2, 2553.
10. P. Shallhorn, S. Ju and N. Gurnagul, Journal of Pulp and

Paper Science, 2005, 31 3, 143.
11. R.S. Seth, TAPPI Journal, 1985, 68, 3, 98.
12. TAPPI T 811: Edgewise compressive strength of corrugated

fiberboard (short column test), 2011.
13. TAPPI T 839: Edgewise compressive strength of corrugated

fiberboard using the clamp method (short column test), 2012.
14. F.J. Plantema in Sandwich Construction, The Bending and

Buckling of Sandwich Beams, Plates and Shells, John Wiley
& Sons, Inc., New York, NY, USA, 1966.
15. R.E. Popil, D.W. Coffin and C.C Habeger, Appita Magazine
& Journal, 2008, 61, 4, 307.
16. R.C. McKee, J.W. Gander and J.R. Wachuta, Paperboard

Packaging, 1961, 46, 11, 70.
17. W.O. Kroeschell, TAPPI Journal, 1992, 75, 10, 79.
18. L.E. Eriksson, Containers International, 1979, 86, 8, 34.
19. L.E. Eriksson, Containers International, 1979, 86, 9, 64
20. C.J. Wilson, C.J. and B. Frank, TAPPI Journal, 2009, 8, 6,

24.
21. M. Schaepe and R.E. Popil, Corrugated International, 2006,

September, 3.
22. T.J. Urbanik, TAPPI Journal, 1990, 73, 10, 263.
132
23. R. Haj-Ali, J. Choi, B. Wei, R.E. Popil and M. Schaepe,

Composite Structures, 2008, 87, 4, 321.
24. R.E. Popil and M. Schaepe, TAPPI Journal, 2005, 4, 8, 25.
25. W.J. Whitsitt, TAPPI Journal, 1988 71, 12, 163.
26. R.C. McKee, J.W. Gander and J.R. Wachuta, Paperboard

Packaging, 1963, 48, 8, 149.
27. TAPPI T 836: Bending stiffness, 4 point method, 2016.
133
134
6
Testing Methods for
Measurement of the Writability
and Printability of Papers
6.1 Background
The current popularity of rollerball and fibre-tip pens, which use
liquid and gel inks, results in the penetration of inks through the
surface and into the sheet. Ink penetrating through pores in the
sheet to the other side is called bleed-through and show-through is
writing becoming visible on the unwritten side of the sheet, which
interferes with the readability of paper on the opposite side. This
issue can be ameliorated if there is high paper opacity (the inverse of
transparency) obtained by either high basis weight ‘b’ or the use of
high-scattering power fillers such as titanium dioxide (TiO2). Bleed-
through is the result of ink penetrating through the thickness of the
sheet and is affected by paper porosity, which is minimised through
paper densification via a combination of increased wet pressing and
high fines content and the use of sizing agents. Both these problems
can be reduced in severity by producing paper of higher sheet density,
lower porosity and surface roughness, and the addition of surface or
internal sizing agents. Show-through can also be limited at a given
sheet basis weight by having a high-scattering filler content such as
TiO2 embedded in the sheet.
This chapter describes the development of a test where the volume

and rate of application of a non-viscous liquid ink is applied in a
controlled and reproducible fashion to paper surfaces. The printed
samples are measured on their unprinted undersides for colour
changes attributable to the combination of bleed-through and
show-through. Quantification of the combined show-through and
bleed-through effects using a reproducible, controlled ink-delivery
135
system removes the subjectivity and arbitrariness of visual ranking of

handwritten samples using various types of pens and inks. A series of
physical tests related to paper surface properties and ink absorption
are performed on a sample set for the purpose of developing a
predictive model for the writability of paper.
An extensive review of the mechanisms and testing of the wetting

of paper surfaces by fluids is contained in the paper by Hubbe
and co-workers [1]. It is instructive, for a basic understanding, to
examine the Lucas–Washburn equation, Equation 6.1, which gives
the penetration length ‘L’ of a column of liquid in a pore or radius ‘r’
as being the square root of (surface tension ‘γ’ × cosine contact angle
‘θ’ × exposure time ‘t’ divided by ‘η’ viscosity). This is the governing
principle of ink transfer to paper, namely, ink is drawn into the paper
pores via capillary action:
r cos i c
L (t) = t
2 h (6.1)
Figure 6.1 is an illustration of the capillary effect, where the fluid

surface forms a meniscus with a contact angle determined by the
balance between the surface tension of the fluid and the surface energy
of the wall material. The term ‘√γ/η’ is the characteristic rate for liquid
penetration. Obviously, oil-based viscous inks, as found in ballpoint
pens, will penetrate much less than viscous inks such as in gel points,
tip markers or fountain pens. Time ‘t’ of ink exposure is governed
by the writing or printing speed and is offset by any evaporation or
drying. This equation forms the basis for assessing the writability
of paper as recently investigated in the MSc thesis of Rioux [2]. The
current objective is to describe a means of quantifying the writability
of various candidate papers by the measurement of show-through
and bleed-through under controlled conditions.
136
Testing Methods for Measurement of the Writability and Printability
of Papers
Ink
reservoir
L(t)
Figure 6.1 Schematic representation of a fluid being drawn into a

pore of radius ‘r’ by capillary action
6.2 Controlled Application of Ink: Using the Bristow

Wheel
The standard Bristow wheel apparatus documented by the American
Society for Testing and Materials, ASTM D5455 [3], and supported
by other literature dating back to 1967 [4], provides a means of
applying a swath of ink such that the length of the track is a measure
of the absorbency of the sheet on a timescale relevant to handwriting
with a pen. The apparatus consists of an accurately controlled
revolving wheel onto the periphery of which paper test strips are
applied. A slotted die containing a fixed volume of ink is applied
to contact the test strip as it approaches to produce a swatch of ink
coverage. Figure 6.2 shows a schematic concept of the Bristow wheel
ink application and Figure 6.3 is a photograph of the apparatus.
137
Ink holder slotted

bottom
Ink swatch
Paper sample
strip
Rotating wheel at
4.5 m/min
Figure 6.2 The Bristow wheel concept
Figure 6.3 Bristow wheel equipment
138
of Papers
A set of 16 commercial notebook papers were supplied by a

marketing organisation representing a range of quality and pricing.
Printed ruling lines run along the cross direction of paper (CD) of all
writing papers and this orientation was checked using the Lorentzen
& Wettre (L&W) tensile stiffness orientation (TSO) tester polar
plotting capability [5]. Strips of 25 mm width were cut along the
CD for use on the Bristow wheel, which were affixed at their ends
with adhesive tape. The wheel was set to maintain a linear speed of
4.5 m/min in keeping with typical fast handwriting speed used in
other documented studies [6]. The speed is checked by counting the
number of revolutions of the rotating drum during a time interval
using a stopwatch. Parker Washable Blue Quink™ fountain-pen ink
is measured by pipette to 75 µl volume and placed into the slotted
die (the slot is 1 × 15 mm). The die is manually lowered into contact
with the test strip once it moves into position to print.
The newly inked surface (which is the labelled side of each paper
pad) then passes by two 250 W infrared lamps to facilitate faster
drying of the surface and removal of the sample. The labelled side
of the sample set is the one that is facing the user, once the notebook
is opened. Fourdrinier paper machine papers have a felt facing side
and a corresponding wire side. The directional drainage of pulp
stock during this process produces a fines rich, somewhat denser felt
side and a fines depleted, less dense wire side. Liquid absorption can
therefore be expected to be side-application dependent. As per usual
laboratory practice, six test strips were printed from each sample
pad. Figure 6.4 shows the top side of the printed sample strips. Here,
the variation of the length of the track produced is a function of the
absorbency of the paper sheet. The variation in optical density or
darkness of the strips is dependent on the level of liquid absorption.
Intuitively, porous, high bulk, high roughness papers are expected
to be highly absorbent and thus unsuitable for writing or printing.
However, this expectation may be countered by the presence of sizing
agents or surface energy in a sheet, as the penetration of aqueous ink
into and through a sheet is governed by capillary action.
139
Figure 6.4 Top printed side of the notebook paper sample set using
the Bristow wheel
Figure 6.5 shows the undersides of the printed samples displaying

varying degrees of show-through and mottle. The variation in bleed-
through is dependent on the paper basis weight, density, porosity,
and possible presence of sizing and filler. Variation in show-through
is attributed to paper opacity, basis weight, sizing and filler. The fifth
sample from the right has a very dark and short inked area and a
large amount of show-through, indicating that it is highly porous and
is not a sized paper. By contrast, the fifth sample from the left has
a dark, mottled inked area and comparatively little bleed-through,
indicating the presence of a high amount of sizing agent, which
limits the penetration of ink from the surface into the bulk of the
sheet. What follows is how the show-through/bleed-through of the
undersides of the printed samples are quantified.
140
of Papers
Figure 6.5 Underside of the inked strips shown in Figure 6.4

displaying varying degrees of show-through.
6.3 Bleed-Through/Show-Through Measurement

The unlabelled side of the unprinted paper pads were measured for
Commission Internationale de l'Eclairage (CIE) Lab* coordinates ‘Lu*
au* bu*’, using a Technidyne Brightimeter S5 with a 0.5” diameter
measuring port as per the Technical Association of the Pulp and
Paper Industry (TAPPI) method T 524 [7]. Colour and shade can be
quantified in terms of these colorimetric values, which incorporates
the perception of colour by the human eye. In simple terms, ‘a*’
describes how ‘red’ a shade is, ‘b*’ how ‘blue’ it is and ‘L*’ how ‘light’
it is. Negative values describe the corresponding contrasting qualities,
namely ‘–a*’ is green in contrast to red, ‘–b*’ is yellow in contrast
to blue and ‘–L*’ how dark in contrast to how light. The scheme to
quantify colour and shade is shown pictorially in Figure 6.6.
141
LIGHTER
L*, a*, b* colour
space system +L*
YELLOW
GREEN
-a* +b*
-b* +a*
RED
BLUE
DARKER
Figure 6.6 Concept schematic for the CIE Lab* colour coordinate
system
For many writing and publishing papers, a low ‘b*’ value is desired
which can be obtained by the use of brightening agents or dyes in
the pulp. In mechanical pulp grades, such as used in newsprint or
softcover books, an increasing ‘b*’ is a measure of yellowing caused
by paper brightness reversion over time. Measurements in the TAPPI
T 524 instruments are made using specified directional wide spectrum
white light illumination at 45° incidence and the diffusely scattered
light emanating at 0° incidence is analysed spectrophotometrically
by passing the scattered light through a series of red, blue and green
filters onto a photodiode, shown schematically in Figure 6.7. The
signals obtained by a photodetector behind the alternating red,
green and blue filters are processed numerically by the so-called
tristimulus functions, which account for the perception colour biases
and sensitivities of the human eye. For example, the human eye is
142
of Papers
least responsive to detecting shades of blue, compared with green or

red, and this factor is incorporated into the numerical calculation for
L*, a*, b*. The geometry of 45°/0° is considered to emulate a typical
reading situation and is sensitive to the machine direction of machine-
made paper (MD) or CD orientation of the paper with respect to
the plane of light incidence. This method is now widely supplanted
by isotropic geometry [8], which makes use of an integrating sphere
so that measurements are not directionally dependent, however, the
illuminant, detection method and calculations remain the same.
Paper
Signal processor
45° and controller
Display
Incandescent
lamp Photodetector
Rotating
filter wheel
Figure 6.7 Schematic of the geometry used for colour measurement
The method is also widely used to measure the brightness of paper

samples, which is the reflectivity of paper at the blue wavelength of
457 nm using broad white light spectrum illumination. Brightness
is important in the marketing of publishing writing papers and is
sensitive to pulp-bleaching processes, the application of dyes or
optical brightening agents.
143
Figure 6.8 Photograph of the Brightimeter used for colour

measurements
The instrument shown in Figure 6.8 is allowed a 15 min warm-up

time after power-up to allow the tungsten halogen lamp to reach
full operating temperature, and is then calibrated using a brightness
standard tile prior to each use. The undersides of printed samples
were measured in three different locations along the length of each
strip, and the strips were backed by a pad of unprinted paper.
Optical measurements are all made along the CD of the test sample
in this TAPPI method. The average of three (Lab)* readings were
made for five strips for each sample and an average for the printed
144
of Papers
underside calculated: Lp*, ap* and bp*. The CIE Lab* coordinates were
also measured for the corresponding unprinted samples: Lu*, au* and
bu*. The change in colour and shade of the underside attributable to
printing of the labelled side is called delta ‘E’ (ΔE), thus:
DE = ^L *u - L *ph2 + ^a *u - a *ph2 + ^b *u - b *ph2

(6.2)
Equation 6.2 provides quantification of the bleed-through and show-

through from inking the samples. Readings were taken using pads
of five or more sheets of the unprinted paper, with backing to sheet
being measured to take into account the transparency of the paper,
which is usually measured as opacity and described in the following.
Opacity via method TAPPI T 425 [9] measures the percentage light
that is transmitted through a sheet by measuring the amount of light
reflected directly back from a sample when backed with a black-
absorbing background, compared with when a white background
is placed against the sample. A sample with high opacity will reflect
much of the light when placed against a black-absorbing background,
typically around 90% or more. Highly opaque samples are expected
to have small show-through when ink is applied on one surface. The
samples that have a high density or have a high amount of filler, such
as TiO2, will have a high opacity and therefore, will be less affected
by ink penetrating the surface in terms of show-through to the other
side. Considerable commercial research and development is devoted
to producing cost-effective paper-filler materials with a high visible
light-scattering capability to produce high paper opacity in order
to compensate for decreasing basis weight. A Technidyne BNL-3
Opacimeter measured the TAPPI opacity. A single measurement of
opacity requires two readings of the sample, one with the instrument’s
white standard background placed against the back of a sheet facing
the light source and optics, followed by another reading with the
light-absorbing black-body background placed against the sample.
145
The instrument calculates opacity as the ratio of the two reflected

light levels. Ceramic tiles with calibrated reflectance are used to check
the instrument performance after a 15 min warm-up at each use.
The brightness, colour, opacity, and so on, of all optical instruments
are subject to ageing of the halogen lamps, with spectrometer filter
characteristics also changing over time, and so require periodic
checking of their performance using brightness and colour standard
tiles. Some compensation for ageing can be made by adjustment
of electronic gain and offsets prior to the eventual replacement of
components.
6.4 Water Drop Contact Angle and Angle Change

Rate
For the selected sample set of writing papers, the liquid-absorption
properties play a principal role in determining the level of acceptable
ink-paper performance. Water drop contact angle and contact angle
change rate are introduced to characterise surface-liquid interaction.
The Lucas–Washburn equation (Equation 6.1) contains the term cos

(θ), which is the cosine of the meniscus angle and for a small water
drop, of a few microlitres, is also the contact angle the edge of the
drop forms with the contacting surface, as shown in Figure 6.9. The
contact angle is a measure of the wettability of the surface, i.e., a
hydrophobic low surface energy will have a high contact angle with
a water drop, whereas a highly wettable surface will exhibit a low
contact angle. The rate the contact angle changes as the water drop
is absorbed into the surface is another useful measure [10].
For the writability evaluation of pen on paper, the water drop contact
angle is known to relate to ‘ruling quality’ such that if the angle the
sides of the drop make with the surface are between 110° and 90°,
sharp lines are formed because the ink stays where it is applied.
Angles greater than 110° will cause applied lines of ink to break
up, whereas a contact angle of less than 90° will lead to ‘feathering’
or lateral spreading of the lines. The change in water drop angle
146
of Papers
with time is also known to be related to feathering propensity. The

delivery of a 4 µl drop, video recording of the water drop absorption
and measurement is facilitated by a Fibro PG-3 instrument [10],
shown in Figure 6.10. Other similar instruments for automated
analysis follow the same principles. Water drop contact analysis
can also be performed manually albeit tediously, using low-power
magnification to observe the droplet contacting the surface, as
described in the TAPPI method. The Fibro PG-3 instrument, also
known as the FibroDat pocket goniometer, dispenses a microlitre
drop onto a paper sample strip, and is equipped with a light source
and video camera connected to an analysing computer loaded with
the instrument software.
a) b)
Figure 6.9 Low (a) and high (b) contact angle water droplets. The
surface in a) is hydrophilic, the angle indicated is <90 ° and in b) it
is hydrophobic with a contact angle ≈90°
147
CCD
camera
Droplet
end with
dispenser,
output to
this is lifted
PC USB
up and
dropped in LED light
its holder to source
put a drop
on the
Pump
paper strip
actuator
switch
Water Droplet volume

reservoir selector
Figure 6.10 FibroDat pocket goniometer for analysing the water

drop contact angle (CCD: charge-coupled device, LED: light
emitting diode, PC: personal computer and USB: universal serial
bus)
Image analysis software is used to analyse the water drop once it

is placed onto the surface. The ‘dynamic’ mode setting allows the
recording of the drop becoming absorbed into the sheet surface. A
sample screen output is shown in Figure 6.11a and shows the screen
output at the instant the water drop comes into contact with a
copy paper sheet. Figure 6.11b shows the drop has been completely
absorbed and the contact angle, as a function of time, is displayed.
During the analysis of water drops on writing papers, the drops
are automatically recorded for 30 s after making contact with the
sheet surface. The table of contact angle versus time provided by the
software is analysed for slope of contact angle change or wetting
rate as Δθ/Δt. A calibration sphere comparable to the water droplet
in size is provided with the instrument to ensure, via a calibration
148
of Papers
procedure provided in the software and followed before each use,

that the angles and drop volume are correct.
a) b)
Figure 6.11 Sample screen outputs from the FibroDat pocket

goniometer for a water drop on standard copy paper. a) Droplet
on copy paper data being accumulated and b) final output: contact
angle starts at 59° and drops to 20° in 5 s
6.5 Cobb Water Absorption Test

Another test adopted from Swedish Technical Forestry Institute
(now Innventia) studies [11] of ink-jet paper characterisation is
water absorption (Cobb test) [12], as shown in Figure 6.12, where
a sample is exposed to a pool of water placed on its surface for 30 s
and the absorbed water is measured as a weight difference of the
blotted sample after water exposure. The weights of each test piece
are recorded to three decimal places on an electronic balance. After
30 s water exposure, samples are removed as quickly as possible
from the Cobb tester, placed between kraft pulp blotting papers and
subjected to two passes of a 20 kg metal roller prior to weighing. The
149
test piece weight difference before and after water exposure mutiplied
by 100 provides the absorbed water in g/m2. Papers with embedded
sizing such as rosin will have low Cobb values, i.e., resisting water
absorption, and may be expected to allow less penetration of ink
as well.
Figure 6.12 A Cobb tester set up ready to test for the amount of
absorbed water/area time. Details: 1) rubber pad underneath the
sample, 2) rubber O-ring on top of the sample, 3) screw down
steel ring tightly, 4) pour 100 cc of 23 °C deionised water and
5) start time
6.6 Caliper and Basis Weight

A dense sheet is expected to be more resistant to ink absorption and
penetration therefore, the density of the writing paper sample set is
estimated as basis weight divided by the caliper. It is customary to
use basis weight in units of g/m2 and caliper in units of microns with
the result of density then being in g/cm3.
An L&W digital caliper gauge was used to measure 10 repeats per

sample to 0.001 mm accuracy. Metal foils and shims of known
150
of Papers
thickness are used to check the instrument. The method specifies

flat parallel steel platens descending onto the specimen at a rate of
4.8 × 10-2 m/min providing, by weight, a pressure of 50 kPa over
an area of 2 ×10-4 m2. The weight and drop rate of the platen takes
into account that paper is a viscoelastic compressible material. The
diameter of the platen covers the typical topographic waviness of
most papers that arises from paper formation or mass non-uniformity.
This combination of platen dimensions and clamping pressure as
specified above provide somewhat of an overestimate of the caliper,
as the surface roughness and topography are not compensated for
in this technique.
Punched samples of 53 × 63 mm, as provided by the pneumatic L&W

Elmendorf Tear Tester sample punch, were measured to four decimal
places for basis weight and reported as g/m2. Balances are checked
by weighing calibrated balance weights. Caliper gauges are checked
using metal shim standards.
6.7 Air Permeability or Porosity

Low-density samples which are capable of absorbing a large
proportion of liquid are expected to have a high permeability to the
flow of air through the sheet. Samples that consist of many pores per
unit area, attributable to coarse fibres, low density or a combination
of both, will have a high air permeability and so will also probably let
a lot of ink through, which contributes to bleed-through and show-
through. Air permeability [13] is often synonymous with porosity
and is measured routinely in applications where liquid absorption
is important, such as the adhesion of fluted medium to linerboard
during corrugated board manufacture. Other applications require
high air permeability to allow paper webs to adhere well to rolls,
e.g., during high-speed printing. An L&W Densometer instrument
was used to measure air flow through the samples, which records
the number of seconds required for 100 ml of air, at a low pressure
of 1.2 kPA, to pass through the sheet. The instrument shown in
Figure 6.13 consists of a free-falling weighted cylinder containing
151
air suspended over a sealing clamping arrangement that allows air

to flow through the paper sheet. An optically triggered timer in the
instrument records the time required for the top cylinder to drop to
the level corresponding to 100 ml of air.
a) b)
Figure 6.13 The air permeability tester (a) and sample clamp detail
(b)
The instrument is checked by measuring the time taken for air to

flow through a calibrated metal foil orifice, which is supplied by the
manufacturer.
6.8 Surface Roughness by Air Leak and Contacting

Stylus Profilometer
Surface roughness is generally accepted to affect ink transfer during
contacting processes, as the ink film, of limited thickness, applied
152
of Papers
on the surface cannot reach deep pores. Roughness, related to ink

transfer, is commonly assessed indirectly by measuring the amount of
air leaking through contacting annular rings [1]. Air escapes through
the pores and surface gaps that do not make a sealing contact with
the ring cylinder end surface. This method is considered to measure
an integral of the volume of pores connected to the surface, based on
the geometry of the rings and air pressure difference assuming laminar
Poiseuille flow. Therefore, results can be misleading as different paper
topographies can produce similar leakage air flow, for example, a few
deep pores in an otherwise smooth surface may provide the same air
flow as a rough surface with many smaller pores. Nonetheless, the
method provides a quick and convenient measurement that is useful
for quality control in paper mills
The Sheffield version of the instrument has been applied to predict

letterpress solid print density and offset printability using the
geometry and parameters shown schematically in Figure 6.14, the
instrument is shown in Figure 6.15.
F = 100 kPa Air flow
47 mm Annulus 0.4
mm thick
Paper
Glass plate
Figure 6.14 Geometry of the Sheffield air leak roughness

measurement principle
153
Air flow
manometers
Flow adjustments
for calibration
Figure 6.15 Manometer and testing head of the Sheffield

roughness instrument
The instrument is calibrated prior to each use by adjusting the

manometers using the internal calibration air leaks. Pressure is
provided by the weight of the land assembly and the sample rests on
a smooth glass plate. Although the manometers read in units of ml/
min flow, the conversion to actual air flow is dependent on the range
used and is typically 7–9 times the actual manometer read-out, the
documented method provides a conversion table if actual air flow
is to be recorded.
An alternative more sophisticated form of the air leak method of

roughness measurement, considered to correlate with offset dot
quality, is the Parker Print-Surf (PPS) [15]. For this method, the
154
of Papers
thickness of the annulus is decreased from 380 μm in the Sheffield

method to 50 μm to correspond to offset printed image dot
dimensions. The applied pressure of 2 MPa and a soft rubber backing
of the instrument in the PPS method account for the compressibility
of the paper under test emulating offset printing conditions. Very
smooth surfaces, such as coated printing grades, are also evaluated
using a selected higher land pressure on the instrument and a hard
backing to extend the sensitivity to variations in the smoother range.
Very rough papers lie outside the range of the air leak methods, and
a contacting stylus profilometer provides a useful measurement in
these cases. In the author’s experience, profilometers intended for
use in machine shops, for evaluating ground machined metal surface
finishes, correlate well with the PPS S-10 values when evaluating a
commercial newsprint sample set with a range of surface roughness.
Historically, the Emveco 210-R [16], shown in Figure 6.16, was
developed to measure linerboard for corrugated boxes and its results
correlate well with flexographic print density mottle. The principle of
operation is similar to an audio vinyl record phonograph. A projecting
tungsten carbide or diamond stylus of radius 25 μm is contained in
a 4 N weight paper contacting skid, which is lowered onto a 20 cm
paper strip length cut along the MD or CD. Roughness values are
higher in the CD than the MD, especially for uncoated papers. The
strip is advanced at a speed of 10 mm/s by a drive motor and the
stylus with a load force less than 10 mN either protrudes into paper
pores or rises at protruding fibre crossings. Consecutive readings
are accumulated during a traversing scan and taken at intervals of
250 μm to produce 500 stylus vertical displacement readings along
the surface. A ‘microdeviation’ value is calculated which is found to
be sensitive to abrupt changes and less affected by the larger scale
wavy topography of paper, as in Equation 6.3:
499
microdeviation = 1, 000 # 1 / ^ x -x h2
n - 1 i= 1 i + 1 i (6.3)
155
The calibration uses shims and calculations are in units of thousandths

of an inch. Microdeviation values of 50 or lower are considered to
be excellent surfaces for subjective print quality, whereas values for
untreated linerboard surfaces can be as high as 400. The air leak
and profilometer instruments are checked for performance assurance
using a smooth (lightweight coated magazine) and a rough paper
sample (xerographic copy paper) with recorded historic values.
Figure 6.16 Stylus profilometer (Emveco 210-R) developed for the

prediction of linerboard printability
The PPS is comparatively more sensitive to high smoothness and

requires special attention to ensure that the annular ring lands and
the soft or hard rubber backings are clean and free of loose fibres
prior to use.
156
of Papers
6.9 Sizing Test Ink Penetration – Hercules Size Test

This method, i.e., TAPPI T 530, measures the time taken for a dark
dye solution to penetrate to the bottom of the sample once the
solution is applied to the top [17]. Sizing agents are often dispersed
into the wet-end stock chest of the paper machine to limit the
penetration of fluids through the sheet. The change in reflectance
of the bottom surface of the sample, which is placed on top of the
instrument, is measured to determine the stop time. 10 ml of dye
solution is poured into a sample holder, containing the paper sample,
and the instrument timer is started simultaneously. A photodiode
measures the reflectance of the underside of the illuminated test
sample. The timer stops once the reflectivity decreases to 80% of
the initial level, i.e., before the dye was poured. Naphthalene B dye
solution containing 1% formic acid was used for the Hercules size test
(HST), which was arranged to detect 80% of the initial reflectivity
level of the bottom of the sample surface; dye solution was applied
on the same labelled surface side which was inked by the Bristow
wheel on replicate samples. Different reflectance level settings or
higher concentrations of formic acid can be used depending on the
type of samples under investigation, however, the setting of 80% base
reflectivity and a concentration of 1% formic acid are the standard
values for evaluating writing or publishing papers which are sized
to produce a penetration time of around 180 s. The testing of high
basis weight samples may use a concentration of 10% formic acid
to limit the time required to reach a 20% decrease in reflectivity.
Prior to each use, the HST undergoes a warm-up period and a

reflectivity calibration check using the white and coloured ceramic
tiles provided by the instrument supplier. The calibration tiles, sample
holder and dye dispenser are shown on the left side of Figure 6.17.
The procedure ensures that the loss in reflectivity is truly as set by its
front panel reflectivity dial indicator, e.g., 80%. This is important to
check as the instrument uses incandescent lamps which age. The dye
is mixed in equal parts with 2% formic acid and is not kept longer
than six months.
157
a) b)
Figure 6.17 Ink dispenser, clamping sample holder and instrument

calibration tiles (green and white) a), and the HST tester b)
6.10 Results and Analysis

Plots of variables reveal relationships that can provide insight into
the governing mechanisms that affect writability. The sample set
investigated here is a commercial selection of notebook papers
representing a range of paper quality and different strategies for
limiting ink penetration to the opposite side of the sheet. High sheet
density, high basis weight and high surface smoothness, which limit
ink penetration physically, may also be balanced by other penetration
limiting mechanisms that arise from the use of high-scattering fillers
or internal-sizing agents or surface coatings.
158
Table 6.1 List of commercial notebook writing papers and their physical properties
Sample ID Visual Writability Caliper Basis weight Sheffield roughness Porosity
rank 2
ΔE c.i. μm c.i. g/m c.i. ml/min c.i. s/100 ml c.i.
D4 4 60.50 5.17 101.8 0.9 73.5 0.3 178.5 14.3 24.1 2.3
D3 4 18.55 1.17 76.2 0.2 57.5 0.2 124.4 22.9 17.1 1.4
D2 2 16.20 4.69 78.7 0.7 56.7 0.5 77.4 21.4 25.1 1.8
D1 2 15.72 1.33 77.5 0.6 55.2 0.6 144.8 19.7 21.1 1.6
C4 3 14.95 3.30 97.9 0.7 75.6 0.5 92 12.8 9.6 1.0
C3 4 14.71 4.51 72.5 1.1 57.1 0.9 56.7 7.4 35.6 3.4
C2 3 11.89 1.18 80.4 0.7 56.2 0.3 41.8 4.5 25.9 2.5
C1 2 10.47 0.83 97.2 1.3 74.8 2.6 147.3 21.4 11.4 1.3
B4 2 9.93 0.51 87.9 1.0 74.2 1.0 50.3 2.3 38.1 1.7
B3 2 8.98 0.46 75.7 0.8 56.5 0.6 97 15.2 19.8 1.8
B2 3 7.90 0.36 102.3 1.8 80.3 0.5 83.8 16.2 35 3.0
B1 1 7.90 0.36 75.5 0.3 51.6 0.2 102.5 8.1 3.9 0.2
A4 4 7.65 0.43 102.5 1.1 77.6 0.5 142.5 12.1 11.4 0.9
A3 1 6.35 0.60 76.1 0.6 58.9 1.0 100 14.4 10.2 0.5
A2 1 5.29 0.17 78.9 0.6 54.4 2.6 131 12.4 23 2.2
A1 1 5.24 0.61 96.8 0.5 80.5 0.2 47.2 6.6 33.8 1.1
159
of Papers
Table 6.1 lists the sample set with a subjective visual ranking of
the ink show-through on the underside, the writability quantified
as the change of colour, ‘ΔE’ from optical measurements, caliper,
basis weight, and air permeability or porosity. Apparent density of
the sheet is simply basis weight/caliper. Using units of g/m2 for basis
weight and microns for caliper, yields density in units of g/cm3. Good
writability, meaning a ‘1’ in visual rank or low ΔE, is expected to be
associated with a high density leading to a low Sheffield roughness
and low porosity.
Figure 6.18 shows that there is not a direct correlation between

surface roughness and sheet bulk, which is the inverse of sheet density.
A sample set consisting of the same furnish and pulp treatment is
expected to have increasing roughness with increasing sheet bulk,
as higher wet pressing along with dry-end calendering in paper
manufacturing both result in lower bulk and lower smoothness.
The scatter of the points in Figure 6.18 indicate that the relationship
between roughness and bulk is altered within the sample set through
using different pulp furnishes, fillers or coatings. Sheffield roughness
was historically used as a primary predictor of newsprint letterpress
solid print ink density. Rougher papers prevented the viscous ink film
on the printing plate from penetrating into the sheet surface, resulting
in solid print areas with many visually objectionable uninked pores.
Figure 6.19 shows porosity versus bulk and many of the samples
display the expected trend, i.e., showing a higher porosity with
increasing bulk. Sample B1, which is a notable outlier from the
trend, has low show-through but a high bulk and low porosity,
which suggests that it may be the result of coarse fibres that are
highly fibrillated due to refining. The fibre coarseness, which can
arise from using softwoods, leads to high bulk but the fines from
the fibrillation would impede ink flow-through and lower porosity.
In contrast, sample A1, a comparatively expensive premium writing
paper, also has little show-through, like sample B1, but has high
porosity at low bulk, which suggests that this sample is probably
hardwood that is not highly refined but the paper may also contain
sizing to limit fluid flow-through the sheet.
160
of Papers
200
180 y = 198.81x - 163.91
R2 = 0.161
160
Sheffield roughness
140
120
100
80
60
B1
40 A1
C2
20
1.100 1.200 1.300 1.400 1.500 1.600
Bulk (cm3/g)
Figure 6.18 Sheffield roughness versus bulk (1/density) for the

notebook writing paper set (roughness is in Sheffield units, bulk
has units of cm3/g)
40
C3
B4
35
B2
A1
30
Porosity (s/100 ml)
25
20
15
10
5
B1
0
1.15 1.25 1.35 1.45 1.55
Bulk (cm3/g)
Figure 6.19 Porosity (s/100 ml) versus bulk (cm3/g) of the

notebook writing paper sample set
161
Table 6.2 Measurement results for the commercial notebook paper set
162
Sample Opacity Hercules size Contact angle Angle rate Cobb30 Cobb30/basis wt
ID % c.i. s c.i. θ° c.i. Δθ°/Δt c.i. g/m2 c.i. g/m2/g/m2 c.i.
D4 90 0.6 0.5 0.1 56.4 3.0 19.98 0.022 87.3 2.5 1.188 0.084
D3 82.7 0.3 9.5 0.8 91.2 2.0 0.459 0.021 50.1 1.6 0.871 0.028
D2 88 1.0 14.3 2.7 104.5 7.8 0.779 0.035 56.3 1.6 0.993 0.029
D1 87 0.8 15.9 2.0 111.7 2.3 0.282 0.067 33.7 1.8 0.611 0.082
C4 90.4 0.4 5.9 1.3 89.8 3.2 1.302 0.027 79.5 4.1 1.052 0.054
C3 88 0.6 38.2 3.7 106.1 3.5 0.148 0.088 18.8 1.5 0.329 0.027
C2 90.2 0.8 22.5 3.8 108.5 4.2 0.188 0.024 43.9 6.0 0.781 0.107
C1 91.4 0.9 3.3 0.6 85.3 4.2 0.526 0.045 56.7 1.8 0.758 0.035
B4 90.1 0.6 6.7 1.4 78.2 2.0 0.459 0.044 65.4 2.4 0.881 0.035
B3 88.4 0.6 89.3 6.0 110 3.0 0.129 0.032 17.9 1.3 0.317 0.023
B2 92.8 0.2 42.7 5.4 96.3 1.6 0.336 0.051 37.9 2.6 0.472 0.032
B1 84.6 0.3 27.8 2.7 111.8 1.7 0.159 0.084 16.6 0.6 0.322 0.012
A4 90 0.3 23.7 6.6 89.3 4.2 0.231 0.080 45.9 2.9 0.591 0.07
A3 85.6 0.3 118.9 13.9 110.9 3.0 0.173 0.091 16.5 0.3 0.280 0.007
A2 85.3 0.6 110.9 7.4 112.8 3.0 0.103 0.190 14.4 0.7 0.265 0.018
A1 89.4 0.2 225.3 4.1 111.7 1.1 0.198 0.039 19.1 0.5 0.237 0.006
of Papers
The physical properties of the sample set when examined in relation

to writability indicate that the relationships are not straightforward.
Although physical properties are expected to influence the absorption
of ink flow into a paper, they cannot be used alone to predict
performance. Therefore, the absorption of fluid by the paper is
examined using various tests, as described earlier and presented in
Table 6.2.
For the data presented in Table 6.2, the liquid-absorption data shows
relationships with writability quantified by ΔE. Sample D4 has a ΔE
of 60, whereas the rest of the sample set range from 5 to 20 therefore,
Figure 6.20 and Figure 6.22 plots are semilogarithmic for vertical
scale compression.
Figures 6.20–6.22 show the HST, contact angle and the contact angle
change rate correlate with ΔE. HST and contact angle change rate are
measures of fluid flow into the sheet, whereas contact angle measures
how receptive the paper surface is to the aqueous ink.
20.00
∆E = 29.894(sec)-0.326
∆E
R2 = 0.698
2.00
0 50 100 150 200 250
Hercules size (seconds)
Figure 6.20 DE show-through versus HST values for the notebook

paper sample set
163
70.00
60.00
50.00
∆E = 55.22ln(θ) + 266.48
R2 = 0.61
40.00
∆E
30.00
20.00
10.00
0.00
50 60 70 80 90 100 110 120
Contact angle
Figure 6.21 Writability or show-through/bleed-through measured

as ΔE versus the contact angle of ink droplets
20
∆E
∆E = 16.67 (dθ/dt)0.397
R2 = 0.71
2
0.02 0.2 2 20
Contact angle change rate (deg/sec)
Figure 6.22 ΔE versus contact angle change rate for the notebook
writing paper sample set
164
of Papers
In combination, HST, contact angle and contact angle change rate

were used in a best fit model with least error in a linear multiple
regression of the form:
DE = a 1 x 1 + a 2 x 2 + g + a n x n + b (6.4)
The ‘regression’ function in the Excel Data Analysis Tools package

provides a convenient means of doing this. The significance of the
fit of the model variables is assessed by an overall high correlation
coefficient ‘r’, low statistical ‘p’ significance values for the ‘xi’ variable
coefficients ‘ai’ and a low calculated significance ‘F’ value. In this
sample set of notebook and writing tablet papers, the best regression
equation was determined to be:
β (6.5)
with an average error of ± 1.91 points. A comparison plot of the

model versus actual ΔE values is shown in Figure 6.23. A ΔE of 10
points or less is visually acceptable, the model allows a prediction of
writability without the need for a printing test. The equation shows
that good writability requires a low surface energy (high contact
angle), with high wettability corresponding to a high contact angle
change rate, good absorption properties (high Cobb/basis weight)
and sizing in the sheet (high HST).
165
70
60
50
Linear model
40
30
∆E model = 0.99 (actual ∆E)
R2 = 0.96
20
10
0
0 10 20 30 40 50 60 70
Actual ∆E
Figure 6.23 Comparison between the linear model and actual ΔE

values. Visually acceptable values of show-through have a ΔE
value of 10 or less
6.11 Physical Testing for Bank Cheque Ink-Jet

Printability
A different set of physical characteristics were found to be of
importance for the commercial high-speed ink-jet printing of magnetic
ink character recognition font (MICR) on bank cheques. In this
case, printed characters along the bottom margin, representing bank
account details, must remain precise and intact on the sheet to within
specific tolerances determined by MICR reading instruments, such as
the Delphax GTX or an RDM Corporation MICR qualifier reader.
An acceptable reading from these instruments is a reject value of
5% or less.
166
of Papers
If the printing is such that the characters become misaligned or

misshapen and result in a poor electronic signal when read, the
cheque has to be processed manually incurring delays and costs. Paper
properties must be tailored for optimal ink-jet print transfer such that
the ink will transfer where intended. Similarly, as was shown with
the notebook writing paper set, the objective is to determine what
physical tests will correlate best with printability measured here as
magnetic ink character percentage (MICR%).
Ink-jet inks in this application are iron oxide particles dispersed in

solvents, such as ethylene glycol, and dried using hot air impingement.
Surface energy and water-absorption properties that were found
to be relevant for notebook writability were found not to be of
consequence in this sample set. Data for the sample set is shown in
Table 6.3, which contains the respective MICR% values, surface
roughness in terms of PPS and Emveco 210-R stylus profilometry,
microdeviation and HST.
Table 6.3 Results of testing an ink-jet bank cheque

printability sample set
Sample ID MICR% PPS S-10 Microdeviation HST
μ c.i. ×10 inch
-3
c.i s c.i.
Best 0.67 5.42 0.22 59.69 5.2 69.4 20.4
Good 0.86 5.23 0.12 54.26 6.32 79.6 27.6
Bad 4.67 4.95 0.19 38.27 6.29 91.7 9
Worst 21 4.5 0.15 33.31 1.6 148.3 20.8
The plot of HST, PPS roughness and microdeviation in Figure 6.24

show relationships as expected. The roughness, measured by PPS and
microdeviation, falls with increasing MICR% misregister, indicating
that a rougher surface is better for printed character integrity. This
167
observation of a rougher surface leading to better ink-jet printability

is consistent with the conclusions of Lyne and Aspler [18]. Higher
HST values correspond to higher MICR% levels, indicating that
increased sizing agent is interfering with ink-jet MICR printing. A
multiple regression model combining the values from the tests shows
the best available predictor to be:
(6.6)
with an average error of +/– 0.07.
160 5.8
140 5.6
HST (sec.) or stylus microdeviation
Stylus 5.4
120
PPS S-10 (µm)
microdeviation
5.2
100 HST (seconds)
5
PPS roughness
80 4.8
4.6
60
4.4
40
4.2
20 4
0 5 10 15 20 25
% MICR
Figure 6.24 HST and surface roughness related to ink-jet

printability measured by MICR% misregister
168
of Papers
6.12 Summary
How paper reacts to applied liquids determines how writing or
printing will appear. Writability or printability can be quantified by
optical measurement of inked areas or the undersides of inked sheet
surfaces. As paper is comprised of layers of fibres, its porous nature
will affect fluid flow. The porosity of paper can be indirectly quantified
by measuring low pressure differential air flow through the sheet.
Internal-sizing agents limit fluid flow and in the case of low viscosity
aqueous inks, improve writability by confining applied ink largely to
the paper surface. HST along with the water contact angle relate well
to the degree of ink penetration in these cases. Surface roughness is
known to affect the transfer of ink films and is indirectly measured
by air flow along the surface of a sheet under compression by a stylus
profilometer. Linear regression models can be used to combine the
results of relevant physical tests to provide a best fit predictive model
relating to the printability. Ink show-through for writing papers was
found to be related to the combination of HST values, and level of
water absorption measured by contact angle change rate and Cobb.
Ink-jet printability of solvent-based MICR ink related best to surface
roughness and was adversely affected by the presence of sizing. Thus,
the measurement of relevant physical properties of paper sample sets
can be used to predict their performance for ink to paper applications.
References
1. M.A. Hubbe, D.J. Gardner and W. Shen, BioResources,
2015, 10, 4, 8657.
2. R.W. Rioux in The Rate of Fluid Absorption in Porous

Media, University of Maine, Orono, ME, USA, 2003.
[MSc Thesis]
3. ASTM Standard D 5455-1993: Short-term liquid absorption

into paper (Bristow Test), ASTM International, West
Conshohocken, PA, USA, 1993.
169
4. J.A. Bristow, Svensk Papperstidning-Nordisk Cellulosa, 1967,

70, 19, 623.
5. G. Lindblad and T. Furst in The Ultrasonic Measuring

Technology on Paper and Board, Lorentzen & Wettre, Kista,
Sweden, 2001.
6. J.M. Hollerbach, Biological Cybernetics, 1981, 39, 2, 139.
7. TAPPI T 524: Color of paper and paperboard 45°/0° geometry,

2013.
8. TAPPI T 527: Color of paper and paperboard d/0° geometry,

2013.
9. TAPPI T 425: Opacity measurement of paper.
10. TAPPI T 558: Surface wettability and absorbency of sheeted

materials using an automated contact angle tester, 1997.
11. a) E. Blohm and P. Åslund in Papers Designed for High Speed

Ink-Jet Printing, Swedish Technical Forestry Institute (now
Innventia), Report CW 248, 2004.
http://www.innventia.com/Documents/Rapporter/STFI-
Packforsk%20report%20cw248.pdf
b) E. Blohm in High Speed Ink-Jet Printing of Newsprint,

Swedish Technical Forestry Institute (now Innventia), Report
CW 119, 2005.
http://www.innventia.com/Documents/Rapporter/STFI-
Packforsk%20report%20119.pdf
12. TAPPI T 441: Water absorptiveness of sized (nonbibulous)

paper, paperboard, and corrugated fiberboard (Cobb test),
2013.
13. TAPPI T 460: Air permeability paper, 2006.
170
of Papers
14. TAPPI T 538: Smoothness paper and paperboard (Sheffield

15. TAPPI T 555: Smoothness paper and paperboard (Print-Surf

16. R. Xu, A. Pekarovicova, P.D. Fleming and V. Bliznyuk in

Proceedings of the 2005 TAPPI Coating Conference and
Exhibit, 17–20th April, Toronto, Ontario, Canada, 2005,
p.17.
17. TAPPI T 530: Size test for paper by ink resistance (Hercules-
type method) (Revision), 2002.
18. M.B. Lyne and J.S. Aspler, TAPPI Journal, 1985, 68, 5, 106.
171
172
7
‘Beater Curves’ to Evaluate Pulp
Potential: Burst, Tensile,Tear,
Opacity
7.1 Background
The potential of pulps for papermaking can be assessed by the
preparation of a series of handsheets following the Technical
Association of the Pulp and Paper Industry (TAPPI) T 205 method
[1]. Many paper mills have their pulps provided as dry sheets which
are converted to a dilute stock and processed by refining, a technique
borrowed from the food industry to produce starch from grain.
Refining wood fibres is achieved by subjecting fibres in aqueous
suspension to high shear via the mechanical action of the rotating
blades. Such action causes fibres to be cut, or more preferably, to
become frayed, producing attached and detached filamentary fibrils
from the fibre walls. The thrashing of the refining action also causes
fibres to be softened producing more collapsible, better bonded
fibres, along with a higher density of the sheet due to the fibrils
(fines) generated during the process. The effects of refining a pulp for
papermaking are assessed by ‘beating’ the pulps to various levels in
a laboratory Norwegian Paper and Fiber Research Institution (PFI)
mill device (Papir- og fiberinstituttet AS) to achieve varying levels
of pulp freeness. The latter term refers to how freely water drains
from the resulting pulp. A highly refined pulp will have considerable
fibrillation, causing a more tortuous path for water to drain. The
Canadian Standard Freeness (CSF), TAPPI T 227 method [2] in
fact measures the amount of water that readily drains away from a
fixed volume of pulp suspension at a fixed consistency. The physical
characteristics of handsheets are measured and related to the level
of pulp freeness in order to determine how much refining may be
173
required to reach a certain level of desired end properties. Lower

freeness implies the generation of more fines, higher sheet density
and shorter fibres, which will result in sheet decreased tear, decreased
opacity, but increased burst and increased tensile properties.
Tensile testing was discussed in Chapter 2. Opacity, the reflectance

of paper at 0° incidence using white light and a back-scatter
detector filtered for visible optical wavelength transmission, has
been previously mentioned in Chapter 6 in the characterisation of
writing papers. Here, we will cover the popular testing methods of
burst and tear of handsheets used to characterise pulps, producing
the so-called ‘beater curves’.
7.2 Tear Testing of Paper

Curiously, it is the out-of-plane tear test, the Elmendorf test [3] detailed
in TAPPI T 414 that is commonly used for paper characterisation,
although this mode of tearing is not directly relevant for most end-use
fracture properties such as web breaks in a press room. However, it is
still regarded as a useful way to rank the relative toughness of different
products or variations in quality. The test consists of initiating a crack
or a cut along one edge of 4 plies of a rectangular sheet sample. The
wire side or top side of the 4-ply sample assembly should all face the
same way and 4 plies is the standard assembly. These factors appear
to make a difference, as explored by Seth and Blinco [4]. A slitting
knife incorporated into the instrument makes an initial cut 20 mm in
length along the edge and the firmly clamped 4-ply assembly is then
torn out-of-plane by a downward swinging circular sector pendulum
attached to one of the test piece clamps. The tear length through the
test piece assembly is fixed at 43 mm. The energy expended in the
tear is measured by the rise of the pendulum, shown schematically
in Figure 7.1. The pendulum energy divided by the fixed tear length
of 43 mm is reported in mN, generally around 400 to 900 mN for
most typical handsheets.
174
‘Beater Curves’ to Evaluate Pulp Potential: Burst, Tensile,Tear,
Opacity
Centre of mass
of pendulum at
test start
∆h
W = mg∆h = F × d
Figure 7.1 Principle of the measurement of the Elmendorf tear test
Figure 7.2 shows a modern digital version of the Elmendorf tester

with panel-activated pneumatic sample clamps, pendulum release,
digital read-out adjusted for the number of plies being tested, and the
ability to calculate and display the average and standard deviation of
multiple test results. Pendulum travel in a test is read by an encoder
transducer at the pendulum pivot and the particular pendulum type is
read by an optical transducer. Different weight pendula are supplied
to ensure the instrument is within measuring range for a variety of
basis weight papers. In most circumstances, the medium-weight
pendulum of 1,600 g will suffice. Specimens are preferably cut with
a punch for greatest accuracy, shown in Figure 7.3, of dimensions
53 × 63 mm, the length being in the tear direction of interest. Most
paper samples produce a slightly higher tear in the cross direction
of paper (CD), compared with the machine direction of machine-
made paper (MD), meaning it takes more work to cut across the
fibres than along the fibres. At least five replicate measurements are
made in either the MD or CD to characterise the tear strength for
any particular sample. Handsheets have no MD/CD orientation, so
five replicate tests suffice.
175
a) b) c)
Figure 7.2 a) A digital Elmendorf tear test unit with its pendulum
prior to a test, b) 4 plies of a test sample in the divided clamps and
c) pendulum in rest position after a test
Pneumatically operated punch for tear

samples 63 × 53 mm, also useful for
quick basis weight measurements using
the 5 place balance
CD tear specimens cut by the

punch, the convention is for the
length of a sample sheet to be in
the MD
Figure 7.3 Punch and orientation of samples for the tear test
176
Opacity
The tear test essentially measures the combined work in fracturing

and pulling out fibres in the prescribed geometry. Therefore, test
values are dependent on the combination of fibre quality and fibre
bonding. The fibre-bond strength is usually less than the fibre-fracture
strength, which is why in many cases individual fibres can be seen
projecting along the edge of a tear. If a process such as refining causes
the bond strength to increase, the tear will be initially observed to
increase but upon increasing the refining level (shear energy imparted
to the pulp) the fibre strength will become affected and the tear
strength decreases. What happens for a given pulp at various refining
levels in terms of tear strength values, depends on the fibre length, as
longer fibres will have more bonds with other fibres, and of course,
the fibre strength is largely dependent on the pulp type and wood
species used.
7.3 Burst Testing of Paper

One of the most simple but misleading tests for paper is the burst
test [5]. Evolved originally from the textile industry and invented
in 1887, the test has been used prevalently for quality control in
the production of papers made for bags, corrugated packaging
and envelopes. Burst strength is the pressure required to puncture
a sheet and is measured using a rubber diaphragm secured under
an annular ring expanding hydraulically against a paper sample
supported firmly against the diaphragm by a circular clamp. An
electric motor drives a piston inside a cylinder containing glycerol
at the end of which is the rubber sheeting diaphragm, shown
schematically in Figure 7.4.
A pressure transducer inside the cylinder connected to a peak-and-

hold signal digital read-out records peak pressure attained when
the sample ruptures as one or more tears. The circular geometry of
the test is amenable to mathematical analysis and the main result,
pointed out by van den Akker [6], is that the peak rupture pressure
‘P’ is proportional to the MD and CD tensile strengths St,MD, St,CD,
and the square root of the strain to break ‘εs’:
177
S t,MD + S t,CD
P+ fs
2r (7.1)
with ‘r’ being the curvature of the diaphragm, which has been
experimentally shown to be true. Therefore, the burst test is a
combined tensile test in terms that the measured peak pressure at
burst is the tensile strength(s) (St) multiplied by the square root of
stretch to break. The in-plane strain of the paper during the test is
proportional to the radius of curvature. Therefore, the side of the
sheet facing away from the diaphragm is subjected to a slightly higher
strain than the side next to the rubber diaphragm. As the surface
structure of many papers is different, especially those made on a
fourdrinier sheet paper machine, the test is two-sided. The practice is
to test a sheet five times with the sheet wire side facing the diaphragm,
then five times again with the top side facing the diaphragm. A sample
sheet of any convenient size is placed over the rubber diaphragm and
the upper clamp ‘tripod’ is pneumatically actuated to firmly secure
the sample over the diaphragm. A lever is used to manually actuate
the piston-geared motor drive and to end the test immediately upon
test piece rupture, the instrument is shown in Figure 7.5.
Expanding rubber
diaphragm Paper
‘Tripod’ clamping ring
Increasing glycerol pressure
Figure 7.4 Schematic cross-section diagram of the burst test

method
178
Opacity
Peak pressure electronic read-out
Pneumatic
actuator
switch for
tripod clamp
Samples placed
here over the
diaphragm
Gear piston drive

with lever
Figure 7.5 A Mullen-style burst tester used for paper and

paperboard
Corrugated boards are similarly tested for burst, however, the

specification for the test piece clamping is different and uses a larger
surface diameter circular ring [7] to ensure no in-plane movement
of the sample board during the test. The fluted medium in this case
does not contribute to corrugated board burst, so samples can be
crushed flat prior to testing without affecting the results.
7.4 Pulp Beating

In this example, a pulp mill submitted three fully bleached softwood
kraft samples for evaluation. Bleached hardwood kraft is the most
responsive pulp to refining followed by bleached softwood kraft,
which is generally about half as responsive, whereas groundwood
pulp is least affected. The objective is to attain optimal properties
179
with as little beating as possible to minimise energy costs while

optimising quality. The mill produces market pulp in the form of
sheets for conversion to other products, such as tissue or writing
papers, at other manufacturing facilities. Pulp samples are tested at
10% consistency with up to 30 g placed into a laboratory PFI mill
[8], shown in Figure 7.6.
Figure 7.6 Photograph of a PFI mill rotor separated from the

beater housing
The upper rotor, consisting of chiseled blades, is inserted into the

lower housing and rotates at approximately 1,400 rpm while the
housing rotates at a lower speed causing a shearing action on the
pulp. The number of revolutions are counted and portions of the
pulp are removed and diluted to a lower 3% consistency and tested
for CSF, which measures the amount of millilitres of water to freely
180
Opacity
drain from 1 litre of the pulp. The results from beating in the PFI
mill and testing the CSF are shown in Table 7.1.
Table 7.1 CSF (ml) as a result of PFI mill revolutions for

3 pulp samples
PFI revolutions Sample A Sample B Sample C
0 748 751 742
1,500 652 632 663
3,000 570 534 544
4,500 413 353 363
6,000 278 230 225
With more PFI revolutions, more work is done on the pulp, which
progressively decreases the freeness. Pure water has a freeness of
over 900 ml compared with the initial test volume of 1 litre. As the
pulp fibres become increasingly swollen and frayed with increased
revolutions they are able to retain more free water in the CSF test.
The increase in freeness is also expected to translate into increased
sheet density due to the increased presence of submillimetre-sized
fines and fibrils, and larger strength values in handsheets due to
increased bonding as a result of the fibres becoming swollen, softer
and less rigid.
The physical properties of the handsheets, due to the pulp being

beaten to various freeness levels, are shown in Table 7.2. As the
number of revolutions increase, the freeness decreases and generally
the basis weight of the handsheets increases. The pulp consistency
during handsheet making is constant, as per TAPPI T 205, so this
observed increase in basis weight with lower freeness is a result of
increased retention of the amount of fibre during drainage and more
particulate matter in the sheet resulting in higher density.
181
Table 7.2 Physical properties of handsheets from three pulps

beaten to different levels. Number of PFI revolutions are
appended to the sample ID letter
Sample CSF Basis Tear index Burst Opacity BL (km)
ID (ml) weight (mN-m2/g) index (%)
(g/m2) (kPa-m2/g)
A 0 742 65.5 ± 4.4 15.14 ± 1.02 1.00 ± 0.05 75.7 ± 1.7 2.02 ± 0.08
1,500 663 68.4 ± 3.9 15.47 ± 0.93 3.63 ± 0.30 66.8 ± 2.2 5.31 ± 0.20
3,000 554 69.0 ± 3.0 12.50 ± 0.54 4.59 ± 0.25 63.8 ± 1.5 5.93 ± 0.37
4,500 363 64.1 ± 3.6 11.13 ± 0.62 5.17 ± 0.11 58.1 ± 2.1 7.11 ± 0.22
6,000 225 65.8 ± 1.9 9.82 ± 0.28 5.36 ± 0.09 60.0 ± 1.4 7.32 ± 0.45
B 0 751 67.0 ± 1.0 12.92 ± 0.19 0.84 ± 0.05 76.4 ± 0.9 1.85 ± 0.05
1,500 632 69.5 ± 2.3 15.16 ± 0.51 2.83 ± 0.30 59.8 ± 0.6 4.56 ± 0.20
3,000 534 62.1 ± 2.4 13.12 ± 0.52 3.88 ± 0.32 57.6 ± 2.3 5.11 ± 0.54
4,500 353 62.0 ± 4.1 11.07 ± 0.74 4.24 ± 0.32 57.6 ± 2.3 6.45 ± 0.58
6,000 230 63.9 ± 1.6 10.11 ± 0.25 4.64 ± 0.18 59.0 ± 1.5 6.28 ± 0.37
C 0 748 68.0 ± 4.3 10.17 ± 0.64 0.64 ± 0.02 77.8 ± 0.9 1.95 ± 0.17
1,500 652 69.9 ± 1.2 20.33 ± 0.35 2.47 ± 0.11 58.0 ± 0.6 5.20 ± 0.43
3,000 570 63.0 ±2.0 15.89 ± 0.50 3.57 ± 0.22 61.4 ± 0.9 6.51 ± 0.68
4,500 413 60.4 ± 2.1 12.92 ± 0.44 4.40 ± 0.10 56.9 ± 1.9 7.53 ± 0.37
6,000 278 58.9 ± 1.8 12.31 ± 0.38 4.58 ± 0.24 56.9 ± 1.4 8.19 ± 0.61
BL: Breaking length
7.5 Results of Physical Properties from Pulp Beating

Of the three pulps, sample C produced the most effective and
desirable results. The values of tear, burst and St are all proportional
to basis weight, so all the strength values are divided by the average
basis weight of the handsheets in each sample set. St is converted to
BL (km) by dividing by the basis weight and multiplying the result
by a constant, as shown in Chapter 2.
Figure 7.7 shows that the tear index (tear strength divided by basis
weight) is improved by 100% with beating to 1,500 revolutions,
producing a freeness of 663 ml. Increased revolution decreases the
freeness but then the tear index decreases as bonds become stronger
182
Opacity
but the fibre becomes weaker. Increasing the bonds with increased
PFI revolution increases the BL and burst, which have similar trends
with CSF, as burst is dependent on St. Note that the increase in burst
and BL saturate with refining level at a point where the increase in
bonds from increased particulate matter becomes offset by shortened
and weakened fibres. Smoothed lines drawn through the points in
Figure 7.7 indicate the trends. The beater curve shows that refining
beyond a freeness of 660 ml is not worthwhile.
Tear index 20
Burst index
BL
15
10
0
800 700 600 500 400 300 200
CSF (ml)
Figure 7.7 ‘Beater curve’ results for pulp sample C [tear index
(mN-m2/g), burst index (kPa-m2/g) and BL (km)]
The opacity of paper is affected by the combination of the absorption

and scattering properties of the fibres, as well as the amount of fibres
per unit area or basis weight, which is quantified by the Kubelka–
Munk theory [9] for light interacting with paper. Scattering is
increased by the number of refractive surfaces per unit volume, so
183
the addition of fines and fibrils to a paper can be expected at first

to increase the opacity, and indeed such is the case when a surface
coating is applied containing fine small-sized pigment particles of the
order of microns or less in diameter. In the case of beaten fibres in wet
suspension, the fines and microfibrils bond to paper surfaces and to
each other resulting in a higher density but not an increased refractive
index gradient, which would otherwise increase the scattering power.
Much like the compaction of snow on roads underneath tyre tracks
reduces the opacity such that the underlying pavement shows through
the overlying snow, the increase of fibrils and fines and increased fibre
bonding all result in a higher sheet density that occurs with lower
freeness causing a loss in paper opacity. The linear dependence of
opacity on basis weight is removed by dividing opacity values to
calculate the opacity index, shown as a function of freeness for the
three pulp samples in Figure 7.8.
1.25
1.2 Opacity index (%-m2/g)
Sample A 1.15
Sample B 1.1
Sample C 1.05
1
0.95
0.9
0.85
0.8
800 700 600 500 400 300 200
CSF (ml)
Figure 7.8 Opacity index for all three pulp samples versus freeness
184
Opacity
All three pulp samples show a similar declining opacity trend with
decreasing freeness, consistent with the concept of loss of scattering
due to increased sheet density. Similar losses in opacity from paper
densification affect the measurement of paper formation (areal
mass uniformity) when lightweight paper is calendered to attain
caliper and surface smoothness [10]. This effect, which has led to
misleading attempts of formation optimisation in mills, prompted
the development of alternative paper formation measurement
methods. These include imaging of the transmission of electrons
from radioactive sources or electron microscopes or long wavelength
X-rays [11] to avoid the optical scattering effects that occur otherwise
when using visible light transmission for paper imaging.
7.6 Summary
Pulps are often evaluated for their response to refining using a
laboratory beater, which is used to prepare samples with a progressive
lower amount of freeness. A lower freeness pulp is the result of the
amount of refining energy that has been imparted from beating,
resulting in higher water retention in the pulp when drained through
a screen. Higher sheet density resulting from softer more collapsed
fibres and the presence of fines and microfibrils creates more stress
transfer between fibres when the sheet is under load so that St
increases with refining work. The common ‘beater curve’ follows
pulp development with increasing beating work by measuring the
out-of-plane tear and tensile or burst properties. Increased fibre
bonding with refining level causes the tear strength to increase with
some refining but with further refining is observed to decrease, while
the St steadily increases with refining. The objective of the ‘beater
curve’ is to determine what freeness level is required to obtain the
optimal balance between tear strength and St.
If other properties were to be followed with changes in pulp freeness,

the expectation from first principles emphasised in this book, is that
the elastic modulus, measured either mechanically or by ultrasonic
propagation, will be observed to follow the same trend as either
185
St or burst strengths. Compression strength measured by the short-

span method is also expected to increase with refining work based
on the tensile–compression strength relationship and their mutual
dependence on elastic modulus. Bending stiffness, being Et3/12,
should be expected to decrease as although modulus ‘E’ increases
with sheet density, the corresponding decrease in caliper from softer
more collapsible fibres and higher bonding will make the handsheet
less structurally rigid.
Increased sheet density with refining produces less scattering, resulting

in a loss of opacity and decreased brightness, an undesirable result for
writing or printing papers. Since most paper products are required to
meet stringent marketing specifications in terms of St or tear strength
properties, opacity or brightness, it is obvious how the combined
engineering and balancing of all these interrelated and interdependent
properties in the papermaking production process is often deservedly
called an art. It is an art that is certainly amenable to physical testing
measurement and quantified characterisation.
References
1. TAPPI T 205: Forming handsheets for physical tests of pulp
(Revision), 2012.
2. TAPPI T 227: Freeness of pulp (Canadian standard method)

(Revision), 1999.
3. TAPPI T 417: Internal tear resistance of paper (Elmendorf-

type method).
4. R.S. Seth and K.M. Blinco, TAPPI Journal, 1990, 73, 1, 139.
5. TAPPI T 403: Bursting strength of paper (Revision), 1997.
6. J.A. van den Akker, Research Bulletin, 1938, 4, 3, 46.
7. TAPPI T 810: Bursting strength of corrugated board, 1985.
186
Opacity
8. TAPPI T 248: Laboratory beating of pulp (PFI mill method),

2015.
9. J.S. Arney, J. Chauvin, J. Nauman and P.G. Anderson,

Journal of Imaging Science and Technology, 2003, 47, 4 339.
10. A. Komppa and K. Ebeling in Proceedings of the 7th Role

of Fundamental Research in Papermaking Symposium,
September, Cambridge, UK, 1981, Mechanical Engineering
Publications, London, UK, 1983, p.603.
11. H. Tomimasu, D. Kim, M. Suk and P. Luner, TAPPI Journal,

1991, 74, 7, 165.
187
188
I
ndex
A
A-fluted medium, 102
A-fluting, 118
Absorb(ing), 1, 16-17, 19, 28, 49, 137, 139, 145-146, 148-151,
170
Absorption, 136, 139, 146-147, 149-151, 163, 165, 167, 169,
183
of fluid, 163
of ink, 163
properties, 146, 165, 167
Additive(s), 27, 37, 45, 48-51, 53, 79, 90
Adhesion, 9, 21, 151
Adhesive, 8, 123, 139
application, 8
tape, 139
Ageing, 146
Agglomeration, 8
Air, 6, 12, 17, 151-156, 160, 167, 169
flow, 12, 17, 151 -154, 169
leak, 152-156
permeability, 151-152, 160, 170
pressure, 153
American Society for Testing and Materials (ASTM), 137, 169
ASTM D5455, 137
Annulus, 153, 155
Aqueous, 17, 139, 163, 169, 173
ink, 139, 163
suspension, 173
193
Arid environment, 10
Asymmetry, 16-17, 120
Attenuation, 50-53
B
B-flute boards, 124
Back-scatter detector, 174
Backing tape, 45
Backing wire (BW), 18
Basis weight, 8-9, 13, 15-16, 18-19, 30, 34, 38, 45, 47, 50-51,
61, 73, 79-80, 83, 85-90, 93-94, 97-99, 101-102, 106, 112, 119,
125-126, 135, 140, 145, 150-151, 157-160, 165, 175-176,
181-184
Beam, 44, 57-59, 65, 68, 77, 85, 104, 106-107, 109-110, 115
bending, 44, 104, 109-110
buckling, 104, 106, 109
mechanics, 115
Bending, 27, 44, 47-48, 55, 57-61, 63, 65-67, 69-71, 73, 75,
77-78, 85-86, 89, 91, 96-97, 101, 103-104, 106, 109-110, 112,
115-116, 126, 132-133, 186
action, 57
angle, 58, 66-67
failure, 47, 101
load, 104
moment, 44, 58, 116
resistance, 55, 58-59, 77
stiffness (D), 27, 44, 51, 57-61, 64-73, 75, 77-78, 104-106, 111,
115-116, 125-127, 129, 131, 133, 186
Billerud cut square test piece, 113
Black-absorbing background, 145
Bleached, 4-5, 18-19, 29-30, 61, 179
chemical kraft pulp, 18-9
hardwood kraft, 179
kraft copier paper, 29
softwood kraft, 179
Bleaching, 18, 143
Bleed, 135-136, 140-141, 145, 151, 164
194
Index
-through, 135-136, 140-141, 145, 151, 164

Blue, 139, 141-143
wavelength, 143
Board, 8-9, 12, 21, 45-46, 54, 57, 59-61, 63, 65, 67-73, 75, 77,
79, 81, 83, 85, 87, 89, 91, 93, 95-121, 123-129, 131-133, 151,
170, 179, 186
Bond, 5, 18, 177, 184
strength, 177
Bonded, 1-2, 5-6, 14, 17-18, 81, 173
fines, 17
Bonding, 5, 8, 19, 23, 38, 45, 49, 51, 53, 79, 82, 96, 126, 177,
181, 184-186
degree, 49, 82
level, 38
Bottom side, 16-17
Bowing, 98
Box compression test strength (BCT), 75, 77, 128-129, 131
Box plants, 96, 98, 102
Breaking, 35, 182
length (BL), 35, 182-183
Brightening agents or dyes, 142
Brightimeter, 141, 144
Brightness, 142-144, 146, 186
Bristow wheel, 137-140, 157
apparatus, 137-138
concept, 138
ink application, 137
British Standard Sheet Machine handsheet former, 49
Buckle, 109, 115, 117-118, 123-124
Buckled, 117
Buckling, 79-80, 82, 84-88, 90-92, 100, 104, 106, 109-120,
123-128, 131-132
column height, 84
condition, 114
load, 84, 91, 104, 106, 109-115, 118-119, 123, 127-128
model, 124-128, 131
of the facings, 115, 124
195
pattern, 84, 117, 131

stress, 86-88, 92
test cylinder, 90
Bulging, 57, 67
Bulk, 49, 139-140, 160-161
Burst, 17, 173-175, 177-179, 181-183, 185-187
properties, 185
test, 17, 177-178
Bursting strength, 186
C
C-flute, 70-71, 73, 96, 99, 102-103, 106, 108-112, 115, 117-118,
120-121, 123-124, 126
board, 70, 106, 108-109, 112, 123
medium, 117, 126
simulation, 118
Calendering, 15, 160
Calibration, 148, 154, 156-158
check, 157
Caliper, 8, 26, 38, 43-44, 50-51, 54, 57-59, 61-69, 71-73, 75, 77,
81, 86-87, 98, 102-104, 107, 119-120, 123, 125-126, 129, 150-
151, 159-160, 185-186
gauge, 150
pressure, 50-51
Canadian Standard Freeness (CSF), 173, 180-184
Capillary, 136-137, 139
action, 136-137, 139
effect, 136
Carton packaging, 17, 79
Cell, 23-24, 59, 70, 85
Cellulose, 1-2, 5, 9, 18-19, 80
Ceramic, 146, 157
white tiles, 157
coloured tiles, 157
Chemical, 1, 4-5, 18-19, 51, 53
pulp, 4-5, 18-19
196
Index
Circular, 39, 43, 58-59, 65, 83, 174, 177, 179

array, 39
platen, 43
Clamp, 24, 85, 99, 101-103, 109-111, 113-114, 132, 152, 177-
179
method, 99, 101, 103, 110, 114, 132
Clamped, 23, 59, 85, 105, 110, 113, 174
Clamping, 70, 101, 114, 151-152, 158, 178-179
fixture, 114
pressure, 151
restraint, 70
test, 101
fixture, 101
Clumping, 8, 16, 28, 32, 43
Coarse(ness), 4, 80, 151, 160
Coarser, 80
Coated, 5, 18-19, 61-62, 155-156
Coating, 17, 171, 184
Cobb, 149-150, 165, 169-170
test, 149, 170
tester, 149-150
Coefficient, 8, 81, 90, 165
of variation (cv), 8, 15-16, 90
Colour, 19, 135, 141-146, 160
measurement, 143
printing, 19
change, 145
Colorimetric, 141
Commercial corrugated boards, 119
Commercial linerboard, 123
Commercial medium, 123
Commission Internationale de l’Eclairage (CIE), 141-142, 145
Compaction, 5-6, 8, 184
Composite, 77, 133
Compression, 12, 15, 29, 45-47, 50, 53, 57, 73, 75, 79-91, 93,
95-99, 101, 103, 105-113, 115, 117-119, 121, 123-129, 131,
133, 163, 169, 186
197
buckling model, 131

crease failure, 118
failure, 79, 101, 124, 131
strain, 81, 97
strength, 12, 15, 46, 53, 57, 73, 75, 79-83, 85, 88-89, 91, 93,
96-97, 106, 111, 113, 115, 118-119, 123-129, 131, 186
stress ratio, 86
test, 12, 45, 80, 82, 85, 96, 99, 115, 118, 128
tester, 45, 83, 89, 97, 108
Compressibility, 19, 27, 38, 43, 50-51, 155
Computer, 71, 129, 147-148
-aided design (CAD), 71-72, 129
Concentration, 5, 157
Conditioning, 12, 21
Contact angle, 136, 146-149, 162-165, 169-170
Contacting stylus profilometer, 152, 155
Containerboard corrugated packaging, 19
Conveying wire side of the paper (CW), 18, 80, 170
Correlation, 44-48, 50, 86, 94, 127, 160, 165
Corrugated, 8-9, 12-13, 18-19, 45-47, 57, 59, 61, 63, 65, 67-71,
73, 75, 77, 79, 87, 89, 96, 102, 104-107, 115-116, 119-120,
123-124, 126-128, 132, 151, 155, 170, 177, 179, 186
board, 8-9, 12, 45-46, 57, 59, 61, 63, 65, 67-71, 73, 75, 77, 79,
87, 89, 96, 102, 104-107, 119-120, 123-124, 126-128, 151,
179, 186
box, 12, 46
medium test (CMT), 45, 47, 55
Corrugating, 48, 55, 83, 90, 123
medium, 48, 55, 83, 90
operating parameters, 123
Crack(ing), 174
Crease, 79, 99-100, 116-118
Critical buckling load, 91, 106, 115, 119
Cross-direction, 6, 19
Cross-head, 24, 70, 109
velocity, 109
Cross-machine reel profile, 39
198
Index
Cross-reel strip, 40
Cross-section, 3-4, 18-19, 178
Crowding factor, 32
Crush, 45-47, 55, 83, 100-103, 103, 131, 179
damage, 101
Crushing, 10, 72, 83, 96, 101-103
Curl, 10, 28, 90
Cut, 3-4, 7, 23, 26, 39, 59, 69-72, 97, 99, 109, 112-113, 124,
139, 155, 173-176
Cutter, 59-60, 90, 98
Cutting, 48, 83, 90, 98, 100, 102, 114
Cylinder, 84, 90, 151-153, 177
D
Debonder, 49-50
dosage, 50
Deflection, 58-59, 63, 65-66, 68, 70, 77
Deformation, 26-27, 29-30, 97, 109, 115
rate, 29
Deionised water, 150
Dense, 43, 139, 150
printing, 43
sheet, 150
Density, 18-19, 34, 38, 51, 61-62, 79-81, 90, 120, 123, 125-
127, 135, 139-140, 145, 150-151, 153, 155, 158, 160-161,
173-174, 181, 184-186
high, 18-19, 79, 80-81, 125-126, 145, 160, 173, 181, 184
low, 18-19, 62, 80, 151
Dewatering, 6, 17, 79
Diamond stylus, 155
Diaphragm, 17, 177-1799
Die, 102, 137, 139
-cutting scoring, 102
-cutting slitting, 102
Dilute stock, 173
Dimpling, 116-117
199
Dip, 40-41
Dipped, 109
Dipping, 98
Disperse, 2
Dispersed, 1, 157, 167
Dispersion, 3
Displacement, 24-27, 29-30, 70, 81, 116-118, 155
Double-wall board, 72-73
Lightweight, 72
Drain(s), 5, 173
Drainage direction, 17, 139
Drainage wire, 7
Draining, 5, 16, 18
Drape, 52, 57
Draw, 39, 79
Drop, 42, 146-149, 151-152
rate, 151
volume, 149
Droplet, 147-149
Dry, 3, 10, 12-13, 15, 35, 39, 45, 49, 79, 160, 173
end, 15, 79
calendering, 160
paper, 12-13, 39
state, 12
weight, 49
Dried, 2, 5, 9, 12, 167
Drying, 6, 14, 38-40, 136, 139
strategies, 39
temperature, 14
Dye solution, 157
E
E-flute, 112-114, 120-121, 124
board, 113, 124
Edge, 1, 39-42, 52, 75, 79, 84, 89, 96-97, 99-100, 108, 146, 174,
177
200
Index
compression strength test (ECT), 96-97, 99-104, 106-120, 123-

129, 131
clamp fixture, 109
measurement, 115
flow, 40
roll, 100
Elastic, 14, 23, 25-26, 37, 43, 48, 53, 57, 59, 63, 65-67, 79-81,
83, 85, 87, 89, 91, 93-95, 97, 99, 101, 103-105, 107, 109, 111,
113, 115, 117, 119, 121, 123, 125, 127, 129, 131, 133, 185-186
beam, 85
modulus (E), 14, 25-28, 34, 37-39, 43, 45, 53, 57, 63, 65-68,
77, 79-81, 83, 85, 87, 89, 91, 93-95, 97, 99, 101, 103-105,
107, 109, 111, 113, 115, 117, 119, 121, 123, 125, 127, 129,
131, 133, 185-186
properties, 23
region, 25, 59, 65
Elasticity, 44, 57-59, 65, 67, 86
theory, 44, 57-58, 65, 67, 86
Elastomericity, 28
Elmendorf tear test, 175-176
Elmendorf test, 174-175
Elongation, 27, 31-32, 34
rate, 34
Emveco 200A, 61
Emveco 210-DH, 62
Emveco 210-R, 155-156, 167
Energy, 28, 125, 136, 139, 146, 165, 167, 174, 177, 180, 185
Environment, 10-13, 116
Epoxy, 108-110
resin, 108-110
Equation, 10, 14, 25-28, 33-34, 37-38, 44-45, 51, 57-59, 63, 65-
66, 68, 70-71, 75-77, 80-81, 84-86, 91-92, 95-97, 104, 106, 109,
114-115, 119, 123-124, 126-129, 136, 145-146, 155, 165, 168
Equilibration, 10-11
Equilibrium point, 12
Euler beam buckling, 104
Euler mechanics elasticity analysis, 58
201
Evaporation, 136
Expanding rubber diaphragm, 17, 178
Exposure, 9, 136, 149-150
time, 136
F
F-flute, 111-112
Fabric, 16-17, 40
Facing, 68, 90, 98, 104, 115, 117-118, 123-124, 139, 145, 178
side, 139
stiffness, 104
Failure mode, 110
Failure strength, 28
Failure stress level, 118
Feathering, 146-147
propensity, 147
Felt, 16-18, 139
side, 16-18, 139
Fibre, 2-5, 7-8, 14-19, 23, 27-28, 34, 38, 45, 49, 51, 79-82, 126,
135, 155, 160, 173, 177, 181, 183-185
axial failure, 80
axial strength, 80
bonding, 5, 23, 45, 79, 126, 177, 184-185
buckling, 79-80
clumping, 8
coarseness, 80, 160
collapse, 8
compaction, 8
contact, 2
direction, 7
elastic modulus, 14, 80
Euler buckling, 80
fines, 18
fracture, 79
strength, 177
mat, 5, 16, 18
202
Index
modulus, 79-81
network, 28
orientation, 38
pull-out, 79
quality, 23, 27-28, 34, 45, 82, 177
relative bonded area (RBA), 81
slurry, 5
species, 19
stock, 15, 18
strength, 177
tensile strength, 28
thickness, 80-81
wall, 80
thickness, 80
width, 2, 80
Fibril(s), 173, 181, 184
angle, 81
Fibrillation, 160, 173
FibroDat pocket goniometer, 147-149
Filler, 18, 90, 135, 140, 145
content, 135
Film(s), 45, 61, 118, 152, 160, 169
Filter, 143, 146
Fines, 17-18, 90, 135, 139, 160, 173-174, 181, 184-185
content, 135
rich, 139
stock fraction, 17
Finite element analysis (FEA), 117-118
Flexographic print density mottle, 155
Flexography, 5
Flexural rigidity, 67-68
Floc, 32
Flow, 12, 17, 40-41, 151-154, 160, 163, 169
-through, 160
Fluid, 136-137, 160, 163, 169
absorption, 169
flow, 160, 163, 169
203
surface, 136
Flute, 70-71, 73, 96-104, 106-118, 120-121, 123-124, 126, 131
board, 70, 106, 108-109, 112-113, 116, 123-124
glue-line spacing, 115
lines, 104
size, 97-98, 117, 124
spacing, 104, 118
Fluted, 4, 45, 68, 96-97, 101-102, 106, 112, 117-118, 151, 179
medium, 68, 96, 101-102, 106, 112, 117-118, 151, 179
buckling, 118
strip, 45
Fluting, 8-9, 102, 104, 118
weight, 102
Fold, 17, 90
Folding, 17, 57, 79, 99, 111
Folder-gluer, 71
Food industry, 173
Force, 26, 44, 52, 58-59, 70, 81, 85, 116, 155
Forestry, 149, 170
Formation, 5, 7, 32, 85, 117-118, 131, 151, 185
Formette prefix, 123
Formic acid, 157
Four-point, 69-73, 75, 77, 106-107
bending, 69-70
stiffness, 71, 73, 77
method, 106
stiffness, 75
Fourdrinier paper, 16, 39, 139
machine, 16, 39, 90, 139
-made, 16, 18, 81
linerboard, 81
sheet, 17, 178
paper machine, 178
Fourier analysis, 51
Fracture, 23, 28, 79, 174, 177
properties, 174
toughness, 28
204
Index
Fracturing, 177
Free span, 85, 109, 111-114
height, 111-114
Free water, 16, 181
Freeness, 173-174, 181-186
Friction, 51
Fully bleached softwood kraft, 179
G
Geometry, 17, 84, 143, 153, 170, 177
Glass, 153-154
plate, 153-154
Gloss, 18
high-gloss surface, 18
Glue, 8-9, 70, 115, 124
line, 115
Glued single-wall, 123
Grams per square metre (GSM), 30-32, 126-127
Groundwood, 5, 18, 179
Guillotine-style cutter, 90
H
Handle-o-meter, 52
Handsheet, 48-49, 121-123, 127-128, 181, 186
Handwriting, 137, 139
speed, 139
Hard, 44, 61-62, 68, 155-156
caliper, 61-62, 68
Hardwood, 4, 19, 90, 160, 179
Head, 24, 70, 109, 154
Headbox, 5, 15, 39, 53
Heavyweight C-flute corrugated board, 71
Heavyweight linerboard facings, 117
Heavyweight single-wall board, 72
Heavyweight unbleached kraft, 18
205
Height, 84, 97-98, 104, 106-115, 131

Hemicellulose, 19
Hercules size test (HST), 157-158, 163, 165, 167-169
Hooke’s law, 26, 81
Hot air impingement, 167
Hot plate press, 123
Human eye, 141-142
Humid(ity), 10-13, 116
ambient, 10-12
environment, 11, 13, 16
high, 10-11, 116
environment, 11, 116
low, 10-11
Hydrodynamics, 5
Hydrogen bonding, 8
Hydrophilic, 1, 9, 11, 19, 147
Hydrophilicity, 1, 9-10
Hydrophobic, 146-147
Hygroexpansivity testing, 29
I
Illumination, 116, 142-143
Image analysis software, 148
Imaging, 21, 185, 187
In-plane, 7-8, 17, 19, 28, 38-39, 42-43, 45, 178-179
ductility, 17
movement, 179
shear sonic wave propagation, 38
strain, 178
tear, 28
ultrasonic specific stiffness, 45
Inelastic mechanical strength, 48
Infrared, 8, 139
Ink(s), 8, 17-18, 135-140, 145-146, 149-153, 157-158, 160, 163-
164, 166-171
absorption, 136, 150
206
Index
dispenser, 158
exposure, 136
film, 152, 160
flow-through, 160
penetrating, 135, 140, 145, 150
transfer, 136, 152-153
Inked surface, 139
Inking, 145
Ink-jet MICR printing, 168
Ink-jet paper, 149
Ink-jet printability, 166, 168-169
Ink-paper performance, 146
Innventia, 149, 170
Institute of Paper Science and Technology (IPST), 120-121, 123,
127-128
Instron 1122 universal test machine, 63
Interflute buckling, 104, 123-124
Interflute linerboard panel, 119
Intrafibre bonding, 53
Isotropic geometry, 143
J
Jet, 39, 53, 79, 149, 166-170
impingement angle, 79
Joining, 116-117
Jumbo roll of paper, 14-15
K
Knife, 39, 99, 174
Kraft, 3-5, 12-13, 18-19, 29-32, 45, 61, 73, 149, 179
linerboard, 12-13, 29-32, 61
linerboard paper, 32
pulp, 5, 19, 45, 149
207
L
Laboratory beater, 185
Laboratory roller press nip, 102
Laboratory-made board, 127
Laminar Poiseuille flow, 153
Latewood, 3
Length, 2, 16, 23-25, 28-29, 39, 44, 58-60, 63, 69, 75, 77, 81,
83-85, 96-97, 104, 110, 112, 119, 128, 136-137, 139, 144, 155,
174-177, 182
direction, 23, 28
Letterpress, 5, 153, 160
solid print density, 153
Light, 2, 8-9, 32, 141-143, 145-148, 174, 183, 185
-absorbing black-body background, 145
interacting with paper, 183
reflected, 145
-scattering, 145
source, 145, 147
Linear, 14, 25, 27, 44, 47-48, 51, 59, 65, 67, 81, 89, 92, 96, 117,
127, 139, 165-166, 169, 184
elastic region, 59, 65
elasticity, 44, 59, 67
regression, 51-52, 165, 168-169
multiple, 51
reversible bending stiffness, 27
speed, 139
strain, 65
Liner, 47, 73, 87, 107-109, 112, 120, 127
Linerboard, 8, 12-13, 29-32, 39, 47-48, 61, 68, 70, 72, 79, 81-83,
90, 95-96, 104, 115-120, 123-124, 126, 151, 155-156
buckling, 115, 120, 123
facing, 115
printability, 156
Liquid, 135-136, 139, 146, 151, 163, 169
absorption, 139, 146, 151, 163, 169
penetration, 136
208
Index
Load(ing), 4, 23-28, 57, 59, 67-68, 70, 75, 83-85, 91, 101, 103-
104, 106, 109-120, 123-124, 127-128, 155, 185
cell, 23-24, 59, 70, 85
displacement, 25, 27, 116
Lorentzen & Wettre (L&W), 38-39, 45-46, 59-60, 63-67, 69, 86,
116, 132, 139, 150-151, 170
Densometer instrument, 151
digital caliper gauge, 150
Elmendorf Tear Tester, 151
TSO tester, 39, 45
Lucas-Washburn equation, 136, 146
M
Machine direction, 5-6, 23, 37, 59, 81, 104, 143, 175
of machine-made paper (MD), 5-7, 15, 19, 23, 26-28, 30-32,
34, 37-42, 59-61, 64, 66, 69-72, 75, 81-83, 87, 100, 105-107,
115, 119, 143, 155, 175-178
modulus, 82
orientation, 6, 143
profile, 39-40
shear, 70, 75, 107
specific stiffness ratio, 40
stiffness, 30-31
strength, 30-31
tension, 40, 82
cross direction of paper (CD), 6-7, 14, 19, 23, 26-28, 30-32,
34, 37-40, 47, 59-61, 64, 66, 69-72, 75, 81-82, 84-87, 94-96,
104, 106-107, 115, 119, 139, 143-144, 155, 175-178
cross-direction, 6
modulus, 82
orientation, 143
profile, 39-40
ratio, 39-40
specific stiffness ratio, 40
stiffness profile, 40
Machine-made linerboard, 82
209
Machine-made paper, 5-7, 19, 23, 37, 59, 82

roll, 7
Magnetic ink character (MICR), 166-169
Manometer, 154
Manufacture, 8, 12, 38, 102, 151
Manufacturing, 14, 23, 40, 160, 180
operations, 23
process, 14, 40
Mass distribution, 28, 32
Material, 4, 15, 26, 59, 91, 117, 119, 136, 151
resistance, 26
McKee, 75, 77-78, 111, 119, 128-129, 131-133
equation, 77, 119, 129
formula, 75, 128
model, 131
Mean square-root error (MSE), 127-128
Measuring platen caliper, 43
Mechanical deformation, 26, 29
Mechanical properties, 27, 34, 37, 44
Mechanical pulp, 3-5, 18, 142
Mechanical strength, 47-48
Mechanical testing, 37, 45, 55, 77
Mechanics, 25-26, 37-38, 43, 58, 77, 84, 115, 131
Medium, 18, 30, 45, 47-48, 55, 61, 68, 73, 79, 83, 86-90, 96-97,
101-102, 104, 106-109, 112, 117-120, 123-124, 126, 129, 151,
175, 179
weight unbleached kraft linerboard, 30
Metal, 38, 52, 149-152, 155
foil orifice, 152
paddle, 38
roller, 149
shim standards, 151
surface, 155
Microdeviation, 155-156, 167-168
Microflute, 113-114
Microscopy, 2-4
Microwave oven, 12
210
Index
Mill, 12, 90, 94, 173, 179-181, 187

production, 12
Miniflute, 111
Mix, 2
Mixed, 18, 30, 49, 157
Model, 4, 28, 77, 91-95, 109, 117-119, 124-128, 131, 136, 165-
166, 168-169
Modelling, 80
Modulus, 14, 26, 37-39, 42-45, 49-50, 53, 57, 63, 65-68, 77, 79-
85, 87, 89, 91, 93-95, 97, 99, 101, 103-105, 107, 109, 111, 113,
115, 117, 119, 121, 123, 125, 127, 129, 131, 133, 185-186
Moisture, 9-14, 19
content, 10-14
percentage in paper (%M), 10-11, 13-14
correction, 12
corrective factor, 12
-dependent property, 13
exposure, 9
history, 10-11
hysteresis, 11
level, 19
low, 11
-proof, 10
Molten, 97-98, 109
wax, 97-98, 109
Mottle, 8, 140, 155
Mottled, 8, 32, 140
appearance, 8, 32
inked, 140
Mullen-style burst tester, 179
Multiwall board, 120, 123
consisting of 2 B-flute medium layers (B-B), 73
N
N-flute, 111-112
board, 112
211
Naphthalene B dye solution, 157

Near infrared, 8
Neoprene, 42, 58
rubber, 42
Newsprint, 3-7, 19, 29, 61, 142, 155, 160, 170
letterpress solid print ink density, 160
Nip, 102-103, 123
crushing, 102
roller press, 102
rolling, 103, 123
Noble and Woods (N&W), 121-123, 125
prefix, 123
Nomura Shoji SST 250, 38
Non-linear, 65, 117
strain, 65
Non-uniformity, 7-8, 28, 32, 43, 85, 151
Non-viscous liquid ink, 135
Norwegian Paper and Fiber Research Institute (PFI), 173, 180-
183, 187
mill, 180-181, 187
revolution, 183
Notebook, 30-32, 139-140, 158-159, 161-165, 167
and kraft linerboard papers, 31
writability, 167
writing paper, 161, 164, 167
O
Offset, 5, 86, 127, 136, 153-155, 183
dot quality, 154
printability, 153
Opacity, 135, 140, 145-146, 162, 170, 173-175, 177, 179, 181-
187
index, 184
Opaque, 145
Optical, 3-4, 81, 139, 143-144, 146, 160, 169, 174-175, 185
density, 139
212
Index
microscopy, 4
scattering, 81, 185
transducer, 175
wavelength transmission, 174
Orientate, 41
Orientation, 5-6, 19, 34, 38, 40-42, 72, 86, 139, 143, 175-176
Oriented, 6, 81
Orthogonal direction, 6, 26
Orthotropic, 37-38, 81
solid, 37-38, 81
mechanics, 38
Out-of-plane, 5, 8-9, 18, 26, 37, 45, 47, 49, 59, 79, 115, 117-118,
131, 174, 185
buckling pattern, 131
corrugated strip crush, 47
crush, 45
displacement, 117-118
shear strain, 59
structure, 8
tear, 174, 185
Out-of-plane z direction of paper (ZD), 5-6, 16, 19, 37-38, 42-43,
49-52
attenuation, 50-52
modulus, 42, 49-50
sonic propagation, 51
transducer, 42-43
P
Packaging, 17-19, 78-79, 132-133, 177
Page equation, 28, 80
Panel, 67, 70, 115, 118-119, 128, 157, 175
-activated pneumatic sample clamps, 175
buckling, 115, 118
Paper, 1-32, 34-73, 75, 77-129, 131-158, 160-164, 166-180, 182-
186, 188
analogy, 2
213
bag, 16
bending, 104
boards, 12
brightness, 142
caliper, 8, 43, 62
characterisation, 37, 149, 174
coated magazine journal, 62
commercial newsprint, 155
commercial notebook paper, 162
copy paper, 4, 24, 148-149, 156
cylinder, 84
densification, 135, 185
fibre mat, 5
film, 45
formation, 85, 131, 151, 185
fracture, 79
imaging, 185
industry, 1, 12, 60-62, 64, 66-67, 80, 83, 88-89, 141, 173
lightweight, 185
coated magazine, 156
lined bleached white notebook paper, 30
lined kraft, 29
machine (PM), 5, 7, 14-16, 39, 41, 53, 79, 90-91, 119, 139,
157, 178
high speed, 14
operator, 41
rolls, 15
manufacturing, 8, 38, 160
mass, 8, 16
modulus, 37
moisture, 10-11, 14
newsprint, 3-7, 19, 29, 61, 142, 155, 160, 170
letterpress solid print ink density, 160
notebook, 30-32, 139-140, 158-159, 161-165, 167
writability, 167
writing, 161, 164, 167
opacity, 135, 140, 145, 184
214
Index
office copier, 4, 18
porosity, 8, 135
premium, 160
printing, 186
speed, 136
product, 5
property(ies), 4-5, 8, 14, 23, 34, 167, 169
quality, 12, 158
reflectivity, 143
roll, 7, 18
roughness, 8, 156
sample, 15, 23, 138, 140, 147, 150, 156-157, 161, 163-164,
169, 177
sheet, 2, 12, 15-18, 32, 34, 37-38, 42-45, 48-50, 52-53, 57-58,
60, 79-83, 85, 123, 125, 135, 137, 139-140, 145, 148, 150-
152, 157-158, 160, 163, 165-166, 169, 173-174, 176-178,
181, 184-186
bonding, 82
compressibility, 43
consolidation, 82
density(ies), 34, 80, 123, 125, 135, 158, 160, 174, 181, 184-
186
elastic modulus, 81
quality, 37, 45, 48
strength, 2
surface, 148, 160
test piece, 44
smoothness, 10, 153-156, 158, 160, 171, 185
strength, 16, 20, 27
stretched, 27
strip, 26, 138, 147-148, 155
structure, 1, 5-6, 38
surface, 1-2, 4, 6, 8, 17-18, 27, 38, 43, 50-51, 58, 115, 135-
136, 139-140, 145-149, 151-153, 155, 157-158, 160, 163,
165, 167-170, 178-179, 184-185
coating, 184
energy, 136, 139, 146, 165, 167
215
friction, 51
-liquid interaction, 146
properties, 136
roughness, 43, 50-51, 58, 135, 151-152, 155, 160, 167-169
smoothness, 158, 185
tension, 136
unevenness, 27
tensile testing, 25, 28
test(ing), 9-11, 14, 38, 53, 137, 150
thickness, 3-4, 18, 27, 43, 54-55, 59, 62-65, 80-81, 84-85, 135,
151-1523 155
tissue, 53
topographies, 153
unprinted, 141, 144-145
unwrapped rolls, 11
web, 40
wetting, 136
writing paper, 29, 150, 160-161, 164, 167
Xerographic copy paper, 156
Paperboard, 34, 54-55, 77-79, 131-133, 170-171, 179
Papermaking, 3, 20, 90, 173, 186-187
operations, 90
production process, 186
stock dispersion, 3
Papir- og fiberinstituttet AS, 173
Parker Print-Surf (PPS), 154-156, 167-168
S-10, 155, 167-168
Particulate, 18, 181, 183
Pattern, 84, 115, 117, 131
Patterned buckling, 118
Peak load, 83-84, 103, 116-117
Peak pressure, 177-179
Peak stress, 85
Pendulum, 105-106, 174-176
pivot, 175
release, 175
travel, 175
216
Index
Penetration, 135-136, 139-140, 150, 157-158, 169

length, 136
Performance, 47, 146, 156, 163, 169
Permeability, 151-152, 160, 170
Photodetector, 142-143
Physical property(ies), 8, 44, 53, 119, 124, 159, 163, 169, 181-
182
Physical testing, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28,
30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52-54, 56, 58, 60, 62,
64, 66, 68, 70, 72, 78, 80, 82, 84, 86, 88, 90, 92, 94, 96, 98,
100, 102, 104, 106, 108, 110, 112, 114, 116, 118, 120, 122,
124, 126, 128, 132, 134, 136, 138, 140, 142, 144, 146, 148,
150, 152, 154, 156, 158, 160, 162, 164, 166, 168, 170, 172,
174, 176, 178, 180, 182, 184, 186, 188
Physics, 1, 54
Pine ‘straw’, 2
Plantema equation, 106
Plastic, 10, 16, 27, 42, 45, 118
deformation, 27
film, 45, 118
creasing, 118
Plate, 3-4, 118, 123, 153-154, 160
failure load, 118
Platen, 43-44, 49-50, 55, 63, 77, 83-84, 89, 151
Pneumatic, 151, 175-176, 178-179
Poisson constant ratio, 38
Poisson ratio, 26, 68, 84
Polar plotting, 139
Polymer, 54, 56
Polymeric, 19, 29
Polysaccharide, 19
Pore, 136-137
Porosity, 8, 135, 140, 151, 159-161, 169
Porous, 17-18, 139-140, 169
Preparation, 15, 48, 96-97, 173
Press, 10-11, 15, 18, 20, 34, 53, 102, 121-123, 125, 174
room, 11, 174
217
Pressed, 3, 16, 80, 123, 125-126

Pressing, 15, 79-81, 90, 123, 125-127, 135, 160
Pressure, 15, 42-43, 49-51, 81, 99, 125, 151, 153-155, 169,
177-179
high, 49, 125
low, 49, 151, 169
Print, 7, 139, 153-156, 160, 167, 171
direction, 7
Printability, 135, 137, 139, 141, 143, 145, 147, 149, 151, 153,
155-157, 159, 161, 163, 165-169, 171
Printed, 1, 8, 57, 135, 139-140, 144, 155, 166-167
character integrity, 167
image, 155
ruling lines, 139
sample, 139
Printing, 10, 17-19, 39, 43, 54, 57, 136, 139, 145, 151, 155,
160, 165-170, 186
high-speed, 151
speed, 136
test, 165
Profilometer, 152, 155-156, 169
Propagation, 7, 37-38, 42, 45, 49, 51, 66, 185
Pulp, 3-5, 12, 18-21, 27, 35, 37, 45, 53-54, 61-62, 64, 66-67,
80-81, 83, 88-90, 95, 131-132, 139, 141-143, 149, 160, 173,
175, 177, 179-187
beaten, 181-182, 184
fibres, 184
beater curve, 183, 185
beater housing, 180
beating, 173, 179-182, 185, 187
coatings, 160
fillers, 160
freeness, 173, 185
mill, 179
stock, 27, 53, 90, 139
suspension, 173
treatment, 160
218
Index
Pulsation, 15
Punch, 59-60, 90, 151, 175-176
cutter, 59-60, 90
Puncture, 17, 177
Q
Quality, 8, 12, 23, 27-28, 34, 37, 44-45, 48, 53, 82-83, 90, 96,
98, 139, 146, 153-154, 156, 158, 174, 177, 180
check, 28, 44, 53
control, 23, 96, 98, 153, 177
screening, 48
testing, 37
Quantify colour, 141
R
RDM Corporation MICR qualifier reader, 166
Reading direction, 7
Recycle, 93-94
Recycled, 9, 15, 95
pulp linerboard, 95
Reel, 14, 39-41
Refining, 80, 90, 125, 160, 173, 177, 179, 183, 185-186
energy, 125, 185
level, 177, 183, 185
pulp, 173
Reflect, 38, 52, 145
Reflectance, 146, 157, 174
Reflected, 8-9, 145-146
light, 8-9, 146
Reflecting, 72, 97
Reflectivity, 143, 157
dial indicator, 157
of paper, 143
Refractive index, 184
Relative, 10, 13, 26, 37, 50, 53, 59, 65-66, 70, 79, 104, 174
219
humidity (RH), 10-14

Research, 1, 20-21, 34-35, 55, 131, 145, 173, 186-187
Resin, 108-110, 112, 116
-embedded end method, 110
Retention, 181, 185
Rigid, 55, 131, 181, 186
Rigidity, 57, 67-68, 105-107, 114-115
Ring, 47, 55, 83-84, 86, 90, 131, 150, 153, 156, 177-179
crush test (RCT), 47-48, 83-84, 86, 88-96, 131
cylinder end surface, 153
structure, 47
Roll, 7, 14-15, 17-18, 23, 39, 100, 102
Room temperature (RT), 10, 84
Rotation, 59
Rotating knife fixture, 99
Rotogravure, 5
Rough, 18, 153, 155-156
Rougher, 17, 62, 160, 167-168
surface, 167-168
Roughness, 8, 43, 50-51, 58, 135, 139, 151-155, 159-161, 167-
169
Rubber, 17, 42-43, 150, 155-156, 177-178
backing, 155
diaphragm, 17, 177-178
Rupture, 177-178
pressure, 177
Rush/drag ratio, 41
S
Sanitary tissue, 19
Scatter, 94, 160, 174
Scattered light, 142
Scattering, 32, 81, 135, 145, 158, 183-186
filler, 135
high-scattering filler content, 135
properties, 183
220
Index
Scored, 17, 71-72

Scoring, 102
Screening, 48
Setterholm, 43, 55, 63-64, 77
thickness, 64
Shear(ing), 38, 42, 59, 65, 70, 72, 75, 77, 104-107, 111, 114-115,
173, 177, 180
energy, 177
rigidity, 105-106, 114-115
stiffness, 72, 75, 105, 111
Sheffield, 50, 153-155, 159-161, 171
air leak roughness, 153
method, 155, 171
roughness, 50, 154, 159-161
Short-span compression, 12, 29, 45, 80
test or strength (SCT), 12-13, 45-48, 80, 83-97, 115, 119-120,
123-128, 131
Show-through, 135-136, 140-141, 145, 160, 163-164, 166, 169
Shrinkage, 40
Single-face, 123-124
Single-facer corrugator, 123
Single-wall, 72-73, 109, 113, 118-119, 123
A-flute, 118
board, 72, 119
C-flute, 118
Size, 8, 14, 28-30, 32, 34, 37, 59, 71, 97-98, 117, 124, 148, 157,
162-163, 171, 178
Sizing, 135, 139-140, 150, 157-158, 160, 165, 168-169
agent, 140, 168
Slack, 25
Slenderness ratio, 85, 97-98
Slitter-scorer, 71
Slotted die, 137, 139
Slurry(ies), 2, 5, 7-8, 45, 49
Soft caliper, 61-64, 66-67
Soft-platen, 43-44, 50, 55, 63, 77
caliper, 43-44, 50, 63, 77
221
measurement, 43
-platen method, 55
Softness (tissue and towel), 5, 19, 49, 51-53
Softwood, 4, 12, 45, 86, 90, 179
/hardwood ratio, 90
kraft linerboard, 12
unbleached chemical pulp, 4
Solid, 37-38, 81, 153, 160
Sonic, 37-39, 45, 48-49, 51
in-plane testing, 39
propagation, 37-38, 45, 49, 51
signal, 51
testing, 37
SoniSys, 38, 42-43, 49
instrument, 42-43, 49
Sound speed, 38, 44-45, 53
Specific modulus, 39
Specific stiffness, 40, 45
Specimen height, 98, 106, 108-109, 114
Spectrometer, 146
Speed, 14, 27, 37-39, 41-45, 53, 79, 102, 136, 139, 151, 155,
166, 170, 180
of sound, 27, 37-38, 42-43, 53
Standard deviation, 16, 90, 175
Starch, 8, 79, 123, 173
application, 79
Stein-Hall starch adhesive, 123
Sticking, 1, 8
Stiffness(es), 25, 27, 30-32, 34, 38-40, 43-45, 47-48, 55, 57, 59,
61, 63, 65, 67, 69, 71-73, 75, 77-78, 86-87, 104-105, 111, 115,
117, 119, 126, 133, 139, 186
orientation, 38, 40, 86, 139
Stock jet, 39, 53
Stock preparation, 15
Stock slurry, 5, 49
Stock to wire speed ratio, 41
Stock velocity, 14
222
Index
Strain, 1-2, 26-28, 45, 59, 65, 77, 81-82, 97, 116, 177-178
hardened, 27
to break, 177
to failure, 28
Straining, 5, 16
mesh belt, 5
wire mesh, 16
Strength, 2, 4, 8-20, 23, 26-32, 34-35, 45-48, 50-51, 53, 55, 57,
73, 75, 79-83, 85, 88-89, 91, 93, 96-97, 106, 111, 113, 115,
118-119, 123-129, 131-132, 175, 177-178, 181-182, 185-186
additive, 45
moisture dependence, 14
properties, 4, 9-12, 14, 16, 19, 23, 29, 45, 51, 82, 186
stiffness, 34
Stress(es), 6, 17, 26, 38, 40, 45, 57, 59, 65, 81-82, 85-88, 92, 104,
118, 185
transfer, 185
Stress-strain, 45, 59, 65, 81-82
curve, 45, 59
tensile test, 65
Stressed, 17, 29
Stressing, 28
Stretch, 16, 27-28, 178
to failure, 16
Strip, 23-28, 30-32, 39-41, 45, 47, 81, 83-85, 89, 105, 137-139,
144, 147-148, 155
length, 24, 85, 155
width, 84
Stylus, 152, 155-156, 167-169
profilometer, 152, 155-156, 169
Sumitomo, 99, 109, 111, 113
clamp, 111, 113
Swedish Technical Forestry Institute (STFI), 149, 170
Synthetic paper, 61
223
T
Taber, 44, 55, 58-60, 64-67, 77, 116, 119-120
instrument, 65, 67, 116
stiffness, 67
Tactile softness, 5, 19
Tear(ing), 1, 7, 16, 28, 151, 173-177, 179, 181-183, 185-187
direction, 7, 175
index, 182-183
length, 174
resistance, 28, 186
test, 174-177
Technical Association of the Pulp and Paper Industry (TAPP), 12,
21-22, 28-29, 34-35, 46, 50, 53-55, 61-64, 66-67, 77-78, 80, 83,
88-89, 97-102, 109-110, 113-114, 120, 131-133, 141-142, 144-
145, 147, 157, 170-171, 173-174, 181, 186-187
TAPPI T 205, 173, 181, 186
TAPPI T 227, 173
TAPPI T 411, 54, 62
TAPPI T 414, 174
TAPPI T 494, 29, 34, 63
TAPPI T 524, 141-142, 170
TAPPI T 530, 157, 171
TAPPI T 811, 97-98, 101, 109, 132
TAPPI T 826, 80
TAPPI T 838, 100-101
TAPPI T 839, 98-99, 101-102, 109-110, 113-114, 132
Technidyne BNL-3 Opacimeter, 145
Technidyne Brightimeter S5, 141
Temperature, 10, 14, 144
Tensile energy, 28
Tensile load, 26-28
Tensile load displacement, 27
Tensile properties, 23, 25, 27, 29-31, 34-35, 174, 185
Tensile stiffness (Sb), 25-27, 32, 34, 38-39, 43-48, 57, 59, 61, 63-
66, 67-69, 71-73, 75, 77, 86-87, 94, 104, 126, 129, 139
Tensile stiffness index (TSI), 39-40, 87, 92
224
Index
Tensile stiffness orientation (TSO), 38-39, 41, 45-48, 86, 92, 139
Tensile strain, 81
Tensile strength(s) (St) 15, 23, 25, 28, 34, 53, 63, 79-80, 82-83,
86, 94-95, 177-178, 182-183, 185-186
Tensile stress, 81
Tensile stress-strain, 81
Tensile test(ing), 23-32, 34, 57, 65, 70, 77, 79, 81-83, 85, 87, 89,
91, 93, 95, 97, 99, 101, 103, 105, 107, 109, 111, 113, 115, 117,
119, 121, 123, 125, 127, 129, 131, 133, 174, 178
Tension, 6, 8, 16-18, 24, 27, 39-40, 57, 82, 90, 136
Test frame, 23-24
Test orientation, 34
Test parameters, 34
Test piece, 28-30, 37, 44, 47-48, 70, 77, 90, 98-100, 103-104,
107-109, 113, 149-150, 174, 178-179
assembly, 174
board, 70
height, 104, 107, 109
structure, 48
Test sheet, 38, 42, 48, 85
Test size, 29, 34, 37, 59
Test span, 28, 75, 77, 131
Test specimen, 8, 14, 28, 58-59, 66, 68, 82, 84-85, 97-98, 101,
106, 108, 111, 131
size, 8, 14, 28, 97
Test strip, 24, 28, 81, 83, 89, 137, 139
edge, 41
Testing time, 29
Thermomechanical pulp, 18
Thermomechanical refiner pulp, 5
Three-point, 70-73, 75, 77
bending, 70, 75
stiffness, 71-73, 75, 77
test span, 75
Time, 1, 5, 11-13, 15, 24, 29, 38, 49, 51, 90, 116, 136, 139, 142,
144, 146-148, 150, 152, 157
Tissue, 5, 19, 42-43, 49, 51-53, 180
225
handfeel measurement, 52
sheet, 42
softness, 5, 19, 49, 51-53
tactile softness, 5
Titanium dioxide (TiO2), 135, 145
Top printed side, 140
Top side, 16, 90-91, 139, 174, 178
Topography, 17, 151, 155
Torsion, 105-106
pendulum, 105-106
Toughness, 28, 174
Towel, 19, 42-43, 49, 52-53
paper, 52
sample set, 52
sheet, 42
softness, 5, 19, 49, 51-53
Transducer, 38, 42-43, 175, 177
Transmission, 50, 174, 185
Transmitted light, 2, 8-9
Transparency, 61, 135, 145
Transverse, 23, 28
Tristimulus functions, 142
Tsai-Wu failure criterion, 118
Twin-wire formed sheet, 18
Twin-wire forming, 18
Twisting, 105
Two-point, 44, 58, 60, 65, 77, 116
bending, 58, 65, 116
strain, 65
stiffness, 60, 77
U
Ultimate strength, 27, 118
Ultrasonic(s), 37, 39, 41-45, 47-49, 51, 53-55, 86, 92, 132, 170,
185
measured, 45
226
Index
measurement, 42, 47, 132

out-of-plane z direction of paper, 49
propagation, 185
stiffness, 47-48, 86
tensile strength, 48
Unbleached kraft, 18, 29-30
linerboard, 29-30
Universal testing machine (UTM), 45, 108-109
V
Variability, 8, 14-16, 34, 85, 90, 131
Variation, 8, 15-16, 39, 48, 85, 90, 109, 139-140
Velocity, 5, 14, 38, 51, 53, 109
Vertical, 42, 49, 67, 75, 84, 90, 97-98, 104, 108-109, 114-116,
118, 124, 155, 163
compression, 90, 115
displacement, 118, 155
load, 67, 75, 84, 115, 118, 124
-stacking load, 116
Vertically loaded beam, 104
Vertically loaded buckling load, 104
Viscoelastic, 29, 45, 151
compressible material, 151
properties, 29
Viscosity, 136, 169
Viscous, 18, 135-136, 160
Visible light transmission, 185
W
Wall, 4, 72-73, 80, 109, 113, 118-119, 123, 136
material, 136
Warp, 9
Water, 1-2, 5, 7, 9-10, 16-18, 146-150, 167, 169-170, 173, 180-
181, 185
absorption, 149-150, 167, 169
227
drained, 5, 7, 18, 173

drop, 146-149
angle, 146
contact angle, 146, 148
droplet, 148
exposure, 149-150
/paper mix, 2
Wavelength, 8, 143, 174, 185
Waxed-end method, 103
WC 3526C, 116
WC 4226C, 108-109, 120
Web, 39-40, 174
Weight, 8-10, 13, 15-16, 18-19, 30, 34, 38, 45, 47, 49-51, 61, 73,
79-80, 83, 85-90, 93-94, 97-99, 101-102, 106, 112, 119-120,
125-126, 131, 135, 140, 145, 149-151, 154-155, 157-160, 165,
175-176, 181-184
difference, 149-150
Weighing, 149, 151
Wet, 2, 5, 10, 16, 18-19, 32, 79-81, 90, 123, 125-127, 135, 157,
160, 184
-end stock chest, 157
-laid, 2, 5, 16, 18-19, 32
fibre stock draining process, 18
process, 5, 32
stock straining process, 16
strained slurry, 2
straining, 16
pressed sheet density, 80
pressing, 79-81, 90, 123, 125-127, 135, 160
density, 90, 126
strength, 19
suspension, 184
Wettability, 146, 165, 170
Wettable, 146
Wetting, 9, 136, 148
Wheel, 137-140, 143, 157
White, 19, 30, 142-143, 145, 157-158, 174
228
Index
appearance, 19
background, 145
light, 142-143, 174
spectrum illumination, 143
standard background, 145
Width, 1-2, 25-26, 28-29, 32, 38, 44, 57-58, 63, 69, 75, 80-81,
84, 86, 99, 115, 117-118, 139
Wire, 5, 7, 14-18, 39, 41, 53, 79, 90, 139, 174, 178
fabric, 17
side, 16-18, 90, 139, 174, 178
speed ratio, 41, 79
Wood, 1-6, 15, 18-20, 173, 177
fibre, 18-19
-based materials, 18
wet-laid paper, 19
Wood species, 177
Writing, 17-19, 29, 43, 135-136, 139, 142-143, 146, 148, 150,
157, 159-161, 164-165, 167, 169, 174, 180, 186
Writability, 135-137, 139, 141, 143, 145-147, 149, 151, 153,
155, 157-161, 163-165, 167, 169, 171
X
Xerographic copy paper, 156
Y
Yellow, 141-142
Yellowing, 5, 142
Yield, 27, 118
density, 160
point, 27
Young’s modulus, 81
229
230
Publ
ishedbySmi
t
h er
sPi
ra,2017
ThePhysi
calTest
ingofPaperr
efl
ect
sdecadesoft
heaut
hor
’sexper
ienceas
a research s cient
is
ta nd lab ma nager provi
ding industry cli
ents
,
manufact
urers,productdevel
opers,ma r
keti
nga nddi
s t
ri
buti
ono rgani
sati
ons
wi
thd at
at oa nswerq uer
iesregardi
ngp roductqual
it
yc oncerns,vari
abi
li
ty
,
r
unnab i
l
it
y,con ver
ti
bil
i
tyandp ri
ntabi
li
ty
.Th eb asi
cprinci
plesund er
lyi
ngthe
var
ioustest
ingme thodsareusedtoil
lustr
atehowtheiri
nter
relat
ionshi
pslead
t
ov al
idat
edfindingsands ol
vi
ngp r
o bl
ems .
Thisb ookc overstheb asica ccept

eds tandardindustr
yme chani
caltests
suppleme nt
edb yult
rasoni
c me thodsap pl
iedtoexamp l
esofcomme r
cialand
l
ab orat
oryhan dsheetsamplesets,pr
e sent
ingthetest
ingtechni
que,dataa nd
analysi
s. Fo cusi sc oncentr
atedo nt het est
st hatare mo stfr
e quent
ly
requir
ed,sucha stensi
lea ndcomp r
essionstrengt
hs,sti
ff
nessforpapersa nd
corr
u gat
edb oard,andre l
evantwa t
era bsor
p t
ioncharact
eri
sti
cs.Itisa i
me d
attheinter
e st
edp aperindust
rytechnologi
storresearcheratanintr
od uct
ory
l
eve lwh o wishestoe stabli
shaf undame nt
alu nder
standi
ngo fwh a tthe
physicaltest
ingr esul
ts me an,ho wt oa voi
dc ommo np i
tf
all
sa nd mo st
i
mp ortant
ly
, howtointer
p r
ettheresult
sfrom ap aperphysi
csp oi
nt-
of-
view.
Shawbury,Shr
ewsbury,shrops
hir
e,SY44NR,UK
Tel
ephone:+44(0)1939250383
Fax:+44(0)1939251118
Web:www. pol
ymer-
books .
com

Physical Testing of Paper

Uploaded by

Copyright:

Available Formats

Physical Testing of Paper

Uploaded by

Document Information

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

Physical Testing of Paper

Uploaded by

Copyright:

Available Formats

Physi

Roman E. Popil, Ph.D.

A Smithers Group Company

Shawbury, Shrewsbury, Shropshire, SY4 4NR, United Kingdom

©Smithers Information Ltd., 2017

All rights reserved. Except as permitted under current legislation no part

ISBN: 978-191024-292-6 (Softback)

Typeset by Argil Services

The ‘Physical Testing of Paper’ is a broad term which covers the

submitted paper samples that disallows them to be nicely printed,

The book is written from a personal view based on my initial

2. Handbook of Physical Testing of Paper, 2nd Edition, Eds.,

3. K.J. Niskanen in Paper Physics Book 16 of Papermaking

4. J.E. Levlin and L. Söderhjelm in Pulp and Paper Testing,

5. J.D. Peel in Paper Science and Manufacture, Angus Wild,

6. G.A. Smook in Handbook for Pulp and Paper Technologists,

1 Introduction –What is Paper?................................................ 1

3 Ultrasonic Testing of Paper.................................................. 37

3.3 Correlation of Ultrasonic Results with Mechanical

4 Bending Stiffness of Paper and Corrugated Board: The

5 Compression Testing of Paper, Board and Boxes:

6 Testing Methods for Measurement of the Writability and

7 ‘Beater Curves’ to Evaluate Pulp Potential: Burst, Tensile,

One of my former physics professors upon hearing that I had started

process is reversed, i.e., a copious amount of water is added to paper,

Figure 1.1 A naturally occurring paper analogy can be mats and

Here, the aspect ratio of fibre width to length is similar to paper

be said to resemble a plate of cooked fettuccine pasta. The wood

Figure 1.2 A confocal 2-photon fluorescence scanning laser

63.9 µm 52.0 µm 60.6 µm

Figure 1.3 Optical microscopy cross-section of newsprint

So, here is a model to bear in mind, paper is reconstituted wood

Some undergraduate structural-engineering classes are tasked with

The properties of paper are similarly determined, to a large extent, by

comparison to chemical pulp fibres and are found in newsprint and

1.1 Paper Structure, Paper Models: Fettuccine, Straw

Most paper is made using the wet-laid process at speeds of 1,000 to

various states of dewatering, drying and compacting, is pulled

Fibres: former wood cells, typically

Paper consists of 30–50% air and a network of bonded fibres

Figure 1.4 Schematic representation of paper structure showing

This directionality in paper may be appreciated by attempting to

newspapers are pulled left to right in the reading direction. A paper

Figure 1.5 A machine-made paper roll with the MD indicated

1.2 Formation of Paper: In-plane Lumpiness

Paper consists of clumps of intertwined fibres sticking to each other

high to avoid samples that are below a set product specification.

Figure 1.6 Left is a transmitted light image through corrugated

1.3 Hydrophilicity – Paper Really Sucks – Water That Is!

wet – dry (weight)

In an arid environment, such as Mexico City in winter, the paper

Accordingly, paper is manufactured to specifications for an

Figure 1.7 A hypothetical paper moisture curve showing the

If paper is left to equilibrate to a high-humidity environment (>50%),

The question often arises as to what the duration of preconditioning/

Figure 1.8 shows the exponential change in time of a moisture-

Figure 1.8 Hypothetical %M, strength, basis weight versus time

1.3 SCT at 50% RH

0.8 SCT%M/SCT50%RH = 0.531e0.089%M

Figure 1.9 SCT of kraft linerboard measured at various humidities