PEER REVIEWED

PRESSING

A systems approach to felt

conditioning

ABSTRACT

DOUGLAS F. SWEET

HE EXTENT OF FELT CONDITION-

ing varies by grade, machine

speed, furnish (recycle or
virgin fiber), paper machine
age, and, obviously, papermaker
knowledge and experience. Sometimes, a little knowledge is dangerous, and the frequently misused solution to a problem, more is better,
quite often is applied. Many mills
have experienced felt conditioning
problems and make changes in the
process, such as more vacuum, more
slot width, more showering, and
even more uhle boxes. Sometimes
the process is improved, but frequently there are negative results.
There have been several press
rebuilds which occurred partly due
to poor application or lack of felt
cleaning equipment. A discussion of
the above items and variables, as well
as some actual case studies, should
lead to a better understanding of the
entire felt conditioning process.
FELT MOISTURE RATIO VS. PERCENT MOISTURE
Typically, felt moisture is surveyed
with a Scan-Pro, a device which
measures the amount of water in the
felt as g/m2. This raw data is then
reported in two ways: moisture ratio
and percent moisture. A problem
exists here because these values are
often used synonymously, and they
are not the same. Two simple equations define these terms.
Moisture ratio =
water/Grams of felt

Grams

of
(1)

% moisture = Grams of water x

100/Grams of water + Grams of felt
(2)
Since both properties are unitless,
one may not be aware when an error
is made. For example, a felt with a
moisture ratio of 0.40 has the same
amount of water as the same felt
with 28% moisture. Often the felt
moisture studies are quickly looked
at, and the papermaker will see 28 is
less than 40 and not realize the felt
moisture is equal.
A worse situation exists when a
felt is reported to be at, say, 32%
moisture, and it is believed to be
drier than one with a moisture ratio
of 0.38. It is not. Typically, these two
values will not be used in the same
moisture study, but both will be used
in studies reported by different felt
vendors.
Furthermore, regarding moisture
studies, it is important to obtain
moisture values before and after the
uhle boxes. Even if a CD profile is not
possible due to press geometry, an
MD measurement should be taken. A
moisture level measured in a felt only
before a uhle box is about as practical as trying to estimate reel moisture
based on a sheet moisture measurement in the middle of the dryer section.
UHLE BOX SLOT WIDTH VS. DWELL
TIME
Uhle boxes, or felt boxes and felt
tubes, as they also are called, are the
devices which remove water and
contaminants from a wet or press

Felt conditioning is a very important

part of the total papermaking
process, but it is often misunderstood.
There are many significant felt conditioning variables, including felt design,
machine speed, vacuum level, vacuum
factor, dwell time, uhle box and cover
design, showering, vacuum system layout, and chemical cleaning. Often the
cause/effect relationship of these variables is not obvious.This results in
inadequate sheet dewatering in the
press, crushing and other quality
defects, short felt life, and sheet moisture variations.This paper reviews the
key factors in felt conditioning and
presents actual case studies where
measurable results have been
achieved.
Application:
This paper would be useful in troubleshooting and optimizing press felt
conditioning, resulting in improved felt
performance.

felt. They appear simple and have no

moving parts. However, a lot of good
engineering goes into the correct
design and application of uhle
boxes.
The first two variables in selecting a uhle box are slot width and felt
width. Slot width is proportional to
machine speed and is responsible for
dwell time over the vacuum zone.
Dwell time is the time, usually
expressed in milliseconds, that the
felt is exposed to vacuum.
A revision of TAPPI TIS 014-55,
Air Flow Requirements for Conditioning Press Felt at Suction Pipes,
has been underway for the last three
or four years. The revised Technical
Information Sheet, TIS 0404-27, was
published in early 1996. Extensive
research was conducted to predict
moisture levels in modern press fabrics. One of the most important variables was found to be dwell time.

VOL. 80: NO. 7 TAPPI JOURNAL

103

PRESSING

Felt/machine speed,
m/min (ft/min)

Total slot width,

mm (in.)

Below 300 (980)

300-450 (100-1500)
450-600 (1500-2000)
600-750 (2000-2500)
750-900 (2500-3000)
900-1050 (3000-3500)
1050-1400 (3500-4500)
Above 1400 (4500)
I. Guide for determining slot width

A dwell time of 24 milliseconds

yielded successful water removal (1,
2). The formula for determining slot
width is:
Slot width (mm) = Dwell (milliseconds) x Machine speed (m/min)/60(3)
Slot (in.) = Dwell time (s) x Machine
speed (ft/min)/5
(4)
This formula is used to determine
the recommended total slot width
on each uhle box, not each slot. For
example, a calculated slot width of
25 mm (1 in.) would typically be
applied using a double slotted uhle
box with two 12.5 mm (0.5 in.)
slots. Rearranging the equation to
calculate dwell time on an existing
uhle box:
Dwell time (milliseconds) = Slot width
(mm) x60/Machine speed (m/min)
(5)
Dwell time (s) = Slot width (in.) x
5/Machine speed (ft/min)
(6)
A dwell time significantly less than 2
milliseconds can lead to incomplete
water removal by the uhle box. Conversely, a dwell time over 4 milliseconds does not increase water removal
enough to justify additional vacuum
capacity.
Slot widths vary typically between
12 and 25 mm (0.5 and 1 in.). A slot
less than 12 mm could plug or bridge
over. A slot width of 12 mm is usually

104

TAPPI JOURNAL JULY 1997

12 (0.5)
15 (0.6)
20 (0.8)
25 (1.0)
30 (1.2)
35 (1.4)
40 (1.6)
46-50 (1.8-2.0)

Uhle box capacity,

m3/min (ft3min)
20 (700)
34 (1200)
54 (1900)
79 (2800)
110 (4000)
160 (5600)
230 (8000)
290 (10000)

Uhle box diameter,

mm (in.)
150
200
250
300
350
400
450
500

(6)
(8)
(10)
(12)
(14)
(16)
(18)
(20)

II. Recommended uhle box diameters

suitable for machine speeds up to

1000 ft/min. A slot greater than 25
mm could cause excessive felt wear.
Usually, after the desired total slot
width on a uhle box exceeds 25 mm,
a double slotted box is used.
The research involved in the new
Technical Information Sheet showed
there was no meaningful difference
in felt moisture exiting the uhle box
due to single or double slots of the
same total width. A guide for determining the total slot width, per uhle
box, based on a dwell time of 2 milliseconds, is shown in Table I.
UHLE BOX DESIGN AND VACUUM
FACTORS
The next item after determining slot
width and resulting area based on
felt width is to apply the vacuum
factor. TAPPI TIS 0502-01 recommends a minimum of 660 m3/min/m2
(15 ft3/min/in.2) open slot area (3).
Furthermore, it is stated that factors
have ranged from 750 to 970
m3/min/m2 (17 to 22 ft3/min/in.2),
especially on felts heavier than 1375
g/m2 (4.5 oz) and long nip or shoe
presses. The uhle box vacuum capacity is determined by the formula:
Vacuum = slot area x vacuum factor
capacity
(7)
Now that correct vacuum capacity
has been determined, an often forgotten item is to properly size the
uhle box diameter. On new or
upgraded vacuum systems, much

effort is spent to minimize vacuum

losses within the vacuum system
piping. However, uhle box diameters
are usually ignored. Since uhle boxes
are conveying vacuum capacity
across the felt width while carrying
liquids and solids, it makes sense
to size them as is done with
vacuum piping. Typically, 1820 m/s
(35004000 ft/min) are velocities
used for vacuum piping with liquid
water (prior to a separator), and
2830 m/s (55006000 ft/min) are
velocities for vacuum piping with
relatively dry air (after a separator).
Therefore, it is recommended to size
uhle boxes using the wet air velocities. For high vacuum capacity flows,
it is necessary to increase these
velocities further because of size
restrictions within the press. Recommended uhle box diameters for various capacities are shown in Table II.
The quantity and placement of
uhle boxes is less of a science and
more of an applications experience.
Theoretically, one uhle box per felt is
adequate, assuming correct dwell
time and vacuum factor. However,
two uhle boxes may be used on felts
carrying more water, where sheet
fillers are used, on the fastest
machines, and for assuring good
moisture profile on high quality,
lightweight sheets. Machines operating below 610 m/min (2000 ft/min)
usually do not require two uhle
boxes per felt. Also, on the driest felt
positions, only one uhle box is
needed (3rd or 4th presses).

Placement of uhle boxes can be

from horizontal to vertical felt runs.
There have even been specially constructed uhle boxes placed upside
down in difficult press sections. Care
should be given to allow enough
space to accommodate oscillating
needle showers, chemical showers,
and lube showers ahead of the uhle
box. Only a lube shower is required
ahead of the second of two uhle
boxes.
Uhle box cover or wear strip
material and configuration also vary
widely. High molecular weight polyethylene is most common, but it has
many different formulations. These
differences affect life and cost and
should be evaluated. Ceramic cover
material has a relatively high initial
cost, but it can be justified over time.
Ceramic wear strip/cover material is successfully used on uhle
boxes. Its advantage is long life, especially on higher speed machines. Be
aware that not all uhle boxes will
accommodate ceramics due to structural and space limitations or lack of
rigidity. The price of ceramic is initially high and ranges from five to ten
times the price of polyethylene. Additionally, there are different grades of
ceramic with corresponding differences in prices. The lowest cost
ceramic is aluminum oxide; silicon
nitride is the most expensive.
Another advantage to using ceramics
is reduced or no pitch buildup on
the ceramic strips. This buildup
varies with mill and region.
The last major topic on uhle box
design is the design of the slot
opening. Traditionally, a straight slot,
multiple slots, or a herringbone pattern were the only choices available.
With the development of seamed
felts, other designs have evolved to
minimize seam wear. Some of the
new slot configurations have also led
to reduced felt wear in general and
may be worth consideration.

SHOWERING
Felt showering has experienced
extensive changes over the last ten
years. Some changes are due to the
need for cleaning modern press felts.
Other new developments are due to
new equipment technology. Developments of new designs and materials for press felts have resulted from
the requirements of long nip and
heavily loaded presses. These major
innovations in pressing occurred in
the early 1980s, leading to significant
innovations in felt design and construction. These innovations spread
through most paper and board
grades. Cleaning these modern felts
requires modern showering techniques. Continuous, not intermittent,
showering with a combination of
low pressure fan and lube showers
plus high pressure, oscillating needle
showers must be applied properly.
A properly designed felt conditioning system will allow continuous
use of low- and high-pressure showers (4). A common belief is that a
uhle box only extracts water that
was absorbed by the felt. Actually, in
a well designed system, only onethird of the water removed by the
uhle box comes from the sheet. The
other two-thirds of this water is from
low and high pressure showers. This
point leads up to the fact that contaminants in a felt cannot be removed
only by applying vacuum. Water is
required to loosen and help convey
debris from the felt.
Fan showers are for evenly wetting the entire width of the felt. They
operate at relatively low pressures of
34 bar (4060 psig). Most often
these are placed directly ahead of
uhle boxes as lubrication showers.
The shower must be designed to provide complete and even coverage of
the felt. Some mills are oscillating
these showers to minimize effects of
plugged or poorly operating nozzles.
It is important to have all fan nozzles
spraying uniformly. This also leads to

the use of showers with internal

brushes to aid in cleaning the nozzles on the run.
High-pressure needle showers
produce a high energy spray which
is concentrated on a very small area.
However, only a thin strip of clothing
is cleaned as it moves past the nozzle. It is not practical to have enough
needle nozzles on the shower pipe
to completely cover the felt. This
would also be prohibitive due to
pumping costs of high-pressure
water. The solution is to move the
shower axially, back and forth across
the felt. This movement should be at
a constant rate with little or no dwell
at the end of the stroke.
The mechanisms used to produce
the shower movement have included
pneumatic and hydraulic cylinders,
gear motors with crank-arms and,
more recently, self-contained, electromechanical shower oscillators. The
apparently simple, cylinder-driven
and crank-arm designs were limited
as they did not provide the uniform
stroke rate and instant directional
change for the shower. These oscillating showers provide a random cleaning of the felt and may completely
bypass some areas. This non-uniform
cleaning causes felt streaking and CD
sheet moisture variations in the
worst cases.
These self-contained, electromechanical shower oscillators were
introduced in the early 1980s. They
also had the feature of speed control.
This new feature achieved complete
coverage cleaning of the felt by moving the shower across the felt a distance equal to the width cleaned by
a single nozzle, with each full revolution of the felt. This motion can be
calculated by the equation:
S T/L = R

(8)

where

VOL. 80: NO. 7 TAPPI JOURNAL

105

PRESSING

KEYWORDS
Equations, equipment, felt conditioning,
hydraulic equipment, moisture, moisture content, problem solving, showers.

S = felt speed in m/min (ft/min)

L = loop length of the clothing in m (ft)
T = nozzle cleaning width in mm (in.)
R = stroke rate in mm/min (in./min)
The stroke rate determined by this
method is relatively much slower
than that observed with previously
used oscillators. The stroke rate
could vary from less than 6 mm/min
(0.25 in./min) for a long, wet felt on
a slow cylinder machine to 100 or
150 mm/min (4 or 6 in./min) for a
typical press felt on a high-speed
newsprint machine. Many times the
speed control for the electromechanical oscillator will be interlocked to paper machine speed controls so the machine speed for different grades can be tracked.
Shower stroke length is based on
the shower nozzle spacing. Most
nozzles for felt cleaning are placed
on 150 mm (6 in.) centers. The recommended stroke length is two
times the nozzle spacing, i.e., 300
mm (12 in.) stroke with nozzles on
150 mm (6 in.) centers. The reason
for this relationship between stroke
length and nozzle spacing is to provide 100% nozzle overlap in the
event a nozzle is plugged. Since needle showers have relatively small
nozzle orifices, typically 11.5 mm
(0.0400.060 in.) diameter, the use
of clean, filtered water is important.
Additionally, as with fan showers,
internal brush mechanisms are also
used with needle showers. In some
very critical applications, a pipewithin-a-pipe configuration may be
used to allow removal and maintenance of the shower and nozzles on
the run.
Usual pressures for needle show-

106

TAPPI JOURNAL JULY 1997

ers vary from 10 to 17 bar (150 to

250 psig). Higher pressures can damage the felt, and lower pressures do
not provide adequate cleaning
energy. Placement of these showers
is also important. Cleaning modern
press felts requires needle showers
to be located on the sheet side of the
felt. It is practically impossible to
clean these multi-layer, heavy felts
from the inside as was done only 10
or 15 years ago. The angle of the needle spray with respect to the felt will
vary with felt construction, shower
design, and papermaker experience.
There are many opinions on this
matter, so a new one will not be
introduced here.
The final and most important
point on shower placement is the
proper distance from the shower to
the felt. The best cleaning is
obtained at distances between 75
and 150 mm (3 and 6 in.) from the
felt. Much greater distances will
allow the needle stream to break up
and form small droplets, producing
small energy pulsations on the felt
and even damaging the felt.
VACUUM SYSTEM
The vacuum system consists of piping, separators, valves, separator
removal pumps, seal/weir tanks, and
the vacuum pumps. An excess of
vacuum pumps will not replace a
poorly designed system, undersized
piping, or lack of separation equipment. This paper began with determining the vacuum capacity and the
sizing and selection of uhle boxes.
This is the logical order of events
leading up to the system design and
vacuum pump selection.
One of the most important
points in the system design is to
have separate vacuum sources for
uhle boxes on separate felts. One
vacuum pump per felt will not allow
two or more felts of different porosities to affect each other. There is no
problem with connecting two uhle
boxes together on the same felt.

Next, the vacuum piping must be

sized (5). This is based on velocity, as
was done with the uhle box sizing.
As stated earlier, vacuum piping
before a separator should be sized
with a velocity of 1820 m/s
(35004000 ft/min). Piping after the
separator can be sized for 2830 m/s
(55006000 ft/min). The piping runs
should be horizontal or sloping
downward before a separator, never
uphill. After the separator, piping may
be routed as needed to reach the vacuum pumps. Vacuum piping is most
often stainless steel, but many systems have used carbon steel between
the separators and vacuum pumps.
The vacuum separator is an
important element of this system.
With the constant showering, there
are significant amounts of water,
fiber, contaminants, and chemicals to
remove before the vacuum pumps.
Separator configuration may vary,
but most designs employ a tangential
inlet. Since the operation of a separator is primarily affected by the internal velocity, more efficient separation occurs at internal velocities
around 2.54 m/s (500750 ft/min).
Most separators are constructed of
stainless steel. The use of either one
or two separators on a felt with two
uhle boxes is acceptable. The choice
is usually based on space availability.
When two separators are used, the
vacuum piping may be joined prior
to the vacuum pump. However, the
water outlet piping should never be
joined. Each separator should have
its own barometric seal line or water
removal pump.
Water removal from the separators is also important. For most top
felt uhle boxes, a barometric seal line
can be used. Care must be exercised
in routing this piping to minimize
turns. Piping should be vertical
when possible or never more than
45 from vertical. Typical velocities
for sizing this piping are 12 m/s
(46 ft/s). A minimum pipe diameter
of 100 mm (4 in.) is recommended.

The other end of the seal line is

at the seal tank. Most modern installations use multi-compartment seal
tanks with one compartment per
separator and calibrated weirs for
flow measurement. This allows simple observation and assessment of
felt water removal. The seal tank
must be located at a sufficient elevation below the separator to overcome the effects of vacuum. This distance must be the vertical distance
measured between the bottom of
the separator and the liquid level of
the seal tank. The direct conversion
from kilopascals to meters of water
is 3.4 kPa/m of water (1.0 in. Hg vacuum per 1.14 ft of water). Since this
is a direct conversion, an additional 1
m (3 ft) should be added to allow for
friction and as a safety factor. Also
important is the volume of the seal
tank. A simple rule is to size the seal
tank volume for two times the volume of the seal line piping.
The final and also very important
step is to select the vacuum pump.
Heavier felt designs on todays
machines require higher vacuum levels for dewatering. As discussed earlier, vacuum factors between 660
and 880 m3/min/m2 (15 and 20
ft3/min/in.2) will cause the uhle box
vacuum level to reach 60 kPa (17 in.
Hg) and often 68 kPa (20 in. Hg).
Therefore, it is important to select a
vacuum pump capable of a fairly
constant capacity up to this maximum vacuum level. Additionally, it
may be necessary to install a vacuum
relief valve to limit maximum vacuum levels should the felt become
excessively filled and compacted.
CHEMICAL CLEANING
Some felts require more than
mechanical conditioning and cleaning with showers and uhle boxes.
Some paper and board grades have
furnishes and fillers that can adhere
to the felt and may be difficult to
remove unless chemical solvents or
detergents are used. An optimum felt

conditioning system utilizes both

mechanical and chemical means to
keep todays modern, synthetic felts
open and functioning.
Contaminants often found in the
analysis of old felts include soluble
and insoluble material (6). The shortest list is the insolubles such as fiber
and fines, clay, talc, titanium dioxide,
and silica. The mechanical action of
showers and uhle boxes can remove
most of these. The soluble materials
come from the pulp mill furnish,
recycled fiber supplies, and the broke
system. They can be separated into
three groups: acid, alkaline, and solvent solubles. These groups include
aluminum hydroxide, calcium carbonate, wet- and dry-strength additives, lignin, starch, size, fatty acids,
glue, latex, oil, grease, and wax.
Removing these and other contaminants with chemicals will require
examination by the chemical supplier. Treating these conditions may
be through batch, intermittent, or
continuous introduction of chemicals. In most cases, the chemicals are
applied to the felt through a stationary fan shower. Depending on the
application, the shower may be
located on the sheet or back side of
the felt. For continuous or intermittent chemical cleaning, the shower is
usually on the back side of the felt
and positioned as close as possible to
the point after which the sheet
leaves the felt surface. This provides
the most residence time to allow the
chemicals to work before the uhle
box extracts them. Again, the chemical supplier familiar with the mills
wet-end chemistry should be
involved.
CASE STUDIES
Case 1
The pick-up felt on a 7.1 m (280 in.)
linerboard machine was operating
with two 300 mm (12 in.) diameter
uhle boxes with herringbone covers
and a vacuum factor of 480
m3/min/m2 (11 ft3/min/in.2). This low

vacuum factor prevented the vacuum

level from reaching beyond about 40
kPa (12 in. Hg). The vacuum pump
was originally sized for two uhle
boxes with double 16 mm (0.625 in.)
slots and was capable of operating at
up to 68 kPa (20 in.Hg).However, the
slotted covers were replaced with
herringbone covers when this position was changed to a seamed felt.
The vacuum piping was properly
sized and installed and included a
vacuum separator. The separator
flowed through a barometric seal
line to a seal tank with a calibrated
weir. Showering was good, with an
electromechanical oscillator driving
the needle shower. A lube shower
was installed on each uhle box. The
primary problem with the system
was the excessive open area of the
herringbone covers. In addition, the
uhle boxes were undersized.
This system was improved with
the replacement of the uhle boxes
with new 350 mm (14 in.) diameter
units. Uhle boxes 400 mm (16 in.) in
diameter were recommended, but
350 mm (14 in.) units were the
maximum size that would fit.
More importantly, the covers were
designed to have about half the open
area of the previous herringbone
covers. This yielded a vacuum factor
of 880 m3/min/m2 (20 ft3/min/in.2).
The notable item in this study was
that there were no changes to the
system other than new, larger uhle
boxes with covers of less open area.
Results were significant and
immediate. Vacuum level at the uhle
boxes started at 34 kPa (10 in. Hg) on
new felts and was at 50 kPa
(15 in. Hg) within a week, due to
initial compaction. Water removal
increased by about 380 L/min (100
gal/min), as measured over the seal
tank weir. The moisture ratio of this
pickup felt dropped to 0.41, where
previously it was 0.55 to 0.60. The
measurably drier felt absorbed more
water from the sheet and increased
sheet dryness from the press section.
VOL. 80: NO. 7 TAPPI JOURNAL

107

PRESSING

Steam pressure was reduced to the

dryer at the same production rate.
Case 2
Another linerboard machine, in this
case a 100% recycled, 4.6 m (180 in.)
wide machine, was converted from a
conventional,two-bottom felt press to
a tandem bottom felt. The conversion
was done to improve runnability and
avoid a major press rebuild. This conversion led to a study of the existing
felt conditioning system. The system
was inadequate, with limited showering capability, undersized uhle boxes
and low vacuum capacity.Vacuum factors were below 440 m3/min/m2 (10
ft3/min/in.2). All uhle boxes were connected on a common vacuum header.
The entire felt conditioning system was replaced, including vacuum
pumps, separators, uhle boxes, and
showers. The system was designed to
allow separate vacuum sources for
uhle boxes on each felt. New vacuum
factors for the system were 790
m3/min/m2 (18 ft3/min/in.2).
Within 36 h after startup, production rates increased by 31%.
Case 3
A 5.6 m (220 in.) coated board
machine was operating with poor
felt life, 2830 days, and sheet defects
caused by high felt moisture. Felt
showering had to be intermittent

108

TAPPI JOURNAL JULY 1997

because the felts were already too

wet. The felt conditioning system
was studied, and several problems
were identified. These included a
common vacuum header for all uhle
boxes, undersized uhle boxes, an
excessive slot open area resulting in
a vacuum factor of less than 220
m3/min/m2 (5 ft3/min/in.2), two uhle
boxes per felt, no vacuum separators,
inadequate showering, and undersized vacuum piping. However, the
vacuum pumps were in good condition and were the correct size. They
would be reused.
The uhle boxes were replaced
with just one per felt and slotted covers yielding a vacuum factor of 660
m3/min/m2 (15 ft3/min/in.2). Vacuum
separators with low NPSH removal
pumps were installed, undersized
vacuum piping was replaced, and felt
showers were replaced with complete coverage designs. Again, the
vacuum pumps were not changed.
Two new felts were installed, and
immediate results were observed
shortly after startup. Machine speed
increased by 15 m/min (50 ft/min)
with continuous showering, and
many sheet defects disappeared. Both
felts were cut off after 40 days and
were determined to still have usable
life. Felt life is now 4550 days.

CONCLUSION
A little theory and common sense
applied to most felt conditioning systems can produce significant results.
Often these are not expensive solutions, and payback is fast. Gathering
and interpreting data correctly is the
key to problem identification. The
solutions are relatively easy. TJ
Sweet is regional sales director, Nash Pulp &
Paper Div., Birmingham, AL.
Received for review June 12, 1996.
Accepted Aug. 10, 1996.

LITERATURE CITED
1. TIS 0404-27, Air flow requirements
for conditioning press felts at suction
pipes.
2. Bennett, J. and Tehan, J., TAPPI 1994
Papermakers Conference Proceedings,
TAPPI PRESS,Atlanta, p. 535.
3. TIS 0502-01,Paper machine vacuum
selection factors.
4. Swett, G., Mechanical felt conditioning, Nash-CVN Systems, 1989.
5. Sweet, D. F., TAPPI 1988 Papermakers
Conference Proceedings, TAPPI PRESS,
Atlanta, p. 279.
6. Dickens, J. H., TAPPI 1989 Papermakers
Conference Proceedings, TAPPI PRESS,
Atlanta, p. 111.