Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
1045866
FIELD OF T~E INVENTION
The present invention relates to a method of improving
the properties of a mechanical pulp and of paper made there-
from. It relates particularly to the improvement of the
bonding properties of such pulp and the resulting surface
properties of such paper.
sackground of the Invention
By mechanical pulps are understood pulps produced prima-
rily by mechanical processing with or without auxiliary steps
of a chemical or physical nature. Such pulps include conven-
tional (stone) groundwood and refiner groundwood and pulps
produced by an array of chemi-mechanical and thermo-mechanical
processes. Such pulps generally have lower bonding properties
than chemical pulps and conventionally have been used, e.g.
in the making of newsprint, with substantial admixtures of
chemical pulp. The tendency has been to reduce more and
more such admixtures of chemical pulp and even to use mecha-
nical pulp alone. This trend has been greatly encouràged by
the development of the more recent processes of thermomechan-
ical and chemi-mechanical pulping and the improvements in the
properties of the resulting pulps.
However, the surface properties of paper produced from
furnishes of preponderantly mechanical fibers often present
problems, particularly accentuated in connection with changes
in printing technology, such as the growing acceptance of
off-set printing. One such problem is ~linting" i.e. the
phenomenon of fibers being picked out of the sheet in the
process o~ printing and accumulating on the printing press.
Related phenomena are scuffing and dusting and also "picking~
in both, wet and dry webs; and in most cases where the term -
"linting~ is used, it is meant to cover all these related
phenomena. A high linting propensity of the paper or
. ~', '~ . ,. . '
.
-
1045866
newsprint will be a nuisance to the printer and in certain
cases may cause the paper to be rejected by the customer.
The linting propensity ancl the associated other defi-
ciencies of the paper are essentially a surface phenomenon
and are due to the relatively weak bonding of and between
certain fibers on or close to the surface of the paper. As
such, the phenomenon is connected in some way with the type
of paper machine, the condition of forming, draining and so
on. But on any given machine there will still be differences
between paper obtained at different times and these differ-
ences are accounted for primarily by differences between the
pulps used.
It has been proposed to reduce the linting of paper by
applying adhesive materials to the surface of the paper,
however, this procedure adds considerably to the expenses
and is not always effective. It is of course, known to
improve the bonding properties of fibres by beating or re-
fining, with the attendant decrease in the freeness of the
pulp. However, this method of improving bonding properties
does not always solve the problem of linting. Furthermore,
excessive refining of the pulp may degrade its properties
and may also lengthen the drainage time beyond what is
acceptable in the papermaking operation. It is also known
that the bonding properties of a mechanical pulp are con-
nected in a positive way (and conversely, the freeness value
in a negative way) with the specific surface of the pulp,
a property which is measurable and which expresses the
ratio of the surface of the fibers to their weight (e.g.
in square metres per gram). However, pulps having substan-
tially the same average specific surface often have differ- -ent linting properties and some pulps of quite low freeness
give paper which still lints badly.
1045866
The specific surface may vary with the method by which
it is determined. The specific surface as referred to
herein is obtained using the Robertson ~ Mason method des-
cribed in "Specific Surface of Cellulose Fibres by Liquid
Permeability Method" Pulp & Paper magazine of Canada, page
103 December l9, 1949.
It is also known to separate from a pulp a "rejects"
fraction by means of a hydrocyclone, such fraction being gene-
rally very small and consisting of the relatively unrefined
particles, and to subject such "rejects" to further refining.
BRIEF DESCRIPTION OF THE INVENTION
We have found that the tendency to lint (insofar as it
is caused by the characteristics of the pulp and not by those
of the paper machine1 is connected not with the overall or
~average value of the specific surface of the pulp but with
the fractional distribution of fibers of different specific
surface in the pulp. The distinction is important. The same
average value may result from a summation of values which
are all close to the average but also of values which deviate -~ -
.. . .
widely from the average toward both extremes. A pulp mayhave a high average specific surface, yet if it contains a
substantial fraction of low specific surface, it will still
have a tendency to lint. To overcome the linting propensity
it i~ necessary that the pulp contain as small a fraction as
possible of lOw specific surface.
- Accordingly, the present invention broadly provides a
process for reducing the linting propensity of a mechanical
pulp and the linting of paper made therefrom in which said
pulp is fractionated into at least two fractions, one of said
fractions having a specific surface lower than the other
fraction, said fraction of lower specific surface is sepa-
rated and subjected to mechanical processing thereby to
increase the specific surface thereof, and the processed
. . .
- 5 -
--` i045866
fraction is recombined with said other fraction.
The present invention more specifically provides a
process wherein a mechanical pulp is fractionated into at
least two fractions, one of said fractions having a
selected average specific surface value between 1.2 and 4m2/g
and the other fraction having a specific surface greater
than said one of said fractions, said one of said fractions
is subjected to mechanical processing thereby to increase
the specific surface to above 4 and generally to not above
10 m2/g and said mechanically processed fraction is recomb-
ined with said other fraction.
The present invention also broadly relates to a method
of determining specific surface distribution of the fibre
in a mechanical pulp by fractionating in a hydrocyclone
system to provide a plurality of overflow and underflow
fractions and analyzing at least some of said fractions to
determine the specific surface of the selected fractions.
Brief Descriptions of the Drawings
The invention will be further illustrated by means of
the attached drawings in which
Figure l-represents a schematic flow sheet of one
embodiment of the invention integrated with a refiner ground
mill operation,
- Figure 2 represents the specific surface distribution
of fibers in a pulp, the curve being a plot of specific
surfaces versus weight fraction (in % of total pulp).
- Figure 3 represents the relationship between the linting
propensity of a pulp and the presence in the pulp of a frac-
tion of low specific surface;
Figure 4 shows a comparison between the specific surface
distribution in a thermomechanical pulp before and after
treatment in accordance with the present invention.
-- 6 --
- 1045866
Figures 5 a, b, and c illustrate schematically methods
of fractionating on the basis of specific surface by means
of systems of hydrocyclones.
DETAILED DESCRIPTION OF THE INVENTION
According to the invention, a mechanical pulp, e.g. one
produced from wood fragments in a refiner, is divided into
at least two fractions, one being characterized by a low
specific surface, and the fraction of low specific surface
is passed through a second refiner. The fractionation of
the pulp on the basis of specific surface i8 conveniently
- achieved by means of hydrocyclones. These are familiar
devices in the pulp and paper industry where they have been
used for many years to remove from pulp a ~rejects" fraction,
~.
usually consisting of shives, fibre bundles and heavy
particles. Usually the ~rejectsn from the cyclones are
kept as low as possible; in mechanical pulp they rarely
exceed about 10% of the pulp. Normally the material reject-
ed by the cleaners in a conventional system would have an
average specific surface of about .8 to 1, and would not
exceed 1.2 m2/g.
Fractionation by means of hydrocyclones takes place on
the basis of various physical and geometrical characteristics
of the fibers but we have found that for mechanical pulps
the predominant measurable characteristic which differentia-
tes the "overflow" from the "underflow" fraction issuing
from a hydrocyclone is the specific surface.~ (By "underflow~
fraction i8 understood the fraction issuing from the apex
and the hydrocyclone). Accordingly, in the present invention,
the pulp is passed into a hydroclyclone or a set of hydro-
cyclones adapted to produce an underflow fraction having a
specific surface not exceeding a certain value; the overflow
fraction will then have a value higher than said value. The
value of specific surface which is thus chosen as criterion
` ' .
- 7 -
1045866
may vary, depending on how exacting are the specification~
for the paper, the latter in turn depending, among othexs,
on the method of printing to be used, the type of ink and
so on. It has been found by studying lint obtained from a
S blanket of a commercial offset printing press from a printing
job of medium difficulty (two color, medium tack ink) that
the lint fibres have an average specific surface of approxi-
mately 2.5 m2/g. This average specific surface for lint
fibres would change depending on the difficulty of the prin-
ting job but normally is within a range of about 1 to 4 m2/g,.
In all cases individual lint fibres will be present having
specific surfaces above and below the average value, however,
the lint material will not be expected to contain significant
numbers of fibres with specific surfaces more than about
1 m2/g above the average specific surface of the lint mate-
rial for a given paper machine and printing operation since
material of this increased specific surface would tend to
bond into the sheet. Based on these findings it was deter
mined that to improve the lint qualities of a paper or a pulp
those fibres with the lower specific surface should be further
worked by mechanically refining so that the amount of lint
candidate material in the pulp for the specific application
for which the paper is intended would be reduced to accepta-
ble limits.
Hydrocyclones are generally set to fractionate a pulp
into fractions according to a ratio of underflow to overflow
fractions, such ratios being determined by the geometrical
features of the hydrocyclone (relative sizes of input to
output openings, cone angle etc.) as well as by conditions
of operation (pressure, consistency). With the present in-
vention the hydrocyclones will be set to produce an underflow
fraction of an average specific surface in the range of 1.2 -
4 m2/g preferably 1.5 - 4 m2/g is separated. The size
. ' ' :
-. , ~ . :
- . . ~ .
- 8 - 1~866
of this fraction will necessarily vary with the conditions
of pulping, type of wood, etc. In the case of refiner ground-
wood and in the freeness range within which most refiner
plants operate, the underflow fraction will amount to between
- 5 15-70% preferably 25-50% of the total pulp. The desirable
size of the underflow fraction may be determined in the mill
by varying the reject rate and determining the linting of the
paper produced or by preliminary tests in the laboratory,
- e.g. by multiple fractionation of a sample of the pulp using
a hydrocyclone and measuring the average specific surface of
each fraction to obtain the specific surface distribution of
the fibers in the sample. A method for determining the
fractional distribution of specific surface in a mechanical
pulp will be described in more detail hereinbelow.
The underflow fraction is passed through a refiner where
it is subjected to sufficient refining to increase the aver-
age specific surface to between 4-lOm2/g preferably to bring
its specific surface to a value at least equal to that of the
initial pulp and even more preferably to a value equal to
or above that of the overflow fraction. Refined underflow
fraction is then recombined with the overflow fraction and
the recombined pulp of improvea properties is further processed
as desired in a conventional manner and forwarded to the
paper machine.
Referring to Figure 1, WOOd chips, or other comminuted
cellulosic raw material, are fed through line 11 to refiner
10 in which the chips are mechanically disintegrated to form
a pulp. Refiner 10 represents diagramatically a refiner
plant, which may be operated in any of the known ways and
may thus include features which are not specif1cally shown
in the figure but are known ln tne art, SUCh as pre-steam1ng
the chips, maintaining tne requisite pressure in an enclosure
-- 9 --
- 1045866 around the refiner or otherwise controlling the conditions
of refini~g, chemical treatment accompanying any of the stages
of refining and so on. The resulting mechanical pulp is
usually passed to a tank (not shown) for latency development
and also to dilute the pulp to a suitable consistency and is
conveyed through line 12 to screen 13 which rejects oversize
particles or particularly long fibres in a conventional
manner. The rejects constituting generally about 5 - 10% of
the pulp are removed through line 21, and can further be
processed, if desired. The portion of the pulp which passes
through the screens is passed through line 14 into hydro-
cyclones, e.g. Centricleaners sold by C-E Bauer Company,
which fractionate the pulp into an overflow fraction removed
from the hydrocyclones via line 17 and an underflow fraction
removed via line 18. A two stage hydrocyclone fractionating
system is shown, represented respectively by hydrocyclones
15 and 16, but a single stage or a multiple stage system
greater than two stage may be used. The overflow fraction
in line 17 is the accepted fraction of the pulp having the
requisite specific surface diQtribution; as mentioned, it
will constitute about 30 - 85% percent of the pulp passed
through the screens. The underflow fraction in line 18 is
the fraction of low average specific surface (1.2 to 4 m2/g)
and will constitute about 15 - 70~ of the pulp (after the
screen). The underflow fraction is then cleaned, if desired,
in hydrocyclone 19 (Magnacleaner) to remove, in a conven-
tional way, dirt and other alien particles (ngrit~) and the
thus cleaned underflow fraction is passed through line 22
where, if desired, is combined with the screen rejects from
line 21. Obviously, the cleaning by means of a magnacleaner
may be omitted, if no cleaning is necessary, and similarly
the underflow fraction need not be combined with the screen -
, ' ' ' . , : .- '.............. - : ' ~ - , I
- , . . .
-- 10 -- .
~45866
rejects if it is preferred to treat these streams separately.
The underflow fraction from line 22 is then thickened, e.g.
in a press 23, to a consistency suitable for refining in a
refiner and passed into refiner 20 where it is refined to
an acceptable average specific surface value, preferably
one substantially equal to the specific surface of the
overflow fraction in line 17. The pulp from refiner 20 is
conveyed through line 24 and, either directly mixed with
the pulp in line 17 or subjected to further treatment
before the paper machine.
The fractional distribution of specific surface in a
thermomechanical pulp, i.e. the percentage by weight of
fractions having a given specific surface, is shown in
Figure 2. ~oth Pulp I and Pulp II are samples collected
from line 12 in Figure 1, but they differ as to fractional
-~ distribution of specific surface, Pulp I containing close
to 60~ of fibers of a specific surface of 2.5 or less, while
Pulp II has about 40% of such low specific surface fibers.
In fact, Pulp I has a much higher linting propensity. The
linting-propensity is measurable by means of a simple appa-
ratus, similar in principle to a printing press, in which the
lint picked out from the sheet of paper is collected under
controlled conditions and subsequently weighed or otherwise
measured to establish the relative proportion of fibers so
picked out from the sheet. This procedure permits the com-
parison of the degree of linting of various papers by means
of a linting index, the latter being expressed in terms
in weight or number of fibres (lint) picked out from the
sheet per unit area of the paper. The direct relationship
existing between the weight of a low specific surface
fraction in the pulp, e.g. less than 2.5 m2/g, (which we
shall call ~the linting propensity index~) and the linting
1045866
index of the paper, is shown in Figure 3. This is a plot of
the linting propensity index (weight fraction of fibres with
specific surface of equal to a less than 2.5 m2/g) against
the paper linting index (number of linting fibres per unit
area of paper surface). Curve B provides such a plot for
one particular paper machine and curve A for a second
machine. The values on the vertical axis represent the
fraction (as measured by the method described hereinbelow)
of the fibers in various thermomechanical pulps having an
average specific surface of 2.5 m2/g or less, while the
values on the horizontal axis represent the paper linting
index. It can be seen that, while for different paper
machines the same relative content of low specific surface
fibers may result in different values of the paper linting
index, for any given paper machine the paper linting index
is in a direct relationship to the content of such low
specific surface fibers.
Figure 4 shows the specific surface distribution of,
respectively a pulp directly from a refiner, a pulp treated
according to the present invention and one of the prior
art. Curve A indicates the specific surface distribution
for pulp as discharged from the refiner (before screening
and cleaning, line 12 of figure 1), curve B indicates the
pulp to the paper machine when treated according to the
present invention, and curve C is the pulp to the paper ma-
chine when treated according to a conventional process.
The data for these curves is provide in table 1.
.. .. .
- . . . ': . .
- 12 -
1~4S866
TABLE 1
AVERAGE SPECIFIC SURFACE m /g
PULP SAMPLEConventional ProcessPresent Invention
3% hydrocyclone 14% hydrocyclone
underflow and underflow and
10% Screen rejects10% Screen rejects
Refiner
Discharge 4.8 4.8
(before screening
and cleaning)
(line 12, figure 1)
Screen Rejects
(line 21, figure 1) 1.6 1.6
Underflow fraction
(line 22, figure 1) 0.9 1.2
Total Rejects - Before
Refining 1.5 1.3
(Press 23,
figure 1)
Rejects - After 6.7 6.2
Refining (line 24,
figure 1)
Centricleaner Accepts 5.2 5.8
(line 17, figure 1)
Pulp to Papermachine 5.4 5.9
Linting Propensity
; Index of pulp to the
machine (weight
. fraction of spec~fic 0.41 0.28
surface of 2.5 m /g
or less)
Paper Linting Index 62 47
(Curve A figure 3)
It will be seen that in this example the original pulp
had close to 50g of fibres with a specific surface of 2.5
or less and that by separating only 14% of the pulp in the
cyclone and mechanically processing the percentage of
fibres with such a low specific surface is reduced about 28%
in the pulp to the machine. This reduced the linting index
of the paper significantly (from 62 to 47) Curve C represents -
the specific distribution of a pulp obtained from pulp A -
by separating from said pulp in a cyclone an underflow
' . , ' . . .
- 13 -
1~)45866
fraction of about 3% and processing this fraction in the
refiner. It is apparent that no substantial reduction in
the quality of fibre with a specific surface below 2.5
(linting candidate material) is obtained.
- 5 No method for readily obtainin- specific surface dis-
tribution of a pulp is available and therefore it was neces-
sary to develop a method of fractionating pulp into frac-
tions each of which has a relatively narrow range of specific
surface. Hydrocyclones have been found to provide an appro-
priate means for separating fractions of pulp of relatively
narrow specific surface distribution. Conventional methods
of measuring specific surface are relatively complicated.
There is however, a correlation between the measurements of
specific surface and freeness for a wide range of these
two properties (correlation is not 100% accurate and there
may be significant deviation if the pulp properties vary
widely). Thus while the basic criteria for characterizing
the linting propensity is fractional distribution according
to specific surfa~e, in practice, the freeness value may also
be used instead of specific surface because of the inverse
- relationship of specific surface to freeness.
Several arrangements of hydrocyclones may be used to
fractionate pulps by specific surface, for example, parallel
arrangements as illustrated in figure 5A, underflow cascade
arrangement as illustrated in figure 5B, and an overflow cas-
cade arrangement as illustrated in figure 5C.
In each of these arrangements the cyclones will be de-
signed to reject different fractions, for example in a pa-
rallel arrangement as illustrated in figure 5A for a 2 inch
diameter cyclone the apex outlets cyclone may be 10/32 in.,
7/32 in. and 5/32 in. which will provide underflow fractions
of about 60% to 40% and 12%.
:.
- 14 -
104S86~
The cascading arrangements shown in figures SB and SC
wherein the underflow or overflow fraction of the first
cyclone becomes the feed to the second, similarly the flow
from the second becomes the feed to the third and so on step
by step down the line provide a narrow fractionation on the
basis of specific surface.
In order to obtain the reading of the specific surface,
e.g. in figure 5A, the flow from the apex of each of the
cleaners is measured, consistency of the underflow is mea-
sured and the specific surface is measure. As indicatedabove, freeness values may measure instead of specific
surface, keeping in mind, however, the limitations on the
correlation between specific surface and freeness.
In figure 5B, the specific surface distr~bution is
determined by measuring the rate, consistency, and specific
surface (or, with the above reservations, freeness) of the
underflows from each of the cleaners, while the index of
the specific surface distribution with the arrangement shown
in figure SC is obtained by measuring the flow, consistency
and specific surface of the overflow of each of the cleaners.
Obviously, in each of the systems the flow, consistency and
specific surface of the whole pulp is measured before it is
fed to the cleaners.
It has been found that these arrangements provide
adequate indication of the specific surface distribution of
the given pulp.
In using the equipment of figure 5, the flow to the
cleaners should be at low consistency say about 0.15%.
It will be understood that various modifications may
be made without departing from the scope of the invention
as defined in the appended claims.
