14 Guests are online
Total Listings: 1046



Register an Account or

What would you like to do?

Sale/Auction: Sprout Bauer Fibre Classifier

Send an email to sales@papermillsurplus.com to inquire about this item.


 Andritz Sprout Bauer Model 203


Bauer-McNett Finer Classifier

Model 203

A tree is made of fibres that differ greatly
in length, width and coarseness. The differences
between fibres within a tree can be as
great as the differences between species. However
the traditional approach of preparing fibres
for papermaking is to pulp the wood and process
the fibres as a collective – refining the
whole pulp to shorten the average fibre length,
for example. This approach simplifies the process
design, but neglects the opportunity to exploit
the inherent benefits of the individual fibre
Fibre fractionation is a process that segregates
a blend of pulp fibres into different
streams based on some physical property of the
fibres, such as their length or flexibility. The
fibres of a particular type can then be directed to
the most appropriate process and product. A
pulp producer with two pulp machines, for example,
could fractionate the fibres and increase
the content of long fibres on one machine to
provide a high-value reinforcing pulp. Alternatively
a paper producer with a multi-layer
headbox could direct the shorter fibres to the
surface layers to improve sheet smoothness,
while placing the longer fibres in the core to
provide strength. Within the mill system, one
could concentrate long, stiff fibres in the feed to
the reject refiner to save energy and increase
capacity, while avoiding the degradation of
fibres that are already acceptable.
The opportunities for fibre fractionation
are clear, but the use of fractionation technology
in industry remains limited. The board industry
uses fractionation to optimize the
character of the different layers [1–3]. In recycled
fibre production, fibre fractionation and
subsequent processing have been shown to improve
strength and brightness, and to reduce energy
consumption [4]. However the absence of
equipment which can precisely, efficiently and
economically fractionate fibres has limited the
widespread practice of fractionation in mill operations.
The promise of fractionation has been
established, in part, through the use of laboratory
fractionation equipment to separate individual
pulp fractions and form paper with
optimal fibre blends. The Bauer-McNett
classifier (BMC) is the most widespread of the
laboratory devices, and it is commonly used to
assess the level of fractionation obtained with
pressure screens and other process equipment.
The BMC operates by the selective passage
of fibres through a screen mesh. It separates
fibres mainly on the basis of length,
although the fundamental principles of operation
are not well understood [5–12]. The practice
of using this laboratory screen to evaluate
pressure screens raises several important questions:
— Do a BMC and a pressure screen operate using
the same principle of separation?
— To what extent does a BMC provide ideal,
precise, length-based fractionation?
— Does a BMC provide insights on how pressure
screens can be optimized for optimal
To answer these questions, this study reviews
how fractionation is assessed, and how
indices of fractionation are applied to industrial
pressure screens. A flow model of a BMC is
then developed, and experimental measurements
are used to assess the validity of the
model and the level of fractionation obtained.
Finally, the extent of fractionation with a BMC
is compared to published values for industrial
pressure screens.
The passage of fibres through screens is
described by their passage ratio (P), which is
the concentration of a particular class of fibres
in the flow passing through a screen plate aperture
relative to the concentration approaching
the aperture [13]. For the fractionation of fibres
on the basis of length, one wishes to have the
short fibres passing freely through the screen (P
equal to 1) while no long fibres pass through (P
equal to 0). It is appropriate to assess fractionation
by measuring passage ratio as a function
R.W. Gooding* and J.A. Olson**
Pointe Claire, QC, Canada
H9R 3J9
* Now with:
CAE Forestry Systems
4635 Patricia Ave.
Montreal, QC, Canada
H4B 1Z2
** Now with:
Dept. Mechanical Engin.
Univ. British Columbia
2324 Main Mall
Vancouver, BC, Canada
V6T 1Z4
An analytic model of the Bauer-McNett classifier (BMC) was derived and tested.The study showed that the BMC classifies fibres mainly by length.
Overlap in the length distribution of classes reflects the statistical nature of the BMC operation rather than the influence of some other fibre quality.
The fundamental quality of fibre length fractionation was similar to that found in an industrial pressure screen.
Un modèle analytique du classificateur Bauer-McNett (BMC) a été dérivé et mis à l’essai.L ’étude a démontré que le BMC classe les fibres surtout
selon leur longueur.Le chevauchement dans la répartition des longueurs des classes reflète la nature statistique du fonctionnement duBMCplutôt que
l’influence d’une autre qualité de la fibre.La qualité fondamentale du fractionnement de la longueur des fibres était similaire à celle que l’on retrouve
dans un classeur sous pression industriel.
of length (P(l)).
The passage of fibres through screens,
and P(l) in particular, has been studied theoretically
and experimentally [14–16]. A characteristic
curve has been proposed by Olson to
describe this relationship [17]*:



where  and  are dimensionless constants with
specific significance, as shown in Fig. 1. The
value of  can be seen as a “shape constant” inasmuch
as Olson has reported a value of  equal
to 0.5 for slotted pressure screens and a value of
1.0 for holed screens. Higher values of  indicate
that the screen approaches the ideal form of
a step curve. The value of  is a “size constant”
and experimental tests have shown that larger
holes and slots lead to larger values of  [17,18].
A key teaching of Fig. 1 is that, for a
* While reference [17] is the first public appearance
of this equation, it was derived independently
by Robert Gooding and
published in an internal Paprican report in
March 1992.
0 1 2 3 4 5 6 7
Fibre length (mm)
Passage ratio
 = 0.5
0 1 2 3 4 5 6 7
Fibre length (mm)
 = 1.0
4 =10
 = 5.0
Fig. 1. Characteristic forms of the P(l) curve for a range of  and  values.
A value of  = 0.5 (A) is typical for screens with slots;  = 1.0is
typical for screens with holes (B) and high values of  (C) denote nearideal
Fig. 2. Photograph of a Bauer-McNett classifier.
Fig. 3.Schematic ofBMCchamber in plan view.The inlet flowis shown
entering from the top. There is a circulatory flow which is driven
(clockwise) by the rotor at the right. Part of the circulatory flow splits
off and passes through the screen mesh (dashed line) at the bottom.
Sample size of pulp 10 g (o.d.)
Active volume of a chamber1 9.8 L
Standard flow-through rate 12 L/min
Upstream velocity2 ~ 2 m/s
Total area of screen mesh 0.033 m2
Open area of screen mesh 14 mesh 0.014 m2
28 mesh 0.012 m2
Aperture velocity (average) 0.02 m/s
Aperture size 14 mesh 1.2 x 1.2 mm
28 mesh 0.6 x 0.6 mm
Wire diameter 14 mesh 0.65 mm
28 mesh 0.39 mm
1. This includes the additional volume resulting from the head driving
the flow to the following tank.
2. Estimated as the tip speed of the impeller.
fixed value of the shape constant, , smaller
values of the size constant, , will reduce not
only the passage of long fibres, but short fibres
too. Thus purifying the stream of short fibres
through the screen plate is done at the expense
of contaminating the stream of long fibres with
short material. Increasing the value of  however,
leads to increased passage of short fibre
and reduced passage of long fibre. Thus  is the
essential parameter in assessing the quality of
To apply P(l) to the performance of industrial
pressure screens, one needs a model of
how the flow through individual apertures relates
to the overall flows through a pressure
screen. A pulp screen takes a single feed stream
and divides it into accept and reject streams.
The quality of fractionation is measured by
considering the extent that the long and short
fibres go to their target streams, i.e. that long
fibres go to the reject outlet and the short fibres
go to the accept outlet. To simplify the analysis,
one can measure the fraction of short fibres
leaving via the reject stream instead of the fraction
of short fibres going to the accept stream
since the accept and reject streams are complementary.
A removal function e(l) has been defined
as the mass fraction of fibres of length l that
pass from the feed to reject outlet [19]. By assuming
plug flow axially through the screen,
one finds that:
e(l) RV( )
P l
where RV is the volumetric reject ratio, equal to
the volumetric reject flow rate divided by feed
flow rate. The plug flow assumption is supported
by several experimental studies
The removal function, e(l), is used to
provide a detailed understanding of fibre fractionation
in screening. It can also be related to
the parameters commonly used to assess fractionation
in industrial terms. These parameters
are the “long fibre removal ratio” (also known
as the “long fibre removal efficiency”) and the
“short fibre removal ratio” (which is related to
the mass reject rate). The long fibre removal efficiency,
EL, is defined as the mass fraction of
long fibres (i.e. fibres greater than the specified
“cut-off” length) that go from the feed to the reject
outlet. The value of EL can be obtained by
using a lab device such as the BMC to measure
the mass fraction of long fibres in feed and reject
samples, or by averaging e(l) over the range
of fibre lengths above the “cut-off” value. Similarly,
the short fibre removal ratio, Es, is obtained
by averaging e(l) over the short fibre
In ideal fractionation, EL = 100% and ES
= RV. Following from Eq. (2), ideal fractionation
implies PL = 0 and PS = 1, which is consistent
with the near-ideal step-curve relationship
indicated in Fig. 1C. While it would be desirable
to have ES = 0, the limiting condition in
screening has very small particles following the
flow split, which leads to ES = RV.
Fractionation parameters have been proposed
to combine the impact of EL and ES. One
parameter,  has been defined as the difference
between EL and ES [18]:
  EL  Es (3)
The value of  is equal to 0 when there is no
fractionation, and  approaches 1 as fractionation
increases. This parameter is a “value characteristic”
for the specific case where the
economic benefit of increased EL is equal to the
penalty of increased ES. It is of particular use for
control purposes.
Another fractionation parameter,  is
defined in terms of the passage ratios of the
long and short fibres:
The value of  is a “performance characteristic”
for a particular screening configuration, and its
use complements that of . Assuming a plug
flow model of flow through a pulp screen, EL
and ES are related as follows:
EL  E1S (5)
The value of  increases from a value of 0,
where PL = PS, EL = ES and there is no fractionation,
to a value of 1 for ideal fractionation,
where PL = 0 and EL = 1. An important attribute
of  is that it is independent of the reject ratio,
which is the chief operating variable of a screen.
Thus  is a fundamental reflection of the pulp
character and screen plate configuration.
Regardless of whether one monitors  or
 the key point is that the values of  and  determine
PL and PS, which have a direct link to
EL and ES and the performance of industrial
pressure screens. Since the BMC has been used
both to assess fibre length (for measurements of
EL) and as a model of fractionation, there is a
keen interest in assessing the values of  and 
obtained with this device.
The BMC is a laboratory device that is
used to assess the size distribution of fibres.
The device is shown in Fig. 2 and standard
methods exist for its use [21,22]. It consists of a
series of chambers, typically four or five, each
with a progressively smaller screen mesh. At
the start of a test, a prescribed amount of pulp is
poured into the first chamber. A continuous
flow of water passes through this chamber and
the series of chambers that follow. Smaller
fibres are carried by the flow through the mesh
and to the following chamber or, in the case of
the final chamber, to the sewer.At the end of the
test period the flowis stopped, the chambers are
drained, and the mass in each chamber is measured.
Thus a pulp with smaller fibres will have
more mass in the chambers with the smaller
mesh sizes.
For the sake of simplicity, the analysis
and experiments in this study only consider the
performance of a single chamber in the BMC,
though they could be extended to the performance
of the overall unit. The elements of a single
chamber are shown schematically in Fig. 3
and key parameters are listed in Table I. An industrial
pressure screen and the BMC both
cause undersize fibres to pass through a screen
plate. The apertures in each device are roughly
the same size. However, there are some important
differences between these two devices. The
BMC has an upstream velocity of about 2 m/s
and aperture velocity of 0.02 m/s, while a pressure
screen has an upstream velocity of about 5
m/s and aperture velocity of 2 to 4m/s. The
maximum consistency in the BMC is 0.1% versus
a typical feed consistency of 1.5–3% in a
pressure screen. The BMC has a screen mesh
while a pressure screen uses a solid metal plate
with holes or slots. Also, the BMC has an impeller
which moves the flow past the mesh surface
and causes mixing in the chamber but,
unlike a pressure screen rotor, does not produce
pressure pulsations which backflush the mesh
apertures. Thus the BMC aperture flow is relatively
steady, while the pressure screen has substantial
flow reversals.
In a typical experiment, the residual
oven-dry mass was measured using a hot-plate
technique, and the fibre length distribution was
assessed using a Kajaani FS-200 Fibre Length
Analyzer, which measures fibre length with a
precision of 0.05 mm. The length-weighted fibre
length distribution was then used to estimate
the total mass of fibres in various length
classes. To determine the passage ratio of fibres
in each length class, a mixed flow model was
applied to the chamber. Thus the amount of
fibres passing through the mesh (dM) in a period
of time (dt) was assumed to be proportional
to the volumetric flow rate (Q), the
passage ratio (P) and the instantaneous concentration
of fibres in the chamber (C):
dM  QPCdt (6)
Fibre concentration is the mass (M) divided by
the volume of the chamber (V). Substituting this
definition into Eq. (6) and integrating from the
initial mass installed in the chamber (M0) at
time equal to zero yields:


e (7)
Equation (7) may be rewritten in terms of the
number of replacements (R) that pass through
the chamber:
 e (8)
where “replacements” represent the total flow
through the chamber divided by the volume of
the chamber, and are assumed in this analysis to
be the essential factor in depleting mass in the
One can also re-arrange Eq. (8) to provide an
expression for P:


ln (10)
By measuring the initial and final mass of fibres
in each size (i.e. length) fraction, Eq. (10) can be
used to generate P(l) from experimental
Seven experiments were conducted in
this study and the experimental program is
given in Table II. A reslushed, softwood TMP
from the British Columbia interior was used for
all experiments. The pulp had a freeness of 200
CSF and was subjected to standard procedures
for disintegration and latency removal.
Degree of Segregation
The first test was made to establish the
quality of length separation achieved with the
BMC. As seen in Fig. 4, the BMC sorts fibres
on the basis of length but does not provide a
precise and complete segregation of the fibres.
Chambers with larger mesh sizes retain longer
fibres, as expected. However the range of fibre
lengths in a particular class is substantial and
there is significant overlap between different
classes. For example, 1.5–2.5 mm fibres comprise
one of the largest classes of fibre length
within the R28 classification, but significant
amounts of these fibres are also found in the
R14and R48 classes. These observations are
consistent with those by Jackson [23].
Figure 5 presents the combined results
of Trials 1 to 3 for the R14class. One sees that
certain fibres, such as those in the 3.5–4.5 mm
class, are almost completely retained, irrespective
of the time spent in the chamber. Others,
such as the 1 mm fibres, have almost completely
passed out of the chamber in the 20 min
recommended operating time. Fibres in the
very important 2–3 mm class appear to be diminishing
in a gradual way suggestive of a random
or statistical passage.
Statistical Passage
A statistical model of passage was assumed
in the model of the BMC. This led to Eq.
(7), which has an exponential form. Figures 5 to
7 show data points that reflect the decay in the
mass of fibres from particular classes as a function
of time. Also shown in Fig. 6 are curves fit
to the data using the exponential form of Eq.
(7). The good fit provides support for the statistical
model of the BMC.
The statistical model presented above
prescribes that the diminution of mass is determined
by the number of replacements that pass
through the chamber (Eq. 9). This was investigated
by also changing the flowrate through the
BMC in Tests 4and 5. Figure 7 includes the
data from the changing flow rate as well as the
data in Fig. 6. The residual mass fraction is
Test No. Mesh Flow Rate Duration
12 L/min
20 min
1. The standard set of BMC screens are: 14, 28, 48, 100 and 200 mesh.
Fibre Length (mm)
Mass density (g/mm)
0 1 2 3 4 5 6
Fig. 4. Fibre length distribution (mass density) of pulp from individual
BMC classes.
Fibre Length (mm)
Mass density (g/mm) 0
5 min
10min (replicate)
Fig. 5. The mass density–fibre length distribution in the R14 chamber
as a function of time. Note the excellent reproducibility of results for
the replicate trials of 10 min duration.
Time (min)
Mass fraction
0.00 5 10 15 20
0 - 1 mm
1 - 2 mm
2 - 3 mm
3 - 4 mm
Fig. 6. Exponential decay of mass in the R14 container for different fibre
length ranges. Mass is expressed as a fraction of the initial mass
for each fibre length. Replicate measurements (at 10min) appear almost
coincident. The fit curves use the form of Eq. (7).
Mass fraction
0 5 10 15 20 25
0 - 1 mm
1 - 2 mm
2 - 3 mm
3 - 4 mm
Fig. 7. Exponential decay of fibre mass in the R14 class as a function
of replacement volumes.Fit curves are of the form of Eq. (8).Data were
obtained by varying both the duration of the trial and the flow rate.
plotted as a function of replacements rather
than time. The additional data fit the exponential
form of the fit curves well.
P(l) Curves
The data in Figs. 5 to 7 support the use of
a statistical/perfect-mixing model for the flow
of fibres from the BMC. It follows that Eq. (10)
can be used to provide the P(l) curve which
characterizes the quality of fractionation obtained
in a BMC. These data are shown in Fig.
8, which includes a curve of the form of Eq. (1).
The excellent match between the trends in the
data and the curve form supports the application
of Eq. (1) (proposed for industrial pressure
screens) to the BMC and suggests that it may
have a general applicability to pulp screening
The value of  obtained by a leastsquares
fit was 0.97, which is very close to the
value of 1.0 proposed by Olson et al. [17] for
pressure screens with holes. It is not overly surprising
that a square hole (in the BMC mesh)
had the same shape constant ( as a round hole
in the pressure screen. More curious is that the
absence of pressure pulsations did not have an
effect on  However, qualitative inspection of
the flow patterns within the BMC suggests substantial
mixing in the flowadjacent to the mesh,
which might be producing the same effect as
the pulses. A value of  equal to 0.82 was obtained
for the R14mesh (with 1.2 mm square
openings) which, compared to the values reported
by Olson et al. [17] for pressure screens,
suggests that the apertures in the BMC act like
smaller holes than in a pressure screen.
P(l) data for the R28 BMC class are
shown in Fig. 9 along with the curve fit using
Eq. (1). Note: Unlike standard BMC operation,
the initial pulp charge was placed directly in the
R28 chamber at the start of this test and the R14
chamber was not used. The excellent match of
the data to the curve form gives further support
to the use of Eq. (1) for characterizing fractionation.
Moreover the good fit made using   
indicates the fundamental quality of fractionation
did not change with the size of the aperture.
The lower value of  is consistent with
previous findings for pressure screens that indicate
lower values of  for smaller screen apertures.
The conclusions of this study can be
stated in the context of questions posed at the
start of this report.
Do a BMC and a pressure screen operate
using the same principle of operation?
Despite the significant differences in the
overall configuration and operation of a BMC
and pressure screen, the essential mechanism
that governs fibre passage at an aperture appears
quite similar. This study demonstrated
that the mixed-flow model works well to describe
the performance of the BMC, and the exponential
model of fractionation fits the P(l)
fractionation curve very well. Like an industrial
pressure screen, the BMC operates by “probability
screening”. This raises questions about
the use of one probability screen (i.e. the BMC)
to evaluate another, such as an industrial pressure
screen. However, it establishes that the
BMC may be used to better understand the fundamentals
of screen operation.
To what extent does a BMC provide ideal,
precise, length-based fractionation?
The high degree of dilution and low reject
ratio in the BMC provides good segregation
of certain sized fibres, but there are
substantial ranges of fibres that are only partly
segregated in the BMC. This mitigates against
the use of theBMCas an absolute measurement
of fibre lengths. The degree of fractionation, i.e.
the values for the shape constant () and size
constant (), is comparable to that found in an
industrial pressure screen.
Does a BMC provide insights on how
pressure screens can be optimized for
optimal fractionation?
The finding that  for the square mesh
apertures in the BMC is the same as that for a
pressure screen with holes reinforces the
thought that  represents a shape constant for
screen operation. Since fractionation in the
BMC and in pressure screens operates by probability
screening, the shape factor may represent
the degree of restraint that the aperture
imposes on the somewhat random way that
fibres approach an aperture. A hole imposes a
higher degree of constraint than a slot because,
for a slot, long fibres have the potential of being
aligned with the slot length and passing
through. Ideal fractionation (i.e. high values of
 would appear to require a mechanism for applying
further degrees of constraint. Increasing
the degree of constraint in the BMC or some
other lab-scale fractionation technique, which
can be modified relatively easily and operated
under controlled conditions, may be a useful
step towards the larger goal of increasing fractionation
on an industrial scale.
This study has provided a cautionary
message about using the BMC as an absolute
measure of fibre measurement. At the same
time, it has revealed an opportunity to use the
BMCas a device to explore the fundamentals of
1. COX, M.T., “Application of Hot Stock Pulp
Fractionation at Mead Coated Board”, Proc.
1989 TAPPI Pulping Conf., 1–4.
2. KOHRS, M., “The Applications for Fractionation
Technology”, Paper Technol. 33(3):10–12
3. MUSSELMANN, W. and MENGES, W., “Concept
and Function of a Wastepaper Fibre Fractionation
System and Practical Operating
Experience”, Woch.fur Papierfab. 11/12:
368–379 (1982).
4. MAYOVSKY, J., “Fractionation of OCC. How
Can It Help You?”, Recycling Symp., 407–416
5. ANDERSSON, O., “An Investigation of the
Hillbom and Bauer McNett Fibre Classifiers”,
Svensk Papperstidn. 56(18):704–709 (1953).
J.S., “A Comparison Between Fibre Length
Measurement Methods”, Pulp Paper Can.
95(4):41–45 (1994).

Video Not Available

©2007 Paper Mill Surplus