AIRICUL1URE
ROOM
SCREEN ANALYSIS AS AN AID I N
PULP EVALUATIO N
Revised November 193 9
M
SCHOOL OF FORESTR Y
OREGON STATE COLLEG E
CORVALLIS, OREGO N
UNITED STATES DEPARTMENT OF AGRICULTUR E
FOREST SERVIC E
FOREST PRODUCTS LABORATOR Y
Madison, Wisconsi n
In Cooperation with the University of Wisconsin
SCREEN AlgLYSI .S AS AN .AID IN PULP EVALUATIO N
A, resume of work at the :Forest Product s - Laborator y
By
E . R . SCHAFER ,
Enginee r
This article supercedes a publication of the same title, (1) ; :-.and ..
brings together the results of several'indepe'identinvestigations, some o f
which have not been previously reported . The ' Laboratory first became iinterested in the fractionation .offiber masses into length classes as, a
means of pulp , evaluation in a study of insulating board from sawmil l
waste . The fractionation was effected with an experimental flat-plat e
diaphragm screen provided with a series of interchangeable slotted plate s
of the type ordinarily used in pulp mill screening operations . Th e
apparatus was later modified for application to evaluating groundwoo d
ahd chemical pulps by substituting . wire screens for slotted plates .
Further study brought improved ef`iciency• through the development o f
maltiple"plate screeriin which the stock flowed through four compartment s
each fitted`with a:screen of different site openings . Experiments will be described of the screen analysis of insulatin
gboardpuls,groundw puls,andchemialpuswithbo esingla d
multiple diaphragm plate sereea-as well as a limited amount of work with . :.
the Bauer-McNett Classifier .
SCREEN ANALYSIS . OF INSULATING BOARD STOCK S
Single Plate Scree n
The plates used in the single-plate diaphragm screen (see fig . 1 )
were 10 inches wide .by l2 inches long .and contained 6o slots, eack 4 inches long . The widths of the slots in the series of plates . were 0 .020 ,
0 .015, 0 .012, and o .OC g inch, respectively . The material was firs t
screened on the screen having the smallest slots . The fraction of th e
material passing through the screen was caught in a drain box with a
"0
Acknowledgement is made to the Appleton Machine Company, Appleton, Wis . ,
for cooperation in working out detail s , of, desig n, ,of ; this screen an d
for building models in the various stages of development . . The machin e
is commercially known as the Appleton Selective Screen ,
R gg4 .
Fourdrinier wire bottom .. Thescreen ,plate was then replaced by the nex t
larger in the series and the screening rep'a tads _ ..':T e
ter ,lduring eac h
screening was vigoroi .sly .,agitated on the screen with water sprays . On e
hundred grams of the original
'b er' ' required about- lO r minirtes screening o n
each plate to effect a practical separation . Continued screening cause d
the various sized fractions to blend into each other . The various fractions obtained in this way were then dried and weighed .
fi
The screen analysis and other 'propertie's-of three well-known commercial insulating boards and one of an 'experimental Douglas fir board ar e
given in table 1 . These boards are composed of various materials that hav e
been processed in different ways . It is, therefore, not surprising t o
find no general relationship between the strength properties and th e
screen analysis . Nevertheless, a change in the fibrous composition o f
any one of them would undoubtedly influence not only the strengt h
properties but its density, formation, insulating value, and appearance .
Having developed a, product of, satisfactory quality, the screen analysis ,
therefore, presents, a means of Oontrol'in manufacture . Plate 1 shows th e
fiber aggregates obtained from insulating boards .
The Bauer =McNett -Classifie r
The .Bauer-McNett Classifier is a ` four compartment apparatus in whic h
the stock flews . by gravity through a'-mes h ' wire screen from one compartmen t
to the next . The screen plate"is set vertical in the . compartment and a
propeller agitates the stock suspension in a=manner claimed•to orient th e
fibers parallel to the screen surface . Ten grams of material are used i n
each test . The fiber .distribution of seven commercial insulating boar d
stocks as obtained :on this classifier are given in table 2 . .
SCREEN ANALYSIS OF GROL'NDWOOD PULP
Since groundwood pulp will practically all pass through a plat e
with the usual 0 .00 g -inch width slots, the separation of this type o f
pulp must be made with screens having openings smaller than those obtain able in the form of slots . For this purpose a series of screens consisting .of various mesh wires of the Tyler standard screen scale wer e
soldered to perforated brass plates . The screens were 24, 32,-42, and 60
mesh of . the Tyler series which, respectively, have screen openings o f
0 .701, 0.495, 0 .351, and 0.246 mm. in width . The ratio of the width o f
screen opening in one screen to that 'of the next one i n . -this series. i s
approximately equal to the square root of 2 .The procedure for the single-plate screen consisted in using thes e
plates in the apparatus in the order of their size, starting with th e
smallest, The fraction of the material passing through the screen wa s
caught in a catch screen the bottom of which was made up of a piece o f
R gg4
-2-
r•
a)
0
O
P4
•
1
-fI
a)
rCC
tti
a)
o
0
rd
a)
O
b
+)
N
4-1
O
41
•ri
.F.,
4.C
0J
'. l
a)
i
I
0
1
U)
a)
U)
I
0
$-t
i
0I
I
rd
1
I
o
I
1
I
U1
A
+'
••
4A
0
fi
I
•
I
i
I
H
E-1
..
..
w
. .
1
r-e I~
I
I
1
I
I
I
I
F4
O
o
I
I
{
0
a)
(••
•
I
f
I
1
I
I
I
O
I P-1
r~
m
I ~.
q I O
..
a)
I
y 1 rd
a) 1 Cd 1
Q)
q
I
ri
I
1)31
0
I
I
1
f
m
I
o
I
1
O
O .
d .
N N
LC\ En
LO
NN
N
. .
. .
O to
00
O O
. .
LO IX1
0
O
p0
l0 ~D
0) CS)
v
v
00
O O
N rI
.. ..
.. .•
MM
.. ..
LC \
..
0
O
H r-1
O
0
0
MM
.. ..
N
N
-
r-1
H
r-•I
. .
r-1
.. ..
r--1
111
LC1
.~'
..
' ..0
..
I •-
.r
CV A 4-,
riO O
•d I r-i
0 r-1
I LCl
▪
..
M
..
0 '•
Ea
t
to _0 +'
IN OU O
I -i O 0 r-1
• .r-I U#
Cj
rn
4-4
0)
..
r--N N
I
I
0
to
0
~I
a)
I
I
r-H
..
Lc)
to
M
~L.0
I
..
i-3
.•
.•
-
.. ..
1
I
0 .'i
L
Pal
a) •rl
U7
d
>~
u) Fa • O
a)
a)
..
MN
to LO
Fr
I
. . ..
• 'a) tc)
42
•r-1 0 O A +'
0
CG
NMM
!,
. .
ri
0
NN
La
I
I
l
U) +~
~~
f
Gd •rl
I r-♦ c)
1 W •r+
1
+'
.1. . .
.. . .
I
a)
R
a)
a) F-t
F4
4
0
s
i U)
♦• . .
I
1
1
I
f
C)
N-
H
0
ty
to
f`
0
I U
C)t
. . ..
f-'ii
a)
U
Fi
a)
Pi
I
r-I
N
.
..
lC)
•
0
r~
r-I
H
Lc)
..
..
..
N-
N
0
to
rI
N
Ql
01
r-
..
1P-
..
L.D
0
r4
I
.0 -P I N
O LCl U O
IIN ri
r-1
1 0 0 • r .1 m
1
f io
1 Q)
I
H
H
0
•O
t O O
No
cd o
a'
0
P,
r-f
p
0
•r~1
O
-1-
o
1
a) a)
P-
f-
M -'
M
•
+)
4•
• -1
m LAO
.
0 us
a)
0
O
O
a)
+3
O
q
a) 0 -P
En
I
f-i
-a)
4-1
4-3
O
•
0
•-1
Pi
Ea
4-4
+) H
q a)
ct
0
cd 4-1
a)
q ra
rd
+
0
5
C~)
0cd
o •-4
R1
U
O
rd
a)
c'.
R
c
0
•H
a)
Retained on 0 .020 inc h
width slots
Passing through 0 .020 inc h
width slots and retaine d
on 0 .015 inch width slot s
Passing through 0 .015 inch
width slots and retained
on 0 .012 inch width slots
Passing through 0 .012 inc h
width slots and retaine d
on 0 .008 inch width slot s
Passing through 0 .008 inc h
width slot s
PLATE 1
Fibers obtained in the fractionation of a pulp used for insulating board. Reduced to 0 .9 natural size .
zM Iz984F
•
q1
. M
• . r-l
4'
O
U
LC\ 0 I t-i
r-1 H r a)
O•
CT to
O
I
r
ri
re
ri
\
H r-i
to LO rO 0 0 LO L0 LO
t0 t0 t0
r-4
r-I
r1
0 Lh r- al 0
to N- N- LO O
r-i
r-1 al LCl r
LO LO LO r-i
rl .-I-
to
r-I
DO
0 0
r-I
r~i
O ~
99
99 99 99
▪
I +'
zn
a)
to
LC1 0 I 0F-i
LO N a
O•
O~
99 99
N- LCI/O
N NN
N- t0
to to to
to to to
r-i i-f r-I
.
H H H r-i ri r-I
99
99
ti
.
r \r l r \
e
r
e
.9 99
99
N\ t0 0 150 0 al
,
Lf ' LCl O re\ H
N N N
N N N
99 99 99
99 99
•
r-1 al 0.
LO 1` NH HH
9 9
99
99
t
•
N 01 01
LC' -i I
r1
r 1
a)
NCV co
(i
O
r
r
LO N
LC \ LC\ LC \
NNN
9
ti N al
N r \N
NN N
NN N
N N N
99 99 9 9
9
.
e
1
•9
•
CR
I
a) ;
r
Lc\
0. . '.-[
0N
r5l
U
▪ :
▪
)
4
99
• • •
:
•
U)
! U
F!
~N
N
..
•
N-0\
H N
r \ r \r \
e
e
e
..
0
..
•
..
rl O LC\
Lr - -d-
re\ r \ r \
e
..
..
99
..
1- LO LC\
LC \ LCl Lc\
N NN
•6
99
..
N
r \
ref r \ re\
N N N
e
e
. .
. .
M
tO
O
I~-LD t-HH H
. .
99 99 9 9
LC1 N to
Lcllp LC\
H H r-1
.
• •
▪
0 al al
re\ r \ re\
e
9
9
9
•
9
..
.
I
! +)1
0
' OI U
C
H r I ` NFi
•
LC) I
.6
e
I
1
90
•
r- M 0
ti to to
'sr,
•
q
fi
i
U)i
e
99 99
Pi
.S'
r \
O re\
N to
▪
90 99 99
-FD
I O
to al
N
•9
.
.
.
:
•
:
▪
•
:
:
:
••
.
Ol 0 al
N Lc\ re\
re\ M re\
LCI
al
.
Lf 1 N
IYl
N LO .zlN
to 0
Lll_71 111
•
Table 3 .--Fiber size distribution of the various screen fractions o f
groundwood pul p
Fiber length
range
From
mm.
0
0 .073
.147
.220
.294
.36s
.44o
:
: mnn.
:
.
:
:
.515 :
.588
.6o2
.736
.810
.884
.956
1 .030
To
:
:
r
:
: Passing 60
: mesh screen
Fibers
measured
.
:
:
:
.
1 .618 •
0 .073
.147
.220
.294
.368
: 489
: 139
: 78
: 15
.
8
.44o :
4
.515 :
7
.588 .
1
.662 .
1
.736
.810 :
.804 :•
.956 :•
1 .322
1 .398 :
: ■
:
:
:
:
:
:
:
:
1 .691 :
:
1 .691 : 1 .765
1 .765 . 1 .839 :
:
:
:• .•f . . . .
:
1 .618
Fibers
measured
: 65 .70 :
: 18 .70 :
: 10 .50 :
: 1 .92 .
: 1.07 :
:
.54 :
:
. 94 :
:
.15 :
:
.15 :
:
:
1 .470
1 .545 :
:
:
:
1 .839 : 1 .912 :•
1 .912 : 1 .986
1 .986 : 2 .060
2 .060 : 2 .115
2 .115
2 .205
2 .205 : 2 .382
2.382
2 .375
2 .375 : 2 .430
:
Total
Retained
:
between
:
and 60 mesh:32
screens
:
Retained
Retained
between
:
between
and 42 mesh; 24 and 32 mes h
screens
:
screens
?.ers
Fiber s
mei.;_17-. ed
measured
------------------ :Number : Percent : Number : Percent: Numbe :c• J. °cent : Number : Percen t
1.030 :•
: 1 .102 :•
1 .102 . 1 .175 :
1 .175 : 1.250 :
1 .250
1 .322
1 .398
1 .470
1 .545
:
:
: 142
:
:: 2
19
41
: 0.40
: 3 .76
:
.
37
• 7 .33
:10.70
: 9 .70
:1o.50
:14.20
: 5 .52
6 .52
: 4 .36
: 2.58
: 3 .76
: 3 .97
: 1 .38
.
: 2 .18
: .59
: .59
.59
: .20
: .59
:
:
:
:
.
:
:
:
:
:
:
54
49
53
72
28
35
22
13
19
15
7
9
11
3
3
3
1
3
o
1
3
2
1
:
:
8 .12
1 .78
:
:
0
.20
. 59
.
:
:
:
:
.
:
:
:
:
:
:
:
.
.4o
:
.20
:
3
:
8
:
9
8
t,
:
22
22
46
21
34
42
26
:
:
:
:
:
:
:
10
11
:
:
11
11
17
23
12
:
:
:
:
0 .82 :
2019 :
2147 :
2 9
3 .02
3 .52
6 . 0.2-41 :6 .02 :
9 .85 :
5 .75 :
9 .32 :
11 .50 .
7 .13 :
4 .65 :
2 .74 :
3 .02 :
6 .03 :
10 =? ,
8
3 ~;
.
5 $ . J- .-7 :
2
,
5 ,
3 3= : ~
2
.
1 :
.27 :
1 .
.27 :
. . . . . . . . . .
:
:
:
.
. . . . . . ... . . . . . .. . . . . . . .•
10
14
1
2
:
. . . . . . . . . . . . . . :
:
2
:
:
:
:
3
:
:
3
Above 0 .662 :Above 1 .986 irrn . :Above 2 .060 mm.:,Above
3 : 0 .45 ;
1 : 0 .20 :
4 : 1410
---------W
5
: 745 : 100 : 505 : 100
: 365 : 100 : 342
:
1 .1 7
.29
:
.5 9
:
.5 9
.88
:
:
.88
2 .430 mm .
:
1 .4 6
:
10 0
80 mesh wire . Ordinarily no attempt was made to catch the material passin g
through the 60-mesh screen . One hundred grams on the basis of weigh t
of oven-dry pulp was used for the test . After screening, all othe r
fractions were dried and weighed and the portion cassing through th e
60-mesh screen was obtained by subtraction . When it was desired to obtain a auantity of the fine fraction to determine its properties, a piec e
of linen toweling was laid over the wire screen in the catch box . B y
this method more than 75 percent of the fiber passing the 60-mesh scree n
was reclaimed .
The appearance of the different fractions obtained with this serie s
of screens is characteristic and indicates a fairly definite separatio n
of the fibers into groups according to their size (3) .
Table 3 shows the results of fiber measurements made to determin e
the fiber-size distribution in the various fractions . The data represen t
the average of five commercial groundwood pulps . The five pulps wer e
fractionated and the fiber measurements made on each fraction . The fiv e
sets of fiber measurements for a given fraction were then combined . Fro m
350 to 500 measurements were made on each of the fractions retained o n
the 24, 32, 42, and 60-mesh screens, and because of the inaccuracy involved in measuring very short fibers, about 750 measurements were mad e
in the material passing the 60-mesh screen .
The manner in which the fiber lengths tend to group themselve s
about the average, or the freauency distribution, was determined by th e
number of fibers falling within a given range of length instead of th e
__number of fibers of a given single length . This was due to the fac t
that even with the apparently large number of measurements taken, th e
relation of the percentage of fibers of a given length to that of th e
next higher or lower length was very irregular and therefore, a curv e
drawn through such data would not present a true picture of the norma l
distribution of the fiber lengths . A much larger number of measurement s
would have smoothed out these irregularities and permitted the use o f
smaller ranges than reported here . The average fiber length, the standar d
deviation, and the coefficient of variability of each fraction are show n
in table 4 . A high degree of variability is indicated in all fractions ,
that for the material passing the 60-mesh screen being almost twice a s
great as the other fractions .
Table L+ . ;--Average fiber length, standard deviation, and coefficient o f
variability of groundwood screen fraction s
Fraction
•
Passing 60-mesh
Retained between :
42 and 60-mesh
32 and 42-mesh
24 and 32-mesh 4
R884
: Average fiber : Standard
length
: deviation
:
mm .
0 .079 7
:
+0 . 0796
.567 5
.812 0
1 .11-1.
0
. 334
.365
.513
-3 -
: Coefficient o f
variability
- ..
.
Percen t
99 . 9
58 . 9
45 . 0
56 . 1
The abscissas of the pofints on the curves show :n in . figure '3 are:
plotted as the midpoints of the range of length chosen . In spite •o-th e
efforts to insure .a complete 'separation, of the fibers there is a ' considerable overlapping of sizes in every fraction . This does not, however ;
invalidate the usefulness of the method . In commercial work the series'o f
screens needed will depend on the characteristics of the pulp and th e
degree of separation desired . In some cases a sharp separation into fibe r
classes may not be necessary . If•a sharp separation is desired, a-stud y
such as described above, although requiring a large amount of careful work ,
will be needed to determine the number and mesh of the screens required .
Detailed studies of the physical and chemical properties of th e
screen fractions of .groundwood pulp have been presented in other •publica tions (4,) . These investigations have indicated some very definite relationships between , fiber size distribution and pulp properties . .
SCREEN ANALYSIS OF CHEMICAL PUL P
11 '
The screen analysis has been found useful in studies of the effec t
of beating . Then beating a mixture of spruce and birch sulfite pulps fo r
glassine stock, samples were taken from the beater at intervals an d
development of hydration and strength of-the pulp followed by means o f
the'freenessand strength tests . The cutting action of the beater ril l
was followei .b'y-the screen analysis, in this case with the single-plat e
screen . The results of the analysis are shown in figure :4 .
It may be noted that the cutting action of the roll, which was o f
stone Was rapid 4uring the first .120 minutes . Cutting was less rapid, during the . next80 minutes but increased again during the last period of 4 0
minutes : It may be noted further that an increase in the amount o f
material that would p ass the 60-mesh screen was accompanied with a n
approximately equal decreas .e in the amount retained on the 24-mes h
screen . The intermediate fractions remained practically constant i n
amount throughout . This example is cited to illustrate a use the scree n
analysis may serve in determining That is happ ening to pulp in processing .
Another example of the . aid of the screen analysis in the processing of pulp is shown by the data in table 5 . This was a study of th e
production of bond paper from bleached sulfite pulpi Th4 4creening wa s
done on the four-compartment screen .showh in figure 5- . The analysi s
indicates the trend of the processing effect on fiber distribution .
The percentage of fiber retained on the ' 24-mesh screen was redu.ced - proportionately as much by beating as by jordaning although the beatin g
treatment was very mild . . The effect of beating on increase of th e
amount of material passing 115-mesh screen was nearly twice as great a s
the effect of jordaning the beaten pulp . Generally speaking more tha n
two-thirds of the fiber size reduction ofthe "on 24-mesh" material b y
beating appeared as increase in the amount of "through 115-mesh" materia l
whereas only a little mbre than one-third of the size reduction of "24 mesh material ft by jordaning the beaten stock appeared as fine material .
8884
-1 -
N
■■MIlIEN111■ I
n1111N■■11 .11..1119
nnnn
f0
11111■1011 NINE
//E L
.11
r■
i
__
MIR
INN
u
u
u)
IN
nnn
u)
a)
u)
a)
-
et:1
z
-
\J N
iti N Fi
u)
a)
a)
a)
a)
4-3
4-1
-p
a)
a)
a)
-0
0
a) rU
a) u1
,o -0 0
00
acd
0
ifi
k ;-l
OOO
~■
9
g
N-N._1-
-d
N
Pi
o
.H
0 0 0
cd cd cd cd
r-=-f Fri
00 <1
I frIl I A
t)
%I%
.
IMO
E
LE
U
a)
r-1
glom . EEM■ ;
44
4i Fi
O U
U)
,
1
■
IR
sm..
nnn
z
l
_
I
r-~
}
--
■■.alBr 41
>F
~~ I t,
.. , _
INN
X11_-a_E
■ 11
1.111•1=11_n.
N
cH
O
q
•
O
-3
N
I Fvl
1
4-0
0
.H
u)
_D
,d
P
a) z3
N
a)
•H
va H
Fi
a)
a
a)
_
N
o
a)
a)
Fa
U
O
a)
a)
UO
a)
.U
O
0
rc1
.W
O
a)
a)
a)
U
o
U
O
w
~
N
tv l
N
d
d
U
0
~O
cd
f
N
N
N
rrl ~-
s~
m
a)
0
a)
+
-p
P4
f~
N
H
U)
11
cd
a)
EiD
a)
a)
a)
O
al
a)
a)
0
N
cd
Pi
IN=
O
O
N
1.11
0
w
0
uaaxos
txo
pautt;qa .T
4txnoult)
.C
-'
o
I
Ei D
cti
a)
cd a
,O
O
`rS
Cl]
F-i
a)
a) p
U
U)
1
L-
rr-i
a)
P
~
II
(4uao .zacl --
•H
o
Ill■=
!
NAM= •=Il
-.
L
O
a)
1
0
•
U
C
Table 5 .--Screen apalysis .of_unbeaten and processed bleached sulfite pul p
Processing:
treatment
,_ := Retained.
on
42 and : 80 and
80-mesh :115-mesh
24 and
U2=mesh
Pexrcent'
Percent
Unbeaten
'Retained between
: -=
Percent . Percent
:
c
75 .0
6 .)4.
After beating . . . . :
58 .4
10 .1
:
6 .5
3 .2 .
After jordaning . . :.
45 .0
14 .o
:
10.0
4 .0
.5 r
:
4+ .3
: Passin g
:,115-mes h
.
,percent
10 .4 ,
'319
22. 0
, 27 . 0
It . .was also observed in this investigation that the percentage o f
fine material in the furnish was associated with the porosity of the
finished paper . The data in table 6 show that a higher porosity valu e
(more dense sheet) was obtained with furnishes containing the highe r
proportions of fines .
Table 6 .--Influence of amount of fines in the furnish on the porosity
of sulfite bond pape r
: Porosity of shee t
by Gurley
:
:
densomete r
: After jordaning : At head box
:
-------------------------------------------------------------------- Number
Second s
Percent
Percent
Machine
run
:
1257
1272
1261
1267
1259
:
R884
:
Proportion of furnish
passing 115-mesh
:
1y .2
:
33 .2
35 .6
30 .0
9 .3
360
230
143
56
50
STUDIES ON THE SCREENING PROCEDURE '
In the ori-Omal screen analysis test employing' asingle-plat e
diaphragm screen it was necessary to interchange screen plates for eac h
fraction . This appaus was improved upon by building_ four diaphragm
screens into- olie . machine . In the improved machine - each compartment contains a,: $cregn plat e: ;with different size- of . openings than the other three .
The screeni - shown -in figure 5 . I-n operation, from 50 to 100 grams o f
the pulp (depeh$ .ng on the-,fiber size distribution Of the sample), in a
dilute suspension., flows by gravity from one compartment to the nex t
through the series . of screen plates, the openings in the plate ineac h
successive compartment being . smaller than in the one preceding . As wit h
the single ..-plate screen ; mesh wires of the Tyler series have been used ,
but with the .'introduction of-one or two screens of finer•mesh•to th e
series : The fibers under the action of . the vibrating diaphragms and
agitating water sprays in .eaeh compartment, pass through . the- screen s
until they arrive on one, the openings of which are too small to allo w
their-passage . The fibers retained in•each compartment'at the end of th e
screening are washed out, filtered, dried, and-weighed ,
Analyses of several commercial groundwood pulps and one experimenta l
grow:ndwoodpulp were made on both screens in order to compare the result s
ofthe multiple-plate and'the single-plate screens . The results shown-i n
table 7 ' indicate that-the two screens are not comparable except perhap s
in the amount passing the 60-mesh wire .
The effect of the amount of pulp used in the test is shown in th e
The analyses obtained by the two screen s
analysis of'pulp Gi21 (table 7) .
are'more comparable with 50 grams of pulp than when greater amounts ar e
used. In general varying the amount of pulp causes greater variations i n
the -results"from the .,single-=plate screen than from the ■ four dilate ' so '
This isshown in table - . -The variability is greater in the four-pla t
screen for the n on 24-mesh" and passing "60-mesh n but is less for th e
int g rtnpdiate fractions . ' :
,
I
Table 7 .--Comparison of screen analyses made with the single-plate and
the four-plate fractionating screen s
Samplel :
Screen
Retained between
:Passing
Amount :Retained :
:o0-mes h
taken for :
on
:
:analysis :24 mesh :24 and :32 and :42 and :
:32 me sh : 42 me sh: 60 mesh:
Number
Grams
Percent :Percent :Percent :Percen t
:Percent
: 4 .0
: 10 .4
: 60 . 4
; 62 .3
9.1
2 .3
:
:
7,0
6 .9
: 65 . 9
: 62 . 2
.
:
7 .5
5 .3
:
:
5 .8
8 .8
; 66 . 2
: 65 . 4
5 .4
.5
:
:
7 .4
4 .3
:
:
6 .6
8 .2
: 53 . 1
: 58 . 3
:
:
5 .3
.9
:
:
6 .9
5 .0
:
9 .2
11 .1
: 69 . 3
: 64 . 5
:
:
8 .7
3 .7
:
10 .3
6 .0
: 10.2
: 10 .6
: 60 . 5
; 65 .4
:
3 .3
7. g
.
5 .8
4 .3
: 6 .7
: 11 .3
: 4g .2
: 33 . 1
19 .0
24 .5
:
:
4.9
.4
: 11 .7
: 2 .4
l4.o
28 .1
:
4 .o
.5
:
100
100
12 .2
19 .6
:
:
8 .2
.9
: Single plate
Four plate
.
100
100
27 .4
28 .7
:
.
P 901
: Single plate
Four plate
.
100
100
9 .0
18 .5
P 902
: Single plate
Four plate
.
100
100
10.3
24 .3
21
: Single plate
Four plate
.
100
100
36 .0
43 .5
P 888
: Single plate :
Four .plate
100
100
P 895
: Single dilate
Four plate
;:;
100
100
P 896
Single plate
s Four plate
.
P 097
G
:
G
21
: Single plate
Four plate
.
72
75
37 .6
4707
:
3 .8
11 .4
:
.
7 .4
5 .0
: 8 .5
: 14 .4
: 42 . 7
: 21 . 5
G
21
: Single plate
Four plate
.
50
50
38.0
35 .6
: 6 .0
: 10 .9
:
9 .0
4 .o
: 12 .5
: 13 0 5
; 34 . 5
; 34 .0
-Commercial groundwood pulp indicated by "P", experimental by "G" .
R88 4
Figure 5 .--Laboratory size four-plate diaphrag m
fractionating pulp screen .
zM 16105 F
O
Table g .--Comparison of single-plate and four-plate fractionating screen s
in regardtovariability in results caused by varying th e
amountofsampl e
Variability-
Fraction
Single plate screen
Retained on 24 mesh Retained
24 and
32 and
42 and
between :
32 mesh
42 mesh
60 mesh
Passing 60 mesh
:
: Four plate scree n
Percent
Percen t
3 .6
11 . 2
41 .2
2g . g
42 .0
23 . 9
21 .g
2g . 2
15 . 0
15 . g
-Calculated by dividing the average of all differences between value s
by the average of all values and multiplying by 100 . The amount s
of sample were 50, 75, and 100 grams of pulp from grinder run No . 21 .
(See table 7 . )
Table 9 shows a number of duplicate analyses made on the four plate screen picked at random from a large number of tests on groundwoo d
pulp . The results show duplicability to be good for this type of pulp .
It is emphasized, however, that althougn the procedure to be followed i s
not difficult, it is not possible to attain good duplication withou t
attention to manipulative details . Experience has further shown that
analytical precision cannot be, expected from a fractionating screen an d
it is for this - ,reason that an apparatus which may employ a fairly larg e
sample is to be preferred . The sample taken should be sufficiently larg e
for the smallest fraction obtained to be weighed on a balance reading t o
a minimum of one-hundredth of a gram . . A11 tests should be run i n
duplicate .
The degree of dupli,cabil6ety that may be expected on the four compartment screen of the Forest Products Laboratory is a maximum difference of 2 in fractions amounting to 10 percent or more of the pulp ,
a maximum difference of 1 in fractions amounting to between 1 and 9 per cent, inclusive, and a maximum difference of 0 .3 in fractions amountin g
to less than 1 percent, of the whole pulp .
R gg 4
- 7-
METHODS OF USING SCREEN ANALYSIS DATA
A screen analysis as ordinarily made gives from 2 to 5 Values ,
depending on the number of screens used . These may be used for•ctmparison
with a similar analysis on a pulp chosen as a standard . The use o f
several numbers for a test result is somewhat inconvenient and severa l
methods have been devised to make the data more useful . Discussion o f
the relative merits of these methods follow .
•
Screen Analysis Curve s
Screen analyses may be conveniently shown by curves . The plo t
of the percentages of the various fractions against the widths of th e
openings of the screens upon which they were retained is usually s o
irregular that a smooth curve cannot be . drawn. through the points . Ho yvever, if the cumulative percentage 's are. plotted against the screen opening a smooth curve may generally be drawn which rises continuously fro m
the amount retained on the coarsest screen opening to the amount retaine d
on a,hypothptical screen having openings of zero widths, : Since th e
cumulative percentage is defined as the amount of material that would b e
retained on ; a . given screen if it were the only one used in the analyses ,
the amount retained on the screen with openings of zero width is, o f
course, 100 percent . This type of curve is ., therefore, useful i n
estimating the amounts of material that, would . be retained on screen s
not used in the testing series ,
Either Cartesian or semilogarithm coordinates may be used fo r
plotting the widths of the screen openings . Both have certai n
advantages . Owing to a constant ratio between the widths of opening s
in the Tyler screen series the ordinates representing them on th e
uniform or arithmetical scale may become very close together as th e
widths decrease, whereas on a logarithmic scale these ordinates are no t
so compressed and thus offer a more : convenient method of plotting . •
Since the•logarithm of the screen open-lag equal to zero is infinity ,
it is impossible to draw the curve on- stiemilogarithm coordinates betwee n
the smallest screen opening used and screen with zero openings . Thu s
it is impossible to estimate the amounts that might be retained o n
screens smaller than that used in the testing series when the plo t
is made on semilogarithm coordinates . Figure 6 shows several typica l
cumulative curves for groundwood pulp drawn on uniform or Cartesian
coordinates* .
Screen analysis data shown in the form of curVes may o y be •
compared visually . However, certain properties of the curves )
be
expressed numerically and used as a means of comparison . A di®u's+io A
of these factors follows .- .
R884
Table 9 .--De ree of du)licabilit of screen analyses made with the
four-plate fractionating scree n
1
Sampl el
: Passing
: Retained :
Retained between
------------------------_150 mesh
on
42
and
:
g0
and
.
: 24 mesh : 24 an d :
42 mesh : 80 mesh : 150 mesh :
Number
: Percent
: Percent
: Percent
:
Percent
:
Percent
1 .6
1 .4
16 .4
14.5
19 .8
21 .2
58 . 2
59 . 2
6.
g
7 .2
20 .0
19.6
23 .4
23 .6
35 . 8
374 2
2 .1
2 .9
7 .9
10 .0
21 .5
20 .2
217 a
b
4 .0
4.0
19 .0
19.4
11 .6
11 .6
312 a
b
15 .2
14 .2
7 .3
g .4
999 a
b
17 .2
18 .8
4 .8
4 .0
P 1101 a
b
18 .9
17 .2
P 1196 a
b
11 .9
12 .4
G
57 a
b
4 .o
3 .7
G
80 a
b
14.0
12.4
G
148 a
b
G
G
.
:
18 .4
19 .5
50 . 1
47 . 4
50. 0
49 . 2
15 .4
15 .8
15 .4
15 .4
:
49 . 2
48 . 9
13 .4
13 .6
20 .0
20 .8
:
44 . 6
42 . 8
12 .1
12 .7
12 .1
12 .8
16 .6
16 .3
40 . 3
40 . 9
17 .3
15 .9
14 .8
40 . 1
16 .1
15 .0
18 .1
:
12 .9
13 .1
:
Commercial pulp indicated by 'P') experimental by "G" .
Rgg 4
:
38 .4
.
1'
-a-I .
1-
--p -1---t''.
'D
1
E
o
o
i
1
L
0
o
(41190 .10a
O
to
o
peLIT"eq.G .1 4T.12TGn2
o
0
OAT1.13TTLUMD)
0
no
i
=•vr_gla. ,
•
•
Screen Analyses Expressed in Single Number s
A number termed the coarseness modulus has been discussed in previous publications (1,2,1) . This factor is the sum of the cumulative percentages of the screen analysis (omitting the fraction that passes th e
finest screen) divided by 100 . The number is approximately proportional
to the area beneath the screen analysis-cumulative percent curve on semi logarithm coordinates, between the limits of the widths of the opening s
of the coarsest and finest screens used .
The area under the curve drawn on uniform coordinates may als o
be calculated . This is more complete than the area under the semi logarithm plot because the fraction passing the finest screen may b e
included in the calculation . Although several ways may be used t o
determine this area the following approximation based on the summatio n
of the areas of rectangles is sufficiently accurate and does not involv e
an undue amount of work : Cumulative percentages are read from the curv e
at intervals 0 .05 mm . width of screen opening :
Let :
a represent the amount retained on the coarsest screen used ;
b_ represent the cumulative amount for a screen opening 0 .05 mm . less i n
width than that of the coarsest screen ; n represent the cumulative amoun t
for a screen opening of 0 .05 mm . in width ; and s represent the sum of al l
cumulative percentage readings taken from the curve at 0 .05 mm . interval s
between b and n :
The approximate area under the curve is :
A = (19a + 56b + 50s + 56n +
1875)
100 0
The area under the semilogarithmic curve, as represented by th e
coarseness modulus, may be calculated directly from the screen analysi s
data whereas the curve must first be drawn before calculating the are a
under the uniform coordinate curve . Under normal circumstances th e
values of these areas increase as the proportion of long fibers increas e
and the proportion of the short fibers decrease . Therefore, in suc h
cases they may be used as a measure of average fiber length . The relationship is not rigid, however . It is evident that any change in the
shape of the screen analysis curve always indicates a change . in fiber size distribution, but a change in shape does not always denote a chang e
in the area under the curve or a change in average fiber length of th e
pulp .
1111
A value found to be more closely associated to the average fibe r
length of the pulp has been called the "average-screen opening for th e
pulp"(4) . It is proposed that this rather awkward phrase now be change d
to the "fiber length index," which is a more descriptive term . Thi s
index may be defined as the average of the widths of the openings in a
pair of screens of which one will just permit fibers of the averag e
Rg 8 1+
-9-
length to pass and the other just permit fibers of average length t o
be retained . It is calculated from screen analysis data by assumin g
that the summation of fiber lengths per unit weight of pulp is a constant and that the number of fibers in a unit weight of a given fraction relative ' to the number in'another fraction is in inverse rati o
to the widths of the screen openings . Table 10 illustrates one metho d
of calculation .
Table ' 10 .--1%:ethod of- calculatin• the fiber-leri th index of a pul • fro m
the screen analysi s
Range of screen
i .-esh- : Widths of
. opening
f
:Relative :Relative : Relativ e
:Average : Factor
total
: of
for
:weight of : number
of
lengt h
: widths :relative : pulp
of
of
:mumber of : retained : fibers
fiber s
:openings : fibers
: between :
screen s
B
A
D
mm .
BD .
:
AB D
: Percent .
16- 2 )4 : 0 .991-0 .701 : 0 .846 : :
1 .00
: 12 .3
: 12 .3
: 10 . 4
214- 42 :
.701- .351 :
.526
:
1 .61
:
2 .7
4 .3
2: 3
42- 80 :
.351- .175 :
.263
:
3 .22
: 20 .0
: 64 :4
. 16 : 9
80-150 :
.175- .10 14 :
.139
.
6 .09
.
63 .0
: 384 .•o
53 . 3
.10 1+- .o
Total
.052
:
16 .30
:
2 .0
: 100 .0
:
32 .6
: 1+97 .
1.7
. L8 1+ . Z
150- 0 :
:
Fiber-length index of the pulp . .8 14 .6/497 .6 = 0 .170 mm .
-Tyler standard . The 16-mesh screen represents, in this case, th e
smallest mesh through which all of the pulp w ill pass .
?It will be noted that for a given set of screens the summation o f
ABD will be a constant equal to 100 times the average of width s
of openings of the two coarsest screens .
R88 1 +
-lC-
yam., .
.. - : •v
LITERATURE CITED
(1) Schafer, E .R . and Carpenter, L.A . "Screen Analysis as an Aid i n
Pulp Evaluation ." Paper Trade J .
, No . 19, 57 (May g , 1930) .
(2)
Schafer, E .R. and Carpenter, L .A. "Groundwood Pulp Evaluation b y
Means of Static Bending Screen Analysis and Rate of Flow Tests ."
Tech . Assoc . Papers li,(l) :267 (May 15530) . Paper Trade J . 91 ,
No . 3, 61 (July 17, 1930) .
(3)
Schafer, E .R. and Heinig, Melburn . "Further Studies on Groundwoo d
Pulp Evaluation ." Tech . Assoc . Papers 14, 1, (May 1531) . Paper
Trade J .
, No . 10, 55 (Sept . 3, 1931) .
(4)
Schafer, E .R . and Santaholma, Matti . "Effect of Different Size d
Fibers on the Physical Properties of Groundwood Pulp ." Tech .
, No . 1 9
Assoc, Papers
(1) :442 (June 1934) . Paper Trade J.
4o (Nov . 9, 1933) .
(5)
Schafer, E .R . and Santaholma, Matti . "Chemical Properties of Scree n
Fractions of . Black Gum and Slash Pine Groundwood Pulps ." Tech .
Assoc . Papers 13 (1) :#31 (June 1934) . Paper Trade J. 0j, No . .19
46 (Nov . 9, 1 933)
{