Professional Documents
Culture Documents
1 s2.0 S0263876297800032 Main PDF
1 s2.0 S0263876297800032 Main PDF
00
IChem E
6. Pt
L
(I)
!:iP
= K1 L
. . . . . . . . .. . . ..
(3)
d2 M
'1}I;
K2
(4)
S32
S33
T~ROUGH
~.
GRANULAR BEDS.
U
. Iess groups, -RlO2
d -8'
pu are
equa Is -,
t h e d'ImenSlon
2 an
lO
pu
7J
.
'.
R=t;,.Py lO t;,.p.y. lO3
pu
obtamed, or, smce
L .8
' L . pu28 and 7J8 are
obtained. These are the groups recommended by
Blake 4 for plotting in the region of turbulent flow.
For viscous flow, the method of Blake gives rise
to the following form of the D'Arcy equation,
t;,.p.y. lO3
LU7J82 =k
(6)
Ue
lO3 t;,.p.y
or, U=kYJ 8 2 -y;-
(6a)
Shape.
I. Circle ..
2. Ellipses(a) Major axis = twice minor
axis.
(b) Major axis = 10 by minor
axis.
3. Rectangles(a) Length = breadth, i.e.,
square.
(b) Length=2 by breadth ..
(e) Length = 10 by breadth ..
(d) Length is infinite ..
4. Equilateral Triangle ..
Coresl9-
5. Pipes with
(a) Core set concentrically
(b) Core set eccentrically
(c)
ko
Remarks.
20
Poiseuille's law.
----
213
245
178
194
265
30
167,
20-30
1,7-3,0
1'2~2'0
Eccentricity <07
>07
S34
U e,
length,L.,atavelocity,~.~e,
corresponds to that
t,
= ~~2. t.~.g.
~.
and
equatio~ (7),
(tY
(8)
-:r-
(~) 2.
(i)
2,
is about 50.
fe =~.
o~r =4,9,
~,
(~)
~=V2 is
portional to (1
Substance.
22
(9)
{l":'-
He also
E)2'
)2'
one obtains:
Range of e.
Range of
Range of
permeability permeability
(observed). (calculated) .
1 : 239
1: 239
..
1 : 359
1 : 347
Silica powder
0375-0,493
II.
Diameter
of sphere.
nr in.
1
'8
In.
12 in.
l
4 'In.
is proportional to
can be as large as
(9a)
(I":'-E)2"
in good
6(~~E)
"
f.rin.
Porosity
k enrr.
0408
0396
0392
0405
0417
'----516
546
522
555
613
499
510
481
491
511
Average 498
2<
S35
-n
d2
3
D.(P2). {/ Po
G=kTJ' 36(1-)2'2L . Po
that is,
keGT1
(6b)
+~ .
III.
d2
01025
00528
0'04i3
0,02.';2
Porosit.y
o
D.(P2)=(po+P l ) (PO-PI) = (Po +P I ) . D.P.
TABLE IV.
Green and Ampt's Data for Small Glass Spheres.
Diamet.er (d)
em.
504
518
519
521
03889
03779
03689
03603
506
490
507
509
00709
03958
0384H
03715
03552
496
512
495
514
00497
03934
03806
0369
03597
484
506
4H5
516
Porosity
(d
----~-
00928
00319
0025
0387
03777
03653
03533
pm . D.P . (/
L
........ (13)
where pm is the density of the gas at the arithmetic
(Po+P l ) .
Po
pm
d
mean pressure
2
,Slllce 2P (PO+P ) an
( .)
G=kTJ' 36(1-)2'
(12)
21
26
loco cit.
loco cit.
k (Air).
k (Water).
--------
-----
572
0391
0370
587
03675
595
03635
574
520
0400
03925
535
545
0388
0384
537
555
0373
518
0376
0373
531
529
0363
521
0361
488
03895
0384
5'11
03795
502
518
03745
03705
515
0368
513
03625
520
03H05
529
0384
539
0379
53H
542
0373
0370
534
0366
549
Average value of k=534.
605
5'43
5'11
536
529
532
5M
S36
(15)
m= pgh
ko
(i) 2. 2 =5.0,
S.
E
that for spheres packed in the most loose arrangement (cubic packing), the porosity is =0476, while
the fractional free area in a plane of centres for any
layer is only 0215. In the plane parallel to this and
distant by half a diameter, the fractional free area is
unity. This would mean a great contraction and
expansion in every distance, d, for a fluid flowing
through the bed. On the other hand, Graton and
Fraser12 have pointed out that such packings should
offer different permeabilities according to the direction
of flow, whereas the characteristic of natural, granular
beds is that they present random packings, and that
the permeability is the same in all directions. In a
random packing, it may be assumed that the voids
are so evenly distributed throughout the bed that
the fractional free area at any cross-section is constant
and equal to the porosity, E.
Thus, instead of considering that the rate of flow
in a pore chan:qel is alternately increasing and
decreasing, it seems nearer the true state of random
packing to assume that it is constant. Within the
bed there cannot be any isolated pore channel, since
the whole system of voids is inter-connected so that
where the section of one void is decreasing in the
direction of flow the velocity does not increase, but
the excess of the fluid escapes to a neighbouring void,
the section of which is enlarging in the direction of
flow. While this emphasises the constancy of the
rate of flow at each point of the bed, it also makes the
sinuous character of the flow clear. It is obvious that
every flow-line of the fluid, in the continual division
and rejunction with other flow-lines, must follow a
loco cit.
1210c. cit.
21 10c. cit.
Furnas, Bull. U.S. Bur. Min., 1929, No. 307.
'4 Chilton and Colburn, J. Ind. Eng. Ohern., 1931,23, 913 ;
MacLaren White, Trans. Amer.1nBt. Ohern. Eng., 1935,31,390.
11
S37
~e
has
TABU; V.
Values of '" cnlculated from Permeability, using Kozeny's
Equation.
been intro-
=!.... = Epgh
m
a
by the capillary rise method, or
8=':"'=e /e!::>Pg
m
ty
5uYJL
(20)
Sand
Sand
Flint sand
Flint sand
35
065
095
"
"
060
TABLE
Chalmcrs, Taliaferro
and Bawlins."
Muskat and Botset."
Grcen and Ampt.'
Donat. 2 '
"
"
VI.
O'S\!
O'SI
065
069
073
065
038
089
055
028
(21)
075
(Nature
0n5
unknown.)
0'70-0'75
Angular.
Roundcd.
083
067
Jagged.
043
Jagged flakes.
.
.
.
......... (19)
\\'orkers.
1>
----------
.
~
~
!::>P.g
thatrs, U=k .-.;j36(I-e)
L
Nature of
Grain.
Material.
"
"
"
"
"
"
"
.
.
.
..
..
..
(:Ubp8.
ltcorr.
._--
0377
502
0426
428
402
0344
0397
0448
529
471
421
4n2
437
392
471
---
loe. cit.
loco cit.
3. Fancher and Lewis, J. Ind. Enq. Chem., In33, 25, 1I3H.
51 Chalmers, Taliaferro and Rawlins. Trans. Amer. Inst.1'.1in.
Metall. Eng., Pet. Div., 1932, 98, 375.
38 loc. cit.
39 Uchida and Fujita, J. Soc. Chern. Ind., Japan (suppl.
binding), 1934, 37, 724B and 791B.
00 J. M. Pirie, private communication.
27
28
S38
WI
~. (WI)
.
dl
where
=fractional weight of particles of size, d ll
and this, substituted in equation (6b), should give
keorr .=5O. The values obtained by Coulson are given
TABLE
VIII.
% Smaller
106
403
620
803
-
0418
0391
0395
0394
-
591
512
505
52
--
--
_. .
I
I
in. and
I
I
:& in.
Smaller
size.
25
lac. cit.
lac. cit.
533
53
5'8
518
5,)5
517
k eorr.
0397
0386
0383
0380
0383
0390
-
.,s. in.
Ratio 2 : 1.
1l1
199
335
500
667
809
---
k eorr,
size.
i\r
Ratio 125 : 1.
and
Smaller
size.
10 in.
f
1l7
190
29'1
346
43,)
533
642
R21
0350
0313
0313
0286
0293
0320
0334
0374
Ratio 5 : 1.
k eorr.
329
40
36
5,)
503
536
548
533
33
S39
Assuming that
E'
=~,
Ue
42
2loc. cit.
.. Bozza and Secchi, G. Chim. Ind., 1929, 11, 443 and 487.
4. Almy and Lewis, J. Ind. Eng. Chem., 1912, 4, 528;
A. J. V. Underwood, J.S.C.I., 1928, 47, 325T.
46 P. C. Carman, J.S.C.I., 1933, 52, 280T; 1934, 53, 159T
and 301T.
S40
General Correlation for .flow through Granular Beds.The decrease of the permeability as ~~ exceeds 20
appears, at first sight, to be analogous to the onset of
turbulence in a straight pipe. The change from
streamline flow, however, takes place very gradually.
I00I-
~-_+------+------_I_'---------+_-----_____1-------
IOI-------+-----~-+-------+------+_-----__ir__------
t~liS~
,
,.D/-------+-------+----~""-+----=:.-.,.,;:-_-_:-"....
-ID--il-------j-------
~.~
II
'~
DII--------+------+------+----~-+___-----+__-----
B
OOI~----+--------__+_----_+----_+---~-t+_------=~
DDD~'n,-----n-'r-----_,';;_----~;;_----_._hc_____---___.h----001
01
10
10
100
1000
pu!".s
FIG. 1.
IDe. cit.
S41
Workers.
Porosity
Nature of bed.
Schriever26
1
2
3
4
5
6
7
8
9
10
"
" 48
Arnould
Uchida and
Fujita 39
" Ray
Furnas",
and Kreisinger 49
A MachO!
o BerJO
0,361-0,391
0317
0,392-0,417
0,303-0,421
0347
0346
090
0844
0384
0,38-0,414
rp
!::l.Pg3
2S
pu
0353
"
"
Air
"
"
"
0,694-0,785
0353-0396
Nature of fluids.
f.
pu
and -S'
7J
"
Range of
e.
'IS
Approximately 0'15-0'7
Range not given. Probably
between 0,1-1,0.
14-20
0,01-1,0
02-250,0
100-4,000
500-9,000
600-13,000
100-350
0'5-250
06-250
Range not given
40-1,100
cit.
cit.
cit.
cit.
cit.
cit.
cit.
L
S42
0/=5 pu
0.1
(24)
8.
s +1
Lpu1l 8
81
Lpu1l 8 1
It may be concluded, therefore, that no elaborate
correction for wall effect is necessary, provided that
represents the average porosity of the whole bed,
including the region at the wall, and that SI is used in
the calculation of 0/.
The following example illustrates these points.
Chilton and Colbum34 published data for 16 and 0.9 6
If.
-60~_/_:..--~------_+_---__"__f----f
,,
I
I
I
I
I
I
"0
zo
\0
3'0
100
15-0
(DId)
FlO. 2.
Relation between Por08ity of Bed and Size of Container.
x Experimental.
Calculated.
pUliS
'
'"loc. cit.
st IDe. cit.
S43
10
50
20
........
........
1'0
....
..................
~t::::::-rr-;
8~
16
0'5
--
-~~
f-I
-..::=..4
18-
-- ---~
~
~
9
~
13
12
II
01
.10
20
50
100
pulqS
Curve.
Workers.
Tower packing.
1
2
Blake'
4
5
6
7
8
" 4
Arnould
Beri liO
10
..
11
12
13
14
15
16
17
18
Macha!
Rose and Higby"
III
, loco
loco
n loco
ooloc.
iii loco
500
FIG. 3.
ReBultB Jar Rings.
Key to Fig. 3.
3<
200
cit.
cit.
cit.
cit.
cit.
1000
2000
S44
(~1)
For a
!;;
=k o' (_TJ_) (for streamline flow) ... (25)
pu.
puem
and
pu e
puem
li~' g . u
lipg
or, R= L'S
=R . S . u.
(L)
L
e
and equations (25) and (26) take the forms
(27)
lip g . 3 _ _
(Le) 2 (TJS)
Lpu2S -!{i-ko L
pu .(2S)
_.
(TJS)0'25
ifJ-O
028 (Le)2.75
L
pu
and
,
(29)
e
of (\ L ) were taken as 25 and V j2,- respectively,
whence
0/=5 (:~)
(30)
and
TJS) 025
!{i=0'073 ( pu
(31)
37
S45
I~, where de is
7J
the
diameter of the pipe and de is the diameter of curvature, and thereby he was able to reduce all his results
to a single curve. Full confirmation has been given
more recently by Adler 56 The resulting curve is
reproduced in Fig. 4 as curve 1; in the same figure,
The values of
2U
~=15, the
e
:---~
____a__
,/
10
2.
..
/
,,,' /
20
---
~~
~ r.:----
50
V
~ -------
20
10
50
~
3a
zoo
100
soo
1000
2000
(1)
White's Curve
(2)
(3)
Elbows
Richter
(!)
Lorenz
(5)
Granular Bcds
(~~) 0'9
015
(Pu)
7]S
. "
(32)
(33)
., ~.A!'!!:.
~
. ~ de
~
7]
=400-18,000. Down to
~dc =
e
approaches
5"
5000
FIG. 4.
Comparison oj C Jor Curved Pipes and Jor Granular Bed8.
>l;L.
ABSCISSAE
L.---
between
i. decreased below
~ = 1,7,
the lowest
S46
G=9'0(~mr032
(35)
(d)
. i
(36)
as given by this
~=1'7-are plotted
(35)
as curves
("18)'1
t/s=Ll:1Ppu'(1.3
28 =0,4 1
pu
(24a)
SUMMARY.
(~~=O'OI-lO,OOO).
7.
17
S47
L.
ApPENDIX
P",~,ity, ~ 1~-it
"
II.
Now,
,~['-~.(~r- V~-IJ
which gives
as a function of
(~)
or of
~.
-~I
FIG. 5
FIG. 6
lies between
10 and 20
where
(1- :~l2~)'
=cP-(D-d)2
d2+nrd2 =d (1 +nr)
I
d+nrd
I+nr
and the number of tubes,
d+nrd (l+nr)2
m=--dl
I+nr
whence the total flow through the tubes, Qv is given by
III.
h2=OA2_(OD-AC)2
ApPENDIX
7T
./_
L =y2.
6
D2 h
7T
D=dia. of tube.
d =dia. of sphere.
~=1'866.
~ =20 is fairly
easily
S48
D
At l=20.
LIST OF SYMBOLS.
in gm./cm.'
in (gm./cm.')', where Po and P 1 are the
absolute pressures, on the two sides of the bed.
g=acoeleration due to gravity, = 981 cm./(sec. 2).
R=frictional force per unit area of particle surface, in
dynes/(cm. 2).
h=height of capillary rise, in cm.
L=depth or thickness of bed, in cm.
~(?)=(PO'-P1')
VIan View
~=
FIG. 7.
(S+~).
~~
FIG. 8.
='!!..d3
3
'TTD 2d 'TT 3
Free vol. = 4 V2"-3"d
l:~;-id3l'
4\/2' (!:.)2
-'TTD2d
-1-D
34V2"
(1-
u= f2(1-.) . ~P 'g
kl7JS2
L
=surface factor, such that, for any shape of particle,
S=6(1-f)fV;d. For spherical particles, <!>= 1.0, and,
for all other shapes, <!> is less than unity.
~P'g
~2) =0'528
'f3
1J;= Lpll'Sl
C = ~atio of resistance in curvilinear flow to resistance
in rectilinear flow for a given Reynold's number.