dr. r. nagarajan professor dept of chemical engineering iit madras
DESCRIPTION
Advanced Transport Phenomena Module 4 - Lecture 16. Momentum Transport: Flow in Porous Media & Packed Beds. Dr. R. Nagarajan Professor Dept of Chemical Engineering IIT Madras. FLOW IN PORUS MEDIA AND PACKED BEDS. - PowerPoint PPT PresentationTRANSCRIPT
Dr. R. Nagarajan
Professor
Dept of Chemical Engineering
IIT Madras
Advanced Transport PhenomenaModule 4 - Lecture 16
Momentum Transport: Flow in Porous Media & Packed Beds
Volumes of interest may contain a solids fraction, ,
made up of:
Granular particles (sand, pebbles)
Wool (steel wool, fiberglass, etc.)
Gauzes, screens (woven metals)
Porous pellets (absorbent, catalyst support)
Void fraction, = 1 – Usually, packing geometry is random
FLOW IN PORUS MEDIA AND PACKED BEDS
> 0.8:
Large void fraction
Flow about each object may be considered as “perturbed”
external flow
Each wetted object contributes drag
< 0.5:
View as internal flow through tortuous ducts between
particles
Used in following module
FLOW IN PORUS MEDIA AND PACKED BEDS
In case of flow through a straight duct of noncircular
cross-section, effective duct diameter
For flow through a packed bed, effective average
interstitial (duct) diameter
where
4,eff
Ad
P
,
4 / 4,
/ '''i eff
volume available for flow total volumed
wetted area total volume a
'''
,
. 1
6. 1
p
p
p eff
Aparticle surface area particlevolumea
particlevolume total volume V
d
FLOW IN PORUS MEDIA AND PACKED BEDS
Effective diameter of each particle in bed
Therefore:
Appropriate Reynolds number for internal flow
where interstitial mass velocity
, 6 pp eff
p
Vd
A
, ,
2. . ,
3 1i eff p effd d
, ,,Re . . ,i i eff i i eff
bed eff
v d G dconst const
FLOW IN PORUS MEDIA AND PACKED BEDS
i iG v
Empty-duct (superficial) mass velocity
and
and
0 ,i
GG
0 ,
,Re . ,1
p effbed eff
G d
0 0 0/G v m A
FLOW IN PORUS MEDIA AND PACKED BEDS
fbed dimensionless momentum transport coefficient
Function of Re
For a single straight duct of short length, dimensionless
momentum-transfer coefficient
where2 2
41 12 2
eff
wf
d dPdz
CU U
| |zP p g z
FLOW IN PORUS MEDIA AND PACKED BEDS
Hence:
and
Correlates well with experimental data (next slide)
Ergun’s approximation:
3
,
20
1
/
p eff
bed
dPd
dzf
G
1501.75
Rebedbed
f
,
2
1/
2.12
i eff
bed
i
d dP dzf const
v
FLOW IN PORUS MEDIA AND PACKED BEDS
Experimentally determined dependence of fixed-bed friction factor fbed on the bed Reynolds’ number(adapted from Ergun (1952))
FLOW IN PORUS MEDIA AND PACKED BEDS
Laminar flow region:
Rebed < 10
fbed ≈ 150/Rebed
Darcy’s law:
Linear relationship between G0 and (-dP/dz)
Effective local permeability = G0 (-dP/dz)
>> intrinsic permeability of each particle
Fully-turbulent asymptote: (Burke-Plummer)
Rebed > 1000
fbed ≈ 1.75
FLOW IN PORUS MEDIA AND PACKED BEDS
Above equations are basis for most practical pressure-
drop calculations in quasi-1D packed ducts
Can be used to estimate incipient fluidization velocity
(by equating –p to bed weight per unit area)
Can be generalized to handle multidimensional flows
through isotropic fixed beds
Can be used to estimate inter-phase forces between a
dense cloud of droplets and host (carrier) fluid
FLOW IN PORUS MEDIA AND PACKED BEDS
For a combustion turbine materials-test program it is
desired to expose specimens to a sonic but atmospheric
pressure jet of combustion-heated air with a post-
combustion (stagnation) chamber temperature of 2000 K.
a.What should the pressure in the combustion chamber
(upstream of the nozzle) be?
b.What should the shape of the nozzle be?
c. How much air flow (g/s) must be handled if the exit jet must
be 2.5 cm. in diameter?
PROBLEM 1
d. What will the exit (jet) velocity be?
e. What will be the flow rate of axial momentum at the
nozzle exit?
f. By what factor will the gas density change in going
from the nozzle inlet to the nozzle outlet?
PROBLEM 1
a. Upstream (Chamber) Pressure
Isentropic gas flow from T0 = 2000 K to sonic speed @
1 atm. Assume
Therefore,
and
1.28
10.140
2
14.571, 1.14
1
SOLUTION 1
Now,
Also,
/ 120*
*
4.571200
p 11 Ma
p 2
pTherefore, 1 0.140 1 p 1.82 atm
1
0*
*
T 11.14 T 1754 K
T 2
SOLUTION 1
For a perfect gas EOS:
b. Nozzle shape: see figure
d. Exit (Jet) Velocity
Therefore
0 0 *
* * 0
p T 11.82 . 1.6
p T 1.14
( gas density ratio across nozzle ),
1/ 281/ 2
** *
1.28 0.83144 10 1754RTu a
M 28.97
4* *u a 8.03 10 cm / s
SOLUTION 1
c. Mass Flow Rate
But, for a perfect gas:* * *m u A
4 3**
*
4*
22 2*
*
4 4
1 28.97p M2.01 10 g / cm
RT 82.06 1754
u 8.03 10 cm / s Part d
and
dA 2.5 4.91 cm
4 4
Therefore m 2.01 10 8.03 10 cm / s 4.91
79.1 g / s
SOLUTION 1
e. Flow rate of Axial Momentum, at exit:
f.
J
4*
6 2
J mu 79.1 g / s 8.03 10 cm / s
6.35 10 g.cm / s
0
*
1.6 calculated above
SOLUTION 1
Consider a dilute acetylene-air mixture at atmospheric pressure
and 300 K with a C2H2 mass fraction of 4.5%. If the heat of
reaction of acetylene is 11.52 kcal/gm C2H2, estimate:
a.The speed with which a Chapman- Jouguet dectonation would
propagate through this mixture (km/s);
b.The stagnation pressure (atm) immediately behind such a
detonation wave;
c. The corresponding Chapman –Jouguet deflagration speed
(km/s)
PROBLEM 2
SOLUTION 2
2 2 1C H
32 2
2 2
2
0.045
g C H 11.52 kcal 10 calq 0.045
g mix g C H 1 kcal
calq 5.184 10
g
1
M 28.97
1.35
R 1.987
T 300 K
2
1
2 2
1 MqH .
2 RT
28.97 5.184 101.35 1H .
2 1.35 1.987 300
H 7.675or
1 2
1/ 2 1/ 2
1 CJ det on
1 1 1
1
Ma 1 H H 5.7156 5.716
kmu a .Ma 0.3472 5.716
s
u 1.98 km / s
2Calculation of T
2 2
effect of '' heat addit"
2 22 1
2 0 2 0 1 1p p p p
since u a
RT uq qT T T T
2c M c 2c c
SOLUTION 2
therefore
Since
we calculate:
21 1
2
p
1T u q
2TR / M
12c
1
2
p
u 1.98 km / s
q 5.184 10 cal / g
c 0.263 cal / g.K
2T 3453 K
SOLUTION 2
This fixes since:
or, equivalently:
But
therefore
2 1/ 2
21 1 2 2 2
RTG .u u .
M
1/ 2
1 1 12 1 2 1
2 2
a Ma T. Ma
a T
1 1 3 32
1
p M 1 28.971.177 10 g / cm
RT 82.06 300
3 32 1.983 10 g / cm
SOLUTION 2
0
Therefore
And, since Ma 2 =1(Chapman-Jouguet Condition)
conclusion: Detonation Speed=1.98 km/s ((Ma)1 =5.7). Stagnation Pressure:
2 12
RTp 19.39 atm;
M
/ 1
0 2 2.
1p p
2
19.39 1.863 36.1 atm
0 2p 36.1 atm;
SOLUTION 2
Consider the steady axisymmetric flow of hot air in a
straight circular tube of radius aw and cross sectional area
A
Conditions (at exit):
p=1 atm (uniform)
T= 1500 K (uniform)
aw= 5 mm
PROBLEM 3
Suppose it has been observed that the axial-velocity profile
is, in this case, well described by the simple
equation:
where U=103 cm/s. Using this observation, the conditions
above, answer the following questions:
a. If the molecular mean-free-path in air is approximately
given by the equation:
zv ( r,z( exit ))
2
zw
rv 2U 1
a
PROBLEM 3
1.2T 1
l( air ) 0.065 ( in m ),300 p
Estimate the prevailing mean free path l and the ratio of l to
the duct diameter- i.e., relevant Knudsen number for the gas
flow:
What conclusions can you now draw concerning the
validity of the continuum approach in this case?
b. Calculate the convective mass flow rate (expressed in
g/s) through the entire exit section. For this purpose assume
the approximate validity of the “perfect “ gas law, viz,:
m
PROBLEM 3
w
lKn
2a
3pMg / cm
RT
Here p is the pressure (expressed in atm), M is the
molecular weight (g/g-mole)(28.97 for air), R=82.06 (univ
gas const), and T is the absolute temperature (expressed
in kelvins). Also note that for this axisymmetric flow a
convenient area element is the annular ring sketched
below
( where is the unit vector in the z- direction).
zd 2 r.drA e
ze
PROBLEM 3
c. Also calculate the average gas velocity
at the exit section and the corresponding Reynolds’
number;
avgU m / A
avg avg avg w
avg
w
U U U 2aRe
U
2a
PROBLEM 3
d. Calculate the convective axial momentum flow rate
(expressed in g. cm/s2) through the exit section. Is your result
equivalent to Why or why not?
e. Calculate the convective kinetic-energy flow rate (expressed in g.
cm2/s3). Is your result equivalent to Why or
why not?
f. If, in a addition to the axial component of the velocity vz, the air in
the duct also has a swirl component how would this
influence your previous estimates ( of mass flow rate,
momentum flow rate, kinetic energy flow rate)? Briefly discuss.
avgmU ?
avg
2m U / 2 ?
v r,z
PROBLEM 3
g. If the local shear stress is given by the following
degenerative form of Newton’s law:
at what radial location does maximize? Calculate the maximum
value of and express your result in dyne cm -2 and Newton m-
2. Calculate the skin-friction coefficient, c f (dimensionless), at
the duct exit. At what radius does take on its minimum
value? Can- be regarded as the radial diffusion flux of axial
momentum? Why or why not? Does the rate at which work is
done by
PROBLEM 3
zrz
v
r
rz
rzrz
rz rz
the stress maximize at either of the two locations
found above? Why or why not?
h. Characterize this flow in terms of flow descriptors and
defend your choices.
rz
PROBLEM 3
well into the continuum regime
1.2
1.2
w w
6 2w
4
T 1a. l( air ) 0.065 . m
300 p
1500 1l( air ) 0.065 . m 0.448 m
300 1
d 2a 2 0.5 cm 1.00 cm
Kn l / d 0.448 10 m / 1.00 10 m
Kn 0.448 10 ( )
SOLUTION 3
b.
w w
43
2a a
z0 0w
22 4 3w
1 28.97 gpM / RT 2.35 10
82.06 1500 cm
rm v ( r ).2 r dr 2U 1 2 r dr
a
m U a 2.35 10 10 0.5 cm
therefore m 0.185 g / s
SOLUTION 3
c.
d.
laminar range
3z ,avg
4 3
w4
2
v m / A U 10 cm / s
2.35 10 10 1.0UdRe
5.40 10
4.36 10
w w
22
a a
z z0 0w
1 22 2 2 2 2w w0
rJ v v 2 r dr 2U 1 2 r dr
a
4J 4 U a 1 2 d U a
3
SOLUTION 3
6( where 540 10 poise )
But since
e.
2w
32
m U a we find :
4 4 g.cmJ mU 0.185 10 246.5
3 3 s
Note that m / A J / m
w
w
2a
zz0
32
a
0
vK v 2 r dr
2
2U 1 r / a.2 r dr
2
SOLUTION 3
Note that:
that is,
f. would not influence
Discuss.
1/ 2m J 2K
A m m
v ( r,z ) 0 zm,J ,K if v were
313 2 2w 0
2 6
5 2 3
4 U a . 1 2 d
2m U / 2 2 0.185 10 / 2
K 1.85 10 g cm / s
1/ 24U U 2 U
3
2
w2U 1 r / a
SOLUTION 3
g.
Thus
rz rz www
4 U0 at r 0 and
a
4 3
w 2
wf
2
4 5.40 10 10 dyne4.32 0.452 Pa
0.5 cm
16C
1 ReU2
SOLUTION 3
therefore
h. Descriptors:
Continuum Laminar
“Incompressible” Quasi-one-dimensional
Newtonian (viscous) Internal
Steady Single-Phase
2f 2
16C 3.67 10
4.359 10
SOLUTION 3
Estimate the drag force (Newtons) per meter of length for
each of the following long objects of transverse dimension 5
cm if placed in a heated air stream with the following
properties.
a. A circular cylinder.
b. A thin “plate” perpendicular to the stream (i.e., at 90o
incidence).
c. A thin plate aligned with the stream (i.e., at 0o incidence).
T 1200 K , p 1 atm, and U 10 m / s.
PROBLEM 4
In each case qualitatively discuss how the drag is apportioned
between “form” (pressure difference) drag and “ friction” drag
For Part ( c), is the application of laminar boundary-layer
theory likely to be valid? (Briefly discuss your reasoning) If so,
what would be the estimated BL thickness, , at the trailing
edge of the plate, i.e., at x=L? Suppose two such adjacent
plates were separated by a distance much greater than
--- would they strongly “interact” with respect to momentum
transfer?
( L )
PROBLEM 4
d. Justify the use of an incompressible Newtonian fluid
CD(Re, shape) curve to solve Part (a) ( involving the gas
air) by showing that is small enough under these
conditions to neglect .
21Ma
2 air
/
PROBLEM 4
Momentum Transfer to (Drag on) Immersed Objects
Drag/meter of axial length=? for objects of transverse
dimension 5 cm. in U =10 m/s, air @ 1 atm., 1200 K.
a. Cylinder in Cross flow
SOLUTION 4
4
43
2
4.7 10 poise
1 28.97pM g2.94 10
RT 82.06 1200 cm
v 1.60 cm / s
3
w2
3
D graph2
10 cm / s 5 cmUdRe
v 1.60 cm / s
3.13 10
Drag / LengthC 0.95
1U Frontal Area / Length
2
SOLUTION 4
SOLUTION 4
Frontal area/meter=5 cm x 100 cm = 5 x 102 cm2 /m
Therefore Drag/Length
22 4 3 22
1 1 dyneU 2.94 10 10 1.47 10
2 2 cm
2 2 2
4
4 5
1.47 10 0.95 5 10 cm
dyne6.99 10
meter
dyne N6.99 10 . 10
m dyne
0.699 N / m.
SOLUTION 4
Most of this drag is due to the p() distribution- that is, “ form” drag.
b. Plate Normal to flow: check literature
c. Plate Aligned with flow:
SOLUTION 4
In this case
and
Since ReL<106 (approx.) we expect flow in the momentum
defect Boundary Layer to be laminar. Then
where
f 1/ 2
L
1.328C ( Blasius )
Re
1/ 21/ 2 3 1LRe 3.13 10 5.59 10
L
ULRe
v
3LRe 3.13 10
SOLUTION 4
and
But total wetted area/meter=(2)(5x102)=103 cm2/m. Therefore
or
2f 1
2
2
1.328C 2.37 10
5.59 10Drag ( both sides )
therefore dimensionless1
U total wetted area2
2.37 10
SOLUTION 4
2
2 2 3 52
dyne cm NDrag 2.37 10 1.47 10 10 10
cm m dyne
2Drag 3.49 10 newtons
This drag is entirely due to - i.e., it is “friction drag”
2plates could be
brought to within
1cm apart w/o
interference
1/ 2 1L
5 55LL 0.45 cm therefore
Re 5.59 10
w x
SOLUTION 4
Reconsider the sonic jet test facility specified in problem1
from the viewpoint of turbulent jet momentum exchange
(mixing) with the surrounding atmosphere, and the
entrainment of that atmosphere
a. Calculate the Reynolds’ number at the
nozzle exit and compare it to the value above
which such jets almost certainly lead to turbulent mixing
with the surrounding atmosphere.
PROBLEM 5
j j jRe Ud / v43 10
b. Estimate the appropriate value of by assuming that the relevant
density is about the arithmetic mean between . How much
larger is the effective turbulent momentum diffusivity, , than the
intrinsic momentum diffusivity of the jet fluid ?
c. Estimate the downstream distance at which the time-averaged
velocity (axial momentum per unit mass) along the jet centerline
will be reduced to 10% of the initial jet velocity (axial
momentum per unit mass) as a result of momentum diffusion.
Compare this to the result that would have been obtained had the
jet
PROBLEM 5
tv
j and
tv j j jv /
z
jU
remained laminar (with kinematic viscosity ).
d. At this location what is the approximate ratio between the
entrained (laboratory air) mass flow and the “primary”
(combustion-heated air) jet?
j
PROBLEM 5
Momentum Transfer: Turbulent Round Jet
SOLUTION 5
Momentum Transfer: Turbulent Round Jet
is determined by . Using
we obtain:
4jet
6 2
cmJ mU 79.125 8.028 10
s
6.35 10 gm.cm / s
1/ 2J / j / 2
4 36.9 10 g / cm
SOLUTION 5
t
Therefore
that is,
Where does drop to Uj/10. For a turbulent jet:
1/ 2 5 3 2t
2j
J / 0.96 10 v 1.55 10 cm / s
cf . v 2.96 cm / s la min ar ( fluid ) momentum diffusivity
2t jv / v 5.2 10 zv 0,z
j
1/ 2
z
U / 10
3 1 J 1v 0,z . .
8 0.0161 z
SOLUTION 5
We find: z=88.6 cm =0.89 m
(This would have been 0.46 km if there had been no
turbulent enhancement in momentum diffusion.)
where
entrained primary jetm / m at this location?
SOLUTION 5
4 3
1/ 2 5 2
6.9 10 g / cm
J / 0.96 10 cm / s
1/ 2
entrained
Jm 8 . 0.0161 z ( turbulent case ),
therefore
so that at
cf.
Therefore
23
t
cmv 1.55 10
s
3entrained
2j
z 88.6 cm
m 4.05 10 g / s
m 0.792 10 g / s
jentrainedz
jet exit
Um51 at location where v 0,z
m 10
SOLUTION 5
Exercises:
1. Calculate the time-averaged profile at this
location.
2. Can you estimate the centerline, time averaged
temperature and CO2(g) concentration at this point?
(Itemize and discuss the underlying assumptions.)
SOLUTION 5
zv r,z