Sie sind auf Seite 1von 15

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/267562795

Validity of Darcy's Law in Laminar Regime

Article  in  Electronic Journal of Geotechnical Engineering · January 2011

CITATIONS READS
6 1,859

1 author:

Olusegun Olalekan Alabi


Osun State University
27 PUBLICATIONS   24 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Thermal Rating of Nigerian Made Bare Overhead Aluminium Conductors View project

Assessment of Water Quality in the City of Osogbo View project

All content following this page was uploaded by Olusegun Olalekan Alabi on 02 October 2015.

The user has requested enhancement of the downloaded file.


Validity of Darcy’s Law in Laminar
Regime

Alabi, O. O.
Department of Mathematical and Physical Sciences,
College of Science, Engineering and Technology,
Osun State University, Osogbo.
geosciencealabi@yahoo.com

ABSTRACT
It is difficult to predict the exact range of the validity of Darcy’s law. The best method to
ascertain the range is to conduct experiments and determine the actual relationship between
the velocity, v, and the hydraulic gradient, i. For Darcy’s law to be valid, the flow through
soils must be laminar, which is true for Reynolds number less than one. However, it has been
found that digressions from Darcy’s law occur, even in the laminar regime, when inertial
forces become effective. It was found that Reynolds number increases with increase in
characteristics length, D, which represents the geometry of the passage of water in soils and
the digression from Darcy’s law increases with decrease in characteristics length with
polynomial equation of second order.
KEYWORDS: soil-water characteristics curve, reddish brown tropical soils,
compaction water content, compactive effort.

INTRODUCTION
The flow of free water through soil is governed by Darcy’s law. In 1856, Darcy
demonstrated experimentally that for laminar flow in a homogenous isotropic porous medium
completely saturated with a single fluid, the velocity of flow (v) is proportional to hydraulic
gradient, where the constant of proportionality is the coefficient of permeability (or hydraulic
conductivity (Bear, 1990; Ehigiator and Anyata, 2008). The seepage of free water in a permeable
medium is due to gravitational force (Arora, 2009).

Darcy’s law is valid if the flow through soil is laminar, and this depends in the dimension of
interstices, which in turn, depends upon the particle size (Cedergren, 1976, Bell, 1978). In fine-
grained soils, the dimension of the interstices are very small and the flow is laminar, while the
flow may be turbulent in very coarse- grained soils. It has been found that flow through soil is
laminar when Reynolds number is less than unity (Arora, 2009).

- 27 -
Vol. 16 [2011], Bund. A 28

Previous studies of fluid flow through porous media in laminar regime have shown that the
definition of permeability as constant characteristics of porous medium fails at low pressure
and/or small pores (Carman, 1937, 1938, 1939, 1948 1950, 1956; Coulson, 1949; Huang and
Ramsey, 1968 and Klinkenberg, 1941). Also, it was observed that there is disproportionally for
flow velocity versus hydraulic gradient for certain types of porous media consisting of or
containing clay. (Swartzendruber, 1962). However, the previous studies considered the theory of
gas flow in porous media (flow of Nitrogen gas in Natural Rocks) and the results did not
adequately describe the relation to the Reynolds number, which is the measure of the regime of
the fluid flow pattern to the digressions of Darcy’s law.

In this present paper, the Reynold’s number of different grain sizes of sands will be found at
different hydraulic gradients and the validity of the Darcy’s for these sand samples shall be
verified. Thereafter, the relationship between characteristics length of porous media at laminar
flow regime and digression from Darcy’s law will be established.

THEORY
It was established by Darcy through experiments that the velocity of a fluid through a porous
medium varies linearly with the loss of head hf, which indicates that the flow through porous
media is laminar.

Let us consider a circular pipe of Length L and diameter D completely filled with porous
material of grain diameter ds. The flow takes place through the interstices of the porous material.
If porosity is φ, the diameter of the passes through the particles is φds. The loss of head when
liquid flows through a porous medium can be determined by using the general expression for
head loss in laminar flow.

The loss of head for laminar flow through a pipe is given by (Rajput, 2007)

32 μ uL
hf = (1)
wD 2

Similarly, the loss of head for laminar flow through parallel plate is given by (Rajput, 2007)

12 μ uL
hf =
wD 2 (2)

Hence, the general expression for laminar flow may be expressed as

Kμ u L
hf =
wD 2 (3)
where,
h f = the loss of head in length L,
K = a constant, the value of which depends on the shape of the passage,
μ = dynamic viscosity of the fluid,

u = average velocity of flow,


w = weight density of the fluid, and
D = a characteristic length representing the geometry of the passage
Equation (3) can be used for laminar flow through porous media. The diameter of the passage
through particles is given by

d = φd (4)
s

Substituting this value of d for D in equation (4), we get

Kμ uL
hf =
wφ d s
2 2


wφ d s
2 2
 hf 
or , u=  
Kμ  L  (5)


or, u = ki (6)
where,
K = a constant, called the coefficient of permeability, and

 hf 
i =   the hydraulic gradient

 L 
Equation (6) is well known Darcy’s equation for flow of water through soil. The equation is
applicable for the Reynolds number less than 1 (Rajput, 2007)

Osborne Reynolds in 1883, with the help of a simple experiment demonstrated that the
existence of the laminar flow and turbulent flow in a closed conduit depends upon the following
factors: (Rajput, 2007).
(i) Diameter of the pipe (D),
(ii) Density of the fluid (p),
(iii) Viscosity of the fluid (μ ) , and
(iv) Velocity of flow ( ν )

By combining the above variables, Reynolds determined a non-dimensional parameter


(Arora, 2007)

- 29 -
Vol. 16 [2011], Bund. A 30

(7)

which is known as Reynolds number (Re). It is the critical parameter that determines when
transition from laminar to turbulence flow occurs.

In general case D is replaced by L, known as characteristic length and we have


ρ VL
Re = (8)
μ
It may also be expressed as
VD
Re = (9)
ν
where

v = Kinematic viscosity =

In pipe flow, the transition from laminar to turbulent flow is characterized by well-known
values of the Reynolds number which expresses the ratio of inertial to viscous forces. By analogy
with pipe flow a Reynolds number has been established in flow through porous media. In which
V is the specific discharge, v is the kinematic viscosity of the fluid, and D is a characteristic
length. If the flow at low velocities is laminar, it is expected to become turbulent at higher
velocities. Turbulent flow, if the analogy with pipe flow has to hold, would require a nonlinear,
near quadratic or quadratic relationship between velocity and head loss. Hence, Darcy’s law
would no longer be valid. Experiments conducted by Lindquist (1935), however, have shown that
digressions from Darcy’s law occur, even in the laminar flow regime, when inertial forces
become effective. Lindquist defined a special value of the Reynolds number Re* at which
digression from Darcy’s law starts because the inertia forces become important. He found Re* to
be the order of 4 in the case of a medium of uniform grain size, with a porosity, φ, of 38 percent .

Darcy’s law for liquid flow through a horizontal linear bed has been experimentally
established to be (Kelvin, 1962)
kA P1 − P2
Q= . (10)
μ L
where Q = the volume rate of flow,
k = permeability,
μ = viscosity,
A = cross-sectional area of the bed,
L = length of the bed, and
P1 and P2 = pressure at the left hand and right-hand ends of the bed.
However, a differential form of Darcy’s law is most useful. In fact, Darcy’s law for linear
bed is of very little value, all-important uses today start from the differential from (Kelvin, 1962)

For simplicity sake we choose to replace L by the coordinate of the two ends relatively to
some arbitrary coordinate system. This gives
kA P1 − P2
Q = . (11)
μ x 2 − x1

Equations (10) and (11) are correct under the assumption that steady-state flow is obtained
(that is ∂Q / ∂x = 0, ∂Q / ∂t = 0 ) that the liquid is incompressible and that k, and A are independent
of x. since Q/A has the unit of velocity, the equation becomes

k Δp
u=− (12)
μ Δx

where u may be interpreted as the bulk velocity, a pseudo velocity, which could be maintained by
the liquid of there was no rock matrix.

As Δx becomes small, equation (12) becomes

 k   dp 
υ = −    (13)
 μ   dx 

which is a differential form of Darcy’s law. The usual extension to three-dimension is written as
(Kelvin, 1962)
k
υ =− ( ∇ P + ρg ) (14)
μ
The Darcy flux is defined as

Q
q= (15)
A
where, q = Darcy flux (ms-1),
Q = Volumetric flow rate (m3s-1), and
A = Cross-sectional area (m2)

The Darcy flux is the volumetric flow rate per unit area.

Then
dh
q = −K (16)
dl

- 31 -
Vol. 16 [2011], Bund. A 32

METHODOLOGY
The mechanical analysis, known as particle size analysis of the processed samples were done
in order to obtain sand of different particle size. The soil is sieved through a set of sieves. Sieves
are generally made of spun brass and phosphor bronze ( or stainless steel or sieve cloth.). The
sieves consisting of sieve 63, 150, 212, 300 and 425 μm).

Thereafter, the saturated sample was transferred into permeameter, one after the other. The
constant -head reservoir was connected to the drainage and cap. Water is allowed to flow out
from the drainage base for sometime till a steady-state is established. The water level in the
constant-head chamber in which the sand is placed is kept constant. The head causing flow (h) is
equal to the difference in water levels between the constant-head reservoir and the constant-head
chamber. The specific discharge q is equal to the volume of water discharge per unit area, per
second, while cross-sectional area which is the area permeameter (Eq. 15)

Then, Reynolds number was determine for each sample at different hydraulic gradient to
verify if they are within the range of laminar regime as proposed by Reynolds. (Eq. 9).

RESULTS AND DISCUSSION


Table 1 presents grain size, porosity and characteristics length. Table 2 presents experimental
determined values of volume flux rate Q for samples at various. Table 3 presents Characteristics
length, hydraulic gradient, Reynolds number and digression.
Table 1: Grain size, porosity and Characteristics length

Characteristics
Sample Grain size (μm) Porosity
Length (D)(m)
A 63 0.25 + 0.010 49.950x10-6
B 150 0.333+0.002 49.950x10-6
C 212 0.364 + .001 77.680x10-6
D 300 0.400 + .001 120.000x10-6
E 425 0.420+ 0.010 178.500x10-6

Figures 1 – 5 are the plot of volume flux against hydraulic gradient for samples A – E. (Table
2) The slope of each indicates the hydraulic conductivity. Figures 1 – 5 show that volume flux is
linearly related to hydraulic gradient with coefficient of relations 0.99, 0.99, 0.99, 1.0 and 0.98
for samples A, B, C, D and E respectively.
Table 2: Experimentally determined values of volume flux rate Q for samples at various
Volume flux Volume flux Volume flux Volume flux Volume flux
Hydraulic rate q* 10-4 rate q* 10-4 rate rate q* 10-4 rate q* 10-4
Gradient (ms-1) (ms-1) q* 10-4 (ms-1) (ms-1) (ms-1)
A B C D E
1.87 1.11+ 0.02 1.98+ 0.02 2.23+ 0.04 2.94+ 0.06 5.84+ 0.07
3.75 1.98+ 0.03 3.84+ 0.04 4.40+ 0.03 6.02+ 0.05 11.7+ 0.09
7.5 2.60+ 0.07 7.44+ 0.03 8.18+ 0.05 12.20+ 0.12 23.64+ 0.06
15.000 5.08+ 0.11 14.51+ 0.10 15.87+ 0.04 24.47+ 0.04 47.30+ 0.10
30.000 14.14+ 0.14 31.25+ 0.12 34.47+ 0.14 49.07+ 0.13 94.81+ 0.15

Furthermore, plots of the experimental data show that generally volume flux, q is related to
hydraulic gradient, i for all samples (Fig. 1 – 5) as:

q = c1i − c 2
(17)

where q = volume flux


i = hydraulic gradient, and
c1 and c2 are constants.

The presence of constant c2 could not be taken as experimental error or random error because
it followed a definite pattern for all the samples. It was found to be negative for all the samples
(Figs. 1 – 5) in which sample A has the highest value and sample E has the least value. The
threshold gradients were obtained by setting v equals zero from equations of the graph of seepage
velocity against hydraulic gradient for the samples.

That is,
d′ = c 2 c1 (18)

It was found that digression, d increases with decrease in characteristics length, D (Table 3).

This is in support of Stearns’s results, in which it was found that the increase in percentage of
clay in a porous medium decreases the porosity of that medium and this in turn increases the
minimum gradient required (that is threshold gradient) before flow can occur in that medium.

The plot was prepared between maximum threshold gradient and porosity and found to be
related in polynomial of degree two (Fig. 6) with relation

d = 15.77 D2 ‒ 14.78 D + 3.46

with coefficient of correlation 0.9948. This shows that they are not linearly related. Also this
relation could be of help in predicting the range of threshold gradient of a medium of a known
characteristics length. In addition this relation can be incorporated into Darcy’s equation for
laminar flow for accurate prediction.

- 33 -
Vol. 16 [2011], Bund. A 34

Following the earlier suggestions of Von Engelhardt and Tun [1955]; Low [1961] and Arora
[2009], the non- Darcy behaviour of flow of water in fine grained sands was attributed to effect of
absorbed water. This fine-grinded soils have a layer of absorbed water strongly attached to their
surface, which is not free to move under gravity. It causes an obstruction to flow of water in the
pores and hence reduces the permeability of soil.

In addition, the hydraulic gradient which is the loss of head or energy of water flowing
through the ground is due to the friction resistance of the ground material (Bell, 2007). This
friction resistance is greater in fine-grand soils than coarse – grained soils. This could be one of
the factors that responsible for the non -proportional behaviour of the velocity – gradient
relationship in fine – grained soils.

Consequences of non-Darcy behaviour are of potential interest in several disciplines, for


example, soil water movement to plant roots and the infiltration of water into soil, are of interest
hydrologically, biologically and agriculturally. Thus, the non-proportional behaviour of velocity –
gradient relationship must be dealt with regardless of the cause.

16

14 q = 0.458i - 0.347
R² = 0.968
12
volume flux (ms-1)

10

0
0 5 10 15 20 25 30 35

hydraulic gradient

Figure 1: Volume flux against hydraulic gradient (sample A)


Table 3: Characteristics length, hydraulic gradient, Reynolds number and digression
Characteristic Hydraulic Vol. Flux
Digression
Length (D) gradient. (E-04) Re
(d)
A 1.6E-05
1.875 1.11 1.75E-03 0.76
3.75 1.98 3.12E-03
7.5 2.6 4.10E-03
15 5.08 8.00E-03
30 14.14 2.23E-02

B 5.00E-05 0.26
1.875 1.98 9.89E-03
3.75 3.84 1.92E-02
7.5 7.44 3.72E-02
15 14.51 7.25E-02
30 31.25 1.56E-01

C 7.7E-05 0.20
1.875 2.23 1.72E-02
3.75 4.4 3.40E-02
7.5 8.18 6.31E-02
15 15.87 1.22E-01
30 34.47 2.66E-01

0.08
D 1.20E-04
1.875 2.94 3.53E-02
3.75 6.02 7.22E-02
7.5 12.2 1.46E-01
15 24.47 2.94E-01
30 49.07 5.89E-01

E 1.79E-04 0.03
1.875 5.84 1.04E-01
3.75 11.77 2.10E-01
7.5 23.64 4.22E-01
15 47.36 8.45E-01
30 94.81 1.69E+00

- 35 -
Vol. 16 [2011], Bund. A 36

35

30
q= 1.037i - 0.261
R² = 0.998
25
volume flux (ms-1)

20

15

10

0
0 5 10 15 20 25 30 35

hydraulic gradient
Figure 2: Volume flux against hydraulic gradient (sample B)
40

35 q = 1.140i - 0.232
R² = 0.997
30

25
volume flux (ms-1)

20

15

10

0
0 5 10 15 20 25 30 35

hydraulic gradient

Figure 3: Volume flux against hydraulic gradient (sample C)

60
q = 1.639i - 0.123
50 R² = 1

40
volume flux (ms-1)

30

20

10

0
0 5 10 15 20 25 30 35

hydraulic gradient
Figure 4: Volume flux against hydraulic gradient (sample D)

- 37 -
Vol. 16 [2011], Bund. A 38

100

90

80 q = 3.163i - 0.09
R² = 1
70

60
volume flux (ms-1)

50

40

30

20

10

0
0 5 10 15 20 25 30 35

Hydraulic gradient

Figure 5: Volume flux against hydraulic gradient (sample E)

0.8

0.7 d= 15.77D2 - 14.78D+ 3.46


R² = 0.994
0.6

0.5
Digression (d)

0.4

0.3

0.2

0.1

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45

Characteristic length (D)

Figure 6: Digression against characteristics length


CONCLUSION
It has been found experimentally that there is digression or deviation from Darcy’s law even
at Reynolds number less than 1. Thus, it is not completely true to say that Darcy’s law is obeyed
by all samples wherever Reynolds number is less than 1. This is so because the influence of
absorbed water and friction resistance is greater in fine-grained sand than in coarse grained sands

It was also found that apart from zero-gradient, seepage velocity was also found to be zero at
higher gradients. This confirms the existence of threshold gradient range and these threshold
gradient range, i′, increases with decreasing porosity, φ,for extremely fine grained soils.

REFERENCES
1. Arora, K. R. 2009. Soil mechanics and foundation Engineering. (Geotechnical 7th
Engineering) Edition. Standard publishers distributors, Delhi.
2. Bell, F.G 1978 Some Petrographic factors relating to porosity and permeability in the Fell
sandstone of Norththumberland. Quarterly Journal Engineering Geology, 11:113-126.
3. Bear, J. and Arnold, V. 1990. Modelling Groundwater flow and pollution. Reidel
Publishing Company.
4. Carman, P.C, 1937.Fluid flow through a porous wall, Trans. Inst. Chem. Engr. London.
15,150.
5. Carman, P.C, 1938. The determination of the specific surface of powders, 1.J.Soc.
Chem. Ind.67,225.Fluidflow through a porous wall, Trans. Inst. Chem. Engr. London.
15,150.
6. Carman, P.C, 1939. The determination of the specific surface of powders, 2.J.Soc.
Chem. Ind.58,1.Fluidflow through a porous wall, Trans. Inst. Chem. Engr. London.
15,150.
7. Carman, P.C, 1948. P.C Some physical aspects of water flow in porous media,
Discussions Faraday Soc, 3,72.
8. Carman, P.C, 1950. P.C Diffusion and flow of gases and vapours through micropores,
proc. Roy. Soc. London, 203A, 55.
9. Carman, P.C,1956. P.C Diffusion and flow of gases through porous media, 182 PP.,
Academic press, New York.
10. Coulson, J.M 1949. The flow of fluids through granular beds: Effect of Particle shape and
voids in streamline flow, Trans. Inst. Chem.
11. Cedergren, H.R., 1976. Seepage, Drainage and Flow nets. Willey- Interscience
Publication, New York.
12. Ehigiator, O. A. and Anyata, B.U. 2008. Saturated hydraulic conductivity in a tropical
afisol in south- west Nigeria. Environmental Research Journal, 2(2): 70-75
13. Huang, J.H and Ramsey, T.L 1968. Fluid flow in porous media-diffusion and flow of
Nitrogen gas in natural rocks. Journal of Geophysiscs Reseach, 73(24): 7685-7686.

- 39 -
Vol. 16 [2011], Bund. A 40

14. Kelvin, R.J. 1962. On the differential form of Darcy’s law. Journal of Geophysical
Research, 67(2):731
15. Klinkenberg, L.J, 1941. The permeability of porous media to liquids and gases, Drilling
and production practice, 200.
16. Lindquist, E. 1935. On the flow of water through porous soil: Stockholm, premier.
Congres des Grands Barrages. 81 – 101.
17. Rajput, R. K. 2007 Fluid mechanics. 3rd Edition, S. Chand. and company Ltd., New
Delhi.
18. Swartzendruber, D. 1962. Non- Darcy flow liquid- saturated porous media. Journal of
Geophysical Research, 67: 2508-2510

© 2011 ejge

View publication stats

Das könnte Ihnen auch gefallen