ARMA 13-520

Non-linear Fluid Flow through Rough-walled Fractures: the Role of Shear

displacement on the Critical Reynolds Number
Javadi, M. and Sharifzadeh, M.
Department of Mining & Metallurgical Engineering, Amirkabir University of Technology, 15875-4413 Tehran, Iran
Copyright 2013 ARMA, American Rock Mechanics Association
This paper was prepared for presentation at the 47th US Rock Mechanics / Geomechanics Symposium held in San Francisco, CA, USA, 23-26
June 2013.
This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and critical review of
the paper by a minimum of two technical reviewers. The material, as presented, does not necessarily reflect any position of ARMA, its officers, or
members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of ARMA
is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 200 words; illustrations may not be copied. The
abstract must contain conspicuous acknowledgement of where and by whom the paper was presented.

ABSTRACT: This paper experimentally examines the role of shear slip on the variation of critical Reynolds number and nonlinearity of fluid flow through rough-walled rock fractures. First, a criterion was developed by comprehensive using of
Forchheimers law to quantify the onset of non-linearity of fluid flow. Hydromechanical laboratory experiments were performed on
different initially closely mated rock fractures undergoing shear sliding to evaluate effects of mechanical displacements on nonlinear flow. At each shear sliding step, several high-precision water flow tests with different hydraulic gradients were carried out
and the critical Reynolds number was determined based on the developed criterion. The results show that; (i) the quadratic
polynomial regression in the form of the Forchheimers law provides a good accuracy, (ii) during shear sliding, the coefficients of
viscous and inertial pressure drop experiences about four and seven orders of magnitude reduction, respectively, and (iii) the
critical Reynolds number varies from 0.001 to 25 and experiences about four orders of magnitude enlarging by increasing shear
displacement from zero to 20 mm.

1. INTRODUCTION
Fluid flow through fractured rocks is a considerable
interest process in many areas of the geosciences and
engineered applications ranging from groundwater
hydrology, petroleum reservoir extraction, and
geothermal storages to safe design of hazardous waste
isolation. In many of such situations, fractures in rock
are typically existent in interconnected networks.
Therefore, before more complicated field-scale fracture
networks hydraulic behavior can be addressed, it is
necessary to investigate thoroughly the physical
processes that govern fluid flow through the primary unit
of these networks: a single rock fracture.
The general description of motion of Newtonian fluids in
a rough-walled fracture is given by of the Navier-Stokes
(NS) equations composed of a set of coupled nonlinear
partial derivatives of varying orders [1]. Unfortunately,
the existence and singularity of closed form solution of
the full NS equations in three dimensions has not yet
proven. The complexity of the NS equations combined
with the complicated geometry of rock fracture renders
them difficult to solve. On the other hands, the
simplified forms of the NS equations such as linear
Stokes equation [2, 3] and, the Reynolds lubrication
equation [4, 5] are just valid for laminar flow.

The linearity of Darcian or laminar flow comes from the

diminishing the inertia effects (i.e., kinetic energy) that
can only be anticipated for low flowrates. By increasing
the flux, inertial terms cannot be neglected with regard
to viscous forces and the pressure drop increases more
than the proportional increases in the flux that is known
as the non-linear fluid flow [6]. The critical Reynolds
number characterizes the onset of flow transition to nonlinear. However, a wide range of critical Reynolds
numbers from 4 to 2300 has been suggested for fractures
[7-11]. The possible explanations of the discrepancy
between these arguments can be addressed by the
mismatched definitions of the flow regimes transition,
non-unique criteria of the flow non-linearity, and a
sufficiently wide range of the geometrical characteristics
of the fracture and flow conditions that covered the data
and correlations of these studies. For instance, by
increasing the fracture roughness, the critical Reynolds
number will rapidly decrease [1, 10]. However, it is
questionably, whether these critical Reynolds numbers
are valid for natural rough-walled fractures that are
subjected to stress field or mechanical displacements.
The main aim of this paper is to experimentally elucidate
the role of shear slip on the variation of critical Reynolds
numbers and non-linearity of fluid flow through roughwalled rock fractures. To reach this goal, first it is

necessary to define a proper criterion to quantify the

onset of non-linearity of fluid flow. This criterion was
developed by comprehensive using of Forchheimers
law for non-linear flow. A series of hydromechanical
laboratory investigations were performed on different
initially closely mated rock fractures undergoing shear
sliding to evaluate effects of mechanical displacements
on the hydraulic behavior. At each shear sliding step,
several high-precision water flow tests with different
hydraulic gradients were carried out and the critical
Reynolds number was determined based on the
developed non-linearity flow criterion. Finally, the
variation of critical Reynolds number with shear sliding
was analyzed using mechanical behavior of fractures
recorded
during
hydromechanical
laboratory
investigations.

2. THEORY AND BACKGROUND

The general description of fluid flow in a single fracture
is given by the NS equations which express momentum
and mass conservation over the fracture void space.
Considering the steady laminar flow of a Newtonian
fluid with constant density and viscosity through a
fracture with impermeable walls, the NS equations may
be written in vector form as [12]

u. u 2u p

(1)

where is the fluid density, is the fluid kinematic

viscosity, u is the flow velocity vector, and p is the
hydrodynamic pressure. Eq. (1) is composed of a set of
coupled nonlinear partial derivatives of varying orders.
The complexity of these equations, combined with the
complicated geometry of rock fracture, renders them
difficult to solve. Therefore, further simplifications are
usually adopted to circumvent the difficulties of working
with the NS equations. By assuming that inertial forces
in the fluid are negligibly small compared with the
viscous and pressure forces, the NS equations can be
reduced to solvable forms such as linear Stokes equation
(or Reynolds lubrication equation. For viscous flow,
linear Stokes equation, under uniform one-dimensional
pressure gradient between two smooth parallel plates,
well-known as Poiseuille flow, the total volumetric flow
rate through the fracture, Q , is given by [13]

kAf
wa3
Q
p
p
12

(2)

where w is the fracture width (perpendicular to the

pressure gradient), a is the aperture of the idealized
parallel smooth fracture in the z-direction, k is the
fracture permeability, and A f is the cross-section area.
This equation is the well-known cubic law [14], which
predicts that the flow rate is proportional to the cube of

the idealized parallel smooth fracture aperture. Since

natural rock fractures consist of rough surfaces with
partially in contact and variable aperture, many
modifications have been implemented to improve the
applicability of cubic law, or Darcy's law, for roughwalled fractures.
Fundamental understanding of the physical nature of
non-linear flow is still limited and needs further
investigation. The most classical approach for the
mathematical description of these non-linear flows
through fractures is the use of the Forchheimers law as
[15]

vx vx vx

p AQ BQ 2

(3a)
(3b)

where is called non-Darcy coefficient or inertial

resistance and vx is the velocity. A and B are
coefficients that describe energy losses due to viscous
and inertial dissipation mechanisms, respectively [16]. It
is generally accepted that the Forchheimers law
adequately describes the non-linear fluid flow through
rough-walled rock fractures, especially for strong inertia
regime. The Forchheimers law can be applied over the
entire range of flow rate because it successfully reduces
to Darcys law at low flow rates.
The transition between flow regimes is usually
expressed in terms of the dimensionless Reynolds
number that gives a measure to compare the ratio of
inertial forces to viscous forces and consequently
quantifies the relative importance of these two types of
forces for given flow conditions. The Reynolds number,
Re, is defined as

Re

vD QD

Af

(4)

v is the average velocity that equals to

v Q / Af , and D is the characteristic dimension of

where

the flow system. For flow through fractures, the

characteristic dimension equals to mean aperture and the
Reynolds number can be defined as,

Re

Q
w

(5)

The critical Reynolds number characterizes the onset

of flow transition to non-linear. In recent studies, the
critical Reynolds number has been defined the point at
which the non-linear pressure drop contributes, say, 10%
of the overall pressure drop [10]. Considering the
Forchheimers law, the role of non-linear pressure drop
(or pressure gradient) reaches to 10% of the overall
pressure drop where,

BQ 2
0.1
AQ BQ 2

(6)

By simplifying Eq. (6) and introducing into Eq. (5), the

critical Reynolds number, Rec , can be defined as,

Rec

A
9 B w

(7)

This equation shows a simple measure to quantify the

critical Reynolds number for non-linear fluid flow
through rough-walled fractures.

3. EXPERIMENTAL PROCEDURE
Granite fracture specimens of 180 mm in length, 100
mm in width, and 80 mm in height have been used for
hydromechanical experiments. Specimens were provided
from intact hard granite (unit weight 2.61 g/cm3,
porosity 0.37%, uniaxial compressive strength 172 MPa)
cut precisely to small blocks of dimensions 210 mm
110 mm 80 mm. An artificial fracture was created at
mid-height of each intact rock specimens. To create this
fracture correctly and easily, two saw-cut slits with 1
mm width and 10 mm depth are initially generated at the
mid-height of the intact rock specimens.
Artificial fractures were created using a generator
apparatus consists of horizontal and normal loading
jacks and a steel guide box. First, the intact specimen
with saw-cut slits is set in the fracture generator
apparatus and a normal load about 200 kN is applied on
the upper face of specimen. A pair of steel wedges are
placed in the saw-cut slits on the opposite faces and a
constant horizontal load about 120 kN is applied through
this steel wedges. Then the normal load is gradually
decreased until the wedges penetrate into the specimen,
and the split is extended by tensile failure. During the
fracture generation, horizontal load is kept constant.
Thus, specimen is fractured smoothly under controlled
conditions without causing violent vibrations and
crashing.
An special apparatus was used for investigating the
coupled shear-flow behavior of rough-walled rock
fractures. The scheme of the shear-flow coupling test
apparatus is shown in Fig. 1a. This apparatus consists of
three main unites: mechanical testing unit, hydraulic
testing unit, and a control and data acquisition system.
This apparatus has the ability to measure precisely and
continuously the shear and normal stresses, shear and
normal displacements, and water flow rate. All these
values can be measured for large shear displacement in
the residual state of shear stress. The advantage of
application of this apparatus is its simplicity in operation
and measurement, easy control of combination of normal
and shear stresses and possible measurement of large
shear deformations.

Fig. 1. Schematic view of the experimental set up: a) fracture

shear-flow coupling apparatus, b) hydraulic sealing around
specimen.

The hydraulic testing unit consists of fracture sealing

system, inlet water supply into the fracture, discharged
outlet water collection, and measurement system. This
testing unit has been developed for linear flow
experiments through rock fracture undergoing normal or
shear loading. Therefore, lateral sides of the specimen
must be sealed during flow tests. The perspective view
of the sealing system of fracture specimen is shown in
Fig. 1b. An special rubber material, so called gel sheet
is placed on all sides of the fractured specimen to
prevent water leakage during flow test. This rubber
material is attached to the steel bars that are pushed to all
four lateral sides of lower and upper specimen. Inlet
water is supplied over the entire width of the fracture
specimen. Inlet water head, or pressure, is regulated
through the pressure gauge according to required head.
During the test, water pressure is recorded by data logger
within defined time intervals. Water discharge from
outlet is collected over the entire width of fracture.
The coupled shear-flow tests were conducted on three
different fracture specimens CI, CII, and CIII under
constant normal stress of 1 MPa, 3 MPa, and 5 MPa,
respectively. Direct shear tests were performed based on
controlling shear displacement. Shear displacement was

applied with a rate of 0.1 mm/min up to the maximum

displacement of 20 mm. Normal displacement and shear
stress were continuously recorded with 0.05 mm of shear
displacement increments. Hydraulic tests were
performed stepwise with 0.5 mm, 1 mm, and 5 mm of
shear displacement increments depending on the
sensitivity and importance of shear region. Steady-state
fluid flow rate throughout the joint was measured at each
shearing step by keeping the normal stress and shear
displacement constant. Various hydraulic tests were
performed with different pressure drops between fracture
inlet and outlet ranging from 3.6 kPa to 30 kPa. For C3
specimen, the upper limit of pressure drop extended to
54 kPa for the first 2 mm shear displacements due to the
higher normal stress applied on the specimen.

4. RESULTS
4.1. Mechanical Behavior
The shear stressshear displacement behavior and
corresponding dilatancy of initially matched and fresh
rough-walled fractures can be divided into four steps: (i)
Elastic pre-peak step, (ii) Non-linear pre-peak step, (iii)
Transient post-peak step, and (iv) Residual post-peak
step. Variation of shear stress with respect to shear
displacement under constant normal stress for all cases is
shown in Fig. 2a. For all the cases, shear stress grows up
quickly from zero to peak stress highly depending on the
applied normal stress. The peak shear stress for CI, CII,
and CIII are 0.882 MPa, 5.724 MPa, and 8.386 MPa,
respectively. By increasing normal stress, the shear
stress reaches the peak value at smaller shear
displacement, where the peak shear stress is observed in
shear displacement of 2.98 mm, 2.046 mm, and 1.968
mm for CI, CII, and CIII, respectively. This means that
shear stiffness increases with increasing the applied
normal stress. After peak, the shear stress falls suddenly
for CII and CIII and decreases gradually to reach the
residual. During residual region, shear displacements are
continued with stick-slip phenomena. But for CI, shear
stress follows by a gradual decrease after peak.
Moreover, Difference between peak and residual shear
stresses increases with increasing the applied normal
stress.
Fig. 2b shows the shear displacement vs. dilation of
fractures during direct shear sliding. Dilation was
calculated as arithmetic mean of the normal
displacements recorded by four displacements
transducers. As shown in Fig. 2b, fracture dilation shows
a non-linear descending-ascending behavior during shear
sliding. As shear stress is applied, fractures experience
negative dilation or closure due to the surfaces
interlocking. In elastic pre-peak step, fracture dilation
continues to decrease until reaches its minimum. The
minimum dilation for CI, CII, and CIII are -0.0175 mm,
-0.014 mm, and -0.017 mm, respectively. The

corresponding shear displacement of minimum dilation

for CI, CII, and CIII are 1.207 mm, 0.914 mm, and 0.856
mm, respectively, decreasing with increasing the applied
normal stress. Fracture dilation begins to grow up with
starting non-linear pre-peak step and reaches the zero
value at shear displacement of 1.43 mm, 1.73 mm, and
1.353 mm for CI, CII, and CIII, respectively.
As shown in Fig. 2b, the dilation corresponding to peak
shear stress for CI, CII, and CIII are 0.267 mm, 0.131
mm, and 0.104 mm, respectively, decreasing with
increasing the applied normal stress. The rate of dilation
increment reaches to maximum after peak shear stress,
and then decreases steadily until approaches to a
constant value. Comparison of the rate of dilation
increment after peak shear stress shows that the highest
rate occurs for CI, and the lowest is for CIII. Moreover,
the fracture dilation corresponding to residual step
decreases with increasing the applied normal stress,
where the smallest and largest fracture dilation at shear
displacement of 20 mm are observed for CIII and CI,
respectively.

Fig. 2. Mechanical behaviour of fractures during shear sliding

under constant normal stress: a) shear stress, b) fracture
dilation

4.2. Non-linear fluid flow

During shear sliding of fractures, one dimensional
steady-state flow tests were performed stepwise with 0.5
mm, 1 mm, and 5 mm of shear displacement increments.
At each shearing step, several hydraulic tests were
performed with different pressure gradients between
fracture inlet and outlet and quadratic polynomial
regression in the form of Eq. 3b was fitted to the
hydraulic results. The quadratic polynomial regression
was properly fitted to the experimental results. An
exception to this was the data measured for CIII at zero
shear displacement. This data was found to be erratic in
scattering possibly because of measurement errors and
therefore was not considered in the analysis. The
residual squared R2 is greatest 0.99 for all shearing step
in the cases of CI and CIII. For CII, the regression
provides R2 about 0.921, 0.973, and 0.957 for shear
displacement of 2 mm, 3 mm, and 5 mm, respectively.
Other results for CIII give R2 greatest 0.99.
Fig. 3 shows the quadratic polynomial regression for
hydraulic tests, for instance, on CII in 4 mm shear
displacement. The fitted quadratic polynomial in the
form of the Forchheimers law consists of a linear term
(AQ) and a non-linear term (BQ2) describing viscous and
inertial pressure drops, respectively. The role of linear
and non-linear terms in overall pressure drop is shown in
Fig. 3. At low flow rates, both the Forchheimers law
and linear prediction propose close pressure drops. Since
the Forchheimers law successfully reduces to Darcys
law at low flow rates, it can be applied over the entire
range of flow rates. By increasing flow rate or Reynolds
number, deviation from the linearity in flow through
fracture becomes more evident and inertial pressure drop
cannot be neglected with regard to viscous pressure
drop.

Fig. 4. Variation of coefficient A and B during shear sliding: a)

CI, b, CII, c) CIII.

At each shearing step, the coefficients A and B were

calculated using quadratic polynomial regression and
shown in Fig. 4. During shear sliding, the coefficient A
and B experience about four and seven orders of
magnitude reduction, respectively. From Fig. 4, it can be
found that the variation pattern of coefficients A and B is
very similar. However, the reduction rate of the
coefficient B is much higher than A.
Fig. 3. The role of linear and non-linear terms in overall
pressure drop and quadratic polynomial regression for
hydraulic tests (Results from CII in 4 mm shear displacement).

For CI, the coefficient A shows a descending variation

during shear sliding. For CII and CIII, the coefficient A
shows an ascending-descending variation during shear
sliding. The coefficient A for both CII and CIII reaches

its maximum at shear displacement of about 1 mm and

falls suddenly after 1.5 mm shear displacement. The
coefficient B shows a descending variation for CI and
CIII, and an ascending-descending variation for CII. The
coefficient B shows an abrupt reduction during the nonlinear pre-peak step (for shear displacements between 1
mm and 2 mm). The reduction rate of the coefficient A
and B decreases after shear displacement of about 3 mm.
After shear displacement of about 3 mm, the coefficient
A and B decreases gradually to attain a somewhat
constant value for shear displacements exceeding
about4mm, 7 mm, and 10 mm for CI, CII, and CIII,
respectively.

polynomial regression in the form of Forchheimers law

was fitted to the hydraulic results. The critical Reynolds
number was determined using the developed criterion
and its variation with shear sliding was evaluated.

At each shearing step, the critical Reynolds number was

calculated using Eq. 7. Variation of critical Reynolds
number during shear sliding is shown in Fig. 5. As
shown in Fig. 5, the critical Reynolds number
experiences about four orders of magnitude enlarging by
increasing shear displacement from zero to 20 mm. For
all the cases and shear displacements lower than 1 mm,
the critical Reynolds number is lower than 0.04. The
critical Reynolds number shows an ascending variation
for CI and CIII. For CI, the critical Reynolds number
shows an enormous increase for shear displacements
between 1 mm and 1.5 mm and attains a somewhat
constant value in the range of 25 for shear displacements
exceeding 3 mm (Fig. 5a). For CIII, the critical Reynolds
number increases almost consistently by increasing shear
displacement and reaches a somewhat constant value in
the range of 15 for shear displacements exceeding 5 mm
(Fig. 5c). As shown in Fig. 5b, the critical Reynolds
number shows a descending-ascending variation for CII.
For this case, the critical Reynolds number falls from
0.004 to 0.001 for shear displacements from zero to 1
mm. After shear displacement of 1 mm, the critical
Reynolds number increases gradually to attain a
somewhat constant value for shear displacements
exceeding 15 mm (Fig. 5b).

5. DISCUSSION AND CONCLUSION

This paper experimentally explains the role of shear
sliding on the variation of critical Reynolds numbers and
non-linearity of fluid flow through rough-walled rock
fractures. To reach this goal, first a proper criterion to
quantify the onset of non-linearity of fluid flow was
developed by comprehensive using of Forchheimers
law for non-linear flow. To evaluate the effects of
mechanical displacements on the non-linearity of fluid
flow and critical Reynolds number, hydromechanical
laboratory experiments were performed on different
initially closely mated rock fractures undergoing shear
sliding. At each shearing step, several hydraulic tests
were performed with different pressure gradients
between fracture inlet and outlet and quadratic

Fig. 5. Critical Reynolds number during shear sliding: a) CI,

b) CII, c) CIII.

Shear stressshear displacement behavior of the rock

fractures presents an unstable mode of failure rather than
a stable where the shear strength shows a marked drop
from the peak strength at small displacements to lower
residual one at large displacement. Fracture geometry
and surface roughness play a key role in the shear
characteristic of rock fractures. The unstable mode of
shear stressdisplacements is caused by processes
arising from surface mismatch, surface disturbances and
crushing of asperities. Due to the roughness of the
fracture surfaces, the shear displacement is accompanied
with shear dilation. Depends on the relative shear
displacement, the fracture traverses a range of unmated
conditions, rough surfaces are severely damaged from
their original form, new contacts are being made, and
aperture increases due to the dilation. All these processes
attend with rigorous changes in fracture geometry that
has obvious disturbances on non-linearity of fluid flow
and critical Reynolds number.
Generally, the onset of flow transition to non-linear can
be characterized by two different dimensionless numbers
are known as the Reynolds number and the Forchheimer
number. The Reynolds number quantifies the relative
strength of inertia forces as compared to viscous forces
at microscopic level. It should be noted that the
microscopic inertial effects do not directly infer to
macroscopic inertial effects, where high value of local or
microscopic Reynolds number does not necessarily
deduce non-linear flow at macroscopic level.
Nevertheless, the Reynolds number has been widely
used for characterization of non-linear flow and a
diversified amount of critical Reynolds numbers from 4
to 2300 has been suggested for fractures that makes a
confused condition for practical purposes. The
Forchheimer number (Fo) is defined as the ratio of nonlinear to linear pressure losses or the ratio of the
quadratic coefficient to the linear coefficient in
Forchheimers law that can be formulated as

Fo BQ

, based on the Eq. (3b). Since Forchheimer

number accounts the structure of the medium, it may

suggest a better indication of the degree of inertial
effects on flow than Reynolds number. The proposed
model, Eq. (7), combines both the Reynolds and
Forchheimer numbers that can provide a more
meaningful criterion for identifying non-linear flow
through rough-walled fractures. For instance,
considering the condition in Eq. (6) leads the critical
Reynolds number equal to the Forchheimer number of
0.11.
It is well clear that, the pressure drops highly depend on
the fracture geometry and surfaces roughness. The
fracture geometry and its attributes such as contact ratio,
dilation, relative roughness, aperture, and matedness
experience sever disturbances during forward shear

sliding. Because of contact area and geometry

heterogeneity of rough-walled fractures, the majority of
flow actually tends to coalesce in intricate assembly of
interconnected tortuous channels of varying aperture
along their length. Non-linearity of fluid flow may occur
as a result of inertial losses arising from entrance and
exit losses, changes in flow velocity or direction along
the flow path, and localized eddy formation. Such
inertial losses are generally proportional to the square of
the fluid velocity that form the repercussion of the
coefficient B. Rapid changes in the aperture along the
flow path will necessitate prompt variations in the inplane velocity, in order to maintain conservation of mass
that causes acceleration and deceleration of the flow.
These repeated acceleration and deceleration sequences
cause a departure from a linear relationship between
pressure drop and flow rate, even if the flow remains
laminar. Due to the changes in the shape of the aperture
distribution, contact area, and degree of mismatching
governing by shear process, significant variations are
expected in coefficients A and B and the critical
Reynolds number as a result.
Experimental results on non-linear fluid flow through
rough-walled fractures undergoing shear sliding indicate
that the quadratic polynomial regression in the form of
the Forchheimers law provides a good accuracy where
the residual squared R2 is greatest 0.99 for most of the
cases. The fitted quadratic polynomial in the form of the
Forchheimers law consists of a linear term (AQ) and a
non-linear term (BQ2) describing viscous and inertial
pressure drops, respectively. By increasing flow rate or
Reynolds number, deviation from the linearity in flow
through fracture becomes more evident and inertial
pressure drop cannot be neglected with regard to viscous
pressure drop. During shear sliding, the coefficient A and
B experience about four and seven orders of magnitude
reduction, respectively. Moreover, the variation pattern
of coefficients A and B is very similar. However, the
reduction rate of the coefficient B is much higher than A.
The critical Reynolds number experiences about four
orders of magnitude enlarging by increasing shear
displacement from zero to 20 mm. For all the cases and
shear displacements lower than 1 mm, the critical
Reynolds number is lower than 0.04. The critical
Reynolds number shows an enormous increase during
pre-peak shear steps and attains a somewhat constant
value in the range of 15-25 at residual post-peak.

REFERENCES
1.

Javadi, M., M. Sharifzadeh and K. Shahriar. 2010. A

new geometrical model for non-linear fluid flow
through rough fractures. J. Hydrol. 389 : 1830.

2.

Brown, S.R., H.W. Stockman and S.J. Reeves. 1995.

Applicability of the Reynolds equation for modeling

fluid flow between rough surfaces. Geophysical

Research Letters. 22(18): 25372540.
3.

Mourzenko V.V, J.-F. Thovert and P.M. Adler. 1995.

Permeability of a single fracture: validity of the
Reynolds equation. Journal of Physics II France. 5:
465482.

4.

Brown, S.R. 1987. Fluid flow through rock joints: the

effect of surface roughness. J. Geophys. Res. 92(B2):
1337-1347.

5.

Oron, A.P. and B. Berkowitz. 1998. Flow in fractures:

the local cubic law assumption reexamined. Water
Resour. Res. 34(11): 28112825.

6.

Cooke, C. E. 1973. Conductivity of Fracture Proppants

in Multiple Layers. Journal of Petroleum Technology.
25(9): 1101-1107.

7.

Huitt, J. J. 1956. Fluid Flow in Simulated Fractures.

AIChE Journal. 2(2): 259-264.

8.

Louis, C. 1969. A study of groundwater flow in jointed

rock and its influence on the stability of rock masses.
Rock Mech. Res. Rep. 10, Imp. Coll., London, UK.

9.

Skjetne, E., A. Hansen and J.S. Gudmundsson. 1999.

High-velocity flow in a rough fracture. J. Fluid Mech.
383: 1-28.

10. Zimmerman, R.W., Al-Yaarubi, A.H., C.C. Pain and

C.A. Grattoni. 2004. Non-linear regimes of fluid flow
in rock fractures. Int. J. Rock Mech. Min. Sci. 41: 3,
163-169.
11. Ranjith, P.G. and D.R. Viete. 2011. Applicability of the
cubic law for non-Darcian fracture flow. Journal of
Petroleum Science and Engineering. 78: 321327.
12. Brush, D., and N.R. Thomson. 2003. Fluid flow in
synthetic rough-walled fractures: Navier-Stokes,
Stokes, and local cubic law simulations. Water Resour.
Res. 39: 4, 1085-1099.
13. Zimmerman, R.W. and G.S. Bodvarsson. 1996.
Hydraulic conductivity of rock fractures. Transport in
Porous Media 23: 130.
14. Witherspoon, P.A., Wang, J.S.Y., K. Iwai, and J.E.
Gale. 1980. Validity of cubic law for fluid flow in a
deformable rock fracture. Water Resour. Res. 16: 6,
10161024.
15. Bear, J. 1972. Dynamics of Fluids in Porous Media.
New York: Elsevier, 764 pp.
16. Moutsopoulos, K.N. 2009. Exact and approximate
analytical solutions for unsteady fully developed
turbulent flow in porous media and fractures for time
dependent boundary conditions. J. Hydrol. 369: 7889.