Influence of Spur Dike on Flow Patterns in an Open Channel
USMAN GHANI*, SHAHID ALI**, AND SABAHAT ARIF***
RECEIVED ON 27.12.2012 ACCEPTED ON 20.03.2013
ABSTRACT
Spur dikes are used for river protection purposes. They are typical in-stream structures. There existence
results in a considerable change in the flow structure of open channel flow both upstream and downstream
of the spur dike. This paper presents a numerical work conducted to get the mean and turbulent flow
features under the influence of spur dike. Primary velocity distributions over cross sections and horizontal
planes, streamlines over vertical sections and turbulence kinetic energy were investigated. The presence
of spur dike was found to have disturbed all the investigated flow features along the length of the
channel. The flow separation and recirculation was observed on the down stream side of the dike. The
reversal of flow behind the dike was directed downwards from the surface. On the basis of the results
obtained in this study, an attempt has been made to enhance the understanding of the flow patterns
which exist in case of a spur dike which can further be used for development and improvement of
formulae relevant to spur-dikes.
Key Words:
Spur Dike, Impermeable, Flow Separation, Navior-Stokes Equations.
1.
INTRODUCTION
S
pur dikes are the structures which extend from
concrete, steel or timber piles while the impermeable
the channel banks and project into the flow.
dikes are built with stones, rocks, gravel and soil.
There primary function is to protect natural
Permeable dikes are also termed as pile dikes. The
river channels from erosion. A spur dike is an effective
construction of a dike results in a considerable change
structure for bank protection which is being used world
in flow characteristics both upstream and downstream
wide with confidence. The dikes redirect flow and trap
of the dike. It results in flow separation on the
suspended sediments in back water zones. Dikes also
downstream side of the channel. The detailed
result in the formation of a safe pool for natural habitats.
experimental as well as numerical study of a spur dike
There are two types of dikes, first one is permeable and
needs basic knowledge of hydraulics and CFD
the second one is impermeable dike. The permeable
(Computational Fluid Dynamics) which is available in a
dikes normally consist of several rows of reinforced
number of standard books [1-3].
*
**
***
Assistant Professor, Department of Civil Engineering, University of Engineering & Technology, Taxila.
Senior Engineer, Atomic Energy Commission, Islamabad Pakistan.
Associate Professor, Department of Architecture Engineering & Design, University of Engineering & Technology, Lahore.
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
495
Influence of Spur Dike on Flow Patterns in an Open Channel
Although a lot of work has been done on flow behavior
around a bridge pier but less focus has been paid to spur
dikes so far. Recently Jennefier [4] did experimental work
on fixed flat bed spur-dike in an open channel. She tried
to understand the flow characteristics around the dike
by measuring velocities with the help of an ADV
(Acoustic Doppler Velocity) meter. Similarly Ahmad, et.
al. [5] conducted research on flows behavior around a
wall abutment. Some researchers did work on permeable
dikes [6-9] in the past. Some of these studies included
both flow and scour behavior around the spur-dikes or
similar structures. Research has also been done with the
help of numerical tool to understand flow characteristics
of a spur-dike. For example, Mayere, et. al. [10] developed
a numerical model for this purpose and first verified it
and then used it to get different flow behavior. Tang [11]
utilized numerical technique to investigate secondary
flow and sediment deposition pattern in the presence of
a spur-dike. Kimura, et. al. [12] used a non-linear two
equation turbulence model for flow study around a bluff
body whereas Ho, et. al. [13] modeled flow in the vicinity
of a groyne. Despite all the above efforts, there is still
much need for a comprehensive understanding of flow
in the presence of a dike so that the existing formulae
can be improved in the light of this enhanced
understanding.
In the present work, a numerical model has been used for
simulating the flow behavior around a spur dike and for
enhancing the understanding of flow due to the spur dike.
The spur dike was non-submerged. A 3D (Three
Dimensional) CFD code FLUENT 12 [14] has been used
for this purpose. The model was first validated using
available data from literature. The numerical xperiments
were then performed for a flat bed dike. Primary velocities,
stream lines representing the secondary flow field, and
turbulent kinetic energy were investigated on various
longitudinal and transverse sections both upstream and
down stream of the spur dikes. The results were analyzed
and discussed to improve the knowledge regarding spurdikes.
2.
NUMERICAL MODEL FOR SPUR
DIKES
The fundamental governing equations for all three
dimensional numerical codes are Navior-Stokes equations.
These are 3D continuity and momentum equations. In the
present work, first of all the published data from literature
was used for validating the numerical model of spur dikes
and then numerical experiments were conducted to get
different flow features. The validation is an important
aspect of any type of numerical simulation work because
it shows the ability of the model to handle the problems
under consideration and it is achieved once the modeled
results match the experimental data. For the present work
the data of Zhang, et. al. [15] has been used for validation.
They performed experimentation in an 8 m long channel. It
had a cross-sectional dimension of 40x40cm. At the
upstream, there was a 1.5m long inlet tank. The spur dike
was located at a distance of 4.5m down stream the inlet. It
had a thickness of 1cm and projected perpendicularly into
the channel with a length of 10cm. It was an impermeable
spur dike with painted wooden plate material. The channel
had a slope of 0.001. The Reynolds number and Froude's
number for this flow case were 14,250 and 0.41 respectively.
The mean velocity of the water was 0.29 m/sec. The
experimental channel has been shown in Fig. 1(a-b).
A numerical work is comprised of three steps i.e. preprocessor, solver and post-processor. The pre-processor
used in this work is GAMBIT 6.3. It is used for creation of
geometry, meshing the geometry and for assigning the
boundary conditions to different surfaces of geometry. In
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
496
Influence of Spur Dike on Flow Patterns in an Open Channel
this way, GAMBIT 6.3 has been used for pre-processing
under consideration. In this case, the following selections
the spur dike problem. The mesh was comprised of
were made. The k-ε turbulence model was selected with
unstructured triangular elements. The Fig. 2 shows the
SIMPLE algorithm. The second order upwind scheme has
plan view of the mesh. A fine grid has been used close to
been used. The boundary conditions were given at the
the spur dike while the density was gradually reduced as
walls, bed, entry of the channel, channel exit and at the
the distance from the dike increased. The mesh
free surface. As we know the velocity values are zero at
independence test showed that the results obtained from
the walls and there is zero slippage between the wall and
the present mesh will remain almost unaffected from any
fluid particles, so a no slip boundary condition was
further refinement of the mesh. This employs that our
assumed at the walls. The free surface treatment was
results were mesh independent. For this purpose, three
achieved by assuming a rigid lid boundary condition. The
different meshes were tested and results have been shown
velocity inlet and pressure outlet boundary conditions
in Fig. 3.
were assigned at the entrance and exit of the computational
domain.
FLUENT 12.0 has been used as a solver for the present
research work. Different options are to be chosen for any
As turbulence model will not be used in the regions close
type of numerical modeling keeping in view the problem
to the walls. Instead, it has been accomplished through a
(a)
(b)
FIG. 1. PLAN VIEW AND THREE DIMENSIONAL VIEW OF EXPERIMENTAL SET-UP [15]
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
497
Influence of Spur Dike on Flow Patterns in an Open Channel
standard wall function. The wall function covers the
regions close to bed and modeling is done only in the
turbulent flow regions. The convergence criteria was set
as 1x10-6. The simulated surface velocity results matched
the experimental values, so the model can be used for
further research work of spur-dikes. After validation, the
numerical experiments were done for a flat bed dike and
results have been discussed below.
3.
RESULTS AND DISCUSSION
3.1
Primary Velocity Contours
The velocity values are positive and increasing in Fig.
4(a-b) while these reduce suddenly over section 4.45 (just
upstream of the dike). The velocity distribution immediately
downstream the dike (at section 4.55m) indicates that on
the left region, velocity values have turned negative that
is a reversal of flow has occurred in this region. This results
in flow separation and possible erosion of bed downstream
the dike. As the distance from the dike increases (Fig.4(eg)), this negative velocity intensity keep on decreasing,
till the velocities over the entire section again become
similar to the one which were noticed upstream the dike.
The Fig. 5 indicates the velocity profiles at two different
locations along the depth of channel.
Fig. 4(a-g) shows the distribution of primary velocities
(mean streamwise velocities) over sections along the
channel. The primary velocity contours have been shown
over three sections upstream the dike (at 4, 4.3 and 4.45m)
while over four sections downstream the spur-dike (4.55,
5, 5.5 and 6m). The velocity contours are indicating the
separation of flow and recirculation processes downstream
the dike.
FIG. 3. MESH INDEPENDENCE TEST RESULTS SHOWING
VELOCITY PROFILES FOR THREE DIFFERENT MESHES
FIG. 2. PLAN VIEW OF THE MESH USED IN THIS WORK
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
498
Influence of Spur Dike on Flow Patterns in an Open Channel
3.2
Streamlines Over Longitudinal Vertical
Sections
dike at a lateral distance of 0.05 and 0.07 m from the bank
The stramlines have been plotted in Fig. 6(a-b) over vertical
which continues upto around five times the length of the
longitudinal sections passing perpendicular to the spur-
dike and which contributes to scouring processes.
of dike. The diagrams have indicated the flow reversal
FIG. 4(a-g). MEAN VELOCITY DISTRIBUTIONS OVER DIFFERENT CROSS-SECTIONS ALONG THE CHANNEL
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
499
Influence of Spur Dike on Flow Patterns in an Open Channel
3.3
Velocity in a Horizontal Plane
reversal and circulation of velocity behind the spur dike.
The Fig. 7 shows the velocity distribution over a plane
From the upstream side (from left to right) the flow is
parallel to bed at a height of 0.27m. It clearly shows the
more on the dike regions but behind it, it has a separation
behavior. This is an indication that flow has changed its
direction immediately after passing through the dike and
a reversal of flow has happened. The influence of dike
on the flow features is less in other parts of the domain.
This negative flow reduces in magnitude as the distance
from the dike increases both in stream-wise and lateral
direction.
3.4
Turbulent Kinetic Energy
The TKE (Turbulent Kinetic Energy) is a measure of
turbulence in the flow. The following diagram (Fig. 8) is
representing the distribution of TKE over different
sections both upstream and downstream the spur-dike.
The intensity of turbulence was very low when the flow
FIG. 5. VELOCITY PROFILES UPSTREAM AND
DOWNSTREAM OF THE DIKE
was undisturbed (Fig. 8(a-b)), however the moment it
FIG. 6(a-b). STREAMLINES ON THE LONGITUDINAL VERTICAL PLANES THROUGH THE DIKE
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
500
Influence of Spur Dike on Flow Patterns in an Open Channel
crossed the dike the turbulence increased tremendously
regions just behind the dike as compared to the rest part
as is clear from Fig. 8 (c-e). The turbulence is maximum in
of the cross section.
FIG. 7. VELOCITY CONTOURS OVER A HORIZONTAL LONGITUDINAL PLANE AT Y=0.27m FROM BED
FIG. 8(a-e). TURBULENT KINETIC ENERGY OVER DIFFERENT SECTIONS ALONG THE CHANNEL
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
501
Influence of Spur Dike on Flow Patterns in an Open Channel
4.
CONCLUSION
[6]
Chrisohoides, A., Sotiropoulos, F., Sturm, T.W., Yarnell,
D.L., and Nagler, F.A., ''Coherent Structures in Flat Bed
The paper was aimed at enhancing the understanding of
flow characteristics in the presence of a spur dike in an
open channel. It was revealed that primary velocities
exhibit flow separation and recirculation just behind the
spur dike which might lead to severe erosion of bed in
those regions. It was explored through simulation that the
reversal of flow happens and is directed downward near
the surface and close to the bed. Similarly investigation
indicated that TKE is also affected considerably due to
presence of spur-dike. It was maximum behind the spurdike and distributed over the major part of the crosssection for some distance (up to around three times the
depth of flow) behind the dike.
Abutment Flow: Computational Fluid Dynamics
Simulations and Experiments'', ASCE Journal of Hydraulic
Engineering, Volume 129, No. 3. pp. 177-186, USA,
September, 2003.
[7]
Koken. M., and Constantinescu, G., ''An Investigation
of Flow and Scour Mechanism Around Isolated SpurDikes in a Shallow Open Channel Flows”, Journal of
Water Resources Research, Volume 44, No. 8,
pp. W08406, USA, September, 2008.
[8]
Kwan, R.T.F., and Melville, B.W., ''Local Scour and Flow
Measurement at Bridge Abutments'', IAHR Journal of
Hydraulic Research, Volume 32, No. 5, pp. 661-674,
Netherland, October, 1994.
[9]
ACKNOWLEDGEMENTS
Barbhuiya, A.K., and Dey, S., ''Vortex Flow Field in a
Scour Hole Around Abutments'', International Journal of
Sediment Research, Volume 18, No. 4, pp. 310-325,
This numerical simulation was performed by using the
modeling facilities available at Department of Civil
Engineering, UET (University of Engineering &
Technology), Taxila, Pakistan. The authors are highly
acknowledged to Higher Education Commission, Pakistan
for providing these facilities at UET, Taxila, and Lahore,
Pakistan.
China, October, 2003.
[10]
of Three Dimensional Numerical Model Simulation of
Flow in the Vicinity of a Spur-Dike'', IAHR Journal of
Hydraulic Research, Volume 33, No. 2, pp. 243-256,
Netherland, April, 1995.
[11]
[2]
[3]
Dike'', Journal of Hydrodynamics, Volume 19, No. 1,
pp. 23-29, January, 2009.
French, R.H., ''Open Channel Hydraulics'', McGraw-Hill
Publishing Company, Singapor, 1985.
Tang, X.L., ''Experimental and Numerical Investigations
of Secondary Flows and Sedimentations Behind a Spur-
REFERENCES
[1]
Mayerle, R., Toro, F.M., and Wang, S.S.Y., ''Verification
[12]
Kimura, I., and Hosoda, T., ''A Non-Linear k-ε Model
Yossef, M.F.M., ''Morphodynamics of Rivers with
with Realizeability for Prediction of Flow Around Bluff
Groynes'', Delft University Press, Delft, Chapter-2,
Bodies'', International Journal for Numerical Methods
Netherland, 2005.
in Fluids, Volume 43, pp. 818-837, June, 2003.
Tu, J., Yeoh, G.H., and Liu, C., ''Computational Fluid
[13]
Ho, J., Yeo, H.K., and Ahn, W.S., ''Numerical Modeling
Dynamics, A Practical Approach'', Elsevier Publishing
Studies for Flow Pattern Changes Induced by Single
Company, New York, USA, 2008.
Gryone'',
32nd
International
IAHR
Congress,
Volume 32, pp. 1234-1242, Italy, July, 2007.
[4]
Jennifer, G.D., ''Mean Flow and Turbulence Around a
Laboratory Spur Dike'', ASCE Journal of Hydraulic
[14]
Engineering, Volume 135, No. 10. pp. 803-811, USA,
User Guide Fluent 13. Lebnon: FLUENT Incorporated
Lebanon, New Hamshire, USA, 2011.
October, 2009.
[15]
[5]
Zhang, H., Nakagawa, H., Kawaike, K., and Baba, Y.,
Ahmad, F., and Rajaratnam, N., ''Observation of Flow
''Experiment and Simulation of Turbulent Flow in Local
Around Bridge Abutment'', ASCE Journal of Engineering
Scour Around a Spur-Dike'', International Journal of
Mechanics, Volume 126, No. 1. pp. 51-59, USA,
Sediment Research, Volume 24, No. 1, pp. 33-45, China,
January, 2000.
January, 2009.
Mehran University Research Journal of Engineering & Technology, Volume 32, No. 3, July, 2013 [ISSN 0254-7821]
502