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

net/publication/329626553

Determination of discharge coefficient for broad crested weir

Article · December 2018

CITATIONS READS
2 6,982

2 authors:

Shaymaa Al Hashimi Huda Madhloom

Mustansiriya University Faculty of Engineering Al-Mustansiriya University
35 PUBLICATIONS 126 CITATIONS 21 PUBLICATIONS 169 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Shaymaa Al Hashimi on 16 March 2020.

The user has requested enhancement of the downloaded file.

> PAPER IDENTIFICATION NUMBER (151) < 1

Determination of Discharge Coefficient of

Rectangular Broad-Crested Weir by CFD
Assist. Prof. Dr. Shaymaa Al-Hashimi, Assist. Prof. Dr. Sadiq A. Salman, and Dr. Huda M. Madloom

 crested weir with vertical faces [Fritz and Hager, 1998]. The
Abstract— The hydraulic characteristics of flow over discharge characteristics for vertical face weirs with square
rectangular broad-crested weirs with varying upstream edge or rounded entrance are extensive analyzed by [Azimi
shape were studied. A large number of these studies are and Rajaratnam (2009)]. First investigations on discharge
physical model tests of rectangular weirs. In this study, capacity of broad crested weir were made by [Bazin in 1898].
Computational Fluid Dynamic (CFD) model together with An extensive series of experiment were conducted by [Hager
laboratory model of rectangular broad-crested weir were and Schwalt (1994)] for evaluating the flow feature over a
used to improve the performance of broad crested weirs. broad crested weir. Earlier work involving flow modeling over
The performance of broad crested weir was improved by
a rectangular broad crested weir has been conducted by
introducing an upstream face slope, varying from 90 to 23
[Sarker and Rhodes (2004)] who investigated a weir
degree in order to reduce the effect of flow separation.
Analysis of numerical results showed that introducing an experimentally as well as numerically (used code Fluent
upstream slope face to square edge broad crested weir V.4.4.7). Good agreement was found in the above mentioned
improved the performance and gave higher values of case for the upstream water level (stated as excellent), whereas
discharge coefficient. Two empirical equations were all other numerical results, like rapidly varied flow profile over
obtained to estimate the value of discharge coefficient in the crest, differ slightly from the results of the physical model
terms of effective head to crest height ratio, and upstream study. [Hargreaves et al (2007)] used the volume of fluid
slope face of weir with high correlation coefficients of method (Fluent V6.2) to compute the discharge also over a
0.976 and 0.985. The discharges performances obtained vertical faced broad crested weir.
from CFD analyses were compared with the observed This paper shows the application of computation fluid dynamic
results for various upstream face slopes. The results CFD model to estimate free surface flow over broad crested
obtained from all face sloes are in a good agreement. weir with different sloping of upstream face. In the current
paper, the commercial program Fluent is used to obtain flow
Index Terms— CFD, Laboratory Model, Rectangular Broad- field and pressure distribution over a broad crested.
Crested Weir, Discharge Coefficient
2. THEORETICAL ASSESSMENTS
1. INTRODUCTION The broad crested models are based on the following
Broad crested weirs are defined as structures where the assumptions: flow upstream of weir is steady, sub-critical and
streamline run parallel to each other over the weir crown, and two dimensional; the effects of flow surface tension and
the crest of the weir is horizontal [Bos, 1976]. In most layouts viscosity are eliminated. For rectangular broad crested weir
of broad crested weirs also the hydrostatic pressure is fully assuming the critical flow on the weir crest, the equation for
accomplished in the middle of crest. However in cases where flow rate is written as (Boiten, 2002):
the weir length is too small it might be that the hydrostatic
pressure is not fully accomplished [Hager, 1986]. The broad
crested weir is in addition to irrigation systems used for …….. (1)
highways, railroad and for hydropower structure. Also where,
an application as a simple discharge measurement structure is q = flow rate over the weir per unit width,
possible. An important feature of boarded crested weir is the CD = dimensionless weir discharge coefficient,
up and downstream side slope angle, which may vary between He = effective upstream head above crest including approach
velocity head,
a vertical end and a ramped slope. Sloping embankment have a
higher discharge capacity compared to a traditionally broad

Manuscript received March 20, 2015; this paper has been reviewed for
H1 = upstream head above crest,
publication in the conference July 25, 2015 and final accepted is August 20, P1 = weir height, and
2015. g = acceleration due to gravity.
The University of Al-Mustansiriyah; Engineering College, Civil
Engineering Dept., Baghdad, Iraq ([email protected],
[email protected])
> PAPER IDENTIFICATION NUMBER (151) < 2

For a free flow over broad crested weir with square water, the Volume of Fluid (VOF) method of [Hirt and
upstream corner, a functional relationship can be written as Nichols (1981)] is used. In this approach, the interface is
follows: tracked by introducing the volume fraction, αi, , where i refers
f ( q, He, P1 , g, α, υ) = 0 …………….. (2) to the phase. The volume fraction for the i th phase is the
where, fraction of the volume of a cell occupied by that phase. When
α = upstream weir face slope, and modelling the free surface between water and air, a transport
υ = kinematic viscosity. equation is solved for the water phase,
Using Buckingham Pi-theorem and after certain permissible
manipulations, Eq.(2) becomes:
CD = f (He / P1, α, Re ) ………… (3) …. (7)
Where, Re = Reynolds number. which holds in the absence of any inter-phase mass transfer
Reynolds number will have very large values and hence its and where αw is the volume fraction of water. The Finite
effect on CD will be very little, therefore, Re can be dropped Volume method is used to solve the above equations and relies
and Eq.(3) can be written as: on the flow domain being divided into a grid consisting of a
CD = f (He / P1, α) ……………. (4) large number of cells. In each cell, if it contains only water,
then αw =1; if none, then αw = 0. For cells that span the
3. EXPERIMENTAL STUDIES interface between the air and water, 0 < α w < 1. Various
[Farhoudi and Alami (2005)] produced a notable paper on algorithms have been developed to sharpen the interface
the flow features over a broad-crested weir in which they between the two phases because the interface tends to be
described a comprehensive experimental campaign using a smeared if Equation (7) is used alone.
horizontal rectangular channel which was 0.6 m wide and 0.7 In the present work, the geometric reconstruction method of
m high. A broad-crested weir of height 0.155 m and length [Youngs (1982)] was employed. To model upstream and
0.36 m was placed in the channel, which had a total length of 8 downstream boundaries, CFD software includes boundary
m. The upstream slope of the weir, ranging from 23 to 90 conditions that are specific to the open channel case, at which
degrees, was modified by a separate sloping piece which could the upstream and downstream water levels can be specified. At
be replaced in each run. The upstream face of the weir was a vertical upstream boundary, a pressure inlet is used. A
water tightened using water proof glue with special care. The pressure outlet is applied for downstream.
flow was lead to the flume through upstream tank which was
fed from the main reservoir. The flow was controlled by an 5. CFD MODEL
adjustable valve and re-circulated to the main reservoir passing The simulations model used version 6.2 of FLUENT,
though the flume, over the weir crest, and downstream tank. Fluent Inc. (2005) applied on a hydraulic model produced by
The flow depth was controlled at downstream of the weir by [Farhoudi and Alami (2005)]. The use of a 2D model can be
means of a hinged gate and measured using a pre- calibrated justified on the grounds that [Hager and Schwalt (1994)]
sharp crested V-notch located at the end of downstream tank. indicated that their experiments were essentially 2D in nature
The pressure head was measured taking the readings of 25 and only took measurements on the centerline in the channel.
piezometers along the experimental reach. This model is used With a 2D model it is possible to produce a grid that resolves
to simulate the numerical Fluent model. the vertical and stream wise directions with sufficient
accuracy. A transient numerical model was applied owing to
4. Numerical Modelling the use of the geometric reconstruction surface tracking
The numerical modelling involves the solution of the algorithm. The Re-Normalized Group theory (RNG) k -ε
Navier-Stokes equations, which are based on the assumptions turbulence model of [Yakhot and Orsag (1986)] was used with
of conservation of mass and momentum within a moving fluid. standard wall functions. This is one of a range of turbulence
In the absence of sources of mass and momentum, the models classed as Reynolds-Averaged Navier-Stokes (RANS)
conservation of mass is described by the differential equation, models as defined [by Ferziger and Peric (1997)]. They are
[Ferziger and Peric (1997]): time-averaged approximations that are widely used in
industrial applications. The RNG k -ε has known advantages
when there is strong curvature in the streamlines, as is the case
….. (5) with the accelerating flow over the weir therefore was used. To
where ρ is the density and v is the velocity of the fluid. The complete the description of the CFD modelling: the body
conservation of momentum is similarly described by the force-weighted pressure discretization scheme was used
equation: because of the presence of gravity; second-order discretization
schemes were used for the momentum, turbulence kinetic
energy and dissipation equations; and the PISO pressure
………. (6) velocity coupling algorithm was used, purely because it is
where p is the pressure and τ is the stress tensor. In order to designed specifically for transient simulations. A time step of
represent the effects of turbulence on the flow, additional 1.0×10-4 s was used throughout to keep the simulation stable
transport equations are solved for various turbulence because of the demands of the VOF model. Since the flow
quantities. To represent the sharp interface between the air and fields being modelled take of the order of 10 s to establish, it
> PAPER IDENTIFICATION NUMBER (151) < 3

can be appreciated how many time steps are required to reach 6. RESULTS
steady-state conditions. Fig. (1) shows the dimensions of the Under usual weir conditions, since the free surface level
domain and position of the boundary conditions used in the and total energy head upstream were fixed, the predicted
modelling. At inlet the upstream boundary pressure inlet was discharge and free surface shape were the chief measures of
used. When modelling a free surface in FLUENT, the user accuracy. To see the extent to which the former measure
specifies the free surface height relative to the datum, which in compared, Fig. (3) show the velocity distribution predicted by
this case was set at the weir crest. FLUENT then internally the CFD modelling. For those simulations using the RNG k -ε
calculates the volume fraction and static pressure at the inlet model, the CFD predictions and experimental data agreed very
based on the position of the face, relative to the free surface well.
position. At the downstream pressure outlet, only the free
surface height (or tail water) height was required. The upper
boundary above the air phase was specified as a symmetry
condition, which enforces a zero normal velocity and a zero
shear stress. Use of a symmetry boundary condition in this way
is a standard practice for such distant, open boundaries. All
other unmarked boundaries are set as walls. On the walls, the
no-slip condition was applied and the walls were assumed to
be smooth, since the experimental channel was constructed of
PVC and glass.

Fig. (1) Schematic of CFD domain showing: boundary

conditions.
The mesh used in CFD can have a dramatic effect on the
accuracy of the solution. In this work there were two sets of
unmatched or non-conformal grid interfaces in the domain, as
indicated in Fig. (2). At such an interface, the cell faces do not
match up exactly as they would at a conformal interface. To
demonstrate this, Fig. (2) shows the mesh used. The use of
such boundaries in regions of the flow main in which flow
gradients are low allows abrupt changes in the cell size. This
means that the overall cell count can be reduced (here the grid Fig. 3 Velocity Distribution over Broad Crested Weir
consisted of 21,375 quadrilateral cells). Extensive use was
made of geometric progressions in the meshing of the domain. 7. ANALYSIS OF RESULTS
A. Upstream Face Slope
A way of improving the performance of broad crested weir
is to introduce an upstream face slope to the upstream square
edge. The effect of upstream face slope on the behaviour of
broad crested weirs was studied by testing five weir models of
different cases, the first weir model was broad crested with
vertical upstream face( square edge), i.e., with slope α = 90o,
while, the other four were of slopes(α = 60o, 45o, 30o and 23o).

B. Empirical Expressions of Discharge Coefficient

All numerical model results of broad crested wires of
vertical upstream face and with different slope face were used
Fig. 2 Zoomed mesh regions from CFD model. as input data in a regression analysis computer program to
obtain an empirical power expression for the variation of CD
with both (He/P1) and α.
CD = 2.615 [He/P1]0.949 ……. (8)
> PAPER IDENTIFICATION NUMBER (151) < 4

with a correlation coefficient =0.976. 9. REFERENCE

Another empirical expression for the variation of CD with both [1]. Bos, M. G.: Discharge measurement structure,
(He/P1) and α can written, Laboratorium voor hydraulic an afvoerhydrologie,
CD = - 0.21 – 0.912 Sin (α) + 0.993 (He/P1) ……. (9) LandbouwhogeschoolWageningen,The etherland,rapport
with a correlation coefficient = 0.985. 4, 1976.
[2]. Hager, W.H.: Discharge measurement structure,
8. CONCLUSION Communications I, chairs de construction hydrauliques,
This paper discusses the accuracy of CFD in being able to department de genie Civil, EPFL,Lausanne, , 1986.
predict the important flow variables over a broad-crested weir. [3]. Fritz, H. M. and Hager,W. H. : Hydraulic of embankment
By using a published experimental dataset as a validation set, weirs, journal of hydraulic research,
it has been shown that it is possible to build a CFD model that 1998,124(9),963^971.
replicates the findings to acceptable levels of accuracy. The [4]. Azimi A. H. and Rajaratnam, N.: Discharge characteristics
driver behind this paper was to test the free surface modelling of weir of finite crest length, journal of hydraulic
capabilities of version 6.2 of FLUENT and in that it proved engineering, 2009, 135(12), 1081^1085.
successful. The time taken to produce steady-state free surface [5]. Bazin, H.: Experiences nouvelles sur l’ecoulement en
flows, even in 2D, is prohibitive when compared with other d’eversoir,Annales des ponts et chausses, 1898,
modelling techniques that are routinely applied to modelling 68(2),151^265.
hydraulic structures. However, for the detailed prediction of [6]. Hager, W. H. and Schwalt, M.: Broad crested weir,
flow around hydraulic structures, CFD is increasingly Journal of irrigation and drainage engineering,
becoming the most reliable and accurate modelling option. 1994,120(1), 13^26
Within the numerical modelling results of the present work, [7]. Sarker M. A. and Rhodes, D. G.: Calculation of free
the following main conclusions can be summarized as: surface profile over a rectangular broad crested weir,
1. The discharge coefficient related to overflow energy heads Flow measurement and instrumentation, 2004,
varies with the relative upstream slope. Decreasing upstream 15,215^219.
face slope makes the water surface profile fall into smooth [8].Hargreaves, D. M., Morvan, H. P. and Wrigth, N. G. :
curvature and become flatter. Also, the discharge coefficient Validation of the volume of fluid method for free surface
tends to increase with decreasing upstream weir slope. Thus, calculation, the broad crested weir, engineering
weirs with upstream face slopes of 900 and 23o have the applications of computational fluid dynamics, 2007,
smallest and largest discharge coefficient, respectively. The 1(2),136^147.
weir discharge coefficient value at a 10o slope is about 22% [9]. Boiten, W.: Flow measurement structures. Flow
higher than for a weir with a 90º upstream face slope. In Measurement and Instrumentation, 2002, 3:203^207.
addition, new empirical expressions were presented to estimate [10]. Farhoudi, J. and Alami, S. : Slope effect on discharge
the discharge coefficient of rectangular broad-crested weirs efficiency in rectangular broad crested weir with slope
with different upstream face slopes. upstream face, International Journal of Civil Engineering,
2. Flow separation occurs at the edge of rectangular broad- March 2005, Vol. 3, No. 1, 58^65.
crested weirs and extends to a certain [11]. Ferziger J and Peric M : Computational Methods for
point over the crest. Regarding the separation, flow velocity Fluid Dynamics. Springer, Verlag, Berlin, 1997.
becomes negative through this zone. Decreasing upstream face [12]. Hirt, C. W. and Nicholas, B. D.: Volume of fluid (VOF)
slope reduces the separation zone and so the negative velocity method for the dynamics of free boundaries, Journal
would not occur at the weir entrance. Hence, there was no computational physics, 1998, 39,201^225.
negative velocity in weirs with upstream slopes of less than [13]. Youngs, D. L. : Time-dependent multimaterial flow with
45o. large fluid distortion. Num. Meth. for Fluid Dynamics.
Academic Press.
Notation [14]. Fluent Inc. (2005). Fluent v 6.2, Lebanon, NH, USA.
The following symbols are used in this paper: [15]. Yakhot, V. and Orsag, S.: Renormalisation group
CD = dimensionless weir discharge coefficient, analysis of turbulence. Journal of Science and
Computing, 1986, 1:3^51.
g = acceleration due to gravity,
H1 = upstream head above crest,
He = effective upstream head above crest,
L = weir length,
P1 = weir height,
q = flow rate over the weir per unit width,
Re = Reynolds number,
α = upstream weir face slope, and
υ = kinematics viscosity.
> PAPER IDENTIFICATION NUMBER (151) < 5

1. Researcher Dr. Shaymaa Abdul Muttaleb

Hashim ; Assistant Professor in Civil
Engineering Dept.; College of Engineering ;
Al-Mustansiriyh University; Baghdad; Iraq.
e-mail: [email protected] ;
[email protected];
2. Researcher Dr. Sadiq A. Salman ;
Assistant Professor in Civil Engineering
Dept.; College of Engineering;
Al-Anbar University; Al-Anbar; Iraq.
e-mail: [email protected]

3. Researcher Dr. Huda Mahdi Madloom;

Lecturer in Civil Engineering Dept.; College
of Engineering ; Al-Mustansiriyh
University; Baghdad; Iraq.
e-mail: [email protected]

View publication stats