Author’s Accepted Manuscript

Incompressible Pulsating Flow For Low Reynolds

Numbers In Orifice Plates

Mara N.E. Reis, Sergio Hanriot

www.elsevier.com/locate/flowmeasinst

PII: S0955-5986(17)30001-8
DOI: http://dx.doi.org/10.1016/j.flowmeasinst.2017.01.001
Reference: JFMI1305
To appear in: Flow Measurement and Instrumentation
Received date: 28 February 2016
Revised date: 13 December 2016
Accepted date: 2 January 2017
Cite this article as: Mara N.E. Reis and Sergio Hanriot, Incompressible Pulsating
Flow For Low Reynolds Numbers In Orifice Plates, Flow Measurement and
Instrumentation, http://dx.doi.org/10.1016/j.flowmeasinst.2017.01.001
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
INCOMPRESSIBLE PULSATING FLOW FOR LOW REYNOLDS NUMBERS IN ORIFICE PLATES

Mara N. E. Reis (1)

Sergio Hanriot (2*)

(1), (2): PhD Professor

(1), (2): Pontifical Catholic University of Minas Gerais, Brazil
Av. Dom Jose Gaspar, 500-Belo Horizonte

(1): mara.nilza@terra.com.br
(2): hanriot@pucminas.br (*)

(*) Corresponding author

ABSTRACT
A flow with periodic variations is known as pulsating flow. A particularly important consequence of these flows
occurs in the presence of orifice plates, which are devices related to the determination of discharge in pipes.
Based on an experimental methodology, this work presents a study on the effects of pulsation variations and
temporal inertia on the discharge coefficient. The work includes situations beyond those contained in the
standards, particularly for liquid flow in small diameter pipes with low Reynolds numbers. The experiments
were conducted on a flow bench, capable of producing pulsating flows inside pipes. For the experimental study
of the transient flow, the propagation of a known flow fluctuation was related to a pressure fluctuation, allowing
the dynamic calibration of the measuring system. The value of the dynamic discharge coefficient was
determined, as well as comparing the coefficients of static discharge and quasi-steady. The results showed that
the inertial effects significantly affect the value of the discharge coefficient of the orifice plate, reducing the
discharge values.
Key-words: Orifice plate, pulsating flow, dynamic discharge coefficient.

1. INTRODUCTION

Pulsating flows occur in compressible and in incompressible flows, in one or two states or even in multiple
states. Among the engineering applications , the main flows of interest are those in natural gas and petrochemical
pipes, admission and discharge pipes of reciprocating or rotary compressors, pumps, internal combustion
engines, and particularly in hydraulic and pneumatic systems, where there is a need to measure discharges to
determine the dynamic characteristics of elements such as servomechanisms (FRAWLEY; GRACE, 2011).
Certain forms of pulsating flow are favourable such as those that increase combustion efficiency in combustion
chambers. Others are harmful, such as the pulsation associated with the increase of compressor instability,
leading to an increase in noise levels.
Research dating from the 1920s describes the difficulties of observing the behaviour of pulsating flows and
accurately measuring. Since this decade, several efforts have been made to predict and prevent pulsating flows
(METWALLY, 2009). During this period, the works were limited to reporting the problem, studying the nature
of the pulsation, and finding methods to eliminate it. Problems relating to the measurement of gases and liquids
in pulsating flow conditions have been recognised since the beginning of this century. The severity of this
pulsating flow depends on amplitude, frequency and waveform of the pulsation. The measurement of pulsating
flows is extremely difficult and the resulting errors can be considerable and unpredictable (BERREBI;
DEVENTER; DELSING, 2004; NABAVI; SIDDIQUI, 2010).
The main parameters that affect pulsating flow behaviour are the Reynolds number, the velocity gradient, the
pressure gradient, the pulsation frequency, the Strouhal number, and the Womersley number (METWALLY,
2009).
A pulsating flow is one where the discharge is related to time but maintains a non-null average value over a
period. This can be defined as a constant average flow over time with overlapping regular cyclical changes.
These variations may have a waveform, amplitude, and frequency, since they do not vary with time at a given
flow condition. This type of flow can be split into two parts: a pulsating flow and a periodic random fluctuation
flow (INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2003).
The problems associated with the effects of pulsation are particularly serious with differential pressure
measurement devices such as orifice plates, Venturi and nozzles, since they are the most common devices for
flow measurement. Thus, the aim is to normalise the use of differential pressure measurement systems in this
situation. Most of the investigations on the effects of pulsation refer to orifice plate type flowmeters, because of
their simplicity and frequent use in industry (AMERICAN GAS ASSOCIATION, 2002; GRAVES, 2010).
Among the flowmeters the differential pressure type is the most wide spread. It is estimated that 80% of the
installed flow meaters consists of differential pressure type, among orifice plates, venturis and flow nozzles.
Among these the orifice plate is the highlight, due to its simplicity, low cost and universal safety. It is estimated
that 40% to 42% of flow meters installed in Europe and the United States are orifice plates (READER-HARRIS;
HODGES; RUSHWORTH, 2008; SHAH et. al., 2012). The geometry, location of the pressure sockets and
installation requirements for these meters are well specified by international technical standards for steady flow,
with Reynolds numbers greater than 3150 and pipes with internal diameters above 50 mm (DELMÉE, 2003).
On the other hand, the installation of meters outside the requirements indicated by the standards or outside the
calibration range have been partially solved by using flow rectifiers at various positions upstream the meters.
(MANSHOOR; NICOLLEAU; BECK, 2011).

Because of their wide spread use, many efforts have been made to reduce uncertainty in their measurements
(BERREBI; DEVENTER; DELSING, 2004; READER-HARRIS; HODGES; RUSHWORTH, 2008; GRAVES,
2010; SHAH et al, 2012). However, all of the improvements achieved so far for steady state were based on
obtaining new empirical formulas for more accurate calculation of the discharge coefficients by changing the
parameters of the orifice such as flow, differential pressure, losses, etc., and comparing these values with those
provided by the standards (MANSHOOR; NICOLLEAU; BECK, 2011). On flows in steady and turbulent state,
the discharge coefficient is set based on the corrected Bernoulli equation, and the experimental determination of
discharge coefficients is performed according to the standards.
These attempts have been applied in order to eliminate pulsations and other irregular effects. However, correct
flow measurement using an orifice plate type flow meters depends on several factors. The reduction of
uncertainty pointed out by the new determinations of discharge coefficients and flux rectifiers alone are not
sufficient to ensure a correct measurement of the flow, especially when it comes to expensive fluids such as
natural gas and others, in which an increase of 1% over an uncertainty measurement can represent considerable
losses (READER-HARRIS; HODGES; RUSHWORTH, 2008; GRAVES, 2010).
In a steady state flow, the discharge coefficients of all the different types of primary meters are dependent on the
velocity profile for a given flow rate. The orifice plate tends to be particularly sensitive to variations in the
velocity profile because of the contraction effect.
When the Reynolds number is considered small, loss of pressure through the orifice is primarily a function of the
internal shear forces resulting from the viscosity of the fluid. If there’s an increase on the Reynolds number,
separation occurs in the downstream region of the flow. In this case, the forces of inertia dominate. This makes it
unfeasible to obtain a single empirical expression in order to obtain the discharge coefficients in the laminar flow
range. In this range, the Reynolds number can change strongly and differently from the values of the discharge
coefficients.
As can be seen, the entire success of this method consists of obtaining experimental relationships that provide
more accurate discharge coefficients values.
This work aimed to experimentally study incompressible, laminar, and pulsating flows in a transient state
through a concentric orifice plate with a sharp edge. The experiments were performed on an arrangement called
a flow bench capable of producing pulsating flows inside pipes.
Experimental methodology was used in order to study the behaviour of the discharge coefficient including the
pulsating transient parameters.

2. LITERATURE REVIEW

Since the mid-1970s, several efforts have been made to predict the behaviour of pulsating flows and thereby,
standardise the use of differential pressure measuring systems in this situation.
The pulsation can be transmitted from one point to another, having its origin in both upstream and downstream
positions of the flowmeter. However, amplitudes can be small and, depending on the distance from the pulsation
sources to the flowmeter, they can be reduced by compressibility effects (both in liquids and in gases) to
undetectable levels at the location of the flowmeter. (INTERNATIONAL ORGANIZATION FOR
STANDARDIZATION, 2003).
Since, generally, the effects produced by pulsation in the measurement of flows are not taken into consideration,
accurate determination of the error in the measurement of flows subject to pulsation in industry can prevent
significant financial losses for the consumer of a gas or liquid or the supplier of the product (MILLER, Richard,
1996; AMERICAN GAS ASSOCIATION, 2002; MILLER, Michael, 2003;McKEE, 2009).
The ISO/TR 3313 standard came as a first response to this need by providing the appropriate parameters for
the correction of flow values obtained in the presence of pulsation (INTERNATIONAL ORGANIZATION FOR
STANDARDIZATION, 2003). On the other hand, standard API MPMS 14.3.1 (AMERICAN PETROLEUM
INSTITUTE, 2003) recognises the influence of pulsation on the flow and suggests that care should be taken to
minimise the effects of pulsation on the flow. In addition, according to standard ISO/TR 3313, pulsation
generates errors in the flow measured by differential pressure systems because the average static and dynamic
pressure gradients are different. Systems whose secondary elements have a slow response to flow variation are
less susceptible to the errors caused by the variation in the average of the dynamic differential pressure.
However, such devices are incapable of measuring instantaneous flow, because the response times that are
usually recommended in standards are at least ten times the value of the fluctuation period (INTERNATIONAL
ORGANIZATION FOR STANDARDIZATION, 2003).
Morrison et al (1995) verified that some facilities that follow API 2530 contained errors of around 5%
(AMERICAN PETROLEUM INSTITUTE, 1985). The authors presented a detailed study of the rotational effect
of streamlines on the discharge coefficient. Research has indicated that for diameter ratios between 0.43 and
0.65, the reflection of recirculation on the discharge coefficient decreases with an increase in the diameter ratio.
Jankowski et al (2008) developed a model to predict the pressure drop and the discharge coefficient for
incompressible flows through an orifice with a length / diameter ratio greater than zero for a wide range of
Reynolds numbers. The pressure drop for the flow through an orifice was represented by two consecutive
pressure drops, i.e. a pressure drop for flow through the orifice with sharp edges and a pressure drop for the fully
developed flow in the straight tubing section. These two pressure drops have been represented in the model
using experimental data and correlations for the designed flow and for the flow through the sharp edges of the
orifice. This work is not only useful as a design correlation for products that rely on flow through an orifice, but
also to explain some of the difficulties that previous authors found when comparing experimental observations
and available theories. This model can be easily extended in order to consider a certain number of geometric
effects such as eccentric orifices with an input radius with a value different from zero.
Majid and Siddiqui (2010) studied the most advanced techniques in the past years to measure velocity in
pulsating flows: Hot Wire Anemometry - HWA, Ultrasound Doppler Velocimetry - UDV, Laser Doppler
Velocimetry - LDV, Particle Image Velocimetry – PIV, and X-ray Optical Tomography - XOT. The authors
discussed and presented the advantages and limitations of these techniques. Based on these studies, they noticed
that for the measurement of speed at one point, HWA presents a higher dynamic range and high spatial
resolution and is suitable for measurements at high frequencies. For non-intrusive measurements and greater
accuracy at the measurement point, LDV is indicated. To measure the total velocity field, PIV has good temporal
and spatial resolution and is thus suitable for various applications in pulsating flows.
In the case of turbulent flow, a recirculation factor can be encountered, which is typical to the flow itself.
According to Murakami and Kito (1980), this change in the streamlines can significantly affect the discharge
coefficient. However, the literature shows large discrepancies in the presentation of results. While the work of
Mottram and Rawat (1986) showed an increase of 2% in the discharge coefficient, values higher than 8% were
registered by Brennan et al (1989) for identical values of Reynolds numbers.
Regarding steady state flows through orifice plates, there are many works describing different geometries
and flow conditions based on both experimental research and numerical methods. Most of these works involve
turbulent flows within the same range of Reynolds numbers and diameters established by the standards in order
to be able to validate their numerical or experimental results (ZANKER; GOODSON, 2000; SANNIKOV;
FEDOROV, 2006; WHITSON; STOBIE, 2009; GÜNDOĞDU; YILMAZ, 2010; HOLLINGSHEAD, 2011).
The transient issue in hydrodynamic processes has been considered by several authors, providing increasingly
accurate information as computational techniques have been developed. Thus, more complex flows were treated
not only on an experimental basis, but also numerically. However, although the transient flow through orifice
plates has been treated in several aspects, most of these works deal with cases with very specific conditions,
which hinder experimental or numerical reproduction by a lack of technical information and construction detail.
Among the studies found in the literature that deal with transients flows using differential pressure flowmeter
devices, the following studies can be highlighted: Trengrouse, Imrie and Male (1966), Wright, Taplin and Wylie
(1983), Yokota and Nakano (1988), Pedroso (1988 and 1990), Bajura and Jones (1991), Botros, and Jungowski
Petela (1992), Yakhot, Grinberg and Nikitin (2004), Ünsal and Durst (2006), and Durst et al (2007).
In several studies, it was possible to verify the concern about the element responsible for the excitement of
the system. Thus, Alpay (1971), Mottram (1974), Mottram (1981), Mohammad and Mottram (1981), Gajan et al
(1992), Stroll et al (1993), Valle (1994), Jungowski and Weiss (1996), Washio et al (1996), Addison et al
(1997), Beaulieu et al (2011), and Svete et al (2012) used sinusoidal generators in their experiments to generate
fluctuations around a known flow.
Mottram (1974, 1981); Mohammad and Mottram (1981) performed an experimental study considering
incompressible pulsating flow through an orifice plate with small variations in density and expansion factor, ε,
usually higher than 0.99., and pulse frequency range of 5-500 Hz. The Strouhal number was less than about 0.02.
In this study, the flow was almost permanent and the authors concluded that a constant discharge coefficient
could be applied under steady flow conditions.
For high Strouhal number values, the inertia term gives a significant contribution to the instantaneous differential
pressure and is not sufficient to eliminate the square root error.
The dynamic behaviour of a flow depends on the pulsation frequency and is expressed by the Womersley
number, W. The Womersley number is used to characterise the periodic nature of the flow, and represents the
relation between inertial forces and viscous forces, but emphasises the dependent terms of time instead of the
stationary inertial terms. Pulsating flows are classified into quasi- steady and dominant inertia regarding the
Womersley number and the limits are indicated by W = 1.32 and W = 28, respectively. The values quoted for the
Womersley number are considered critical values on which the nature of the flow shows radical changes
(GÜNDOĞDU; YILMAZ, 2010; ALFREDSSON et al, 2012).
Beaulieu et. al. (2011) developed a numerical model for pulsating flow measurement for application in
mechanical respiratory systems using a Venturi. The results were experimentally validated, the prototype was
tested using water, for quasi-permanent flow. The velocity profiles were measured by particle velocimetry
images (PIV) and the equipment was calibrated by comparison with an ultrasonic meter. The venturi discharge
coefficient was determined as a function of the Reynolds number for steady flow and for low frequencies (≤ 4
Hz) with good agreement of the results.
Svete et. al. (2012) developed a flow bench for experimental investigations on the effects of sinusoidal pulses on
flow meters using water as a fluid at the frequency of 10 Hz to 1 kHz. The experimental dynamic characteristics
of the pulsating flow were obtained from the differential pressure measurements through an orifice plate and
were compared with a mathematical model using the characteristics method. The frequency and the amplitude of
the pulse were able to be reproduced, but with some limitations, as example, the variation in the discharge
coefficient.
Ünsal and Durst (2006) and Durst et. al. (2007) presented a technique for measuring the transient mass flow in
pulsating flows, and its performance was demonstrated experimentally in fuel injector nozzles. The application
of the mass flow measurement technique is described for pulsating air flows. The performance tests were
performed for different time variations of the pulsations of the mass flow. The authors have shown that the mass
flow rate of the average flow and pulsating flow can be separated and both can be accurately measured.
Ribas (2004) developed a numerical tool capable of predicting the behavior of turbulent, incompressible,
pulsating flows, and fully developed in pipes. Three models of turbulence were tested for the turbulent
magnitudes: the Launder and Sharma k-ε model, the Wilcox k-ω model and the Durbin model. In the paper, the
author considered the modeling of transients associated with the flows in suction / discharge systems of
reciprocating compressors and in the gap between piston and valve plate, during the emptying of the cylinder. In
the case of suction and discharge systems, the modeling was aimed at evaluating flow parameters, such as shear
stress in the wall and velocity profile, under different levels of transients. The governing equations were
discretized according to the finite volume methodology and solved through the TDMA algorithm. Numerical
results, validated by comparison with experimental data, showed that transients cause significant changes in
shear stress in the wall and other flow parameters.
Yakhot, Grinberg and Nikitin (2004) studied numerically the incompressible, pulsating laminar flow through a
tube with a restriction. The fluid is driven by a senoidally imposed variable pressure difference, and governed by
the balance between local acceleration, inertia and viscous forces. For pulsating flow through the orifice in the
tube, an oscillating recirculation bubble develops behind the orifice. The induced flow, the length of the
recirculation bubble, as well as the phase angle , with respect to the imposed pressure difference, Womersley’s
number and Reynolds’ number were computed for different relationship of restriction diameter. The presented
results allowed the distinction between the regimes with insignificantly non-linear effects and restriction those
that can be considered as almost permanent.
Lee et. al. (2007) performed numerical simulations for pulsing sinusoidal laminar flow in a tube with a smooth
restriction. The second order finite volume method was developed to solve the governing equations of the
problem. The effects of Reynolds’ and Womersley’s numbers, pulse amplitude, diameter ratio, and restriction
diameter were studied. In the study, the authors demonstrated that the dynamic nature of the pulsating flow
depends heavily on the frequency. They also concluded that the recirculation zones occur simultaneously
upstream and downstream of the restriction as the instantaneous flow approaches zero for each Womersley’s
number. The deceleration of the flow in the pulsating cycles tends to increase in the recirculation region, and this
effect becomes more significant with the increase in pulse amplitude.
Doblhoff et al (2011) reviewed the theoretical basis of pulsating flow measurement through orifice plates,
providing a summary of the work done in this field and showing the limits of applications of existing theories.
Special attention was given to temporal inertia and the application of its expressions previously developed and
found in the literature. Other factors that influence the measurement of pulsating flow through orifice plates,
such as change of velocity profiles and the effects of the assembly between pressure transducers and the orifice
plate were analysed. An experiment was conducted in order to investigate the application of the developed
theories and identify the magnitude of the measurement errors, taking into consideration the reverse flow and
temporal inertia of the pulsating flow with relative pulsation amplitudes around 1 and frequencies between 12.5
and 50 Hz. In their work, the authors found that the previously suggested equation may provide acceptable
results if the ratio of the velocity component and the product of the angular frequency and the orifice diameter, is
not greater than 10. For high ratios, the results were not explained by the suggested equation. The authors found
that different theoretical expressions correlating the differential pressure with the mass flow through orifice
plates are found in several studies that differ only quantitatively.

3. THEORETICAL BASIS

In general, the calculation in steady state for an orifice plate consists of finding the orifice diameter from the
required flow, the properties of the fluid, the geometric characteristics of the pipe, and the differential pressure.
Thus, the steady state flow through an orifice plate is obtained from the difference of the pressure measured
before and after the narrowing and is given by Equation (1):

√ (1)
√( )

It is common to use a discharge coefficient, Cd, which takes into consideration all the other factors that
influence the correct determination of the flow. From Equation (1), it follows that each flow has a corresponding
discharge coefficient. From a theoretical point of view, the value of this coefficient is only accurate for the flow
rate used to determine it at the time of calibration. In practice, however, the coefficient obtained covers a range
of flow rates obtained for the orifice plate within the calibrated range, even though small errors can occur. For
the correct determination of the required flow rate, the value of the discharge coefficient, Cd, depends on the
density of the fluid and on the measured pressure difference. Another issue is related to a change in the value of
the discharge coefficient based on the Reynolds number, when the variation range of the system’s pressure is
very large. The type of flow is defined by the Reynolds number in a way that the coefficient will suffer
significant changes, especially when the variation in the Reynolds number implies a change in the flow regime
in the passage (laminar to turbulent or vice versa) since the flow regime change implies a variation in the
velocity profile. When the Reynolds number is very small, the pressure loss through the orifice is primarily due
to internal shear forces resulting from the viscosity of the fluid. If the Reynolds number increases, separation
occurs in the region downstream of the flow and therefore, inertial forces dominate. This makes it impossible to
find a single empirical expression to obtain the discharge coefficients in the laminar flow range. In this range, the
Reynolds number can change sharply and in different ways with varying values of discharge coefficient
(AMERICAN PETROLEUM INSTITUTE, 1985; Mattingly and Yeh, 1994; Morrison et al, 1995; Branch, 1995)
According to Valle (1994) and Gajan et al (1992) for a flow in transient regimes in orifice plates, the expression
used for the calculation of the flow rate can be written in the same way as for steady states, as in Equation (2):
 ( ) ( )
( ) √ (2)
√( )

Thus, experimentally obtaining Q(t) and ∆p(t) allows us to find the dynamic discharge coefficient Cd (t). In
this case, in addition to predicting the influence of geometric parameters and the variation of the flow regime, the
transient parameters must also be taken into consideration. Although the discharge coefficients for flows in
steady state have been accurately measured for Reynolds numbers above 4500, the information on these
coefficients for low Reynolds numbers, notably, lower than 2300 are limited (TRIP et al, 2012).
Since a variation of the flow regime implies a variation of the velocity profile, it can be shown that the variation
of the discharge coefficient when the regime changes is mainly due to a variation of flow, but also to a variation
in the incident velocity profile. Thus, the Reynolds number can greatly alter the values of the discharge
coefficient, depending on the speed range in which it is working. Therefore, any disturbance in the flow that
could change the speed profile and the magnitude of this speed will also change the discharge coefficient and
therefore, the value of the flow rate.
The correct discharge coefficient should be able to fix the theoretical equation to allow for the influence of the
speed profile and the geometric parameters, assuming there is no loss of energy between the pressure
extraction points and their locations. The incident speed profile depends not only on the variation of the
Reynolds number but also on the transient parameters and the presence of any disturbing device upstream of the
orifice. Therefore, it can be stated that a failure to consider these parameters when calculating the discharge
coefficients implies that there will be no reduction in all the possible errors in this type of meter. Generally, due
to the existing difficulty in the location of the installation, only the fixed geometric parameters and the influence
of the Reynolds number are considered during calibration.
Many hydrodynamic factors affect the discharge coefficient of the flow in orifice plates that cause
measurement errors. Although discharge coefficients for flows in steady state have been measured very
accurately for a wide range of conditions, the information on these coefficients for turbulent and transient flows
are rather limited and from an experimental origin (BERREBI; DEVENTER; DELSING, 2004; SANNIKOV;
FEDOROV, 2006; READER-HARRIS; HODGES; RUSHWORTH, 2008; WHITSON; STOBIE, 2009;
GRAVES, 2010; HOLLINGSHEAD, 2011; SHAH et al, 2012).
The equation used to express the flow rate with differential pressure measured through a constraint originates
from the one-dimensional equation of linear momentum. The equation contains a temporal acceleration term and
a term of convective acceleration, and both contribute to the differential pressure. Nevertheless, it is likely that
the convective acceleration of the fluid passing through a constraint is much greater than the temporal
acceleration, considering that the flow fluctuations are not fast.
A mathematical analysis of the theory of flow in quasi-steady state can be made taking into account the
temporal acceleration term (also called temporal inertia). However, the theory does not take into consideration
the possible effects of pulsations on the discharge coefficient at the flowmeter. In the theory of flow in a quasi-
steady state, the discharge coefficient is assumed to be constant and has its value determined under steady state
flow conditions. In fact, it must have a cyclic variation (GAJAN et al, 1992; MOTTRAM, 1992;
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2003; SVETE et al, 2012).
3.1 The theory of the flow in quasi-steady state

This theory says that the relation between the flow rate and differential pressure remains constant during the
pulsation cycle (GAJAN et al, 1992; MOTTRAM, 1992; INTERNATIONAL ORGANIZATION FOR
STANDARDIZATION, 2003; SVETE et al, 2012).
If this is assumed, the equation becomes the relation to the theory of flow in quasi-steady state, which is valid
when applied to an instant of time during pulsation conditions.
In this case, the mass flow rate is given by Equation (3):

̇ ( ) (3)

The average mass flow rate, Equation (4), is given by integrating with respect to time in Equation (3).

̇ ∫ ̇ ( ) ∫ (4)

̇ ( ) (5)

Equation (5) can be applied to pulsating flows, considering flow conditions in a quasi-steady state. The average
mass flow rate is obtained using the average value of√ . However, consideration of flow in quasi-steady state
is valid only for very low frequency pulsations in incompressible fluids. For a better understanding of the
pulsating flow behaviour through orifice plates, it is also necessary to consider the effects of temporal inertia, the
effects of compressibility, and other factors that affect the discharge coefficient (GAJAN et al, 1992;
MOTTRAM, 1992; INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2003; SVETE et al,
2012).

3.2 Theory of the flow in quasi-steady state, including the effects of temporal inertia

To obtain the equation of the instantaneous differential pressure, the equations of linear momentum and
continuity are integrated over the constraint length, assuming the fluid to be incompressible, the constraint length
is short compared with the wavelength of the pulsation, and the temporal acceleration term needs to be included
(GAJAN et al, 1992; MOTTRAM, 1992; INTERNATIONAL ORGANIZATION FOR STANDARDIZATION,
2003; SVETE et al, 2012).
( )
̇ ∫
(6)
( )

The first term, on the right side of Equation (6), is the pressure differential associated with the convective
acceleration term of the fluid through the orifice, which is identical to the expression for the differential pressure
for flow in a steady state. The second term is the differential pressure associated with the temporal acceleration
of the fluid. Its value increases with pulsation rate and is zero for a steady state flow. An expression for the
second term is given by Equation (7) (GAJAN et al, 1992; MOTTRAM, 1992):

∫ ( ) ∫ (7)

̇
After replacing the speed with , Equation (7) can be written:
( )

( ) ̇
( ) ̇ [ ] (8)
( ) ( )

Where Cc is

(9)

If the flow is permanent with the same mass flow, the differential pressure can be found from Equation (10):
( )
̇
(10)
( )

Equation (8) can thus be rewritten as:

( ) ( )
( ) (11)

In Equation (11), the first term on the right side corresponds to the temporal inertia of the fluid between the two
pressure extraction points and the second term is that of the convective inertia, which can be associated with the
discharge coefficient (GAJAN et al, 1992; MOTTRAM, 1992; INTERNATIONAL ORGANIZATION FOR
STANDARDIZATION, 2003; SVETE et al, 2012). The temporal inertia term is a function of dimensionless
frequency, known as the Strouhal number, Srd, and associated with the diameter of the orifice, Equation (12):

(12)

In the theory of flow in a quasi-steady state, the influence of the inertia of the fluid is assumed to be negligible
and the K2 parameter maintains the value of the flow in steady state, which is directly related to the discharge
coefficient (GAJAN et al, 1992; MOTTRAM, 1992; INTERNATIONAL ORGANIZATION FOR
STANDARDIZATION, 2003; SVETE et al, 2012). Thus, at any time during the pulsation cycle, the transient
volume flow is derived from the dynamic differential pressure corresponding to Equation (13):

( )
( ) √ (13)
Where K2 is defined as:
( )
(14)
( )

In the theory of flow in a quasi-steady state including the effects of temporal inertia, the pulsation does
not affect the value of the coefficient K2 during the cycle. However, the term of temporal acceleration is
considered constant, according to Equation 15 (GAJAN et al, 1992; MOTTRAM, 1992; INTERNATIONAL
ORGANIZATION FOR STANDARDIZATION, 2003;SVETE et al, 2012):

(15)

The effective length of the constraint was considered equal to the diameter of the orifice, (GAJAN et al, 1992;
MOTTRAM, 1992; INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2003; SVETE et al,
2012).
The dynamic behaviour of the flow depends on the pulsation frequency and is expressed by the Womersley
number, . The Womersley number is used to characterise the periodic nature of the flow, and represents the
relationship between inertial forces and viscous forces, but emphasises the dependent terms of time instead of
the stationary inertial terms, and is found from equation (16):

√ (16)

The average dynamic discharge coefficient is obtained from Equation (17).

(17)
( ) ∫ ( )

4. EXPERIMENTAL METHODOLOGY

The flow bench consisted of an experimental bench for the study of laminar and turbulent flows that are
incompressible and in a steady or transient pulsating state. The flow measurement system was an orifice plate
type. The flow was studied in an acrylic pipe of small diameter (D <50 mm). A flow pulse generator was used to
create a sinusoidal wave flow around an average flow where pressure transducers with dynamic characteristics
were used to monitor the pressure fluctuations over time. For this study, the propagation of a known flow
fluctuation was related to a pressure fluctuation, allowing dynamic calibration of the measuring system. The
fluctuation of flow around an average flow, also known, was produced by the harmonic motion of a piston. The
variation in pressure across the orifice plate was experimentally determined by pressure transducers attached to a
dynamic data acquisition system.
The general schematic of the static and dynamic system built is shown in Figure 1. The description of the
operation of the flow bench is described with the aid of Figure 1.
The oil contained in tank (l) is pumped by the vane pump (2) at a constant flow rate . The pump flow
rate is controlled by a flow control valve with pressure and temperature compensation (3), which allows
variation of the pump flow rate, , with an average flow rate, but still constant. An accumulator (5), using
nitrogen gas at 4-bar pressure, is responsible for minimising the pulsation caused by the pump in the line. The oil
passes through the acrylic pipe (6) to the static discharge measurement system (7). The purpose of this system is
to confirm that the length of the straight pipe used in the dynamic flow rate measurement system is long enough
to not interfere with the flow values found in a steady state. After passing through the resonator (8), the oil
comes into contact with the flow pulse generator (10) through an existing ramification in the pipeline. The flow
pulse generator (10) is activated by an AC motor (9) and its rotation is changed by a frequency inverter (11).
Then, the oil goes to the dynamic flow rate measurement system (12), consisting of a concentric orifice plate
with sharp edge and three pressure transducers with dynamic characteristics. When the flow pulse generator is
turned off, both measuring systems, (7) and (12), transmit voltage signals corresponding to a differential
pressure value equivalent to the average flow rate, . When the flow pulse generator is activated, the flow rate
measurement system (12) indicates a pressure difference equivalent to an average flow, Q0, increased by a
damped and lagged flow fluctuation that actually passes through the orifice plate, Qa (t). The resonator (8)
associated with the inclination of the tube that carries the oil to the dynamic flow measurement system, aims to
prevent the propagation of this fluctuation to the measuring system (7). The fluctuation generated by the flow
pulse generator Qp (t) has a delay and damping, so that the fluctuation of a damped and lagged flow that actually
passes through the orifice plate is, Qa (t). After passing through the dynamic flow measurement system (12), the
oil passes through a directional valve (13) which can switch the oil flow between the measuring container (14)
and the reservoir (1). In order to calibrate the measurement systems in continuous operation and determinate the
average flow, Q0, a timer (15) is used acting on the aperture of the directional valve, allowing the circulating oil
to enter the container over a given period. The oil temperature in the container is measured by the PT-100 (16).
A heat exchanger (17) is used to keep the oil temperature within the reservoir (l) at the desired temperature.
The flow fluctuations at various frequencies are produced by the harmonic motion of a piston driven by an
electric motor with controlled angular frequency. The flow pulse generator uses a Scotch Yoke mechanism,
capable of generating sinusoidal wave flows around the average flow, Q0. This value is provided by the vane
pump (2) and it is measured by the flow measurement container or the flow measuring system (7). The
fluctuation generated by the flow pulse generator, Qp (t), is produced by the flow pulse generator (10), driven by
the electric motor (9). The transient volumetric flow rate, Q(t), constituted by the floating damped and lagged
flow that actually passes through the orifice plate, Qa(t), and the average flow, Q0, is measured by the dynamic
flow measurement system (12) . A series resonator (8) prevents the propagation of the pressure fluctuation into
the upper orifice plate, which is used to measure the average flow, Q0.
The dynamic flow measurement system (12) is installed at a distance of 590 mm from the flow pulse generator,
LG, and at a distance equal to 315 mm of straight pipe from the sloped tube, LE. This distance influences the
wave flow produced by the flow pulse generator, Qp(t), which is damped and lagged in the orifice plate, Qa(t),
after covering the distance LG.
The experimental methodology consists of directly measuring the transient volumetric flow, Q(t) through an
orifice plate type flow measuring system. In this case, the dissipative effects of fluctuations are implicit in the
values of the transient volumetric flow rate, Q(t), obtained experimentally. The average flow, Q0,is measured
instantly by the static flow measurement system (7).
The fluid used in the flow bench was Mobil DTE 24 hydraulic oil (ISO VG 32). This choice was due to its
common applications, especially in industrial hydraulic systems where pulsating flow situations similar to those
studied in this work are always present. The specific mass () and the kinematic viscosity () of the oil are
857.61kg/m3 and 31.5cSt, respectively. In this study, the oil temperature was maintained at 40  1 ºC.
For the operation of the dynamic measurement system as well as the flow delivered by the pump and the
parameters relating to the orifice plate (dimensions and geometry), it is necessary to know the static differential
pressure (pressure transducer IFM PN 3003 - calibrated range: 0 to 25 bar), the dynamic pressure in the orifice
plate (pressure transducer PCB1501 - calibrated range: 0 to 13.8 bar), the angular frequency of the flow pulse
generator, oil temperature (ECIL PT-100 - calibrated range: 0 ° C to 100 ° C), and the piston position of the flow
pulse generator (Proximity sensor ALLEN-BRADLEY872CD2NN8E2 - calibrated range: 0-2 mm). The
pressure transducers had a manufacturer’s measurement uncertainty equal to ± 0.25% of the final scale value.
Considering that the measured pressure range was between 1.19 and 10.75 bar, and in accordance with the
calibration certificate, the range of uncertainty in the pressure measurement was 5.58% of the measured value for
the lowest pressure and 0 74% of the value measured for the higher pressure. The uncertainty for the
temperature, according to the calibration certificate, was ± 1° C. All errors regarding the pressure and
temperature measurement system were quantified by the calibration of the transducers.
The data acquisition system used was a NATIONAL INSTRUMENTS, model NI USB-6229 with a time
resolution of 50 ns and sampling rate of 250 kS/s.
The measurement of the volumetric flow was performed using a graduated container, in litres. After filling the
container for 10 seconds, the collected oil volume was then measured and the volumetric flow rate was
determined and the oil temperature was measured by a PT-100, installed in the container. The percentage of
uncertainty of the flow measurement container can be estimated to be half the value of a scale division, which is
± 1.2% of average volumetric values used in the tests for the lower flow rate and ± 0.4 % for the higher flow.
Table 1 shows the values of predetermined average flow rates and temperatures for the experiments.
Eight measurements were performed for five flow values, ranging from 0.430 x 10-3 m3/s to 1.250 x 10-3 m3/s,
and for each flow, 11 pulsation frequencies were used in a range between 4 and 24 Hz, with increments of 2 Hz.
The hydraulic bench was composed of a reservoir, a pump with a variable displacement vane, a gas accumulator
(pure N), a flow control valve with pressure and temperature compensation, a heat exchanger, a flow
measurement container, and a directional valve connected to a timer to the filling container. The pump used was
a BOSCH RACINE, mod. WI 53406, Pmáx = 140 bar, Vmáx = 45 cm3/rev., Qmáx = 80 Lpm (with a 1760 rpm
engine). The heat exchanger was able to exchange up to 20 kW and was installed in the return circuit to the
reservoir to stabilise the oil temperature during the tests.
For the dynamic calibration of the orifice plates and the acquisition of the delayed and damping values of the
sinusoidal wave flow, a device is required that provides a known flow and periodic signal. Therefore, a system to
generate flow pulsations was used that was composed of a compressor, an electric motor, and a frequency
inverter. A device known as a scotch yoke was chosen to generate sinusoidal motion of the piston, and the
motion can be represented by Equation (18).

̇( ) ( ) (18)

The fluctuation generated by the flow pulse generator, ( ), is given by the Equation (19):

( ) ̇( ) ( ) (19)

Equation (19) represents the wave in the flow produced by the flow pulse generator, ( ). However, it was
necessary to adjust this pulsation in magnitude and phase at the orifice plate in order to consider the dissipative
effects of the fluctuations that occur in the existing conduit between the pulse generator and the orifice plate.
A three-phase AC electric motor of 30 hp, combined with a frequency inverter (WEG model CFW-06) was used
to allow the speed of rotation to be varied between 120 to 3600 rpm, covering the whole the range of expected
frequencies. The combined uncertainty of the frequency inverter and the electric motor was ± 0.1 Hz for the
entire operating range, equivalent to ± 6 rpm. As the operation range was between 240 and 1440 rpm, or between
4 and 24 Hz, the greatest error was at the frequency of 4 Hz, equivalent to 2.5%.
The imposed transient flow, ( ), which passes through the dynamic flow measurement system orifice plate is
given by the sum of fluctuations generated by the flow pulse generator, ( ), and the average flow, Q0 , is
measured at the orifice plate for the static flow measurement system. Therefore, in Equation (20) there is:

( ) ( ) (20)

The fluctuation generated by the flow pulse generator, ( ), and the average flow, , are values that are
imposed by the system.
From the theory of flow in quasi-steady state, the measurement of transient volumetric flow, ( ), was obtained
experimentally through the dynamic flow measurement system, measuring the transient differential pressure on
the orifice plate, ( ), and using Equation (21).

( ) [ ( )] (21)

This relation is similar to that normally used for flow measurement in steady state. The difference being the time
dependence of flow and pressure.
For flows in small diameter pipes, standard ISO/TR15377 recommends using corner taps and an excellent
finishing of the inner surface of the pipe. For this reason, in addition to using corner taps, acrylic was chosen as
the material used for the pipe, since this material can be very easily polished to produce a high quality finished
surface (INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2007).
The straight sections upstream and downstream of the orifice plate complied with the limits recommended by the
consulted standards.
The orifice plate for the dynamic flow measurement system had a concentric orifice with sharp edge, a diameter
(d) equal to 11.6 mm, with a ratio between the orifice diameter and the diameter of the pipe () equal to 0.53 and
was built according to ISO 5167 (INTERNATIONAL ORGANIZATION FOR STANDARDIZATION,2003a-
b).
The orifice plate was made from stainless steel and was polished until its surface was mirrored on the upstream
side. All other specifications listed in the standard, such as the minimum thickness and diameter tolerances were
respected.
The calibration of the orifice plate in a steady state consisted of an experimental determination of a transfer
function that relates the flow rate to be measured with the pressure difference measured at the orifice plate. The
pressure transducers provide current signals and / or voltages in response to the deformation imposed by the
fluid. These were calibrated before calibrating the orifice plate.
After determining the system transfer function, the calibration of the orifice plate in a steady state was
performed. With the flow pulse generator disabled, an average flow, , was imposed on the system, using
software to select a time of 12s to collect the fluid, according to Table 1.
To determine the volumetric flow rate, eight measurements were taken under identical conditions. A maximum
systematic error corresponding to half the value of a scale division was considered. Since the minimum scale
division was 0.1 litres, the maximum constant systematic error is 0.05 litres for each reading.
Therefore, the function that relates the average flow, Q0, to the pressure difference across the orifice plate can be
obtained.
By using Equation (1), the discharge coefficient for the steady state using the Bernoulli equation can be
determined, leading to Equation (22):
√( ) ( )
(22)
√ √

The ISO/TR 3313 standard suggests the use of Equation (23) for calculating the error due to the presence of
pulsation in the flow, assuming there are no errors in the secondary instrumentation (INTERNATIONAL
ORGANIZATION FOR STANDARDIZATION, 2003).

( ) (23)

The calculations of the results’ uncertainties obtained in the experimental tests were referenced in ISO 5167
(INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2003a-b).
All errors regarding pressure and temperature measurement system were quantified by calibrating the
transducers. In the development of the experimental part of this work at least 8 measurements were made,
requiring an analysis of the uncertainties about the obtained results.
In order to develop such analysis, the measurement systems (metrological characteristics), the procedures
adopted to perform measurements and data treatment, and the action of the most significant influence quantities
were evaluated.
The discharge coefficients for various types of primary elements were experimentally determined through a large
number of tests, and then presented in the standard as equations and graphs. The test data allows the statistical
determination of the discharge coefficient uncertainty.
The uncertainty of the discharge coefficients was calculated by the equation:

( ) √( ) ( ) ( ) ( ) ( ) ( ) ( )

The terms from the uncertainty of the expansion factor, ( ), of the specific mass uncertainty, ( ), the

uncertainty of orifice diameter ( ), and the uncertainty of the pipe diameter( ) were not considered.

In order to assure that the average flow, measured by the static flow measuring system wouldn’t be influenced by
the pulsating created by the flow pulse generator, measurements were performed under identical conditions
before and after each set of dynamic tests. The flow measured by the static flow measurement system during the
dynamic tests were not different from those obtained in steady state, so that the pulsating influenced only the
flow rate measured by the dynamic flow measurement system. The error due to the presence of pulsation in the
average flow was calculated through Equation (23). The maximum error calculated was 3.9%, for the lowest
average flow and lower pulse rate, and 1.02% for the higher average flow and higher pulse rate.
The errors increased as the pulse rate increased and decreased as the average flow increased within the analyzed
flow range. For higher average flows and, consequently, for higher pressure levels, the total errors obtained
begin to present a decreasing behavior for higher frequencies.
The Reynolds number can greatly change the values of the discharge coefficient, depending on the speeds range.
Therefore, any flow disturbance that changes the velocity profile and velocity magnitude will modify the
coefficient.
According to ISO / TR 3313, the pulsation generates errors in the flow measured by differential pressure systems
because the average gradients of static and dynamic pressure are different. Systems whose secondary elements
have a slow response to the flow variation are less susceptible to the errors caused by the average variation of the
dynamic differential pressure. However, such devices are unable to measure instantaneous flow since the time
response normally recommended in the standards is at least ten times the value of the fluctuation period.
(INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 1998).

5. RESULTS

In the literature, there are few studies related to the discharge coefficient for fluids that are more viscous and low
Reynolds numbers, especially in laminar flow conditions. The discharge coefficient values shown in the
consulted literature and standards are for high Reynolds numbers and turbulent flow conditions at extraction
points between D-D/2 and at the flanges, but not for corner tap extraction points.
In steady state flows, the discharge coefficient is defined based on the Bernoulli equation, and the experimental
determination of the discharge coefficients is performed under static conditions and high Reynolds numbers. The
discharge coefficients from all different types of primary meters are dependent on the speed profile for a given
flow rate. The orifice plate tends to be particularly sensitive to variations in the speed profile because of the
contraction effect. The discharge coefficient of the concentric orifice plate with sharp edge did not show great
variation for the range of Reynolds numbers used. According to the consulted standards, a concentric orifice
plate with sharp edges can be used in pipes with diameters between 25 and 50 mm, as long as some special care
is taken regarding the roughness of the pipe, the finishing of the orifice plate, the centralisation of the plate in the
pipe, and the position and geometry of the pressure extraction points. As a result, the use of the concentric orifice
plate with sharp edge for measurement of high viscosity fluids, such as oil, was satisfactory.
The Reynolds numbers, which range from 1498 to 4356 when based on the orifice diameter or from 794 to 2308
when based on the diameter of the pipe for which the experimental data was obtained, are small compared with
the range of data from the consulted literature and standards.
For the transient regime, several curves were determined in order to characterise the behaviour of the pulsating
flow measurement. These curves show the variation of pressure in the dynamic flow measurement system caused
by the flow pulse generator, upstream and downstream the orifice plate.
From the experimental pressure values, dynamic flow curves were generated using Equation (20) for an average
flow rate of 1.020 x 10-3 m3/s and 10 Hz, as shown in Figure 2.
Eight measurements were performed for five pressure values between 0.430 x 10-3 m3/s and 1.250 x 10-3 m3/s
and for each flow rate, 11 pulsation frequencies were chosen in a range between 4 and 24 Hz, with 2 Hz
increments.
The behaviour of the Strouhal number as a function of Reynolds number for the entire frequency range
used is shown in Figure 3. From the figure, it can be observed that the Strouhal number presented a downward
trend in relation to the Reynolds number, as the inertia effects become significant due to variations in flow
velocity because the temporal acceleration term becomes more important in relation to convective acceleration
term.

The dynamic behaviour of the flow depends on the pulsation frequency expressed by the Womersley number,
which is used to characterise the periodic nature of the flow, and represents the relation between the inertial
forces and the viscous forces. With an increased pulsation frequency, an increase in the Womersley number is
expected, so that the inertial forces become dominant compared to the viscous forces, as shown in Figure 4. The
Womersley number was between 9.75 and 23.95, a range in which inertial forces are present (GÜNDOĞDU;
YILMAZ, 2010; ALFREDSSON et al, 2012).
Figures 5 and 6 show flow curves obtained at the concentric orifice plate with a sharp edge with an average flow
of 0.810 x 10-3 m3 / s and the corresponding curve of the imposed transient flow. From Figure 5, it can be seen
that for a frequency of 4 Hz, the input signal is very close to the flow on the plate, showing that the wave is in
phase and with little damping at the orifice plate. By comparing Figure 6 (for a frequency of 16Hz) to Figure 5,
it can clearly be observed that damping of the wave also increases at a higher frequency. When the average flow
rate is increased, the corresponding increase in the system pressure makes the experimentally generated pulse
more stable.
Figure 7 shows the efffects of pulsation on the dynamic discharge coefficient and the value corresponding
to an average flow of 0.810 x 10-3 m3/s and 6 Hz.
All results presented hereinafter refer to the average dynamic discharge coefficient obtained from Equation (17).
Figure 8 shows the average dynamic discharge coefficient behaviour for some experimental frequency ranges
and studied flow rates. The average dynamic discharge coefficient decreased with an increase in both the
frequency and the flow rate compared to the static discharge coefficient. It can also be observed that the average
dynamic discharge coefficient for all the flow rates is lower for the highest pulsation frequencies. With an
increase in flow rate, the average dynamic discharge coefficient decreases due to the increase in the differential
pressure and due to the pressure drop corresponding to the acceleration of the fluid. When the flow rate
decreases, the average dynamic discharge coefficient increases due to the recovery of the pressure resulting from
the deceleration of the fluid. The pressure difference is a result of the energy required to accelerate and
decelerate the fluid through the orifice.

Figure 9 shows the discharge coefficients for flow in steady state (static coefficient), for the theory of flow
in quasi-steady state (quasi-constant coefficient) and quasi-steady state including the effects of the temporal
inertia (dynamic coefficient) for a given pulsation rate and varying average flows. The inertial effects are not
taken into consideration when determining the static discharge coefficient, and its value is the highest of the
three, Equation (22). In addition, the quasi-constant constant discharge coefficient is shown, i.e. the one that
considers the flow equation related to pressure, Equation (13) for steady state but with both flow rate and
pressure terms varying depending on time. Finally, the average dynamic discharge coefficient is shown, in which
the inertial terms are taken into consideration, and therefore, its value is the lowest of the three, Equation (11).

The uncertainty was ± 0.025 for the lowest flow rate and frequency and ± 0.011 for the highest flow rate
and frequency for the average dynamic discharge coefficient using the theory of flow in quasi-steady state. The
uncertainty was ± 0.028 for the lowest flow rate and frequency and ± 0.015 for the highest flow rate and
frequency for the average dynamic discharge coefficient for the theory of flow in quasi-steady state including the
effects of the temporal inertia.
In the flow rate range between 0.430 x 10-3 m3/s and 1.250 x 10-3 m3/s and pulsation frequency range between 4
and 24 Hz, the uncertainty was ±0.028 for the lowest frequency and flow rate, corresponding to 3.9%, and the
uncertainty was ±0.015 for the highest frequency and flow rate, corresponding to 2.2%.
According to the ISO/TR 3313 standard, pulsation generates errors in the measured flow rate by differential
pressure measurement systems, because the average gradients of static and dynamic pressure are different
(INTERNATIONAL ORGANIZATION FOR STANDARDIZATION, 2003; BERREBI; DEVENTER;
DELSING, 2004; NABAVI; SIDDIQUI, 2010).
In order to ensure that the average flow, Q0 , measured by static flow measurement system was not influenced by
the pulsation created by the flow pulse generator, measurements were performed under identical conditions
before and after starting each round of dynamic tests. The flow rate measured by the static flow measurement
system during the dynamic tests were not different from those obtained in a steady state, so that the pulsation
only influenced the flow measured by dynamic flow measurement system. The error due to the presence of
pulsation in the average flow was calculated using Equation (23). The maximum error calculated was 3.9% for
the lowest average flow rate and lowest pulsation frequency and 1.02% for the highest average flow rate and
highest pulsation rate.
The flow in a concentric orifice plate with sharp edge was analysed under conditions not yet presented in
the consulted literature, primarily between pulsation frequencies from 4 to 24 Hz and flow rates between 0.430
x10-3 and 1.250 x 10-3 m3/s for low Reynolds numbers and more viscous fluids. In the consulted literature, there
are results for frequencies below 10 Hz for incompressible fluids and above 50 Hz using air as the fluid.
The Reynolds number range for which the experimental data were obtained is small compared with the range of
data used in the consulted literature and standards.
The biggest gap in knowledge about pulsation effects in the orifice plate flowmeters is the variation of the
discharge coefficient with regard to the various pulse parameters in a given flow system.
The practical approximation for calculating the flow in pulsating flow conditions consists of using a constant
value for the discharge coefficient, preferably the same value used in steady-state flow conditions.
For flows with low Reynolds numbers, even in steady state, there is no standardisation, and most studies
found in the literature present results for very specific flow conditions.
In the consulted standards, there is no reference to dynamic discharge coefficients. The results found in the
literature are very limited. Uncertainty coefficients are presented, which correct the flow values obtained by
conventional measurement systems due to the influence of pulsation in the line. These coefficients are based on
average squared values of differential pressure and flow velocity as well as on the pulsation frequency.
The dynamic behaviour of the flow depends on the variations related to the inertia of the fluid and the pulsation
effects that can alter the pressure gradient and the velocity profile. The friction (viscous forces) tends to decrease
the discharge coefficient, as well as the inertial forces. In the steady state flow, only viscous forces tend to
influence the determination of the discharge coefficient, while in a pulsation situation, both forces due to inertia
and pulsation of the fluid must be considered. It can be seen that these forces tend to decrease the discharge
coefficient, resulting in an increase in the total resistance in order to reduce the flow rate.
One of the main assumptions of the quasi-steady state theory is that the discharge coefficient can be constant and
have the same value as under steady state flow conditions. The theory does not take into consideration the
possible effects of pulsations on the discharge coefficient of the flowmeter. From the results shown, it can be
observed that this consideration can present conflicting results.
The influence of pulsation using the flow theories, particularly the discharge coefficient, was analysed in this
study. The main result of the investigation was that, under the influence of dynamic effects, the dynamic
discharge coefficient, ( ), progressively decreased from its value of the static discharge coefficient over the
discharge range and the investigated frequencies, and it was also dependent on the pulsation frequency.
With an increased pulsation frequency, the Womersley and Strouhal numbers increased as the inertial forces
become dominant compared to viscous forces, resulting from the temporal acceleration term becoming more
important compared to the convective acceleration term.
For a correct measurement of the transient pressure, not only the pressure transducer must have a fast response,
but so must the entire system consisting of pressure transducers and pipes connecting the transducers with the
orifice. The lengths of the connecting tubes must be short compared to the length of a quarter of the wave pulse
in order to preclude resonances that can seriously distort the indication of the pressure transducers. A method for
reducing the effects of the resonance is the placement of electronic filters during the acquisition of the pressure
signals, not eliminating the pulsations, but reducing the effects of pulsations when receiving data from the
transducers.

6. CONCLUSIONS

In the literature, there are few studies related to the discharge coefficient for more viscous fluids with low
Reynolds numbers, especially in laminar flow conditions. The discharge coefficient values shown in the
consulted literature and standards are for high Reynolds numbers for turbulent flow for extraction points D-D/2
and for extraction points on flanges, and not for corner tap extraction points.
For high Reynolds values in a steady state, the discharge coefficient for the concentric orifice plate with sharp
edge is set based on the Bernoulli equation, which is quasi constant. However, for the measurement of high
viscosity fluids and transient pulsation state for low Reynolds values, there is no standardisation regarding the
use of concentric orifice plate with sharp edge, not even in a steady state, and most of the work found in the
literature presents results for very specific flow conditions. Using the value for the static discharge coefficient in
situations of transient pulsating flow results in significant errors.
The average dynamic discharge coefficient varies under the influence of the pulsation frequency, the differential
pressure as well as the Reynolds, Strouhal and Womersley numbers should not be considered constant and
having the same value as under steady state flow conditions. Since the variations in the flow velocity profile
affect the pressure difference across the orifice, it is important to measure them in order to make it possible to
examine their influence on the dynamic discharge coefficient and consequently, on the flow value.
The dynamic discharge coefficient of an orifice plate in a pulsating flow is different from the static discharge
coefficient and must be taken into consideration for flow measurement.
 The static discharge coefficient of the concentric orifice plate with sharp edge did not show great
variation for the range of Reynolds numbers used nor for a pipe with a diameter smaller than
recommended by the consulted standards, as long as special care is taken regarding the roughness of
the tube, the finishing of the orifice plate, the centralisation of the plate in the pipe, and the position
and geometry of the pressure extraction points.
 With an increase of pulsation frequency, an increase in the Womersley and Strouhal numbers occurred
as the inertial forces become prevalent in relation to the viscous forces. This is due to variations in the
flow velocity because the temporal acceleration term has become more important in relation to the
convective acceleration term. When the Womersley and Strouhal numbers were small, the effects of
the temporal inertia were neglected. When the frequency increased, the inertial term became significant
and could not be eliminated, thus having an important role on the value of the dynamic discharge
coefficient.
 Increased frequency caused a decrease in the dynamic discharge coefficient because the transient
differential pressure across the orifice plate increased. With an increase of flow rate, the dynamic
discharge coefficient decreased due to the increase of the differential pressure and due to the pressure
drop corresponding to the acceleration of the fluid. When the flow rate decreased, the discharge
 coefficient increased due to the recovery of pressure resulting from the deceleration of the fluid. The
pressure difference was due to the energy required to accelerate and decelerate the fluid through the
orifice.
 The dynamic discharge coefficient gradually decreased from its value of static discharge coefficient
when the frequency increased over the investigated frequency range. With an increase of pulsation
frequency, more energy was required to accelerate and decelerate the fluid, resulting in a decrease in
the energy necessary to move the fluid through the orifice.
 The biggest coefficient of all was the static coefficient, because it did not take into consideration the
inertial effects. The quasi-steady state was next largest, which considered the flow rate equation as a
function of pressure for steady state, but with the terms of both flow and pressure varying in time.
Finally, the dynamic coefficient, which considered the inertial terms, had the lowest value among the
three.
NOMENCLATURE
= sectional area of the flow pulse generator cylinder (m2);
= 3.46x10-4 m2;
= constant obtained from the calibration of the orifice plate;
= sectional area of the orifice plate (m2);
= static discharge coefficient;
= contraction coefficient;
d = diameter of the orifice plate (m);
= Error due to the presence of pulsation in the flow;
= pulsation frequency (Hz);
= constant obtained from the calibration of the orifice plate;
= effective length of the constraint (m);
̇ = mass flow rate (kg/s);
( ) = pressure upstream of the orifice plate (Pa);
( ) = pressure downstream of the orifice plate (Pa);
( ) = transient volumetric flow rate (m3/s);
( ) = fluctuation generated by the flow pulse generator (m3/s);
= average flow rate, measured in a steady state (m3/s);
= volumetric flow rate (m3/s);
r = radius of the handle eccentric of the flow pulse generator (m);
= 4.91x10-3 m;
= internal radius of the tube (m);
= time (s);
= time range (s);
= initial time (s);
= end time (s).
 = angular frequency (rad/s);
= angular position (angle which the handle is located) (rad);
= relation between the diameter of the orifice and the diameter of the pipe;
= fluid expansion factor. For incompressible fluids,  = 1;
= specific mass of the fluid (kg/m3);
= Dynamic viscosity (kg/m.s);
= differential pressure at the orifice plate (Pa);
( ) ( ) ( );
( ) = transient differential pressure at the orifice plate (Pa);
= average differential pressure measured during pulsation (Pa);

= average differential pressure measured under static conditions (Pa);

= average speed at the orifice (m/s);
ACKNOWLEDGMENTS
The authors thank the Brazilian National Council for Scientific and Technological Development (CNPq) and the
Minas Gerais State Research Support Foundation (FAPEMIG) for their financial support of this project.

REFERENCES

ADDISON, P. S. et al. An experimental investigation into the breakdown of low Reynolds number pulsed flows
at a pipe orifice. Journal of Fluids Engineering, v.119, n.2, p. 347-353, June 1997.

ALFREDSSON, P. H. et al. A flow facility for the characterization of pulsatile flows. Flow Measurement and
Instrumentation, v.26, p. 10-17, 2012.

ALPAY, S. A. The flow coefficient of a hydraulic control valve under dynamic conditions, 1971, University of
Waterloo. p. 1227-1234.

AMERICAN GAS ASSOCIATION. Pulsation effects on orifice meters. Presented at the Operations Conference.
Chicago, Illinois, May, 2002.

AMERICAN PETROLEUM INSTITUTE. API 2530: Orifice metering of natural gas and other related
hydrocarbon fluids. 2nd ed. Washington: API, 1985.

AMERICAN PETROLEUM INSTITUTE. API MPMS 14.3.1: Manual of petroleum measurement standards -
Chapter 14 - Natural Gas Fluids Measurement - Section 3 - Concentric, Square-Edged Orifice Meters. 3th ed.
reaffirmed. Washington: API, 2003.

BAJURA, R. A.; JONES, E. H. A numerical analysis of pulsating laminar flow through a pipe orifice. Journal
of Fluids Engineering, v.113, n.2, p. 199-205, June 1991.

BEAULIEU A. et al. A flowmeter for unsteady liquid flow measurements. Flow Measurement and
Instrumentation, v.22, n.2, p. 131-137, 2011.

BERREBI, J.; DEVENTER, J. van; DELSING, J. Reducing the flow measurement error caused by pulsations in
flows. Flow Measurement and Instrumentation, v.20, p. 1-5, May 2004.
BOTROS, K. K.; JUNGOWSKI, W. M.; PETELA, G. Gauge line effects and DP transmitter response to flow
pulsation through orifice plate. Flow Measurement and Instrumentation, v.3, n.3, p. 130-144, 1992.

BRANCH, J. C. The effects of an upstream short radius elbow and pressure tap location on orifice discharge
coefficients. Flow Measurement and Instrumentation, v.6, n.3, p. 157-162, July 1995.

BRENNAN, J. A. et al. The influence of swirling flow on orifice and turbine flowmeter performance. Flow
Measurement and Instrumentation, v.1, n.1, p. 5-8, Oct. 1989.

BURDEN, T. L.; MCLAURY, B. S. Laser Doppler velocimeter measurements to characterize turbulence in a

constriction with sharp and rounded inlets. Experiments in Fluids, v.32, n.4, p. 472-480, 2002.

DELMÉE, G. J. Manual de medição de vazão. 3. ed. rev. e atual. São Paulo: Edgard
Blücher, 2003. 366 p.(In portuguese).

DOBLHOFF-DIER, K. et al. Time resolved measurement of pulsating flow using orifices. Flow Measurement
and Instrumentation, v.22, n.2, p. 97-103, Apr. 2011.

DURST, Franz et al. Method for defined mass flow variations in time and its application to test a mass flow rate
meter pulsating flows. Measurement Science and Technology, v.18, p. 790-802, 2007.

FRAWLEY, P.; GRACE, Andrew. Experimental parametric equation for the prediction of valve coefficient for
choke valve trims. International Journal of Pressure Vessels and Piping, v.88, p. 109-118, 2011.

GAJAN, P. et al. The influence of pulsating flows on orifice plate flowmeters. Flow Measurement and
Instrumentation, Toulouse, v.3, n.3, p. 118-129, 1992.

GRAVES, D. Effects of abnormal conditions on orifice measurement accuracy. Pipeline & Gas Journal, v.237,
n.7, p. 1-4, July 2010.

GÜNDOĞDU, Mehmet Y.; YILMAZ, Fuat. Experimental and computational investigation of velocity field for
intermediate region of laminar pulsatile pipe flow. Journal of Thermal Science and Technology, v.30, n.2, p.
21-35, 2010.

HOLLINGSHEAD, Colter L. Discharge coefficient performance of Venturi, standard concentric orifice

plate, v-cone, and wedge flow meters at small Reynolds numbers. 2011. 93f. Thesis (Master of Science) –
Utah State University, Civil and Environmental Engineering, Logan, Utah.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. ISO 2186: Fluid flow in closed conduits -
Connections for pressure signal transmissions between primary and secondary elements. 2nd ed. Switzerland:
ISO, 2007.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. ISO 5167-1: Measurement of fluid flow

by means of pressure differential devices inserted in circular cross-section conduits running full – part 1: General
principles and requirements. 2nd ed. Switzerland: ISO, 2003a.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. ISO 5167-2: Measurement of fluid flow

by means of pressure differential devices inserted in circular cross-section conduits running full – part 2: Orifice
plates. Switzerland: ISO, 2003b.
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. ISO/TR 15377: Measurement of fluid
flow by means of pressure-differential devices – Guidelines for the specification of orifice plates, nozzles and
Venturi tubes beyond the scope of ISO 5167. Switzerland: ISO, 2007.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. ISO/TR 3313: Measurement of fluid

flow in closed conduits – Guidelines on the effects of flow pulsations on flow-measurement instruments. 3th ed.
Switzerland: ISO, 1998.

JANKOWSKI, T. A. et al. A series pressure drop representation for through orifice tubes. Journal of Fluids
Engineering, v.130, n.5, p. 1-7, May 2008.

JUNGOWSKI, W. M.; WEISS, M. H. Effects of flow pulsating on a single-rotor turbine meter. Journal of
Fluids Engineering, v.118, n.1, p. 198-201, Mar. 1996.

LEE, T. S. et al. Numerical study on sinusoidal fluctuated pulsatile laminar flow through various constrictions.
Communications in Computational Physics, v.2, n.1, p. 99-122, Feb. 2007.

MAJID, N.; KAMRAN, S. A critical review on advanced velocity measurement techniques in pulsating

Flows. Measurement Science and Technology. 21 (2010) 042002 (19pp). doi:10.1088/0957-0233/21/4/042002

MANSHOOR, B.; NICOLLEAU, F. C. G. A.; BECK, S. B. M. The fractal flow conditioner for orifice plate
flow meters. Flow Measurement and Instrumentation, v.22, p. 208-214, Feb. 2011.

MATTINGLY, G. E.; YEH, T. T. Pipeflow downstream of a reducer and its effects on flowmeters. Flow
Measurement and Instrumentation, Gaithersburg, v.5, n.3, p. 181-187, July 1994.

McKEE, R. J. Pulsation effects on orifice metering considering primary and secondary elements. Gulf Coast
Measurement Short course, Houston, Texas, Southwest Research Institute, p. 1-10, September 2009.

METWALLY, M. Review of compressible pulsating flow effects on system performance. In: 13th
INTERNATIONAL CONFERENCE ON AEROSPACE SCIENCES & AVIATION TECHNOLOGY, ASAT-
13, 2009, Cairo, Egypt. p. 1-21.

MILLER, Michael R. Minimizing the effects of pulsation. Appalachian Gas Measurement Short Course, p.
125-131, 2003.

MILLER, Richard W. Flow Measurement Engineering Handbook, 3rd ed. New York, United States of
America: McGraw-Hill Publishing Company, 1996.1168 p.

MOHAMMAD, W. A.; MOTTRAM, R. C. High frequency pulsating effects on orifice meter accuracy. In:
INTERNATIONAL CONFERENCE ON ADVANCES IN FLOW MEASUREMENT TECHNIQUE, University
of Warwick, BHRA Fluid Engineering, 1981.

MORRISON, G. L. et al. Beta ratio, axisymmetric flow distortion and swirl effects upon orifice flow meters.
Flow Measurement and Instrumentation, v.6, n.3, p. 207-216, July 1995.

MOTTRAM, R. C. Measuring pulsating flows with a differential pressure meter. In: 2nd INT. SYMP. ON
FLUID FLOW MEASUREMENT, 1981, St. Louis, v.2, p. 347-361.

MOTTRAM, R. C. The behaviour of orifice and Venturi-nozzle meters in pulsating flow. 1971. 266f. Thesis
(Doctor of Philosophy) – University of Surrey, Faculty of Engineering, Department of Mechanical Engineering,
Guildford.
MOTTRAM, R. C. The measurement of pulsating flows using orifice plate meters. In: FLOW, ITS
MEASUREMENT AND CONTROL IN SCIENCE AND INDUSTRY, 1974, Pittsburgh, v.1, p. 197-208.

MOTTRAM, R. C.; RAWAT, M. S. The swirling damping properties of pipe roughness and the implications for
orifice meter installation, proceedings of flows measurement. In: THE MID 80’s INTERNATIONAL
CONFERENCE, 1986, National Engineering Laboratory, Scotland. p. 1-29.

MURAKAMI, M.; KITO, O. Effects of swirling flow component on discharge coefficients of constriction
meters in case of orifice. Japan Society of Mechanical Engineers, v.23, n.175, p. 34-42, Jan. 1980.

NABAVI, Majid; SIDDIQUI, Kamran. A critical review on advanced velocity measurement techniques in
pulsating flows. Measurement Science and Technology, v.21, p. 1-19, 2010.

PEDROSO, J. L. Estudo experimental da atenuação de ondas orientado aos transitórios ocasionados pela reação
explosiva sódio-água em reatores rápidos, In: 60 SIMPÓSIO BRASILEIRO SOBRE TUBULAÇÕES E VASOS
DE PRESSÃO, 1990, Rio de Janeiro, n. 18, p. 303-321.

PEDROSO, J. L. Experimental investigation of the pressure loss through perforated structures in unsteady flow.
Journal of Structural Mechanics, VTT, Finland, v.21, n.1, p. 42-54, 1988.

READER-HARRIS, M.J., McNAUGHT J., Impulse lines for differential-pressure flowmeters best practice
guide.NEL, East Kilbride, Glasgow(2005).

READER-HARRIS, M.; HODGES, G.; BARTON, N. The effect of contaminated orifice plates on the discharge
coefficient. Flow Measurement and Instrumentation, v.25, p. 2-7, June, 2012.

READER-HARRIS, M.; HODGES, D.; RUSHWORTH, Ronnie. The effect of drain holes in orifice plates on
the discharge coefficient. In: 26th INTERNATIONAL NORTH SEA FLOW MEASUREMENT WORKSHOP,
Oct. 2008, p. 1-19.

RIBAS, F. A. J. Análise Numérica de Escoamentos Transientes em Compressores. 2004.

122f. Dissertação (Mestrado em Engenharia Mecânica) - Universidade Federal de Santa
Catarina, Programa de Pós-Graduação em Engenharia Mecânica, Florianópolis.(In
portuguese).

SANNIKOV, M. S.; FEDOROV, A. V. Study of pressure drop pulsations in a standard orifice device of
measurements systems for natural gas. Measurement Techniques, v.49, n.6, p. 592-598, 2006.

SHAH, Manish S. et al. Analysis of flow through an orifice meter. Chemical Engineering Science, v.71, p.
300-309, 2012.

STROLL, H. et al. Study of laminar, unsteady piston-cylinder flows. Journal of Fluids Engineering, v.115, n.4,
p. 687-693, Dec. 1993.

SVETE, A. et al. Development of a liquid flow pulsator. Flow Measurement and Instrumentation, v.23, p. 1-
8, 2012.

TRENGROUSE, G. H.; IMRIE, B. W.; MALE, D. H. Comparison of unsteady flow discharge coefficients for
sharp-edged orifices with steady flow values. Journal of Mechanical Engineering Science, v.8, n.3, p. 322-
329, Sept. 1966.
TRIP, R. et al. An experimental study of transitional pulsatile pipe flow. American Institute of Physics -
Physics of fluids, v.24, p. 1-17, Jan. 2012.

ÜNSAL, Bülent; DURST, Franz. Pulsating flows: experimental equipment and its application. Japan Society of
Mechanical Engineers, v.49, n.4, p. 980-987, 2006.

VALLE, Ramón Molina. Escoamento laminar em placas de orifício: análise teórica e experimental em regime
permanente e transiente. 1994. 250f. In Portuguese. Tese (Doutorado) - Universidade Federal de Santa Catarina,
Curso de Pós-Graduação em Engenharia Mecânica, Florianópolis.(In portuguese).

WASHIO, S. et al. Study of unsteady orifice flow characteristics in hydraulic oil lines. Journal of Fluids
Engineering, v.118, n.4, p. 743-748, 1996.

WHITSON, Richard J.; STOBIE, Gordon J. An overview of non-invasive flow measurements methods. In: THE
AMERICA WORKSHOP, Feb. 2009. p. 1-23.

WRIGHT, S. J; WYLIE, E. B; TAPLIN, L. B. Matched impedance to control fluid transients. Journal of Fluids
Engineering, v.105, n.2, p. 219-224, June 1983.

YAKHOT, A.; GRINBERG, L.; NIKITIN, N. Simulating pulsatile flows through a pipe orifice by an immersed-
boundary method. Journal of Fluids Engineering, v.126, p. 911-918, 2004.

YOKOTA, S.; NAKANO, K. On extending the intelligent potential of a cylindrical choke type instantaneous
flow meter. In: 2nd INTERNATIONAL CONFERENCE ON FLOW MEASUREMENT, 1988, London, UK. p.
35-46.

ZANKER, K.; GOODSON, D. Qualification of a flow conditioning device according to the new API 14.3
procedure. Flow Measurement and Instrumentation, v.11, p. 79-87, Jan. 2000.

Table 1 – Pre-established average flows and temperatures for the

experiments

Flow x 10-3 ReD based on the Red based on the Temperature (o C)

(m3/s) Pipe diameter orifice diameter
0.430±0.005 40 ± 1
794 1498
0.620±0.005 40 ± 1
1145 2160
0.810±0.005 40 ± 1
1496 2822
1.020±0.005 40 ± 1
1883 3554
1.250±0.005 2308 4356 40 ± 1

Highlights
 Effects of pulsation variations and temporal inertia on the discharge
coefficient of an orifice plate were studied.

 Inertial effects significantly affect the value of the discharge

coefficient of the orifice plate, reducing the discharge values.

 Dynamic discharge coefficients were compared with the

conventional static discharge coefficient, revealing large differences.

 Equation models have been formulated, and were capable of

capturing the observed effects.
Figure 1

1. Tank 11. Frequency inverter

2. Pump with variable displacement vanes 12. Dynamic flow measurement system
3. Flow control valve 13. Directional valve
4. Manometer 14. Container for flow measurement Qo
5. Accumulator 15. Timer
6. Acrylic pipe 16. PT-100
7. Static flow measurement system Qo 17. Heat exchanger
8. Resonator 18. Proximity sensor
9. Electric engine 19. Data acquisition system
10. Flow pulse generator SP1 to SP7 – Pressure transducers

Figure 1 – General schematic of the flow bench

Figure 2

1.200E-03

1.000E-03

Flow (m3/s) 8.000E-04

6.000E-04
Almost permanent
transient flow
4.000E-04
Imposed transient
2.000E-04 flow

0.000E+00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (s)

Figure 2 – Flow rate versus time for the average flow rate of 1,020 x 10-3 m3/s
(Red = 3554) and 10 Hz – EXP.
Figure 3

0.080
0.070 f = 4 Hz

0.060 f = 8 Hz
0.050
Strouhal
f = 12 Hz
0.040
f = 16 Hz
0.030
0.020 f = 20 Hz

0.010 f = 24 Hz
0.000
0.00 1000.00 2000.00 3000.00 4000.00 5000.00
Reynolds

Figure 3 – Strouhal number versus Reynolds number – EXP.

Figure 4

30.00

25.00

20.00

Womersley 15.00

10.00

5.00

0.00
2 4 6 8 10 12 14 16 18 20 22 24
Frequency (Hz)

Figure 4 – Womersley number versus frequency – EXP.

Figure 5

1.200E-03
Almost permanent
transient flow
1.100E-03
Imposed transient
flow
Flow (m3/s) 1.000E-03

9.000E-04

8.000E-04

7.000E-04

6.000E-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (s)

Figure 5 – Flow rate versus time for an average flow of 0.810 x 10-3 m3/s
(Red = 2822) and 4 Hz – EXP.
Figure 6

1.200E-03
Almost permanent
transient flow
1.100E-03
Imposed transient
flow
1.000E-03
Flow (m3/s)

9.000E-04

8.000E-04

7.000E-04

6.000E-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Time (s)

Figure 6 – Flow rate versus time for an average flow of 0.810 x 10-3 m3/s
(Red = 2822) and 16 Hz – EXP.
Figure 7

0.6774

0.6773

0.6772

Discharge
Cd (t)

0.6771
coefficient
Mean value
0.677

0.6769

0.6768
1.55 1.6 1.65 1.7 1.75 1.8 1.85 1.9 1.95 2 2.05
Time (s)

Figure 7 – Behaviour of the dynamic discharge coefficient for an average flow of

0.810 x 10-3 m3/s (Red = 2822) and 6 Hz – EXP.
Figure 8

0.775
0.765
0.755
0.745
0.735
0.725
Static Cd
Cd (t)

0.715
0.705 Dynamic Cd (t) - INERTIAL
0.695
Dynamic Cd (t) - ALMOST
0.685 PERMANENT
0.675
0.665
0.655
1.400E+03 2.400E+03 3.400E+03 4.400E+03
Red

Figure 8– Average dynamic and static discharge coefficient versus flow rate for
frequencies – EXP.
Figure 9

0.775
0.765
0.755
0.745
0.735
0.725 Static Cd
Cd (t)

0.715
Dynamic Cd (t) - INERTIAL
0.705
Dynamic Cd (t) - ALMOST
0.695
PERMANENT
0.685
0.675
0.665
0.655
1.400E+03 2.400E+03 3.400E+03 4.400E+03
Red

Figure 9 – Average dynamic and static discharge coefficient versus flow rate for 8
Hz for the theories of flow in quasi-steady state and flow in quasi-steady state
including the effects of temporal inertia – EXP.