Theoretical background

Chapter Abstract
This chapter provides an overview of the theoretical background that is necessary for this thesis. A
description of flowmeter calibration systems leads step by step to the presentation of control valves and
the requirements that have been set for this investigation. After the selection of the control valve, the
tools that support this research are presented. The reasons that orifice plates will stand as a guideline
for the investigation and the optimization of the control valve are documented. Also, some fundamental
fluid mechanics principles are presented and an introduction to computational fluid dynamics is provided.

1. Flowmeters & Calibration

Sensors and meters are devices that assist us to better understand how a system works. As
the industry is expanding and the demands become bigger, it is crucial to produce meters that
can provide information that describe accurately a process. Most branches of the industry are
dependent on the flow measurement. So, it is a necessity to manufacture flowmeters that are
capable of calculating precisely the flowrate at every moment.
One fundamental problem, that is inherent to all flowmeters, is that despite the standardized
process of manufacturing, their behavior is initially different at the same operating conditions.
The process of calibration aims to resolve this issue. So, calibration can be defined as the
procedure, during which a comparison between the reading of the flowmeter and a standard
is made, in order to establish the relationship that will determine how the flowmeter will perform
in practice. Special care is taken, since there are numerous influencing factors on the flow
measurement, the process is carried out under strictly specified conditions.
Flowmeter calibration systems are very complicated because of they consist of numerous
components. En route to the presentation of the control valve, which is the topic of this thesis,
it is considered important to describe in detail key elements of the calibration system. As a
result, calibration and verification methods, the ways that the liquid is collected, and all the
subgroups of the system will be presented. The modules of the calibration system are listed
below:
a) Tank
The start and finishing point of a calibration point is a tank that stores the fluid, that returns
from the measuring module, until it is fed again to the system to continue or repeat the
process. Lime deposits and contaminants can be inspected in this phase, in order to ensure
that the fluid will be pure during the calibration.
b) Pump
The pump module has the task of pumping the water through the pipes of the calibration
system. The fact that the calibration is done in a number of different flowrates, this demands
the adjustment of the pressure head. Furthermore, the calibration is performed under the
same operating conditions in terms of pressure at the outlet of the flowmeter that is under
test. So, the pump has to ensure that the required operating pressure in the system for
each flow rate by means of speed control. An alternative to the most common method of
throttling with a valve is to use variable speed to change the pump curve and meet the
pressure head demands of the system. If the pressure drop in the flowmeter under test is
considered negligible, the pressure head is maintained until the inlet of the control valve.
The change in the position of the control valve, will demand from the pump to vary its speed
in order to maintain constant pressure head for the process. For this purpose, the pump
must be coupled with the control valve.
 c) Flowmeters
From the pump the water is led into the measuring section. The flow meter to be calibrated
is clamped in this module. To perform the calibration more efficiently, two measuring
sections are implemented. A special challenge of the measuring section is to optimize the
flow profile to the point so that the measuring result is not influenced by the measuring
device. Flow conditioners or sufficiently long inlet pipes are, thus, usually used. The volume
flow now passes from the measuring section into the master module. Master flowmeters
are used as reference during calibration and serve as verifiers. A single flowmeter can be
used in series with the DUT (Device Under Test). In the same way, multiple meters in
parallel can be assembled to achieve a flowrate range in excess of what may be achieved
through the primary system. Sets of six, eight or more master flowmeters in parallel can be
assembled to test and calibrate flowmeters that handle very large flowrates.

d) Control valve system

After the measuring section and the Master-flowmeters module, the volume flow enters the
control valve system, which consists of two main parts. The first one is section of the control
valve. When the valve actuates, changes its position and the effective cross-sectional area
of the flow. Consequently, the flowrate is adjusted to the new conditions. After the outlet
section of the control valve, a variable nozzle exists. Its aim is to fully remove turbulence
and straighten the flow to exit the control valve system.
e) Fluid collection & measurement
At the last part of the calibration system, the reference flow measurement is made. This is
done mostly in two different ways. The first option is to conduct gravimetric calibration. In
this case the fluid is collected and weighted in a vessel while the process is ongoing. At
the end of the calibration at a flowrate, the time and mass differences are calculated to
produce the reference measurement. The second option is to conduct volumetric
calibration, where in this case, respectively, the difference in the level of the fluid inside the
vessel is measured.
One more important decision that shapes this last section is related to the method that the
fluid will be collected. Again, there are two methods that are mostly used. The first one,
known as “Standing start & stop”, is very simple and requires the entire fluid to be collected,
from the beginning until the end of the calibration. When the container is empty, the flow
starts, and when the container filled the flow stops. The quantity, that was collected, is
measured and compared with the indication of the flowmeter; the time to fill provides the
flowrate. However, on the other hand, there is the option to maintain a continuous flow in
the DUT. This is achieved in the “Flying start & finish” method with a diverting mechanism.
The diverter aims to guide the flow in the vessel when the measurement is underway or
lead the fluid to the storage tank. The pulse, that changes the position of the diverter,
signals the beginning and the end of the measurement each time. At the end of calibration,
the water from the vessel returns with a piping network to the storage tank to be ready for
the next one.

2. Control Valve

i) Introduction
The topic of this thesis focuses exclusively on the control valve system, and this will be the
sole point of discussion in this and the following chapters.
Control valve can be defined as the device that has the ability to throttle and regulate a flow
with its movement. They are key components in closed loop control systems, because as the
final element of the loop, they are responsible for causing a desired change in the system. The
control loop sends a signal, which then, via an actuator, changes the position of the control
valve. The regulation of the flow is achieved with the control valve’s translation or rotation. In
this way, the effective cross-sectional opening varies. As the fluid approaches the control valve,
it accelerates in order to maintain the mass flow through the control valve, until it reaches its
maximum velocity at “vena contracta”. At the same point, the pressure of the fluid is reduced
its lowest value. Then, the fluid recovers again, as the velocity gradually decreases and the
pressure once more increases to its final levels.

Figure 1: Vena contract at control valves & velocity and pressure

fluctuation

ii) Inherent flow characteristics

An initial attempt to categorize the control valves is made by dividing them into groups
according to their inherent flow characteristic curves. These curves display the relationship
between the position of the control valve in the available range of their movement and the
resulting flowrate at the respective position. To determine them, the control valves are tested
under constant pressure drop conditions across the restriction at various positions. The most
common resulting curves are presented in Figure 2, and lead to the characterization of control
valves as:
a) Quick opening valve
In these characteristic curves, the flowrate increases abruptly for the initial travel of the
control valve. So, only with a few steps, almost the maximum flow rate can be achieved in
these control valve, and then there is negligible flow adjustment. So, these control valves
are primarily useful for “on-off” applications, where there is need for tight shut-off.
b) Linear valve
These control valve have a characteristic curve which is approximated by a straight line.
So, the relationship between the position of the control valve and the flowrate it produces
is directly proportional. For a specific travel of the valve, percentagewise, there will be the
same flowrate variation. This kind of control valves is suitable for some flow-control
applications.
c) Equal-percentage valve
 This category of control valve characteristics is the opposite of the Quick-opening valves
and the curve resembles a parabola. In this case, at the initial travel of the control valve,
despite the movement, the flowrate is hardly increased. However, as the control valve
keeps on actuating and increases the cross-sectional area of the restriction the increase in
the flowrates takes place in a faster pace. However, in total, the change in the flowrate, as
the name implies, is of equal percentage for the same change in the control valve’s
position.

Figure 2: Inherent flow characteristic curves of control valves

iii) Flow phenomena

In the majority of the control valves, the flow is subjected to big changes, as the fluid abruptly
accelerates and decelerates across the control valve. These conditions are the ones that make

Figure 3: Process of cavitation & Formation of bubbles

the control valves prone to the occurrence of cavitation and flashing. These are two flow-
induced phenomena that can lead to extensive damage of the control valve geometry.
Cavitation is a phenomenon, during which a homogenous liquid, under constant temperature,
is partially transformed into vapor. The characterization of partial transformation is chosen,
since cavitation appears to take place locally, in regions where the pressure is lower than the
respective value, which defines the fluid’s vapor point. So, at the process of cavitation, the fluid
is initially subjected to a reduction in the pressure, which is followed by an increase, such that
the final pressure value is higher that the corresponding vapor pressure.
Although it was previously referred as bubble, cavitation may occur in different ways. The flow
regime and the geometry of the obstacle that lead to the formation of cavitation are two
significantly contribution factors. The most common types cavitation are bubble, sheet and
vortex cavitation. It is also possible that cavitation may consist of more than one type or it can
be transitioned from one type to another.

• Bubble cavitation consists of separate bubbles of cavitation which travel along with the
liquid of the free stream. This type mainly occurs in low flow rate conditions with uniform
pressure distribution on the surfaces. This type is considered to be very harmful since
the collapse of cavitation bubbles leads to significant amounts of energy being
exchanged with the solid surfaces.
• Sheet cavitation is a type that corresponds to even lower cavitation numbers than the
ones of bubble cavitation. It primarily appears in hydrofoils and is formed from large
bubble which are merged to shape a cavitation sheet.
• Vortex cavitation is formed in regions of high vorticity. Centrifugal forces on the fluid have
a significant impact on the formation off cavitation vortices. The magnitude of these
forces determines the appearance of the vortices since bigger forces push bubbles to
the center of the vortex and form a smoother looking cavitation.
In its various forms, cavitation does not have the same intensity and effect on the flow and the
control valve system every time. Categorizing and identifying the intensity of the cavitation,
could serve as precaution to avoid extensive damage. So, cavitation is classified in increasing
intensity order, accordingly:
1) Incipient cavitation, which is hardly detectable with the cavitation structures being few.
The level of vibrations and noises can be observed only with measurement devices,
and consequently no damage occurs to the system.
2) In critical cavitation, the occurrence of cavitation structures is steady, and the impact
becomes more intense as the flowrates increase. The noise and the vibrations that
critical cavitation causes are noticeable, but still there is not any obvious damage in the
geometry.
3) In the stage of incipient damage, either the geometry of the control valve or the
downstream piping is harmed. Cavitation has managed to intensify and cause damage
to the system. Nevertheless, it is difficult to detect this stage. So, it is considered as a
reasonable designing limit, in order to avoid further damage.
4) Choking cavitation, at which, as the name implies, the flow is choked. The flowrate is
not increasing anymore and damage at its most severe level occurs before the flow
gets choked. Very intense vibrations and noise characterize this level of cavitation.
To avoid such problems from cavitation, parameters and procedures have to be determined.
One very common practice is the use of the cavitation number:
𝑃𝑢 − 𝑃𝑣
𝜎=
Δ𝑃
where Pu is the pressure of the flow upstream the control valve, Pv is the vapor pressure of the
fluid and ΔΡ is the pressure drop across the control valve. So, this parameter is the ratio of the
forces that try to avoid cavitation, to the forces that cause it. So, it is important, when designing
a control valve, to ensure that the conditions will be such, that the number will get higher values,
because when this parameter decreases, the danger of cavitation occurrence and resulting
damages is higher. However, the cavitation number is a function not only of the conditions of
the flow, i.e. its pressure, but also a function of the entire system geometry, including the size
of the control valve and its design. A standard experimental procedure to detect the occurrence
of cavitation is by installing accelerometers to the measurement rig. By altering the operating
conditions, this will lead to a change in the flowrate and then in the reading of the
accelerometer. At the conditions where the acceleration of the vibration will start increasing is
the point where incipient cavitation is occurring.
Hydraulic systems are designed to operate under the desired ideal, non-cavitating conditions.
Nevertheless, sometimes it is not possible to completely avoid the formation of cavitation
cavities, so the most important action to take the necessary precautions to restrict the formation
and consequences of cavitation. The primary expected consequences of the formation of
cavitation in the flow are:

• The performance of the system is alternated. In hydraulic systems this alternation is

interpreted as reduced capacity in the pipes. The flow downstream the region where
cavitation occurred, contains bubbles, which occupy some space. This means that there
are head losses, which affect the ability of a valve to convert pressure drop into flow rate.
As a result, a valve is unable to handle larger flow rates in such cases.
• The local geometry may be physically damaged. The bubbles that are carried with the
flow impact on the solid walls and accelerate the appearance of damage.
• Production of noise and vibrations, which is dependent on the intensity of the cavitation.
In order to avoid the negative consequences of the occurrence of cavitation, certain guidelines
have been set, especially for the control valves. These actions, that should be taken, are:

• Ensure that hard facing materials are used to resist the damaging effects of cavitation.
• Increase the outlet pressure, so that the flow pressure adequately recovers to a level
that does not allow cavitation structures to be created.
• Prevent wall contact with the cavitation bubbles after the sudden pressure drop, so that
the bubbles collapse in the free stream.
• Dissipate the turbulent kinetic energy of the resulting flow after the pressure drop, by
either turbulent mixing or dividing the pressure drop in multiple stage. The approximated
stages 𝒏, which are necessary, are calculated according to the following expression:
𝑝2
𝑛 = −6.45 log
𝑝1

Where: P1 = upstream pressure & P2 = downstream pressure.

• Use of the principle of flow partition. The division of the flow in multiple jets improves the
dynamic behavior of the medium and leads to more uniform and better downstream flow.

The recovery of pressure of pressure sometimes is not adequate in order to exceed the vapor
pressure, and the final value is lower than the vapor pressure. In these cases, the respective
phenomenon is called flashing. When flashing occurs, the bubbles do not collapse, but instead
they keep travelling and are generated along with the liquid flow. The consequences of flashing
are identical to the ones of choked cavitation. So, there is extensive physical damage in the
control valve and the capacity of the valve is reduced. The bubble occupy a large part of the
flow stream area and do not allow to the flow to increase its rate.
 Figure 4: Process of flashing - Generation & travelling of bubbles

iv) Control valve types

Control valves can be classified in different ways, according to the purpose of the selection
that has to be made. In this thesis, it is aimed to develop a control valve that will regulate the
flow and will be classified according to their motion. This preliminary classification is shown in
Figure 5.
Control
valve

Linear Rotary
motion motion

Globe Ball

Gate Butterfly

Diaphragm Plug

Needle Rotary disc

Pinch

Figure 5: Classification of control valves according to their motion

Each of these control valve types will be elaborated, by presenting the necessary information
on the way the work, their advantages and disadvantages, and, more importantly, the suitable
applications to each type.

• Globe valves
Globe valves have a very compact and simple structure. They consist of a ring seat and a body
connected to a rotating shaft, which transfers the signal of the actuator to the internal structure.
The shape of the main plug body varies, and it can be a semi-globe, a semi-disc or even a
needle. When the shaft starts rotating, the fluid passes through the area between the edges of
the seat and the plug. The lift of the main body determines how large the flowrate, that crosses
the control valve, will be.
 Figure 6: Globe valve design

In this type of valve, due to the symmetric shape of the opening, the forces that are exerted on
the plug are equally distributed, and thus lead to an even and reduced wear. Also, their simple
design, along with the construction of the main body and its seat, provide very good flow
regulation and throttling characteristics. More sophisticated globe valve designs are able to
provide fine flow control, too. However, the complex path that the fluid has to follow, results in
higher turbulence levels and pressure drop. Consequently, the flow capacity of the valve is
reduced and the pump’s demands for pressure head are increased.
The broad use of the globe valves is directed in applications, where the control valve has to
operate frequently, and the flow control demand is high. They are primarily used in two
positions, to start and stop the flow, and although it can be a throttling valve, in cases where
fine adjustment is required, they are not the most suitable.

• Gate valves

Probably one of the simplest designs for a control valve, the wedge gate valve is a wedge plate
with a shaft attached to its highest point, so that they can slide in a direction perpendicular to
the pipe, as it is displayed in Figure 3. As the gate is lifted, it is stored in a chamber, in order
not to interfere with the flow.

Figure 7: Flow at different positions of the gate valve

This characteristic results in minimum resistance to the flow, and thus, increases the capacity
of the valve. The simple design significantly decreases their cost and its structure is assisting
towards providing very good shut-off characteristics at the fully closed position. However, the
control over its motion is poor and the fact that in the partially open positions, the gate valve is
prone to extensive vibrations and damage, because of the extreme reduction of the fluid’s
pressure, sets this valve as unsuitable for throttling or flow-regulating applications. As a
consequence, they can be used for their tight shut-off characteristics to start and stop the flow,
with a recommendation not to be used very often.

• Diaphragm valves

This is a distinctive valve type because the part that is responsible of the variation of the flow
passage is an elastic diaphragm that works as a seal in the fully closed position. The flexible
diaphragm is moved by a solid part that deforms it. The selection of the flexible material and
the demand for the range of the diaphragm’s stroke determine its type, which is either weir or
straight-through.
This special type of control valve has the great advantage that, because of its shape, it can
provide very good flow characteristics, as the flow is streamlined and hardly leads to mixing.

Figure 8: Weir & straight-through diaphragm valves

On the other hand, the material of the diaphragm introduces certain limitations to the use of
this control valve. Depending on its properties, it is determined which are the maximum
temperature and pressure conditions, under which the diaphragm can be used.
However, when the conditions are appropriate, it can be used in tight shut-off applications and
it is capable of throttling the fluid in a very narrow range. The most important characteristic of
this type, however, is that it can handle almost all types of fluids and slurries, despite of the
corrosive or erosive properties they may have.

• Needle valves

Needle valves are a specific type of globe valves, where the main body that determines the
opening, instead of a disc shape, is a conical piston that look like a needle. The thread of the
needle body acts against leakages and it provides suitable geometrical characteristics, so that
needle valves can handle regulate and precisely adjust the flow. Moreover, they are suitable
for systems, which operate under high pressure conditions, since no vibrations or cavitation is
detected. Nevertheless, their used is limited in controlling very low flowrates and clean fluids,
in order to take advantage of their assets.
 Figure 9: Different configurations of needle valves

• Pinch valves

Pinch valves are another type of valves that displays a lot of similarities with the
aforementioned types, and more specifically with the diaphragm valves. The main difference
between the two is that, in pinch valves, the entire body is flexible. Also known as clamp valves,
the flexible valve is usually an elastomeric tube or sleeve. The opening, and thus the flowrate,
is regulated in the same way, from a metal closure member. These valves can be controlled
more precisely with multiple screwing or pressurized-air mechanisms.
Similarly, pinch valves are capable of handling fluids that are carrying solid particles, slurries
and aggressive fluids that could harm standard metal valves with their properties. Also,
because their uniformly pressurized and deformed, they enable the flow to remain laminar and
streamlined across the restriction. Finally, the fact that in some pinch valves, the relationship

Figure 10: Pinch valve mechanism

between the sleeve travel and the resulting flowrates is directly proportional, is another
advantageous characteristic. Nevertheless, because of their ability to handle demanding fluids
and corrosive material, they are mainly used in the chemical industry or in the water treatment
industries.

• Ball valves

Being the most wide-known control valves, because of their uses in home applications, ball
valves are the least expensive valves and incorporate a very simple operating mechanism,
too. They consist of two metal rings, that put in place the main body, which is a spherical ball
with a through hole. The ball valve transitions from the fully closed position, where the ball is
perpendicular to the flow, to a fully open position in a move of 90o travel.
Their broad use in so many different kinds of occasions is that their cost is very low, in
combination with the fact that they can efficiently operate under a variety of pressure and
temperature conditions, while they can handle a large range of flowrates. Furthermore, with
their design, leakages are sufficiently restricted and at the fully closed position they provide
tight shut-off conditions. However, at partially open positions, it is very likely that cavitation will
occur, and they cannot handle liquids with solid particles or slurries.
So, their main use is not for flow control and they restricted to “on-off” applications, where
limited throttling abilities are needed. Finally, they are used in applications where the working
fluid has a high temperature.

Figure 11: Ball valve

• Butterfly valves

Another simple design for control valves, is the one of butterfly valves, which consists of a disc
and an axis that rotates the disc. At the fully open position, the disc of the valve is parallel to
the flow, and perpendicular to the flow at the fully closed position, respectively. So, like ball
valves, the range of butterfly valves’ motion is limited to 90o. However, in contrast with the ball
valve, can be used for throttling applications, if tight shut-off is not a requirement.

Figure 12: Butterfly valve

design

As in the case of ball valves, this valve type has also low cost and high flow capacity, but also
manages to provide good flow control characteristics with reduced pressure drop. Although
they can be manufactured even in very large sizes, if the valve is used for flow-control
purposes, there are limitations in the range of the ranges that they can handle. Also, a problem
that comes with butterfly valves, is that they cannot be used for tight shut-off needs, as there
is always even a small amount of leaking fluid.
For these reasons, it is preferred to use butterfly valves in application where regulation of the
flow is a requirement. Also, their broad use is justified from the wide range of materials that
may compose the fluid, varying from gases and standard liquids to slurries, powder and
granules.

• Plug valves

Plug valves are one more modified design of the primary ball valve design. In this type the
main body of the valve is a cylinder with a flow port, that enables under certain position to the
flow to pass across the valve. Similarly, with a travel of 90o, the valve can fully obstruct the
flow. As a combination of the characteristics of the valves that have already been presented,
plug valves can provide tight shut-off and are primarily used to start and stop the flow.
However, they can also be used for throttling and regulating flow purposes. Because of their
design, they are useful in cases, where the flow has to be diverted. Nevertheless, there are
certain limitations that are related mostly with the temperature and the pressure of the fluid.
Preferably, both of these parameters should have low values. Otherwise, the tightness of the
system is reduced.

Figure 13: Plug valve design

• Rotating Disc valves

Finally, the rotating disc valve is simple design that uses the surface of the disc to incorporate
the shape of the opening. The flow passage is determined by the cross-section of the inlet pipe
and the drilled hole on the disc.
This type of valve can provide continuous rotation of the disc and an extended range of motion
and precise timing in the movement and the opening of the port. Also, the design ensures that

Figure 14: Rotating disc valve type design

transitional flows occur, and as a result, the transition from a fully closed to the fully open
position is smoother. Furthermore, because of the continuous movement that can take
advantage of all the flowrates in the working range, problems of deadhead in pumps are
eliminated. Also, its design provides reduced pressure drop in the fully open position. However,
standard designs struggle with friction and leakage related problems.
Despite the simple configuration and design of this valve type, they have been classified as
valves that were not frequently used. Nevertheless, the realization that there are very good
characteristics in terms of flow control, has led to an increase in their use recently, and
expanded the range of the applications that they are used in.

v) Requirements & Control valve type selection

The presentation of information for various control valve types provided a clear picture of their
characteristics. For the purposes of the thesis, the focus will be on only one of them. The
selection process will be done according to the following considerations:

o The fluid that has to be handled may pose limitation in the selection, because of its
properties and their interaction of the material that the valve can be manufactured.
o There are functional requirements that can significantly influence the decision on the
control valve type.
o The operating conditions mainly referring to the pressure and temperature of the
system are another limitation in the valve selection.
o Frictional losses and flow related requirements are one more aspect of the problem of
the selection.
o The size that the valve can have is one more restriction, since the valve is just another
part of a large system.
o According to the application, more special requirements may apply, that make the
selection process more complicated.

In a calibration process, with a system as extensive as the one that has been described, with
numerous components, the rigidness and stable operation of the control valve is a key to
overall process efficiency. So, some further requirements that will specify the characteristics
of the control valve and the behavior of the flow across it, are presented below:

o Linearity for the control valve system

The variation of a control parameter should lead to the same variation in the measuring
parameter, for a small region of the investigated range of flowrates. This requirement does not
add limitations on the selection on the inherent characteristics of the control valve. Its purpose
is to ensure that even locally the variation of the flowrate will be minimal and predictable without
unforeseeable changes. To achieve this goal, the gradient of the resulting flowrate should be
as low and constant as possible at the various positions of the control valve.

o Repeatability of the control valve

A critical parameter that significantly determines the operation of the system is the water’s
pressure at the inlet of the control valve system. From a pressure of 3 bars after the flowmeter
that is calibrated into the ambient conditions that the fluid returns, it is important to maintain
the necessary pressure for the whole measuring range, and at the same time, design the valve
in such a way that enables the flow measurement to have a resolution of 1 %.

o Robustness of the system

The proper operation of the control valve system is a complicated problem, that demands the
control of multiple parameters. The control valve must be able to operate under different
conditions. The effect of lime depositions and increased hardness of the water, manufacturing
tolerances, false operating conditions and variation of the ambient conditions are aspects of
this problem, that have to be considered. Precautions have to be provided and designing
decisions, that will lead provide warnings when problems arise or, preferably, reduce the
parameters’ impact on the performance of the control valve to negligible levels, have to be
made.

o Flow Stability

The flow is in the turbulent regime, starting even from the smallest flowrates. The nature of
turbulence introduces variances to the measured mean flowrates and pressure of the system.
The goal is to introduce critical parameters in the design of the control valve and select the
proper values, so that these fluctuations are minimized to a range of 0.2 % of the mean flow
for every measuring point.
When all these requirements and restrictions are considered, according to the presented
characteristics of the various valve types, selecting which control valve will be investigated in
this thesis is not very difficult. The simplicity of the design along with the ability to provide
precise flow control over a wide range sets the rotary disc type as the favorite for this selection.
The guidelines for the selection, that were earlier mentioned, and it cannot be yet ensured that
are met as limitations, will be attempted to be achieved as goals in the course of this thesis.
The investigation, improvement and, eventually, optimization of the control valve will be
conducted from a flow and geometrical perspective. Issues that are related to friction or other
structural and purely technical matters do not belong in the context of this project and they will
not be further elaborated.

vi) Orifice plate similarity

The analysis of the control valve has been based on one approximation that serves as a guide
to the research that has been conducted in this thesis. The geometry of the disc is very similar
to the one of orifice plates. Orifice plates are normally used as differential-pressure flowmeters.
Differential-pressure meters partially obstruct the flow and hence operate by creating a
difference in static pressure between the upstream and the throat of the device. They are
normally installed in line, i.e. with pipework both upstream and downstream of them, but they
may be installed at the inlet to, or sometimes at the outlet from, a length of pipe.
An orifice plate is fundamentally a plate with a hole machined through it which is inserted into
a pipe. As flow passes through the hole it produces a pressure difference across the hole, with
some of the pressure being recovered downstream the plate. The pressure difference is
proportional to the square of the mass or volume flowrate. To design an orifice plate, given the
range of mass flowrates, the density, the desired range of differential pressures, the discharge
coefficient, the expansibility factor and the desired pipe diameter (D), the hole diameter (d) can
be determined from Eq. 1:

𝑚̇ 𝑚̇
𝐶𝑑 = ⇒ 𝐶𝑑 =
2𝛥𝑝 𝜋𝑑2 2𝛥𝑝
𝐴√ 4 √𝜌(1 − 𝛽 4 )
𝜌(1 − 𝛽 4 )

A few shapes and designs have been produced for which the discharge coefficient can be
predicted. Of these only for a round sharp-edged centrally located hole are there sufficient data
to allow prediction of flowrate with an uncertainty which is low enough to use them broadly with
the various fluids. Designing properly an orifice demands precision and special attention to
every aspect of the plate (the circularity of the bore, the flatness, the parallelism of the two
faces, the surface condition of the upstream face, and, above all, the edge sharpness), the
pipe (the pressure tappings, the pipe roughness, the effect of upstream steps, the concentricity
of the orifice plate in the pipe and the circularity and cylindricality of the pipe) and the pressure
loss.
However, in the view of the of rotary disc control valve design, certain differences can be
observed. The standard orifice plate is a very simple geometry of a thin disc that has a central
hole, which has been installed in a suitable position, and interferes with the flow in a rectilinear
tube. The first difference is related to the position of the opening. The first difference is that the

Figure 15:Orifice plate types

hole is not concentric with the disc. In a standard orifice, the center of the hole is coincident
with the plate's center. In contrast with the orifice, in this control valve design, whichever
opening area might exist is in an eccentric position. To create the essential conditions for a
variable, growing orifice, the opening is not possible to be placed constantly in the center of
the disc. Even from a structural point of view this is not possible, a priori, because in this
position is placed the shaft of the control valve disc. The second difference results from the
fact that the variable -depending on the position- area is the outcome of a spiral's and a circle's
intersection. Consequently, the opening area is always not circular, with the only exception
being the position where the opening area is maximum.

Consequently, the control valve disc eventually resembles a combination of an eccentric and
a segmental orifice plate. The various types of orifice plates are presented in Figure 2.
Segmental and eccentric orifice plates are functionally identical to the concentric orifice. The
circular section of the segmental orifice is concentric with the pipe. The segmental portion of
the orifice eliminates damming of foreign materials on the upstream side of the orifice when
mounted in a horizontal pipe. Eccentric orifice plates shift the edge of the orifice to the inside
of the pipe wall. So, the control valve plate is similar to some orifice plate type in terms of
shape, with the geometry and the respective parameters in both cases seeming to strongly
influence the performance, i.e. the discharge coefficient. Finally, the configuration of the control
valve system and the orifice plate with its piping sections are similar, too.

vii) Fluid Mechanics

Part of the analysis in the thesis is conducted with analytical methods, utilizing fundamental
principles of fluid mechanics, in order to acquire data that will develop some guidelines at the
design of the control valve or assist in better understanding and modifying, when necessary,
results from experiments and simulations. For this reason, the flow regimes will be shortly
described, reporting some characteristics of laminar and turbulent flows, accompanied by a
brief presentation of the dimensional analysis and the primary unidimensional parameters that
are used here. Along with them, the main equations, that will be presented, are the Bernoulli
equation and the conservation of mass.

• Dimensional analysis

As humans continuously build ever more complicated systems, the fluid-mechanics problems
they have to cope with, become proportionally more difficult. The majority of the engineering
problem cannot be solved anymore in an analytical manner. Experiments to investigate a new
system are necessary. Another option is to use numerical tools and conduct simulations.
However, in both cases, the results are acquired that does not help engineers to easily
understand the phenomena that take place or a tendency of an investigated parameter.
Dimensional analysis is a traditional and powerful tool that engineers use in the field of fluid
mechanics in order to describe results of experiments in a compact manner, transform or fill
theories and computer models.
Dimensional analysis is a method that condenses a multi-variable physical phenomenon into
a problem with less variables. Reducing the complexity of a problem provides a series of
advantages, as in the case of an experiment for instance, the investigation of a multi-variable
phenomenon would be cost-ineffective and extremely time-consuming. Furthermore,
dimensional analysis is assisting into structuring theories in an alternative, dimensionless way.
By rejecting variables and transforming the problem, deeper insight into the physical
relationships of the studied phenomenon can be gained. Finally, due to the dimensionless
parameters that are introduced, scaling laws can be provided. Tests and experiments in full-
scale models are not worthy, and primarily, extremely expensive, when there is the capability
to test a phenomenon or the performance of a prototype.
In fluid mechanics, although, there are numerous dimensionless parameters that are used, the
Reynolds number is considered to be the most commonly used, because of the correlations
that it provides between geometry, the properties of the fluid and the operational conditions in
the flow. Reynolds number is defined as the ratio of the inertial to the viscous forces on the
fluid.
𝜌𝑉𝑎𝑣𝑔 𝐿 𝑉𝑎𝑣𝑔 𝐷
𝑅𝑒 = =
𝜇 𝜈

At large Reynolds numbers, the fluid's behavior is primarily affected by the inertial forces and
relatively larger to the viscous forces. Consequently, the viscous forces cannot prevent the
random and rapid fluctuations of the fluid. However, at lower Reynolds numbers, however, the
shear stresses caused by the fluid's viscosity are large enough to suppress these fluctuations
and to keep the fluid “in line.” The type of the flow is considered as turbulent and laminar in
each case, respectively. The transition from a laminar to a turbulent flow is not occurring in a
specific Re, but in range of Re values. Thus, 3 main flow regimes are defined, depending on
the value of Re, as presented in Table 1:

Type of flow Re
Laminar Re ≤ 2300
Transitional 2300<Re<4000
Turbulent Re ≥ 4000

• Laminar Flow

In fluid mechanics, a flow can be characterized as laminar when the fluid particles following
smooth paths in layers, without getting mixed with the adjacent layers. At low velocities, fluid
particles of neighboring streamlines are flowing without lateral mixing. There are no
crosscurrents in different flow directions, nor eddies or swirl effects. So, the motion of the
particles of the fluid is described by straight lines parallel each other. Laminar flow is a regime
distinguished for being stable against disturbances introduced from outside, resulting in flows
with a high degree of order and in which diffusion phenomena are characterized by high
molecular diffusion and low momentum convection.
Laminar flows can be dependent on time, as the von Karman vortex shows, which is depicted
in Figure 1. As long as the flow maintains its characteristic order, it is laminar. This means that
the viscosity of the fluid is able, in stable laminar flows, to attenuate sufficiently fluctuations of
the flow properties that would otherwise disturb the orderliness of the flow. Perturbation
attenuations of this kind usually occur at low Reynolds numbers of all flows.

Figure 16: Example of laminar flow

• Turbulent Flow

Turbulence is commonly observed in everyday phenomena such as surf, fast flowing rivers,
billowing storm clouds, or smoke from a chimney, and most fluid flows occurring in nature or

Figure 17: Turbulent flow example - wake of flow around a cylinder

created in engineering applications are turbulent. For a flow to be turbulent, the kinetic energy
of the fluid has to be very high, and thus its inertial forces, so that the damping effect of the
viscosity is overcome. For this reason, turbulence is mainly observed in low viscosity fluids. In
general terms, in turbulent flow, unsteady vortices appear of many sizes which interact with
each other, consequently, drag due to friction effects increases.
In flows that characterized by high Reynolds numbers, flow phenomena such as the von
Karman vortices, visually perceivable thanks to flow visualization techniques, lose their
“regularity,” as indicated in Figure 2. These exaggerated fluctuations immensely influence the
mean flow characteristics. The fluid motions, known to be of high regularity for laminar flows,
do not exist any longer and, in turbulent flows, the state of the fluid displays strong irregularity,
in connection with an extremely high diffusivity which can exceed the molecular-dependent
transport processes by several orders of magnitude. On top of that, the increased intermixing
of the fluid and transport rate of the momentum, result in higher heat and mass transport, too.
In order to sufficiently describe a turbulent flow field, it is essential to measure instantly the
velocity and derive its time-dependent behavior. At a point in space, the signal is characterized
by strong fluctuations of the flow velocity in time, which can be stated as deviations u'j(xi, t)
from a mean value Uj(xi), the latter being a constant with respect to time. Here, the time mean
value Uj(xi) is defined as follows:

1 𝑇
̅𝑗 (𝑥𝑖 ) = lim
𝑈 ∫ 𝑈̂𝑗 (𝑥𝑖 , 𝑡)𝑑𝑡
𝑇→∞ 𝑇 0

where 𝑈 ̂𝑗 (𝑥𝑖 , 𝑡) indicates the instantaneous value of the velocity and T is the integration time
over which the indicated time-averaging takes place. Thus, 𝑈 ̂𝑗 (𝑥𝑖 , 𝑡) can be taken as a quantity
which allows one to consider the local flow velocity, varying over time, as the sum of a quantity
that is constant with respect to time and a quantity that is fluctuating in time. This
decomposition of the instantaneous velocity 𝑈 ̂𝑗 (𝑥𝑖 , 𝑡) into a time-averaged part 𝑈 ̅𝑗 (𝑥𝑖 ) and a
′
fluctuating part 𝑢𝑗 (𝑥𝑖 , 𝑡) has advantages. It was introduced by Reynolds to treat turbulent flows.

𝑢𝑗′ (𝑥𝑖 , 𝑡) = 𝑈
̂𝑗 (𝑥𝑖 , 𝑡) − 𝑈
̅𝑗 (𝑥𝑖 )

The time average of the turbulent velocity fluctuations 𝑢𝑗′ is equal to zero per definition. Hence,
there is a way to present turbulence in local, time-varying quantities, in a form such that the
turbulent fluctuations of all flow quantities, that are introduced into the considerations, show a
time mean value that is zero. To quantify the intensity of turbulence, the RMS values of the
fluctuating terms are calculated. So, the
̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
√1 (𝑢 ̅̅̅2 ̅̅̅2 ̅̅̅2
2 1 + 𝑢 2 + 𝑢3 )
𝑇𝑢 =
̅𝑡𝑜𝑡
𝑈

When the value is around 10% and more, the flow is defined as highly turbulent. Highly
turbulent flows occur mostly in industrial flow systems. It is the task of fluid mechanics to
develop and bring to application measurement techniques and numerical solution methods
which allow investigations of turbulent flows with low and high turbulence intensities. In
practice, it is often sufficient to only have information on time mean values of turbulent
quantities of an investigated flow field.

• Bernoulli Equation

The Bernoulli equation is an approximate relation between pressure, velocity, and elevation,
and is valid in regions of steady, incompressible flow where net frictional forces are negligible.
Despite its simplicity, it has proven to be a very powerful tool in fluid mechanics.
The key approximation in the derivation of the Bernoulli equation is that viscous effects are
negligibly small compared to inertial, gravitational, and pressure effects. Since all fluids have
viscosity (there is no such thing as an “inviscid fluid”), this approximation cannot be valid for
an entire flow field of practical interest. However, it turns out that the approximation is
reasonable in certain regions of many practical flows.
However, since in general, frictional effects are always important very close to solid walls
(boundary layers) and directly downstream of bodies (wakes), the Bernoulli equation is
typically useful in flow regions outside of boundary layers and wakes, where the fluid motion is
governed by the combined effects of pressure and gravity forces.
The motion of a particle and the path it follows are described by the velocity vector as a function
of time and space coordinates and the initial position of the particle. When the flow is steady
(no change with time at a specified location), all particles that pass through the same point
follow the same path (which is the streamline), and the velocity vectors remain tangent to the
path at every point. The Bernoulli equation can be written between any two points on the same
streamline as:

1 1
𝑝1 + 𝜌𝑉12 + 𝜌𝑔𝑧1 = 𝑝2 + 𝜌𝑉22 + 𝜌𝑔𝑧2
2 2
However, the Bernoulli equation is one of the most frequently used and misused equations in
fluid mechanics. Its versatility, simplicity, and ease of use make it a very valuable tool for use
in analysis, but the same attributes also make it very tempting to misuse. One of the restrictions
that must not be violated is that the flow has to be frictionless. A component that disturbs the
streamlined structure of flow and thus causes considerable mixing and backflow such as a
sharp entrance of a tube or a partially closed valve in a flow section can make the Bernoulli
equation inapplicable. Nevertheless, since this equation is useful, it will be attempted to be
used under rational assumptions and approximations that will provide important information.

• Continuity – Mass Conservation Equation

The conservation of mass relation for a closed system undergoing a change is expressed as
msys = constant or dmsys/dt = 0, which is a statement of the obvious that the mass of the system
remains constant during a process. For a control volume (CV), mass balance is expressed in
the rate form as conservation of mass:

𝑑𝑚𝐶𝑉
𝑚̇𝑖𝑛 − 𝑚̇𝑖𝑛 =
𝑑𝑡

where min and mout are the total rates of mass flow into and out of the control volume,
respectively, and dmCV/dt is the rate of change of mass within the control volume boundaries.
In fluid mechanics, the conservation of mass relation written for a differential control volume is
usually called the continuity equation. More specifically, the amount of mass flowing through a
cross section per unit time is called the mass flow rate, and, in average value terms over a
cross section of a pipe is expressed as:

𝑚̇ = 𝜌𝐴𝑐 𝑉𝑎𝑣𝑔

In many practical applications, the density is essentially uniform over the pipe cross section.
Velocity, however, is never uniform over a cross section of a pipe because of the no-slip
condition at the walls. Rather, the velocity varies from zero at the walls to some maximum
value at or near the centerline of the pipe, as it was presented earlier.

viii) Computational Fluid Dynamics

Engineering problems that focus on the study of a flow field can be solved in two ways, either
experimentally or analytically. The latter approach demands the solution of differential
equations analytically or numerically. The complexity of most problems leads to a wide use of
numerical solutions with computational fluid dynamics. CFD is employed to shorten the design
cycle through carefully controlled parametric studies, thereby reducing the required amount of
experimental testing. The current state of computational fluid dynamics is that CFD can handle
laminar flows with ease, but turbulent flows of practical engineering interest are impossible to
solve without invoking turbulence models. Unfortunately, no turbulence model is universal, and
a turbulent CFD solution is only as good as the appropriateness of the turbulence model. In
spite of this limitation, the standard turbulence models yield reasonable results for many
practical engineering problems.
The simulation of a turbulent flow field requires the solution of certain differential equations,
which are presented below:

o Mass Conservation
𝜕𝜌
⃗)=0
+ ∇(𝜌 ∙ 𝑉
𝜕𝑡

o Navier-Stokes
 1
⃗ ∙ ⃗∇)𝑉
(𝑉 ⃗ = − ∇𝑃⃗ + 𝑔 + 𝜈∇2 𝑉
⃗ + ⃗∇𝜏𝑖𝑗
𝜌

̅̅̅̅
𝑢′2 ̅̅̅̅̅
𝑢′ 𝑣′ 𝑢̅̅̅̅̅̅
′ 𝑤′
where 𝜏𝑖𝑗 = [ ̅̅̅̅̅
′
𝑢 𝑣′ 𝑣 ̅̅̅̅
′2 ̅̅̅̅̅̅
𝑣 ′ 𝑤′]
̅̅̅̅̅̅
′
𝑢 𝑤′ 𝑣 ̅̅̅̅̅̅
′ 𝑤′ ̅̅̅̅̅
𝑤 ′2

CFD simulations can be conducted with different methods. The method of direct numerical
simulation (DNS) attempts to solve the entire flow field by calculating the eddies of all scales.
It is clear that this method demands extremely high computational power, and, hence, is not
frequently used. Another approach in CFD is to conduct large eddy simulations (LES), where
the larger structures of turbulent flow are solved, while the smaller ones are modelled. Although
it requires significantly less resources, it is still challenging. For most engineering problems,
the next method to be presented, is proven totally sufficient. In this final method, none of the
eddies is resolved and the mixing and the diffusion that turbulence causes, are included and
calculated with the use of mathematical models. Moreover, in order to simplify the solution, the
flow is considered to be only steady and incompressible. When using a turbulence model of
this method, the steady Navier–Stokes equation is replaced by what is called the Reynolds-
averaged Navier–Stokes (RANS) equation for steady, incompressible, turbulent flow. The
turbulence models are classified in:

1. Algebraic models
2. One-equation models
3. Two-equation models
4. Reynolds Stress Models

The solution of the aforementioned equations is not possible, because of the imbalance in the
number of the unknown parameters and the available equations. For this reason, the majority
of RANS models, except for the RSM, are based on the Bousinnesq approximation, which
introduces the parameter of eddy viscosity. This parameter does not truly exist, but it is defined
as a computational parameter in order to deal with the closure problem. Its goal is to create an
expression that correlates the terms of disturbances caused by turbulence and the average
values of the velocity components. The RANS models have been classified in the categories
that were earlier mentioned, because of the method that is used to calculate the eddy viscosity.
Algebraic models use an expression to complete the calculation, while in the case of 1- and 2-
equation models, 1 and 2 equations are used for the calculation, respectively.
The two-equation models have proven to be very useful in a wide range of applications, and,
hence, are the most popular nowadays. Two of the most popular turbulence models are the k-
ε model and the k-ω model. These so-called two-equation turbulence models add two more
transport equations, which must be solved simultaneously with the equations of mass, linear
momentum and also energy, if this equation is being used. The first equation involves the
kinetic energy of turbulence and the second equation incorporates either the rate of dissipation
ε (of the turbulent kinetic energy into thermal energy) or the specific rate of dissipation ω.
Besides the differences in the equations, there is one more fundamental difference between
the models. Since no turbulence model is universal and the parameters, that they contain, are
calibrated to enable the model work properly for a variety of flows, it can be understood that
each model is suitable for different applications. Depending on the location of the
computational cell or node, the models are classified as low-Reynolds or high-Reynolds
models. In this case, the Reynolds number is a dimensionless parameter that refers to the
distance between the wall of the geometry and the cell. This difference leads also to different
boundary conditions that are used from the model for the turbulence and the velocity field.

ix) Bibliography

Bhatia, A. (2014). Control Valve Basics - Sizing & Selection.

Cengel, Y. A., & Cimbala, J. M. (2006). Fluid Mechanics - Fundamentals and Applications.
McGraw-Hill.
Dickenson, T. C. (1999). Valves, Piping and Pipelines Handbook. Elsevier Science Ltd.
Durst, F. (2008). Fluid Mechanics - An Introduction to the theory of Flui Flows. Springer.
Mobley, R. K. (2000). Fluid Power Dynamics. Butterworth-Heinemann.
NEL. (kein Datum). Good Practice Guide - The Calibration of Flow Meters.
Reader-Harris, M. (2015). Orifice Plates and Venturi Tubes. Springer.
Twort, A. C., Ratnayaka, D. D., & Brandt, M. J. (2000). Water Supply. Elsevier Ltd.
White, F. M. (2011). Fluid Mechanics. McGraw-Hill.