Vous êtes sur la page 1sur 74

INDUSTRIETECHNIK

SRI LANKA INSTITUTE of ADVANCED TECHNOLOGICAL EDUCATION

ELECTRICAL and ELECTRONIC


ENGINEERING
Instructor Manual

Training Unit

Control Systems 3
Theory

No: EE 066

Training Unit
Control Systems 3
Theoretical Part
No.: EE 066

Edition:

2008
All Rights Reserved

Editor:

MCE Industrietechnik Linz GmbH & Co


Education and Training Systems, DM-1
Lunzerstrasse 64 P.O.Box 36, A 4031 Linz / Austria
Tel. (+ 43 / 732) 6987 3475
Fax (+ 43 / 732) 6980 4271
Website: www.mcelinz.com

CONTROL SYSTEMS 3

CONTENTS

Page

LEARNING OBJECTIVES - THEORETICAL.......................................................................3

INTRODUCTION TO CONTROL MECHANISMS ........................................................5

SELF-ACTING CONTROLLERS..................................................................................8

2.1

The self-acting pressure controller.......................................................................8

2.2

Hydraulic powered controller (not seif-actino)......................................................9

2.3

Integral action ....................................................................................................10

2.4

Proportional plus integral action.........................................................................10

2.5

Advantages of hydraulic controllers ...................................................................11

2.6

Disadvantages of hydraulic controllers ..............................................................11

REMOTE CONTROLLERS MOTION BALANCE ..................................................12


3.1

The operation of a controller ..............................................................................12

3.2

Added integral control action .............................................................................13

3.3

Types of delay system used in pneumatic controllers........................................14

3.3.1

3.3.1 Dead-ended delay system Pressure in ................................................15

3.3.2

Open-ended delay system .............................................................................15

3.3.3

Derivative control action.................................................................................16

FORCE BALANCE PNEUMATIC CONTROLLERS ...................................................18


4.1

Proportional-plus integral action "stack" controller.............................................19

4.2

Derivative control action.....................................................................................20

ELECTRONIC CONTROLLERS ................................................................................21


5.1

P-Control unit .....................................................................................................21

5.2

I-Control un it .....................................................................................................22

5.3

PI-Control unit ....................................................................................................23

5.4

PID-Control unit .................................................................................................24

MANUAL CONTROL AND "BUMPLESS TRANSFER" ..............................................26


6.1

Integral tracking .................................................................................................27

6.2

The standby mode operation .............................................................................27

ADJUSTMENT OF AUTOMATIC CONTROLLERS ...................................................30


7.1

Proportional action only .....................................................................................30

7.2

Proportional plus Integral action ........................................................................30

7.3

Proportional plus derivative action .....................................................................30

7.4

Proportional plus integral plus derivative action.................................................31

7.5

Setting of the control parameters accordinq to the transfer function (reaction

curve) 31
7.6

Adjusting a cascade control ...............................................................................32

INTRODUCTION TO FINAL CONTROL ELEMENTS................................................34


8.1

The control valve................................................................................................35

POSITIONERS ...........................................................................................................41

10

CONTROL VALVE CHARACTERISTICS..............................................................42

10.1

Butterfly valve characteristics ............................................................................42

10.2

Quick opening and equal percentaqe characteristics ........................................43

10.3

The globe-type valve..........................................................................................44

11

INSTALLED CHARACTERISTICS ........................................................................46

12

VALVE SIZING ......................................................................................................50

13

MAINTENANCE OF CONTROL VALVES .............................................................52

14

DYNAMIC RESPONSE .........................................................................................53

15

CONTROL VALVES AS PLANT EQUIPMENT......................................................54

16

CHOICE OF VALVE ACTION................................................................................57

17

SPLIT-RANGE OPERATION.................................................................................58

18

18. DIGITAL COMPUTER .....................................................................................59

18.1

Operation ...........................................................................................................59

18.2

Application of the computer in process controlling.............................................59

CONTROL SYSTEMS 3

LEARNING OBJECTIVES - THEORETICAL

The student should

know how the "flapper nozzle" mechanism is employed in feedback control as an


error detector and an amplifier in order to generate the required control actions.

understand how the control actions are generated in common pneumatic and electric
three-term analogue "controllers".

understand the problems which can arise from control action interaction in
pneumatic controllers and relate this to integral saturation (wind up) in particular.

understand how integral "tracking" is achieved in pneumatic systems.


understand how a "control valve" works, appreciate the significance of the terms:
a)

"flow to open" or "flow to close"

b)

"air to open" or "air to close"

and their importance with respect to actuator size, safety and process operation.

understand the meaning of the term "valve characteristic" and the considerations
which determine the choice of characteristic.

understand the rote of the positioner in a feedback control loop, how it works and its
application to:
a)

valve characterisation

b)

"split range" operation

a)

flow control

b)

feed-forward control.

know how to "size" a control valve from given process data, and the significance of
"trim" and "body" sizes.

appreciate the advantages/disadvantages of:


a)

"globe" pattern valves

b)

"butterfly" pattern valves

c)

"ball" pattern valves

when used as "final control elements".

be familiar with the characteristics, advantages and disadvantages of hydraulic and


electrohydraulic valve operators.

know how a hydraulic jet pipe controller functions.


understand the modus operandi of a digital feedback control loop and appreciate the
advantages/disadvantages over analogue loops.

understand how electronic digital computers can be interfaced with pneumatic or


electronic feedback control loops (with special reference to stepper motor
Operation).

CONTROL SYSTEMS 3

INTRODUCTION TO CONTROL MECHANISMS

The previous Training Unit "Control Systems 2" dealt with the principles of control. lt is
now necessary to understand the mechanisms which are used to apply these principles in
the plant. The three mechanisms shown below and overleaf can all be used for the same
purpose but they have different degrees of flexibility and precision.

In all three mechanisms the level in a vessel is measured by a float.


In mechanism (a) the bouyancy of the float moves the final control element, (the valve)
through a system of mechanical levers to a position at which
the inflow to the vessel equals the outflow, and the level is constant. However, the
bouyancy of the float must also balance the forces generated by out of balance pressures
in the valve. This effects accuracy of measurement as the float position is not then an
accurate measure of level.

Mechanism (b) uses pneumatic pressure to operate the valve, thus avoiding this problem,
and is more accurate.

In both of these mechanisms the proportional action factor can be adjusted to suit the
process, but they are not capable of any other form of control action.

Mechanism (c) only uses the float as a measuring device and employs
a separate mechanism to generate control action. Not only is measurement not affected
by the control valve out of balance forces, but the controller can generate any combination
of control actions. In addition, the controller can be sited at any reasonable distance from
the plant.

SELF-ACTING CONTROLLERS

For many control purposes a simple local "self-acting" (does not require any electrical,
pneunnatic or hydraulic power supply) mechanism (such as mechanism (a) on page 7) is
adequate. Some mechanisms, like the simple pressure regulator shown below, combine
measurement, generation of control action and application of control action (final control
element).

2.1

The self-acting pressure controller

The pressure in the pipeline acts on the diaphragm and is opposed by


the range spring. The movement of the final control element is proportional to the
compression or extension of the spring. The set point can be adjusted by "loading" the
spring, and the proportional action factor can
be changed by replacing the spring with a stronger or weaker one. However, this type of
mechanism is very crude, inaccurate and inflexible and is only used where this is
acceptable. Such a self-acting mechanism does have one advantage, however, in that it is
very fast acting.

2.2

Hydraulic powered controller (not seif-actino)

Sometimes the fast response of a local control mechanism is essential, such as when
controlling the speed of a gas turbine set. Then a hydraulic powered controller (not seifacting) such as that shown diagramatically
below is often used. Compared to pneumatic power hydraulic power has the advantage of
being faster acting, hydraulic fluid being a liquid, and that it is incompressible. Besides,
the delays which occur in "filling" the large
space of a pneumatic valve "motor" are avoided and can develop much greater forces for
a given size of valve motor.

Any change in pressure in the process pipeline acts an the measuring diaphragm to move
the "jet" pipe. Hydraulic fluid is pumped at very high pressure and velocity from the jet
pipe onto a "splitter" plate, which converts the velocity energy into pressure again. As the
jet pipe moves to one side or the other, the pressure at one side of the control piston
increases whilst the pressure at the other side decreases. The force generated by any
difference in pressure across the control piston balances the out of balance forces in the
final control element, so that when the jet pipe moves, the piston moves the final control
element to a new position.

2.3

Integral action

The trainee should appreciate that the control action thus generated is integral action. The
force generated by a very small change in position of the jet pipe is very large, and will
cause the valve to continue to move as long as the jet pipe remains in the displaced
position. Thus, control action continues until the pressure in the pipeline is restored
to the set value and the jet pipe returns to its equilibrium position. In theory, therefore, no
offset is possible with this type of control. In practice, however, to the extent that the outof-balance forces in the control valve (final control element) vary under different process
conditions and some offset does occur. Since control action is integral saturation (or wind
up) will occur, and the control valve will always go to the fully open or fully closed position
whenever the set value cannot be attained, such as blockage in the process line or when
the plant is shut down. This is not usually a serious disadvantage, (may actually be an
advantage sometimes) as the control action is very fast. However,
it would obviously be a mistake to use this type of controller to regulate a slowly changing
process variable.

2.4

Proportional plus integral action

10

A variation of this type of control mechanism is shown an the previous page, in which
proportional control action has been added to the basic integral action.
The pressure in the second cylinder Z1 increases when the jet pipe moves downward,
moving the piston upward against the spring.
lhe spring force will be such that there is no proportional action when the jet pipe is in the
central position (when there is no error between measured and set values). Under these
conditions the position of the final control element is determined solely by the integral
piston Z 2 .

2.5

Advantages of hydraulic controllers

The advantages of hydraulic controllers are as follows:


a) High speed response, because the liquid control medium is incompressible.
b) High power gain, because liquids can be converted readily to high pressure through the
use of various types of pump.
c) Simplicity of final actuator, normally two hydraulic lines connect the hydraulic
controller's output to a straight-type cylinder.
d) Long life as the seif-lubricating properties of most hydraulic controllers, particularly
when petroleum oils are used.

2.6

Disadvantages of hydraulic controllers

The disadvantages of hydraulic controllers are as follows:


a) Maintenance of the hydraulic fluids. The following items are required:
1. Filteration to keep the fluid clean
2. Fire-resistant fluid
b) Seals and connections must be used to prevent leakage, especially hydraulic fluid
leakage
c) Power supply.

11

REMOTE CONTROLLERS MOTION BALANCE

Modern control systems can be divided into two categories:


a) Local control
b) Remote control.
Local systems or mechanisms are most often used to regulate a variable (such as steam
pressure or turbine speed), the set point of which does not need to be changed frequently.
They have the advantage of faster response to disturbances than "remote" systems.
However, the majority of control systems today have the controller sited in a central
control room, where the process operators can vary the set points as necessary to
operate the plant. Such controllers cannot operate directly from the measurement impulse
signal, but receive analogue signals from pneumatic or electronic transducers and
generate an analogue control output to operate the final control element. Since the full
scale input and output are always the same (0.2 to 1.0 bar or 4 to 20 mA), such controllers
can be absolutely standard in this respect for every application. The design of these
controllers is based on the force balance principle described in the earlier unit on signal
transmission. However, as in the case of transmitters, early examples of pneumatic
controllers were motion balance mechanisms added to (usually large) impulse driven
indicating or recording Instruments. A typical example of a controller is shown
diagramatically overleaf.

3.1

The operation of a controller

lt can be seen that, as in the case of a motion balance transmitter, the motion of the
recorder pen is opposed by the feedback bellows (termed the proportioning bellows).

12

The only difference between this controller and a transmitting Instrument, is that the ratio
of the output to the measured value signal input of the pneumatic mechanism (the
proportional action factor) can be varied. Integral and derivative control actions can be
added to this type of controllers as outlined in the following paragraph.

3.2

Added integral control action

13

The loading spring is replaced by the integral bellows, with the result that whilst an error
persists, air continues to flow into or out of the integral bellows through the variable
integral restriction.This is to slow up the rate of flow. An error will cause the flapper to
move closer to or further away from the nozzle. Whereas the consequent increase in
pressure in the proportional bellows will be progressively opposed by the loading spring in
the case of the proportional-only-controller, the integral bellows will continue to expand, (in
the case of a positive error) causing the flapper to continue moving closer to the nozzle.
Thus, whilst the proportionalaction-only controller will develop an output pressure
proportional to the error, the proportional-plus-integral-controller will continue increasing or
decreasing the output pressure at a rate proportional to the error. Eventually the control
action will result in the error being eliminated and at this point the flapper and nozzle will
be in exactly the same position as originally, and the pressure in the integral and
proportional bellows will again be equal, though different from the original value.

3.3

Types of delay system used in pneumatic controllers

The trainee zhould appreciate that the generation of integral is achieved by building
resistive/capacitive delay of exactly the same kind that is found in the process itself. Such
delays are constructed, in terms of pneumatic equipment into the control mechanism, by
connecting fixed or adjustable restrictors in series with capacity chambers which may be
bellows in some cases.

14

3.3.1

Dead-ended delay system Pressure

3.3.2

Open-ended delay system

15

The diagrams on the previous page illustrate two types of delay system used in pneumatic
controllers to generate control actions. Capacity is measured in units of pressure rise per
unit of mass flow of air into the chamber. The resistance of the restrictors is measured in
units of mass flow of air per unit of pressure difference across the restrictor. In most
cases, it is not possible to easily change the capacity, so the time constant of the delay is
altered by varying the setting of the variable restrictor.

3.3.3

Derivative control action

The diagram below illustrates how derivative control action is added to a motion balance
controller.

The derivative restriction delays the air flow into the proportional bellows when an error
causes the flapper to move closer to the nozzle, thus, causing the pressure to increase
behind the nozzle. Therefore, the feedback bellows is slow in expanding and moving the
nozzle in the opposite direction (negative feedback). This reduces the pressure rise
behind the
nozzle. In this manner the proportional action factor is increased temporarily. This is why
derivative control action was originally termed "delayed proportional".

16

ldeally the control actions should be generated separately and then added together as
shown diagramatically below.

However, it will be obvious that the control actions are not really independent when
generated in the manner described above. For instance, the derivative restriction, in
delaying air flow into the proportional bellows, also delays the transfer of air from the
proportional bellows
to the integral bellows, and thus, changes the integral action factor.

17

FORCE BALANCE PNEUMATIC CONTROLLERS

The force balance pneumatic controller was developed to be used with analogue
transmitted signals, in a remote location, (usually a central control room). This is in
contrast to the motion balance controller which is suitable for mounting in an indicating or
recording instrument only, and is usually mounted local to the process. The force balance
controller has the same advantages of precision as the force balance
transmitter. For this reason, when local mounted controllers are required, the force
balance type is often chosen. This principle is shown below.

Instead of the input force generated by transducer, the input force is


the error between the measured value analogue pressure and the set value pressure
generated by a mechanism within the controller. Integral and derivative actions can be
added in exactly the same way as in motion balance controllers (the integral bellows is
replaced by a spring in the proportional-action-only version).

18

4.1

Proportional-plus integral action "stack" controller

Flexible diaphragms can be used instead of bellows as shown in the proportional-plus


integral action "stack" controller below. Air flows through the fixed restrictor( j) to the
flapper/nozzle and through the adjustable proportional restrictor to chamber( b) . The
pressure in chamber( b) acts on the diaphragm to produce a force on the centre post. The
difference in the measured value and set value pressures
(the error) also produces a force which acts on the centre post (chamber (C) and (d) ).
The flapper/nozzle adjusts its position to produce a pressure in chamber(e ) which applied
through the lowest diaphragm, provides
the negative feedback.

Thus the "error force" is balanced by the force generated in the proportional and feedback
chambers(b) and (e) in the same manner as the beam balance/bellows type. The
pressure in chamber (b) depends on the pressure drops through the adjustable
proportional restrictor, the fixed restrictor(ji) and the output pressure. Integral action is
generated when the adjustable integral restrictor is opened to allow air to flow into
chamber( a) , forcing the top diaphragm down. This restricts the flow out through the vent(
iii) and increases the pressure in chamber( b) .

19

4.2

Derivative control action

Derivative control action is generated in a separate unit as shown above. This avoids the
interaction between integral and derivative restrictors mentioned in connection with the
beam balance bellows type of controller. The output of the proportional and integral
controller becomes the input of the derivative unit, increasing pressure in chamber ( f) ,
acting on the upper diaphragm, forcing the "flapper" closer to the nozzle, causing the
output pressure to rise. The increase in output pressure causes air to flow into chamber
(g) slowly through the derivative restrictor. This increases the pressure in chamber( g) and
tends to oppose the increase in (f) .
If the input pressure (output from proportional and integral controller) is increasing, the
pressure in( g) will always be less. This is due to the delay caused by the restrictor valve,
and this difference will be proportional to the rate at which the input is increasing. The
output of the derivative unit will, therefore, be higher than the input by an amount which is
proportional to the rate of increase of the input. Thus, in effect, the proportional action is
increased by an amount proportional to the rate
of change of the error (derivative control action).

20

ELECTRONIC CONTROLLERS

These are composed of an operational amplifier with input and feedback resistors.
The command variable (voltage of the set point adjuster) and the controlled variable
(measured voltage) are connected to the input resistors Ril and Rj2.

5.1

P-Control unit

When there is a deviation, a voltage difference between the command variable and
controlled variable Ux is created.
A corresponding current if flows over the feedback resistor.
Output voltage Uy=if x Rf

Wiring diagram

Transfer function

21

the

5.2

I-Control unit

A deviation allows a current if for flow again.


The capacitor Cf in the feedback path charges up and causes a voltage uy .

Wiring diagram

Transfer function

22

5.3

PI-Control unit

An additional resistor is wired into the circuit in the feedback path.

Wiring diagram

Transfer function

23

5.4

PID-Control unit

Wiring diagram

Transfer function

24

The operation of the circuit is exactly the same as that explained for individual control
action. In practice, the set-point and measured value voltage are subtracted to give the
error voltage in the same way that the set-point and measured value pressure are
opposed in the pneumatic force balance controller.
The output voltage "Lly " will normally be converted to a 4 - 20 mA analogue signal for
transmission to the final control element.
In most systems, it will be converted to a pneumatic analogue signal to drive the final
control element (control valve). The current flowing in the controller circuit must not be
allowed to disturb the error measuring circuit. A "buffer" circuit is used to overcome this
problem (a differential amplifier or bridge circuit).

25

MANUAL CONTROL AND "BUMPLESS TRANSFER"

Under certain operating conditions or circumstances (for instance start-up, or for


maintenance) automatic control action is stopped. The manual regulator is used simply to
position the final control element without reference to the measured value at all. This is
"open loop control" and is usually referred to a "manual mode". When the controller is
being used in this way, lt is very difficult to stop the generation of control actions even
though the process no longer responds to them. In any case, lt is important to arrange that
the output of the controller (which is the sum of the control actions) does not change
suddenly when it is switched
back into "automatic mode". When there is no error, the output of any controller is only the
integral control action term, since there is no proportional action without an error and no
derivative action if the error is not changing. To ensure "bumpless transfer", therefore,
from manual mode operation to automatic mode, the following two points should be
observed:
a) The set point must equal the measured value.
b) The integral action must be equal to the controller output.
In order to achieve these two points, the measured value signal is connected to the set
point system inside the controller (to the set point bellows in the case of a pneumatic
controller), and the output of the controller (generated by the manual regulator in manual
mode) is connected to the Integral action generating mechanism (to the integral bellows in
the case of a pneumatic controller). Thus, when the controller is switched from manual
mode Operation to automatic mode, the set point is equal to the measured value whatever
the latter may be. There is, therefore, no error, and the integral action is equal to the
output of the controller which does not change. After transfer, the set point mechanism
can be set by the Operator to any value required. The controller will vary the output to the
final control element smoothly until this value of measured value is attained.

26

6.1

Integral tracking

The principle of bumpless transfer, from manual mode Operation to automatic mode, can
be extended to the transfer of control from one controller to another alternative or standby
controller. This is not an uncommon need in process control. Only one of two (or more)
controllers is connected to the final control element, and the other controller(s) must
"track" the output of the operational controller so that transfer can take place bumplessly.
Such an arrangement is often required where a computer is used to carry out the
functions of a number of controllers simultaneously (direct digital control - DDC). Analogue
controllers, provided as back-up, must operate in "standby" mode.

6.2

The standby mode operation

The diagram overleaf shows how this is employed by using a "stack" type pneumatic
controller. The external selector switch is positioned to select the required controller. Both
will be in automatic mode as selected by the position of the "auto-man" switch. In the case
of the standby controller, pneumatic pressure will operate the switch associated with the
set point generating unit, putting it into the "tracking" mode. Its output will then be equal to
its Input (the measured value). Hence, as far as thestack controller is concerned, the set
point and measured value inputs are exactly the same and there is no error. Supply air
pressure is also applied via the l'auto-man" switch to the pneumatically operated switch
associated with the manual valve loading unit, putting this also into "tracking" mode.

27

The output of this unit is, thus, connected via the stacked diaphragm transfer switch to the
output of the controller and into the common connection of both controllers with the control
valve. At the same time, supply air pressure is applied to a switch at the top of the
controller, which causes the Integral restrictor to be bypassed.

28

Thus, since the output to the manual valve loading unit follows the input, and is connected
in an unrestricted loop with it, the pressure in this part of the system would "float" except
that it is connected into the output of the operating controller. The output of the standby
controller, therefore, "follows" that of the operating controller. When the standby controller
is selected its set point equals its measured value and ist output of the other controller. As
a result, the process is not disturbed by the transfer operation and the "new" controller
controls the process to maintain its measured value (whether this is the same or different
from the measured value of the other controller) at the value it was at the moment of
transfer. The Operator can then change the set point to a new value in the normal way.
When the selector switch is operated to select the other controller, the set point
generating unit of the selected controller (which up to that time has been the standby)
resumes ist normal function, and the integral restrictor is no longer bypassed. The
controller, therefore, generates both integral and proportional control actions. The manual
valve loading units of both controllers remain in "track", however, since both controllers
are in "auto" mode, and both follow the output of the operating controller.
Integral tracking is very important not only for standby arrangements, but also for systems
where control is exercised on a different basis at different stages of the process operation,
particularly in batch control. The trainee should particularly notice that in the diagram on
the previous page, only the control valve, not the measurement transducer, is shown
connected to both controllers. Each controller can have its own measurement which does
not have to be the same variable. In principle, several controllers each with a different
measured value, can be connected in such a way that any one of them can be selected as
the operating controller at any time. Not all controllers operate in the same manner, and lt
may be found that a particular design will not allow bumpless transfer from one to another,
as previously described. Obviously, a good understanding of the principles of operation of
the controller mechanisms is essential in selecting the correct equipment for a particular
application.

29

ADJUSTMENT OF AUTOMATIC CONTROLLERS

7.1

Proportional action only

1.

Set the controller proportional band at 100 %.

2.

Disturbe the plant by increasing or decreasing set point value from desired value
(about 5 % of full scale).

3.

Bring the setpoint value to the initial value.

4.

Record the time which is required for disturbance to be settled.

5.

Decrease the proportional band in successive steps (50 per cent, 20 per cent, 10 per
cent) and repeat the steps 2 through 4.

6.

Proportional band is reduced so much until the oscillations do not die away.

7.

lncrease the proportional band slightly so as to give the desired stability.

7.2

Proportional plus Integral action

Setting of this type of controller is done as follows:


1.

Set the integral-action time to a maximum.

2.

Adjust the proportional band as explained for proportional-action only.

3.

Decrease the integral-action time so that it is halved at each adjustment.

4.

The integral-action time is reduced so much until oscillations occur. These


oscillations indicate that the integral-action time is reduced too far, and it should be
increased to obtain the desired stability.

7.3

Proportional plus derivative action

Optimum setting of this type of controller is obtained as follows:


1.

The derivative-action time is adjusted to the minimum.

2.

The proportional band adjustment is done as explained for proportionalaction only,


but when the oscillations occure the proportional band is not increased.

30

3.

By increasing the derivative-action time, eliminate the oscillations which are


produced by too narrow proportional band.

4.

Decrease the proportional band slightly until oscillation occurs.

5.

Increase the derivative-action time so that oscillations are eliminated.

6.

Repeat steps 4 and 5 until increasing the derivative-action time fails to eliminate
oscillations produced by decreasing the proportional band.

7.

Increase the proportional band slightly to obtain required stability.

7.4

Proportional plus integral plus derivative action

1.

Adjust integral-action time to a maximum.

2.

Adjust controller as explained for a proportional plus derivative-action controller.

3.

Adjust the integral-action time equal to the derivative-action time.

7.5

Setting of the control parameters accordinq to the transfer function (reaction


curve)

The parameters of a controller are determined as follows:


1.

Switch the controller on manual control position.

2.

Apply a sudden step change in process Input by opening or closing the control
valve.

3.

Record the reaction curve by recorder.

4.

Measure T9, Tu values from reaction curve.

5.

Calculate the controller parameter from the overleaf rules-of tumb formulae.

31

7.6

Adjusting a cascade control

1. Set both primary and secondary controllers in manual operation.


2. Control and stabilize the process by manual operation control.
3. Place the secondary controller in automatic.
4. Adjust the secondary controller as described previously.
5. Place the primary controller in automatic and adjust the controller.

32

33

INTRODUCTION TO FINAL CONTROL ELEMENTS

The final control element applies the control action, generated in the controller
mechanism, to the process. If lt does not perform its function adequately there would be
little point in measuring or in generating control action. Hence, the final control element is
in some respect the most important part of the control system. Certainly, lt is often given
too little attention, with the result that the control system performs very badly.
Processing is usually controlled by regulating the flow of gases, vapours, liquids or
fluidised powders through pipelines. For this reason, the final control element is almost
always a throttling valve. However, lt may be a thristor control element for regulating the
flow of electrical energy to an electrical machine, a damper in a ventilation or forced
draught duct, or any regulating mechanism. Whatever it is, lt will have a characteristic
relationship between its input (the control action signal generated by the controller), and
its output (the effect it has on the process) which will not usually be one of simple
proportionality. In addition, its response is unlikely to be instantaneous. There will be
delays, as there are in both process and measurement. The trainee should appreciate,
therefore, that the final control element will inevitably modify the control action generated
in such a way, that control will be different under different processing conditions. As in the
case of the process and the controller, lt is necessary to understand how the input/output
relationship of a control valve effects the closed loop.
To enable a fluid to flow through a given piping system, there must be a pressure
difference between the input and the output of the system. In order to adjust the flow rate,
the pressure difference must be manipulated (reduced to reduce flow rate and increased
to increase flow rate). Less power is required to drive the lower flow rate, and the function
of the throttling control valve is to waste the additional power available in the system from
a pump or other driving force provided to cater for the maximum flow rate required.

34

In fact much less energy is wasted manipulating a speed control system an the pump as
the final control element, rather than throttling, however, the capital Investment is much
higher.
In the case of a gas or vapour, it would always be less energy wasting to manipulate the
pressure of generation (a steam boiler for instance). However, this is not usually
practicable. In the majority of systems, therefore (with the exception of those which use a
very large amount of power, such as long distance pipelines), unneeded power is simply
wasted by throttling flow through a valve. A control valve is a device
for wasting energy so that the required measured value can be maintained. In doing so, it
must convert the analogue signal, which is the output of
the controller, into a corresponding change in the process variable.

8.1

The control valve


a) The "double-seat" qlobe valve

35

The control valve is simply a variable restrictor in the process pipeline. lt can take many
forms as shown below and on page 39 and 40.

b) The ball valve

36

c) The butterfly valve

Whatever form lt takes, however, it must have certain essential elements:


a) A means of varying the opening through the valve
b) A mechanical linkage
c) An Operator or motor, which transiates the analogue control signal into motion.

In the case of the globe valve, a) takes the form of a plug and seat, or for the "double
seat" valve two plugs and two seats. b) is a stem, and c) is a diaphragm "motor". In the
case of the ball valve, a) takes the form of a ball with a large hole drilled through it, which
seats against a seal ring. The butterfly uses a disc, pivoted about an axis, so that
in one position it presents its narrowest profile to the flow, whilst pivoted through 90 it
completely blocks the pipe.
Both the ball and butterfly valves are operated by an arm and spindle, which rotates in
bearings. The operator may also be a diphragm motor or a pneumatically, electrically or
hydraulically powered motor. All valves also have a gland at the point of exit of the
mechanical linkage.

37

The operator or motor shown on the globe valve, positions the stern. The pneumatic
analogue control signal is connected into the diaphragm chamber. The pressure acting
across the area of the diaphragm provides a force which is opposed by the spring and
whatever forces the process pressures acting on the valve plugs produce. As the
analogue signal
pressure increases, the spring is compressed and the stern moves downwards closing the
valve and "throttling" the flow of process fluid through it. Since the spring force is
proportional to the distance through which it is compressed, the movement of the valve
stern is proportional to the
change in the analogue signal pressure. The two points listed below cause errors in this
relationship.
a) The force acting on the plug of the valve, which vary with the process conditions and
with the position of the valve (these are called the out-of-balance forces).
b) "Friction" between the stern and the gland, which is inevitable.
The "double-seat" valve as shown in a), page 37, is two valves which are mechanically
linked. This is to reduce the out-of-balance forces as much as possible. Flow is upwards
through the top plug, and downwards through the lower plug. As there is a pressure loss
through the valve (that is its function), the pressure acting upwards over the crosssectional area of the plug produces a different force from that acting downwards. The
resultant out-of-balance force is, however, upwards in one case, and downwards in the
other. Provided the cross-sectional areas are carefully balanced, the resultant out-ofbalance force on the stern of the valve, which will produce error, is very small. If the spring
force and the force produced by the diaphragm are much greater than this out-of-balance
force, the error in positioning of the valve stern will be small. Some smaller globe-type
control valves are "single-seat", that is, they have
only one plug and seat. The out-of-balance forces in a single-seat valve will obviously be
much greater than in a double-seat valve.
To ensure that this does not result in large errors of positioning, a larger motor and
stronger spring are used.

38

Nevertheless, single seat valves cannot be used where there are large pressure drops
through the valve, unless a positioner is used to overcome these errors of positioning.
Single seat valves do have the advantage that they can control accurately at much
smaller openings. Any inaccuracy of manufacture, damage or differential thermal
expansion in a double seat valve is likely to result in one plug "seating" before the other.
The problem of out-of-balance forces is much worse in the case of the butterfly valve as
can be seen from the graph below. This illustrates a plot of torque (rotary force) against
the angle of opening of the valve. This clearly shows, that the force required from the
Operator merely
to hold the valve plug in position, varies from almost nothing when the valve is slightly
open, to a very large maximum valve when it is 80 open. A very considerable decrease in
the out-of-balance torque can be achieved by using a "fishtail" plug, the "tail" breaks up
the air flow in the same way as a "spoiler" on a car and, thus, reduces the pressure acting
on the face of the plug.

fishtail butterfly valve

39

The ball valve also has very variable out-of-balance torque characteristics which cannot
be overcome.

40

POSITIONERS

The effect of out-of-balance forces (or torques) together with the static friction of the
gland, introduces hysteresis and repeatability errors into the relationship between input
signal and plug position. When these errors
are unacceptable, the size of the Operator can be increased so that its force and that of
the spring are muck greater than the error producing forces. However, with the exception
of the double-seat globe pattern valve, this
may require an unacceptably large operator. In such cases, a valve positioner can be
fitted. The valve positioner is really a controller which has, as its measured value, the
position of the stem of the valve, which
is transmitted to it by mechanical linkage, and it is mounted on the frame or "yoke" of the
valve itself. The control action is usually proportional
only, so the errors are reduced but not eliminated altogether. The linkage can be
"characterised" (usually by using a cam) to change the relationship of the stem movement
of the valve to the opening between plug and seat (the advantage of this will be seen
later).
The positioner is in fact a "slave" controller to the process controller, and the two are in
cascade, as described in the previous Training Unit (Control Systems 2).
In order that no tendency to oscillatory behaviour in the slave loop of a cascade system is
being transferred into the master controller (in this case from the positioner to the process
controller) or the other way round,
it is essential to observe the rule that the speed of response in the minor or slave "loop"
should be at least 4 times faster than that in the major or master loop (the "interference"
which would otherwise occur is
similar to that which occurs between radio stations which operate on "frequencies" too
close together). lt is rarely possible to keep this rule in the case of flow control, which is
why positioners are not normally used on flow control valves.

41

10 CONTROL VALVE CHARACTERISTICS

lt will be recalled from the Training Unit, "Control Systems 2", that the relationship
between the movement of the stern of the valve, or other linkage, and the increase or
decrease in the opening between the plug and seat, is
not constant, but varies with the valve position. The valve characteristic is the name given
to that relationship.

10.1 Butterfly valve characteristics

It can be seen that when the valve is nearly wide open or closed, a given movement of the
spindle causes the opening to change in area less than the same spindle movement does,
when the valve is near the middle of its movement. If the characteristic plot were a straight
line, it can be seen that the change in valve opening for a given movement of the spindle
would be constant, no matter in what position the valve was at the time. Such a
characteristic is termed "linear", and it may be that
this is the ideal characteristic, since the valve opening will be proportional to input (the
control signal) throughout the valve movement.

42

However, two other characteristics are common, and have considerable advantages in
certain situations. These are "quick opening" and "equal percentage" characteristics.

10.2 Quick opening and equal percentaqe characteristics

The "quick opening" is seif explanatory. The "equal percentage" characteristic is such,
that at any point in the valve movement, a given stern or spindle movement results in an
equal fractional increase in
the opening. Thus, when the valve is only slightly open, the change in opening for the
given stern or spindle movement is small, but when nearly wide open the change in
opening for the same movement of the stern or spindle is much greater. One of the main
advantages of this characteristic is that the valve positioning becomes progressively less
critical as it approaches the closed position, which effectively extends the range of
opening over which the valve can operate adequately.

NOTE
The butterfly valve characteristic is a mixture of these three characteristics, and is,
therefore, not likely to be very suitable for any application.

43

10.3 The globe-type valve


The main advantage of the globe-type valve is that the plug shape can be made such as
to give any characteristic.

Examples of characteristics
a) Equal percentaqe

b) Linear

c) Quick-openinq

The flow rate through the control valve depends an the valve opening and the pressure
difference between the inlet and the outlet (the pressure drop across the valve). lt will also
depend an such factors as the "flow profile": For instance the capacity of the right angle
bodied valve (a), will not be as large as that of the straight through valve. This is because
the flow has to change direction in passing through the valve. On the other hand, the
valve (b) will have a larger capacity because of the "flared" exit.

44

All valves have a capacity co-efficient for any particular opening:

Where f is the volumetric flow rate of the process liquid through the valve with a pressure
drop of dp across it and SG is the specific gravity of the liquid.
Any given valve can best be described, therefore, by its characteristic plot of Cv against
stem or spindle movement.

45

11 INSTALLED CHARACTERISTICS

As the flow rate through the valve depends on the pressure drop
across it, lt is not necessarily true that a linear characteristic is ideal. Ideally the change in
the process variable resulting from a given change in the valve stern or spindle position
should always be the same, as only then will the control valve not modify the control
action generated by
the controller. However, in most systems the pressure drop across the valve changes with
flow through the valve, and lt is, therefore, not sufficient to ensure that the valve opening
varies linearly with stern movement. The valve characteristic must be chosen so that lt
compensates for changes in pressure drop. This can be understood by referring to
diagrams (a) and (b).

46

In the diagrams (a) and (b) the process system is such that the pressure drop across the
valve reduces as the flow rate increases and until the valve is almost fully open, it is very
small indeed. lt is obvious, therefore, that the valve should open slowly when it is nearly
closed, as it must increase its opening by a large amount for each unit movement of the
stem as it approaches the fully open position. By calculating the required Cv for different
flow rates, and plotting these in such a manner that
each increment in flow rate is proportional to the corresponding increment in valve stem
movement (and hence to the control signal) as shown, the "ideal" valve characteristic can
be found.
In this case, it can be seen that it is very similar to the equal percentage characteristic. If a
linear characteristic valve were used in this case, the proportional action factor would be
very much smaller at high flow rates than it would be at low flows, and the system could
not be tuned for
the best response at both high and low flow rates.

47

48

In diagram (c) on the previous page, the pressure across the valve is provided by a static
head rather than a pump, and although the pressure lost by friction in the pipe increases
as flow increases, the pressure drop across the valve changes little between maximum
and low flow rates.
In this case, therefore, the flow rate will be approximately proportional to the valve
opening and the ideal characteristic can be close to the linear form. lt would be just as
incorrect to install an equal percentage valve in this system as it would be to install a
linear valve in the first system.
The correct choice of valve characteristic will differ with every system, and as is seen from
the examples on the previous pages, this depends on the process characteristic as well
as the valve characteristic. The ideal installed characteristic is that which makes the
(steady state) change
in the process variable proportional to the valve stem movement.
Often the ideal installed characteristic lies between equal and linear percentage. On other
occasions it is more important to obtain the maximum rangeability, which favours the
equal percentage characteristic. This is why equal percentage characteristic valves are
chosen more often than linear ones.

NOTE
If the characteristic does not suit the process, it will only be possible to tune the system for
correct response under one set of operating conditions.
Often a compromise solution is acceptable, but in cases where both maximum
rangeability and correct tuning are essential (as when the process system must operate
over a wide range of operating conditions) the valve
positioner offers a partial solution. As mentioned in paragraph 4, the linkage transmitting
the valve stem position to the positioner can be characterised
in such a way as to compensate for mismatch between process and valve characteristics.
This is, however, a very difficult design problem, and is only resorted to in exceptional
cases.

49

12 VALVE SIZING

A control valve which is too large or too small cannot properly apply the control action
generated by the controller from measurement of the process variable. The capacity of the
valve at any opening is given by its Cv , and these must be matched to the design normal,
maximum and minimum requirements. The smallest flow rate which can be effectively
controlled will obviously be related to the leak rate across the plug and seat of a fully
closed valve. All valves leak, and the rate depends on the "shut-off" pressure drop. The
type of valve will also have a considerable influence on the leak rate. lt was previously
outlined, that differential thermal expansion may cause one of the plugs of a double seat
globe valve to seat before the other, which will obviously increase the leak rate
considerably. The seating of a butterfly valve is very poor due to the difficulty in accurate
guiding, and high leak rates are inevitable. However, a ball valve in good condition has a
very low leak rate. Obviously the condition of the valve and the severity of service affect
the leak rate and limit the lower end of the valve range. Taking all these factors into
consideration, it is probably reasonable, and, therefore, good practice, to limit the
operating range at the lower end to 10 % of the full open capacity in the case of double
seat globe valves and butterfly types, and 5 % in the case of single seat globe valves and
ball valves. Even then, it must be remembered that in order to control the process in the
face of disturbances to the steady state "design" conditions, a control valve must be
capable of increasing or decreasing the flow rate about a mean steady state at all
operating conditions.
Thus, in matching the valve capacity to the process, these limits become 20 % and 10 %.
For the same reason the upper limit is usually taken as no more than 80 % of the full open
value. Thus the rangeability (the ratio of the maximum useable Cv to the minimum
useable Cv ) lies between 5 : 1 and 10 : 1 depending on the type and condition of the
valve.

50

As outlined in the previous section, when the pressure drop across the valve rises as the
flow rate falls, this range is reduced still further, and it is not uncommon to find that the
effective rangeability is no more than 3 : 1 in practice.
One way in which errors of sizing can be corrected after commissioning of the plant is
finished (and it can be seen whether the valve is too large or too small) is to install a valve
body with a "reduced" size "trim". Most globe pattern valves can be obtained with a seat
and plug which is one size smaller than the body of the valve would normally contain.
Thus, in effect a smaller valve is installed with the option of increasing it later (it is much
cheaper and easier to replace "trim" than to remove the whole valve from the pipe once
the plant has started up). However, this option is only possible in the case of globe type
valves (single or double "seat"), as can be appreciated from the construction of other
types. This is one reason why this type of valve, though expensive, is still most widely
used. The other reasons are more precise positioning and greater flexibility of
characterisation.

51

13 MAINTENANCE OF CONTROL VALVES

lt is obvious that maintenance of the seats and plugs, the stems or spindles and the
glands of control valves must be maintained in good conditions at all times. Failure to do
so will result in poor control usually long before breakdown results. Care should be taken
to ensure that valves are installed in the pipeline in such a location (having due regard to
control requirements, especially response), that access for maintenance is good. Often, in
designing the pipework, it will be overlooked that sufficient clearance must be provided,
not only for valve access, but to withdraw its internal parts (trim for instance), for
inspection and repair without removing the body from the pipeline (which can usually only
be carried out at a shut-down). This often prevents adequate nnaintenance with
disasterous results.

52

14 DYNAMIC RESPONSE

lt was mentioned earlier in this Training Unit that in addition to modifying the steady state
relationship between control action signal and process variable, the control valve
introduces delays because of the time it takes for Instrument supply air to "fill" or "empty"
the !arge capacity of the Operator or motor. This is likely to be a problem with
"fast" process loops such as flow and some pressure controls, particularly where the
controller is located in a central control room a long
distance from the control valve (which is common today). A solution is often found by
fitting an amplifier relay similar to that used in the controller (and usually referred to as a
"booster" relay) at the valve itself. This isolates the controller output signal from the valve
motor, so that air from the controller travelling through the considerable restriction of the
signal line from the remote control room, has only to fill the very small capacity of the relay
chamber, with the result that hardly any delay is generated in that part of the system. Air
to fill the motor chamber is supplied at high pressure (5 - 7 bar) close to the valve and is
subject to much less restriction. Thus the delay would still be too great, high pressure
motors can be obtained, with the advantage that the same force can be generated with a
smaller capacity, using higher pressure air.

53

15 CONTROL VALVES AS PLANT EQUIPMENT

The control valve is the only part of the control loop which is also an integral part of the
plant itself (measurement sensors which are also installed into the plant, do not take any
part in Operation). lt is to
be expected, therefore, that the proper specification of a control valve is as much
concerned with process and mechanical considerations as
with control.
A control valve must never be used as an isolating valve. lt is not designed for this
purpose, and its seats and plugs will probably be damaged to such an extent that control
will be badly affected if it is used as a shut-off device. Where the plant operation requires
them, isolating valves must be installed in addition to control valves!
The direction of flow through the valve is of great importance, except possibly in the case
of a simple butterfly valve. The seal an a ball valve must face the flow. The tail of a fishtail
butterfly must be downstream when the valve opens. In the case of a globe valve, both
the operator and the capacity of the valve will be affected if the flow
direction is changed.

54

The diagram on the previous page represents a single seat globe-type control valve. If
flow is downward through the plug and seat, the upstream pressure, which must always
be the higher, acts on the smaller cross-sectional area as pressure on the valve stern
does not act downwards. Due to this, the out-of-balance forces can be reduced. In fact lt
may be possible, if the stern is of a Iarge enough diameter, to virtually eliminate them.
Thus, the actuator can be smaller than would be required if the same valve were installed
so that flow is upwards, in which case the highest pressure acts on the largest area, and lt
is not possible to reduce the out-of-balance forces.
Other considerations may dictate that flow must be in the direction which maximises outof-balance forces, and this may be a reason for selecting a double seat valve, despite its
greater cost and leakage potential. For instance, it may be vital that if the pneumatic air
supply fails, the valve should open. The trainee will appreciate from the diagrams of the
two right-angled body valves (one having a venturi high recovery outlet throat) why the
capacity of valves may vary considerably according to direction of flow.
In considering the mode of failure of the valve it should be noted that the plug is only
driven in one direction by pneumatic force. The spring provides the force to drive lt in the
other direction. Apart from the direction of flow through the valve, there is another option.
Spring or pneumatic pressure is employed to open or close the valve if there is a failure in
the pneumatic supply pressure. Finally, lt may be important to consider the result if the
control signal should fall but not the pneumatic air supply (if electronic control and
transmission is used or if either a positioner or "booster" relay is used at the valve). The
answer to all these questions will decide the size of actuator required.
The ,process conditions under which the valve has to operate will decide the type of
construction appropriate for the internal construction of the valve.

55

For instance, the stem may be top guided only, or may be supported top and bottom if
turbulence or "flashing" (sudden change from liquid vapour as the pressure drops through
the valve) is expected. The effect
of these options an out-of-balance forces and capacity, can be appreciated from the
diagrams below.

The body construction material and particularly the seat and plug material, are of
considerable importance. One reason why it is not desirable for the valve to be oversized,
is that it would then be operating with a small clearance between the plug and seat for
much of its life. This will inevitably accelerate wear by erosion, particularly where the
process fluid contains solid matter (dirt) or flashing may occur, and this in turn will
adversely affect control.

56

16 CHOICE OF VALVE ACTION

The action of a valve, whether "air to open" or "air to close", depends on the nature of the
process. For example, if the valve is used to control the flow of steam to a heater which
keeps the temperature of process material in a liquid state, then lt is essential for the valve
to open if the air supply to it falls, i. e., an air-to-close valve would be used.
In this event the amount of steam flow may be more than that needed, but the effect of
overheating would be less serious than the result of the material solidifying in the process
pipes and vessels.

There are two types of valve acting:


a) Air-to-close or reverse acting
b) lncrease in the output pressure from the controller will cause the valve to close and so
reduce the flow through the valve.
c) Air-to-open or direct acting
d) An increase in output pressure of the controller will cause the valve to open and
increase the flow through the valve.

57

17 SPLIT-RANGE OPERATION

It has been explained that the useful range of operation of any control valve is often quite
small. Fortunately most process plants tend to be operated close to design conditions.
However, it is not uncommon to find that the rangeability available, when all the
constraints have been allowed for in the design and selection of equipment, is not
adequate. In such cases it is possible to install two control valves in parallel, one normally
being larger than the other, to greatly extend the operating range. Each valve must have a
positioner, which can be adjusted so that it operates over its full range of opening in
response to only a part of the control signal range:

The operating ranges of the two valves overlap to ensure that there cannot be a part of
the total operating range where no increase in capacity occurs for an increase in control
signal. However, selection of the most appropriate valve characteristic can pose a
considerable problem. In very special cases, it would be possible to characterise both
valve positioners so that the combined character of the two valves would be close to ideal.
This is so difficult in practice that it is
rarely carried out, and increasing rangeability is normally achieved at the expense of
installed characteristic and, therefore, poor tuning over the operating range.

58

18 18. DIGITAL COMPUTER

18.1 Operation

In control application the measured analogue quantities are generally converted into
digital form.
For example, in temperature nneasurement the millivolt obtained by a thermocouple is
converted to 4 to 20 mA then into digital signal. Several types of analogue digital
converters have been devised. In one type, the input analogue quantity (e. g. voltage) is
applied to an oscillator where the frequency is a function of the Input voltage amplitude.
The digital output is obtained by counting the number of cycles occuring in a fixed period.
After converting the input analogue
signal into a digital signal, the digital signal is then fed into the process computer which
has been appropriately programmed. The output of the computer is converted to a
suitable analogue signal for the actuator.
lt is possible to transmit the output of the computer to the final control element in digital
form, to drive a stepper motor for instance.

18.2 Application of the computer in process controlling

a) Multi-variable process control


b) lt is most economical to use a single digital computer rather than a number of
analogue controllers.
c) The computer can be used to carry out the control functions of several interconnected
control loops corresponding to interrelated process variables which is not possible with
single-loop analogue controllers, enabling the Optimum combination of process
variables to be obtained.

59

EE 066

Control Systems 3
Theoretical Test

60

EE 066
CONTROL SYSTEMS 3
TEST 1

1.

What effects the accuracy of measurements in a ball float level controller with direct
connection to the control valve?

2.

Explain the self-acting control mechanism.

3.

Draw the diagram of the self-acting pressure controller.

4.

Explain in the diagram below:


a)

How the set point is adjusted

b)

How the proportional action factor changes.

5.

What is the advantage of the self-acting controller?

6.

State why a hydraulic power system is faster than pneumatic power.

61

7. State the disadvantage of a hydraulic powered controller.

8. Draw a dead ended delay system used in pneumatic controllers.

9. State which type of controller is shown below.

10. Explain how the derivative action is generated in a proportional controller.

62

EE 066
CONTROL SYSTEMS 3
TEST 2

1. Draw the diagram of a force-balance three term pneumatic controller.

2. What is the relationship between the output and the Input of the following diagram?

3. Draw an electrical circuit for the following pneumatic dead ended delay system.

63

4. Draw the transfer function of the following circuit.

5. Draw the diagram of an electrical integrator circuit.

6. Name an electronic buffer-circuit.

7. State under which operating condition the controller is used as a manual regulator
(manual mode).

8. State the two points which must be observed when transferring from manual mode of
controller to automatic mode until a bumpless transfer takes place.

9. Why is it necessary when a computer is used to carry out the function of a number of
controllers simultaneously, the analogue controllers, provided as back-up must
operate in "stand by" mode?

10. State the function of final control elements.

64

EE 066
CONTROL SYSTEMS 3
TEST 3

1. Name two final control elements.

2. State the three essential elements of a control valve.

3. State two factors which cause error in positioning of a control valve stern.

4. Which one of the following control valves has the least out-of-balance forces?
a) Double-seat globe valve
b) Single-seat globe valve.

5. In order to reduce the out-of-balance forces in a control valve using a diaphragm type
operator, what relation must exist between the spring force and the out-of-balance
forces?

6. Which one of the following control valves is used to control the flow at small openings?
a) Single-seat globe
b) Double-seat globe valve.

7. Does a problem of out-of-balance forces exist in butterfly control valves?

8. Which type of plug must be used in a butterfly control valve in order to reduce the outof-balance torque?

65

9. What overcomes the effect of the out-of-balance forces (or torque) together with the
static friction of the gland?

10. How can the relationship of the stem movement of the valve to the opening between
plug and seat be "characterised"?

66

EE 066
CONTROL SYSTEMS 3
TEST 1
(Solution)

1. The forces generated by out of balance pressure in the valve.

2. They are simple local controllers which do not require any electrical, pneumatic or
hydraulic power supply.

3.

4. a) The set point can be adjusted by the "Ioading" range spring. b) By replacing the
range spring with a stronger or weaker spring.

5. lt is very fast acting.

67

6. Because hydraulic fluid being a liquid, cannot be compressed and the delay which
occurs in "filling" the large space of a pneumatic valve "motor" is avoided.

7. a) Maintenance of the hydraulic fluids.


b) Leakage - care must be taken with seals and connections.
c) Requirement of power supply.

8.

9. Motion balance three term pneumatic controller.

10. By adding a restriction between the output of the amplifier relay and the feedback or
proportional bellows.

68

EE 066
CONTROL SYSTEMS 3
TEST 2
(Solution)

1.

2. The output signal is the proportional plus derivative of the Input signal.

3.

69

4.

5.

6. Differential amplifier or bridge circuit.


7. In start-up or during maintenance.
8.

a) The set point must be equal the measured value.


b) The integral action must be equal to the controller output.

9. Because the transfer can take place bumplessly.


10. The final control element applies the control action, generated in the controller
mechanism, to the process.

70

EE 066
CONTROL SYSTEMS 3
TEST 3
(Solution)

1.

a) Control valve
b) Damper

2.

a) A means of varying the opening through the valve.


b) A mechanical linkage.
c) An Operator or motor, which translate the analogue control signal into motion.

3.

a) Out-of-balance forces.
b) Friction between the stern and the gland.

4. Double-seat globe valve

5. The spring force must be a lot larger than the out-of-balance forces.

6. Single-seat globe valve

7. Yes

8. "Fishtail" plug

9. Valve positioner

10. By using a valve positioner.

71

KEY TO EVALUATION

PER CENT

MARK

88 100

75 87

62 74

50 61

0 49

72