Vous êtes sur la page 1sur 25

Competences

The smart choice of Fluid Control Systems


Contents

1. Open-loop and closed-loop control Page 4


1.1. Function and sequence of an open-loop control system Page 4
1.2. Function and sequence of a closed-loop control system Page 5

2. The control-loop Page 6


2.1. The elements of the control loop Page 7
2.2. The controlled system Page 8
2.3. The controller Page 11
2.3.1. On/off controllers Page 12
2.3.2. Continuous-action controllers Page 14

3. Adapting the controller to the controlled system Page 18


3.1. Selecting the suitable controller Page 19
3.2. Determining controller parameters Page 19
3.2.1. Setting guidelines in line with Ziegler and Nichols Page 19
3.2.2. Setting guidelines in line with Chien, Hrones and Reswick Page 21

4. Rating and selection of control valves Page 22


4.1. Introduction and definition of terms Page 22
4.2. Rating and selection of pilot valves Page 22
4.3. Rating and selection of control valves Page 23
4.3.1. Fluidics fundamentals Page 23
4.3.2. Characteristic curves Page 24
4.3.2.1. Valve characteristic Page 24
4.3.2.2. Flow characteristic and rangeability Page 24
4.3.2.3. Operating characteristic and pressure ratio Page 25
4.3.3. Rating and selection Page 26
1. Open-loop and closed-loop control

The terms “open-loop control” and The room temperature PV (output vari- the control unit (e.g. in the form of
“closed-loop control” are closely inter- able of the open-loop control system) characteristic curves). Use of such a
linked. is to be maintained at a constant value control unit allows the influence of the
by adjusting the electrically operated outside temperature on the room
mixer valve and thus, the temperature temperature to be eliminated.
1.1. of the heating supply line or radiator.
Function and Besides the outside temperature, the
sequence of an open- If the outside temperature changes, example shown in Figure 1 also con-
loop control system the room temperature changes as a tains other disturbance variables which
consequence of this. The outside tem- also affect the room temperature:
An open-loop control system is char- perature is referred to as the disturb- – opening a window or a door
acterized by the fact that one or more ance variable and identified with the – changing wind conditions
input variables of a system influence letter z (z1 in the example). – presence of persons in the room.
its output variables in accordance with The task of the open-loop control sys- Since it is not detected by the control
the system’s own interrelations. tem is to counteract the influence of unit, the effect of these disturbance
the “outside temperature” disturbance variables on the room temperature is
One everyday example of an open- variable. not compensated for by the open-loop
loop control system: For this purpose, the outside tempera- control system.
ture is measured via the outside temp- The use of such an open-loop control
The inside temperature of a room is to erature sensor. system is practical only if it can be as-
be maintained at a constant value as a The mixer valve is adjusted or the sumed that there is a low influence of
function of the outside temperature. temperature of the radiator varied via (secondary) disturbance variables.
the control unit.
The interrelationship between the out- The block diagram in Figure 2 shows
side temperature and the heating out- the open action sequence characteris-
put required for maintaining a constant tic of an open-loop control system.
room temperature must be stored in
Disturbance variable z2 PV
Opening the window Room
temperature
Disturbance variable z1
Change in outside
Disturbance variable z4
temperature Window
Radiator Number of persons in the room
supply line

Disturbance variable z3
Heating

Outside Wind conditions


temperature
sensor Disturbance variable z2
Opening the window
return line

Disturbance variable z1 Room


Heating

Change in outside temperature


temperature Outside Control unit Mixer valve Radiator Room PV
M temperature
sensor
Electrically
Control operated
unit mixer valve
Figure 2: Block diagram of the open-loop control system for room temperature
Heating
boiler

Figure 1: Open-loop control of the inside


temperature of a room

4
However, if the effect of the other dis- Disturbance variable z1
Change in outside temperature
turbance variables is so strong that it Disturbance variable z2
Opening the window
also needs to be compensated for, it
Disturbance variable z3
becomes necessary to control the room Wind conditions
Disturbance variable z4
temperature on the basis of a closed- Number of persons in the room
Room
loop control system. The block diagram Set-point
temperature
Controller Mixer valve Radiator Room PV
Set-point value +
in Figure 3 shows the closed action programm. –

sequence that is typical of a closed- Actual value

loop control system. Outside


temperature
sensor

Figure 3: Block diagram of the closed-loop control system for room temperature
1.2.
Function and
sequence of a closed- In this case, the effects of disturbance with the set-point value. On the basis
loop control system variables on the room temperature of the comparison between set-point
should be eliminated. and actual value, the controller adjusts
The fundamental difference with res- the temperature of the heating supply
pect to an open-loop control system is The effects of all disturbance variables line or radiator via the mixer valve until
that the output variable of the system influencing the room temperature: the required room temperature is
(the actual value) is constantly meas- – change in outside temperature reached.
ured and compared with another vari- – opening a window
able (the set-point value or the refer- – changing wind conditions The table below contains application
ence variable). If the actual value is are registered by measurement of the recommendations for open-loop and
not equal to the set-point value, the room temperature and comparing this closed-loop control.
controller responds to this. It changes
the actual value by adjusting it to the Open-loop control Closed-loop control
set-point value.
Application • If no or only one essential • If several essential disturb-
and measurable disturbance ance variables are present.
One everyday example of a closed-
variable is present. • If disturbance variables
loop control system:
cannot be detected or can
The inside temperature of a room is to
be only poorly detected with
be maintained at a preset temperature.
measuring systems.
• If unforeseeable disturbance
variables may occur.
Advantages • Low implementation effort. • Disturbance variables are
Disturbance variable z2 PV
Opening the window Room • No stability problems detected and compensated
temperature
Disturbance variable z1 due to the open action for.
Change in outside
temperature Window sequence. • The preset value (set-point SP)
Setpoint Radiator
is more precisely complied.
supply line

adjuster
Room
Heating

temperature
Set-point value

sensor Disadvantages • Disturbance variables • The equipment effort and


Actual value

return line

that occur are not detected complexity are greater than


Heating

automatically. with open-loop control.


M

Control Electrically • Measurement is required • A measurement is always


operated
unit
mixer valve for each disturbance variable required.
Heating
to be compensated for.
boiler • All interrelationships of the
system to be controlled must 5
Figure 4: Closed-loop control of the be known in order to design
inside temperature of a room
the control unit optimally.

Table 1: Application, advantages and disadvantages of open-loop and


closed-loop control
2. The control loop

Closed-loop control is a process that One example: Figure 5) so that the process value ad-
is used in more than just technical The fluid level in a tank is to be main- justs to the set-point value, i.e. so that
applications. Closed-loop control tained at or adjusted to a preset value. the control deviation becomes zero. If
systems run virtually everywhere and In order to implement a closed-loop the process value is higher than the
always. The process of setting the control system, it is necessary to con- set-point value, the control valve must
required water temperature when tinually measure the filling level in the be closed further. If the filling level is
showering or complying with a speed tank (process value). too low, the valve must be opened
limit when driving a car involves This is done here, for example, by an wider.
closed-loop control. These two exam- ultrasonic level transmitter. The process
ples demonstrate the task of a closed- value is constantly compared with the Control deviations in a control loop
loop control system: adjusting, a preset target filling level (set-point va- are caused by two factors:
specific variable such as temperature, lue), which is set e.g. on a control unit – disturbance variables
speed, flow rate or pressure to a re- via buttons or selector switches. – changes in the set-point value.
quired value. The comparison between process value
and set-point value is performed by the In our example, the following two dis-
In principle, a closed-loop control pro- controller. If a deviation occurs be- turbance variables may occur:
cess appears to be a very simple one. tween the process and set-point value – outflow from the tank, occurring
However, when implementing techni- (control deviation), the controller must abruptly due to opening of one or
cal closed-loop control systems, prob- respond to it. The controller has to ad- more ON/OFF valves
lems are very quickly encountered. just a suitable final control element or – slow filling-level change due to
The precondition for correct function- actuator (a continuous-action control evaporation of the fluid from the
ing of a closed-loop control system is valve in the example shown in tank.
the interplay of the individual compo-
nents involved in a closed-loop control
system. The totality of components of
a closed-loop control system is referred
to as the control loop. In the following,
the control loop is explained in further
detail.

A control loop consists of the following


components:
Contoller
– the measuring instrument or sensor Set-point value = required value
for detecting the variable to be Controller
Sensor
Process value = measured value
controlled
– the controller, the core of the Sensor
Continuous-
closed-loop control system action level
transmitter
– the system to be controlled (this Continuous-
part is referred to as controlled action control
valve Tank
system).

ON/OFF
valves

Controlled system

6
Figure 5: Hardware representation of a closed-loop filling-level
control system
2.1. SP: Set-point value or reference The basic structure of this block dia-
The elements of the variable (required filling level) gram corresponds to that of the closed-
control loop PV: Process value or controlled loop room temperature control struc-
variable (measured filling level) ture. Thus, the following general block
Block diagrams are used to represent PVd : Control deviation diagram (shown in Figure 7) can be
control loops. This mode of represen- (actual value – set-point value) used to summarize the closed-loop
tation affords the advantage that it CO: Manipulated variable or con- control engineering.
concentrates on the control-engineer- trol output (output value of In this case, it is assumed that the ac-
ing problem. The interplay of the in- the controller) tion point of the disturbance variables
dividual components of the control z1: Disturbance variable 1 does not always need to be at the out-
loop is represented graphically. For (outflow from the tank) put of the controlled system, but that
the example of a closed-loop filling- z2: Disturbance variable 2 the action of the disturbance variables
level control system, the block diagram (evaporation of fluid from the can be converted to this point.
looks as follows: tank) Usually however, the simplified block
diagram shown in Figure 8 is used.
z1 z2
The disturbance variables are combi-
Reference
point
ned and their action point is at the
SP PVd Controller CO Continuous- Tank PV output of the controlled system.
action position-
ing valve Block diagrams are used to create a
+ closed-loop control engineering model
Measured of a real system. The main compo-
value pick-up
plus nents of the control loop are represen-
transmitter
ted by function blocks, frequently also
referred to as transfer elements. The
Figure 6: Block diagram of the closed-loop filling-level control system
functional relationship between the in-
dividual blocks and in regards to the
z1
environment is shown by action lines.
Disturbance
var. system 1 Each function block is characterized
+
by the dependence of its output signal
+
z2 Disturbance
var. system 2
on the input signal. This dependence
Reference
point is described by the response. There
SP – PVd Controller CO Final control Controlled + + PV are numerous possible ways of repre-
element/ system
+ actuator senting the response. The most con-
PV ventional way is stating the step re-
Measured sponse or transfer function. It is plot-
value pick-up
plus ted as a simple timing diagram in the
transmitter
relevant function block.
Figure 7: General block diagram of the control loop The step response is the characteristic
of the output signal, which occurs
when the input signal changes abrupt-
ly as a function of time. The transfer
Control system, controller
Reference
z function (designated h(t)) is the step
point response standardized with respect to
SP – PVd Controller CO Final control Controlled + + PV
element/ system the magnitude of the input step or in-
+ actuator
put signal (h(t) = PVo(t)/PVi0).
PV

Measured
value pick-up
plus
transducer 7
Figure 8: Simplified general block diagram of the control loop
2.2.
PVo
The controlled system
PVi PVo
Input signal Output signal
In order to select a suitable controller
PVi Time PVo and be able to adapt it to the control-
led system (the system or equipment
PViø
to be controlled), it is necessary to
have precise information on the behav-
Time Time
ior of the controlled system. Factors
that must be known include to what
Figure 9: Step response of a transfer element
extent and in what timeframe the out-
put signal of the controlled system re-
z2 z1
sponds to changes of the input signal.

co xo xo
SP – PVd CO PV Real transfer elements differ from ideal
?
+ Time Time Time ones by virtue of the fact that they al-
PV Controller Final control element/ Controlled system most always feature a time-delayed
actuator
xo response. This means that a certain
Time
time elapses until the output signal re-
Measured value pick-up
sponds to a changing input signal.
+ transmitter

Figure 10: Signal flow diagram of the closed-loop filling-level control system Controlled systems can be subdivided
into two categories in terms of their
time response or steady state condi-
If, in the block diagram, we replace the The signal flow diagram or the block tion:
individual control loop elements or diagram is an important aid to design-
function blocks by the form of repre- ing control loops and for adapting the Controlled systems with compen-
sentation shown in Figure 9, we obtain controller to the controlled system. In sation:
the signal flow diagram of the control many cases, adaptation of the control- In the case of controlled systems with
loop. Figure 10 shows the signal flow ler is one of the most demanding tasks, compensation, the output variable of
diagram of the closed-loop filling-level and one that requires basic knowledge the system reassumes steady-state
control system. of controlled systems and controllers. condition within a specific period. One
This is covered in the following sec- example of a controlled system with
tions. compensation is the flow rate in a pipe.
If the degree of opening of a continu-
ous-action control valve is changed, a
constant flow rate is established after
a specific period assuming constant
pressure conditions. The transfer ele-
ment shown in Figure 11 symbolizes a
controlled system with compensation.

PVo
PVi PVo
Input signal Output signal

PVi Time PVo

PViø

Time Time

8
Figure 11: Transfer element with plotted step response of a controlled system with
compenastaion
Controlled systems without com-
pensation: PVo

In the case of controlled systems with- PVi PVo


Input signal Output signal
out compensation, there is no steady-
state condition. Even with a constant PVi Time PVo
input variable (greater than 0), the out-
PViø
put signal changes at a constant rate
or acceleration. In the example of
Time Time
closed-loop filling-level control in a
tank, this relates to a controlled system
Figure 12: Transfer element with plotted step response of a controlled system without
without compensation. If the valve for compensation
filling the tank is open and the ON/OFF
valves are closed, the filling level in-
creases continuously without assuming Designation Transfer element Application
a steady-state condition. In the case
P-system Closed-loop pressure
of real controlled systems without com- PVo control with fluids
PVi PVo Closed-loop flow-rate
pensation, there is generally a limita- Input signal Output signal control with fluids
tion: in the example of closed-loop
filling-level control, this results from Time

overflowing of the tank.


The transfer element shown in Figure 1st-order time-delayed
PVo
Closed-loop pressure
system control with fluids
12 symbolizes a controlled system PVi PVo and gases
Input signal Output signal Closed-loop flow-rate
without compensation. control with gases
Closed-loop rotational-
When selecting and setting the con- Time speed control

troller, the aspect of whether the con-


trolled system is a controlled system 2nd-order time-delayed Closed-loop
PVo
with or without compensation is of system
PVi PVo
temperature control

Input signal Output signal


crucial importance.

Time
The most frequently occurring control-
led systems and their transfer func-
3rd-order time-delayed Closed-loop
tions are described below in greater system
PVo temperature control
PVi PVo (steam via heat
detail. Table 2 provides an initial over- Input signal Output signal exchanger)

view.
Time

Tt-system Closed-loop conveying


PVo quantity control on
PVi PVo conveyor belts
Input signal Output signal

Time

T1 1st-order time-delayed Closed-loop pH control


PVo Closed-loop conductivity
system
PVi PVo control
Input signal Output signal Mixing two fluid streams
in one pipe

Time

I-system Closed-loop filling level


PVo control
PVi PVo Closed-loop conductivity
Input signal Output signal control

Time

Table 2: Overview of typical controlled systems

9
The P-element:
On the P-element or proportional ele- PVi PVo
Input signal Output signal
ment, the output signal follows the in-
put signal directly, with no time delay.
Input and output signal are propor-
tional to each other. There is no time ∆PVi ∆PVo = Ks· ∆PVi
delay. Figure 13 shows the behavior or
step response of a P-element. Time Time

Input step Step response

Figure 13: Step response of a P-element

The 1st-order time-delay element:


PVi PVo
On the 1st-order time-delay element, Input signal Output signal

the output signal follows the input sig-


nal with a time delay. In this case, the
output signal changes immediately,
but increases continuously to the full ∆PVi ∆PVo = Ks· ∆PVi

scale value with a time delay.


Time Time
An analog response is shown by the T1

voltage through a capacitor when Input step Step response

charging via a series resistor.


Figure 14: Step response of a 1st-order time-delay element

The 2nd-order time-delay element:


The 2nd-order time-delay element is a PVi PVo
Input signal Output signal
controlled system with two delays
(two 1st-order time-delay elements Flex tangent

connected in series). 2nd-order time-


delay systems are characterized by ∆PVi Flex point ∆PVo = Ks· ∆PVi
three parameters, the system gain Ks
Time Time
and the two time constants Tu and Ta. Tt Tu Ta
Unlike the 1st-order time-delay ele- Input step Step response
ment, the step response is initially
Figure 15: Step response of a 2nd-order time-delay element
characterized by a horizontal tangent,
features a flex point and then runs
asymptotically towards the full scale
value.
Real 2nd-order time-delay elements
are controlled systems with two (ener-
gy) stores, such as those occurring
when tempering a tank via a heat ex-
changer.

What follows is a consideration of the


controllability of controlled systems PVi PVo
Input signal Output signal
with compensation (with time
delays/dead time). These controlled
systems can be approximated by the
approximate model shown in Figure 16. ∆PVi ∆PVo = Ks· ∆PVi

Time Time
Tt Tu Ta
Tte
10 Input step Step response

Figure 16: Approximate model for controlled systems with compensation and
dead time
On the basis of practical experience, it
Tte/Ta Controllability Control engineering
is possible to make an approximate
effort
statement on the controllability of a
controlled system with compensation < 0.1 Very well controllable Low
and equivalent dead time via the ratio 0.1 … 0.2 Well controllable Moderate
Tte/Ta. 0.2 … 0.4 (Still) controllable High
0.4 … 0.8 Poorly controllable Very high
> 0.8 Barely controllable Special measures or controller
structures required
Table 3: Estimation of the controllability of a system with compensation

PVi PVo
The I-element: Input signal Output signal

On the I-element (integral element),


the output variable is proportional to PVo = Ks· PVi · t
the time integral of the input variable.
In the case of a constant input varia- ∆PVi

ble, the output variable increases con-


Time Time
tinuously.
Input step Step response

Figure 17: Step response of an I-element

PVi PVo
The lag element: Input signal Output signal
On the lag element, there is a similar
behavior to that on the P-element with
system gain 1 (Ks = 1). However, the
lag element does not respond immedi- ∆PVi ∆PVo = ∆PVi

ately to changes of the input value. In


Time Time
the case of a stepped change of the Tt
input variable PVi, the same stepped Input step Step response

change of the output variable occurs


Figure 18: Step response of a lag element
upon expiration of time Tt.

2.3. In order to meet these requirements


The controller when designing a closed-loop control
system or a control loop, the appro-
A closed-loop control system must priate controller must be selected for
ensure that the process value is equal the given controlled system and must
to the set-point value or is adjusted to be matched to the controlled system.
the set-point value under all circum- In addition to knowledge of the dy-
stances, even under the influence of namic and static behavior of the con-
disturbance variables, i.e. the control trolled system, this also necessitates
deviation must be 0. knowledge of the characteristics of
In addition, the closed-loop control the individual controller versions or
system must operate stably and the controller types.
process value may neither drift from
the required operating point nor oscil- In the following, the individual control-
late around it as a consequence of a ler types are described in greater de-
change of disturbance variable or set- tail.
point value. Controllers can initially be subdivided
into two main groups:
– continuous-action controllers and 11
– on/off controllers.
2.3.1. On/off controllers Figure 19 illustrates the principle of The described two-point controller can
operation of a two-point controller. be used, e.g., for tempering a room.
On/off controllers are frequently used The control response of the two-point
in temperature, filling level, pressure, Ideally, the switch-on point and switch- controller is illustrated below on the
pH value and conductivity control off point of the two-point controller basis of this example.
loops. On/off controllers are also used would coincide. In practice however,
in day-to-day applications or appli- the switch-on point and switch-off The diagram in Figure 21 shows the
ances such as automatic coffee ma- point are reciprocally offset. The inter- principle characteristic of the process
chines, irons, refrigerators or building val between the two switching points value (temperature in the room).
heating systems. is referred to as (switching) hysteresis
and is identified with PVh. If the pro- The upper diagram in Figure 21 shows
Two-point controller: cess value PV drops below the preset the process value (temperature) and
A two-point controller operates in the set-point value SP minus half the hys- the set-point value (desired tempera-
same way as a switch. Its output can teresis, the output of the controller is ture) as a function of time and the lower
assume only two states: switched on switched on (CO = 100 %). diagram shows the manipulated vari-
or switched off. This means that the able (control output). At instant t = 0
controlled final control element or ac- If the process value rises above the (switch-on instant of the closed-loop
tuator is either switched on or opened set-point value plus the hysteresis, the control system), the controller switches
or is switched off or closed. A two- output is switched off again (CO = 0 %). its output on and thus opens the heat-
point controller can be seen as a Half of the hysteresis prevents a con- ing valve. For the period of the dead
P controller (continuous-action con- stant switching on and switching off at time (Tt), the actual value initially does
troller) with very high gain. the same point owing to very minor not change. After the dead time ex-
A two-point controller can only be disturbances. pires, it increases. The characteristic
used in conjunction with time-delayed of the process value corresponds to
systems (1st-order or 2nd-order time- the step response of the controlled
delay systems) or controlled systems system (characteristic of the process
without compensation (I-systems). value shown with a dashed line).

Control output CO
CO = 100 Control Set-point value
Output output
switched on Process value
Temperature
sensor

CO = 0 Heating
Output system
switched off SP Heating valve Room
Process value PV
Set-point value
Radiator
PVh
Hysteresis

Figure 19: Principle of operation, characteristic of a two-point Figure 20: Hardware representation of a closed-loop
controller temperature control system

Process value Ts
displacement of the

Set-point value
Tt
Tt Tt
process value
Peak-to-peak

Set-point
+ 0,5* Hysteresis

Set-point
Set-point
– 0,5* Hysteresis
Tt Tt Tt

Time

Control
output CO
100%
On

0%
Off
Time

12 Figure 21: Process value and manipulated variable or


controller output of the closed-loop temperature control
system as a function of time
Tt: Dead time
Ts: System time constant
If the process value reaches the set- The peak-to-peak displacement of the
point value plus half the hysteresis, actual value is chiefly dependent on On most modern two-point controllers,
the controller switches its output off two aspects: it is possible to set the circuit function
and thus closes the heating valve. For – the controller hysteresis (this can so that they can be used for both ap-
the duration of the dead time, the pro- generally be set on the controller). plications, depending on the setting.
cess value initially still rises. After the The peak-to-peak displacement in-
dead time elapses, it drops. If the pro- creases with increasing hysteresis 3-point controller:
cess value drops below the set-point – the system time constant (this ge- A 3-point controller is a switch, like a
value minus half the hysteresis, the nerally cannot be changed and is two-point controller. In contrast to the
controller switches its output back on determined by the structure of the 2-point controller, its output may as-
and the heating valve opens. In the controlled system). sume three switch positions. 3-point
same way as with the rise in tempera- The lower the system time constant, controllers are used, for example, in
ture, it initially still drops before it rises the greater will be the peak-to-peak the following applications:
again, due to the time delay or dead displacement.
time of the system. – Closed-loop temperature control
The control cycle then starts again. The lower the time constant of the con- Closed-loop temperature control of
The dead time of the controlled system trolled system (system time constant) a room on which the disturbance
and the hysteresis of the controller and the hysteresis of the two-point variables can be counteracted by
result in periodic fluctuation of the pro- controller are, the more frequently the heating and cooling.
cess value about the set-point value. controller will switch. Depending on – Closed-loop pH control systems
the design of the controller, the final For neutralization of media. The
In order to assess the controllability of control element/actuator and the sys- required pH value can be set by
a controlled system with compensation tem, increased wear on the control adding acid or lye.
and dead time using a two-point con- loop components will occur in the case – Control of motorized actuators
troller, it is possible to use the ratio of of frequent switching. Consequently, a Motorized actuators operate valves
equivalent dead time to system time two-point controller cannot be used such as butterfly valves, ball valves
constant (Tt/Ts). Figure 21 shows how on controlled systems without time or gate valves via mechanisms. The
the two times are determined from the delay (P-systems). valves can be opened, closed or
step response of a controlled system. stopped at any position.
The above-described response of the
two-point controller is referred to as Figure 23 demonstrates the principle
the “heating function”. Besides this of operation of a 3-point controller.
Tte/Ts Controllability “heating function”, two-point control-
< 0.1 Well controllable lers are also used for the “cooling
0.1 … 0.3 Controllable function”. The principle of operation is
similar in this case but the controller
> 0.3 Poorly controllable
output is switched on when the set-
Table 4: Estimation of the controllability point value is exceeded. Figure 22
of a system with compensation and dead
time using a two-point controller
shows this principle of operation.

Control output CO Control output CO


CO = 100 CO = 100%
Output Output
switched on switched on

PVh PVh
Hysteresis Hysteresis
CO = 0 CO = 0%
Output Output
switched off SP
Process value PV switched off Process value PV
Set-point value

PVh
Hysteresis
CO = –100%
Figure 22: Principle of operation, characteristic curve of
Output
a two-point controller, cooling function SP switched on
Set-point value

13
Dead band

Figure 23: Principle of operation, characteristic curve of


a three-point controller
If the process value PV rises above the 2.3.2. Continuous-action The individual types of controller are
preset set-point value SP plus half the controllers explained in greater detail below.
dead band plus half the hysteresis,
the output of the controller is switched Continuous-action controllers are used P controller
on (CO = +100 %). If the process value for demanding control engineering The P controller is a purely propor-
drops below the set-point value plus tasks. Unlike on/off controllers, the tional-action controller. Its control out-
half the dead band minus half the hys- manipulated variable of these control- put is directly proportional to its input
teresis, the output is switched back lers may assume any value within the variable, the control deviation PVd, in
off again (CO = 0 %). The difference range of the manipulated variable/con- stationary state. The control output of
between the switch-on and switch-off trol output (i.e. the range between the the P controller is calculated as fol-
point is referred to as the hysteresis maximum and minimum possible value lows:
(as on the 2-point controller). of the control output, e.g. generally
The controller displays the same prin- between the positions “OPEN” and CO = Kp · PVd
ciple of operation in the other direction. “CLOSED” on a control valve; this CO = Kp · (Process value –
If the process value PV drops below then generally corresponds to a range Set-point value)
the preset set-point value SP minus of 0 ... 100 %). These controller types
half the dead band minus half the hys- are characterized by the fact that they Depending on Kp, the control output
teresis, the output of the controller is respond to any change in control devi- may drop below (Kp < 1) or increase
switched on (CO = –100 %). If the pro- ation (PVd = set-point value – process above (Kp > 1) the control deviation.
cess value rises again above the set- value) at the output. Kp is referred to as proportional gain
point value minus half the dead band factor or proportional coefficient.
plus half the hysteresis, the output is There are different types of continuous- As can be seen from the above calcu-
switched back off again (CO = 0 %). action controller: lation formula for the control output,
– P controller the P controller requires a control de-
A 3-point controller is shown by the – PI controller viation (PVd = PV – SP) for forming a
following symbol: – PD controller control output (CO = Kp • PVd). For
– PID controller this reason, control loops with P con-
trollers feature a permanent control
These controller types differ by virtue deviation which decreases with in-
of their dynamics, i.e. by virtue of their creasing Kp. On the basis of dynamic
SP – PVd CO
time response of their control output aspects however, it is not possible to
PV
as a function of the control deviation. achieve Kp of any arbitrary magnitude.
The various controllers are characteri- This may lead to instabilities of the
zed by their step response, i.e. by the control loop.
Figure 24: Symbolic representation of time characteristic of their control out-
a 3-point controller
put after an abrupt change in input va- Figure 26 shows the step response of
riable, the control deviation PVd. the P controller.
In principle, the 3-point controller com-
prises two 2-point controllers whose
set-point values are mutually offset.
One of the controllers must be oper- PVd CO
ated in “cooling” mode and the other Input signal

must be operated in “heating” mode.


CO = Kp · PVd

CO1
Time Time

Input step Step response

Figure 26: Step response of the P controller


SP – PVd

PV

CO2

14
Figure 25: Symbolic representation of
a 3-point controller comprising 2-point
controllers
The P controller is represented by the
following symbol:

Kp
PVd CO
CO = Kp · Tv/T1 · PVd

SP – PVd CO CO = Kp · PVd

PV
Time Time
T1
Input step Step response

Figure 27: Symbolic representation of Figure 28: Step response of the PD controller
a P controller

Characteristics of the P controller: portional gain factor Kp to be set high- for example, by disturbances in
– The P controller operates without er than on the pure P controller. electrical transfer of the process
delay and very quickly; it responds value (e.g. by standardized signals),
immediately to changes in the Figure 28 shows the step response of lead to constant fluctuations of the
control deviation. the PD controller. On real PD control- control output.
– Control loops with P controller have lers, the D-component is time-delayed – Setting parameters:
a permanent control deviation. (time constant T1), which is allowed Kp (proportional gain factor)
– Setting parameter: Kp (proportional for in the transfer function shown. The Td (derivative-action time)
gain factor). time constant T1 can, however, not be
set directly on most controllers. PI controller
PD controller The PI controller consists of a propor-
On the PD controller, not only the con- The PD controller is represented by tional component and an integral com-
trol deviation, but also its rate of change the following symbol: ponent. The integral component calcu-
is used to form a control output. The lates its share of the control output via
controller thus already responds when the time integral of the control devia-
a control deviation occurs and counter- Kp, Td tion. If there is a control deviation, the
acts the occurrence of a higher control integral component increases the con-
deviation. The control output increases trol output. This avoids a permanent
all the faster the control deviation control deviation as occurs on P con-
SP – PVd CO
changes. The control output of the PD trollers and PD controllers. The control
controller is calculated as follows: output of a PI controller is calculated
PV
as follows:


d(PVd(t)) 1
CO = Kp·(Td· +PVd(t)) CO = Kp ·( · (PVd(t)dt)+PVd(t))
dt Figure 29: Symbolic representation of Tr
a PD controller
PVd = PV – SP: Control deviation
Kp: Proportional gain factor PVd = PV – SP: Control deviation
Td: Derivative-action time Kp: Proportional gain factor or
Characteristics of a PD controller:
proportional coefficient
– Like the P controller, the PD con- Tr: Reset time
As can be seen from the above calcu- troller operates without delay and
lation formula for the control output, responds immediately to changes As can be seen from the above calcu-
the influence of the D-component is de- in the control deviation. lation formula for the control output,
termined via parameter Td. The higher – The PD controller responds to the the influence of the I-component is de-
Td becomes, the higher the D-compo- rate of change of the control devia- termined by parameter Tr. The lower Tr
nent becomes when calculating the tion and thus counteracts the build- becomes, the greater the I-component
control output. up of a higher control deviation. becomes when calculating the control
As is also the case on the P controller, – Control loops with PD controller output. Reset time Tr is the time which
control loops with PD controller have have a permanent control deviation. the controller requires to generate a
a permanent control deviation which – The D-component of the controller control output of the same magnitude
decreases with increasing Kp. How- may lead to a situation in which as that which occurs immediately as
ever, the D-component produces a minor fluctuations of the process the result of the P-component by 15
stabilizing effect which allows the pro- value, and thus minor fluctuations means of the I-component.
of the control deviation, as caused,
Figure 30 shows the step response of
the PI controller.

PVd CO
Input signal

CO = 2 · Kp · PVd

CO = Kp · PVd

Time Time
Tr
Input step Step response

Figure 30: Step response of the PI controller

The PI controller is represented by the PID controller


following symbol: The control output of the PID control-
ler is calculated from the proportional,
integral and differential component.
Kp, Tr
The control output of the PID control-
ler is calculated as follows:


SP – PVd CO
1 d(PVd(t))
CO = Kp · ( · (PVd(t)dt)+ Td · +PVd(t))
PV Tr dt

PVd = PV – SP: Control deviation


Kp: Proportional gain factor or
Figure 31: Symbolic representation of proportional coefficient
the PI controller Tr: Reset time
Td: Derivative-action time

Characteristics of the PI controller:


– The PI controller is advantageous
in that it responds quickly due to its
P-component and eliminates per-
manent control deviations owing to
the I-component.
– Since two parameters can be set
on the PI controller (Kp and Tr), it is
possible to better adapt it to the
dynamics of the controlled system.
– Setting parameters:
Kp (proportional gain factor)
Tr (reset time)

16
Figure 32 shows the step response of
the PID controller.

PVd CO
CO = Kp · Td/T1 · PVd

CO = 2 · Kp · PVd

CO = Kp · PVd

Time T1 Time

Tr
Input step Step response

Figure 32: Step response of the PID controller

The PID controller is represented by On the basis of practical experience, it


the following symbol: is possible to provide the following
estimation of the suitability of the var-
ious continuous-action controllers for
Kp, Tr, Td closed-loop control of important tech-
nical controlled variables.

SP – PVd CO Controller type

PV
P PD PI PID
Controlled Permanent No permanent
variable control deviation control deviation
Figure 33: Symbolic representation of Temperature Conditionally Conditionally Suitable Suitable for
the PID controller suitable suitable stringent demands
Flow rate Unsuitable Unsuitable Suitable Over-dimensioned
Characteristics of the PID controller: Pressure Suitable Suitable Suitable Over-dimensioned
– The PID controller unites the char- Filling level Suitable Suitable Suitable Over-dimensioned
acteristics of the P controller, PD
Table 5: Suitability of various continuous-action controllers for controlling important
controller and PI controller. technical controlled variables
– Setting parameters:
Kp (proportional gain factor)
Tr (reset time)
Td (derivative-action time)

17
3. Adapting the controller to the
controlled system

There are two requirements made on a When planning a closed-loop control The control loop is stable; the process
controller or control loop. system, there is the risk that the con- value is corrected to the new set-point
trol loop becomes unstable due to the value.
Variable command control: selected combination of controller and
Set-point value SP
In the case of variable command con- controlled system or owing to the se-
Process value PV
trol, the set-point value is not constant lected parameters of the controller. The Maximum overshoot PVm

but changes over the course of time. following behaviors may occur, e.g.
Tolerance band
The process value must be corrected after occurrence of a set-point change
Permanent control
to the set-point value. The behavior of or disturbance variable change. deviation PVb
a closed-loop control system in the
case of changing set-point value is The control loop is at the stability limit;
referred to as response to set-point the process value oscillates at constant
Time
changes. amplitude and frequency. Rise time
Trise

Settling time Tset


Fixed command control: Set-point value SP
In the case of fixed command control, Process value PV Figure 36: Process value characteristic
after a set-point change in the case of a
the set-point value is constant. In this stably operating control loop
case, the closed-loop control system
has the task of maintaining the process The behavior of a well-tuned or well-
value at the value of the set-point. Dis- set control loop after a disturbance
turbance variables acting on the con- variable change is similar.
trolled system should be compensated
for in this case. The behavior of a clos- Time
Set-point value SP
ed-loop control system with changing Figure 34: Process value characteristic,
Process value PV
control loop at the stability limit
disturbance variables is referred to as Disturbance
Variable z
disturbance response. The control loop is unstable. The pro-
Maximum overshoot PVm
cess value oscillates with increasing
In addition to having a good response amplitude. Tolerance band
to set-point changes, a closed-loop
Set-point value SP Permanent control
control system should, in most cases, Process value PV
deviation PVb

feature a good disturbance response.


If a disturbance occurs in a control
loop, this leads to a control deviation Time
Trise
which the controller must compensate
for. When planning a closed-loop con- Tset

trol system, disturbance variables are


Figure 37: Process value characteristic
of special significance. If several dis- Time after a disturbance variable change in
turbance variables act on a controlled Figure 35: Process value characteristic, the case of a stably operating control
control loop unstable loop
system, the individual disturbance var-
iables generally have a different time
response. Many disturbance variables
occur abruptly and others less abrupt-
ly. Even the magnitude of the influence
on the process value differs with the
individual disturbance variables.

18
The quality of a closed-loop control or 3.1. 3.2.1. Setting guidelines
a control loop is assessed on the basis Selecting the suitable in line with Ziegler and
of the following parameters. controller Nichols
(oscillation method)
Permanent control deviation PVb The controller must be matched to the
The permanent control deviation oc- controlled system in order for a control With this method, the controller para-
curring after the adjustment process loop to operate optimally. meters are set on the basis of the be-
has decayed. havior of the control loop at the stabili-
Suitable combinations of controllers ty limit. The controller parameters are
Overshoot PVm and controlled systems on which a initially set so that the control loop
Maximum value of the process value stable control response can be achiev- starts to oscillate. Critical characteris-
or of the controlled variable minus the ed by appropriate setting of the con- tic values then occur which allow con-
process value in steady state. troller parameters: clusions to be drawn in terms of the
– Kp (proportional gain factor) controller parameters. The precondi-
Rise time Trise – Tr (reset time) tion for using this method is that the
The time which elapses after a set- – Td (derivative-action time) control loop can be caused to oscillate.
point or disturbance variable change can be established on the based on
until the process value occurs for the the dynamics and stability of control
first time in an agreed tolerance band loops and allowing for empirical values. Procedure:
(e.g. 2 % or 5 %) about its stationary There are, of course, also control loops – Set controller as P controller (i.e.
end value. necessitating other combinations of Tr = 9999, Td = 0), initially select a
controlled system / controller. low value for Kp.
Settling time Tset Table 6 provides an overview of suit- – Set the required set-point value.
The time which elapses after a set- able combinations of controllers and – Increase Kp until the process value
point or disturbance variable change controlled systems. executes an undamped
until the process value occurs and sustained oscillation (see Figure 38).
permanently remains in an agreed
tolerance band (e.g. 2 % or 5 %) ab- 3.2.
out its stationary end value. Determining the Process value

controller parameters
On the basis of these parameters, it is
possible to formulate the requirements After a suitable controller has been se-
made of an optimally tuned control lected, a second step is to match the
loop: parameters of the controller to the
Tcrit
– permanent control deviation PVb = controlled system.
0 wherever possible
Time
– maximum overshoot PVm as low as A number of setting guidelines with
Figure 38: Process value characteristic
possible which a favorable setting of the con- of the control loop at the stability limit in
– settling time Tdet as low as possible troller parameters can be determined order to determine the control para-
meters in line with Ziegler and Nichols
– rise time Trise as low as possible. experimentally are cited in control-
engineering literature. The proportional gain factor set at the
In order to avoid incorrect settings, stability limit is designated Kcrit. The
the conditions under which the rele- resultant period of oscillation is desig-
vant setting guidelines were establish- nated Tcrit.
ed must always be followed. Besides
the characteristics of the controlled
system and of the controller them-
selves, other important factors include
whether a disturbance variable change
or a reference variable change is to be
compensated for optimally.

19
Controlled system Continuous-action controllers On/off controllers
P PI PD PID 2-point 3-point
P-element PVo Unsuitable Response Unsuitable Unsuitable Unsuitable Unsuitable
to set-point
changes
well-suited
Time
Disturbance
response
well-suited
PTt- PVo Unsuitable Response Unsuitable Unsuitable Unsuitable Unsuitable
element to set-point
changes
suitable
Time
Disturbance
response s.
1st-order PVo Response Disturbance Unsuitable Unsuitable Suitable Suitable
time-delay to set-point response
element changes well-suited
Time well-suited

1st-order PVo Unsuitable Response Unsuitable Response Condition- Condition-


time-delay to set-point to set-point ally ally
changes changes
element suitable if suitable if
suitable well-suited
with dead Time
Disturbance Disturbance hysteresis hysteresis
time response response is low is low
suitable well-suited
2nd-order PVo Unsuitable Response Response Disturbance Suitable Suitable
time-delay to set-point to set-point response
changes
element changes well-suited
suitable
Time
Disturbance well-suited
response s.
2nd-order PVo Unsuitable Response Unsuitable Response Unsuitable Unsuitable
time-delay to set-point to set-point
changes changes
element
suitable well-suited
with dead Time
Disturbance Disturbance
time response response
suitable well-suited
3rd-order PVo Unsuitable Response Unsuitable Response Unsuitable Unsuitable
time-delay to set-point to set-point
element changes changes
suitable well-suited
Time
Disturbance Disturbance
response response
suitable well-suited
I-element PVo Response Disturbance Response Disturbance Suitable Suitable
to set-point response to set-point response
changes well-suited changes suitable
Time well-suited suitable

I-element PVo Unsuitable Unsuitable Response Disturbance Suitable Suitable


with to set-point response
1st-order changes well-suited
delay Time well-suited

20
Table 6: Suitability of continuous-action and on/off controllers for combination with various types of controlled system
The controller parameters can then be
Setting the parameters in line with Ziegler and Nichols:
calculated in accordance with Table 7
from Kcrit and Tcrit.
Controller type Setting parameters
The Ziegler and Nichols setting were
determined for P systems with 1st- P controller Kp = 0.5 · Kcrit
order time delay and dead time. They PI controller Kp = 0.45 · Kcrit Tr = 0.85 · Tcrit
apply only to control loops with dis-
PID controller Kp = 0.6 · Kcrit Tr = 0.5 · Tcrit Td = 0.12 · Tcrit
turbance response.
Table 7: Controller parameters in line with Ziegler and Nichols

3.2.2. Setting guidelines


Control output CO
in line with Chien, Hrones Procedure for determining the step
and Reswick (control out- response of the controlled system:
put step method): – Switch the controller to manual
operating mode.
With this method, the controller para- ∆CO – Emit a control putput step and
meters are set on the basis of the tran- record the process value with a
sient response or step response of the recorder.
controlled system. The controller emits – In the case of critical controlled sys-
Time
a control output step. The times Tu tems (e.g. in the case of the risk of
and Tg are read off (see Figure 38) overheating), switch off in good time.
from the characteristic of the process Process value PV It must be noted that the process value
value or the controlled variable. The may increase again after switch-off on
control output step must be selected thermally sluggish systems.
so that an process value is produced
that lies within the range of the subse- Table 8 lists the controller parameter
∆PV
quent operating point of the controlled settings as a function of Tu, Tg and Ks
system. Ks is the proportional coeffi- for response to set-point changes and
cient of the controlled system. It is disturbance response and for an ape-
calculated as follows: Tu Tg Time riodic control process and a control
∆PV Figure 38: Step response of a controlled
process with 20% overshoot. The figu-
Ks =
∆CO system for determining control para-
meters in line with Chien, Hrones and
res apply to systems with P-response,
Reswick
with dead time and with 1st-order de-
∆PV: Magnitude of the process value step lay time.
∆CO: Magnitude of the control output step

Setting the parameters in line with Chien, Hrones and Reswick:

Controller type Parameters settings


Aperiodic control process Control process with overshoot
(approx. 20% overshoot)
Resp. to set-p. changes Disturbance response Resp. to set-p. changes Disturbance response
P controller Tg Tg Tg Tg
Kp = 0.3 · Kp = 0.3 · Kp = 0.7 · Kp = 0.7·
Tu · Ks Tu · Ks Tu · Ks Tu · Ks

PI controller Tg Tg Tg Tg
Kp = 0.35 · Kp = 0.6 · Kp = 0.6 · Kp = 0.7·
Tu · Ks Tu · Ks Tu · Ks Tu · Ks

Tr = 1.2 · Tg Tr = 4.0 · Tu Tr = Tg Tr = 2.3 · Tu

PID controller Tg Tg Tg Tg
Kp = 0.6 · Kp = 0.95 · Kp = 0.95 · Kp = 1.2 ·
Tu · Ks Tu · Ks Tu · Ks Tu · Ks

Tr = Tg Tr = 2.4 · Tu Tr = 1.35 · Tg Tr = 2.0 · Tu


Td = 0.5 · Tu Td = 0.42 · Tu Td = 0.47 · Tu Td = 0.42 · ˘Tu
21
Table 8: Controller parameters in line with Chien, Hrones and Reswick
4. Rating and selection of
control valves

In addition to controllers and sensors, Valves can also be classified in accord- 4.2.
actuators or final control elements ance with the distinction between the Rating and selection
which intervene in the process to be main functions of final control elements/ of ON/OFF valves
controlled as a function of the signals actuators in compliance with DIN
preset by the controller and which 19226, dividing them into CONTROL- This kind of valves can either open or
change the process variable to be con- final control elements and ON/OFF-final close a line (ON/OFF valve) or switch
trolled are required for constructing control elements. over a material stream from one line to
closed-loop control systems. another.
ON/OFF valves having only two or a
4.1. few circuit states are used for open- An important criterion for the valve to
Introduction and loop control tasks. Control valves which be selected is initially that the required
definition of terms are able to continuously set the fluid fluid quantity be able to flow through
stream are used for closed-loop pro- the valve at a given pressure differen-
Valves are final control elements or ac- cess control tasks. ON/OFF valves tial, i.e. the valve cross-section must
tuators for influencing fluid streams in and control valves have extremely dif- be adequately large. The following rule
pipe systems. In accordance with DIN ferent tasks in some cases, so that the of thumb frequently applies: the line
IEC 534, a positioning valve is a device, rating and selection of both valve types cross-section is equal to valve (fluidic
operated with auxiliary energy, which necessitate greatly different procedu- connection) cross-section. The next
varies the flow rate in a process. It con- res. requirement is that the valve be able to
sists of a valve fitting, connected to the switch against the maximum pressure
actuator, which is able to change the differential, i.e. that the valve actuator
position of the restrictor in the valve as be adequately powerful. The maximum
a function of the controller signal (con- switchable pressure differential is spe-
trol output). Generally, a control system cified in the data sheet. If the type of
is required between the actuator and auxiliary energy has been defined and
controller to act as a signal transducer the material suitability has been
and/or amplifier. In the case of many checked, it is already possible to define
positioning valves, the control system a specific valve type and to select the
is integrated as far as a field bus inter- specific valve.
face in the actuator. In accordance with
DIN IEC 534, positioning valves are sub-
divided on the basis of the following
types:
Valve type Restrictor
Lift-type valve The restrictor is generally designed as a cone.
Through-way valve It moves perpendicular to the seat plane.
3-way valve
Angle valve
Gate valve The restrictor is a flat or wedge-shaped plate.
Diaphragm valve A flexible restrictor performs the function of
of isolation and sealing.
Ball valve The restrictor is a ball with a cylindrical bore
or a segment of a ball.
Butterfly valve A disc mounted in such a way as to allow it to rotate.
Plug valve The restrictor may be a cylindrical, conical or
22 eccentrically mounted ball segment.

Table 8: Classification of positioning valves in accordance with DIN IEC 534


4.3.
Rating and selection
of control valves

Control valves are able to constantly Range of the Flow form Interrelationship
change their opening cross-section Reynolds number between ∆p and Q
and thus continuously influence fluid
Re low Laminar ∆p  Q
streams. They thus represent variable
flow resistors. Re high Turbulent ∆p  Q7/4
Re  Recritical Transitional form To be determ. experimental.
4.3.1. Fluidics
fundamentals
From this, we can conclude that the In the case of a high Reynolds number,
Flow resistances occur in process in- flow resistance R is constant only in i.e. in turbulent flow, the following ap-
stallations in various forms: the case of laminar flow owing to plies to cross-sectional resistances in
– as resistances in capillaries, gaps, ∆p ≈ Q. Otherwise, a non-linear rela- non-compressible fluids: ∆p  Q2.
nozzles, diaphragms and tionship always applies between pres-
valves sure drop ∆p and fluid flow Q. The flow-rate variable kv which is de-
– as line resistances in pipes, hoses fined as follows is used to identify
and ducts The following applies to fluid resist- valves as orifice-type fluidic compo-
– as leakage resistances in gaps and ances in the case of cross-sectional nents:
porous components. variation in non-compressible fluids the kv value (in m3/h) is the volume
and with turbulent flow: flow of water at +5 to +30 °C passing
In general, the ratio of pressure drop through the valve at the relevant
∆p to fluid flow Q can be defined as The permanent pressure loss ∆ploss is stroke s with a pressure loss
the flow resistance R of a component. taken as the basis for the flow resist- ∆p0valve = 100 kPa. (1 bar; 14.5 psi)
ance R. The flow resistance coefficient
∆p  is introduced as a “non-dimensional Analogous to this, the flow-rate coef-
R= pressure loss” by referring the perma- ficient cv is described in the American
Q nent pressure loss to the dynamic literature, defined as follows:
pressure. the cv value (in US gal/min) is the vol-
Basically, a distinction must be made ume flow of water at 60 °F which pas-
∆ploss
between two types of resistance on = ses through at a pressure loss of 1 psi
 · u2 (
the basis of the physical causes:
– frictional resistances due to flow
involving friction
( 2
with the relevant stroke s.

Moreover, the QNn-Wert (in l/min) is


– cross-sectional resistances owing The following applies to the model specified as the flow characteristic for
to variations in the flow cross- case of flow through an orifice plate: compressed air under standardized
section. conditions for pneumatic valves.
Apipe 2

The following distinction between


cases for dependence between ∆p
= [( · Aorifice
–1
[ (
The following conversion factors apply:

and Q must be made for frictional re-   contraction coefficient Kv ↔ cv: kv = 0.86 cv
sistances in non-compressible fluids and Kv ↔ : kv = 4 · d2/()1/2
as a function of the Reynolds number Kv ↔ QNn: kv = 1078 QNn
∆ploss = (1 – mb) · ∆pB
∆pB  effect. press. through the orifice plate
u·Dh Aorifice
Re = mB 
Apipe
 opening ratio
v
u  mean flow velocity
Dh  hydraulic diameter,
and, after introduction of a flow coef-
Dh = kA/U ficient a, we obtain the flow-rate equa-
v  kinematic viscosity
tion

Q =  · Aorifice  2 ·∆p
 flow coefficient 23
 density of the fluid
4.3.2. Characteristic The equal-percentage valve characte- Linear flow characteristic
curves ristic approximates practical require- In the simplest case, the flow charac-
ments to a greater extent than the li- teristic is linear, i.e. kv = K1 · s:
4.3.2.1. Valve characteristic near characteristic, since
The valve characteristic represents the – low variations in stroke ∆s cause
dependence of the aperture cross- low ∆A, i.e. fine infeed movements
section A on the stroke s of the valve – high variations in stroke ∆s cause

Flow rate kv
spindle: A = f(s). high ∆A, i.e. coarse infeed move-
In the simplest case, the valve charac- ments.
teristic is linear, i.e. A = K1 · s: In the case of s = 0, a minimum aper-
ture cross-section A0 is present. The
valve closes only with an additional
Stroke s
sealing edge.
Aperture area A

Figure 42: Linear flow characteristic


Various valve characteristics are imple-
mented by the contour of the closure
elements, the valve cones. Convention- Note: Unlike the linear valve character-
al designs include, for example, par- istic, the interrelationship between the
abolic cone, lantern cone, perforated volumetric flow rate Q through the
Stroke s
cone, V-port cone and many others. valve and the valve stroke s is linear in
Figure 39: Linear valve characteristic of the case of the linear flow characteris-
a control valve 4.3.2.2. Flow characteristic and tic.
rangeability
Note: Despite the linearity of the valve The most conventional characteristic Equal-percentage flow characteristic
characteristic, there is no linear inter- curve required for valve selection is The equal-percentage flow character-
relationship between the volumetric the flow characteristic. The flow char- istic is described by a constant per-
flow rate Q through the valve and the acteristic represents the dependence centage increase in the flow rate kv
valve stroke s due to non-constant of the standardized flow rate kv on the with the stroke s (referred to the rele-
pressure drops through the valve. stroke s: kv = f(s). vant flow rate kv present).
The equal-percentage valve charac-
teristic is described by a constant per- ON/OFF flow characteristic
centage increase in the aperture cross- Due to their plate cone, ON/OFF valves
section A with stroke s (referred to the initially have a linear flow characteristic dkv(s) 1
· = K2
relevant aperture cross-section A pre- in the range of small stroke (up to ap- ds kv(s)
sent). prox. 30 % stroke). At an opening angle
of 30 to 40 % stroke (s), such valves kv = kv0 · eK2s
already achieve approx. 90 % flow
dA(s) 1 rate (kv). As the aperture opens even
· = K2 further, the flow rate rises only very
ds A(s)
slowly through to the maximum, at full
A = A0 · eK2s
Flow rate kv

stroke. Since the total stroke of the


ON/OFF valves is low in relation to the
stroke of control valves, the actual
task of producing a high change in
Aperture area A

flow rate with low stroke is performed.


Stroke s

Figure 43: Equal-percentage flow


characteristic
Flow rate kv

At s  0 a minimum aperture cross-


Stroke s section A0 is present, causing a mini-
Figure 40: Equal-percentage valve
mum flow kv0. The valve closes only
characteristic of a control valve with an additional sealing edge.

Stroke s

24 Figure 41: OPEN/CLOSE flow


characteristic
The maximum aperture cross-section The operating characteristic curve Examples of operating
· 1 1
Amax is reached at maximum stroke s. (V = f (s)) thus differs from the flow characteristics (f. rangeability =
 30
The related kv value is referred to as characteristic curve kv = f (s) of the and various ):
kvs value. The range between kv0 and valve considered in isolation.
kvs is the total range of the manipulat- Control valve with linear flow
ed variable of the valve. The magnitude of the difference (the characteristic:
degree of characteristic distortion) is
The ratio of kv0 to kvs is referred to as represented by the pressure ratio . ·
·
V
Vmax
the rangeability and defines a valve The pressure ratio  is stated for the
1.0
characteristic value: fully open valve:
0.8
= 0.1 0.3 1.0
1
=
kv0 ∆pvo ∆pvo 0.6

 = =
kvs ∆po ∆pvo+∆pLo+∆p'o 0.4
1
=
1
 30
0.2
∆po: Pressure drop over entire installa-
s
tion smax
Conventional values are as follows: ∆Pvo: Pressure drop at fully opened valve 0.2 0.4 0.6 0.8 1.0
(max. flow)
∆pLO: Pressure drop at tubes, fittings... Figure 46: Control valve with linear flow
1 1 1 1 ∆p'o: Pressure loss at pump (at max. characteristic
= ; ; flow rate)
 25 30 50
Control valve with equal-percentage
The behavior of the system as an flow characteristic
4.3.2.3. Operating characteristic interplay between source (pump) cha-
and pressure ratio racteristic and load (valve) characteris- ·
·
V
Vmax
The operating characteristic identifies tic can be shown in the characteristic
1.0
the flow behavior of the valve under map (see Figure 45):
0.8
operating conditions in the installa-
= 0.1 0.3 1.0
tion. It represents the dependence of The following standardized equation 0.6
·
the volume flow V on the stroke s of applies to the operating characteristic 0.4
1 = 1
the valve spindle. 0.2  30
·
· V 1 s

V = f(s) · = 0.2 0.4 0.6 0.8 1.0


smax

V [ (
[(kvkv
max 2
s
The following main elements of the in- 1+ -1 Figure 47: Control valve with equal-per-
centage flow characteristic

stallation influence the operating be-


havior: An approximation of a linear operating
– the pump; the pressure generated i.e. the operating characteristic de- characteristic can be achieved
by the pump drops as ∆p over the pends on the pressure ratio  and the – in the case of linear flow character-
entire installation flow characteristic istic with high pressure ratio
– the tubes with the pressure drops – in the case of equal-percentage
∆pLi flow characteristic with low pres-

∫( s
– and other resistances ∆p'i in the in-
kvs
=
s ( sure ratio .
stallation (shut-off valves, heat ex- kv max The non-linearities for both valve types
changers, pipe elbows, branches, have approximately the same magni-
changes in cross-section and other tude at  0.3.
installed fittings).

∆p Positioning
∆p valve Operating points
Closes

Pump
Tube
Valve
Ventil
Tube
∆pL0 ∆pL0
Opens ∆p = ∆pvmin + ∆pL
∆pL1 ∆pv ∆pL2 ∆p0
∆pL1 + ∆pL2 = ∆pL ∆pvmin
∆pv ∆pv ∆pL
∆pv
Figure 44: Pressure losses in ∆pL0 ·
an installation V
∆pL ∆pL

Range of the control output


25
Figure 45: Characteristic map, source
characteristic and load characteristics
A linear operating characteristic is The following formulae are helpful for differential pressure across the valve
achieved only if the valve features an practical application: and at stroke s. The kvs value, analog-
optimum flow characteristic as the ously, is the quantity at stroke
result of a special valve contour. Fluids: s = 100 %.

Flow and operating characteristics for NW = 0.42 · Q Analogous to the kv value the flow rate
1 1
valves with  0.3 and = : coefficient cv is described in the Ame-
 30 NW: Connection nominal diameter
Q: Volumetric flow rate in l/h
rican literature. The following conver-
kv
kvs sion factor applies: kv = 0.86 · cv.
1.0 See also chapter “4.3.1. Fluidics fun-
0.8
Gases: damentals”.

0.6
lin. opt. ep.
QN The kv value must be calculated for the
0.4 NW = 4.2 · current operating data. A distinction
Flow
characteristics p1
kv0
0.2 must be made between maximum load
= 0.33 s QN: Volumetric flow rate in Nm3/h
kvs
smax p 1: Pressure upstream of the valve in
(maximum quantity Qmax, minimum
0.2 0.4 0.6 0.8 1.0
bar absolute ∆pmin ➾ kvmax) and minimum load (mi-
Figure 48: Flow characteristics: nimum quantity Qmin, maximum ∆pmax
linear, optimum, equal-percentage ➾ kvmin) Both load cases must be cal-
Steam: culated individually and be adjusted
on the basis of the valve rangeability.
·
V
·
Vmax NW = 2.8 · G · v"
1.0 The following applies to cold water:
0.8 G: Mass flow rate in kg/h
v": Specific volume in m3/kg 1
kv = Q ·
0.6
lin. opt. ep. ∆p
0.4
Operating
characteristics Q: Volumetric flow rate in m3/h
0.2 General: ∆p: Pressure differential at the valve in
0.061 s bar
smax
0.2 0.4 0.6 0.8 1.0
QB The following applies in general to
NW = 18,8 · fluids (sub-critical):
Figure 49: Operating characteristics: c
linear, optimum, equal-percentage
QB: Volumetric flow rate in m3/h p2 > ps2
c: Flow velocity in m/s
4.3.3. Rating and ps2: Saturated steam press., in bar abs.,
related to the temperature downstream
selection
of the valve
In the case of simple control valves on
Control valves must be rated and se- which a connection nominal diameter 1
kv = Q · 0.032 ·
lected with a view to their specific task is assigned directly to a kvs value, the ∆p
in order to be able to ensure a fault- anticipated flow velocity should at mi-
less control function. nimum be checked. 1: Density of the medium in kg/m3
∆p: Pressure differential at the valve in
bar
Initially, the connection nominal dia- The nominal pressure stage results
1
meter must be defined in accordance from knowledge of the valve material, kv = G · 0.032 ·
with the medium and the related, effi- the operating temperature and the max. 1 · ∆p
cient flow velocity. operating pressure, e.g. from DIN 2401, G: Mass flow rate in kg/h
or from a valve data sheet. ∆p: Pressure differential at the valve in
bar
The following guideline values apply:
– 2 m/s for fluids The actual closed-loop control func- The following applies to fluids in
– 20 m/s for gases tion, i.e. setting the fluid flow rate of a general (super-critical):
– 45 m/s for steam. given temperature and given pressure
while simultaneously producing a de- p2 < ps2
fined pressure loss, is determined by ps2: Saturated steam pressure, in bar
the flow characteristic, the kv value absolute, related to the temperature
downstream of the valve

The kv value is a reference variable and The kv value is calculated here in two
26 is defined as follows: kv value = quan- steps: the kv value for the evaporating
tity in m3/h of cold water (+5 … +30 °C) steam quantity kvD and the kv value for
which flows through the valve at 1 bar. the fluid kvF are calculated separately
and both values are added.