Beruflich Dokumente
Kultur Dokumente
written by
Peter Devine
additional information
Karl Grießbaum
type setting and layout
Liz Moakes
final drawings and diagrams
Evi Brucker
Devine, Peter
Radar level measurement - The user´s guide
1. Radar
2. Title
621.3´848
ISBN 0-9538920-0-X
Foreword ix
Acknowledgement xi
Introduction xiii
Part I
1. History of radar 1
2. Physics of radar 13
3. Types of radar 33
1. CW-radar 33
2. FM - CW 36
3. Pulse radar 39
Part II
4. Radar level measurement 47
1. FM - CW 48
2. PULSE radar 54
3. Choice of frequency 62
4. Accuracy 68
5. Power 74
5. Radar antennas 77
1. Horn antennas 81
2. Dielectric rod antennas 92
3. Measuring tube antennas 101
4. Parabolic dish antennas 106
5. Planar array antennas 108
Antenna energy patterns 110
6. Installation 115
A. Mechanical installation 115
1. Horn antenna (liquids) 115
2. Rod antenna (liquids) 117
3. General consideration (liquids) 120
4. Stand pipes & measuring tubes 127
5. Platic tank tops and windows 134
6. Horn antenna (solids) 139
B. Radar level installation cont. 141
1. safe area applications 141
2. Hazardous area applications 144
Foreword
To suggest that any one type of level The purpose of this publication is
transmitter technology could be regar- quite specific, and that is to explain
ded as 'universal' would be unrealistic some of the principles involved, and to
and potentially irresponsible due to the show that by applying some simple
variation and complexity of available guidelines, what is obviously a sophi-
applications when liquids, powders and sti-cated technology can be simple and
solids are all considered. However, the reliably used in an enormously wide
rate at which radar based level trans- range of industrial and process applica-
mitters have established themselves tions.
over the last couple of years would We make no apology for including a
tend to suggest that this technology is chapter on Vega specific products, and
closer to that definition that any princi- hope this guide stimulates a radar user,
ple has ever been. or some greater depth of knowledge if
I have personally been involved in you have some experience, we look
the development, applications, sales forward to hearing from you.
and marketing of level transmitters,
controllers and indicators of most types Mel Henry
over the last twenty years. In that time Managing Director
nothing has, in my opinion, come close Vega Controls Ltd.
to matching the significance of radar in
terms of its overall suitability, for not
just conventional but extreme process
conditions applications for the vast
majority of substances in vessels of vir-
tually any size or complexity.
This unique principle combined with
current reflections processing software,
materials of construction, simplicity of
installation and transmitter digital com-
munications allows this to be conside-
red as a day to day 'first consideration'
for level, whereas only a very short
time ago it was regarded as expensive
and specialised - this is no longer the
case.
ix
Acknowledgements
xi
Introduction
xiii
1. History of radar
1
Prior to World War II, radar was
being developed independently in a
number of different countries, includ-
ing Britain, Germany, the United
States, Italy, France and the Soviet
Union.
In 1934, following a series of exper-
iments at the Naval Research
Laboratory in the United States, a
patent was granted to Taylor, Young
and Hyland for a ‘System for detecting
objects by radio’.
In February 1935, British scientist,
Robert Watson-Watt presented a paper
on ‘The detection and location of air-
craft by radio methods’ to the Tizard
Committee for the Scientific Survey of
Air Defence.
Christian Hülsmeyer
produced the first practical radar
patented in 1904
(Pic. 1.3 - D.M.M)
2
1. History of radar
3
mounted sea search radar transmitters
called Seetakt. These were delivered as
early as 1938 with a frequency of 366
MHz (wavelength 82 cm) and were
installed on many vessels including the
famous battleships, Bismarck and Graf
Spee.
German Naval developments also
produced the Freya range of search
radars operating on 125 MHz (wave-
length 2.4 metres). These were found to
be effective for tracking aircraft at long
range, and were subsequently supplied
to the Luftwaffe for early warning.
However, they could not provide alti-
tude information.
Other German radars in wide use
British Chain Home Radar aerials - were the parabolic antenna Würzburg
Radar was instrumental in the defence and Würzburg Riese (Giant Würzburg)
of Britain during the second world war transmitters. The standard Würzburgs
(Pic. 1.6 - I.W.M) were generally used for directing
transmitters. They operated on a fre- searchlights and flak batteries and the
quency of 200 MHz (wavelength 1.5 Würzburg Riese for tracking individual
metres). intruders and directing night fighters to
It is well documented that the CH intercept them.
and CHL network of radar stations In a similar fashion to the British
were a crucial factor during the Battle Chain Home system, the Germans built
of Britain in the summer of 1940. It a defensive network of ‘Himmelbett’
enabled the fighters of the Royal Air radar stations. The literal translation of
Force to be deployed when and where Himmelbett is four poster bed. The
they were needed and rested when the four ‘posts’ of the bed consisted of a
threat receded. The limited resources in Freya early warning radar, a Würzburg
men and machines were not wasted in radar for tracking the intruding aircraft,
long standing patrols. a Würzburg radar to guide the night
German radar research was also con- fighter to the intruder and a Seeburg
ducted in secret in the late 1930’s. plotting table (Seeburgtisch) to monitor
Whereas the development effort in the interception.
Britain was focused on air defence, in This defensive radar system became
Germany separate radar developments known by the British as the
were carried out for the Navy, Army ‘Kammhuber Line’ after the German
and Luftwaffe. general in charge of night fighters.
Companies involved in German
naval research produced a range of ship
4
1. History of radar
5
Both Britain and Germany devel- It became clear to radar researchers
oped airborne radar for fighter inter- that a shorter ‘centimetric’ wavelength
ception by night. British airborne radar would be more useful for a number of
trials started in 1937 with the produc- applications. This would enable a more
tion AI Mark 1 taking to the air in May focused airborne radar that would not
1939. The first practical British suffer from the ground returns that
Airborne Interception radar was the AI restricted capabilities of the first air-
Mark IV which was first tested in borne radars. The higher frequency
August 1940. could be used for a ground mapping
In Germany the Lichtenstein air- radar unit to locate towns and other
borne radar was available in mid 1941. geographic features.
The characteristic external radar aerial The problem was how to find a
array of the Lichtenstein caused signifi- method of generating sufficient power
cant aerodynamic drag. This could at the desired wavelength of 10
reduce the aircraft speed by as much as centimetres.
40 kilometres per hour. By 1943 the
range had been extended to 6000
metres.
6
1. History of radar
The Cavity Magnetron was used in centrimetric ‘microwave’ airborne radar and pro-
duced a quantum leap in performance. The radar dish was protected inside a
plastic nose assembly
(Pic. 1.12 & 1.13 - H.R.A)
7
Britain also used the cavity mag- function.
netron in the development of a ground In the same way in Germany,
mapping radar called H2S. This device radar was disguised as ‘Dezimeter
enabled aircraft to be accurately navi- Telegraphie’ or ‘De-Te’, translated as
gated to their destinations without the decimetric telegraphy
aid of ground based beacons or beams. It was the Americans who intro-
Britain shared this secret microwave duced the now universally used palin-
technology with the United States drome, RADAR or RAdio Detection
where additional development took And Ranging.
place at the Radiation Laboratory with- The history of the development of
in the Massachusetts Institute of radar during the course of the Second
Technology. From the work carried out World War is a huge subject in
at MIT, further airborne interception itself. Many devices were developed.
radars and gun laying radars were mass Measures and counter measures were
produced and delivered to the allied taken in the radar war.
forces. The American SCR-720 (known Since 1945, radar has been used for
as AI Mark X in Britain) was first an increasing number of peaceful appli-
delivered to the USAAF by late 1942. cations. The giant Würzburg parabolic
This radar unit became a standard radar transmitters of the Second World
device long after the war had finished. War became post war radio telescopes.
War time secrecy meant that radio The basic designs were developed and
detection devices were given coded enlarged and can be seen at the well
names. In Britain, the early chain home known Jodrell Bank Observatory near
radar was called RDF after the existing Manchester which has a dish diameter
Radio Direction Finding systems in the of 75 metres.
hope that it would mislead their real Viewed from Earth, the planet Venus
Modern radar systems are exemplified by this ‘AWAC’ airborne early warning aircraft.
Multiple targets can be detected at extreme range
(Pic. 1.14 - P.D)
8
1. History of radar
Detection by radar is not always desirable. Huge sums of money have been spent
reducing the radar signature of the F117 stealth fighter
(Pic. 1.16 - P.D)
9
Radar technology is part of our Later, lower accuracy FM - CW radar
everyday lives. The cavity magnetron transmitters became available for the
is used in microwave ovens. process industry.
Continuous wave (CW) radars are used In the late 1980’s, pulse radar level
in automatic door detection and vehicle transmitters were developed for process
speed measurement. Other well known measurement applications. The avail-
civilian radar applications include air ability of suitable crystals and solid
traffic control, shipping and weather state components such as GaAs FET
radar. oscillators enabled cost effective radar
Radar altimeters developed in the level transmitters to enter the market.
1930’s use a form of radar called In 1997 a significant improvement
FM - CW or Frequency Modulated in the specification of radar level trans-
Continuous Wave radar. mitters was achieved. VEGA produced
In the 1970’s, the same FM - CW the world’s first two wire, loop pow-
measurement technique was used in ered, intrinsically safe radar level trans-
the production of the first radar level mitter. For the first time low cost, high
tank gauge. Initially these radar level specification radar level transmitters
transmitters were used to measure became available.
petroleum products in supertankers. It is likely that these advances will
Further developments of FM - CW continue into the new millennium and
level transmitters led to their use on that radar level transmitters will
shore based storage tanks in the mid become a commodity item in the same
1980’s. Originally these were expen- way as differential pressure transmit-
sive, high accuracy systems for fiscal ters.
measurement of petroleum products.
10
1. History of radar
A raw oscilloscope echo trace had to be interpreted by skilled operators using the British
war time Chain Home Low radar
(Pic. 1.18 & 1.19 - I.W.M)
Comprehensive information is available on the PC echo trace of the latest two wire loop
powered radar level transmitters
(Pic. 1.20 - Vega Pic. 1.21 - Vega)
11
Inhalt
Foreword ix
Acknowledgement xi
Introduction xiii
Part I
1. History of radar 1
2. Physics of radar 13
3. Types of radar 33
1. CW-radar 33
2. FM - CW 36
3. Pulse radar 39
Part II
4. Radar level measurement 47
1. FM - CW 48
2. PULSE radar 54
3. Choice of frequency 62
4. Accuracy 68
5. Power 74
5. Radar antennas 77
1. Horn antennas 81
2. Dielectric rod antennas 92
3. Measuring tube antennas 101
4. Parabolic dish antennas 106
5. Planar array antennas 108
Antenna energy patterns 110
6. Installation 115
A. Mechanical installation 115
1. Horn antenna (liquids) 115
2. Rod antenna (liquids) 117
3. General consideration (liquids) 120
4. Stand pipes & measuring tubes 127
5. Platic tank tops and windows 134
6. Horn antenna (solids) 139
B. Radar level installation cont. 141
1. safe area applications 141
2. Hazardous area applications 144
2. Physics of radar
Electromagnetic waves
Th e velocity of light in free space is The velocity of an electromagnetic
299,792,458 metres per second, but wave is the product of the frequency
who is timing? For the purposes of the and the wavelength.
calculations in this book, we will call it
300,000 kilometres per second or c = f xλ
3 x 108 metres per second. [Eq. 2.2]
Maxwell’s theories of electro-
magnetism were confirmed by the c velocity of electromagnetic
experiments of Heinrich Hertz. These waves in metres / second
show that all forms of electromagnetic f frequency of wave in second -1
radiation travel at the speed of light in
free space. This applies equally to long
λ wavelength in metres
wave radio transmissions, microwaves,
infrared, visible and ultraviolet light The original cavity magnetron had
plus X-rays and Gamma rays. a wavelength of 9.87 centimetres.
Maxwell showed that the velocity of This corresponds to a frequency of
light in a vacuum in free space is given 3037.4 MHz (3.0374 GHz).
by the expression : The frequency of a pulse radar
Examples :- level transmitter may be 26 GHz
1 or 26 x 108 metres per second.
co = The wavelength is 1.15 centimetres.
(µ o x εo) [Eq. 2.1] The electromagnetic waves have an
electrical vector E and a magnetic vec-
co velocity of electromagntic wave
tor B that are perpendicular to each
in a vacuum in metres / second other and perpendicular to the direction
of the wave. This will be discussed and
µo the permeability of free space
(4 π x 10 -7 henry / metre) illustrated further in the section on
polarization. The electrical vector has
εo the permittivity of free space the major influence on radar applica-
(8.854 x 10 -12 farad / metre)
tions.
λ direction of wave
amplitude
Fig 2.1
13
The Electromagnetic spectrum
10 8 10 7 10 6 10 5 10 4 10 3 10 2 10 1 10 0 10 -1 10 -2 10 -3 10 -4
10 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 9 10 10 10 11 10 12
3m 0.3 m 3 cm 3 mm
14
2. Physics of radar
10 13 10 14 10 15 10 16 10 17 10 18 10 19 10 20 10 21 10 22 10 23 10 24 Hz
15
Permittivity Permeability µ and relative
In electrostatics, the force between permeability µr
two charges depends upon the magni- The magnetic vector, B, of an elec-
tude and separation of the charges and tromagnetic wave also has an influence
the composition of the medium on the velocity of electromagnetic
between the charges. Permittivity ε is waves. However, this influence is neg-
the property of the medium that effects ligible when considering the velocity in
the magnitude of the force. The higher gases and vapours which are non-mag-
the value of the permittivity, the lower netic. The relative permeability of the
the force between the charges. The product being measured has no signifi-
value of the permittivity of free cant effect on the reflected signal when
space (in a vacuum) εo, is calculated compared with the effects of the rela-
indirectly and empirically to be: tive permittivity or dielectric constant.
8.854 x 10-12 farad / metre. For the non-magnetic gases above the
product being measured, the value of
Relative permittivity or the relative permeability, µr = 1.
dielectric constant εr
The ratio of the permittivity of a Frequency, velocity and wave-
medium to the permittivity of free length
space is a dimensionless property As we have already stated, the fre-
called ‘relative permittivity’ or ‘dielec- quency (f), velocity (c) and wavelength
tric constant’. For example, at 20° C (λ) of the electromagnetic waves are
the relative permittivity of air is close related by the equation c = f x λ.
to that of a vaccum and is only about The frequency remains uninfluenced
1.0005 whereas the relative permittivi- by changes in the propagation medium.
ty of water at 20° C is about 80. However, the velocity and wavelength
(Dielectric constant is also widely can change depending on the electrical
known as DK.) properties of the medium in which they
The value of the dielectric constant are travelling. The speed of propaga-
of the product being measured is very tion can be calculated using equation
important in the application of radar to 2.3.
level measurement. In non-conductive co
products, some of the microwave ener- c =
gy will pass through the product and
the rest will be reflected off the surface.
(µ x ε )
r r [Eq. 2.3]
16
2. Physics of radar
solvent echo
Fig 2.3 - Effect of dielectric constant on the running time of a microwave radar
17
The same effect can be experienced when looking at interface detection using
guided microwave level transmitters to detect oil and water or solvent and aqueous
based liquids.
reference echo
Fig 2.4 Oil/water interface
(water without oil) detection using a
guided microwave
level transmitter. Note
that the water echo
oil echo water echo has a reduced ampli-
tude and appears to be
further away. The
running time of
microwaves in oil is
slower than in air
18
2. Physics of radar
From equation 2.4 and equation 2.3, differ but they have only a very small
we can calculate the percentage error effect on the accuracy of radar.
caused by variations in the dielectric Radar level transmitters are usually
constant of different gases and vapours calibrated in air. For this reason, the
and the relative effects of changes in following tables show
process temperature and pressure.
1. Dielectric constant of different gases
Gases and vapours at normal temperature and pressure
By definition, the dielectric constant (273K, 1 Bar A)
in a vacuum is equal to 1.0. The dielec- 2. Percent error in the running time in
tric constants of the gases and vapours the gases compared with air
that may be present above the product
Table 2.1 The dielectric constants under normal conditions, εrN and the error caused by
the dielectric constant of typical process gases under normal conditions
19
Temperature
High temperature or large temperature gradients have very little effect on the
transit time of microwaves within an air or vapour space. At a temperature of
2000° C the variation is only 0.026% from the measurement value at 0° C. Radar
level transmitters with air or nitrogen gas cooling are used on molten iron and steel
applications.
0.03
0.025
0.02
% error
0.015
0.01
0.005
0.0
0 250 500 750 1000 1250 1500 1750 2000
Temperature in ° C
Fig 2.5 Temperature effect on radar measurement of air at a constant pressure of 1 BarA
20
2. Physics of radar
Pressure
Pressure does have a small but more significant influence on the velocity of
electromagnetic waves. At a pressure of 30 Bar, the error is only 0.84%. However
this becomes more significant and at a pressure of 100 Bar there is a velocity
change of 2.8%. If the pressure is varying constantly between atmospheric pressure
and 100 Bar, the velocity variations can be compensated using a pressure transmit-
ter.
10
8
% error
0
0 50 100 150 200 250 300 350 400
Fig 2.6 The influence of pressure on radar measurement in air at a constant temperature
of 273 K
21
Waveguides, stilling tubes & bypass tubes
In the preceding equations, we have within a metallic tube the running time
assumed that the microwaves are appears to slow down because the
travelling in ‘free space’ in a vacuum. microwaves travel further bouncing
However, in practice the proximity off the inside wall of the tube and
of metallic vessel walls and other currents are set up on the inside surface
structures will have an influence on of the tube.
the propagation velocity of the This effect is discussed in more
microwaves. This is particularly true detail in the chapters on antennas and
when microwave radar level transmit- mechanical installations. The wave-
ters are fitted inside bypass tubes or guide effect can be compensated during
stilling tubes or when a horn antenna is calibration and the use of stilling tubes
fitted with a waveguide extension. and bypass tubes can be beneficial in
When microwaves are propagating some level applications.
22
2. Physics of radar
Transmitted power: W1
Reflected power: W2
Dielectric constant: εr 4 x εr
Then the percentage of reflected Π = 1- 2
power at the dielectric layer, (1 + ε )r
W2
Π = [Eq. 2.5]
W1
Toluene Acetone
Solvent with a low dielectric constant, Solvent with a dielectric constant,
εr = 2.4 εr = 20
4x (2.4) 4x ( 20 )
Π = 1- Π = 1-
2 2
(1 + (2.4)) (1 + (20) )
4.46% power is reflected 40 % power is reflected
100
Π x 100% power reflected
80
60
40
20
0
0 10 20 30 40 50 60 70 80
Dielectric constant, εr
Fig 2.7 Reflected radar power depends upon the dielectric constant of the product
being measured
23
In radar level measurement the reflected energy from a product surface becomes
more critical at a dielectric constant (εr) of less than 5. The following graph shows
this important region.
20
Π x 100% power reflected
15
10
0
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Dielectric constant, εr
Fig 2.8 Reflected radar power depends upon the dielectric constant of the product being
measured. This graph shows the critical region where care must be taken over
choice of radar antenna
- 10
Loss L, dB
- 20
- 40
- 60
1.02.51.5 3.0
2.0 3.5 4.0 4.5 5.0
Dielectric constant, εr
Fig 2.9 Reflection loss in dB: loss L = 10 log Π
24
2. Physics of radar
Polarization
Electromagnetic waves have an gram. The direction of the linear polar-
electrical vector E and magnetic vector ization is set by the orientation of the
B that are in phase but perpendicular to signal coupler from the microwave
each other. The direction of propaga- module. The properties of the polariza-
tion of the waves is perpendicular to tion of microwaves can be important in
the electrical and magnetic vectors as the application of radar to level mea-
shown in the diagram below. surement.
Polarization defines the orientation In television and microwave com-
of the electromagnetic waves and refers munications, linear polarization is also
to the direction of the electrical vector referred to as horizontal or vertical
E. Most process radar level transmitters polarization depending on the relative
exhibit linear polarization as in the dia- orientation of the aerials or antennas.
direction of wave
Fig 2.10 Diagram showing linear polarization and the relative orientation of the electric
vector E, the magnetic vector B and the direction of propagation of the
microwaves
25
Another form of polarization is Circular polarization can also be
elliptical polarization. A specific form used in search radars to separate the
of elliptical polarization is circular reflections from aircraft or ships from
polarization where the electrical vector interference echoes from rain. The
E and magnetic vector B rotate through almost spherical shape of the rain drops
360° within the space of a single wave- causes a definite reversal of polariza-
length, when a linear or circular polar- tion which can be easily rejected by the
ized signal is reflected the direction of receiving antenna. However, the scat-
polarization is reversed. With circular tered reflections from the ship or air-
polarization it is possible to use the craft provide roughly equal amounts of
reversal of polarization to distinguish reversed and un-reversed energy that
between a direct echo and an echo that enables detection.
has made two reflections.
Fig 2.11 Circular polarization involves rotation of the electrical and magnetic vectors
through 360° within a wavelength
26
2. Physics of radar
Direction of wave
Large echo
Fig 2.12 If a metallic or high dielectric object is orientated in the same plane as the
electrical vector of the polarized microwaves, the radar level transmitter will
receive a large amplitude echo
Direction of wave
Small echo
Fig 2.13 If the same object is orientated at right angles to the plane of the electrical vector,
the received echo will have a smaller amplitude
27
Diffraction
Beam angle is often discussed in which accounts for most of the radiated
relation to radar transmitters. This can power, there are also weaker side lobes
give the impression that the radar of energy. This phenomenon is caused,
antenna can direct a finely focused in part, by diffraction. In addition to
beam towards the target. Unfortunately this, destructive interference causes the
this is not the case. null points or notches that form the
In practice, although they are characteristic side lobes.
designed to produce a directed beam, a Chapter 5 provides a detailed expla-
radar antenna radiates some energy in nation of beam angles, side lobes and
all directions. As well as the main lobe types of antennas.
side lobes main lobe
antenna
Fig 2.14 The lobe structure of antenna beams is caused by diffraction and destructive
interference
Refraction
In the same way as light is refracted The angle of refraction depends on
at an air/glass or air/water interface, the angle of the incident wave and also
microwaves are refracted when they on the ratio of the dielectric constants
encounter a change in dielectric. This at the interface.
could be a low dielectric window It is possible to utilise the refractive
(PTFE/glass/polypropylene) or a non- properties of electromagnetic waves to
conductive low dielectric liquid such as construct a dielectric lens that will
a solvent. focus microwaves.
interface
28
2. Physics of radar
Interference - Phase
Problematic interference effects are caused primarily by the inadvertent mixing
of signals that are out of phase. The microwave signals have a sinusoidal wave-
form.
Phase angle
45°
Fig 2.16 In this illustration both of the sine waves have an identical frequency and
amplitude but the second wave has a 45° phase lag
29
Interference
Microwaves can manifest interfer- to interference of the signal. The chap-
ence effects in exactly the same way as ter on mechanical installation should
light. Potentially this can cause mea- help a radar level user to avoid this
surement problems. The causes of potential problem.
interference should be understood and However, we use destructive inter-
avoided by design and installation con- ference to our advantage when we
siderations. apply pulse radar level measurement
The wrong choice of antenna, instal- through a low dielectric ‘window’ to
lation of an antenna up a nozzle, posi- measure conductive or high dielectric
tioning transmitters too close to vessel liquids.
walls or other obstructions can all lead
+ =
C
B’
B B”
Fig 2.18 Interference caused by positioning an antenna too close to the vessel wall. If a
radar level transmitter is installed too close to the vessel wall it is possible that
interference will occur. With indirect reflection A B’ B’’ C, the phase may be
altered by 180° when compared with the direct reflection A B C. For this reason
the microwaves may partially cancel out due to destructive interference
30
2. Physics of radar
The thickness of the dielectric win- other out. This type of installation
dow must be a half wavelength of the is explained more fully in Chapter 6
window material. When the half wave- on the mechanical installations of
length is used, there is destructive inter- radar level transmitters together with
ference between the reflection off the a table showing the optimum thickness
top surface of the window and the of most important plastics and glasses
reflection off the internal second surface which are suitable for penetration with
of the window. radar sensors.
There is a 180° phase shift between
these reflections and they cancel each
reflection with
emitted wave phase shift from top
surface
reflection without
phase shift from
internal surface
plastic vessel ceiling
emitted wave
Fig 2.19 Destructive interference is a benefit when using pulse radar to measure through
a low dielectric window. The reflection from the top surface and the reflection
from the internal second surface cancel each other if the thickness is a half
wavelength
31
Contents
Foreword ix
Acknowledgement xi
Introduction xiii
Part I
1. History of radar 1
2. Physics of radar 13
3. Types of radar 33
1. CW-radar 33
2. FM - CW 36
3. Pulse radar 39
Part II
4. Radar level measurement 47
1. FM - CW 48
2. PULSE radar 54
3. Choice of frequency 62
4. Accuracy 68
5. Power 74
5. Radar antennas 77
1. Horn antennas 81
2. Dielectric rod antennas 92
3. Measuring tube antennas 101
4. Parabolic dish antennas 106
5. Planar array antennas 108
Antenna energy patterns 110
6. Installation 115
A. Mechanical installation 115
1. Horn antenna (liquids) 115
2. Rod antenna (liquids) 117
3. General consideration (liquids) 120
4. Stand pipes & measuring tubes 127
5. Platic tank tops and windows 134
6. Horn antenna (solids) 139
B. Radar level installation cont. 141
1. safe area applications 141
2. Hazardous area applications 144
3. Types of radar
1a. CW, continuous wave radar
In continuous wave or CW Radar, a note is higher as it approaches the lis-
continuous unmodulated frequency is tener and lower as it recedes. The
transmitted and echoes are received doppler effect is also used by
from the target object. If the target astronomers to monitor the expansion
object is stationary, the frequency of of the Universe. By measuring the ‘red
the return echoes will be the same as shift’ of the spectrum of distant stars
the transmitted frequency. The range of and galaxies the rate of expansion can
the object cannot be measured. be measured and the age of distant
However, the frequency of the return objects can be estimated.
signal from a moving object is changed In the same way, when an object that
depending on the speed and direction has been illuminated by a CW Radar
of the object. This is the well known approaches the transmitter, the frequen-
‘doppler effect’. The doppler effect is cy of the return signal will be higher
apparent when the siren note of an than the transmitted frequency. The
emergency vehicle changes as it speeds echo frequency will be lower if the
past a pedestrian. The pitch of the siren object is moving away.
yv
tvelocit
targe
f + f dp
requ ency t
rece ived f
hλ
ave lengt
yf w
frequenc t,
m itted
trans
33
In Fig 3.1, the aircraft is travelling same speed, the received frequency
towards the CW radar. Therefore the would be ft - fdp.
received frequency is higher than The velocity of the target in the
the transmitted frequency and the sign direction of the radar is calculated by
of fdp is positive. If the aircraft was equation 3.1
travelling away from the radar at the
c is the velocity of microwaves
v is the target velocity
λ x fdp c x fdp ft is the frequency of the
v = = transmitted signal
2 2 x ft fdp is the doppler beat frequency
which is proportional to velocity
[Eq. 3.1] ft+fdp is received frequency. The sign
of fdp depends upon whether the
target is closing or receding
34
target
Fig 3.2 The effect of low flying aircraft on television reception is similar to the method of
detection by CW wave-interference radar
3. Types of radar
35
2. FM-CW, frequency modulated continuous wave radar
Single frequency CW radar cannot If the distance to the target is R,
be used for distance measurement and c is the speed of light, then the
because there is no time reference mark time taken for the return journey is:-
to gauge the delay in the return echo
from the target. A time reference mark ∆t = 2xR
can be achieved by modulating the fre- c
[Eq. 3.2]
quency in a known manner.
If we consider the frequency of the We can see from Fig. 3.3 that if
transmitted signal ramping up in a we know the linear rate of change of
linear fashion, the difference between the transmitted signal and measure the
the transmitting frequency and the difference between the transmitted and
frequency of the returned signal will be received frequency fd, then we can
proportional to the distance to the calculate the time ∆t and hence derive
target. the distance R.
frequency
cy
u en
r eq
df
m itte y
ns e nc
tra qu
re
e df
c eiv
re
∆t
fd ∆t = 2xR
c
time
36
3. Types of radar
4.2GHz
frequency
Fig 3.5 Triangular wave
Used on FM - CW
radar transmitters
time
37
If we look at a triangular wave tive value independent of whether the
form we can see that there is an inter- modulation is increasing or decreasing.
ruption in the output of the difference The diagram below makes the
frequency , fd. In practice, the received assumption that the target distance is
signal is heterodyned with part of the not changing. If the target is moving,
transmitted frequency to produce the there will be a doppler shift in the dif-
difference frequency which has a posi- ference frequency.
frequency
time
difference
frequency
fd
time
Fig 3.7 & 3.8 The change in direction between the ramping up and down of the frequency
creates a short break in the measured value of the difference frequency.
This has to be filtered out. The transmitted frequency is represented by the
red line and the received frequency is represented by the dark blue line.
The difference frequency is shown in light blue on the bottom graph
38
3. Types of radar
3. Pulse radar
a. Basic pulse radar
Pulse radar is and has been used of electromagnetic energy, but are in
widely for distance measurement since fact a short wave packet. The number
the very beginnings of radar technolo- of waves and length of the pulse
gy. The basic form of pulse radar is a depends upon the pulse duration and
pure time of flight measurement. Short the carrier frequency that is used.
pulses, typically of millisecond or These regularly repeating pulses
nansecond duration, are transmitted have a relatively long time delay
and the transit time to and from the tar- between them to allow the return echo
get is measured. to be received before the next pulse is
The pulses of a pulse radar are not transmitted.
discrete monopulses with a single peak
The inter pulse period (the time If the pulse period t is 500 microsec-
between successive pulses) t is the onds, then the pulse repetition frequen-
inverse of the pulse repetition cy is two thousand pulses per second.
frequency fr or PRF. The pulse duration In 500 microseconds, the radar pulses
or pulse width, τ, is a fraction of the will travel 150 kilometres. Considering
inter pulse period. the return journey of an echo reflected
The inter pulse period t effectively off a target, this gives a maximum the-
defines the maximum range of the oretical range of 75 kilometres.
radar. If the time taken for the return
Example journey is T, and c is the speed of light,
The pulse repetition frequency then the distance to the target is
(PRF) is defined as
1 Txc
fr = R=
t 2 [Eq. 3.3]
39
b. Pulse doppler radar
The pulses transmitted by a standard The velocity of the target in the
pulse radar can be considered as a very direction of the radar is calculated in
short burst of continuous wave radar. equation 3.4:
There is a single frequency with no
modulation on the signal for the dura- λ x fdp c x fdp
tion of the pulse. c = =
If the frequency of the waves of the 2 2 x ft
transmitted pulse is ft and the target is [Eq. 3.4]
moving towards the radar with velocity
v, then, as with the CW radar already
described, the frequency of the return This is the same calculation as for
pulse will be ft + fdp , where fdp is the CW radar. The distance to the target is
doppler beat frequency. Similarly, the calculated by the transit time of the
received frequency will be ft - fdp if the pulse, equation 3.3.
target is moving away from the radar.
Therefore, a pulse doppler radar can Txc
be used to measure speed, distance and R =
2
direction. [Eq. 3.3]
The ability of the pulse doppler
radar to measure speed allows the sys- As well as being used to monitor
tem to ignore stationary targets. This is civil and military aircraft movements,
also commonly called ‘moving target pulse doppler radar is used in weather
indication’ or MTI radar. forecasting. A doppler shift is measured
In general, an MTI radar has accu- within storm clouds which can be dis-
rate range measurement but imprecise tinguished from general ground clut-
speed measurement, whereas a pulse ter. It is also used to measure the
doppler radar has accurate speed mea- extreme wind velocities within a torna-
surement and imprecise distance mea- do or ‘twister’.
surement.
40
Pulse doppler radar
ft
f t + f dp
Fig 3.10 Pulse doppler radar provides target speed, distance and direction
3. Types of radar
41
c. Pulse compression and ‘Chirp’ radar
With pulse radar, a shorter pulse radiated energy and therefore range but
duration enables better target resolution (with a standard pulse radar) at the
and therefore higher accuracy. expense of resolution and accuracy.
However, a shorter pulse needs a sig- Pulse compression within a ‘Chirp’
nificantly higher peak power if the radar is a method of achieving the
range performance has to be main- accuracy benefits of a short pulse radar
tained. If there is a limit to the maxi- together with the power benefits of
mum power available, a short pulse using a longer pulse. Essentially, Chirp
will inevitably result in a reduced radar is a cross between a pulse radar
range. and an FM - CW radar.
With limited peak power, a longer
pulse duration, τ , will provide more
frequency
f1
f2
time
t1 t2
τ
amplitude
time
Fig 3.11 Chirp radar wave form. Chirp is a cross between pulse and FM - CW radar
42
3. Types of radar
Filter
Time lag
Frequency
Long frequency modulated echo pulse
Compressed
signal
Fig 3.12 Pulse compression of chirp radar echo signal
43
Part II
45
Contents
Foreword ix
Acknowledgement xi
Introduction xiii
Part I
1. History of radar 1
2. Physics of radar 13
3. Types of radar 33
1. CW-radar 33
2. FM - CW 36
3. Pulse radar 39
Part II
4. Radar level measurement 47
1. FM - CW 48
2. PULSE radar 54
3. Choice of frequency 62
4. Accuracy 68
5. Power 74
5. Radar antennas 77
1. Horn antennas 81
2. Dielectric rod antennas 92
3. Measuring tube antennas 101
4. Parabolic dish antennas 106
5. Planar array antennas 108
Antenna energy patterns 110
6. Installation 115
A. Mechanical installation 115
1. Horn antenna (liquids) 115
2. Rod antenna (liquids) 117
3. General consideration (liquids) 120
4. Stand pipes & measuring tubes 127
5. Platic tank tops and windows 134
6. Horn antenna (solids) 139
B. Radar level installation cont. 141
1. safe area applications 141
2. Hazardous area applications 144
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 47
The benefits of radar as a level mea- a radar signal to travel one metre
surement technique are clear. and back takes 6.7 nanoseconds or
Radar provides a non-contact sensor 0.000 000 006 7 seconds.
that is virtually unaffected by changes How is it possible to measure this
in process temperature, pressure or the transit time and produce accurate ves-
gas and vapour composition within a sel contents information?
vessel. Currently there are two measure-
In addition, the measurement accu- ment techniques in common use for
racy is unaffected by changes in densi- process vessel contents measurement.
ty, conductivity and dielectric constant They are frequency modulated continu-
of the product being measured or by air ous wave (FM - CW) radar and PULSE
movement above the product. radar
The practical use of microwave In this chapter we explain FM - CW
radar for tank gauging and process ves- and PULSE radar level measurement
sel level measurement introduces an and compare the two techniques. We
interesting set of technical challenges discuss accuracy and frequency consid-
that have to be mastered. erations and explore the technical
If we consider that the speed of light advances that have taken place in
is approximately 300,000 kilometres recent years and in particular two wire,
per second. Then the time taken for loop powered transmitters.
47
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 48
f2
Transmitted signal
fd
frequency
∆t
Received
signal
f1
t1 time
48
Voltage Controlled Oscillator VCO
Directional Directional f (t + Dt)
Coupler Coupler
f(t)
Linear sweep
V(t) control loop f (t + Dt)
Mixer
Voltage Frequency f(t)
Control Measurement
Filter
Intermediate
frequency
Amplifier
Signal sampling
Front end control function and
Fast Fourier transforms (FFT)
Signal Microprocessor
Fig 4.2 Typical block diagram of FM - CW radar. A very accurate linear sweep is required
4. Radar level measurement
49
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 50
50
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 51
f2
f1
t1
Transmitted signal
Real echo signal
In an active process vessel, the vari- dure which converts the jumbled array
ous echoes are received as frequency of difference frequencies in the time
differences compared with the frequen- domain into a frequency spectrum in
cy of the transmitting signal. These fre- the frequency domain.
quency difference signals are received The relative amplitude of each fre-
by the antenna at the same time. The quency component in the frequency
amplitude of the real echo signals are spectrum is proportional to the size of
small compared with the transmitted the echo and the difference frequency
signal. A false echo from the end of the itself is proportional to the distance
antenna may have a significantly high- from the transmitter.
er amplitude than the real level echo. The Fast Fourier Transform requires
The system needs to separate and iden- substantial processing power and is a
tify these simultaneous signals before relatively long procedure.
processing the echoes and making an It is only when the FFT calculations
echo decision. are complete that echo analysis can be
The separation of the various carried out and an echo decision can be
received echo frequencies is achieved made between the real level echo and a
using Fast Fourier Transform (FFT) number of possible false echoes.
analysis. This is a mathematical proce-
51
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 52
52
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 53
frequency
Fig 4.4 FM - CW frequency spectrum after Fast Fourier transform. The Fast Fourier
transform algorithm converts the signals from the time domain into the frequency
domain. The result is a frequency spectrum of the difference frequencies. The
relative amplitude of each frequency component in the spectrum is proportional to
the size of the echo and the difference frequency itself is proportional to the
distance from the transmitter. The echoes are not single frequencies but a span
of frequencies within an envelope curve
Complex process vessels and solids higher than the real echo. The proces-
applications can prove too difficult for sors that carry out the FFT analysis are
some FM - CW radar transmitters. swamped by different amplitude sig-
Even a simple horizontal cylindrical nals across the dynamic range all at the
tank can pose a serious problem. This same time. As a result, the FM - CW
is because a horizontal tank produces radar cannot identify the correct echo.
many large multiple echoes that are As we shall see, this problem does
caused by the parabolic effect of the not affect the alternative pulse radar
cylindrical tank roof. Sometimes the technique.
amplitudes of the multiple echoes are
53
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 54
Fig 4.5 Pulse radar operates purely within the time domain. Millions of pulses are
transmitted every second and a special sampling technique is used to produce a
‘time expanded’ output signal
54
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 55
To illustrate this principle, consider which is slightly longer than T1, then a
the sine wave signal in Fig 4.6. It is a time expanded version of the original
regular repeating signal with a period sine wave is produced as an output.
of T1. If the amplitude (voltage value) The time scale of the expanded output
of the output of the sine wave is sam- depends on the difference between the
pled into a memory at a time period T2 two time periods T1 and T2.
T1
Periodic
Signal
(sine wave)
Sampling
T2
signal
Expanded
time signal
Fig 4.6 The principle of sequential sampling with a sine wave as an example.
The sampling period, T2, is very slightly longer than the signal period, T1. The
output is a time expanded image of the original signal
T1
Periodic
Signal
(radar echoes) Emission Echo
pulse pulse
T2
Sampling
signal
Fig 4.7 Sequential sampling of a pulse radar echo curve. Millions of pulses per second
produce a periodically repeating signal. A sampling signal with a slightly longer
periodic time produces a time expanded image of the entire echo curve
55
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 56
This periodically repeating signal the same for the emission pulse repeti-
consists of the regular emission pulse tion as for any echo pulse repetition as
and one or more received echo pulses. shown.
These are the level surface and any However, the sampling signal
false echoes or multiple echoes. The repeats at period of T2 which is slight-
transmitted pulses and therefore the ly longer in duration than T1. This is
received pulses have a sine wave form the same time expansion procedure by
depending upon the pulse duration. A sequential sampling that has already
5.8 GHz pulse of 0.8 nanosecond dura- been described for a sine wave. The
tion is shown in Fig 4.8. factor of the time expansion is deter-
The period of the pulse repetition is mined by T1 / (T2-T1).
shown as T1 in Fig 4.7. Period T1 is
Example
The 5.8 GHz, VEGAPULS radar level transmitter has the following pulse repeti-
tion rates.
56
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 57
Sample signal
Fig 4.9 Comparison of switch sampling with ‘cross correlation’ sampling. The pulse
radar uses cross correlation with a sample pulse. This means that rapid ‘picosec-
ond’ switching is not required
57
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 58
ExM
max
Integral 0
ExM
min
Fig 4.10 Cross correlation of the received signal E and the sampling M.
The product E x M is then integrated to produce the expanded time curve. The
technique builds a complete picture of the echo curve
The pulse radar sampling procedure This method transforms the high
is mathematically complicated but a frequency received signal into an accu-
technically simple transformation to rate picture with a considerably
achieve. Generating a reference signal expanded time axis. The raw value
with a slightly different periodic time, output from the microwave module is
multiplying it by the echo signal and an intermediate frequency that is simi-
integration of the resultant product are lar to an ultrasonic signal. For example
all operations that can be handled easi- the 5.8 GHz microwave pulse becomes
ly within analogue circuits. Simple, but an intermediate frequency of 70 kHz.
good quality components such as diode The pulse repetition frequency (PRF)
mixers for multiplication and capaci- of 3.58 GHz becomes about 44 Hz.
tors for integration are used.
58
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 59
t1 t 2 t3 t4 t5
time
transmit pulse
Fig 4.12 With a pulse radar, all echoes (real and false) are separated in time. This allows
multiple echoes caused by reflections from a parabolic tank roof to be easily
separated and analysed
Pulse radar operates entirely within Pulse radar takes literally millions of
the time domain and does not need the ‘shots’ every second. The return echoes
fast and expensive processors that from the product surface are sampled
enable the FM - CW radar to function. using the method described above. This
There are no Fast Fourier Transform technique provides the pulse radar with
(FFT) algorithms to calculate. All of excellent averaging which is particular-
the pulse radar processing is dedicated ly important in difficult applications
to echo analysis only. where small amounts of energy are
Part of the pulse radar transmission being received from low dielectric and
pulse is used as a reference pulse that agitated product surfaces.
provides automatic temperature com- The averaging of the pulse technique
pensation within the microwave mod- reduces the noise curve to allow small-
ule circuits. er echoes to be detected. If the pulse
The echoes derived from a pulse radar is manufactured with well
radar are discrete and separated in time. designed circuits containing good qual-
This means that pulse radar is better ity electronic components they can
equipped to handle multiple echoes and detect echoes over a wide dynamic
false echoes that are common in range of about 80 dB. This can make
process vessels and solids silos. the difference between reliable and
unreliable measurement.
59
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 60
60
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 61
Pulse block diagram - (Fig 4.11) frequency of 3.58 MHz minus 43.7 Hz
The raw pulse output signal (inter- and hence a slightly longer periodic
mediate frequency) from the pulse time. GaAs FET oscillators are used to
radar microwave module is similar, in produce the microwave carrier fre-
frequency and repetition rate, to an quency of the two sets of pulses.
ultrasonic signal. The first set of pulses are directed
This pulse radar signal is derived in to the antenna and the product being
hardware. Unlike FM - CW radar, measured. The second set of pulses are
PULSE does not use FFT analysis. the sample pulses as discussed in the
Therefore, pulse radar does not need preceding text.
expensive and power consuming The echoes that return to the anten-
processors. na are amplified and mixed with the
The pulse radar microwave module sample pulses to produce the raw, time
generates two sets of identical pulses expanded, intermediate frequency.
with very slightly different periodic Part of the measurement pulse sig-
times. A fixed oscillator and pulse for- nal is used as a reference pulse that
mer generates pulses with a frequency provides automatic temperature com-
of 3.58 MHz. A second variable oscil- pensation of the microwave module
lator and pulse former is tuned to a electronics.
Pic 3 Two wire pulse radar level transmitter mounted in a process reactor vessel
61
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 62
Choice of frequency
Process radar level transmitters frequency radar and every frequency
operate at microwave frequencies radar in between.
between 5.8 GHz and about 26 GHz. In reality, no single frequency is
Manufacturers have chosen frequencies ideally suited for every radar level
for different reasons ranging from measurement application. If we com-
licensing considerations, availability of pare 5.8 GHz radar with 26 GHz radar,
microwave components and perceived we can see the relevant benefits of high
technical advantages. frequency and low frequency radar.
There are arguments extolling the
virtues of high frequency radar, low
2.6 GHz
5.8 GHz
Fig 4.14 Comparison of 5.8 GHz and 26 GHz radar antenna sizes. These instruments
have almost identical beam angles. However this is not the full picture when it
comes to choosing radar frequencies
62
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 63
Fig 4.13 For a given size of antenna, a higher frequency gives a more focused beam
63
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 64
64
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 65
65
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 66
frequency range
66
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 67
.. condensation
build - up
. steam and dust
.. wave movement
material cones with solids
.
reflection from medium
signal scattered
Fig 4.19 Signal strength from agitated and undulating surfaces and radar frequency
67
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 68
68
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 69
Fig 4.20 Pulse radar range resolution. Fig 4.21 The null to null bandwidth
The guaranteed range BWnn of a radar pulse is equal
resolution is the length of the to 2 / τ where τ is the pulse
pulse. A shorter pulse has a duration. Example a 5.8 GHz
wider bandwidth and better radar with a pulse duration of
range resolution one nanosecond has a null to
null bandwidth of 2 GHz
Fig 4.23 A shorter pulse duration gives better range resolution. The combination of
shorter pulse duaration and higher frequency allows better accuracy because the
leading edge of the envelope curve is steeper
69
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 70
frequency
∆F x 2R
fd =
Ts x c
[Eq. 4.1]
fd
∆F ∆F bandwidth
Ts sweep time
R distance
fd difference frequency
c speed of light
time
Ts
2
∆fd =
Ts
[Eq. 4.2]
∆fd
frequency
70
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 71
amplitude ∆R ∆fd
=
R R fd
2
∆R Ts
= ∆F x 2 R
R
Ts x c
∆R distance
∆R c
=
R ∆F x R
∆R c
=
∆F
Fig 4.24 to 4.26 - FM - CW range resolution
[Eq. 4.3]
71
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 72
amplitude
envelope curves
around echoes
frequency
amplitude
envelope curves
around echoes
frequency
72
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 73
73
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 74
74
radar_applied_to_level_rb.qxd 15.01.2007 18:46 Seite 75
PULSE radar
· Direct, time of flight level measurement
· Uses a special sampling technique to produce a time expanded intermediate
frequency signal
· The intermediate frequency is produced in hardware and does not require FFT
analysis
· Low processing power requirement mean that practical and very capable two
wire, loop powered, intrinsically safe pulse radar can be used in some of the
most challenging process level applications
75
Contents
Foreword ix
Acknowledgement xi
Introduction xiii
Part I
1. History of radar 1
2. Physics of radar 13
3. Types of radar 33
1. CW-radar 33
2. FM - CW 36
3. Pulse radar 39
Part II
4. Radar level measurement 47
1. FM - CW 48
2. PULSE radar 54
3. Choice of frequency 62
4. Accuracy 68
5. Power 74
5. Radar antennas 77
1. Horn antennas 81
2. Dielectric rod antennas 92
3. Measuring tube antennas 101
4. Parabolic dish antennas 106
5. Planar array antennas 108
Antenna energy patterns 110
6. Installation 115
A. Mechanical installation 115
1. Horn antenna (liquids) 115
2. Rod antenna (liquids) 117
3. General consideration (liquids) 120
4. Stand pipes & measuring tubes 127
5. Platic tank tops and windows 134
6. Horn antenna (solids) 139
B. Radar level installation cont. 141
1. safe area applications 141
2. Hazardous area applications 144
5. Radar antennas
77
Farfield E_Abs (Theta); Phi=90,0 deg.
90
120 60
Max.: 20,4 dB
150 30
180 0
0 10 20 30
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 14,9 deg.
side lobe suppression : 21,6 dB
Radiation patterns of different antennas and radar frequencies are compared at the
end of this chapter.
78
5. Radar antennas
isotropic power
directional power
2
πxD 4π x A
G = ηx ( λ
) = ηx λ2
[Eq. 5.1]
79
Also, we can see that the antenna means that a 26 GHz antenna is lighter
gain and hence directivity is inversely and easier to install for the same beam
proportional to the square of the wave- angle. However, as discussed in
length. Chapter 4, this is not the whole story
For a given size of antenna the beam when choosing the right transmitter for
angle will become narrower at higher an application.
frequencies (shorter wavelengths). For For a standard horn antenna the
example the beam angle of a 5.8 GHz beam angle φ, that is the angle to the
radar with a 200 mm (8") horn antenna minus 3 dB position, can be calculated
is almost equivalent to a 26 GHz radar using equation 5.2.
with a 50 mm (2") horn antenna. This
λ
Beam angle φ = 70° x
D
[Eq. 5.2]
The following graph shows horn anten- most common radar frequencies,
na diameter versus beam angle for the 5.8 GHz, 10 GHz and 26 GHz.
80
5.8 GHz
beam angle in degrees (-3dB)
10 GHz
60
26 GHz
40
20
0
50 75 100 125 150 175 200 225 250
antenna diameter, mm
Fig 5.3 Graph showing relation between horn antenna diameter and beam angle for
5.8 GHz, 10Ghz and 26GHz radar
80
5. Radar antennas
1. Horn antennas
The metallic horn antenna or cone At the transition from the wave-
antenna is well proven for process level guide to the horn of the antenna the low
applications. The horn is mechanically dielectric material is machined to a
robust and in general it is virtually pointed cone. The angle of this cone
unaffected by condensation and prod- depends on the dielectric constant of
uct build up, especially at the lower the material. For example, ceramic has
radar frequencies such as 5.8 GHz. a sharper angle than PTFE.
There are variations in the internal The microwaves are emitted from
design of horn antennas. The this pointed cone in a controlled way
microwaves that are generated within and are then focused towards the target
the microwave module are transmitted by the metal horn.
down a high frequency cable for encou- After reflection from the product
pling into a waveguide. The metal surface, the returning echoes are
waveguide then directs the microwaves collected within the horn antenna for
towards the horn of the antenna. A low processing within the electronics.
dielectric material such as PTFE,
ceramic or glass is often used within
the waveguide.
81
Horn antenna design 1
Fig 5.5
1. HF Cable
1 2. Signal coupling
8. PTFE cone
82
5. Radar antennas
Horn antenna design 2
Fig 5.6
1. HF cable
2. Signal coupling
1
3. Waveguide
(PTFE filled)
2
3
4. Process seals Viton
4 or Kalrez
5 5. PTFE cone
6. Metallic horn
antenna
6
With this antenna design, the HF form for the transition into the horn
cable is encoupled into the PTFE mate- antenna. The PTFE cone and the metal-
rial inside the waveguide. The metal lic conical horn focus the microwaves
waveguide is welded to the flange and and collect the return signals in the
there are two process seals between the usual manner.
metal waveguide and the PTFE. These An antenna of this design is capable
seals protect the signal coupler from of withstanding a process temperature
the process. This seal material can be of 200° C + and a process pressure of
Viton for stainless steel horn antennas 40 Bar.
or Kalrez for Hastelloy C horn anten- This antenna design can also be used
nas. on very high temperature, ambient
There is a continuous transition for pressure applications with air or nitro-
the microwaves within a single piece of gen gas cooling of the antenna.
PTFE which is machined into a cone
83
Horn antenna design 2a
Fig 5.7 Very high temperature, ambient pressure applications.
Air/nitrogen cooling through flange
1. HF cable
1 2. Signal coupling
2 3. Waveguide
(PTFE filled)
3
4 4. Tappings for
air/nitrogen keeps
antenna area cool
Air / N2
5. Metallic horn
antenna
This adaptation of the previous furnace burden level and molten iron
antenna allows the antenna to be cooled ladle levels. The microwaves are unaf-
with air or nitrogen gas. fected by the air movement within the
This is achieved by drilling two horn area.
holes, 180° apart, laterally from the In addition to cooling, this air purg-
flange edge into the horn antenna next ing technique is also used for solids
to the PTFE cone. The flow of air or applications where very high levels of
nitrogen prevents hot gases from conductive dust, such as carbon, heavi-
affecting the PTFE and the viton seal ly coat the inside of the horn and cause
and it effectively cools the entire flange signal attenuation.
and horn area. Water purging has also been used
This technique has been used suc- where heavy product build up is
cessfully with very high temperatures, expected.
including 1500° C + in the steel indus-
try with applications such as blast
84
5. Radar antennas
Horn antenna design 3
Fig 5.8 Special enamel coated antenna
1 1. Signal coupling
2
2. PTFE waveguide
3
4 3. PTFE flange face
5
4. PTFE seal
6
5. Lapped flange
7
6. Steel internals of
horn antenna
7. Enamelled coating
This antenna is also a development between the external cylinder and the
of the antenna design in Fig 5.6. internal horn.
The waveguide, PTFE transition The top of the cylinder has a flange
cone and process flange are standard. for sealing between the PTFE transition
The face of the flange is all PTFE. cone and the process flange and also
The difference is in the application between the glassed antenna and the
of a special enamel (glass) coated horn vessel nozzle. External studs hold the
that provides excellent process materi- enamel antenna to the process flange
als compatibility without resorting to and PTFE seals are used to provide
more expensive metals such as internal sealing.
Tantalum. The antenna is manufactured from
The external dimensions of the carbon steel with blue enamel coating
antenna represent a simple cylinder. which is identical to the enamel found
The internal dimensions of the antenna in glass lined vessels. It provides the
are identical to a standard horn antenna efficiency benefits of a horn antenna
(150 mm (6")) is illustrated. At the bot- with first class materials compatibility.
tom of the antenna there is a gradual lip
85
Horn antenna design 4
Fig 5.9 High temperature / high pressure antenna with ceramic waveguide
1. Connection to HF
cable from
1 microwave module
2 2. Coaxial tube to
signal coupling
3. Signal coupling in
ceramic waveguide
3
4 4.Vacon/ceramic
brazing seal
5
5. Graphite seal
6
6. Ceramic waveguide
cone
The above antenna has been stainless steel expands more than twice
designed with both high temperature as much as ceramic. A double graphite
and high pressure in mind. The seal is fitted on the process side of the
mechanical strength and sealing ability ‘vacon’ bush. The entire waveguide
of PTFE degrades at elevated tempera- assembly is laser welded to ensure that
ture and is therefore limited to about the transmitter is gas tight and that
200° C. differential thermal expansion is
This special design of radar has negligible.
a chemically and thermally stable In order to withstand constant pro-
ceramic (Al2O3) waveguide within a cess temperatures of 400° C, the elec-
stainless steel or Hastelloy C horn tronics housing of the radar is mechani-
antenna and flange. The ceramic cally isolated from the high process
waveguide is fused to a ‘vacon’ steel temperature by a temperature extension
bush using a special brazing technique. tube. The microwave module is con-
‘Vacon’ is used because it has a nected via the HF cable and an air
coefficient of thermal expansion that is coaxial tube to the signal coupler in the
similar to ceramic, whereas normal ceramic waveguide.
86
5. Radar antennas
1
1. HF cable (coaxial)
2
3
2. Signal coupling
4
5 3. Ceramic waveguide
6 4. Brazing of ceramic
to vacon
5. Vacon bush
6. Graphite seal
7. Metallic horn
antenna
7
Vacon/ceramic brazing
87
Adapting horn antenna radars
a. Measurement through a PTFE window
Another possible variation of a horn constant of more that εr = 10, then it is
antenna radar is measurement through possible to measure through a low
a low dielectric window. We have dis- dielectric window or lens.
cussed Hastelloy, Tantalum and the Some antennas are manufactured
special enamel coated horn antenna. with a PTFE window as part of the
However, if a liquid is being measured construction.
and it is conductive or has a dielectric
Antenna housing
Horn antenna
Process flange
PTFE window
Fig 5.12 Horn antenna radar is constructed with a metal housing around the antenna
and a PTFE process ‘window’
Fig 5.13 Variations of this design include the use of cone shaped windows. The cone can
point towards the horn or towards the process
88
5. Radar antennas
b. Horn antenna -
waveguide extension
In the first section of Chapter 6,
Radar level installations, we discuss
how horn antenna radars should be
installed. It is recommended that the
end of the antenna is a minimum of
10 mm inside the vessel. A 150 mm
(6") horn antenna is 205 mm (8") long.
If the nozzle is longer than 200 mm,
we should consider a waveguide exten-
sion piece between the radar flange and
the horn antenna. Waveguide exten-
sions should only be used with highly
reflective products.
Fig 5.14 Extended waveguide horn
c. Horn antenna - antenna to enable measurement
bent waveguide extensions in long nozzles or through a
As well as simple waveguide exten-
concrete tank or sump roof
sions it is possible to bend waveguide
extensions in order to avoid obstruc-
tions or to utilise side entry flanges.
A simple 90° bend or an ‘S’ shaped
extension tube are possible.
The waveguide extensions should be
free from any internal welds and the
minimum radius of curvature should be Waveguide
200 mm. extension with ‘S’
bend
89
High frequency radar antennas
The majority of antennas in this A special patented high frequency
chapter are designed for microwave antenna design from VEGA minimises
frequencies of between 5.8 GHz and the potential problems associated with
10 GHz. Later in this chapter, we dis- small waveguide assemblies.
cuss the use of radar in measuring The encoupling is made within a
tubes where there is a minimum critical small PTFE waveguide to establish a
diameter for each frequency. A measur- single mode. As the microwaves travel
ing tube is a waveguide. The minimum towards the horn antenna, there is a
theoretical tube diameter for a 5.8 GHz carefully designed transition that
radar is 31 mm. increases the diameter of the PTFE
At a higher frequency the minimum waveguide while maintaining the single
diameter of a waveguide is smaller. mode.
At this minimum diameter, the The increased diameter of the PTFE
microwaves are established within the waveguide reduces the adverse effects
waveguide with a single mode and of condensation and build up where the
hence a single velocity. tapered cone of the waveguide enters
As the waveguide diameter increas- the metallic horn of the antenna.
es in size, more modes become estab- Compare this design with horn
lished for the given frequency. antenna design 2, Fig 5.6. The 5.8 GHz
Measurement problems will be radar does not need a transition in the
encountered if there are multiple modes waveguide diameter and the angle of
within an antenna waveguide. This is the metallic horn is not as sharp as for
because with different modes the the high frequency radar.
microwaves travel at different veloci- Viton or Kalrez process seals are fit-
ties in the waveguide and therefore a ted between the PTFE and stainless
single target will reflect more than one steel body of the waveguide.
return echo. Measurement will become Extended versions of the high
inaccurate or impossible. frequency antenna design involve
For this reason, the encoupling of a lengthening the HF cable within a
high frequency radar must be made into stainless steel extension tube and weld-
a small waveguide. The small wave- ing the waveguide assembly to the end
guide assemblies of high frequency of the extension tube.
radar are susceptible to contamination
by condensation and build up when
compared with lower frequencies such
as 5.8 GHz.
90
5. Radar antennas
1. HF cable from
microwave module
3. Carefully designed
transition from small
diameter to larger
1 diameter without
affecting the waveguide
mode
2
4. Viton or Kalrez process
seals between PTFE and
3
stainless steel of the
4 waveguide
5. Cone shape of PTFE
waveguide for the
transition into the
5
metallic horn of the
antenna
6. Metallic horn antenna
of high frequency radar.
6
It has a sharper angle
than the lower frequency
radars
91
2. Dielectric rod antennas
The dielectric rod antenna is an The reflected echoes are captured in
extremely useful option when applying a similar fashion for processing by the
radar level technology to modern radar electronics.
process vessels. Dielectric rods can be Rod antennas should only be used
used in vessel nozzles as small as on liquids and slurries and not on pow-
40 mm (1½") and they are manufac- ders and granular products.
tured from PP, PTFE or ceramic wetted There are some important considera-
parts. tions when applying rod antenna
This means that, normally, radar radars.
level transmitters can be retro-fitted First of all, the tapered section of the
into existing tank nozzles and they rod must be entirely within the vessel.
have low cost materials compatibility If the tapered section is in a nozzle,
with most aggressive liquids including it will cause ‘ringing’ noise that will
acids, alkalis and solvents. effectively blind the radar. This is
The design of dielectric rod antennas explained more fully in Chapter 6.
has been refined in recent years. Also, it can be seen from Fig 5.17
Essentially the microwaves are fed that the microwaves rely on the rod
from the microwave module through an antenna being clean. If a rod antenna is
HF cable to a signal coupler in the coated in viscous, conductive and adhe-
waveguide. As with the horn antenna sive products, the antenna efficiency
the waveguide can be air filled or filled will deteriorate very quickly.
with a low dielectric material such as With the horn antenna product build
PTFE . up is not a particular problem.
The waveguide feeds the However, product build up works
microwaves to the antenna. The against the reliable functioning of a rod
microwaves pass down the parallel antenna radar.
section of the rod until they reach the
tapered section of the rod. The tapered
section of the rod acts like a lens and it
focuses the microwaves towards the
product being measured. The size and
shape of the dielectric rod depends on
the frequency of the microwaves being
transmitted.
92
5. Radar antennas
93
Rod antenna design 1
Fig 5.18 Rod antenna for short process nozzles
1
2 1. HF cable
3
2. Process connection
PVDF boss
3. Signal coupling
4 within PTFE/PP
filled waveguide
4. Inactive section
with metallic wave-
guide, PTFE/PP
inner and outer
parts
5. Solid PTFE/PP
active tapered
section of antenna
focuses the
microwaves towards
the product surface
This rod antenna is a simple and low The HF cable from the microwave
cost design that provides a radar level module is coupled into PTFE/PP inside
transmitter with good materials com- a metallic tube that acts as a wave-
patibility. It is ideal for vented and low guide. This metallic tube is totally
pressure vessels such as acid and alkali enclosed within the PTFE/PP parallel
tanks. It is designed for use in short section of the antenna. The microwaves
1½" BSP / NPT process nozzles. The pass down the metallic waveguide
nozzle height should not exceed 60 mm directly to the tapered section of the
(2½"). antenna where they are focused
The process connection is a 1½" towards the product being measured.
PVDF boss and the antenna is
polypropylene (PP) or PTFE.
94
5. Radar antennas
Rod antenna design 2
Fig 5.19 Rod antenna with solid PTFE extendible rod
1. HF cable
1
2. Signal coupling
2
3 3. Air waveguide
4
5 4. PTFE cone
5. Process connection
With this design of rod antenna the If this type of antenna is to be used
signal coupling is into an air filled in a long nozzle, the parallel section of
waveguide. The microwaves are direct- the solid rod is extended to ensure that
ed towards the antenna. There is a tran- the tapered section is entirely within
sition to PTFE via a cone shaped ele- the vessel.
ment. The microwaves continue An extended, solid PTFE rod anten-
through the PTFE waveguide to the na can suffer from ‘ringing’ noise
solid PTFE dielectric rod. The tapered caused by microwave leakage from the
section of the rod focuses the parallel section resonating within the
microwaves towards the product being nozzle. See Fig 5.20.
measured.
95
Fig 5.20 Extended rod antenna in solid PTFE. This design can suffer from ‘ringing’
noise caused by leakage of microwave energy from the parallel section of the
solid PTFE rod resonating in the vessel nozzle
96
5. Radar antennas
Rod antenna design 3
Fig 5.21 Extended rod antenna with inactive section and signal coupling below nozzle
level
1. HF cable
1 2. Rod extension
casting
(metal within PTFE)
2
3. Signal coupling at
the bottom of the
rod extension
4. Inactive section
3
This antenna is designed for use in tapered sections are sealed together and
nozzles of either 100 mm length or are designed to withstand a process
250 mm length. All wetted parts of the temperature of 150° C .
antenna are PTFE. The parallel section This antenna design is used with
that is designed to be within the nozzle 1½" BSP (M) stainless steel bosses or
has a PTFE coating on a cast metal with PTFE faced flanged transmitters.
tube. The flanged version is designed for
Below this parallel section is the maximum chemical resistance to acids,
active, solid PTFE, tapered antenna. alkalis and solvents. The flange face is
The HF cable from the microwave PTFE with a tight seal between the
module is fed through the metal casting flange PTFE and the top of the PTFE
and the signal coupling is made just covered inactive section.
above the tapered rod. The parallel and
97
Extended rod antenna Extended rod antenna
for 250 mm nozzle for 100 mm nozzle
Fig 5.22 Extended rod antenna with inactive section and signal coupling below nozzle
level. All wetted parts are PTFE on the flanged version of this antenna
For less arduous applications a stainless steel extension tube is used instead of the
PTFE covered tube. The tapered section of the antenna is made of polyphenylene
sulphide (PPS).
Fig 5.23 Extended rod antenna with stainless steel inactive section and PPS rod antenna.
This is for less chemically arduous process conditions
98
5. Radar antennas
Rod antenna design 4
Fig 5.24 Extended rod antenna with metallic grid waveguide extension within carbon
impregnated PTFE inactive rod. Tapered active section of virgin PTFE
1. HF cable
1 2. Signal coupling
2 3. PTFE waveguide
3
4 4. Screwed connection
5
5. Carbon impregnated
PTFE antenna parallel
6 section and flange face
7. PTFE waveguide
8 8. Virgin PTFE tapered
antenna
99
Rod antenna design 5
Fig 5.25 This is a high temperature ceramic rod antenna design. There is temperature
separation between the electronics and the signal coupling (similar to the high
temperature horn antenna Fig 5.10). The ceramic rod has a sharper taper than
the equivalent PTFE rod
1. Signal coupling
2. Ceramic waveguide
Rod antennas are available with the be taken when installing ceramic rods
dielectric rod manufactured from because they are brittle and prone to
ceramic (Al2O3). accidental damage.
Ceramic has good chemical and
thermal resistance. However, care must
100
5. Radar antennas
Bypass tube and stand pipes are used for the following reasons:
101
Measuring tube radar 1 - horn antennas
Fig 5.26 Installation of horn antenna radars into stand pipes or bypass tube
∅ 50 ∅ 80 ∅ 100 ∅ 150
Horn antenna radars are most com- For 80 mm and above, the appropri-
monly used in measuring tube level ate horn antenna is attached and this is
applications. Stilling tube internal designed to fit inside the tube.
diameters can be 40 mm (1 ½"), 50 mm As discussed in Chapter 2, Physics
(2"), 80 mm (3"), 100 mm (4") and 150 of radar and Chapter 6, Radar level
mm (6"). Larger tubes are possible. installations, the linear polarization of
Normally, the 40 mm and 50 mm the radar must be directed towards the
tubes do not require a horn. The PTFE tube breather hole or mixing slots, or
or ceramic waveguide impedance towards the process connections in the
matching cone can be installed directly case of a bypass tube.
into the tube.
102
5. Radar antennas
Measuring tube radar 2 - offset rod antennas
Fig 5.27 Offset rod antenna for use on 50 mm and 80 mm measuring tubes
1. HF cable
1
2. Signal coupling
103
Microwave velocity within measuring tube
The speed of microwaves within a The microwaves bounce off the
measuring tube is apparently slower sides of the tube and small currents are
when compared to the velocity in free induced in the walls of the tube. For a
space. The degree to which the running circular tube, or waveguide, the
time slows down depends on the diam- velocity change is calculated by the
eter of the tube and the wavelength of following equation :
the signal.
{ }
2 the measuring tube / waveguide
λ co
cwg = co x 1- is the speed of light in free
( 1.71d )2 λ
space
is the wavelength of the
microwaves
[Eq. 5.3]
d is the diameter of the measur-
ing tube
104
5. Radar antennas
There are different modes of propa- Equation 5.4 shows the relationship
gation of microwaves within a wave- between critical diameter and wave-
guide. However, an important value is length. For example, 5.8 GHz has a
the minimum diameter of pipe that will wavelength λ of ~ 52 mm. The mini-
allow microwave propagation. mum theoretical tube diameter is
The value of the critical diameter, dc = 31 mm
dc , depends upon the wavelength λ of With a frequency of 26 GHz, a
the microwaves: The higher the fre- wavelength of 11.5 mm, the minimum
quency of the microwaves, the smaller tube diameter is dc = 6.75 mm. In prac-
the minimum diameter of measuring tice the diameter should be higher. The
tube that can be used. diameter for 5.8 GHz should be at least
40 mm.
dc = λ
1.71
[Eq. 5.4]
100
% speed of light, c
80
60
40
20
0
0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
Tube diameter / wavelength, d / λ
Fig 5.29 Graph showing the effect of measuring tube diameter on the propagation speed
of microwaves
105
4. Parabolic dish antennas
Fig 5.30 Typical parabolic antenna
2. Parabolic reflector -
secondary antenna
1
3. Primary antenna
4. Focus of parabolic
reflector
2
3
4
106
5. Radar antennas
The reflected energy is captured by antenna. This can reduce the antenna
the dish and focused back to the prima- efficiency.
ry antenna for echo analysis. Parabolic antennas have been
Parabolic antennas are used widely applied to bitumen storage tanks where
in custody transfer applications and are build up on the parabolic dish is said to
well proven in large storage tanks. cause minimum signal attenuation. If
The benefits of parabolic antennas in the primary antenna was coated in vis-
these applications are clear. The good cous product, this would cause a major
focusing of the paraboloid shape problem to the signal strength.
ensures high antenna gain or directivi- In conclusion, the parabolic antenna
ty. Also this narrow beam angle results has a niche application in fiscal mea-
in higher sensitivity. surement of large, slow moving prod-
However, parabolic antennas are uct tanks, but is not suitable for the
large, heavy, relatively complex and arduous conditions that are prevalent in
expensive to manufacture. These fac- the wide variety of vessels within the
tors limit the use of parabolic antennas process industries.
in most process level applications.
The central feed to the primary
antenna at the focus of the dish causes
a blind area directly in front of the
107
5. Planar array antennas
Fig 5.31 Planar antenna - side view
1
1. Electronics housing
2. Process flange
2
3. Antenna feed
5. Microwave absorbing
material
3
6. Microwave patches
5
6
7
108
5. Radar antennas
Fig 5.32 Cut away of planar array antenna for radar level transmitter
3. Microwave absorbing
material prevents
3 ringing from stainless
4 steel back
109
Antenna energy patterns
At the beginning of this chapter we directivity and minimise the effect of
stated that the definition of ‘beam side lobes.
angle’ is the angle at which the The metallic horn (or cone) antenna
microwave energy measured at the cen- and the dielectric rod antenna are the
tre line of the radar beam has reduced most practical for process level mea-
to 50% or minus 3 dB. surement. The following pages show
We discussed directivity and antenna antenna radiation patterns for different
gain and stated that even the best antenna types, frequencies and sizes.
designed antennas have side lobes of These can be summarised as follows :
energy. The aim is to maximize the
180 0
-10 0 10 20
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 32,1 deg.
side lobe suppression : 16,9 dB
110
5. Radar antennas
180 0
-10 0 10 20
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 27,9 deg.
side lobe suppression : 20,9 dB
180 0
0 10 20 30
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 14,9 deg.
side lobe suppression : 21,6 dB
111
2 Comparison of dielectric rod antenna with horn antenna
The following show a 5.8 GHz horn Although the beam angles are
antenna compared with a 5.8 GHz rod similar, the rod has more significant
antenna. side lobes.
180 0
-10 0 10 20
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 32,0 deg.
side lobe suppression : 14,6 dB
180 0
-10 0 10 20
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 27,9 deg.
side lobe suppression : 20,9 dB
112
5. Radar antennas
180 0
-10 0 10 20
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 18,2 deg.
side lobe suppression : 17,2 dB
180 0
0 10 20 30
150 30
120 60
90
main lobe direction : 0,0 deg.
angular width (3dB) : 9,4 deg.
side lobe suppression : 22,1 dB
113
Inhalt
Vorwort ix
Danksagung xi
Einleitung xiii
Teil I
1. Geschichte des Radars 1
2. Physikalische Grundlagen des Radars 13
3. Radartypen 33
1. CW-Radar 33
2. FMCW-Radar 36
3. Pulsradar 39
Teil II
4. Radar-Füllstandmessung 47
1. FMCW-Radar 48
2. Pulsradar 54
3. Frequenzwahl 62
4. Genauigkeit 68
5. Leistung 74
5. Radarantennen 77
1. Hornantennen 81
2. Dielektrische Stabantennen 92
3. Standrohrantennen 101
4. Parabolantennen 106
5. Planarantennen 108
Richtcharakteristik von Antennen 110
6. Installation 115
A. Mechanischer Einbau 115
1. Flüssigkeitsanwendungen - Hornantenne 115
2. Flüssigkeitsanwendungen - Stabantenne 117
3. Allgemeine Einbauhinweise 120
4. Standrohre und Bypass-Rohre 127
5. Messung durch Behälterwand und Radarfenster 134
6. Messung von Schüttgütern mit Hornantennen 139
B. Elektrische Anschlussvarianten 141
1. Nicht-Ex-Anwendungen 141
2. Geräte für Ex-Anendungen 144
6. Installation
Mechanischer Einbau
Der richtige Einbau ist für die Echoverhältnisse zuverlässig auswerten
Funktion eines Füllstandradars von kann, ist dies immer noch die wichtig-
sehr großer Bedeutung. Obwohl die ste Vorausetzung für eine funktionie-
Signalverarbeitungssoftware moderner rende Messung.
Geräte inzwischen auch schlechte
1. Flüssigkeitsanwendungen - Hornantenne
Stutzen / Muffen
Üblicherweise werden Radarsen- Die Hornantenne eines Gerätes mit
soren auf einem Behälterstutzen oder einem Flansch DN 150 (6") ist z.B.
einer Muffe installiert. Referenzpunkt 205 mm lang. Ist der Montagestutzen
für die Messung ist die Unterseite des deutlich länger als 195 mm, sollte eine
Geräteflansches. Hohlleiterverlängerung verwendet wer-
Die Vorderkante der Hornantenne den. So kann garantiert werden, dass
sollte immer mindestens 10 mm aus das Ende der Hornantenne über den
dem Stutzen heraus in den Behälter Stutzen hinausragt.
ragen.
Korrekter Falscher
Einbau Einbau
10 mm
115
Hohlleiterverlängerung und Minimale Messdistanz bei
gebogene Hohlleiter Geräten mit Hornantenne
Eine Hohlleiterverlängerung sollte Mit einer Hornantenne ist es norma-
verwendet werden, wenn ein Radar- lerweise möglich, flüssige Medien bis
gerät mit Hornantenne in einem langen an die Unterkante der Antenne zu
Stutzen installiert wird. Hierfür wird messen. Dies ist allerdings nur
ein Edelstahlrohr zwischen den PTFE / möglich, wenn die Flüssigkeit gute
keramischen Hohlleiter im Flansch und Reflexionseigenschaften hat.
der Hornantenne montiert. Es ist auch Das Eintauchen der Antennen in die
möglich, die Hohlleiterverlängerung Flüssigkeit, eventuell sogar mit Anhaf-
für einen seitlichen Einbau des Gerätes tungen, verursacht insbesondere bei
abzubiegen. Der minimale Biegeradius 6,3 GHz-Geräten kaum Probleme.
für diesen Antennentyp ist 200 mm, der
Winkel sollte nicht über 90° betragen.
Bei der Verwendung eines geboge-
nen Hohlleiters ist die Ausrichtung der
linearen Polarisation des Radars
wichtig. Die Polarisationsrichtung des
Radars sollte horizontal sein, wenn die
Biegung nach unten verläuft.
Verlängerte und gebogene Hohlleiter
sind für Flüssigkeiten mit guten
Reflexionseigenschaften geeignet. Sie
sollten nicht bei Flüssigkeiten mit
niedrigen DK-Werten oder bei
Schüttgütern verwendet werden.
116
6. Installation
2. Flüssigkeitsanwendungen - Stabantenne
Stutzen / Muffen
Eine PTFE-Stabantenne eignet sich Verbindungen wie 1½" (NPT oder G),
gut bei chemisch aggressiven Produk- Flanschanschlüsse von DN 50 (2") bis
ten wie Säuren und Laugen. Sie wird DN 150 (6") oder hygienische Lebens-
oft in der chemischen und pharmazeu- mittelanschlüsse geliefert.
tischen Industrie benutzt, wo Mischun- Beim Einbau ist wichtig, dass der
gen aus Lösungsmitteln, Säuren und komplette konische Teil der Antenne
Laugen alltäglich sind. aus dem Stutzen in den Behälter
Die PTFE-Stabantennen mit Tri- ragt.
Clamp und spaltfreier Dichtkonstruk- Für den Einbau in langen Stutzen
tion sind speziell für Anwendungen in sind Stabantennen mit unterschied-
der Lebensmittelindustrie und für lichen inaktiven Längen verfügbar.
sterile Behälter optimiert. Typische Längen für diesen inaktiven
Die Stabantenne wird für Flüssig- Teil, und somit die maximale Länge
keiten und Schlämme, aber nicht für des Stutzens, sind 100 mm und
Schüttgutanwendungen benutzt. Der 250 mm.
Sensor ist meistens in einem einfachen
Stutzen oder in einer Gewindemuffe
eingebaut. Radarsensoren mit Staban-
tenne werden passend für geschraubte
117
Falscher Einbau einer Stabantenne
Wenn der konische Abschnitt einer wellen ein starkes Rauschen (Klingeln).
Stabantenne in einem Stutzen montiert Dies führt speziell im Nahbereich zu
wird, erzeugen die abgestrahlten Mikro- einer Verringerung der Messsicherheit.
Abb. 6.5:
Richtig:
Antenne mit angepasstem
inaktiven Teil für lange
Stutzen.
Normale Rauschkurve mit
deutlichem Echo.
Abb. 6.6:
Falsch:
Kurze Stabantenne in einem
langen Stutzen. Produziert
hohes „Klingeln“. Im Nah-
bereich kann dies sogar das
Echo überdecken.
118
6. Installation
Mechanische Belastung
Es sollte beachtet werden, dass die
PTFE-Antennen nur beschränkten
mechanischen Belastungen widerstehen
können. Beim Auftreten einer Quer-
kraft kann sie sich biegen und verfor-
men oder sogar brechen. Hat die
Anwendung starke Füllgutbewegun-
gen? Kann die Biegekraft Schaden am
Stab verursachen?
119
3. Allgemeine Einbauhinweise:
Horn- und Stabantenne bei Flüssigkeitsanwendungen
Folgendes sollte bei der Montage eines Radargerätes mit Horn- oder
Stabantenne auf einem Behälter berücksichtigt werden.
Montage in Behältern mit Paraboleffekt
gewölbtem Deckel Wird ein Radarfüllstandmessgerät
Ein Radarsensor sollte nicht im im Zentrum eines gewölbten Deckels
Zentrum eines gewölbten Deckels oder montiert, empfängt der Sensor stark
zu nahe an der Gefäßwand montiert überhöhte Vielfachechos. Der Effekt
werden. Die ideale Position ist unge- dieser Vielfachechos kann deutlich auf
fähr ½ Radius von der Außenwand ent- der Echokurve betrachtet werden.
fernt. Gewölbte Tankdeckel können Abb. 6.8 zeigt, dass das dritte Vielfache
sonst als parabolischer Reflektor eine deutlich höhere Amplitude
wirken. aufweist als das erste, tatsächliche
Ist der Radarsensor im „Brenn- Echo. Dieser Effekt kann auch in
punkt“ eines parabolischen Deckels liegenden Rundtanks vorkommen.
montiert, empfängt er deutlich über- Vielfachechos können bei Pulsradar
höhte Vielfachechos. Dies wird ver- durch die Software erkannt werden, da
mieden, wenn der Sensor wie zuvor sie zeitlich deutlich getrennt sind. Wie
beschrieben eingebaut wird. bereits in Kapitel 4 beschrieben, ist
dies bei FMCW ein größeres Problem.
Echokurve
r/2
Abb. 6.7: Die ideale Position für das Abb. 6.8: Dieser Effekt tritt auf, wenn das
Gerät ist bei Behältern mit gewölbtem Gerät in der Spitze eines gewölbten
Deckel bei der Hälfte des Radius. Deckels montiert werden.
120
6. Installation
Störechos
Ebene Flächen, Einbauten z.B. Ver- mieden werden, sollten diese mit einem
steifungen oder auch Einbauten mit zur Seite ablenkenden Streublech ver-
scharfen Kanten verursachen große sehen werden. Die dann mehrfach
Störechos. An diesen Objekten werden gebrochenen Radarsignale sind in der
hohe Störamplituden produziert. Runde Amplitude deutlich kleiner und deshalb
Profile hingegen produzieren eine dif- von der Software leichter zu verarbeit-
fuse Reflexion und somit nur geringe en.
Störechos. Sie sind deshalb vom Gerät Diese Maßnahmen müssen umso
leichter zu verarbeiten als große gewissenhafter durchgeführt werden, je
Störechos, die von einer ebenen Fläche geringer der DK-Wert des Produkts ist
stammen. und je höher die Genauigkeitsan-
Können flache Reflexionsebenen im forderungen sind.
Messbereich des Radars nicht ver-
Abb. 6.10: Durch diffuse Reflexion an Abb. 6.11: Ein Streublech verteilt die
runden Teilen werden deutlich gerin- Mikrowellenenergie zur Seite und
gere Störechos produziert. reduziert damit die Störechoamplitude.
121
Vermeiden vom Störechos Absätze
Bei der Einbauposition des Radar- Behälterprofile mit flachen Absätzen
gerätes sollte darauf geachtet werden, rechtwinklig zur Hauptstrahlrichtung
dass sich keine Streben und kein des Radars erzeugen starke Störechos.
Befüllstrom im Detektionsbereich des Durch den Einbau eines Streublechs
Radars befinden. kann die Störechoamplitude deutlich
Die folgenden Beispiele zeigen typ- reduziert werden, um somit eine zuver-
ische Messprobleme und wie sie ver- lässige Messung zu ermöglichen.
mieden werden können. Einbauten mit einer rechtwinkligen
Fläche zum Sensor, z.B. Einlässe,
Achsen, sollten mit einem „Dach“ ver-
sehen werden (Abb. 6.13). Hiermit
wird das Radarsignal ebenfalls ge-
streut, die übrigen Störechos können
von der Signalverarbeitungssoftware
herausgefiltert werden.
122
6. Installation
Behältereinbauten
Einbauten wie z.B. Streben, Leitern, schlecht reflektierenden Produkt ge-
Versteifungen und Sonden verursachen fährden. Durch Anbringen von kleinen
oft Störechos. Durch einen gute Wahl Blechen können diese Störechos
der Einbauposition können viele verkleinert werden. Die Störamplitude
Störechos bereits im Vorfeld vermieden sinkt und kann von der
werden. Signalverarbeitung besser verwertet
Auch Schweißnähte im Behälter werden. Bei der Herstellung des
können Störechos produzieren. Speziell Behälters können Störechos durch
bei höherfrequenten Radargeräten, die Verschleifen der Schweißnähte
nahe an der Wand montiert sind, minimiert werden.
können diese die Messung bei einem
123
Anhaftungen
Ist der Radarsensor zu nahe an der Der Sensor sollte deshalb immer etwas
Behälterwand montiert, können Pro- Abstand zur Behälterwand haben. Der
duktanhaftungen Störechos erzeugen. ideale Kompromiss ist ½ Radius.
Polarisation
Wie schon in Kapitel 2 besprochen, von Störechos, z.B. von Streben oder
sind die Mikrowellen der VEGA - der Behälterwand, kann oft durch
Radargeräte linear polarisiert. Drehen des Radarsensors um 45º oder
Obwohl die Polarisation eine grö- 90º reduziert werden.
ßere Bedeutung in Standrohren und Die Richtung der Polarisation wird
Bypassrohren hat, kann sie auch bei durch das Einkoppelsystem festgelegt,
Anwendung in „normalen“ Behältern es ist am Gerät durch die Position des
von Bedeutung sein. Die Amplitude Typenschildes erkennbar.
124
6. Installation
Fließende Produkte
Ein Radarsensor sollte nicht direkt mieden, dass anstelle der Produkt-
über oder in der Nähe einer Befüllung oberfläche der Befüllstrom gemessen
montiert werden. Dadurch wird ver- wird.
125
Sensor zu nah an der Behälterwand
Wird der Radarsensor zu nahe an der messgeräte unterschiedliche Öffnungs-
Behälterwand montiert, kann dies winkel (Kapitel 5: Radarantennen).
starke Interferenzen verursachen. Die Im Allgemeinen sollte darauf ge-
Echos von Anhaftungen, Nieten oder achtet werden, dass sich die Behäl-
Schweißnähten überlagern sich mit terwand nicht innerhalb des 3dB-Öff-
dem richtigen Echo. Es muss aus- nungswinkels der Antenne befindet.
reichend Abstand vom Sensor zur Bei ungünstigen Einbaubedingungen
Behälterwand eingehalten werden, um bzw. Störungen durch die
dies zu verhindern. Behälterwand können die Mess-
Abhängig von der Antennengröße verhältnisse durch Verändern der
haben verschiedene Radarfüllstand- Polarisation optimiert werden.
127
E E E
128
6. Installation
Abb. 6.23: Durch Schlitze wird eine gute Durchmischung von inhomogenen Produkten
erreicht. Die Polarisation muss in Richtung der Schlitze ausgerichtet werden.
129
Anhaftende Produkte Messrohr mit Kugelhahn
Um Messprobleme und Messfehler Zur Abtrennung des Rohrs bzw. des
bei der Messung von anhaftenden Messgeräts vom Prozess kann ein
Produkten in Standrohren zu vermei- Kugelhahn verwendet werden. Mit dem
den, sollte das Rohr einen Innen- Kugelhahn ist es möglich, Wartungs-
durchmesser von mindestens 100 mm arbeiten durchzuführen, ohne den
(4") haben. Sollen inhomogene Behälter zu öffnen. Dies ist bei
Produkte oder Produkte gemessen wer- Flüssiggas und giftigen Erzeugnissen
den die eine Trennschicht ausbilden, besonders wichtig. Bei geöffnetem
muss das Standrohr Löcher oder lange Ventil sollten möglichst keine Kanten
Schlitze haben. Diese Öffnungen stel- im Durchlass zu sehen sein, dies würde
len sicher, dass die Flüssigkeit durch- sonst zu Störechos führen.
mischt wird und, dass sie sich an den
richtigen Füllstand angleicht. Je inho-
mogener das Produkt, desto mehr Öff-
nungen müssen vorhanden sein.
Die Löcher und Schlitze müssen aus
Gründen der Polarisation in zwei um
180º versetzten Reihen positioniert
werden. Der Radarsensor muss so aus-
gerichtet werden, dass die Polarisation
in Richtung der Löcher ausgerichtet ist.
E E
130
6. Installation
131
Diagramm 1
Radarsensor
VEGAPULS 54
Flansch DN 50
Rohrdruchmesser 50 mm
Vorschweißflansch
100%
2.9…6
5…15
Schweißungs der
150…500
Verbindungsmuffe
0.0…0.4
Verbindungsmuffe
Rz ≤ 30
Schweißung des
Vorschweißflansch 2.9 Vorschweißflansches
0.0…0.4
1.5…2
Löcher müssen
gratfrei sein
minimal messbare
0% Füllhöhe (0%)
Alle Abmessungen in mm
Abb. 6.26
132
6. Installation
Diagramm 2
Radarsensor
VEGAPULS 54
Flansch DN 100
Rohrdurchmesser 100 mm
Schweißflansch
Schweißung des
100% Schweißflansches
3.6
5…15
150…500
Schweißung der
Verbindungsmuffe
0.0…0.4
Verbindungsmuffe
Rz ≤ 30
Vorschweißflansch 3.6
Schweißung des
Vorschweiflansches
0.0…0.4
1.5…2
Löcher müssen
gratfrei sein
133
5. Messung durch die Behälterwand und Radarfenster
Die Mikrowellensignale von Radar- Ein solche Messung ist bei Produk-
füllstandmessgeräten durchdringen ten mit guten Reflexionseigenschaften
dielektrische Materialien wie z.B. möglich, sie können bei geeignetem
PTFE, Polypropylen und Glas. Dies ist Behältermaterial direkt von oben,
für einige Anwendungen sehr wichtig, durch die Behälterdecke, gemessen
z.B. bei der Messung von hochreinen werden. Produkte mit guter elektrischer
Flüssigkeiten in der Pharmaindustrie Leitfähigkeit und mit einer Dielek-
oder der Halbleiterfertigung, oder bei trizitätszahl von mehr als 10 sind dafür
hochaggressiven Produkten in der geeignet. Bei Messungen in denen es
chemischen Industrie. In diesen Fällen prozess- oder produktbedingt zu star-
ist es aus Sicherheitsgründen und im ken Niederschlägen oder Kondensation
Hinblick auf die Produktqualität von an der Behälterdecke kommt, ist dieses
Vorteil wenn der Behälter geschlossen Verfahren mit Vorsicht anzuwenden.
bleibt.
Abb. 6.28: Gut reflektierende Medien können direkt durch die Behälterwand oder durch
ein Messfenster gemessen werden.
134
6. Installation
Abb. 6.29: Eine flache Behälterdecke Abb. 6.30: Die Messung über einem
produziert eine Störreflexion direkt angeschrägten Bereich des Behälter-
zurück in die Antenne. deckels verbessert die Messung deut-
135
Messung durch ein dielektrisches Fenster
In einigen Ländern ist es verboten Bei Messungen durch ein Fenster
FMCW-Radar-Füllstandmessgeräte kann eine Verbesserung erzielt werden,
außerhalb eines Metallgefäßes zu wenn die Scheibe eine konische Form
betreiben. In solchen Fällen muss das erhält (siehe Abb. 6.32). Solch eine
Gerät, um die Vorteile eines „dielek- Trennscheibe kann bei geeigneter
trischen Fensters“ nutzen zu können, in Dimensionierung als Linse wirken und
einem metallischen Stutzen über einem die Mikrowellen zusätzlich fokus-
Kunststoff oder Glasfenster installiert sieren. Diese Form begünstigt zusätz-
werden (Abb. 6.32). Dies kann jedoch lich das Ablaufen und Abtropfen von
einen hohen Störpegel verursachen. Kondensat.
Radar-Sensor
metallischer
Stutzen
konische
Teflonscheibe
Abb. 6.32
136
6. Installation
Reflexion mit
Gesendete Phasendrehung
Welle von der Oberfläche
Reflexion ohne
Phasendrehung von
der inneren Oberfläche
Kunststoffdeckel
Sendesignal
{
Reflexion mit
Phasendrehung
Reflexion ohne Gegenseitige Auslöschung
Phasendrehung
Abb. 6.33: Die optimale Dicke des Fenstermaterials beträgt λ/2 der Radarfrequenz.
137
Die Tabelle zeigt die optimale Dicke für die wichtigsten Kunststoffe und Gläser
die zum Durchstrahlen geeignet sind. Es wird die optimale Dicke für
6,3 GHz und 26 GHz gezeigt.
Anmerkung: Die optimale Dicke kann auch durch Aufschichten einiger Lagen
identischen Materials erreicht werden. Die Schichten müssen jedoch ohne Luftspalt
aufeinander liegen. Vielfache der optimalen Dicke führen ebenfalls zu guten
Ergebnissen, jedoch verursacht die Dicke des Fenstermaterials eine
Signaldämpfung.
138
6. Installation
139
Abb. 6.35 und 6.36: Schüttkegel von typischen Schüttgutanwendungen beim Befüllen und
Entleeren.
140
6. Installation
B. Elektrische Anschlussvarianten
In den vergangenen Jahren hat sich Seit ihrer Markteinführung haben
die Auswahl an unterschiedlichen sich eigensichere Zweileiter-
Radar-Füllstandmessgeräten erhöht. Radarsensoren als vollwertiger Ersatz
Zudem haben sich eine Vielzahl von für traditionelle Sensoren wie z.B.
elektrischen Anschlussmöglichkeiten Differenzdruckmessumformer oder
für Standard- und Ex-Anwendungen Verdränger durchgesetzt. FMCW-
auf dem Markt etabliert. Diese Radarsensoren benötigen jedoch noch
umfassen 4 … 20 mA- und ver- immer die erhöhte Energie aus einer
schiedene Feldbussensoren. Bei der Vierleiterversorgung. In diesem
Auswahl eines Radarsensors müssen Abschnitt werden die möglichen
die entsprechenden Verkabelungs- Beschaltungskonfigurationen für alle
kosten berücksichtigt werden. Arten von Radar betrachtet.
1. Nicht-Ex-Anwendungen
a. 4 … 20 mA, Zweileiter-Radarsensor
4 … 20 mA, 24 VDC
Abb. 6.38
Abb. 6.39
c. HART®-Protokoll
Die meisten Zweileiter- und Vierleiter-, 4 … 20 mA Radar-Füllstandmess-
geräte sind mit dem HART®-Protokoll, aufmoduliert auf dem Stromsignal,
verfügbar. Dadurch wird Folgendes möglich:
- Fernparametrierung mit dem HART®-Handheld Programmiergerät
- Einspeisung der HART®-Daten direkt in das Prozessleitsystem
- multi-drop Betrieb mit bis zu 16 Sensoren parallel an einem Strang
141
d. Feldbus (VBUS)
142
bis zu 15 Sensoren parallel auf zwei Drähten
mit VEGALOG 571 und EV-Eingangskarten maximal 255 Messungen zusammenfassbar
VBUS
bis zu 15 Sensoren an
einer Zweidrahtleitung
verschiedene Industrie-Standard-
Kommunikationen
Abb. 6.40
e. Feldbus (Profibus PA)
max. 32 Sensoren gemeinsam an einem Segmentkoppler
Profibus PA
Segmentkoppler
Profibus DP
Abb. 6.41
6. Installation
143
2. Geräte für Ex-Anwendungen
a. eigensicher ia, 4 … 20 mA, Zweileiter-Sensoren mit HART®-Protokoll
Ex ia 4 … 20 mA, 24 VDC
Zener-
barriere
Ex-Bereich Nicht-Ex-Bereich
Abb. 6.42
Ex ia
4 … 20 mA,
Ex d 24 VDC Ex e
Bedienung,
Display und
Elektronik
eigensicher Zener
ausgeführt barrier
Ex-Bereich Nicht-Ex-Bereich
Abb. 6.43
144
6. Installation
4 … 20 mA Ex d 24 VDC, Ex e
eigensicher
Zener
barrier
Nicht-Ex-Bereich
Ex-Bereich
Abb. 6.44
24 VDC, Ex e
Ex d
4 … 20 mA
Nicht-Ex-Bereich
Ex-Bereich
Abb.6.45
Nicht-Ex-Bereich
Ex-Bereich Abb.6.46
145
f. Feldbus (VBUS)
146
- max. 15 Sensoren an zwei Leitungen in Ex e, Verdrahtung in erhöhter Sicherheit
- separate Energieversorgung der Sensoren in erhöhter Sicherheit
Separate
Spannungs-
versorgung
Ex e
VBUS
bis zu 15 Sensoren an einer
Zweidrahtleitung, Ex e
verschiedene Industrie-
Standard-Kommunikationen
Abb 6.47
g. Feldbus (VBUS)
- max. 15 Sensoren Ex-e (je fünf Sensoren pro Strang, drei Stränge) pro VBUS-Karte
- Verdrahtung in erhöhter Sicherheit ohne externe Versorgung VBUS
Fünf Sensoren an jeder Zweidraht-
leitung versorgt durch diese Leitung
Ex e
Verkabelung
VEGALOG 571
verschiedene Industrie-
Standard-Kommunikationen
6. Installation
Abb. 6.48
147
h. Feldbus (VBUS)
148
- max. 15 Sensoren, eigensicher Ex ia, pro Ausgangskarte
- je 5 Sensoren pro Zweileiter- Schleife, max. 3 Schleifen pro EV-Karte
VBUS
VEGALOG 571
verschiedene Industrie-
Standard-Kommunikationen
VBUS
Fünf Sensoren an einer Zweidrahtleitung
eigensicher
Abb. 6.49
i. Feldbus (Profibus PA)
- Ex ia eigensicher, max. 8 Sensoren pro Zweiader-Schleife
- Verbindung über Segmentkoppler zu Profibus DP
Profibus PA
Segmentkoppler
Profibus DP
Acht Sensoren an
einer Zweidrahtleitung eigensicher
6. Installation
Abb. 6.50
149