Escolar Documentos
Profissional Documentos
Cultura Documentos
Ultrasonic Inspection
Level 1,II
Course Literature
Hitan Co.Ltd.
Ground Floor - No. 15 - 2ndSt. - Pakistan St. - Dr. Beheshti St. - Tehran - IRAN
Phone : +98-21-8745034,5 Fax : +98-21-8747685
Email : hitan@,hitanco.corn http://www.hitanco.corn
COMPREHENSIVE PRACTICAL ULTRASONIC
WELD EXAMINATION
NDT 4
INTRODUCTION 4
PRINCIPLES OF ULTRASONIC
TEST1NG 12
GENERATING ULTRASOUND 36
COMPRESSION PROBES 58
1.0 lNTRODUCTlON
The text for this course is laid out in a manner which it is hoped will make r:
easier to follow than conventional course texts.
In general, right hand pages are used for text and left hand pages for flow
charts, diagrams and tables. Looking across the page to the right of a
particular diagram you should find the relevant text.
Enough space is left on the pages to encourage you to add notes from the
lectures.
It is hoped that the flow charts will prove useful to follow the progress of the
course lectures.
Because flow charts are used there is no index. Each flow chart splits a
subject title into several sub-headings, given with a decimal notation for the
paragraph number. Therefore the number 2.2.31 means paragraph numbe:
31, under sub-heading number 2 of subject title 2.
As far as is practical the terms used in these notes are those defined in
BS3683 Part 4, 1985 revision.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
.*
SI UNITS OF HEASUREXENT
Base q u a n t i t i e s
Length metre 1
Mass kilogram kg
Time second sec
Electric current ampere A
Thermodynamic temperature kelvin K
Luminous i n t e n s i t y candela cd
Amount of s u b s t a n c e mole mo 1
Derived u n i t s
Volume li t r e
a s s tonne
Energy electron
volt approx. 1.60219 x 10-' ' J
Prefixes
tera
gigs
mega
kilo
hecto
deca
dec i
cent1
milli
micro
nano
pic0
f emto
atto
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Certainly not all of these units are of relevance to this course, but the table
will be useful reference.
For example:
7.0 x lo3 = 7000
1
but mas-'= m x -
S
Penetrant testing
Radiography
Ultrasonic testing
The origin of modern ultrasonic testing is the discovery of the Curie brothers
in 1880 that quartz crystals cut in a certain way produce an electric potential
when subjected to pressure - the piezo-electric effect, from the Greek
piedzein, to press or strike.
In 1881, Lippman theorized that the effect might work in reverse, and that
quartz crystals might change shape if an electric current was applied to
them. He found this was so and experimented further. Crystals of quartz
vibrate when alternating currents are applied to them. Crystal microphones
in a modern stereo rely on this principle.
When the "Titanic" sank in 1912, the Admiralty tried to find a way of locating
icebergs by sending out sound waves and listening for an echo. They
experimented further with sound to detect submarines during the First World
War. Between the wars, marine echo sounding was developed and in the
Second world War ASDlC (Anti Submarine Detection Investigation
Committee) was extensively used in the Battle of the Atlantic against the U
boats.
In the 1930's the cathode ray tube was developed and miniaturised in the
Second World War to fit small airborne radar sets into aircraft. It made the
ultrasonic testing set as we know it possible.
The first flaw detector was made by D 0 Sproule in 1942 while he was
working for the Scottish firm Kelvin & Hughes. Similar work was done by
Firestone in the USA and by German physicists. Sproule went on to
develop the shear wave probe.
At first, ultrasonic testing was restricted to testing aircraft, but in the 1950's
UT was extensively used in the building of power stations in Britain for
examining thick steel components safely and cheaply.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
UT was found to have several advantages over radiography in heavy
industrial applications:
It did not have the health hazard associated with radiography, and a
UT technician could work next to welders and other employees
without endangering them or holding up work.
It was efficient in detecting toe cracks in boilers - a major cause of
explosions and lack of fusion in boiler tubes.
UT found planar defects like laminations which were
sometimes missed by radiography.
A UT check on a thick component took no more time than a similar
check on a thin component as opposed to long exposure times in
radiography.
Over the years, UT sets have been miniaturised with the availability of
transistors and display features improved. The process has also been
automated and computerised with varying degrees of success.
SCHOOL OF APPLIED NON-DESTRUCTIVE TES77NG
2.1 SOUND
2.3 FREQUENCY
i
2.4 WAVELENGTH
2.0 PRINCIPLES OF
ULTRASONIC TESTING
2.7 DIFFRACTION
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Sound
Sound is made when something vibrates. You can twang a ruler on a
table or flick a stretched elastic band to verify this. The stretched
surface of the rubber band or the ruler vibrates and sats up a series of
vibrations, sound waves, in the air. As the surface of the band or ruler
pushes into the air, the air molecules are forced together and a region
of high pressure forms: compression. As the surface moves back, the
air molecules move apart, forming a low pressure area, or rarefaction.
As the surface vibrates, alternate compressions are rarefactions are set
up in the air and travel out from the surface to form a sound wave.
The air molecules don't move with the wave - they vibrate to and fro in
time with the vibrating surface.
The sound wave so produced travels through the air at a speed of 332
metres per second, at o'C. We hear the sound when it hits a
membrane in our ear and cause it to vibrate.
Sound will travel through any medium that has molecules to move, bl;:
it travels faster in more elastic materials because the vibrations are
passed on more quickly. Sound travels faster in water or metal than i?
does in air as liquids and solids are more elastic than air.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
2.2.1 COMPRESSIONAL OR
LONGITUDINAL WAVE
2.2.2 SHEAR OR
TRANSVERSE WAVE
L . 2.2 MODES OF
SOUNDENERGY
2.2.3 SURFACE OR
RAYLEIGH WAVE
Air
Water
Steel
Aluminium
Perspex
Copper
Brass
testing range
Compressional wave
We cannot hear all sound. What we can hear is sound in a
COMPRESSIONAL mode, where molecules vibrate backwards and
forwards in the same direction as the energy of propagation - rather like
~illiardballs in a line. A COMPRESSIONAL wave of sound is also called
a LONGITUDINAL wave: waves of this type consist of alternate
compressional and dilation in the direction of propagation. As each
particle moves it pushes or pulls the adjacent particle through elastic
interconnection. Gases, liquids and solids have elasticity, so
compressional waves can travel in all of them.
Note that sound can only travel through air and water in the
COMPRESSIONAL mode. Sound can travel through perspex, steel and
duminium in modes other than the compressional modes.
Shear Wave
Sound can travel in solids in a SHEAR mode as well as in a
compressional mode. In the SHEAR mode, molecules vibrate up and
cown, across the direction of propagation, not to and fro, and foi this
-eason the SHEAR mode is also called the TRANSVERSE mode, as
zarticle vibration is transverse to the direction of sound energy.
Tnis type of sound travel can only happen when the molecules through
which it propagates are joined together - in a solid. A solid has rigidity as
well as elasticity. Air and water, like other gases and liquids, do not have
rigidity. SHEAR or TRANSVERSE waves cannot travel in gases of liquids
for this reason.
Compressional
Shear
Direction of propagation
Particle
w vibration
Surface
0 3000m/sec
Fig.2.2.113Modes of propagation
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
LAMB or PLATE waves propagate in thin plate materials when the plate
thickness is about the same as the wavelength.
LAMB or PLATE waves travel at velocities which vary with the plate
thickness and the wavelength.
Frequency
As sound is a series of vibrations, one way of measuring it is to count the
number of vibrations per second - the frequency. Frequency is measured
in Hertz. One vibration in one second is one Hertz. Two vibrations in one
second is two Hertz. Ten vibrations in one second is 10 Hertz and 1000
vibrations in one second is 1000 Hertz or one Kilohertz (kHz). One million
vibrations in a second is one Megahertz (MHz). The higher the frequency
the higher the note sounds - the higher the pitch. If you twang the ruler
or the rubber band hard, the noise is louder, it has greater amplitude, but
the note remains the same. If, however, you shorten the ruler or tighten
the rubber band, they vibrate more quickly and the note given out is
higher. The frequency is greater. To raise the pitch of their instrument,
guitar players move their fingers down the frets, thus shortening the string
and making it vibrate more quickly.
However, there is an advantage for the lower frequencies. The lower the
frequency, the more penetrating a sound wave is - that is why foghorns
give out very low notes and why the low throbbing notes from your
neighbour's stereo set come through the wall rather than the high notes.
SCHOOL OF APPUED NON-DESTRUCTIVE T'ESnNG
1 Wavelength,:
F
I =1h I
i
f = no of wavelengths I
-1
time
f = frequency = Kt
I $ t = time
A= lanbda wavelength
2.4 Wavelength
A wave in the sea is a vibration of energy. As the wave passes a 3xed
point it produces a constant rise and fall of energy.
WAVELENGTH = VFLOCITY
FREQUENCY
-
- 1.66 metres
-
- 0.00296 metres or 2.96 millimetres
-
- 0.001625 metres or 1.625 millime::es
It can be seen that waveiengi-.lof the shear wave is less ihan tha; of the
compression wave.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
We can calculate how much sound is transmitted and how much sound is
reflected back by knowing the acoustic impedance of both materials.
Once you know the acoustic impedances of two materials, you can use a
formula to work out how much sound will be reflected back. The formula
is:
88% of the sound energy is reflected back at the interface. This of course
means that 12% of the energy is transmitted at the interface. Using the
same formula, the figures for other media can be worked out. At a
steelloil interface, 91% of sound energy is reflected back; at a
glycerine/steel interface 90% of energy is reflected back.
In ultrasonic testing a sound pulse is sent into a solid object and an echo
returns from any flaws in that object or from the other side of the object.
An echo is returned from a solid-air interface or any solid-non-solid
interface in the object being examined.
If that material contains a flaw, sound energy is reflected back from the
flaw and appears on the screen as a signal in front of the backwall echo:
the sound reflected from the flaw has not had so far to travel as the sound
reflected from the backwall. See Fig 2.6.a.
Anything that send back sound energy to a probe to cause a signal on the
screen is called a "reflector".
By measuring the distance from the edge of the CRT screen to the signal,
we can calculate how far down in the material the reflector lies.
Laminar
defect
defect
Probe
I ,Q
Test
material
Fig.2.7 Diffraction
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
2.7 Diffraction
When a sound beam passes through an aperture or past a sharp edge the
energy will bend around and behind the edge or edges. This is tre
phenomena known as diffraction and is responsible for some ene:gy not
being able to return to the transmission point.
The phenomena has been used in a defect sizing method once known as
'Crack Tip Diffraction', but now more commonly referred to as Time of
Flight'.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
3.0 GENERATING
ULTRASOUND
3.3 PULSE LENGTH AND DAMPING
3.4 RESOLUTION
Off- TI 4
The piezo electric effect changes mechanical energy into electrical energy.
It is reversible, so electrical energy - a voltage - can be changed into
mechanical energy or sound, which is the reverse piezo electric effect.
The first people to observe the piezo electric effect were the Curie
brothers who observed it in quartz crystals.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
3.1.1. QUARTZ
3.1.l. Quartz
Quartz or silicon oxide (Si 0,) is found in granite as a natural crystal.
Compressional or shear waves can be produced according to the way
quartz crystals are cut. An "Xucut crystal is cut in a direction that directly
crosses the axis joining two angles of the crystal. A "Y" cut crystal is cut
in a direction parallel to the axis joining two angles of the crystal.
For these reasons quartz has been largely superseded by other piezo
electric materials.
3.1.2. Lithium sulphate (Li SO,) crystals grow as a solution of lithium stilphate is
evaporated. Their advantages are:
a. Lithium sulphate is the'most efficient receiver of ultrasound
b. It has very low electrical impedance
c. Lithium sulphate operates well at low voltages
d. It does not age
e. It has very good resolution
f. Lithium sulphate crystals are easily damped to give short pulse
lengths
all of which makes them unsuitable for industrial use, though lithium
sulphate crystals are used for medical ultrasonics for the examhation of
pregnant women and patients suffering from tumours.
Lead metaniobate (Pb Nb 0,) crystals are made in a similar way to barium
titanate. Its qualities are that it:
a. has heavy internal damping
b. gives out very narrow pulses of ultrasound, which gives good
resolution
but is has the disadvantage of being much less sensitive than lead
zirconate titanate or PZT.
Lead zirconate titanate (Pb Zr 0,. Pb Ti 0,) is the best all-round crystal
for industrial testing for its qualities. It :
a. has a high Curie point, up to 350°c
b. has good resolution
c. does not dissolve in water
d. is tough and resistant to ageing
e. iseasilydamped.
3.2.1. COMPRESSIONAL
SINGLE CRYSTAL PROBES
Electrical connections
,Damping
h'iembrane
Transducer
/
Test material
Transducer
S
Test material
i
C - Compressional w a v e
I - incidenr angle
R - R e f r a c e d angle
S - Shear w a v e
Angle Probes
Angle probes utilise compressional probes mounted on a wedge of
perspex. The wedge of such a probe is cut to a particular angle to enable
the beam to refract into the test material at a chosen angle.
Receiver Transmitter
crystal crystal
Focusing
Insulator plastic lens
The signal on the screen caused by the initial pulse may mask signals
from flaws close to the top of the parent material. For this reason the area
on the CRT screen in which flaws may be masked is called the DEAD
ZONE.
A twin crystal probe has two crystals mounted on perspex shoes angled
inwards slightly to focus at a set distance in the test material. Were the
crystals not angled, the pulse would be reflected straight back into the
transmitting crystal.
The perspex shoes hold the crystals away from the test surface and so
that the initial pulse does not appear on the CRT screen. The dead zone
is greatly reduced to the region adjoining the test surface, where the
transmission and reception beams do not overlap.
I '.
I::
SCHOOL OF APPLlED NON-DESTRUCTIVE TES77NG
1 cycle
-- -------
--
Amplitude reduced
to 1/10 of maximum
Pulse length (or width) is also sometimes called "wave train length". It is
defined in a number of ways and even British Standards disagree. We
choose the one in BS4331.Pt.2. which defines it as the first and last
instant at which the value of the pulse reaches 10% of its peak amplitude.
See Fig.3.3.a.
The longer the pulse length, the more penetrating the ultrasound, as it
contains more energy, but the worse the sensitivity and resolution; hence
the need to compromise.
SCHOOL OF APPLIED NON-DESTRUCTNE TESllNG
- -
;'C4,: Amphtude
-
t
\r Short pulse ,?rcCe - ' to 2 Cjcles
a) Resolved 5 ) N o t resolved
3.4 Resolution
Resolution is the ability of an epipment/combination probe to distinguish
between two echoes from ref1ec:crs that are close together. To have good
resolution a probe must present ?NO signals on a CRT screen from two
separate reflectors: if it has poor resolution the echoes from the two
reflectors appear as one signal on the screen.
In the early days of ultrasonic testing we used the 1OOmm, 91mm and
85mm steps, at the radiused end of the block to test resolving power.
However, today this is regarded as much too crude a test and BS4331
pt.3 recommends that we should be able to recognise two discrete echoes
less than two wavelengths apar;. By discrete echoes they mean split by
more than 6dB, see Fig.3.4. or to more than half the total height of the
signals.
For a single crystal probe the lergth of the initial pulse is the dead zone,
for any signal from a reflector e: a shorter distance than this will be
concealed in the ir%tialpulse.
We deliberately delay off the ir:,: al pulse beyond the left of the timebase,
by mounting the transducers of z twin crystal probe onto plastic wedges.
This and the focusing of the crystals reduces the dead zone considerably
and it is only where the transrxsion and receptive beams do not overlap
that we cannot assess flaws.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
4.1 CALIBRATION
4.1 Calibration
Calibration is setting at a timebase on a cathode ray tube screen to
represent a chosen distance.
Adjust the RANGE and DELAY controls on the set until you have four
echoes on the screen.
Adjust the two controls further until the first echo is 25% screen width, the
second echo 5056 screen width, the third echo at 75%, the fourth echo at
full screen width. The first echo now represents 25mm, the second 50mm,
the third 75mm and the full screen width 100mm.
Check your calibration by putting the V1 block on its edge and taking a
reading from one edge to the other; if you have calibrated correctly you
will get a signal at full screen width, 100m.
Put the probe on the plastic insert in the V1 block and you will get an echo
signal midway across the screen at 50mm - sound travels more slowly in
perspex than it does in steel, so the 23mm of perspex gives a signal at the
same distance across the timebase as 50mm of steel.
Adjust the RANGE and DELAY controls until the first echo is at 50%
screen width and the second signal is at full screen width.
Place the probe on the perspex insert in the V1 block and you should get
a signal at full screen width.
SCHOOL OF APPLIED NON-DESTRUCTNE TESnNG
Slot a:
zero :o~nt
T-
See note 1
L
Adjust bc:n controls until the fourth echo is midway across the screen and
the last echo is at full screen width.
Put the peObeon the edge of the V1 block and you should see on echo
midway xross the screen and another signal at full scrsan width.
Note splrcious indications behind the backwall echo resulting from mode
change.
Adjust tke RANGE and DELAY controls until one signal is midway across
the screen at 50% screen width and the other is at full screen width.
Check yoJr setting by checking the signals on the screen against other
th:cknesszs on the step wedge.
SCHOOL OF APPLIED NON-DESTRUCTWE TESTlNG
0
7
'a,
.03
r.
CD
Ln
cy
, m
0
SC8OOL OF APPLIED NON-DESTRUCTIVE TESTING
If you are looking for pitting or corrosion on the inside of a pipe or vessel,
it is better to use a UT set than digital thickness meter. The extent of
pitting is more apparent on a screen.
SCHOOL OF APPLIED NUN-DESTRUCTIVE TESTING
Plate and pipe mills make frequent use of ultrasonic lamination checks,
and quality control usually demands a lamination check on plate coming
into a fabrication shop. It is normal to check parent metal for laminations
in the area where a weld will be made before welding to at least 100mm
from the edge preparation, and it is essential to check for lamination when
making a fillet, 'T' or node weld on areas likely to be affected by welding.
Small laminations can cause lamellar tearing, a type of crack and a
potentially dangerous defect.
Adjust the AMPLITUDE or dB control until the signal from the first backwall
echo (1st BWE) is at full screen height (FSH). Adjust the 2nd BWE to
FSH on weld parent metal.
Apply couplant, usually water or paste, and examine the plate on a grid
system as instructed or as detailed in BS5996. (When checking a weld
preparation examine to 100mm from the edge).
Size defects by the 6dB drop method and report their location, depth and
area. (The 6dB drop method is described in BS5996 and later in these
notes).
Move the probe over the plate; where the echo pattern changes drastically
a lamination or inclusion is likely.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
a Large crystals give out greater energy and so have a longer range
b Small transducers have a smaller near zone
c Large diameter transducers give poor contact on curved surfaces
Single crystal probes are suitable for use on materials over 15mm to
30mm thick, according to dead zone length and should be used in
preference to double crystal probes on anything over 30mm thick.
Double crystal probes are used for scanning thin materials and when
looking for defects close to the test surface. They can be shaped to fit
curved surfaces if necessary.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
5.1 PRINCIPLES
5.7 REJECTISUPPRESSION
5.9 PRESENTATION
SCHOOL OF APPLIED NON-DESTRUC77VE TESTING
Transmission pulse
Generatar
( Timer)
r
Time
Base
This is basically how a UT set works. It transmits energy into material via
a probe and measures the time in microseconds that the sound pulse
takes to return to the probe.
The controls on the UT set are almost entirely concerned with presenting a
display on the CRT screen for the operator to interpret.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESnNG
screen
The X and Y plates above, below and beside the electron stream carry
potentials that move the electron stream from side to side and up and
down, moving the green dot on the screen.
The X plates control horizontal movement and the Y plates control vertical
movement. By altering the potential of the X and Y plates, the dot can be
moved on the screen.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
The pulse generator sends the pulse to the TlME BASE GENERATOR on
the CRT and to the PULSE TRANSMITTER.
The timebase generator sends the green dot moving across the CRT
screen by putting a charge into the X plates in the tube.
At the end of each pulse the green dot on the CRT screen flies back to
the lefthand side of the screen to await the next pulse.
If the test material is thick the dot must ;ravel across the screen fairly
slowly, as the pulse repetition rate is lowered. Only one pulse must be in
the test material at any one time or confusing echoes will result. For this
reason the PRF is lowered when thick specimens are examined.
If the speed of the dot across the screen is not even, as a result of
equipment failure, we say the timebase is not linear. Flaw detectors
should be checked frequently to assess TlME BASE LINEARITY.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
5.5 Delay
The DELAY control makes the timebase generator wait for a while before
sending the green dot moving across the screen.
Twin crystal and angle probes have perspex blocks or wedges between
the crystal and the test material. This need not be shown on the CRT
screen, so we adjust the delay to move it sideways off the display, so the
passage of the ultrasound through the perspex in the probe does not
appear on the screen.
You can also use the DELAY control to wait until the sound has travelled
part of the way through the test piece itself before representation on the
screen. For example, if you only want to look at the bottom 25mm of a
20mm specimen, you can adjust the delay so that the green dot begins to
travel across the screen at 175mm. On thickness checks this can make
for greater accuracy in readings for thick specimens.
From the attenuator, the current goes to the Y plates in the CRT. When
the current hits the Y plates, they pull the electron stream upwards m d the
green dot jumps from the bottom of the screen to make a signal. The
height of the signal is increased or decreased by turning ttle gain up or
down.
and we can find the latter by comparing a signal from the reflector in the
test piece with a signal from an artificial reflector in a reference block.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
20 x 0.301 = 6.02 dB
So the answer is 6 decibels
Test this on the CRT screen. Obtain a signal from a backwall echo on a
test block and increase or decrease the gain until the signal touches the
top of the screen. Take out 6 dBs with the gain control and the signal
should drop to 50% full screen height. If it does not, the vertical linearity
of the UT set is out, the signal height is not changing in accordance with
energy from the probe.
5.8.1 B SCAN
5.8.2 C SCAN
5.8.4 P SCAN
5.8.5 T SCAN
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
5.9 Presentation
The flaw detector or UT set sends ultrasound energy into test materials
and some of this returns to the set to be presented as information on a
CRT screen.
Plan view
x y projection
(c-SCAN)
y z projection x z projection
(6-SCAN) (0-SCAN)
Ref lector
(defect)
The similarity is not complete, however, for the beam has two separate
parts or zones, which is not the case in a beam of light. These zones are
the NEAR ZONE and the FAR ZONE. The NEAR ZONE is always next to
the crystal and is sometimes called the FRESNEL ZONE. The FAR ZONE
is always beyond the near zone and is sometimes called the
FRAUNHOFER ZONE.
The DEAD ZONE is not part of the sound beam but as we have said, is
the region near the surface in the material where flaws cannot be
detected.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTlNG
After a time the shock waves, or pulses, even out to form a continuous
front.
The area between the crystal and the point where the wave front evens
out is what we call the NEAR or FRESNEL ZONE.
Inside the Near Zone signals from a reflector bear no accurate relation to
the size of the reflector, as the sound vibrations are going in all directions.
This affects the accuracy of flaw sizing of small reflectors inside the Near
Zone.
The Near Zone of a crystal varies with the material being tested, but it can
be worked out by a formula:
D2
- or
CRYSTAL DIAMUER~ -
-
41 4 x WAVELENGTH OF SOUND MATERIAL
= NEAR ZONE
to give 21 .I millimetres.
a The greater the probe diameter the greater the Near Zone.
i
far zone
'1
K = Intensity consrant
X = Wavelength
D = Crysral diameter
Half divergere angle = sine -62 -- K-
xX
D
BEAM DIVERGENCE
If we take the extreme edge of the sound beam to work out beam spread
angles, then K is 1.22.
So the angle of beam spread is 14.7" if you take the edge of the beam to
be where energy is 10% of the main energy or 16.6" taking the extreme
edge of the beam.
That is or,e of the reasons why low frequency probes have large diameter
crystals.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
7.3 MEASUREMENT OF
MATERIAL
ATTENUATION
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTlNG
7.0 TOTAL AlTENUATION LOSS
33683 defines attenuation as a diminuition in the level of acoustic energy
z it propagates through material,
s-i is made up of two components:
The loss due to the fact that the bean is a divergent cone and
therefore some energy is reflected away.
I
80% FSH
40% FSH
I Reflector
smaller I \ 80% FSH
20% FSH
0 10 20 30 40 50
Fig.7.1.b. Law of small reflector
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
7.1 Attenuation Due to Beam Spread
LARGE REFLECTORS outside 3 NEAR ZONES obey the INVERSE LAW.
SMALL REFLECTORS outside 3 NEAR ZONES obey the INVERSE
SQUARE LAW.
Echoes from reflectors in the Far Zone cause signals on the CRT screen
which vary in height with the amount of energy being reflected back.
These laws only hold good beyond a distance three times the length of the
Near Zone.
EXAMPLES:
A large reflector at 30mm and beyond 3NZ, gives a signal of 50% FSH.
What will be the signal size from a large reflector at 40mm?
Signal Size = -
302 x 50% = 28%
402
We can plot the edge of a sound beam for compression and angle probes
using side drilled holes in a reference block.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
ABSORPTION occurs as the sound pulse hits the molecules of the test
material and makes them vibrate. The energy lost in vibrating the
molecules turns to heat. The rate of absorption varies from one rrzlerial
to another and even from one type of steel to another. It is very hgh in
perspex, nylon and lead and is low in aluminium.
b Obtain an echo from twice the distance of the BWE used and
increase the signal height until it is at FSH. Note dB difference.
c Subtract 6 dB, due to beam spread loss, from the difference and
divide the remainder by the distance the sound has travelled
between the two echoes. This is twice the distance shown on the
CRT, as the sound has to travel to the backwall and back to trigger
the probe.
d The answer will give the dBs lost per millimetre in attenuation. This
can give an assessment of plate quality and of heat treatment.
signal
1
This method is called the 6dB drop method because the signal drops by
6dB when the probe is moved to the edge of a large reflector.
Note:
The peak of the reflector is normally taken as being the last peak on the
screen before the probe goes off the end of the reflector, not necessarily
the maximum signal from the reflector.
SCHOOL OF APPUED NON-DESTRUCTNE TESnNG
& =
Defect
Defect
I
ItLength of movement A
I
-
b) defect length length of movement minus beam width
(W) at testing range.
Note:
The peak of the defect is normally taken as being the last peak or; t7e
screen before the probe goes off the end of the defect, not necesszily the
maximum signal from a defect.
SCHOOL OF APPUED NON-DEST.RUCTNE TESnNG
-
. - . . ..
centre I I
100% FSH
Mark
centre
of probe - Move probe
sideways
10% FSH
1 Mark
1. centre
of probe
10% FSH
Move the probe until you get the highest signal you can from the hole,
then turn the signal to FSH using gain. See Fig.8.3a. Mark the position of
the middle of the probe on the side of the block. Move the probe to one
side until the signal drops to 10% FSH (-20dB) and mark the centre of the
probe on the side of the block. See Fig.8.3b.
Move the probe to the other side of the hole until the signal drops to
10%FSH (-20dB) and mark the centre of the probe on the block. See
Fig.8.3~.
Use the distances between the marks on the block to plot the beam on a
piece of graph paper. Measure 13mm depth on the paper then mark the
distances of the probe centre at -20dB from the beam centre at 100%
FSH on either side.
Now find the 25mm hole and maximise the signal, turning it to 100% FSH.
See Fig.8.3d.
Move the probe to either side of the hole marking the centre of lthe probe
on the side of the block where the signal drops by 20dB.
Measure 25mm on the paper and use h e distances on the block to plot
the beam dimensions at 25mm. See Fig.8.3e.
Repeat using the 32mm hole. Join up the points marking the probe centre
at -20dB to obtain a beam plot. See Fig.8.3f.
Note that we have only drawn the beam width in one plane, so the probe
must be marked accordinglfy and used to measure defects in this plane.
We use knowledge of the beam spread to size defects, to find the edges
and hence their width, length and sometimes orientation.
SCHOOL NON-DESTRUCTIVE TESllNG
d Move the probe along the edges of the plate and along scan lines.
20 dB drop method
6 dB drop method
Multiple echo method using a time base of 500mm and watching the
shape of the backwall echo panern on the screen. This method is
very good for examining thin plate.
The Distance Gain Size method using the DGS chart. A method
used in Germany and rather complicated.
Probe
Reflector
Sound obeys the law of reflection in that it reflects at the same angle as
the angle of incidence. Therefore a reflector angled at 45" to the scanning
surface will not reflect sound back to a normal beam probe. See Fig
9.0.a.
Transducer
i S
Test m a t e r ~ a l
C - Compressional w a v e
I - Incident a n g i e
R - Refracted a n g l e
S - Shear w a v e
If there is no couplant on the bottom surface of the perspex wedp, all the
energy is reflected back into the probe. If there is couplant and i' the
probe is placed on test material, sound energy passes into the test
material and generates a shear wave.
SCHOOL OF APPLIED NON-DESTRUCTWE 7ES77NG
Compressional
Polystyrene
Shear Compressional
..-
.,
I..
I '.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
The angle of refraction varies with the shear speed of sound in the test material and the
compression speed of sound in the perspex and with the angle the compression wave
makes with the interface. The formula describing this is covered by SNELL'S LAW,
which says that the ratio between sound speed in two materials is the same as the ratio
between the sines of the incident and refracted angles.
-- All angles are measured from a line vertical to the interface called the NORMAL
-
RIAL 1 -
SPEED OF SOUND IN MATERIAL 2 SINE ANGLE OF REFRACTION
v, x sine R
Sine I =
"2
For example, if we want a probe giving a 45" shear wave in steel we must calculated
the angle at which we must cut the perspex wedge - the incident angle. The
compression speed of sound in perspex is 2730 metres per second, the shear speed of
sound in steel is 3250 metres per second and the refracted angle we need is 45".
sine I =
sine R x Velocity 1
Velocity 2
sine I =
sine 45" x 2730
3250
sine I =
0.7071 x 2730
3250
sine l = 0.594
Look this up in your tables or on your calculator and you will find the angle is 36.44"
However, with incident angles less than 27", both compression and shear waves are
generated in the steel. This makes interpretation very confusing. To get a shear wave
on its own, the angle of incidence must be more than 27.4", called the FIRST CRITICAL
ANGLE. This gives a shear wave of 33" (the lowest standard angle probe
manufactured is 35".
If the incident angle is above 57.14". the shear wave is replaced by a surface wave.
This angle is called the SECOND CRITICAL ANGLE.
The largest probe angle you can get from manufacturers without a special order is 80'.
So shear waves on their own in steel are only possible with incident angles between
27.4" and 57.14", using a perspex wedge.
Of course, this is worked out by the probe.manufacturers, but it must be borne in mind
that a probe which gives a refracted angle of 45" in steel will give a different refracted
angle in other materials.
SCHOOL OF APPLIED NON-DESTRUCTWE TESTING
EXAMPLE
Calculate the refracted shear wave angle in steel for an incident
compression wave of 35" in Perspex (2.730 rnlsec) velocity
steel shear = 3.250 mlsec
Velocity, Angle 8 ,
m/sec X lo3 degrees
--- 80
1
--- 60
- 45
- 40
- 35 - Angle in Perspex
2
Perspex
compression
3
Steel shear
* Critical angle
Join velocity and angle known. Point (x) at which this line crosses
reference line gives new datum point.
Join known velocity and point (x) and project to find new angle.
Answer to example " 43".
The FIRST CRITICAL ANGLE is the incident angle at which the compression wave
in the test material is generated at 90". So using Snell's Law:
Use the speed of sound in the material. The speed of sound in perspex is 2730mlsec.
The speed of sound for water is 1480m/sec. The sine of 90 is 1. Work out the answer
with a calculator and sine tables.
The SECOND CRITICAL ANGLE is the incident angle at which a shear wave is
generated in the material at 90". Use Snell's Law again:
Example: What is the first critical angle when testing aluminium in water?
So we must check the probe before we can calibrate the tirnebase to enable its use.
There are a number of performance checks which should also be carried out at
specified intervals and these will be described in Section 16.0.
SCHOOL OF APPLIED NON-DESTRUCTNE TESnNG
To find the probe index place the probe on a V1 block and obtain an echo from the
100mm radius and establish at more than 50% FSH using the gain mntrol. Maximise
the echo by moving the probe backwards and forwards. Mark a line on each side of the
probe directly above the slots which indicate the centre of the 100mm radius. This is
the probe index, where the axis of the beam leaves the perspex shoe.
This procedure can be repeated for a 70" probe but reflecting the energy from the
plastic insert radius is unreliable. Therefore we suggest you use the 1.5mm hole as a
target instead.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTltdG
It is possible on some flaw detectors to calibrate the timebase to 100mm range, from
the V1 block. However, this involves delaying the signal by 100mm and not all
equipments can do this on the appropriate scale expansion setting, so we will confine
ourselves to calibrating for 200mm full screen width.
-
Place the probe on the V1 block and obtain boundary echo from the 100mm radius.
See Fig.lO.1. Establish this signal to more than 50% FSH using the gain control.
Further maximise the echo by moving the probe backwards and forwards. Hold the
probe stationary.
Place the signal from 100mm at 5 (half scale) on the timebase and the one from
200mm at 10 (full scale), using the delay and range controls. The timebase is
calibrated for 200mm. Longer ranges can be catered for in multiples of 1OOmm.
However, the V1 block is bulky, not convenient for site work and it is not always
possible to calibrate for 100mm, so we tend to use the V2 block.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
The timebase now represents 100mm. Check it by turning the probe round and
pointing it at the 50mm arc. If you have calibrated correctly, the signal when maximised
will come up exactly in the middle of the screen at 5.
FSH
I
Fig.11 .O.a. Centre position
'/,, FSH
4
1
Fig.11 .O.b. Forward position
I/,, FSH
Use a 20dB beam edge for 45" and 60" probes, but use a 10dB for 70" probes. With
the 70" probe a 20dB beam spread is so wide and difficult to consWct as to be almost
useless.
1
Find the hole which is 13mm below the top surface with the probe and maximise the
signal to FSH. Mark where the index point comes on the block with a pencil or crayon.
See Fig. 11.d.a. Move the probe forward until the signal drops to one tenth screen
height (20dB drop). Make a second mark on the block where the index point on the
probe stands on the block. See Fig.11.O.b.
Move the probe backwards until the signal maximises and then drops d0.m to 1/10
screen height. Mark where the index point now stands. See Fig.ll.0.c. Draw a
vertical line on the block from the hole to the upper edge. Measure the distances of the
three index point marks from the top of the line and note them doisn. See Fig.11.O.d.
Now find an echo from the 19mm deep hole and repeat the process, noting the
distances.
Take the slide out of a beam plotting chart and draw three faint lires across it at depths
of 13mm, 19mm and 25mm.
Transfer the distances of the index points from the vertical lines to the rekvant pencil
lines on the chart.
Joint the marks up. The centre line represents the main energy of the beam and the
other two marks represent the leading and the trailing edges ofthe beam.
With a 45" probe, use the 19mm, 25rnm and 32mm depth holes as the 13mm hole may
be in the probe's near zone.
Use a 10dB drop with a 70" probe, instead of dropping the signal :o 1/10 FSH for the
leading and trailing edges, use the 3/10 line on the screen.
SCHOOL OF APPUED NON-DESTRUC77VE TESTING
-\\
Stand off 1
I
I
Range \- \ .
00
Fig.11.l.c.
Use the corner of the block as reference point from which to measure stand-offs.
On the cover of the beam plotting chart, use the corner of the block to represent the
centre line.
Obtain a signal from the top hole of the six, maximise it, then push the probe towards
the block corner until the signal drops ?o 1/10 FSH (3110 for a 70" probe). (See fig.
11.la. Mark where the index point occurs on the block and measure the stand-off.
Note the range of the reflector on the screen.
Use the stand-off and the range to plot the defect along the trailing edge of your beam
spread. Mark it on the slide cover.
Now obtain a signal from the bottom hole of the six, maximise it and turn it up to FSH
on your screen. Pull the probe back until the signal drops to 1/10 FSH (3110 for a 70"
probe). (See fig.11 .l
b). Plot the bottom of the defect on your slide cover using the
leading edge of your beam plot.
Lay the transpatent slide over the 10W block and the top and bottom of the drilled holes
should coincide with the marks on the slide. If they do not your beam plot is off or you
are going wrong somewhere. See fig. 11.l c.
In effect you have just sized a defect by the 20dB drop method (10 dB drop for the 70"
probe).
The method described is the one expected from PCN candidates, as only IOW blocks
are provided. However, in SANDT we teach students how to construct a beam plot
using a special block, as illustrated in Fig.11.l .d.
This block is much more convenient, especially for novices, as all the holes are reached
from the same face. The probe only has to move forward and back and a fixed probe
guide stops the probe swivelling. Worst of all the IOW block is a large lump of metal to
carry about.
SCHOOL OF APPUED NON-DESTRUCTNE E S D N G
However, the beam starts to travel in the plastic wedge and is then refracted and
carries on in the material we are testing. We are only concerned with the part of the
beam near zone registering later than zero on the time base, e.g. in the test material.
.- This we call the modified near zone.
Here is an example:
A 5 mHz shear probe has a 10mm diameter crystal. The beam travels in perspex for
10mm. What is the modified near zone?
i.
We must now subtract the perspex wedge part of the beam which is 10mm, mult~plied
by the ratio of the perspex and steel velocities which is 10 x 2730 = 8.4mm
3250
Therefore modified near zone = 38.46 - 8.4 = 30mm.
L a-
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTlNG
P NW block
- --- Max.
Straightedge
\,-
' - - - 10 X FSH
\
Fig. 1l.3a
Probe positions
1 2 -
J
14131211109 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 91011121314
Method:
a. Place the probe to pick up the 1.5mm hole at skip and rnaximise the signal
from the intersection of the hole and the opposite face, see Fig.ll.3.a. Mark on
the straight edge adjacent to the near centre of the probe, to indicate the beam
centre.
b. Position a straight-edge either in front of or behind the probe to hoid the probe in
the fixed transverse position. Scan the probe laterally - sideways - until the hole
signal drops by 20dB. Mark on the straight edge adjacent to the rear centre
position of the probe. Tnis registers half a beam at the % skip range.
c. Scan the probe laterally the other way, through the maximum signal position, until
the hole signal again drops by 20dB. Mark on the straight edge as before.
d. You now have three mxks on the straight edge to indicate the bezrn width at that
range. Transfer these to the beam plotting chart as appropriate ard as shown in
Fig.ll.3.c.
e. Repeat steps a to d but at full skip and 1 l/2 skip for a 45" probe b ~ only
: at full
skip for a 60". Note tha: mode conversion reduces the llh skip sisnal on a 60"
probe to too low a level :o be reliable.
f. Join up the 3 points on either side of the centre-line to complete the beam. Only
take the lines back to the near zone because the edge is not relia!ie before that.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTiNG
There are several systems for setting sensitivity. We have already met one of them,
first backwall echo (BWE) to full screen height for lamination checks. However, when
checking plate adjacent to a weld, the second BWE should be to full screen height.
When setting sensitivity we must be sure that a signal from a defect will be visible on
the CRT screen and that we will be able to distinguish the defect signal from
background 'noise' or 'grass'. All UT sets differ slightly, so we cannot say, "Set the
sensitivity to x dB", as different probes and equipments will give entirely different signals
from the same reflector. It is to attain some uniformity that the different methods have
evolved.
Different methods are used in different places. In SANDT the IOW block is used and it
is the recommended method for PCN examinations. On North Sea contracts the
Distance Amplitude Correction Curve (DAC) method or "ASME curves' is used. The
DAC method is recommended in BS3923, and in Germany the Distance Gain Size
(DGS) system is usually applied, especially when evaluating small reflectors.
The purpose of sensitivity setting is to find a gain level sufficient to find a flaw and
depends on the:
b Flaw detector
Find a hole on the block that approximately coincides with the thickness of the material
you are testing. Double the thickness if you are examining at f ~ lskip,
l i.e. bouncing
your sound beam off the backwall.
Obtain a signal from the hole and turn the GAIN control until the signal is at FSH.
You have now set sensitivity and can be assured that flaws having the equivalent
reflectivity of 1.5mm side drilled holes will appear on the screen.
a It is simple to use
*a ,
M~ncmumof four O 3mm side-driiied through holes
, ,
Obtain extra graticules for the flaw detector. Clear radiographic film is a good
substitute, as you must construct curves for each of your probes and for each
calibration distance i.e. 100mm, 200mm or more.
Obtain a maximised signal from the nearest hole in the reference block which is outside
the probe's near zone and adjust the gain until the signal is at 80% FSH. Mark where
the top of the signal comes on the graticule or clear film.
Obtain a signal from the next hole down but do not alter the gain. Maximise the signal
and mark its top on the graticule.
Obtain another signal from the third hole down, maximise it and mark its top on the
graticule.
Obtain signals from all the relevant holes in the block, reflecting the sound energy from
the backwall if necessary but without altering the gain.
Joint up the marks on the graticule to form a curve and write the probe number and dB
setting on the graticule.
Construct a second curve on the graticule which is 6dB less than the previous gain
setting if required. US specifications like ASME V require this to be done.
Work out a transfer correction, see section 12.5, when you conduct an examination.
Increase or reduce the gain setting used during curve construction as appropriate.
Compare the signals from discontinuities to the curves on your screen. Any signal
above the curve shows a reflector larger than the reference hole.
Accept or reject discontinuities as instructed in the specification you are working to.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Advantages:
a A quick way of accepting or rejecting discontinuities without too much time
consumed in sizing reflectors.
b Some idea can be gained of the discontinuities size in relation to reference holes.
c Uniformity, provided by all technicians constructing their curves from the same
test block.
Disadvantages:
a Curves must be constructed for each probe in conjunction with each set.
DIMENSION A
in inches
-
1
16
i
i
?
3
t1
7
5
1
It
1i
f
3:
DIMENSION B 3:
in inches 4:
1 4:
M
1 5:
M
L 5!
M
DIMENSION A
in m m
3 6
10 20
30 40
50 70
100 150
DIMENSION B
in mm
When setting sensitivities the specification or technique will specify the block to be used
and the amplitude of signal to be obtained from the FBH.
Blocks are cut for use with 0' probes or angle probes in different materials.
Advantages:
a Easy to use
Disadvantages:
a Fairly rigid system for specific applications
d Blocks for angle probes are rarely cut exactly normal to the beam.
You may hear these blocks called H l l T or ALCOA blocks, after the originators.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Advmtages:
a Quick and easy
Disadvantages:
a No accurate sizing of the defect
b Discontinuities near the surface of the test material may be hidden in the 'grass'
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
12.5 TRANSFER
CORRECTION
Transfer Correction
Reference blocks usually have smooth machined surfaces while test objects
frequently have rougher and more uneven surfaces. Also the attenuation of
sound in the reference block might be different to that in the test material.
Usually attenuation in the reference block is less than the test material but not
always. This means that allowance must be made for the differences in sound
energy transfer between probe and test material and probe and reference block.
More energy can be passed into a reference block than into a rougher surfaced
component. Hence the artificial defects in reference block may give higher
amplitude signals (anything up to 6dB or even more) than signals from similarly
sized discontinuities in the test materials.
Allowances have to be made for this and corrections made for different surfaces.
This allowance is named transfer correction, or sometimes transfer loss.
There are several methods of determining transfer correction, some requiring the
constuction of separate DAC curves and some requiring calculation according to
formulae.
Curve f o r r n a t e r ~ c ~
I
I
R
Beam path range (mm)
NOTE, Correction for attenuation and transfer loss at range R equals (G2- G 1) decibels
Place the two probes opposite each other on the reference block with one probe
transmitting and the other receiving, so that the sound energy is bounced off the
backwall and caught by the receiving probe (Pitch and Catch). See Figs.12.5.2.a and
12.5.2.b.
Place the two probes on a piece of test material of the same thickness as the reference
block and repeat the process.
Note the difference between the two gain settings. This is the transfer correction
needed.
Reflector sizes are expressed in terms of the probe diameter and distances from the
probe are expressed as multiples of the near zone.
Now if you have a signal from a flaw at a certain depth, you can compare the signal
size to what the signal of a backwall echo should be at that depth and estimate the size
of flat bottomed hole that would give such a signal at that depth. The defect can then
be sized according to a flat bottomed hole equivalent.
The attenuation factor for the test material must be taken into consideration when using
the DGS system.
EXAMPLE: You are using a 5MHz 10mm diameter compression probe on 100mm thick
steel plate and you find a defect at 60mm depth which gives a signal at FSH with a
30dB gain setting. What is its flat bottomed hole equivalent? First, work out the probe
near zone. It is 21 mm, so the defect is at a distance of 3 near zones.
Now get a backwall echo and find what the dB reading is. Say it is 20dB when the
BWE is at FSH. lOOmm is 5 Near Zones. What will it be at 60mm, 3 near zones?
Refer to the DGS curves. If the BWE is a FSH with 20dB at 100mm, by the law of
large reflectors and according to the BWE line on the DGS curves a BWE at 60mm
should reach FSH at 16dB, 4dB less than at 100mm.
The signal height from the flaw is 30dB, 14dB more than the BWE. Look down the
scale 14dB at 3 near zones from the BWE and you find that the nearest line is 0.5 of
the probe diameter. The probe diameter is 10mm so the nearest equivalent flat
bottomed hole to the flaw had a diameter of 5mm.
By similar working, a sensitivity setting can be worked out for a flat bottomed hole of a
certain diameter at a given range to a given screen height and the flaw detector gain
set accordingly.
b It tells you the smallest defect you can find at a given range
f Flaw surfaces and shapes are not ideal reflectors, therefore signal amplitudes are
not the same as a comparative FBH.
For angle probes, plastic slides have been manufactured by Krautkramer to fit over the
CRT screen. The set is calibrated and gain setting is made by bringing the back wall
echo or the echo from the 1.5mm hole on the V2 block up to marks on the slide. Flat
bottomed hole equivalents for flaws can then be read straight off the slide.
It is quicker and easier, however, to use a flaw location slide and a beam plot or even a
piece of clear plastic film with the probe angle drawn on it. Use the slide as follows:
Draw a cross section of the weld on the transparent outer envelope of the slide. Draw
a mirror image of the weld cross section immediately under it if the sound energy is
going to bounce off the backwall, i.e. using full skip. Use the printed datum line on the
plastic envelope as the centre of the weld and measure all standoffs from it.
Maximise the echo from a defect and mark where the index point falls on the parent
metal. Measure its distance from the centre of the weld.
Move the datum line on the plastic envelope to the standoff distance. Look along the
centre of the beam plot until you come to the range shown on the screen. Make a
mark on the envelope this represents the centre of the defect it shows the defect's
position in the weld body.
Push the probe forward until the signal drops by 20dB (10dB for a 70" probe). Note the
standoff and range.
Plot the edge of the defect on the chart using the trailing edge of the beam plot.
Now pull the probe back until the signal from the defect maximises and falls off again by
20dB (10dB for a 70" probe).
Note the standoff and range, then plot the other edge of the defect along the leading
edge of the beam plot. The measured distance between the two plotted points is the
width of the flaw.
To find the length of the defect, use the 6dB drop as follows:
Move the probe sideways along the defect until the signal peaks for the last time.
Sometimes we may move right off the reflector and then back on, to verify this last
significant echo. Maximise the signal from the last peak, then move the probe sideways
until the signal drops 6dB to 50% FSH. Mark where the centre of the probe coincides
with the parent metal. This shows one end of the defect.
Move the probe to the other end of the defect and maximise the signal from the last
peak. Adjust the gain until the peak signal is at FSH. Move the probe sideways until
the signal drops to 50% FSH. This marks the other end of the defect and all you have
to do is measure the distance between the two marks to find the length of the defect.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
To zit the identification of defects there are four basic probe movements:
b orbital
Ho%eU;zr,at the other extreme, porosity can be considered as a large number of small
spheVsealreflectors which cause the energy to reflect everywhere. Rather like the light
reflecog from a disco ball hanging from the ceiling. Porosity is a DIFFUSE reflector.
By combining these movements and watching the movement of the signal on the screen
you ~ 2 draw
7 conclusions. Characteristics of different defects are shown in the
accomanying diagrams, with explanations adjacent.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Crack I
Multirange Signal
Echo falls rapidly when traversed with 70" probe, also range increases.
Probe Movement
Echo falls rapidly when angle probe traversed forward.
Measurement
Not possible to measure depth with angle probe. Length difficult but usually by 6dB.
Probe Movement
Traverse backwards echo falls more rapidly than lack of penetration.
Measurement
Use centre of beam and 20dB drop (trailing edge) for height. Not always possible to
measure height.
Probe Movement
Orbit, echo held over large angle.
Lateral, echo held with multirange signals and variations on time base.
Measurement
6dB for length.
Traverse forward with 20dB for height.
SCHOOL OF APPLIED NON-DESTRUCTNE TESnNG
Fig.14.1.5. Misalignment
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Probe ?.fovement
Confirr with 70" probe, when traversed.
Large rovement for 20dB drop.
Orbit, echo falls rapidly.
Measurement
Lateral use 6dB or 20dB drop.
Traverse use 20dB or 10dB for 70" probe.
14.1.5 Misalig~ment
Large shgle echo from one side. No echo from opposite side.
Probe r.!ovement
Traversz back echo falls rapidly.
Measurement
Lateral for 6dB drop.
SCHOOL OF APPLlED NON-DESTRUCTIVE TESTING
14.2.2 CRACK
14.2.4 POROSITY
Lack of Fusion
Echo large, single, narrow at time base when side wall. Poor echo from opposite side.
Confirm by skip scan.
Probe Movement
Rotate or orbit, echo falls rapidly. bteral or traverse echo height held.
Measurement
For depth use 20dB. For length use 6dB or 20dB.
Crack
Multiple peak reflector usually high amplitude, but dependent on type of crack and size,
echo with 'fir tree' appearance.
Probe Movement
Orbit - echo held over large angle compared with fusion defects.
Lateral - signal held with varying height.
Measurement
For length use 6dB or 20db
For depth use 20dB.
SCHOOL OF APPLIED NON-DESTRUCTNE TESTlNG
Probe Yovement
Rotate. lateral and traverse echo falls rapidly. Orbit echo height remains.
me as^-ement
Impracxal to measure height and length. Report as isolated reflector. Equate
reflect..ity against disc area or DGS.
14.2.4 Pores?,'
Multipk peak echo. Low intensity 20% CRT broad at time base due to numerous
ranges.
Probe ?,lovement
Orbit - echoes held with amplitude variations.
Meask-ement
Indicax area by pin-pointing last maximum signal from traverse and lateral scans.
SCHOOL OF APPLIED NON-DESTRUCTNE TESTING
Probe Movement
Orbit, echo held with various maxima and minima. Similar with traverse.
Rotational, echo will drop quickly.
Lateral, will produce large variations in height.
Perhaps with total loss of signal for short distances less than beam width.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
15.1 BS3923
15.0 B U l 7 WELD
r-'
EXAM l NATION
15.8 PIPES
15.9 REPORTING
SCHOOL OF APPLIED NON-DESTRUCTNE TES77NG
E x a m i n a t i o n Level 1 2 3
Normal beam
scans I DAC+14dB DAC+8dB I DAC+8dB
Scans for
transverse DAC + 20 dB DAC + 14 dB DAC + 8 dB'
imperfections
I I I
* W h e n requested b y the purchaser (see 3.3(e)).
NOTE 1. Where excesswe grass IS experienced at t h e m l n l m u m s e n s m v t t y settlngs q u o t e d , a l o w e r
frequency p r o b e IS recommended. Where excessive glass 1s st111experienced w ~ t a h lowet flequency
probe. ~t ts ~ e c o m m e n d e dt h a t testlng should o n l y c o n t i n u e b y agreement b e t w e e n t h e
c o n t r a c t t n g partles.
NOTE 2. + means m o r e sensltlve.
15.0 BS3923
British Standard 3923 describes the procedure for examining welds and details beam
paths to be used for welds of different configurations e.g. plates, pipes nozzles and
nodes. To standardize methods of examination BS3923 recommends the use of
distance amplitude curves and different sensitivity settings to match the criticality of
different examinations. Sensitivity is set higher for examining, say, a high pressure
steam pipe in a chemical works than it is for a comparatively low pressure line in a
refinery. A chart detailing sensitivity for the different examination levels is given in
Fig.15.1.
Different procedures are followed on different projects. Usually the test procedure is
formulated before the job starts so the technicians know exactly what is expected of
them.
a A COMPRESSION SCAN to check the parent metal on either side of the weld for
laminations and to check through thickness dimensions.
b A ROOT SCAN to check the root for longitudinal defects such as lack of
penetration, lack of root fusion, cracking or mismatch.
c A WELD BODY SCAN with shear probes to check the sides of the weld and the
weld body for longitudinal defects like lack of fusion, cracks, slag and porosity.
d A TRANSVERSE WELD SCAN to check the weld for transverse and chevron
cracking.
e SIZING defects
f REPORTING
The examination level: Whether the examination is critical, in house, for a client and
other relevant details to enable him to carry out his examination in accordance with the
client's instructions.
The material: Hard steel, forged steel, cast steel, alloy or other metal.
The manufacturing stage at which the examination is to be carried out. Before or after
heat treatment, 48 hours after welding, before or after machining and other
circumstances.
The joint preparation and dimensions: The included angle in the weld is essential
knowledge, the dimensions of the root gap and face through thickness, and width if
applicable.
The welding procedure: This will enable the operator to expect defects common to the
prescribed welding procedure e.g. lack of sidewall fusion in automatic MIG welding, slag
in manual metal arc welds.
The surface condition: Should be better than 6.3 micrometers roughness average or as
agreed by client and customer.
The couplant type: Grease is sometimes forbidden as it can affect paint bonding.
The details of scans and probes to be used: These are detailed in BS3923.
The sensitivity settings: n e s e vary considerably, methods can use the IOW block,
DAC techniques, the DGS system or special reference blocks.
The datum points: Dimensions and positions of defects must ba measured from these.
The reporting requirements: Often a set report form is used on projects and the
operator is expected to fill it out to a set formula.
The acceptance standard: Large companies have their own standards, some common
standards can be applied to set jobs, eg ASME IX. Some techniques are supplied by
manufacturers, particularly in the aerospace industry. Basically the operator must know
what discontinuity or flaw is acceptable and what is to be rejected.
The operator qualification: An operator working on plate should be qualified to PCN 3.1
level or equivalent. Someone examining offshore structures should hold a PCN 3.9
qualification or equivalent. Other qualifications may be required.
SCHOOL OF APPL/ED NON-DESTRUCTIVE TESTING
All these questions should be considered before the job starts and the operators should
be made aware of them. On well run projects a UT procedure is agreed between the
client, customer and inspection company before work starts.
It is essential that the component examined by the operator is correctly identified. The
operator should be shown the component he is to examine and the exact weld
identification written down for him. This avoids misunderstanding, ambiguity and time
wasting. If the weld has been radiographed, checked by MPI or PFD, have a look at
the radiographs or the test result.
SCHOOL OF APPUED NON-DESTRUCTNE TESnNG
- .
Fig.15.3.a. Locating the centreline of the weld
Full skip
Probe
movement
v
A A A A A A
Probe
movement
Full skip
I
r' i
Find the centre of the weld. It is from the centre of the weld that you must measure
stand offs. If the root is detectable, find the root from both sides with a 60" probe.
Mark where the index point falls on both sides of the weld when the root signal is
maximised and the weld centre is midway between the two marks. See Fig.15.3.a.
Note the thickness of the parent metal on either side of the weld. Be sure there is no
counterboring.
Examine the parent metal on either side of the weld with a 0" compression probe with
sensitivity set a 2nd BWE to FSH or the sensitivity given in the procedure. Look for
changes in thickness and for lamination. You should cover the parent metal either side
of the weld to full skip distance for the highest angle probe, that is for a 60" or a 70"
probe. See Fig.15.3.b.
Draw a cross-section of the weld on the slide of your beam plotting chart one you are
sure of the weld dimensions.
If the cap of the weld has been dressed ground flush with the parent metal, examine the
body of the weld with the 0" probe. If a backing bar has been used, check bonding at
the root if possible.
Check the root on a double sided weld if the cap has been ground flush. Lack of
penetration at the root will sometimes show with a 0" probe.
If.you are examining a single V weld with a dressed cap, a 0" probe scan of the weld
can reveal lack of inter run fusion and large pockets of slag.
SCHOOL OF APPUED NON-DESTRUWE TESTlNG
Index
4 Stand off
Stand off
/
Straight edge
Measure out the half skip distance on the parent metal either side of the weld
measuring from the centre line of the weld. Draw lines either side of the weld at this
distance. Note the range of the root centre along the main beam. See Fig.15.4.a.
Set the GAIN from an IOW block so a hole at the depth of the root will give an echo to
FSH. (Or set the GAIN according to the set procedure).
Place the probe index point on the line and put a magnetic ruler or strip behind it as a
guide. See Fig.15.4.b.
Move the probe laterally along the weld. Signals from a good root should be small
while defects will give a large signal. Signals from a fully penetrated root will usually
appear just beyond the range of the root centre.
Mark any defects on the parent metal or on the magnetic guide behind your probe.
Find the height of the defect using a 20dB drop on the trailing edge of your beam.
Record the defect on a sketch and note standoff distance and range.
SCHOOL OF APPUED NON-DESTRUCTNE TESTING
Mark \
'I,skip
/ Mark full skip (
t 'I,weld width
4 Stand off 4
-
I 1 ' 1
1
Range
L
Fig.15.5.c. Locating flaw position on beam plotter
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
If a 60' probe is first used follow up with a 70" probe if the plate thickness is less than
25mm and a 45" probe if it is thicker than 20mm. There is a bit of an overlap between.
Use your beam plot and flaw location slide to work out the half skip distance to the root
centre and full skip distance plus half the weld width to the top of the weld. Mark these
stand off distances on the parent metal on either side of the weld. Note the range of
the root can cap. See Fig.15.5.a.
Set GAIN so a hole at maximum range will be at FSH using an IOW block or use the
sensitivity setting given in the procedure.
Move the 60" probe backwards and forwards along the weld on either side so that the
beam covers the sidewalls and centre body of the weld. Defects should maximise
between the root and cap signals. If any signal occurs near the cap, use a couplant
covered finger to damp any signal that may be coming from the cap, thus identifying its
position.
Use the beam plot and flaw location slide to locate any defect on the walls or in the
weld body maximising the signal and plotting along the main beam. See Figs.15.5.b.
and c.
Size the through thickness of the defect using the 20dB drop method and the leading
and trdling edges of the beam. Find the length of the defect using the 6dB or 20dB
drop method.
Note t5e defect on a sketch and note the stand off and range of the centre of the
defect.
Make a second scan of the weld body with a 45" probe, for materials more than 20mm
thick, or a 70" probe if less than 25mm thick.
Check the area just under the cap with a 45" probe if you are examining thicker
material. You may find porosity that the 60" probe has not revealed. Check any defect
already noted and see if it is longer than shown with the 60" probe.
On thinner materials, a 70" probe will confirm defects already noted and may help you
distingzish between signals from a defect and signals from, say, the weld cap or mode
change.
SCHOOL OF APPLIED NON-DESTRUC77VE TESDNG
If the weld has been dressed, push the probe along the centre line of the weld in both
directions, then pcsh it along both edges of the weld in both directions. See Fig.15.6.b.
SCHOOL OF APPLIED NON-DESTRUCTNE TESnNG
'/, width
Examine the root with a 70" probe straight into the root at 114 skip. Run the probe
laterally along the weld with sensitivity set from a hole in the IOW block at a suitable
range; this should show any lack of penetration at the root. See Fig.15.7.a.
Make body scans with 60" and 45" probes from both sides of the weld if the plate is
thick (over 25mm) using % skip to lh skip stand off. See Fig.15.7.b.
On thinner plate, use a 60" and 70" probe from one side of the weld only between 114
skip and full skip and 1/2 weld thickness stand off positions. Take great care in locating
the defects and reporting, as you will have to use a mirror image on your plotting chart
and this can lead to confusion. You may find a defect and place it on the wrong side of
the weld. See Fig.15.7.c.
SCHOOL OF APPLIED NON-DESTRUCTNE 7ES77NG
Sin 0 --
ID
OD
0-
ID
Angle Sin-' -
OD
15.8 Pipes
Circumferential pipe welds can be examined as butt welds in plates are examined, that
is:
a A compression scan either side of the weld to full skip plus half weld width
distance.
b A root scan at a fixed beam path distance using a flexible strip at the back of the
probe to ensure that the centre of intensity of the beam goes into the centre of
the root.
c Scans of the weld walls and body with angle probes. Here the flexible strip can
be used to mark the extent of tracking distance so you do not have to take your
eyes off the screen and you will not be so confused by echoes from the cap,
particularly on thin walled pipes.
d Transverse scan. Here difficulty may arise, as you must be sure that the beam
reaches the bore of the pipe.
The flexible guide is convenient for marking defects. All defects must be measured
from a datum point. The "button" at the top of the weld is a convenient datum point.
So a 12" pipe with a 1" wall will need what probe angle?
A 60' probe will not reach the bore of this pipe, so a 45" probe is advised.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
15.9 Reporting
Every UT report must contain the following information:
a Job identification
b Material
f Date of test
m Sketch of scanning pattern used and indicating any areas that were inaccessible
n Report on parent metal describing any laminations, surface flaws or areas of high
attenuation
p Sketch of location of flaws in the weld cross section with depth and through
thickness
r Opeiator's signature
Every UT report should include a statement of whether or not flaws are acceptable to
specifications applicable to the component or should accompany a similar statement
from a responsible inspector or supervisor.
Report forms are often provided for an operator to fill out. Any report form should be so
designed as to provide space for the information above.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
16.4 SQUINT
This s-adard details methods of monitoring probes in Part 3 and BS2704 gives
dimensions and manufacturing standards for calibration blocks.
These standards only apply to pulse-echo techniques and not to other ultrasonic
methocs.
In the aerospace industry different standards may apply and different calibration blocks
may be used.
C a l i b r ~ o nof time base, checking index points, angles and beam plotting have already
been czscribed, so'we will show you how to carry out the remaining checks and
resolvicg power, as promised in section 3.4.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Adjust RANGE and DELAY until the first and last echoes coincide with scale marks.
Bring successive backwall echoes up to 80% FSH and check that the leading edge of
each signal lines up with the appropriate graticule line. Deviation should not be more
than 2%.
Vl block
Angle of
squint
I I -1
Ruler
/
This angle should
be 90" if the
probe is aligned
properly
Increase signal height on the attenuatorlgain control until the grass at the same range
reaches 20% FSH and note the difference between the two dB readings.
The first reading gives an idea of the sensitivity of the probe and flaw detector.
The difference between the two readings gives the signal to noise ratio.
16.4 Squint
Direct the probe at the radius of a V1 block and maximise the signal. If the edge of the
probe is not parallel to the edge of the block, the sound beam is not properly aligned
and the probe has "squint".
Place the probe to be checked on the side of a V1 block and obtain a signal from the
edge of the block at half skip.
Lay a ruler along the side of the probe and draw a line along the ruler to meet the edge
of the block. Measure the angle it makes with a protractor. The line should meet the
edge of the block at 90". If it does not, the beam is not aligned, the probe has "squint".
See Fig.16.4
This check should be done weekly or when the probe is considered suspect due to
wear or damage.
SCHOOL OF APPUED NON-DESTRUCTWE TESnNG
STEEL mm
Freq Compressional Shear
STEEL mm
Freq Compressional Shear
Adjust the GAIN until the signal from the hole is 100% FSH.
This check should be done daily in conjunction with the probe index check.
Also the screen trace must show the 2 signals distinctly split to more than half their
height i.e.> 6dB. See Fig.16.6.a.
To measure the resolving power of the probe matched to a particular set, first work out
which are the appropriate steps on the A.7 resolution block. See Fig.16.6.b. Place the
probe in the centre of radii but on the flat face of the block. Obtain echoes from the
steps of interest and see if the signal on the screen separates into two peaks at half
height (6dB). If they do not, find what steps do resolve.
The check described is good for both compressional and shear probes, but if a critical
check is to be done we suggest that the A.6 block is used. See Fig.16.6.c.
SCHOOL OF APPLIED NON-DESTRUC77VE TESTING
Probe
-
I
Fig.16.6.b. Block for resolving power (A.7.)
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
L
17.0 COMPLEX WELDED JOINTS
17.3 NOZZLES
1.7;4 NODES
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
To use a profile gauge, push the gauge onto the profile you wish to reproduce and
make sure all the wires touch the profile surface.
Lift the gauge from the profile and trace the profile onto a piece of paper. Check the
thickness of the parent metal and make an exact cross sectional drawing of the weld
you wish to examine. On a variable geometry weld several drawings are necessary.
A flexible strip of lead or a plastic strip with internal wires is an effective alternative to a
profile gauge.
Use a SANDT Beam Path Location Rule to plan scans. See fig.17.1. If you have not
got one, get a piece of clear plastic film and trace probe beam angles onto it. Trace the
beam paths from the illustration if you wish.
Accurate cross sectional drawings are essential when examining complex welds.
Compression scan of parent material,
for laminations, and areas of
high attenuation.
Other defects in plate 81weld.
a. Lamellar tear.
b. Weld defects, slag, gas pores.
c. Toe cracks.
d. Land - extension cracks.
e. Lack of fusion.
2. Shear scan
To confirm some o f the above
defects, and find planar defects
not orientated for the compression
probe scan.
T butt ,welds are either full penetration welds or partially penetrated welds with a "land of
no we:d' or unfused portion. If the weld is partially penetrated, you must make sure the
land of no weld is no larger than design permits and that cracks are not found at the
edge of the land, as is sometimes the case.
Make 2 careful cross sectional drawing of the weld before attempting to check it. This
will enale you to plan your examination of the weld with different angle probes and to
know here signals from caps, lands and backwalls should appear on the screen so
you czn distinguish these from discontinuities. Beam paths drawn on a piece of clear
film are useful.
Scan b e parent material for laminations with a compression probe with a sensitivity set
at 2nd BWE to FSH. If there is a land of no weld, plot this using the 6dB drop method
and m u k it on the plate.
Scan b e weld area from the flange side with a compression probe set from a 1.5mm
hole in the IOW block at plate thickness to FSH or the sensitivity gain in the procedure ,
this should show any lamellar tearing in the plate and any fusion defects where the weld
meets f i e plate. Scan from the flange as shown with a 60" or 45" probe to find toe
cracks, land extension cracks and porosity in the weld.
Scan h r n both sides of the web to find fusion defects in the web to weld metal fusion
face w2-1appropriate angle probes.
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
Set-in branches
Nozzles
Nozzle welds are simi:ar to T butt welds, but instead of joining two plates, a "branch" is
joined to a "shell", one pipe is joined to another. Again, like T butt welds, nozzles can
be fully penetrated or prtially penetrated.
There are three basic types of nozzle weld, "set in", "set on" and "set throughu. The
technique used for examining these welds ultrasonically follows roughly the same
procedure as the examination of a T butt weld. A lamination check, a compression
check from the inside of the shell, an angle probe check from inside the shell and finally
an angle check from both sides of the branch. On many occasions one or two of these
surfaces may be inaccessible, in which case your examination will not have complete
coverage. If this is so you must state it in your final report.
A problem with many nozzles is their changing geometry. The cross section of the weld
changes its shape as you move round. Make careful cross sectional drawings on the
weld at different points of its circumference using a profile gauge, a "mimic gauge", or a
lead strip which you can bend to reproduce the contours of the weld and then transfer it
to paper. You can now use a beam path on a clear film to plan scans and locate flaws.
Scan the parent material for laminations on the shell and the branch with 1st BWE to
FSH or sensitivity given in procedure and map out the land of no weld, if any.
Change the sensitivity using the IOW block with the 1.5mm hole at shell thickness to
FSH or sensitivity given in procedure then scan the fusion faces from the shell side,
looking for lamellar tearing, fusion defects and defects in the weld body. Scan the weld
from the shell side with angle probes looking for toe cracking, land extension cracks and
defects in the weld body.
Scan from both sides of the branch with angle probes to find fusion defects on the
branch weld fusion face and defects in the weld body.
Measure stand off posdons from the weld toe on the shell or the branch and use a
beam path to locate defects. See figs.17.3.a-c.
SCHOOL OF APPUED NON-DESTRUC77VE TESTING
SCHOOL OF APPLIED NON-DESTRUCTIVE TESTING
17.4 Nodes
Nodes are encountered in structural tubework, for example jacket construction for oil
platforms. Instead of a shell and a branch, we have a "cord", or "main leg" and a
"brace". Access with nodes is usually restricted to full coverage is rarely possible.
Make careful cross sectional drawings of the node at least six point around the weld
using a profile gauge or a flexible strip.
Plan scans using your drawings and a beam chart on a piece of clear film.
Scan the weld from inside the cord if possible and mark the extent of the weld using the
6dB drop method. Use a compression probe with 1st BWE to FSH. Look for
laminations in the parent material. Make a compression check of the cord to weld
fusion face with the 1.5mm hole of the IOW block at FSH at cord thickness. Look for
lack of fusion, lamellar tearing and defects in the weld body.
Scan the root of the weld from the brace side with a 60" or 70" probe at half skip. A
good root will not give an echo. Look for lack of penetration. Scan the volume of the
weld from inside the cord if possible with 45" and 60" probes at half skip or from
outside the cord with 45" and 60" probes at full skip. Look for toe cracking and defects
in the weld body.
Scan the weld body and brace to weld fusion faces with 60" and 70" probes from the
brace. Some areas may be inaccessible, in which case state this in your report.
Plot flaws on your drawings using your beam plot, sizing them as accurately as
possible.