Ultrasonic Back-to-Basic

80
BACK TO BASICS - ULTRASONICS 1 BASIC PRINCIPLES OF SOUND Sound waves are vibrations of the particles of solid liquid or gas through which the sound is passing. Each particle oscillates about a mean position and in doing so causes a similar vibration to be taken up by its neighbour. The resulting disturbance radiates out from the source as a sound wave. Sound waves are therefore a form of mechanical energy that can only exist in a solid liquid or gas and not in a vacuum. Essentially, there are two requirements for sustaining a vibration: there must be something to vibrate and some force that will always try to return that ‘something’ to its original position. In other words, there must be MASS and ELASTICITY. This is illustrated in figure 1.1a below. A weight is suspended from a beam by a spring. The weight (W) provides the MASS and the spring provides the ELASTICITY. At rest, the force of gravity (G) acting on the weight is balanced by the tension (T) in the spring. Fig. 1.1a G T W A C B Fig. 1.1b Down Time Up Amplitude One cycle If the weight is pulled downwards from its rest position (A) to position B, the tension in the spring will increase. When the weight is released, the weight will accelerate back towards position A reaching its maximum velocity at position A when the forces T and G are again equal. The momentum of the weight travelling at speed will cause the weight to overshoot position A. Immediately the tension in the spring is less than the force of gravity and the weight will begin to decelerate until it comes to rest at position C. Because force G is now larger than T, the weight will start to descend again; overshooting position A again until the increasing tension in the spring eventually stops the downward movement. At this time, the whole cycle of events starts again and continues until friction and air resistance losses gradually bring the oscillations to a stop.

description

Ultrasonic by J. C. Druryfrom Silverwing

Transcript of Ultrasonic Back-to-Basic

Page 1: Ultrasonic Back-to-Basic

BACK TO BASICS - ULTRASONICS

1 BASIC PRINCIPLES OF SOUND

Sound waves are vibrations of the particles of solid liquid or gas through which the sound is passing.

Each particle oscillates about a mean position and in doing so causes a similar vibration to be taken

up by its neighbour. The resulting disturbance radiates out from the source as a sound wave.

Sound waves are therefore a form of mechanical energy that can only exist in a solid liquid or gas

and not in a vacuum. Essentially, there are two requirements for sustaining a vibration: there must be

something to vibrate and some force that will always try to return that ‘something’ to its original

position. In other words, there must be MASS and ELASTICITY. This is illustrated in figure 1.1a

below. A weight is suspended from a beam by a spring. The weight (W) provides the MASS and the

spring provides the ELASTICITY. At rest, the force of gravity (G) acting on the weight is balanced by

the tension (T) in the spring.

Fig. 1.1a

G

T

W

A

C

B

Fig. 1.1b

Down

Time

Up

Amplitude

One cycle

If the weight is pulled downwards from its rest position (A) to position B, the tension in the spring will

increase. When the weight is released, the weight will accelerate back towards position A reaching its

maximum velocity at position A when the forces T and G are again equal. The momentum of the

weight travelling at speed will cause the weight to overshoot position A. Immediately the tension in

the spring is less than the force of gravity and the weight will begin to decelerate until it comes to rest

at position C. Because force G is now larger than T, the weight will start to descend again;

overshooting position A again until the increasing tension in the spring eventually stops the downward

movement. At this time, the whole cycle of events starts again and continues until friction and air

resistance losses gradually bring the oscillations to a stop.

Page 2: Ultrasonic Back-to-Basic

Figure 1.1b is a graph of the displacement of the weight, during this up and down motion, against

time. In the diagram, two points on the graph are shown where the weight is doing the same thing,

travelling upwards and passing through position ‘A’ on consecutive passes. The distance (time)

between these two points represent one complete cycle of the oscillation. The number of cycles of

oscillation completed in a given period of time (usually one second) is called the ‘Frequency’ of the

oscillation. The maximum displacement of the weight from its normal rest position is called the

‘Amplitude’ of the oscillation.

One of the best examples of an oscillating source of sound that can be used later in describing the

action of an ultrasonic test probe is the guitar. The strings of a guitar are elastic and pre-tensioned to

produce a particular frequency of vibration. Each string is distorted by the guitarist to stretch the string

and then released. As soon as it is released, the string begins to oscillate about its mean position at

the resonant frequency of that string. Shortening the string using a finger to hold the string against

one of the frets can change the frequency. The human ear recognises the frequency as the ‘Pitch’ of

the note produced. The ‘Loudness’ of the note depends on how far the guitarist distorted the string

before t was released, in other words, the ‘Amplitude’ of that distortion.

The mass of woodwork to which the string is attached amplifies the sound and adds its own harmonic

frequencies to produce a range of notes to give the characteristic richness of tone to the instrument.

The band of frequencies produced is called the ‘Bandwidth’ of the sound in ultrasonics.

THE ACOUSTIC SPECTRUM

Sound waves are described above as the oscillation of particles of solids, liquids or gases. The

human ear can only detect a small range of possible vibration frequencies, roughly between 16 cycles

per second and 20,000 cycles per second. In theory, however, there is a limitless spectrum of

frequencies and that are possible even if humans can’t hear the whole range. The spectrum is

illustrated in figure 1.2 below: -

Typical test range

Subsonic range

Audible Range Ultrasonic range

20MHz 0.5MHz

0 10 100 1000 10,000 100,000 1,000,000 10,000,000 100,000,000 Hz

Fig. 1.2 Acoustic Spectrum

The unit used to denote frequency is the Hertz, abbreviated as Hz, where 1Hz is one cycle per

second. One thousand Hz is written as 1KHz (Kilo Hertz) and one million Hz as 1MHz (Mega Hertz).

The part of the spectrum from zero to 16Hz is below the range of human hearing and is called the

Page 3: Ultrasonic Back-to-Basic

‘Subsonic’, or ‘Infrasonic’ range. From 16Hz to 20KHz is known as the ‘Audible’ range and above

20KHz as the ‘Ultrasonic’ range. Ultrasonic flaw detection uses vibrations at frequencies above

20KHz.

Most flaw detection takes place between 500KHz and 20MHz although there are some applications,

for example in concrete, that use much lower frequencies and there are special applications at

frequencies above 20MHz. In most practical applications in steels and light alloys, frequencies

between 2MHz and 10MHz predominate. Generally the higher the test frequency, the smaller the

minimum detectable flaw, but it will be shown in following articles that higher frequencies are more

readily attenuated by the test structure. Choosing an appropriate test frequency becomes a

compromise between the size of flaw that can be detected and the ability to get sufficient sound

energy to the prospective flaw depth.

MODES OF PROPAGATION

Sound energy travels, or ‘propagates’, outwards from the source of the vibration as the oscillation of a

particle of solid, liquid or gas disturbs the neighbouring particles so that the neighbour takes up the

oscillation. It will take time for the disturbance, called the ‘sound wave’, to reach a given distance from

the source. This is a measure of the velocity of sound in a given medium. It will be shown that this

velocity varies with the characteristics of each material and the way in which the disturbance is

transmitted from one particle to the next. The different ways in which the disturbance may be

transmitted are known as the ‘Modes of Propagation’.

The different modes of propagation come about because solids, unlike liquids and gases, have a

modulus of rigidity as well as a modulus of elasticity. Solids Liquids and gases all show resistance to

compressing or stretching. In the case of solids we refer to this resistance as Young’s Modulus of

Elasticity (‘E’) for the material. The elasticity of a solid is plotted when a tensile test is carried out and

from the resulting graph the ‘Ultimate Tensile Strength’ (UTS) of the material can be derived. The

Modulus of Rigidity (‘G’) is the material’s resistance to a shear load.

COMPRESSION WAVE MODE Because liquids and gases have no modulus of rigidity, sound waves can only propagate by using

their resistance to tension and compression. This type of sound wave is called the ‘Compression

Wave’. Compression waves can exist in solids, liquids and gases because they all have elasticity.

Compression waves are also known as ‘Longitudinal’ waves, and sometimes as ‘Plane’ waves The

individual particles of the solid liquid or gas oscillate about their mean position, and the direction of propagation of the compression wave is in the same plane as the particle motion as shown in

figure 1.3.

Page 4: Ultrasonic Back-to-Basic

Particle Motion Direction of Propagation Fig. 1.3 SHEAR WAVE MODE

Shear waves only exist in solids and rely on the modulus of rigidity of the solid under test, they can

exist on their own or co-exist with compression waves and surface waves. Shear waves are also

sometimes called ‘Transverse’ waves. Again, the individual particles of the solid oscillate about their

mean position, but the direction of propagation of the shear wave is at right angles to the particle motion. This is illustrated in figure 1.4.

Par

ticle

Mot

ion

Direction of Propagation Fig. 1.4

SURFACE WAVE MODE

At the surface of a solid, a complex mode of oscillation can exist in which the particle motion is mainly

perpendicular to the direction of propagation as with the shear wave, and partly in the same plane as

the direction of propagation as with the compression wave. This mode of propagation is called the

‘Surface wave’ or ‘Rayleigh wave’. Surface waves only affect the surface layer of the solid to a depth

of about one wavelength, and have the advantage that they follow the surface contour of the object

and only reflect at an abrupt change such as a corner or crack. For the surface wave, the particle motion is elliptical with the major axis of the ellipse at right angles to the direction of propagation. This is shown in figure 1.5.

Fig. 1.5

Direction of Propagation Elliptical

Particle Motion

Page 5: Ultrasonic Back-to-Basic

LAMB WAVE MODES

Lamb waves, like Surface waves, propagate parallel to the test surface and have an elliptical particle

motion. They occur when the thickness of the test material is only a few wavelengths at the test

frequency and where the test piece is of uniform thickness. Lamb waves fill the wall thickness and

propagate along the major axis of the component. They can travel several meters in steel, so they

can be used for rapid scanning of plate tube and wire. Recent developments for rapid corrosion

monitoring in buried pipes use Lamb waves under the name ‘Guided Waves’. The wall of the

component flexes so that the sound ripples along the material distorting both surfaces. Figure 1.6

illustrates a type of Lamb wave where the crests of the wave on the near and far surfaces coincide.

These are called Symmetrical Lamb Waves. Figure 1.7 shows another type of Lamb wave where the

crest on one side coincides with a trough on the other. These are called Asymmetrical Lamb Waves.

CREEPING (LAT

There is a speci

the surface rathe

Reference: - ‘Ultr Published in INS

Original plate surface Particle motion

Fig. 1.6

Original plate surface Particle motion

Fig. 1.7

ERAL) WAVES

al type of compression wave called a ‘Creeping’ or ‘Lateral’ wave. It sneaks along

r like a surface wave, its use is described later under ‘TOFD techniques’.

asonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury

IGHT magazine November 2004

Page 6: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

2. PROPERTIES OF SOUND WAVES

VELOCITY

Sound travels at different speeds through different materials. This is noticeable when, for

example, a railroad worker is observed from a distance striking a rail with a hammer. Since

the speed of light is much faster than that of sound, the observer first sees the hammer strike

the rail. If the observer is also close to the rail, the next event is the sound of the blow coming

out of the rail and finally the airborne sound is heard.

This tells us that the speed of sound in the rail is faster than the speed of sound in air. It is

true that sound travels faster in liquids than in gasses and faster in metals than in liquids.

However, it is also true that sound travels at different speeds in different metals. There is a

distinct speed of sound for each material and in ultrasonics this is called the VELOCITY of

sound for that material. This being so, it would be useful to have an understanding of the

reasons for the difference.

Imagine two pairs of identical steel balls, one pair joined by a strong compression spring and

the other pair by a weak spring. If one of each pair is moved towards its partner at a constant

speed, the spring joining the pair will start to compress. Eventually there will be enough

compression in the spring to overcome the inertia of the second ball and it will start to move.

As shown in figure 1, the second ball will move sooner for the pair connected by the stronger

spring.

M1 P

P = Constant force for both pairs

M1 = Movement of strong spring second ball

t1 = Time taken to movement M1

M2 = Movement of weak spring second ball

t2 = Time taken to movement M2

M2

t1 0 Time

P

t2

Fig. 1

In the analogy, the balls represent the particles of solid, liquid or gas through which the sound

wave is propagating and the springs represent Young’s Modulus of elasticity ‘E’. The

suggestion made by the analogy is that the disturbance will pass more quickly from one

particle to the next in a material having greater elasticity. In other words, the velocity of a

compression wave will be higher for greater values of elasticity. This is generally the case but

there is another main factor affecting velocity, and that is the density of the material.

Page 7: Ultrasonic Back-to-Basic

Consider another situation in which pair of aluminium balls and a pair of lead balls replace the

steel pairs in the above analogy but with each pair joined by springs of equal strength. The

inertia of the lead ball is greater than that of the aluminium ball and this time it will take longer

to get the lead ball moving. This suggests that the compression wave velocity will be lower for

high-density materials than for low-density materials. Density and elasticity are the dominant

factors affecting velocity and the expression for the compression wave velocity in a fine wire

is taken to be: -

ρEVco =

However, when we have a bulk of material, the sample is more rigid than a fine wire giving an

effective increase in Young’s Modulus and we need to modify the expression to take account

of Poisson’s Ratio. During a tensile test, to measure the strength of a metal sample, the

diameter of the sample reduces as the sample is stretched. The change in diameter divided

by the change in length is Poisson’s Ratio. Considering all these factors, the velocity of a

compression wave in a bulk material can be calculated from the following formula: -

( )( )σσσ

ρ 2111

−+−

⋅=EVc

Where

cV = Compression wave velocity

E = Young’s Modulus of Elasticity

ρ = Material Density

σ = Poisson’s Ratio

Shear waves are able to exist in solids but they do not travel at the same velocity as the

compression wave in a given material. This is because it is the Modulus of Rigidity, rather

than Young’s Modulus, that dictates the velocity, and the modulus of rigidity is lower than the

modulus of elasticity. This means that the shear wave velocity is always slower than the

compression wave velocity in a material. In Liquids and gases, the value of the modulus of

rigidity is so low that shear waves cannot propagate. As a rule of thumb, the shear wave

velocity is roughly half the compression wave velocity. The velocity can be calculated from: -

( )σρ +⋅=

121EVs Or, alternatively

ρGVs =

Where

sV = Shear wave velocity

G = Modulus of Rigidity

ρ = Material Density

Page 8: Ultrasonic Back-to-Basic

σ = Poisson’s Ratio

Surface (Rayleigh) waves also have their own particular velocity, which is generally taken to

be approximately 90% of the shear wave velocity.

Although the velocity for each of these modes of propagation can be calculated, it requires a

precise knowledge of all the parameters, and these are not usually available to the ultrasonic

practitioner. Parameters such as density and strength vary with alloying, heat treatment,

casting, rolling and forging processes – all of which make it difficult to know that the correct

values are being used. Instead, it is more normal to carry out a routine called ‘Calibration’

during the setting up procedure for an ultrasonic inspection. In the calibration procedure the

flaw detector time-base is adjusted to give a convenient scale against a calibration sample of

known thickness and made of the same material as the work to be tested. Table 1 at the end

of this chapter lists the compression and shear wave velocities for a number of materials.

WAVELENGTH

While the particles are completing each cycle of their oscillation, the sound wave is moving

outwards in the direction of propagation at the characteristic velocity for that material. It

follows that during the time taken to complete one cycle of vibration, the sound wave will

move a certain distance depending on the velocity in that material. For a given sound

frequency, this distance is relatively small for liquids and gasses compared to that in metals,

because velocities are higher in metals. The distance travelled by the sound wave during one cycle of vibration is called the WAVELENGTH. In general, if the maximum dimension

of a reflecting surface is equal to or greater than half a wavelength, the reflection will be

detectable. It follows that calculation of the wavelength will help in the choice of test frequency

for a specific application.

Wavelength is given the Greek symbolλ (lambda) and for any material and sound frequency,

wavelength can be calculated from the equation: -

fV

Where λ = wavelength

V = Velocity

f = frequency

At ultrasonic frequencies, the wavelength of sound in metals is relatively short and so it is

usual to express the wavelength in millimetres. This is done at the start of the calculation by

changing the velocity from meters to millimetres a second by multiplying the value in M/sec by

1,000.

Page 9: Ultrasonic Back-to-Basic

Example

Calculate the wavelength of a 5MHz compression wave in steel (Vc = 5960 m/sec).

500000010005960×

mm192.1=λ

ACOUSTIC IMPEDANCE

Acoustic impedance of a material is the product of the material’s density and velocity. At the

interface between two materials, the acoustic impedances either side of the interface will

determine what proportion of the incident sound wave will reflect and what proportion will

transmit into the second material. The symbol allocated to acoustic impedance is ‘Z’ and for a

given material, VZ ×= ρ

REFLECTION

Incident sound

Reflected sound

Material 2 Acoustic Impedance Z2

Material 1 Acoustic Impedance Z1

Interface

Transmitted sound Fig. 2 Figure 2 shows the interface between two materials whose acoustic impedances are Z1 and

Z2 respectively. In the example, part of the energy is transmitted into Material 2 and part is

reflected back into Material 1. The percentage of the incident energy that is reflected can be

calculated from the equation: -

%1002

21

21 ×⎟⎟⎠

⎞⎜⎜⎝

⎛+−

=ZZZZRE

Where: -

RE is the reflected energy

Z1 & Z2 are the acoustic impedances

Example 1

Calculate the percentage of the incident energy that would be reflected at a ‘steel to water’

interface given that Zsteel = 46.7 and Zwater = 1.48.

Page 10: Ultrasonic Back-to-Basic

%10048.17.4648.17.46 2

×⎟⎠⎞

⎜⎝⎛

+−

=RE

%10018.4822.45 2

×⎟⎠⎞

⎜⎝⎛=RE

%88=RE

Note that the remaining 12% is transmitted into the water.

If the example had been given as a ‘water to steel’ interface, the second line of the calculation

would have shown a negative value inside the brackets. However, the square term outside

the bracket would restore the answer to a positive value and the answer would have been the

same 88% reflected, this time in the water, and 12% would have been transmitted into the

steel.

When the interface is between two solids, as in the case of a brazed joint between two pieces

of steel, the reflected energy is much smaller, most of the energy passing across the braze

and into the second steel layer. There are also examples of two very different materials that

have the same acoustic impedance such as Ro-cee rubber and water. Sound travelling

through water and then encountering this particular rubber compound will carry on through

the rubber as if the interface did not exist. Table 1 shows the acoustic impedance for a

number of materials.

COUPLANT

Acoustic impedances for metals tend to have high values whereas those for gasses are low.

From the above example it is clear that at a solid to gas interface, the proportion of energy

reflected is going to be very high and the proportion transmitted will be very low. That is useful

because it means that a discontinuity such as a crack or a void in a metal object will reflect

almost all the sound back to the test surface. However, it is also a nuisance because it means

that air between the ultrasonic probe and the test surface will prevent the sound from entering

the component. A couplant is a liquid or paste used between the probe and the test surface

to try to match the acoustic impedance of the probe to that of the test material. It is not a very

efficient process because the best couplants, for example glycerine, only allows about 15% of

the sound to enter the component, and only the same proportion of any energy coming back

to the test surface can enter the probe to give an echo. At best, then, only a little over two

percent of the energy generated at the probe ever gets back to the display.

There are specially formulated couplants for use in flaw detection as well as water, oils,

greases, glycerine and pastes such as wallpaper paste. The most important considerations

when choosing a couplant are firstly that it is not hazardous to the individual and secondly

that it will not adversely affect the component.

References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury ‘Ultrasonic Flaw Detection in Metals’ – Banks Oldfield & Rawding – ILIFFE 1962

Page 11: Ultrasonic Back-to-Basic

Material Velocity (C) Velocity (S) Density Acoustic Impedance

Units m/s x 103 m/s x 103 kg/m3

Air 0.33 - 1.2 0.0004

Aluminium 6.40 3.13 2700 17.3

Barium Titanate A 5.26 - 5700 30.0

Barium Titanate B 5.53 - 5700 31.5

Beryllium 12.89 8.88 1800 23.2

Brass 4.37 2.10 8450 37.0

Cast Iron 3.5 to 5.6 2.2 to 3.2 7200 25 to 40

Copper 4.76 2.33 8930 42.5

Glass (Crown) 5.66 3.42 3000 17.4

Gold 3.24 1.20 19300 63.0

Iron 5.96 3.22 7850 46.8

Lead 2.40 0.79 11300 27.2

Lead Niobate 2.76 - 5800 16.0

Lead Zirconate Titanate A 3.00 - 7600 22.8

Lead Zirconate Titanate B 3.00 - 7500 22.5

Lithium Sulphate 5.45 - 2060 11.2

Magnesium 5.74 3.08 1720 9.90

Mercury 1.45 - 13550 19.6

Molybdenum 6.25 3.35 10200 63.7

Nickel 5.48 2.99 8850 48.5

Oil 1.44 - 900 1.3

Perspex 2.68 1.32 1200 3.2

Platinum 3.96 1.67 21400 85.0

Polystyrene 2.35 1.12 1060 2.5

Steel (Mild) 5.96 3.24 7850 46.7

Steel (Stainless) 5.74 3.13 7800 44.8

Silver 3.70 1.70 10500 36.9

Tin 3.38 1.61 7300 24.7

Titanium 5.99 3.12 4500 27.0

Tungsten 5.17 2.88 19300 100.0

Tungsten-Araldite 2.06 - 10500 21.65

Tungsten-Carbide 6.65 3.98 10000 to 15000 66.5 to 98.5

Uranium 3.37 2.02 18700 63

Water 1.48 - 1000 1.48

Zinc 4.17 2.48 7100 29.6

Zirconium 4.65 2.30 6400 29.8

Table 1

Page 12: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

3 REFRACTION & MODE CONVERSION

Previous diagrams have shown the incident sound as if it were a single ray of energy, but of course it

is really a beam that has some width, rather like a torch beam. If the incident beam is directed at an

interface between water and steel at an angle other than normal, the angle taken up by the

transmitted beam in the steel will be greater than the incident angle in water. The advancing wave

front in a sound beam can be defined as the plane in which all the oscillating particles are ‘in phase’,

or at the same position in their oscillating cycle. The bottom edge of the beam shown in figure 1

arrives at the interface first and immediately takes up the faster velocity of the steel. As the rest of the

wave front reaches the interface, so the transmitted beam gradually takes up steel velocity. By the

time that the top edge of the beam enters the steel, the sound from the bottom edge has already

travelled four times further than it would have in water. Joining up the ‘in phase’ points on the wave

front at the instant the top edge enters the steel shows the wave front advancing at a new angle. The

beam of sound is said to have undergone ‘Refraction’ as it crossed the interface and the new angle is

called ‘the angle of Refraction.

Transmitted Beam

Water

Steel

Incident Beam

Refraction takes place in the time taken for the beam to cross the interface

Interface Normal Fig. 1

The refraction occurs because of the difference in velocity on either side of the interface and the

proportions of energy reflected in the water and transmitted into the steel remain the same as it would

be for normal incidence. Figure 2 shows the incident, reflected and refracted angles. These angles

are always measured from the Normal to the interface. In the diagram, ‘i’ is the angle of incidence, ro

is the angle of reflection and ‘R’ is the angle of refraction.

Page 13: Ultrasonic Back-to-Basic

Normal

i r

R

Interface Medium 1 Medium 2

Incident compression wave

Reflected compression wave

Refracted compression wave

Fig. 2

The angles and velocities are related and the relationship is expressed in Snell’s Law such that: -

211 VSinR

VSinr

VSini

== Where: -

i = Angle of Incidence

r = Angle of Reflection

R= Angle of Refraction

V1= Velocity in Medium 1

V2= Velocity in Medium 2

MODE CONVERSION

If Medium1 is a liquid and Medium 2 a solid, some of the energy in the solid will change to the Shear

Wave mode. This change is known as Mode Conversion. For small angles of incidence the

proportion of energy changing to shear wave mode is small and can be ignored. However as the

angle of incidence increases the proportion increases and the shear wave becomes significant so that

there can be two types of wave in medium 2 at the same time, both of which can reflect from surfaces

within the object. Since they both travel at different speeds, and Snell’s Law tells us that they will

refract in different directions, the results can be very confusing. This was a restricting factor in

ultrasonics until Sproule developed the first Shear wave angle probes in 1947. Until then it was

unsafe to rely on angles of refraction greater than about 100 since echoes from the compression wave

could not be discriminated from the shear wave reflections. Because of this ambiguity, ultrasonics

tended to be restricted to the detection of discontinuities with surfaces parallel to the scanning surface

such as laminations and cavities. Attempts to detect, for example, weld defects such as lack of

sidewall fusion and root cracks by angling the beam were not reliable.

Page 14: Ultrasonic Back-to-Basic

Sproule realised that the compression wave refracted angle would always be about double the shear

wave refracted angle because the shear wave velocity is about half the compression velocity.

Therefore if the angle of incidence were to be increased progressively there would be a critical angle

of incidence at which the compression wave would refract through 900. Any increase in angle of

incidence beyond this critical angle would leave only a shear wave in medium 2 and the compression

wave would undergo total internal reflection in Medium1. With only a shear wave in medium 2

travelling at a known velocity and at a known angle, the field was open for many new applications of

ultrasonics. The critical angle at which the compression wave is refracted through 900 is called the first critical angle. For a water to steel interface the first critical angle is about 150 and for a Perspex

to steel interface the angle is about 280. At these critical angles, the remaining shear wave is at an

angle of refraction just over 300. Increasing the angle of incidence above the first critical angle causes

the compression wave to be totally reflected in medium 1 and the shear wave refracted angle to

increase so that transducers can be produced at a suitable angle to detect particular defect

propagation directions.

Eventually a second critical angle of incidence will be reached at which the shear wave will be

refracted through 900. The shear wave at this second critical angle will again mode convert, this time

to become a Surface (Rayliegh) wave. This new wave travels at 90% of the shear wave velocity, only

penetrates to a depth of about one wavelength, will follow the surface contour of the object and will

only reflect at an abrupt change in surface direction such as a corner or a crack. If the angle of

incidence is increased beyond the second critical angle, no sound will be transmitted into medium 2.

Ultrasonic transducers having refracted angles between 00 and 100 are likely to be compression wave

probes and those with refracted angles between 350 and 800 will be shear wave probes. Surface

wave probes have a refracted angle of 900. Between 100 and 350, and 800 to 900 it would be possible

to have two simultaneous modes existing in Medium 2 and so it is unusual to find transducers in

these two ranges – exceptions to this rule will be discussed in a later chapter.

0

10

20

30

0 30 60 9

Refracted angle Sreel

Inci

dent

ang

le W

ate

0

r

Shear

Compression

Fig. 3

Page 15: Ultrasonic Back-to-Basic

0

20

40

60

0 30 60

Refracted angle Steel

Inci

dent

ang

le P

ersp

ex

90

Shear

Compression

Fig. 4

Figures 3 and 4 show the relationship between the incident angle and refracted angle for water to

steel and Perspex to steel interfaces. The graphs show that the second critical angle for water to steel

is about 280 and for Perspex to steel about 580. These values would be different if medium 2 were to

be aluminium or some other solid than steel.

Example 1 An incident compression wave in water meets a steel interface at an incident angle of 190, calculate

the shear wave refracted angle in the steel given that the compression wave velocity in water as

1480m/s and the shear wave velocity in steel as 3240m/s.

From Snell’s Law

21 VSinR

VSini

= Therefore: -

1

2

VxSiniV

SinR =

1480325603240 .xSinR =

71280.SinR =

R = 45.46o

From a practical point of view it is more usual to know the refracted angle needed in the test material

in order to detect a particular discontinuity, and so the calculation would be to find the necessary

Page 16: Ultrasonic Back-to-Basic

angle of incidence, in water for immersion testing, or in Perspex for contact scanning. Example 2

shows this version of the application of Snell’s Law.

Example 2

Calculate the angle of incidence required in Perspex in order to produce 45o Shear wave in steel

given that the compression wave velocity in Perspex is 2680m/s and the shear wave velocity in steel

is 3240m/s.

From Snell’s Law

21 VSinR

VSini

= Therefore: -

3240707102680 .xSini =

58490.Sini =

Incident angle = 35.8o

REFLECTIVE MODE CONVERSION Mode conversion also takes place when an ultrasonic beam reflects at internal surfaces in solids

whether these are boundary surfaces, machined features, or discontinuities. The relationship

between incident angle of a given beam and the relative amplitude of the reflected and mode

conversion beams for steel is shown in the following graphs. They allow an assessment to be made

of the potential confusion in any given situation and can be used to determine an alternative test

angle to be chosen to avoid the problem.

900600 30000

400

300

200

100

500

β

β

S

C

0.4

0.6

0.8

1.0 Relative Amplitude

a) Incident Compression Wave

Steel Air

β α

S C C

Fig. 5

0.2

α

Page 17: Ultrasonic Back-to-Basic

A compression wave incident on a steel to air interface will reflect as a compression wave together

with a mode converted shear wave. For example figure 5 shows that at an incident angle (α) of 30o

we find that β is around 15o but the relative amplitudes of the shear wave and compression wave are

90% and 70% respectively. Both will give strong signals if they reach the receiver. In the extreme

case, where α is around 60o and β around 30o we find that the relative shear wave amplitude is 90%

but the reflected compression wave amplitude has fallen to only about 10%. For greater angles of

incidence than 60o, the shear wave rapidly decreases in amplitude and the compression wave

recovers. Clearly we need to take care in our interpretation of signals if we see that a compression

wave in steel is likely to meet a known reflecting surface in that part of the graph where the shear

wave amplitude is significant.

500400 300 200100

1.0

Relative Amplitude

0.8

0.6

0.4

0.2

00

α

C

S

b) Incident Shear Wave α

900 Steel Air

β

C α

S S

600 300

Fig. 6

β A shear wave will reflect as a shear wave together with a mode converted compression wave. Using

the graph in figure 6 we can see that the most severe case is when the incident shear wave meets a

steel to air interface at about 30o. The reflected shear wave amplitude is very low and the mode-

converted compression wave is very strong and almost perpendicular to the test surface.

If the incident shear wave grazes a surface, in other words the incident angle is around 90o, there will

be a mode conversion to Rayleigh wave. This can happen when a shear wave grazes the bore of a

machined hole in the specimen. In that case the Rayleigh wave will follow the bore surface and will

reflect if it encounters a sharp changes to the bore such as a keyway. If you are not aware of the

Page 18: Ultrasonic Back-to-Basic

possibility, you may assume that there is a discontinuity in a false position. An example is shown in

figure 7 below. Of course, when you are aware of this phenomenon, you can use it to advantage to

find a crack in an otherwise ‘blind’ position around a hole or radius.

Assumed reflection point Rayleigh wave reflects from keyway Fig. 7 References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury ‘Ultrasonic Flaw Detection in Metals’ – Banks Oldfield & Rawding – ILIFFE 1962

Page 19: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 4. TRANSDUCERS FOR GENERATING AND DETECTING SOUND WAVES

SWINGING THE LEAD

There is an amusing story about a nautical gentleman at the time when wooden ships were being

superseded by iron ones. This sailor thought up a new way to determine the depth of water under the

hull to replace the old lead weight on a rope method. He decided that it should be possible, with a

large hammer and a stopwatch, to bang on the iron bottom of the ship with the hammer and time the

return echo from the seabed with the stopwatch. The measured time could be used to calculate depth

using the speed of sound in water. Fired with enthusiasm, he gathered together a number of marine

dignitaries in the bilges of his ship, passed a large sledgehammer to a muscular boatswain, took out

his stopwatch and ordered the ‘swain to wallop the floor. This he did with such vigour that the hull

boomed for ten minutes and the assembled observers were deafened for a month! Nobody heard an

echo.

There are parallels with ultrasonic flaw detection in the story; we need our sound pulses to be ‘loud’

enough to penetrate to the depth of the anticipated flaw and we need the duration of the pulse to be

short so that it does not mask any returning echoes. We also need the sound frequency of the pulse

to produce a wavelength short enough to detect the smallest reflector that must be detected to ensure

safety. In this chapter we will discuss various ways of generating and detecting suitable pulses and

some of the limitations we face in terms of penetration and flaw sensitivity.

ULTRASONIC TRANSDUCERS

A transducer is a device that will change one form of energy into another. Ultrasonic transducers

change electrical energy into mechanical energy (sound waves) or vice versa. There are several

methods used to generate and detect ultrasonic pulses in modern flaw detection and the most

common of these makes use of the Piezo Electric effect found in certain materials. Other methods,

such as the Electro Magnetic Acoustic Transducer (EMAT) and Laser technology will also be

described.

PIEZO ELECTRIC TRANSDUCERS

In 1880 the Curie brothers discovered that slices cut in a particular way from certain crystal materials

would generate an electrical potential across the faces of the slice when distorted by a mechanical

force. They called this phenomenon ‘Piezoelectricity’ from the Greek words for ‘Pressure’ and

‘Electricity’. A year or so later Lippman reported that the reverse was true; that a voltage applied

across the slice would produce a mechanical distortion. Quartz was the prime example of a

piezoelectric material, but Rochelle salts and Tourmaline crystals also displayed the same effect.

Page 20: Ultrasonic Back-to-Basic

For the first thirty years of ultrasonic flaw detection from Sokolov in 1929, until the end of the nineteen

fifties, quartz was the most common transducer material. Appropriate slices were cut from a single

crystal. Later new polycrystalline materials were developed that had lower electrical impedance

(resistance to high frequencies) and gave better ultrasonic performance, as much as 60 to 70 percent

more efficient than quartz. These materials have to be ‘Polarised’. During polarization the individual

crystals align themselves in the same direction so that their combined effect is coherent. The

polarisation process involves heating the discs in an oil bath to a critical temperature called the ‘Curie Temperature’, applying a strong electrostatic field across the disc and then allowing the temperature

to cool slowly. Figure 1 illustrates the polarising process.

Electrostatic field

Heated oil - - - - - - - - - - - - -

++++++++++++++

Polycrystalline disc

Fig. 1

The Curie temperature differs for each of the common materials used in ultrasonics, so that the oil

bath will need to be heated to a suitable temperature for the material in use. For Barium Titanate the

Curie temperature is around 120oC whereas for various grades of Lead Zirconate Titanate (PZT) the

temperature is from 190o to 350oC and for Lead Metaniobate (PMN) it is about 400oC. If the material

is subsequently heated to a temperature near to the Curie temperature, the disc will ‘depolarise’ and

lose its piezoelectric properties. It follows that care needs to be taken to avoid depolarisation when

testing hot materials and this will sometimes influence the choice of transducer material.

MODE OF VIBRATION

Whether the transducer disc is made from a naturally occurring piezoelectric crystal, or one of the

polarised polycrystalline materials, we usually refer to the disc as ‘the crystal’ when talking about

probe construction. The crystal ‘disc’ or ‘plate’ may be round or rectangular and for some applications

may be curved plates or concave discs to focus the sound. The way in which the plate vibrates when

stimulated by an electrical pulse depends upon the ‘cut’, in the case of quartz, or the direction of

polarisation in the case of polycrystalline materials. Figure 2 represents a typical quartz crystal

showing the three axes defined by crystallographers, and two plates cut from a crystal, one an X-cut

plate and the other a Y-cut plate.

Page 21: Ultrasonic Back-to-Basic

Z

X Y

X

X-cut crystal X X

Y Y

Y Z Y-cut crystal

Fig. 2

An X-cut plate is taken from the quartz crystal so that the X-axis is perpendicular to the plate and the

Y-cut plate has the Y-axis perpendicular to the plate. If a voltage is applied across the faces of these

plates, an X-cut crystal will distort in the thickness mode whereas a Y-cut crystal will distort in shear

mode. Figure 3 illustrates the changes in shape when an alternating voltage is applied to an X-cut

crystal and Figure 4 shows the shape changes for a Y-cut crystal. The same two modes of vibration

can be obtained using the polycrystalline materials by polarising across the faces of the plate

(equivalent to X-cut), or parallel to the faces of the plate (equivalent to Y-cut)

The X-cut crystal is the one most commonly used in ultrasonic flaw detection, it can generate and

detect compression waves, and can therefore transmit sound through the liquid couplant we use.

Since shear waves cannot exist in liquids or gases, the only way in which a Y-cut crystal could be

used to generate shear waves in a metal object would be to use a solid couplant or high viscosity

Fig. 4 Fig. 3

Page 22: Ultrasonic Back-to-Basic

liquid such as honey; in other words we would need to almost glue the crystal in position. This is done

in a few very special applications.

METHOD OF PULSING AND FREQUENCY

When we generate a short pulse of sound with our ‘crystal’, we don’t ‘drive’ the crystal with an

alternating voltage of suitable frequency; instead, we ‘pluck’ the crystal with a short sharp electrical

shock and allow the crystal to ‘ring’ at its natural resonant frequency. This is rather like ‘plucking’ a

guitar string that also vibrates at its natural frequency. In the case of the piezoelectric plate, the

crystal stretches as the voltage is applied and only produces sound when the voltage is rapidly cut

off. To increase the amplitude (loudness) of the ultrasound we increase the peak voltage (pulse

energy) applied to the crystal. The frequency of our ultrasonic transducer is determined by the

thickness of the crystal. As the crystal is made thinner, so the resonant frequency increases. Quartz

crystals are split in the appropriate and then lapped to the correct thickness for the required

frequency. The polycrystalline materials are made as slurry that is moulded and compacted under

pressure and then sliced and lapped to the required thickness before polarising.

The required thickness for a given frequency can be calculated from the frequency-thickness constant for the crystal material to be used. Since this depends on the velocity of a compression

wave in that material it can be seen that the thickness for a given frequency will not be the same for

PZT and quartz, for example. The frequency-thickness constant is defined mathematically as: -

2Vfxt =

Example

Calculate the re

the compression

2Vfxt =

xfVt

2=

xt

500000023000

=

t = 0.3mm

CONTROL OF P

In ultrasonic flaw

entering the sca

determine the d

Where f = the desired frequency

t = the crystal thickness

V = the compression wave

velocity in the crystal

quired thickness of a PZT crystal to produce a resonant frequency of 5MHz given that

wave velocity for PZT is 3000M/s.

mmx1000

ULSE LENGTH

detection we measure the time taken for each echo to arrive at the receiver after

nning surface of the object. If we know the velocity of sound in the material we can

istance travelled by the sound wave. Suppose that a crack has grown from a bolthole

Page 23: Ultrasonic Back-to-Basic

in the object as in figure 5; some of the sound will reflect from the top of the bolthole, and a little while

later, some will reflect from the crack. The arrival of the two echoes at the receiver will be separated

by a short interval of time (T2 – T1). If the ringing time of the crystal (pulse length) is longer than this

interval of time, then we may not be able to distinguish the crack from the top of the bolthole – we

may miss the crack. We say that we have not ‘resolved’ the two echoes or that the resolution is

poor. In order to improve resolution we need to ensure that the pulse length is as short as possible.

Bolthole Crack

T2 T1

T1T2

T1T2

Short pulse

Long pulse

Fig. 5 In ultrasonics we shorten the pulse duration by applying a weight to the back of the crystal known as

the ‘damping’ or ‘backing’ slug. The damping slug is often made of a mixture of tungsten powder in

an epoxy resin. The amount of damping applied to the crystal will govern the resolution of the probe.

A short pulse probe will have only one or two cycles whereas a longer pulse probe may have from

three to five cycles. An undamped crystal may have twelve or more cycles in the pulse. For a given

number of cycles in a pulse, the duration or space occupied by the pulse will depend on the

wavelength, which in turn depends on the probe frequency and the velocity of sound in the material

being inspected. Thus: -

Pulse length = Number of cycles in the pulse multiplied by the wavelength.

It is obvious that one way to improve resolution would be to increase the test frequency, however,

since the penetration of sound into the object decreases as the frequency increases, this is not

always possible. Choosing a suitable test frequency is often a compromise between resolution

penetration and flaw sensitivity and sometimes ultrasonics will not be able to detect a particular

discontinuity at the critical depth. While resolution is an important consideration in many applications,

it is not always the case and sometimes a longer pulse is preferable. For example, in the examination

of a long shaft such as a railway axle, the screen on the flaw detector may only be 75mm wide and

the display may represent the length of the shaft, say 2.5m long. A short pulse of 2 cycles will occupy

Page 24: Ultrasonic Back-to-Basic

such a small part of the screen that it is too faint to see and it would be better to use a longer, more

visible pulse

PIEZO-COMPOSITE TRANSDUCERS

In a more recent development of the piezoelectric transducer, the active plate in the test probe is

made by slicing piezoelectric crystals into small squares and assembling them into a matrix separated

with an epoxy or a rubber compound as shown in figure 6. The main advantages of this type of

construction are firstly, lower acoustic impedance allowing better acoustic matching. This is an

advantage when testing castings and stainless steel. Secondly, resolution – they tend to provide short

pulses without additional damping, allowing the probes to have a low profile.

Plan view

Epoxy

Crystals

Side view Fig. 6 POLYVINYLIDENE FLUORIDE (PVDF) TRANSDUCERS PVDF was also found to display Piezo electric characteristics and has been used in ultrasonic flaw

detection. These thin plastic films have advantages and limitations compared with conventional

crystals. On the plus side, they can be easily shaped to focus sound, they produce very short pulses

and they give good transmission into water because the acoustic impedance is similar to water.

Against these advantages, the films are fragile and cannot be used in contact scanning, and the

power output is relatively low compared with ceramic crystals. The main application is in high

resolution immersion testing.

ELECTROMAGNETIC ACOUSTIC (EMAT) TRANSDUCERS

EMAT transducers provide a non-contact alternative to piezoelectric transducers. Sound waves are

generated in the surface of a conductive test object by an electrical pulse applied to a flat coil that is

Page 25: Ultrasonic Back-to-Basic

positioned between a strong magnet and the test piece. The interaction between the magnetic field

generated in the coil by the electrical pulse and the fixed magnetic field of the magnet causes a rapid

‘shock’ deformation at the surface of the test piece and an ultrasonic wave travels through the metal

object. The EMAT probe needs to be close to the test surface, but does not need to touch it.

Returning echoes arriving at the scanning surface cause the surface to vibrate in the magnetic field.

This generates eddy currents in the test surface and the coil detects the eddy currents.

EMAT probes can be used with an air gap when testing hot surfaces and on surfaces coated with

non-conducting material such as rubber, paint and fibreglass because the sound wave does not have

to travel through the gap material. The probes can be configured to generate horizontally polarised

shear waves directly into the test object. This is an advantage when testing austenitic welds, castings

and other materials with dendritic grain structure because the shear wave does not mode convert

when it meets a reflecting surface that is parallel to the direction of polarisation. Because shear

waves travel at roughly half the velocity of compression waves and have shorter wavelengths, it is

possible to obtain better near surface resolution and this can be an advantage when testing thin

materials.

However, there are some disadvantages with EMAT probes, they are relatively large and inefficient

compared with conventional probes and they cannot be used on non-conducting test objects unless a

conducting coating is applied.

LASER TRANSDUCERS

Another non-contact method of generating ultrasound uses laser technology. A short burst of a laser

beam on the surface of the test object causes a thermal shock with rapid local expansion of the

surface. The sudden distortion of the surface causes an ultrasonic pulse to travel through the test

object. The returning echo distorts the test surface and this distortion can be detected by a separate

laser interferometer without a couplant, or can be detected with a conventional piezoelectric crystal

and couplant. The gap between the transducer and test surface can be greater than is possible with

EMAT probes and can be as much as 250mm (10 inches). Typical applications include the inspection

of composite materials in the aircraft industry.

References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury ‘Ultrasonic Flaw Detection in Metals’ – Banks Oldfield & Rawding – ILIFFE 1962

Page 26: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 5. PROBE CONSTRUCTION

COMPRESSION WAVE PROBES

Standard compression wave probes can be for contact scanning or for immersion testing. The contact

scanning probes are either single crystal or twin crystal (dual) in construction. The construction of a

typical single crystal contact probe is shown in the diagram in figure 1.

Wear face

Damping Slug

Housing

Cap

Co-axial connector

Crystal

Fig. .1

The thickness of the crystal determines the operating frequency as we described in an earlier article

and the faces of the crystal are coated in silver to make electrical contact.

The damping slug is cast onto the rear of the crystal and bonds to it as the epoxy sets. The amount of

damping used determines the pulse length. A fine wire is soldered to the back of the crystal, using a

solder that melts at low temperature, before adding the damping slug.

The wear face is glued to the front face of the crystal to protect it during contact scanning. The

thickness of the wear face is important. It is made to be one quarter of the wavelength at the test

frequency for the velocity of sound in the wear face material. This thickness gives maximum

transmission of sound out of the probe into the test sample. Some wear faces are made from shim

steel, others from a hardwearing ceramic material. The steel wear faces can be used to earth the

front face of the crystal to the probe housing and are less fragile if you drop the probe, but are

inclined to stretch and disbond from the crystal with use. If a non-conductive wear face is used, an

alternative earthing method must be used.

The wear face, crystal and damping slug assembly are then fitted into the housing, the other end of

the centre wire is soldered to the centre terminal of the connector and the cap and connector fitted to

the housing. Figure 2 is a photograph of a typical single crystal compression wave probe.

Page 27: Ultrasonic Back-to-Basic

Fig. .2

Twin crystal, or ‘dual’ probes are used to eliminate the ‘dead zone’ occupied by the transmission

pulse with a single crystal probe. In this type of probe one crystal acts as a transmitter, the other as a

receiver and the amplifier is isolated from the transmitting crystal. The two crystals are mounted on

acrylic or polystyrene wedges these components are illustrated in figure 3. An acoustic barrier,

usually made of cork, is fitted between the wedges and crystals to prevent cross talk between the

transmitter and receiver. Figure 4 shows a typical twin crystal probe.

Tx Crystal Acoustic barrier

Wedges Rx Crystal

Fig. 3

Fig. 4

Page 28: Ultrasonic Back-to-Basic

Immersion probes are similar in construction to that shown in figure 1 except that it is not necessary

to fit a wear plate and so the silvered face of the crystal is usually visible. Probes can be focussed

and this is achieved by fitting a plastic or epoxy lens to the front of the crystal, or by making a curved

sectioned crystal. Figure 5 shows a 20MHz immersion probe with a small diameter spherically

focussed crystal.

Figure 5

The lens or curvature can also be cylindrical as illustrated in figure 6. The cylindrical version is often

referred to as a ‘Paintbrush probe’ because it allows a wide scan.

Focussed beam

Curved crystal

Fig. 6

Focussing can also be achieved using a technique called ‘Phased Array’, although not with

conventional ultrasonic sets. The phased array probe contains a number of small crystals and the

pulsing circuit is designed to be able to apply a pulse to all crystals simultaneously to produce a

conventional zero degree compression wave, or to pulse each crystal separately with a small time

interval between each. In the diagram shown in figure 7, the outer elements are triggered first and a

progressive delay is used to pulse inner elements, the centre crystals being the last to be triggered.

The result is that the ultrasonic wavefront reinforces in the curved way shown in the diagram to focus

Page 29: Ultrasonic Back-to-Basic

at a region determined by the delay intervals. By changing these intervals, the focal length can be

changed. The principles of constructive and destructive reinforcement will be dealt with later in a later

article.

Fig. 7

Single crystal ‘Delay line’ probes are sometimes used in contact scanning to reduce the ‘Dead Zone’ below the beam entry surface occupied by the transmission pulse and probe noise. The delay line is

usually Perspex or a similar material and provides a stand off just like the water path in immersion

testing. The length of the delay line must be sufficient to allow one or more backwall echoes in the

specimen depending on the application. Figure 8 is an example of a delay line probe.

Fig. 8

SHEAR WAVE PROBES

Since shear waves cannot travel through liquids or gases, angled beam probes use compression

waves in the incidence wedge in contact probes. The incident angle will be an angle between the first

and second critical angles (described in the third article in this series) so that we only have the mode

converted shear wave in the test material. Figure 9 is a sketch of the typical arrangement.

n

Incident wedge

e

Compression Wave Element Test surface

Shear wav

Internal Reflectio

Fig. 9

Page 30: Ultrasonic Back-to-Basic

We not only get a mode converted, refracted shear wave in the test piece, but we also have a

reflected compression wave in the wedge. If this internal reflection manages to get back to the crystal

face as it bounces around the wedge, we would have a standing echo that would be confusing.

Several methods of avoiding this problem have been used over the years. The earliest probes used a

long Perspex shaped ‘Cusp’ so that the reflection would be absorbed before it could return to the

crystal. The Cusp made a rather unwieldy probe and the next design used ‘V’ shaped grooves in the

front and top surfaces of the incident wedge to scatter the internal reflections. Some had a plastic

material moulded onto these grooves to further damp the reflection. In the latest, most compact,

designs the wedge is surrounded by a material that has a good acoustic match to Perspex, but a

much higher absorption of sound. The internal reflections are transmitted easily into this layer and

then absorbed. Figure 10 shows examples of the three designs and illustrates their relative size.

Fig. 10

Figure 11 is a photograph of a sectioned shear wave probe, showing the crystal, incidence wedge

and the blocking medium for the internal reflections.

Coaxial connector

Housing

Crystal Blocking medium

Incidence wedge

Fig. 11 Phased Array transducers, such as the one already discussed (figure 7), are also used to generate

angled shear waves in the test piece. These transducers have the advantage that the phase delay

between the crystal elements can be varied to give different angles of refraction. The delays can be

swept through a range of values to give a shear wave beam that sweeps through a desired range of

shear wave angles rather as a Radar scanner sweeps the skies.

In the fourth article, we said that EMAT probes could generate compression or shear waves, but that

shear waves were often used because they can be directed perpendicular to the test surface (that is

Page 31: Ultrasonic Back-to-Basic

a 00 probe). That has advantages in resolution, because the wavelength for a shear wave is about

half the wavelength for a compression wave and because the velocity of the shear wave is about half

that of the compression wave, we are able to measure thinner sections than we can with conventional

00 probes of the same frequency. The EMAT probe shown in figure 12 is a radially polarised shear

wave probe operating broadband between 1-10MHz, with a centre frequency of about 5MHz.

Courtesy of Ultrasonics Group,

Dept of Physics, University of Warwick

Fig. 12

‘Q’ FACTOR AND BANDWIDTH

Up to this point we may have gained the impression that our transducer produces a pure note at the

calculated frequency, but this is not true. In fact the sound wave produced contains a band of

frequencies related to the thickness of the crystal, its diameter or length and width plus the effects of

the damping medium. In addition the electrical characteristics of the transducer and associated

circuits affects the overall spectrum of frequencies. We refer to this spectrum as the ‘Bandwidth’ of

the probe. In a well-designed probe, the centre of this band should be the desired probe frequency

and the lower and upper limits are usually defined as the frequencies at which the amplitude is

reduced by a given factor. Some people use 30% (-3dB) and others 50% (-6dB) as the factor we will

use 50% in the following examples. Figure 13 illustrates the bandwidth of a 5MHz probe in which the

–6dB bandwidth is equal to the centre frequency, in other words, from 2.5MHz to 7.5MHz.

Page 32: Ultrasonic Back-to-Basic

0

20

40

60

80

100

0 2.5 5 7.5 10

Frequency (MHz)

Am

plitu

de (%

)-6dB points

Peak Frequency

Fig. 13

Figure 14 shows the bandwidth for another 5MHz probe, but this time the bandwidth is only from 3.75

to 6.25MHz.

0

20

40

60

80

100

0 1 2 3 4 5 6 7 8 9 10

Frequency (MHz)

Am

plitu

de (%

)

-6dB points

Fig. 14

The probe shown in figure 13 can be described as having a broad bandwidth whereas the probe in

figure 14 has a narrower bandwidth. In practice, short pulse probes have a broad bandwidth and long

pulse probes are narrow bandwidth. For a given crystal size, material and frequency damping not

only reduces pulse length, but also reduces pulse amplitude, so the narrower bandwidth probes will

have longer pulses but more amplitude in the pulse therefore giving deeper penetration.

Page 33: Ultrasonic Back-to-Basic

Another way of expressing bandwidth that is also common in other branches of electronics is the

‘Quality Factor’ or ‘Q’ of the probe and is defined by the formula: -

12

0ff

fQ

−= Where f0 = the centre frequency

f1 = the upper –6dB frequency

f2 = the lower –6dB frequency Example

Calculate the Q factor for the probe illustrated in figure 13.

12

0ff

fQ

−=

52575

..Q

−=

1=Q

Example 9

Calculate the Q factor for the probe illustrated in figure 14.

12

0ff

fQ

−=

7532565

..Q

−=

525.

Q =

2=Q

Undamped crystals can have a Q factor as high as 20,000 but for ultrasonic flaw detection the Q

factor is normally in the range 1 to 10.

References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury ‘Ultrasonic Flaw Detection in Metals’ – Banks Oldfield & Rawding – ILIFFE 1962

Page 34: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 6. PULSE-ECHO FLAW DETECTOR

The ultrasonic flaw detector is required to provide the voltage pulse to activate the probe

crystal, to amplify received signals from the probe and to display those signals so that the

relative time of arrival and amplitudes of the signal train can be viewed and interpreted. In

order to display the very short intervals of time involved in testing metals, the early pulse echo

systems used a cathode ray tube (CRT) as the display module. More recently, equipment

manufacturers have turned to digital technology and used LCD panels for the display. The

result has been the manufacture of much smaller and lighter ultrasonic equipment. Ultrasonic

sets in the early 1960’s used thermionic valves (vacuum tubes) and weighed 25 to 30 Kg (50

– 60 lbs). From the late 1960’s, transistor technology and smaller CRT’s meant that the flaw

detectors became smaller and lighter weighing between 5 and 10 Kg (10 – 20 lbs). In the new

millennium, the weight has come down to around 3 Kg. Figures 1 to 3 show the progression.

Circa 1960

Fig. 1

Figure 4 is a blo

and the controls

Circa 1980

Fig. 2

ck diagram of a typical analogue flaw detector sho

associated with each component.

Circa 2000

Fig. 3

wing the main components

Page 35: Ultrasonic Back-to-Basic

Depth range coarse Depth range fine Delay

Single/Dual Switch

Tx Rx

Timebase Generator Pulse Generator

Frequency Gain Reject Rectification

ReceiverAm

plifier

Clock

CRT Display

Focus Brightness X-shift Y-shift

PRF

Fig 4 The clock or ‘timer’ is the heart of the flaw detector. It feeds an electrical pulse to the Pulse

Generator and simultaneously to the Timebase Generator. This timer pulse causes the pulse

generator to send a short, high voltage pulse to the crystal and at the same time triggers the

timebase generator to begin to sweep the electron beam in the CRT tube from left to right

between the ‘X’ plates at a constant speed.

As soon as the high voltage pulse at the transmitter crystal is cut of, the crystal starts to

vibrate and an ultrasonic pulse propagates into the test piece. While this sound pulse travels

through the material, the CRT sweep continues to track the time as it moves towards the right

hand side of the display. Reflections from internal surfaces arrive at the receiver crystal,

generate a voltage in the crystal and this voltage is amplified and passed to the ‘Y’ plates

where it causes a vertical deflection of the electron beam proportional to the amplitude of the

received signal.

When the electron beam reaches the extreme right hand side of the CRT it flies back to the

left hand side and waits for the next trigger pulse from the clock. This whole sequence of

events takes place so quickly that we wouldn’t be able to see the trace. The clock repeats the

sequence many times a second and the result is a flicker free trace that increases in

brightness the more times we repeat the process each second. The number of trigger pulses

per second is known as the ‘Pulse Repetition Frequency’ (PRF), or ‘Pulse Repetition

Page 36: Ultrasonic Back-to-Basic

Rate’ (PRR). It is important that we allow enough time between pulses for all the multiple

echoes within the specimen to die away or we will the tail end of these echoes showing as

‘Ghosts’ at confusing positions on the timebase. For this reason the PRF is controlled by the

Depth Range Coarse control in the timebase generator circuit. However, some flaw detectors

have an additional manual control that the operator can use. Ghost echoes are most likely to

be encountered when testing fine-grained light alloy forgings that have very low attenuation of

sound.

The voltages developed in the receiver crystal are very small and need to be amplified. The

‘Amplifier’ circuit needs to be tuned to accept the frequency of the ultrasonic pulse and this

can be by way of switched bands for example, 1-3Mhz, 3-7MHz, 7-10MHz & 10-15MHz, or it

could be a wideband amplifier with the range 1-15MHz. If the former, the set will have a

‘Frequency’ selector switch that should be switched to the appropriate band for the probe in

use just as you would use the tuning dial to select the desired radio programme.

The ‘Gain’ or ‘Sensitivity’ control allows the amplification to be increased or decreased

depending on the strength of the received signals much like the volume control on a radio.

The Gain control is usually calibrated in decibels (dB) and is sometimes called the

‘Attenuator’. Strictly speaking, an attenuator should be calibrated such that increasing the dB

reduces the signal amplitude, but this is seldom the case over recent years. The ‘Bel’ is a unit

that is commonly used in electronics to compare the ratio between two power or voltage

values and is a logarithmic unit so that large ratios can be expressed concisely. The intensity

of sound in a received pulse is a measure of the power or energy in that pulse, and that

mechanical energy is converted into electrical energy by the piezoelectric crystal. If the power

increases from P1 to P2, then the gain can be expressed as: -

BelsPP

LogGain1

210=

However, the Bel is too large a unit for the values we shall encounter in ultrasonics and so we

use a unit of one tenth of a Bel or decibel. The equation then becomes: -

dBPP

LogGain1

21010=

The CRT measures voltage and electrical power is proportional to the square of the voltage: -

dBVV

LogGain2

1

21010 ⎟⎟

⎞⎜⎜⎝

⎛= And, removing the brackets: -

dBVVLogGain

1

21020=

Page 37: Ultrasonic Back-to-Basic

The height of a signal on the CRT is proportional to the voltage applied to the ‘Y’ plates and

so we can change the equation so that it is in terms of signal height: -

dBHHLogGain

1

21020=

Example 10

Calculate the gain ratio in dB between a signal that is 60% full screen height and one that is

only 30% full screen height.

dBLogGain306020 10=

dBLogGain 220 10=

dB.xGain 3010020=

dB.Gain 026=

When we measure depth or thickness from the timebase, we use the left hand flank of the

signal on the screen. Sometimes surface roughness, material grain size, or electronic ‘noise’

create noise signals (grass) that obscure the point where the flank meets the timebase and it

is difficult to make the correct reading. In these circumstances, we can use the ‘Suppression’ or ‘Reject’ control to remove the grass, much as we use a tone control on a radio to cut out

‘hiss’. Because this control can also cut out small relevant signals and make the gain non

linear, a warning light comes on when the control is in use.

The last feature that we need to consider in the amplifier circuit is the one that controls the

degree of rectification and smoothing of the pulse. The received signals are, of course, a few

cycles of alternating voltage. We can display these as they are – ‘Unrectified’ – but it is not

so easy to measure amplitude directly from the screen. It is more usual to display these

signals as ‘Rectified’ and smoothed signals in which the negative half cycles are inverted

and the signal envelope smoothed out. On some equipment, we may also have the choice to

only display the ‘Positive’, or ‘Negative’ half cycles and this may give a sharper flank to the

signal. Figure 5 illustrates the four conditions, but unsmoothed to illustrate the principle.

Unrectified Full Positive Negative

Fig. 5

Page 38: Ultrasonic Back-to-Basic

The ‘Timebase’ circuit controls the sweep speed and delay functions. The sweep speed will

determines the thickness range that can be displayed on the CRT. A high sweep speed (fast

timebase) may only allow a return path from a 10mm thickness in the test piece and at the

other extreme, a low sweep speed (slow timebase) may allow a return path from 5 metres or

more. Two controls achieve the desired thickness range, the ‘Coarse Depth’ or ‘Range’ control switches the range in steps (10mm, 50mm, 100mm, 500mm, 1m & 5m for example)

and the ‘Fine Depth’ or ‘Range’ control is a continuously variable control that allows fine

adjustment during calibration to allow for the specific material velocity. The fine depth range

control is sometimes labelled ‘Material’ or ‘Velocity’.

There are times when we don’t want the timebase generator to begin the sweep when the

crystal is pulsed. For instance, when we are carrying out an immersion test we want the

timebase to start when the sound enters the specimen so that the left hand end of the

timebase represents the top surface of the test piece. Another example might be when we are

testing a long shaft and we want to look in more detail at, say, the last 200mm of the shaft. In

either case, we can delay the start of the sweep with the ‘Delay’ control.

The last component to consider is the display module, the CRT. The image created by the

electron beam (the trace) must be displayed so that the baseline is aligned with the graticule,

extends beyond the left and right hand ends of the graticule, is bright enough to see in the

test environment and is in focus. There are four controls for these functions, the ‘X-shift’ and

‘Y-shift’ controls position the trace, the ‘Brightness’ control can be adjusted for indoor or

outdoor viewing, and the ‘Focus’ control sharpens the trace. On many flaw detectors, only

the focus and brightness controls are provided for operator adjustment.

Digital flaw detectors provide the same PRF, Amplifier and Timebase functions but these are

usually controlled using a combination of menu selection and so called ‘Smart Knobs’ through the controlling central processing unit (CPU). Figure 6 is a representative block

diagram for a digital instrument.

One of the real advantages of the digital instruments is the facility to store calibrations for a

number of inspection procedures and probes, to store whole traces complete with the

calibration data for each trace and to create databases to store thickness readings. Because

the instruments are based on computer technology, it is possible to connect the flaw detector

to a PC through a serial cable and download stored data, for reporting purposes.

The LCD display also has advantages over the CRT. It consumes less power than the CRT; it

can be backlit for viewing in low light conditions and at the same time is easy to see without

backlighting in daylight. In difficult conditions, the trace can be ‘Frozen’ so that the operator

can move to a more comfortable position before reading the timebase.

Page 39: Ultrasonic Back-to-Basic

LCD Display

Display Control

Controller

Acquisition & Digitiser

CPU

Tx Rx

Pulser/Receiver Attenuator Rectification & Filtration

Fig 6

Many flaw detectors, both analogue and digital, have gating circuits that allow signals to be

monitored by the instrument and the output used to trigger audible or visual alarms, or to be

connected to chart recorders or computers. The monitor gates may be displayed in one of two

ways. The timebase may be raised over the gate distance as shown in figure 7, or a separate

‘bar’ may be used as shown in figure 8.

0 1 2 3 4 5 6 7 8 9 10

Gate

0 1 2 3 4 5 6 7 8 9 10

Gate

Fig 7 Fig 8

Page 40: Ultrasonic Back-to-Basic

There are four main functions controlling the gate, these are: -

Gate Start Gate Width Gate Level or Threshold Gate sense (Rising or Falling Signal)

The gate ‘start’ control positions the left hand edge of the gate, the first depth that you want

to start monitoring. The gate ‘width’ control then allows you to set the right hand edge of the

gate, the last depth that you want to monitor. Any signal within that depth range is said to be

‘in the gate’. You may only want signals exceeding a predetermined amplitude to ‘trigger’ the

gate alarm and you do this using the gate ‘level’ or ‘threshold’ control. For those gates that

look like figure 6.7, you set a signal in the gate at the desired amplitude, and adjust the

‘threshold’ until the alarm just triggers. For those gates that look like figure 6.8, you simply

adjust the gate ‘level’ control until the gate is at the desired screen height. For some

inspections, such as when you are using the ‘through transmission’ technique, you may wish

to monitor for a decrease in signal amplitude. The gate sense can be changed using the

‘sense’ control. When ‘falling signal’ has been selected, the alarm does not trigger as long as

there is a signal in the gate that exceeds the threshold level. Instead, the alarm operates as

soon as the signal drops below the gate threshold.

Some flaw detectors have more than one gate. Two gates can be used in several ways; one

can monitor backwall echo amplitude (falling signal) and one can be used to monitor part of

the timebase for discontinuities (rising signal). The two gates can be used to monitor

consecutive backwall echoes and the difference (gate 1 minus gate 2) can be output as the

thickness of the object. The ‘menu’ of a digital flaw detector may allow you the choice to

monitor either signal amplitude or ‘time of flight’ (depth). This is also possible with some

analogue flaw detectors by the appropriate pin selection on the output connecting lead.

Generally, the voltage range for the output signal is about 0V to 5V; this means that the

vertical or horizontal (amplitude or timebase) scales of the display will be proportional to the

output range. If monitoring and recording amplitude, for example, a full screen echo height will

output 5V and a half screen height signal will output 2.5V.

Page 41: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 7. THE ULTRASONIC BEAM

The beam of sound waves emerging from an ultrasonic probe is rather like the beam of light from

a torch. The beam will spread out into an elongated cone shape, and the further away you go from

the source, the weaker will be the beam. So in order to know just how this beam affects our in-

spection, we need to study the shape of the beam in detail, and to study the changes in intensity of

the beam along its axis and across the beam.

As a general principle, we have said that the beam gets weaker as we get further from the

transducer. This weakening, or decrease in intensity represents a loss of energy; we say that the

beam is attenuated as it progresses through a material. The sound beam suffers this attenuation for

the following reasons: -

ABSORPTION - of the energy due to moving the vibrating molecules

SCATTER - of sound waves reflecting from the grain boundaries

INTERFERENCE EFFECTS - close to the transducer

BEAM SPREAD - the energy spreads over a larger area with distance

The amount of energy lost through ‘Absorption’ depends upon the elastic properties of the material

being tested so that steel and aluminium have less absorption than lead, or Perspex. ‘Scatter’ also

depends upon the material being tested, the larger the grain size, the greater the scatter (see figure

1). Forged and rolled materials generally give less scatter than castings or forgings. Heat treatment

may reduce grain size and therefore reduce scatter, making testing easier. Faced with a material

that presents either, or both, high absorption and scatter, you have to resort to a lower test

frequency to overcome the problem. We can either say attenuation (absorption and scatter)

decreases as test frequency decreases, or penetration increases, as frequency decreases.

Fig. 1

Page 42: Ultrasonic Back-to-Basic

INTERFERENCE EFFECTS

Point Source: - If we consider a point source of sound energy, then the disturbance (sound

wave) will radiate outwards from the point in an ever increasing circle, just like the ripples on a pond

spreading out when you drop a stone into it. So sound radiates in all directions from a

point source. (figure 2).

It c

dis

a li

dis

Sound wave expanding outwards from point source

Positive peak wave front

Negative peak wave front

Fig. 2

Finite source: - Our transducer, however, is not a point source, but a plate of

piezoelectric material of finite dimensions. In order to appreciate the way in which sound

spreads out from a finite source, and to help us understand interference effects we will

use Huyghens principle, Huyghens said that you can consider a finite source to be made

up of an infinite number of point sources. When you energise the transducer, sound will

radiate out from each of these point sources, just as it did for the stone dropping into the

pond. Figure 3 shows sound radiating from just one of these point sources and figure 4

shows sound radiating from several point sources.

Time t1

New wave front

Fig. 4 Fig. 3

an be seen from figure 4 that a short time (t1) after the finite source has been energised, the

turbances from each of the point sources will have moved outwards by the same amount. Along

ne equal to the radius of the small circles, running parallel to the face of the transducer, these

turbances re-enforce each other to produce a wave front moving out from the transducer.

Page 43: Ultrasonic Back-to-Basic

Notice also, a little energy ‘diffracts’ around the edge of the transducer and is ‘lost’. A little while

later (t2), sound from each point source will have travelled a little further and reinforce at a new

distance in front of the transducer, thus the sound wave progresses from the source (figure 5).

Time t2

Fig. 5

This wave front may represent the initial expansion of the transducer as it starts to vibrate (a

positive going half cycle). It will tend to push particles of the material away from the source.

Shortly afterwards, the transducer will contract as part of its vibration, and a wave front, drawing

particles into the source (a negative going half cycle) will follow on behind the initial wave front,

followed by another push, then another pull and so on.

In the third article in this series, we discussed refraction of the beam as an angled incident wave

meets an interface. The bottom edge of the beam reaches the interface first and takes up the new

velocity. We can use Huyghens principle to explain what happens. As each point along the

incident wave front reaches the interface, each in turn takes the new velocity and in the new

material, the line of initial wave fronts will determine the direction of the refracted beam. Similarly,

in the fifth article we mentioned phased array probes. The shape of the beam and beam angle will

be determined by the wave front for which the individual wave fronts are in phase.

Now consider a point reflector ‘P’ just in front of the probe centre. Let us consider how this

reflector is affected by just three of the point sources, one in the centre and one at each edge of

the transducer (figure 6).

We energise the source, and a split secon

gives it a push (figure 7). Notice that ene

This will take longer because P is further

P

Fig. 6

d later sound from the middle point source reaches P, and

rgy from the edges of the probe has not reached P yet.

from the edges than from the centre.

Page 44: Ultrasonic Back-to-Basic

P

Fig. 7

By the time sound from the edges of the transducer reaches P (figure 8) and tries to push P away

from the transducer, the energy from the centre may be on the opposite half cycle of vibration, and

be pulling P back towards the transducer. The resultant force acting on point P will be the vector

sum of the forces acting from all parts of the crystal. In our example, the result is that P doesn't

move at all (i.e. the sound intensity=0). The distance between the solid arc (positive peak) and the

dotted arc (negative peak) is half a wavelength. If a different frequency had been used, it may have

been that the second positive half cycle from the centre of the crystal reached point P at the same

time as those from the edges of the crystal. In that case, the forces would have reinforced and

point P would have been given an extra hard ‘push’.

P

Fig. 8

When two solid arcs cross, the forces from those two parts of the crystal are both ‘pushing’ at

the intersecting point and when two dotted arcs cross the forces from that part of the crystal

are both ‘pulling’ at the intersecting point. In both cases we call the effect ‘constructive interference’. When a solid arc cuts a dotted arc, the forces are in opposition and we call the

effect ‘destructive interference’. Of course point P will not always be exactly a multiple of

half wavelengths away from the center and the edges, and constructive interference happens

when the relevant point sits anywhere within the same half cycle. Destructive interference

happens when the relevant point is in dissimilar half cycles.

‘Interference’ occurs whenever energy arrives at different phase (wavelength) intervals at a

particular point. Whether the interference is constructive, or destructive, is determined by the path

difference between P and the centre, and P and the edges. As P gets further away from the front of

the transducer, this path difference becomes negligible compared to the wavelength (figure 9) and

interference problems become insignificant.

Page 45: Ultrasonic Back-to-Basic

P

Next half cycle

Initial wave fronts

Fig. 9

Variations in intensity due to interference effects occur for some distance in front of the

transducer, as we have just seen. This region is known as the ‘Near Field’ and the extent of

the near field, known as the near field distance can be calculated from: -

λ4

2DNF = Where, NF = Near Field Distance.

D = Crystal Diameter.

λ = Wavelength

Example 1.

Calculate the Near Field distance for a 10mm diameter, 5MHz crystal transmitting into steel

(Velocity 5960m/sec. ∴ λ = 1.192mm).

19214102

.xNF =

7684100.

NF =

NF = 21mm (Approx.)

This means that for this probe, in steel, we can expect fluctuations in intensity of sound for the

first 21mm of steel depth due to interference effects. As a result, it is unwise to rely solely on

amplitude as the criterion for acceptance or rejection of the part for discontinuities that are in

the near field region.

The last item on our list of factors affecting attenuation of the sound as it travels through a

material is the ‘Beam Spread’. Because the beam spreads out into a conical shape, intensity follows

the inverse square law just as it would for a beam of light or X-rays. If you double the distance from the

probe, the intensity drops to one quarter of its original value because of beam spread. Of course, it

will actually fall to less than a quarter, because we have to add any absorption, scatter losses to

the beam spread losses. We can now plot a graph of intensity against distance from the

Page 46: Ultrasonic Back-to-Basic

probe, to summarise the previous discussions. Figure 10 show amplitude on the vertical axis and

distance on the horizontal axis. Distance is shown in multiples of the near field distance.

0 1NF 2NF 3NF 4NF 6NF5NF

Am

plitu

de

Fig. 10

D

Fig. 11

θo Near Field

Far Field

Page 47: Ultrasonic Back-to-Basic

The beam profile shown in figure 11 is very much a ‘theoretical’ beam spread. Alongside there

are three ‘slices’ through the beam showing that the highest sound intensity is in the centre of

the beam. The sound gradually fades away towards the edge of the beam until there is no

sound left. It is often more convenient to define the beam to a theoretical edge where the

intensity of sound has fallen to one half (-6dB), or sometimes one tenth (-20dB) of the intensity

at the beam centre. We can consider three theoretical edges; one defining the absolute edge of

the beam, another defining the 6dB edge and the third defining the 20dB edge. These three

edges can be expressed mathematically: -

D.Sin λθ 221

2= Defines the absolute edge

D.Sin λθ 560

2= Defines the 6dB edge

D.Sin λθ 081

2= Defines the 20dB edge

It is often convenient to use the theoretical beam shape shown in figure 11 in order to explain

some concepts in ultrasonic flaw detection. However it is not good practice to use a

calculated beam shape for sizing discontinuities by one of the intensity drop methods. This is

because practical beam shapes seldom match the theoretical model closely enough.

Example 2.

Calculate the 20dB beam spread angle for a 5MHz compression wave in steel from a 10mm

diameter crystal.

D.Sin λθ 081

2=

101921081

2.x.Sin =

θ

10287361

2.Sin =

θ

12873602

.Sin =θ

o472

.=θ

o472 .x=θ

θ = 14.8o

Page 48: Ultrasonic Back-to-Basic

We have used three terms connected with the beam of sound in the test material, namely

‘Dead Zone’, ‘Near Field’, and ‘Far Field’. The dead zone is that part of the timebase

occupied by the initial pulse when using a single crystal contact probe. The near field is the

distance in the material that suffers from interference effects and the far field is the rest of the

beam beyond the near field. The trace shown in figure 12 is calibrated for 100mm of steel

return path using a single crystal 5MHz compression wave probe. The three zones are shown

on the trace.

0 1 2 3 4 5 6 7 8 9 10DeadZone

Near

Field Far Field

Fig. 12

Page 49: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 8. CALIBRATION & REFERENCE STANDARDS

As soon as one starts to carry out practical ultrasonics, it becomes apparent that neither the vertical nor

the horizontal scales of the display have any absolute meaning per se. The horizontal scale can be adjusted

to represent a great variety of different time intervals, and these, for a given material and velocity, can be

translated into depth values. The vertical scale gives an indication of the amplitude of signal being detected,

provided you know how much ‘Gain’ you are using, but it does not necessarily tell you much about the size

of defect causing that reflection. The safest way to get more information about the specimen from the

display is to compare signals from the specimen with those from specially machined blocks. These blocks

we normally classify under one of two headings, depending on the function of the block.

The term ‘Calibration Block’ is defined in British Standard BS 2704 as: - "A piece of material of

specified composition, heat treatment, geometric form, and surface finish, by means of which ultrasonic

equipment can be assessed and calibrated for the examination of material of the same general

composition." Therefore, a calibration block may be a simple step wedge in a particular material to allow the

timebase to be calibrated for accurate thickness measurement, or it may be a more complex block like the

A.2 block described in BS 2704 which allows calibration of timebase, plus calibration of probe index, angle,

resolution etc.

The second heading, ‘Reference Block’ is defined in BS2704 as: - "An aid to interpretation in the form of

a test piece of the same material, significant dimensions and shape as a particular object under

examination, but not necessarily containing natural or artificial defects". So, for example, a section of an

aircraft wing forging may be prepared as a reference block so that a technician may become familiar with

the standard signal patterns from the various changes in section and more easily recognise a defect

quickly when examining the component on an aircraft. More usually, the block would contain artificial

defects from which the sensitivity (gain) used in the test could be set.

In either case, the use of the block ensures that there is adequate timebase to display the reflecting

surfaces that are of interest and that the test is carried out at a reproducible sensitivity

CALIBRATION BLOCKS

The BS 2704 A.2 Calibration Block, also known as the International Institute of Welding (I.I.W.) block, or

‘V1 block’, is illustrated in Fig.1.The block can be used for the following assessments: -

- Calibration of the timebase in terms of thickness. - Assessment of Dead Zone.

- Checking linearity of the timebase. - Checking linearity of the Amplifier.

- Assessing overall sensitivity of probe and amplifier. - Checking Resolution.

- Determination of the angle of refraction. - Determination of probe index

- Determination of Beam Characteristics. - Finding the correct Zero Point.

Page 50: Ultrasonic Back-to-Basic

405060

607075

25.00300.00

100.

00

100.00R

Perspex insert

50mm diameter hole

1.5mm hole

Fig. 1

The A.2 block was derived from the original ‘Dutch Block’ designed by RTD Rotterdam and accepted

by IIW as the ‘IIW V1 Block’. In its original form, the deep slot at the center of the 100mm radius was

a scribed line and a 25mm radius slot was positioned as shown in Figure 2. This design is still used in

some parts of the world, and has the advantage that shear waves can be calibrated for ranges other

than multiples of 100mm. In all other respects it is the same as the A.2 block.

Scribed line 25mm radius slot

Fig. 2

The BS 2704 A.4 Calibration block, also known as the 'V2 block', is a more compact form of the 'V1 block'

suitable for site use, although somewhat less versatile in its functions. Figure 3 illustrates the A.4 block.

40 50 60

65 70

75.00

25.00R

50.00R

12.50

Fig. 3

Page 51: Ultrasonic Back-to-Basic

The Institute of Welding (I.O.W.) Beam Profile calibration block is designed primarily for beam profile

measurement and has four 1.5mm diameter side drilled holes giving eight depths from two scanning

surfaces. These can be examined by direct scan for probes of various angles, and at several more ranges

for each probe, using indirect scans by reflecting from the far surface. There are two series of five holes on

an inclined axis to measure shear wave probe resolution and to simulate an inclined discontinuity. The block

is illustrated in Figure 4.

13

2519

43

62

5056

22

75

w

REFERENCE BLOCKS

Area / Distance reference blocks are m

for defect sizes by reference to echo height.

of blocks contains the same diameter flat-bo

with 3/64” diameter flat-bottomed holes,

holes. The scanning depths can range from

blocks appropriate to your range of work. F

depth, 5/64" flat bottomed hole).

A = Scan depth

B = Hole diameter

Side vie

32

50

305

Plan view

Fig. 4

ainly used for setting sensitivity levels and accept/ reject levels

Blocks are produced in a range of scanning depths and each set

ttomed hole in each block. There are three sets of blocks, a set

a set with 5/64" diameter holes and one with 8/64" diameter

½” to 22”, but at shop floor level, you would only have the few

igure 5 shows a typical block, in this case a 3 x 5 block (3" scan

A

B

Fig. 5

Page 52: Ultrasonic Back-to-Basic

Distance/Amplitude Correction (DAC) reference blocks are made from the same thickness and grade of

material as the work piece. They contain an artificial flaw (a side-drilled hole). The change of echo height

with changes in scanning distance (multiple skips) is noted and plotted on the display as a "DAC" curve so

that a signal amplitude can be specified to cover all depths within the working range for reporting,

acceptance, or rejection purposes. Figure 6 shows a typical ASME DAC block and Figure 7 shows a DAC

curve.

1/4 T

1.5" or T

4" Always

1.5" Always

1.5" Always

Fig. 6

0 1 2 3 4 5 6 7 8 9 10

DAC Curve

Fig. 7 Figure 8 shows an example of a reference block for the examination of a lug in a light alloy structure such as an aircraft fitting. It is made from the same material as the actual part to be inspected, will

Page 53: Ultrasonic Back-to-Basic

have the same surface finish and, in this case contains an artificial defect to aid the setting of sensitivity and to help with identification of signals during interpretation.

35o

Fig. 8 References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury

Page 54: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 9. COMPRESSION WAVE TECHNIQUES

CALIBRATION OF TIMEBASE

The important thing to remember when calibrating the timebase for compression waves is that the left

hand end of the timebase (Zero) must exactly correspond to the entry surface of the beam and the

right hand end represents a known thickness in the material being tested. The exceptions to this rule

are those occasions when you are using delay to expand some distant portion of the material, or

when you are using a multiple echo technique and only noting the decay pattern. For single crystal

probes, the initial pulse contains two elements; the applied square wave voltage pulse and the ringing

of the crystal. The end of the applied voltage pulse represents the top surface at which time the

crystal ringing starts. Unfortunately, the amplitude of this part of the initial pulse is so large that it is

not possible to identify the point at which the ringing starts, nor is it possible to tell from the timebase

line.

There is a similar problem when we calibrate using a twin crystal probe because the initial pulse is at

the start of the Perspex delay line and the sound enters the work piece sometime later. In any case,

because the amplifier is deliberately isolated from the transmitter crystal, there is no signal to mark

the entry surface. Our calibration procedure, whether for a single or twin crystal probe must find some

other way to identify the entry surface or true zero. The most common way to achieve this is to use

two echoes that are a known distance apart, to set one at timebase zero and the other at the right

hand end (10) of the timebase. We do this in the following way: -

- Use the delay control to position the first backwall echo from the desired range on our

calibration block to zero.

- Then use the depth range controls to position the second backwall echo to 10 on the

timebase.

- This may also move the echo from the zero position and so we need to check and adjust this

with delay again.

- These two controls are used alternately until the two echoes are exactly on 0 and 10. We now

know that the timebase is exactly equal to the calibration thickness

- Once we are happy that we have that exact range on the timebase, we lock the depth

controls.

- We then use delay to move the first backwall echo to the right until it is exactly on 10. If the

timebase is exactly a known range and the first backwall echo is on 10; then zero must

coincide with the entry surface.

Figures 1 to 3 illustrate the procedure for calibrating the timebase for 100mm of steel on the A2

calibration block.

Page 55: Ultrasonic Back-to-Basic

405060

0 1 2 3 4 5 6 7 8 9 10

2nd BWE 1st BWE

Depth Delay

Fig. 1

0 1 2 3 4 5 6 7 8 9 10

1st BWE

2nd BWE

DepthDelay

Fig. 2

0 1 2 3 4 5 6 7 8 9 10

1st BWE

Delay only

Fig. 3 In this example we know that the first and second backwall echoes represent 100mm and 200mm of

steel path time because the A2 block is 100mm wide at this probe position. Therefore figure 2

Page 56: Ultrasonic Back-to-Basic

represents exactly 100mm timebase. This timebase remains constant as long as we do not alter the

depth control, so figure 3 represents zero to 100mm exactly.

CHOICE OF COMPRESSION WAVE PROBES

TWIN CRYSTAL PROBES

For conventional techniques twin crystal probes are generally used on thicknesses below 50mm. They

are also in general use for high temperature thickness measurement, where a thermal insulating

material is used instead of Perspex, to protect the crystals. Because the transmitter section of the

probe is isolated from the amplifier, there is no dead zone and so reflections from surfaces close to the

probe can be identified. However since the Perspex shoe absorbs some sound, less damping is used

and the pulses are longer so that resolution is generally poorer than with single probes.

SINGLE CRYSTAL PROBES. Single crystal probes are generally used on thicknesses in excess of 50mm. They are also used below

50mm if resolution is an important factor, since single crystal probes usually have shorter pulse

lengths than twin crystal probes. However, for conventional techniques they can only be used when

the transmission noise (Dead Zone) does not encroach upon the useful part of the timebase for that

job. As a guide, you can expect the shortest Dead Zone from high frequency, heavily damped probes.

PROBES FOR MULTIPLE ECHO TECHNIQUE

These are usually single crystal probes, although in some cases twin crystals can be used. When

dealing with thin walled material it is possible to get resonance and anti-resonance conditions that will

affect the decay pattern and may give false indications. This can be avoided if you choose a probe

frequency such that the plate thickness is more than 1.5 x the wavelength of the compression wave in

the specimen material, and a pulse length that is not more than 3 cycles.

THICKNESS MEASUREMENT

One of the most important uses of ultrasonics is that of thickness measurement. It is particularly

useful because it is relatively quick, simple and accurate, and access to only one surface of the

specimen is required. There are many types of equipment and techniques made exclusively for

thickness measurement. It is not intended to deal with all of them here. We will only discuss the use of

the pulse echo system with an A-scan display.

A-SCAN RECTIFIED DISPLAY

This is the most common display presentation for ultrasonic flaw detection equipment. In the sixth

article we described the display for an unrectified trace and various types of rectification.

a) CALIBRATION

The basic calibration of the timebase should be carried out to ensure proper positioning of the

zero and backwall echo as described above. The calibration block should be made of the same

material as the work piece.

Page 57: Ultrasonic Back-to-Basic

For best results the range chosen for calibration should be the shortest range which allows the

first back wall echo to be displayed. For example, if the nominal wall thickness of the work

piece is 9mm. and your flaw detector is capable of showing 10mm across the full graticule,

then the 10mm range should be used. Since the graticule of most flaw detectors can be

divided into 100 small units it follows that a timebase calibrated such that those 100 units

represent 10mm gives you a reading accuracy of 0.1mm per division. If on the other hand you

calibrate such that 100 units represent 25mm, the reading accuracy is 0.25mm per division.

b) AMPLITUDE (GAIN SETTING)

The amplitude of the calibration echo and the amplitudes of thickness echoes made on the

work piece should be adjusted to the same predetermined amplitude. This is normally

between 1/3 and 1/2 full screen heights.

c) READING THE THICKNESS (SINGLE ECHO).

The specimen thickness is determined from the left hand edge of the backwall echo. This is

normally a steep sloping line. If a small half cycle appears at the left hand edge of the signal

that was not present during calibration, this may be removed by inserting a small amount of

suppression or by choosing ‘positive’ or ‘negative’ rectification. (See figure 4).

3 4 5 6 7

Extra half cycle

Fig. 4

d) READING THE THICKNESS (MULTIPLE EC

If the specimen thickness and calibrated rang

most accurate result can be obtained by readi

and dividing the answer by the number of bac

fifth backwall echo shows at 22 mm. so the true

single crystal probe has been used and the ini

echo.

3 4 5 6 7

After suppression or rectification change

HOES)

e are such that multiple echoes are produced, the

ng the thickness of the last multiple echo displayed

kwall echoes. In the example shown in figure 5, the

thickness is 22 divided by 5 = 4.4 mm. In this case, a

tial pulse is obscuring the start of the first backwall

Page 58: Ultrasonic Back-to-Basic

Sometimes the initia

back echoes. Figure

must be taken to asses

e) USE OF TIMEBASE

Apart from its use

probes, "Delay" ca

example, you may w

thickness is 80 mm

mm of that materia

timebase so that 10

scale represents 0.

the third backwall e

100 scale divisions.

first back echo fro

timebase range.

l pulse obscures the entire first backwall echo and maybe all or part of other

6 shows the same thickness but with the first two back echoes obscured. Care s the number of echoes that have been obscured.

t

n

.

0

2

c

T

m

0 5 10 15 20 25

Fig. 5

DELAY

o correct for Pe

be used as a

ant to accurately

If you calibrate t

l, each small di

scale divisions r

5 mm instead of

ho from the calib

he timebase wo

the work piec

0 5 10 15 20 25

Fig. 6

rspex path distance in twin crystal compression wave

n aid to more accurate thickness measurement. For

measure the thickness of a component whose nominal

he timebase so that 100 scale divisions represents 100

vision represents 1 mm. If instead, you calibrate the

epresent 25 mm of the test material each division on the

1 mm. The delay control can then be adjusted so that

ration block is set at O, and the fourth backwall echo at

uld represent a thickness range of 75mm to 100mm. The

e (80mm) will appear at approximately 1/5th of the

Page 59: Ultrasonic Back-to-Basic

A-SCAN UNRECTIFIED DISPLAY.

There are occasions in thickness measurement, particularly if the scanning surface is rough, when

a lot of unwanted signals, "noise " or "grass" appear on the CRT and make it difficult to determine

the point at which the back echo starts. If the ultrasonic set allows an unrectified trace to be

selected, then measurements can be made using the tip of a particular down going half cycle instead

of using the point at which the signal first leaves the timebase.

a) Firstly, let us identify our measuring point. Figure 7 shows a back echo from the 25mm range on

the V1 block with the timebase calibrated for 50mm using the conventional rectified display. The

presentation has then been changed to ‘unrectified’ and the vertical or 'Y' shift used to raise the

timebase to a level between 1/3 and 175th full screen height. Gain has been adjusted so that the

peak of the longest down going half cycle just meets the graticule. In this case, it is the second

half cycle that is the longest, and we will use the second half cycle as our measuring point.

(Note that sometimes a back echo from the work piece may show the 1st or 3rd half cycle as

the longest - despite this, if you calibrate on the second half cycle you then always measure

from the second half cycle even if it is not still the longest,)

0 1 2 3 4 5 6 7 8 9 10

Fig.7

b) Having identified the half cycle that you are going to use, you calibrate the timebase so that

this part of the signal coincides with the correct point of the graticule. In the case shown in

figure 7, if we wished to calibrate for 50 mm we would use "delay" to move the second half

cycle from 5.15 to 5.0 divisions and check that the second half cycle of the second back echo

coincides with 10.0 divisions (see figure 8).

Page 60: Ultrasonic Back-to-Basic

0 1 2 3 4 5 6 7 8 9 10

Fig. 8

c) Calibration for other timebase ranges would be done in the conventional way but using the

second half cycle instead of the left hand edge of the pulse, as your measuring point.

LAMINATION TESTING

Lamination testing of plates and pipes that are to be welded or machined is a very common NDT

task. It is also a simple application of compression waves in ultrasonic flaw detection, but one that

might give some problems when examining thinner samples. STANDARD PROCEDURE

a) Calibrate the timebase to allow at least two backwall echoes to be displayed.

b) Place probe on the work piece and adjust the gain controls so that the second backwall echo is

at full screen height.

c) Scan the work piece looking for lamination indications that will generally show up near half

specimen thickness together with a reduction in back echo amplitude. In some cases, a

reduction in the amplitude of the second back echo may be noticed without a lamination

signal being present. Care must be taken to ensure that this reduction in amplitude is not due

to poor couplant or surface conditions.

MULTIPLE ECHO TECHNIQUE

Lamination testing of plate or pipe less than 10 mm. in wall thickness may be difficult using the

standard procedure because multiple echoes are so close together that it becomes impossible to pick

out lamination signals between backwall echoes. In such cases, we can use a technique called the

"multiple echo technique" using a single crystal compression wave probe. The procedure is as

follows: -

a) Place the probe on a lamination free portion of the work or calibration piece.

Page 61: Ultrasonic Back-to-Basic

b) Adjust the timebase and gain controls to obtain a considerable number of multiple echoes in a

decay pattern over the first half of the time base. A typical example is shown in figure 9.

c) Scan the work piece, the presence of laminations will be indicated by a collapse of the decay

pattern such as the one shown in figure 10. The collapse occurs because each of the many

multiple echoes is closer to its neighbour in the presence of a lamination.

Fig. 9

0 1 2 3 4 5 6 7 8 9 10

Unlaminated plate

0 1 2 3 4 5 6 7 8 9 10

Laminated plate

Fig. 10

EXAMINATION OF BRAZED AND BONDED JOINTS

Compression waves can also be used for the detection of areas of non-adhesion in brazed or

bonded (glued) joints,

Page 62: Ultrasonic Back-to-Basic

a) BRAZED JOINTS

If the wall thicknesses permit clear separation between back wall echoes, brazed joints can be

examined using the standard procedure for lamination checking. However, since the braze metal

separating the two brazed walls will have a slightly different acoustic impedance to that of the

parent metal, a small interface echo may be present for a good braze. The technique, therefore, is

to look for an increase in this interface echo amplitude. Figure 11 shows the type of indication

when the unbrazed portion is smaller than the beam diameter. If the two brazed walls are too thin to

permit clear back echoes, a multiple echo technique can be used as described above.

Brazed joint Good braze Bad braze

Fig. 11

b) BONDED JOINTS

These may include metal-to-metal glued joints and metal to non-metal glued joints (e.g. rubber

blocks bonded to steel plates). The technique used is a multiple echo technique. Each time the

pulse reaches a bonded interface; a portion of the energy will be transmitted into the bonded layer

and absorbed. Each time a pulse reaches an unbonded layer, all the energy will be reflected. If we

look at the multiple echo pattern for a good bond, the decay will be relatively short because of the

energy loss at each multiple echo into the bond. However, for an unbonded layer each multiple

echo will be slightly bigger because there is no interface loss, and the decay pattern will be

significantly longer. Figure 12 shows a good bond (probe 1) and poor bond (probe 2).

1 2

Metal

Rubber

Decay pattern for good bond Decay pattern for bad bond

Fig. 12

References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury

Page 63: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 10. SHEAR WAVE TECHNIQUES

Shear waves at various angles of refraction between 35o and 80o are used to locate defects whose

orientation is not suitable for detection by compression wave techniques. Some defects, of course,

have volume and their shape enables them to be detected by both compression and angled shear waves.

In this chapter, however, we will be dealing with planar defects whose orientation is such that only

angled shear waves can be used.

Because the beam is travelling through the test piece at a refracted angle other than perpendicular,

we need to distinguish between the beam path length to a discontinuity and its depth below the test

surface. When we encounter a signal, we can measure the beam path length (range) from the

timebase, but we may want to calculate how far in front of the probe (horizontal distance) and how

far below the surface the reflector is located. It is also important when using shear waves to know

where along your probe the beam enters the specimen (beam index). Knowing the beam index

position relative to some datum on the specimen, and the exact beam angle allows you to calculate

the horizontal and vertical distances. There are standard terms for various distances when using

shear waves and these are illustrated in figure 1.

Index.

D

C A

HSD

FSD

θ

FSD = Full skip distance

HSD = Half skip distance

θ = Beam angle

ABCD= Beam path length

AB = Half skip BPL

B Fig. 1

Full skip and half skip distances are measured along the top surface and beam path length (BPL),

along the beam centre. To calculate these, knowing specimen thickness (t) and probe angle (θ) use

the following formulae: -

a) HALF SKIP DISTANCE = t x Tanθ

b) FULL SKIP DISTANCE = 2 x t x Tan 0

c) HALF SKIP BPL = θCos

t

a) FULL SKIP BPL =θCos

t2

Page 64: Ultrasonic Back-to-Basic

Example 1

Calculate the Full skip distance for a 40o shear wave beam in a 20mm thick steel plate.

θxtxTanFSD 2=

40202 xTanxFSD =

83910202 .xxFSD =

FSD = 33.564mm

ESTABLISHING THE TRUE PROBE BEAM INDEX

We need to find the exact beam index for any shear wave probe before measuring the beam angle.

This is because the beam index may not be the one marked on a new probe – it may be a millimetre

or so before or after the marked index. Manufacturers use a standard drawing to make probes, but

the velocity of sound in Perspex varies from batch to batch, and with temperature. Also the beam

index and probe angle change as the probe wedge or ‘shoe’ becomes worn with use. So, the

establishing of beam index and angle will be routine checks throughout the life of the probe. Finding

the beam index is a simple procedure carried out on the A2 or A4 calibration block.

The probe is positioned close to the edge of the calibration block and beaming towards the 100mm

radius (A2 block) or the 25mm or 50mm radius (A4 block) as shown in figure 2 (A2 block) and figure 3

(A4block).

40 50 60

65 70

45

60

45

Fig. 3 Fig. 2 The probe is moved backwards and forwards about the centre mark of the radius with the probe kept

parallel to the edge of the block. As the probe moves, the signal will rise to a maximum and then fall

again as shown in figure 4. When the signal reaches the maximum amplitude, the beam centre is

meeting the tangent to the radius at right angles. This happens when the beam centre is entering the

block at the centre of the radius. The true beam index is now in line with the centre mark of the radius

being scanned. If this does not coincide with the beam index marked on the probe, you would then

Page 65: Ultrasonic Back-to-Basic

either mark the true index on the probe body, or, if the probe body has a millimetre scale, make a

note of the true position in front or behind the marked index.

3 4 5 6 7

Fig. 4 MEASURING THE TRUE BEAM ANGLE

Once the true beam index is known, the true beam angle can be measured on the A2 or A4

calibration block. The nominal probe angle is marked on the probe and is the refracted angle for steel

unless identified for another material. A 45o shear wave probe made for aluminium would be marked

‘45AL’ and for copper ‘45CU’. The actual angle for a new probe may be plus or minus two degrees

from the nominal angle because of the batch velocity variations in Perspex, and will change with

wear. Most of us have an inherent tendency to wear the probe in a particular way, just as we do for

shoes. We may wear the heel of the probe down and so increase the actual angle, or wear the toe

and decrease the actual angle. For this reason the beam angle measurement is also a routine probe

check. If the probe is worn down towards one edge, the beam will be thrown off towards that side and

the condition is called ‘squint’.

Beam angle is measured on the A2 block by aiming the beam at the 45mm diameter hole and on the

A4 block at the small hole. The probe is positioned on the block at a point near the nominal angle and

the gain adjusted to give a signal amplitude of about 50% full screen height. As the probe is moved

forward and back, the signal rises and falls just as it did when finding the beam index. When the

signal reaches its maximum amplitude, the beam centre is aimed at the centre of the hole and the

beam is hitting the tangent to the hole at right angles. The true beam angle can be read against the

true beam index from the graticule on the calibration block. The example shown in figure 5 has the

beam index opposite an angle of about 43o and the nominal angle is 45o. With this probe, we would

have to use 43o in our distance calculations and for defect sizing.

Page 66: Ultrasonic Back-to-Basic

6070

405060

45

Fig. 5 CALIBRATION OF TIMEBASE

The method of calibration of the timebase for shear waves depends on the purpose of the inspection.

If the inspection were to be volumetric, looking for any discontinuities within the scanned volume of

the test piece, then we would calibrate for a suitable timebase range at shear wave velocity. On the

other hand, if the purpose is to look for a specific discontinuity such as a fatigue crack, in a predicted

location, we may well use a ‘Skip’ method or a ‘Reference Block’ method. The calibration for a

known range will be dealt with first, using the A2 block and then the A4 block.

USING THE A2 BLOCK

- Place the probe on the A2 block as shown in figure 6.

- Obtain a maximum echo from the 100mm radius.

- Adjust the gain control to peak the signal at about 80% full screen height.

- Use the delay control to position the 100mm signal at zero on the timebase.

- Use the depth controls to place the second reflection from the 100mm radius at ten on the

timebase.

- Check that the left hand edges of the two signals are exactly at zero and ten.

- Lock the depth controls

- Use delay to move the first signal from zero to ten.

- The time base is now calibrated for 100mm at shear wave velocity, and zero represents the

top surface entry point below the beam index.

45

0 1 2 3 4 5 6 7 8 9 10

Delay Depth

100mm echo 200mm echo

Fig. 6

Page 67: Ultrasonic Back-to-Basic

Sometimes you may see part of the initial (transmission) pulse around the zero, this will depend on

the pulse length and gain setting as shown in figure 7.

0 1 2 3 4 5 6 7 8 9 10

100mm echo

Residual initial pulse

Delay

Fig. 7 The slot that marks the 100mm radius on the A2 block is about 4mm deep so that when the probe is

aligned with the edge of the block, the slot makes a corner for the returning echo to reflect part of the

energy back to the 100mm radius. This is why it is possible to obtain repeat echoes from the radius. If

the slot were not there, the reflected energy from the first returned signal would reflect to the rear of

the probe. Figure 8 shows an exaggerated view of the beam path to illustrate the ‘corner’ effect.

45

4mm deep slotExagerated beam path

Fig. 8 USING THE A4 BLOCK

To calibrate for 100mm using the A4 block: -

- Place the probe on the A4 block as shown in figure 9.

- Obtain a maximum echo from the 25mm radius.

Page 68: Ultrasonic Back-to-Basic

- Adjust the gain control to peak the signal at about 80% full screen height.

- Use the delay control to position the 25mm signal at 2.5 on the timebase.

- Use the depth controls to place the second reflection (from the 50mm radius slot) at ten on

the timebase.

- Check that the left hand edges of the two signals are exactly at 2.5 and 10.

- Lock the depth controls

- The time base is now calibrated for 100mm at shear wave velocity, and zero represents the

top surface entry point below the beam index.

40 50 60

65 70

45

0 1 2 3 4 5 6 7 8 9 10

25mm echo 100mm echo

DelayDepth

Fig. 9

The sound path in figure 9 shows the first echo from the 25mm radius and then the echo from the

50mm radius after reflecting at the scanning surface down to the 25mm radius and back. The total

return path is: -

25mm + 50mm + 25mm = 100mm.

Facing the 25mm radius on the A4 block, signals will arrive at 25, 100, 175, 250mm and so on,

incrementing by 75mm each time. If the probe is turned around to face the 50mm radius, signals will

arrive at 50, 125, 200, 275mm and so on, again incrementing by 75mm.

CALIBRATION USING THE ‘SKIP’ METHOD

If the purpose of the inspection is to detect surface breaking flaws at the bottom surface or top

surface (typical of fatigue cracks), we know that the echoes will arrive at exactly the half skip or full

skip beam path lengths. We could calibrate the timebase for an exact range using one of the methods

described above and calculate the beam path lengths for half and full skip using the formulae. We

would then know exactly where to look on the timebase for the two conditions. We do use this method

to carry out the critical root scan in weld inspection.

However, in many cases there is a quicker and simpler method. Using a piece of plate of the same

wall thickness as the item to be inspected we can point the probe at the end surface (position 1) and

scan back as shown in figure 10 until we see the echo from the bottom corner (position 2).

Page 69: Ultrasonic Back-to-Basic

Scan The signal will rise

timebase and gain

we would adjust t

usually about ’4’. W

is seen (position 3

shows a trace with

over the two critic

display all the time

Another point to no

‘8’ (twice ‘4’). This

the exact timebas

interested to find o

If the plate to be in

use the corners on

45 4545

Full skipHalf

skip

1 2 3

Fig. 10

to a maximum as the centre of the beam moves into the corner. We can adjust the

to make sure that we can see that maximum point. As the maximum is reached,

he timebase range to position the signal at some convenient part of the trace,

e would then continue moving the probe backwards until the top corner reflection

). As this signal maximises, we note its position along the timebase. Figure 11

the half skip and full skip positions marked, and in this example, gates positioned

al locations so that the operator can listen for the alarm rather than watch the

.

te from figure 11 is that the position for full skip is at ‘9’ on the timebase and not at

means that timebase zero is not the top surface, and furthermore, we don’t know

e range. However, for this inspection it doesn’t matter because we are only

ut whether or not there is a bottom or top surface breaking ‘corner’.

0 1 2 3 4 5 6 7 8 9 10

Gates

Half skip zone Full skip zone

Fig. 11

spected has accessible edges, you don’t need a calibration plate because you can

the test piece to set up the two positions on the timebase. However, you have to

Page 70: Ultrasonic Back-to-Basic

be sure that there are no laminations in the beam path because these might reflect the beam back to

the top without reaching the bottom. It is easy to check whether the signal is coming from the

anticipated corner because a shear wave meeting an interface at an oblique angle is easy to ‘damp’.

If you put an oily finger on the expected reflecting corner, the signal will be seen to reduce in

amplitude significantly. In figure 10, if you ‘damp’ the bottom corner when the beam is at the half skip

position (2), the signal will fall and when the beam is at the full skip position (3) you can damp the

signal at the top corner and at the reflecting point on the bottom surface.

PIPE WALLS

If you are going to scan a pipe wall in the longitudinal direction, then you can use any of the above

calibration procedures. However, if you are scanning circumferentially the calculation of beam path

length, and skip distances is more complicated. If you have a segment of pipe of the same outside

diameter and wall thickness as a reference block, you can use the ‘skip’ method for finding the critical

half and full skip positions on the timebase. If you also need to look for discontinuities in the volume of

the object, you calibrate the timebase on the A2 or A4 block for an exact range, and then put the

probe on the ‘reference’ pipe segment and note the half and full skip ranges.

The wall thickness for any given outside diameter is important because the normal range of angled

shear wave probes (45o, 60o,and 70o), when used on thick wall pipe may cut across to the outside

surface again without touching the bore. An example is shown in figure 12 where a 45o shear wave

only reaches about half way through the wall. In other words, for this outside diameter, the thickest

pipe wall we could test with a 45o probe is only half that shown in the diagram. It follows that, when you

are presented with an unusually thick pipe wall for a particular outside diameter, you need to choose

your probe angle carefully in order to inspect the bore properly. For a given angle, the maximum wall

thickness that allows the centre of the beam to reach the bore of the pipe can be calculated from: -

( )2

1 θSindxt −=

Where t = wall thickness d = pipe outside diameter θ = beam angle

Thickest wall that can be tested with this 45o probe

Fig. 12

Page 71: Ultrasonic Back-to-Basic

This formula can be turned round so that you can calculate the best angle given for a wall thickness,

the formula becoming: -

dtSin 21−=θ

Table 1 shows maximum wall thickness that can be tested for three standard angles and a range of

pipe diameters.

Pipe OD

Maximum pipe wall thickness for probe angles

35o 45o 60o

4” (100mm) 21mm 14mm 7mm

6” (150mm) 32mm 22mm 10mm

8” (200mm) 43mm 29mm 13mm

10” (250mm) 53mm 36mm 17mm

12” (300mm) 64mm 44mm 20mm

14” (350mm) 74mm 51mm 23mm

16” (400mm) 85mm 58mm 27mm

18” (450mm) 96mm 66mm 30mm

20” (500mm) 106mm 73mm 33mm

Table 1

Once the correct angle for the pipe size and wall thickness has been chosen, you can establish the

skip and half skip positions using a section of pipe with a drilled hole to produce the required ‘corner’

reflectors – as shown in figure 13.

Fig. 13

CALIBRATION OF THE GAIN

This is often called "setting the sensitivity", and it means that we adjust the gain so that a significant

discontinuity will give a signal that is large enough to see, but small surface scratches will not. Very

often, we use a reference block, similar in shape and material to the specimen, and containing either a

drilled hole or an artificial (machined) crack. The probe is aimed at this reference hole or crack, to

Page 72: Ultrasonic Back-to-Basic

obtain an echo, this is then maximised by probe movement, and then, the gain is adjusted to give the

required signal height known as the ‘reference level’ and the gain is then said to be ‘calibrated’. This

reference level may be 50% or 75% of full screen height, and is often used as the basis for getting

acceptance standards for the inspection. Hence, you may find that you are working to a specification

that says that any signal equal to, or greater than the amplitude of the reference level is cause for

rejection of the component, whereas any signal lower than the reference level may be ignored.

TESTING FOR OUTSIDE DIAMETER SURFACE FLAWS

Discontinuities that break the top surface such as the crack show in figure 14 will cause a reflection to

occur at exactly the beam path distance for the full skip if a suitable angle that will reach the bore is

used. However, as you can see from figure 15, if you are testing a thick wall tube or a solid bar, the

beam may reach the top surface without first reflecting from any other surface. The beam path length

at which a top surface defect will appear in that case can be calculated from the formula: -

θcosDBPL = Where: - D is the outside diameter and θ is the probe angle.

Fig. 15

Crack

Fig. 14 In the sort of application illustrated in figure 15, if there is no crack, the sound will carry on around the

bar or pipe as shown in figure 16. Provided there is enough sensitivity, you may only need to scan

from position ‘A’ to position ‘B’. The beam will sweep the entire circumference during the short scan

and as long as you have enough timebase and gain, echoes from any discontinuities breaking the

surface will appear at predictable positions.

Page 73: Ultrasonic Back-to-Basic

A

B

Fig. 16 References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury

Page 74: Ultrasonic Back-to-Basic

BACK TO BASICS – ULTRASONICS

Part 11. SURFACE & TOFD TECHNIQUES

SURFACE WAVE TECHNIQUES

Surface waves have been used very successfully for a great number of applications, particularly in

the Aircraft Industry. However, it is not so common in the Steel Industry because surface finishes are

often less smooth, and magnetic particle inspection will find most defects detectable by surface

waves. Nevertheless, there are occasions when the use of surface wave techniques can give the

simplest and most positive results and so, in this section we will discuss some general principles that

can be applied when considering a surface wave technique.

ADVANTAGES OF SURFACE WAVES

Surface waves will follow gentle contours without reflection, but will reflect sharply from a sudden

change in contour. Figure 1 shows a typical example of a component having a complex shape that

would make the use of shear or compression waves difficult, if not impossible. Cracks may develop

anywhere along the leading or trailing edge of the blade out to about two thirds of the blade length, or

in the root radius. A surface wave probe placed at the end of the blade, and directed towards the root

will send a beam along the surface, round the radius and reflect from the edge of the root as shown.

Cracks in the suspect areas will give reflections at an earlier time than the root.

Fig.1

The fact that surface waves only penetrate to a depth of about one wavelength can be used to

advantage when testing relatively thin wall sections. Figure 2 shows a pipe with a change of section.

We are told that cracks may occur on the inside or outside diameter of the pipe in the necked region.

An angled shear wave probe might be used, but it would be difficult to predict the skip points as the

beam bounces around the section change. However, if we choose a surface wave at a frequency for

which the wavelength is approximately equal to the wall thickness, then the surface wave will fill the

wall thickness, and follow the section change, reflecting for a defect breaking either surface.

Page 75: Ultrasonic Back-to-Basic

SF

Fig. 2

LIMITATIONS OF SURFACE WAVES

The main limitation of the surface wave technique is that the beam is almost immediately attenuated

if the surface finish is rough, covered in scale, or a liquid (such as the couplant), or has any pressure

applied by another object (such as your hand). For this reason it is normal to use grease as the

couplant for surface wave probes (it doesn't run!), and to apply the grease to the probe, place the

probe on the job and scan forward (away from your own grease trail). Ridges left in the couplant

during scanning, and objects resting on the test surface, often give spurious signals that might be

taken to originate from defects so it is normal to test such indication by rubbing a cloth over the area

indicated by the signal. If after this ‘cleaning’ operation, the signal disappears, then it was a spurious

indication.

CALIBRATION & DEFECT LOCATION

It is not usual to calibrate the timebase for surface waves in the way we would do for shear or

compression waves. This is because we can normally run a finger along the surface in front of the

probe, when we find a defect indication, until the signal is no longer ‘damped’. This happens as we

pass over the defect with our finger, however, there are occasions when access is limited and we are

directing surface waves to a region that is out of sight and cannot be reached with the hand. In these

cases, the timebase can be calibrated using the same procedure as for shear waves on an A2 or

similar block.

The sensitivity can be set from drilled holes or spark-eroded slits in suitable reference blocks. In the

aircraft industry, these reference blocks are usually sections of an actual component with a spark-

eroded slit in the critical location.

DEFECT SIZING USING TOFD

The TOFD technique, first used by Silk in 1977, uses tip diffraction to identify the top, bottom and

ends of a discontinuity in one pass. Silk chose to use an angled compression wave for the TOFD

technique rather than a shear wave, for two reasons. Firstly, the tip diffraction signal is stronger than

Page 76: Ultrasonic Back-to-Basic

a shear wave diffraction signal, and secondly, a lateral wave is produced which can be used to

measure the horizontal distance between the transmitter and receiver.

The tip diffraction signal is generated at the tip of the discontinuity – effectively a ‘Point’ source.

According to Huyghens, a point source produces a spherical beam. Figure 3 shows a typical TOFD

transducer set-up on a component with a vertical discontinuity. There are four sound paths from the

transmitter to the receiver. Path ‘A’ is the lateral wave path travelling just below the surface. Path ‘B’

is the tip diffraction path from the top of the discontinuity. Path ‘C’ is the tip diffraction path from the

bottom of the discontinuity and path ‘D’ is the backwall echo path.

Figure 4 shows a typical unrectified trace for the four signals. Note the phase relationships, A and C

are in opposite phase to B and D. The important difference to note is between B and C – the top and

bottom diffraction signals are in opposite phase. This phase difference allows the practitioner to

identify those points.

RT Lateral wave

Diffraction echoes

Backwall reflection

from top & bottom

Backwall ‘D’

‘C’

‘B’

‘A’

Fig. 3

A B C D

Fig. 4

Page 77: Ultrasonic Back-to-Basic

Assuming that the diffracting tip is centred between the two transducers, the depth of the tip below

the surface can be calculated from: -

22

22⎟⎠⎞

⎜⎝⎛−⎟

⎠⎞

⎜⎝⎛=

HDBPLDepth

Where:

BPL = Beam path length for the signal in question

HD = Beam path length for the lateral wave.

The distance measurements taken from the ultrasonic trace must be made from the same part of

each waveform. In the trace shown in Figure 4, the largest half cycle would be selected. For signals A

& C this is negative and for signal B positive. The advances in computer technology have made it

possible to carry out all the calculations and plotting to be handled automatically and stored for

subsequent evaluation. The method that has been chosen to display this TOFD data presents the

information in a special ‘B-scan’ form that is easy to assimilate. The way in which the positive and

negative half cycles are displayed needs explaining.

In a conventional B-scan image, the ‘slice’ is taken across the weld perpendicular to the centre line.

In the TOFD display, the ‘slice’ is taken along the weld (figure 5). However, whereas the conventional

B-scan is a relatively thin slice, the TOFD image represents the volume between the probes as they

scan along the weld. The presentation is known as a ‘D-scan’.

Cap

Weld length

Root

Fig. 5

An echo arriving at the receiver is a pulse of a certain pulse width and amplitude. In conventional B-

scan displays, this pulse is displayed as a bright spot whose diameter is proportional to the pulse

width and whose brightness is proportional to the signal amplitude. In some ways, it is like a broad

pencil tip that can be used to draw pictures in light or bold broad strokes. The pulse is really a short

burst of a few cycles of alternating waveform. In the TOFD system, the waveform is depicted in

greyscale with positive going half cycles tending towards white, and negative going half cycles

tending towards black (see figure 6). This type of display will allow us to identify phase change so that

we can discriminate between he lateral wave, top and bottom defect signals and backwall.

Page 78: Ultrasonic Back-to-Basic

Fig. 6

This allows particular half cycles to be identified for measurement purposes, and phase changes to

be recognized for determination of top or bottom echo. Figure 7 shows a typical computer screen for

a TOFD inspection. The image shows details of the component (in this case, a weld) as well as the

TOFD D-scan image and an A-scan trace. In this image, left to right represents the component

thickness, and the vertical dimension represents scan length.

The A-scan trace shown corresponds to a slice through the weld at the location indicated by the

‘cross hairs’ of the cursor. The striped band on the left of the TOFD image represents the lateral

wave, and the bold striped band to the right of the image represents the backwall echo. The

difference in boldness is due to the different signal amplitudes. Following the horizontal ‘cross hairs’

and about half way between the lateral wave and backwall ‘stripes’, a series of feint ‘horse shoe’

shaped stripes can be seen. These are diffraction signals from a small discontinuity. The A-scan trace

shows the signal clearly.

Fig. 7

Page 79: Ultrasonic Back-to-Basic

In this example, the discontinuity has a very small dimension in the through-thickness dimension, but

close study of the A-scan shows a small phase shift in the last half-cycle of the discontinuity signal.

This tells the practitioner that the distance from top to bottom of the discontinuity is about the same as

the pulse length for this particular discontinuity.

Fig. 8

A much bolder indication can be seen towards the top of the lateral wave line suggesting a

discontinuity at, or just below the surface. In figure 8, the cursor has been moved to this location. The

lateral wave signal can be seen to be longer and stronger than at the previous location. The fact that

the wave shape stays in phase suggests that the diffraction echo, which is extending the signal, has

the same phase as the lateral wave. In other words, it is a bottom tip signal. However, it is not

possible in this case to see where the lateral wave ends and the bottom tip begins, and so it is not

possible to say how deep the discontinuity extends below the surface. The TOFD method is limited in

its ability to size near surface discontinuities when the arrival time difference between the lateral wave

and the diffraction signal is similar to pulse length. Near surface resolution when using TOFD can be

a bit confusing if you look at it from a conventional ultrasonics point of view. Imagine a top surface

crack 4mm deep. At 5MHz, it represents more than two wavelengths at compression wave velocity

and with a reasonably short pulse of two cycles; you might expect to resolve the bottom of the defect.

However, the path difference between the lateral wave and the tip diffraction signals for a probe

separation of 80mm is only 0.4mm and this is about the same as the wavelength for 15MHz (See

figure 9). You would need a 15MHz transmitter with only one cycle in the pulse to resolve the crack.

80.4mm

80mm

Fig. 9

The transducers used in TOFD techniques are angled compression wave transducers. The common

angles used are 60o and 70o, although other angles may be used if the component thickness makes it

necessary. The design and construction of the transducer is important in order to promote a good

lateral wave. Previous theory has suggested that a shear wave should also exist in the component

Page 80: Ultrasonic Back-to-Basic

and this is true, it does. Figure 10 shows a little more of the trace for the above example. On the

extreme right of both the A-scan and TOFD D-scan, the shear wave can be seen. Since it arrives well

after the other signals, it does not present a problem in this application.

Shear wave

Fig. 10 Scanning with the TOFD system is fast and many scanning systems are motorized. They all require

distance encoders so that the D-scan image can be constructed. The vertical extent of those defects

that can be resolved is many times more accurate than other sizing systems.

References: - ‘Ultrasonic Flaw Detection for Technicians’ - Third Edition, June 2004 by J. C. Drury